Advances in Forming, Machining and Automation: Proceedings of AIMTDR 2018 [1st ed. 2019] 978-981-32-9416-5, 978-981-32-9417-2

Table of contents :
Front Matter ....Pages i-xix
Front Matter ....Pages 1-1
Effect of Heat Treatment on Formability of AA6082 by Single Point Incremental Forming (S. Maharajan, D. Ravindran, S. Rajakarunakaran)....Pages 3-15
Forming Behavior of AA5052-H32 and AA6061-T6 During Single Point Incremental Forming (M. M. Ghadmode, R. R. Pawar, B. U. Sonawane)....Pages 17-28
Pull-Out Forming: Experiments and Process Simulation (S. Kumar)....Pages 29-51
Failure Prediction and Forming Behavior of AA5754 Sheets at Warm Temperature (Sudhy S. Panicker, Kaushik Bandyopadhyay, Sushanta Kumar Panda)....Pages 53-65
Magnetic Pulse Forming and Punching of Al Tubes—A Novel Technique for Forming and Perforation of Tubes (Sagar Pawar, Sachin D. Kore, Arup Nandy)....Pages 67-77
Experimental Investigation on the Forming of AA 5052-H32 Sheet Using a Rigid-Body-Based Impact in a Shock Tube (S. K. Barik, R. Ganesh Narayanan, N. Sahoo)....Pages 79-90
An Experimental Study on Single-Point Incremental Forming of AA5083 Sheet Using Response Surface Methodology (Gautam Kumar, Saurabh, Maharshi Roshan, Kumar Nandan, Kuntal Maji)....Pages 91-103
Study and Establishment of Manufacturing Process of Molybdenum Liners Using Warm Flow Forming Process (P. S. S. R. K. Prasad, Navneet Verma, Narendra Kumar, K. M. Rajan)....Pages 105-116
Front Matter ....Pages 117-117
Influence of Surgical Drill Geometry on Drilling Performance of Cortical and Trabecular Bone (Ramesh Kuppuswamy, Brett Christie-Taylor)....Pages 119-131
Study of Cutting Temperature and Chip Formation in Drilling of AA6351–B4C Composite (S. Thirumalai Kumaran, G. S. Samy, M. Uthayakumar, Tae Jo Ko)....Pages 133-141
A Study of Parameters Affecting Cutting Forces in Minimum Quantity Lubrication-Assisted Cross-Peripheral Grinding of Alumina Ceramic (A. V. Manu, V. G. Ladeesh, R. Manu)....Pages 143-152
Condition Monitoring of Abrasive Waterjet Milling Using Acoustic Emission and Cutting Force Signals (U. Goutham, M. Kanthababu, S. Gowri, K. R. Sunilkumar, M. Mathanraj, J. John Rozario Jegaraj et al.)....Pages 153-164
A Novel Small Quantity Lubrication Method to Assess Grindability of Inconel 718 (Sirsendu Mahata, Manas Bhattacharyya, Bijoy Mandal, Santanu Das)....Pages 165-176
Performance of Carbon Nanotubes Based Cutting Oil on Turning of AISI 1040 Steel (M. Amrita, P. Yogesh Chandra, P. Venkata Ramana, U. Shyam Sai, Chatti Sreeram)....Pages 177-188
Investigations on the Influence of Serration Parameters on Cutting Forces (P. Bari, P. Wahi, M. Law)....Pages 189-204
Effect of Minimum Quantity Lubrication on Tool Wear and Surface Integrity During Hard Turning of EN31 Steel (Jitendra Kumar Verma, Gaurav Bartarya, Jitendra Bhaskar)....Pages 205-218
Temperature Profiling of Microwave–Metal Discharge Plasma Channel Using Image Processing Technique (Anurag Singh, Apurbba Kumar Sharma)....Pages 219-227
Applicability of CaF2 Solid Lubricant-Assisted Minimum Quantity Lubrication in Turning for Sustainable Manufacturing (Mayur A. Makhesana, K. M. Patel, Anand S. Patel)....Pages 229-238
Cryogenic Machining of AZ31B Magnesium Alloy for Bio-implant Applications (Vaibhav Tibrewal, Kalpit Dak, Aundhe Himanshu, Hema Kumar, P. Kuppan, A. S. S. Balan)....Pages 239-251
Experimental Investigation on Machining Parameters of Hastelloy C276 Under Different Cryogenic Environment (S. Vignesh, U. Mohammed Iqbal)....Pages 253-267
Machining of EN-31 Steel and Experimental Analysis of Various Process Parameters Using Minimum Quantity Lubrication (Ashutosh Saini, S. K. S. Yadav)....Pages 269-282
Effect of Tool Material on Trepanning of CFRP Composites (B. R. Jayasuriya, A. Harsha Vardhan, V. Krishnaraj)....Pages 283-294
Effect of Air Delivery Pressure and Flow Rate on Surface Integrity in Minimum Quantity Cooling Lubrication Grinding of Inconel 718 (Anirban Naskar, Amit Choudhary, Biddu Bhushan Singh, S. Paul)....Pages 295-304
Effect of Different Geometric Texture Shapes on Wettability and Machining Performance Evaluation Under Dry and MQL Environments (Sarvesh Kumar Mishra, Sudarsan Ghosh, Sivanandam Aravindan)....Pages 305-314
Evaluation of Surface Morphology of Yttria-Stabilized Zirconia with the Progress of Wheel Wear in High-Speed Grinding (Amit Choudhary, Anirban Naskar, S. Paul)....Pages 315-323
Grindability and Surface Integrity of Nickel-Based Cast Superalloy IN-738 by Vitrified Alumina Wheel (Srinivasa Rao Nandam, A. Venugopal Rao, Amol A. Gokhale, Suhas S. Joshi)....Pages 325-337
Simulating the Effect of Microstructure in Metal Sliding and Cutting (A. S. Vandana, Narayan K. Sundaram)....Pages 339-347
What Do Chip Morphologies Tell Us About the Cutting Process? (Koushik Viswanathan, Anirudh Udupa, Dinakar Sagapuram, James B. Mann)....Pages 349-359
Simultaneous Optimization of Milling Process Responses for Nano-Finishing of AISI-4340 Steel Through Sustainable Production (Muhammed Muaz, Sounak Kumar Choudhury)....Pages 361-374
Assessment of Cutting Tool Reliability During Turning Considering Effects of Cutting Parameters and Machining Time (Gaddafee Mohamad, Satish Chinchanikar)....Pages 375-384
An Experimental Investigation on Productivity and Product Quality During Thin-Wall Machining of Aluminum Alloy 2024-T351 (G. Bolar, S. N. Joshi)....Pages 385-393
An Approach of Minimizing Energy Consumption in the Machining System Using Job Sequences Varying Technique (Shahzar Jawaid, S. C. Srivastava, S. Datta)....Pages 395-405
Comparative Study on the Performance of Different Drill Bits for Drilling CFRP (N. S. Sowjanya, V. G. Ladeesh, R. Manu, Jose Mathew)....Pages 407-419
A Cyber-Physical System Improves the Quality of Machining in CNC Milling Machine—A Case Study (Ganesh Kumar Nithyanandam, Saravana Kumar Sellappan, Selvaraj Ponnumuthu)....Pages 421-429
Challenges in Machining of Silica–Silica Cone for Aerospace Application (Ashish Tewari, T. Srinivasulu, B. Hari Prasad, A. P. Dash)....Pages 431-440
Optimization of Cutting Parameters for Hard Turning of WC–Co–Ni–Cr (15% Binder) Mill Rolls on CNC Lathe with Polycrystalline Diamond (Mahesh J. Hunakunti, Vaishali Jagannath, Ramesh S. Rao, S. Srinivas, S. Shyamsundar, D. Ashokkumar)....Pages 441-449
Multi-response Optimization of End Milling on Al6061–Sicp Metal Matrix Composite–Hybrid GRA-PCA Approach (B. Ravi Sankar, P. Umamaheswarrao)....Pages 451-459
Experimental Investigation on Dewaxed Tungsten Carbide-Based Self-lubricant Cutting Tool Material (A. Muthuraja)....Pages 461-469
An Experimental Investigation on Horizontal Surface Grinding of Mild Steel Using Different Lubricating Oils (Soutrik Bose, Nabankur Mandal)....Pages 471-480
Comparison Between Advanced Cutting Tools to Achieve a Better Cutting Condition for the Machining of Aluminium (S. K. Pattnaik, M. Behera, S. Padhi, S. K. Sarangi)....Pages 481-494
Investigation on Machining Responses during Hard Turning of AISI D2 Steel under Dry, Wet and Nano-based MQL Conditions (Vaibhav Chandra, Sudarsan Ghosh, P. Venkateshwara Rao)....Pages 495-504
Effect of Machining Parameters on Surface Integrity in End Milling of Inconel 625 (Ramesh Rajguru, Hari Vasudevan)....Pages 505-515
Tool Wear Behavior in Milling of Hardened Custom 465 Steel (V. Prasath, V. Krishnaraj, J. Kanchana, B. Geetha Priyadharshini)....Pages 517-525
Experimental Study on Machining of EN24 Using Minimum Quantity Lubrication (Gaurav Tyagi, J. Bhaskar, S. K. Singhal, G. Bartarya)....Pages 527-538
Investigative Study of Temperature Produced During Turning Operation Using MQL and Solid Lubricants (Anand S. Patel, Mayur A. Makhesana, K. M. Patel)....Pages 539-549
Effect of Dressing Infeed on Alumina Wheel During Grinding Ti–6Al–4V Under Varying Depth of Cut (Manish Mukhopadhyay, Souvik Chatterjee, Pranab Kumar Kundu, Santanu Das)....Pages 551-560
Experimental Evaluation of Surface Roughness, Dimensional Accuracy, and MRR in Cylindrical Grinding of EN 24 Steel (Pankaj V. Mohire, Raju S. Pawade)....Pages 561-569
Front Matter ....Pages 571-571
Bio-inspired Knowledge Representation Framework for Decision Making in Product Design (Varun Tiwari, Prashant Kumar Jain, Puneet Tandon)....Pages 573-585
Heuristic Algorithmic Approach for Automatic Generation of Pin Layout for Robotic Unloading of Sheet Metal Parts (A. Ramesh Babu)....Pages 587-600
Voxel-Based Strategy for Efficient CNC Machining (A. Kukreja, H. D. Mane, M. Dhanda, S. S. Pande)....Pages 601-612
Effect of Geometrical and Process Parameters on Utilization of Sheet Material in Plasma and Laser Cutting Processes (N. Venkatesh, S. Sabari Sriram, V. Satish Chandran, A. Ramesh Babu)....Pages 613-624
Development of Manufacturability Indices for Prismatic Parts (Manish Kumar Gupta, Pramod Kumar Jain, Abinash Kumar Swain)....Pages 625-635
A Cyber-Physical System Architecture for Smart Manufacturing (Jitin Malhotra, Faiz Iqbal, Ashish Kumar Sahu, Sunil Jha)....Pages 637-647
Measurement of Bores Using Scanning Mode of Articulated Arm Coordinate Measuring Machines (Ashik Suresh, P. B. Dhanish)....Pages 649-658
A Method for Evaluation of Simple Torus Surfaces (T. S. R. Murthy)....Pages 659-669
An Artefact-Based Continues Performance Verification of Coordinate Measuring Machine (Goitom Tesfay, Rega Rajendra)....Pages 671-678
Development of Welding Fixture for Rocket Motor Casing Assembly (Venkateswarlu Chepuru, P. Kiran, B. Hari Prasad)....Pages 679-688
Generation of Sequence of Machining Operations Through Visualization of End Product (G. V. S. S. Sharma, P. Srinivasa Rao, B. Surendra Babu)....Pages 689-698
Monitoring the Dynamics and Tracking of a Vehicle Using Internet of Things (IoT) (Sri Harsha Dorapudi, Vivek Varma Buddharaju, V. V. Vimal Varma, R. Ramesh)....Pages 699-707
Automated Production of Medical Screws Using Titanium Bar on Indigenous Sliding Headstock Automat (Manohar Bulbule, Naveen Hosamani, S. R. Chandramouli)....Pages 709-721
Process Mechanization and Automation for Hybrid TIG MAG Arc Welding (Onkar S. Sahasrabudhe, D. N. Raut)....Pages 723-731
Two-Sided Assembly Line Balancing—A Company Case Study Solved by Exact Solution Approach (Ashish Yadav, Sunil Agrawal)....Pages 733-741

Citation preview

Lecture Notes on Multidisciplinary Industrial Engineering Series Editor: J. Paulo Davim

M. S. Shunmugam M. Kanthababu Editors

Advances in Forming, Machining and Automation Proceedings of AIMTDR 2018

Lecture Notes on Multidisciplinary Industrial Engineering Series Editor J. Paulo Davim , Department of Mechanical Engineering, University of Aveiro, Aveiro, Portugal

“Lecture Notes on Multidisciplinary Industrial Engineering” publishes special volumes of conferences, workshops and symposia in interdisciplinary topics of interest. Disciplines such as materials science, nanosciences, sustainability science, management sciences, computational sciences, mechanical engineering, industrial engineering, manufacturing, mechatronics, electrical engineering, environmental and civil engineering, chemical engineering, systems engineering and biomedical engineering are covered. Selected and peer-reviewed papers from events in these fields can be considered for publication in this series.

More information about this series at http://www.springer.com/series/15734

M. S. Shunmugam M. Kanthababu •

Editors

Advances in Forming, Machining and Automation Proceedings of AIMTDR 2018

123

Editors M. S. Shunmugam Manufacturing Engineering Section Department of Mechanical Engineering Indian Institute of Technology Madras Chennai, Tamil Nadu, India

M. Kanthababu Department of Manufacturing Engineering College of Engineering, Guindy Anna University Chennai, Tamil Nadu, India

ISSN 2522-5022 ISSN 2522-5030 (electronic) Lecture Notes on Multidisciplinary Industrial Engineering ISBN 978-981-32-9416-5 ISBN 978-981-32-9417-2 (eBook) https://doi.org/10.1007/978-981-32-9417-2 © Springer Nature Singapore Pte Ltd. 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

AIMTDR 2018 Conference’s Core Organizing Committee

Patrons Dr. M. K. Surappa, Vice Chancellor, Anna University Dr. J. Kumar, Registrar, Anna University

President (NAC-AIMTDR) Mr. P. Kaniappan, Managing Director, WABCO India Ltd.

Vice-President (NAC-AIMTDR) Dr. Uday Shanker Dixit, Professor, IIT Guwahati, India

Co-patrons Dr. A. Rajadurai, Dean, MIT Campus, Anna University Dr. T. V. Geetha, Dean, CEG Campus, Anna University Dr. L. Karunamoorthy, Chairman, Faculty of Mechanical Engineering, Anna University Dr. S. Rajendra Boopathy, Head, Department of Mechanical Engineering, Anna University

v

vi

AIMTDR 2018 Conference’s Core Organizing Committee

Chairman Dr. S. Gowri, Honorary Professor, Department of Manufacturing Engineering, Anna University

Co-chairman Dr. P. Hariharan, Professor, Department of Manufacturing Engineering, Anna University

Organizing Secretary Dr. M. Kanthababu, Professor and Head, Department of Manufacturing Engineering, Anna University

Joint Organizing Secretaries Dr. M. Pradeep Kumar, Professor, Department of Mechanical Engineering, Anna University Dr. A. Siddharthan, Associate Professor, Department of Production Technology, Anna University

International Scientific Committee Prof. Prof. Prof. Prof. Prof. Prof. Prof. Prof. Prof. Prof. Prof. Prof.

Abhijit Chandra, Iowa State University, USA Ajay P. Malshe, University of Arkansas, USA Andrew Y. C. Nee, NUS, Singapore Chandrasekar S., Purdue University, USA Dean T. A., University of Birmingham, UK Hong Hocheng, National Tsing Hui University, Taiwan John Sutherland, Purdue University, USA Kamlakar P. Rajurkar, University of Nebraska, USA Kornel Ehmann, Northwestern University, USA Liao Y. S., National Taiwan University, Taiwan McGeough J. A., University of Edinburgh, UK Mustafizur Rahman, NUS, Singapore

AIMTDR 2018 Conference’s Core Organizing Committee

Prof. Prof. Prof. Prof. Prof. Prof.

Philip Koshy, McMaster University, Canada Rakesh Nagi, University of Buffalo, USA Shiv Gopal Kapoor, University of Illinois, USA Srihari Krishnasami, Binghamton University, USA Tae Jo Ko, Yeungnam University, South Korea Tugrul Ozel, State University of New Jersey, USA

National Advisory Committee Prof. Ahuja B. B., Government College of Engineering, Pune Prof. Amitabha Ghosh, BESU Prof. Bijoy Bhattacharyya, Jadavpur University, Kolkata Prof. Biswanath Doloi, Jadavpur University, Kolkata Prof. Chattopadhyay A. K., IIT Kharagpur Prof. Deshmukh S. G., IIT Gwalior Shri. Dhand N. K., MD, ACE Micromatic, Bangalore Prof. Dixit U. S., IIT Guwahati, Guwahati Prof. Jain P. K., IIT Roorkee, Roorkee Prof. Jain V. K., IIT Kanpur Prof. Jose Mathew, NIT Calicut Shri. Lakshminarayan M., WABCO India Pvt. Ltd. Prof. Lal G. K., IIT Kanpur Prof. Mehta N. K., IIT Roorkee Prof. Mohanram P. V., PSG Institute of Technology and Applied Research Shri. Mohanram P., IMTMA, Bangalore Dr. Mukherjee T., Tata Steel Ltd., Jamshedpur Shri. Muralidharan P., Lucas TVS Ltd., Vellore Prof. Narayanan S., VIT University, Vellore Mr. Niraj Sinha, Scientist ‘G’, PSA, GOI Prof. Pande S. S., IIT Bombay, Mumbai Dr. Prasad Raju D. R., MVGREC Prof. Radhakrishnan P., PSG Institute of Advanced Studies, Coimbatore Prof. Radhakrishnan V., IIST, Trivandrum Prof. Ramaswamy N., IIT Bombay (Former), Chennai Prof. Ramesh Babu N., IIT Madras Shri. Rangachar C. P., Yuken India Ltd., Bangalore Prof. Rao P. V., IIT Delhi Dr. Santhosh Kumar, IIT BHU Dr. Sathyan B. R., CMTI, Bangalore Prof. Satyanarayan B., Andhra University (Former), Visakhapatnam Prof. Selvaraj T., NIT Trichy Prof. Shan H. S., IIT Roorkee (Former), Chandigarh Prof. Shunmugam M. S., IIT Madras

vii

viii

AIMTDR 2018 Conference’s Core Organizing Committee

Shri. Shirgurkar S. G., Ace Designers Ltd., Bangalore Dr. Sumantran V., Celeris Technologies Dr. Suri V. K., BARC, Mumbai Shri. Venu Gopalan P., DRDL, Hyderabad Prof. Vinod Yadav, Motilal Nehru National Institute of Technology, Allahabad

Foreword

It gives us immense pleasure to present the Advances in Manufacturing Technology and Design—Proceedings of All India Manufacturing Technology, Design and Research (AIMTDR) Conference 2018. We would like to express our deep gratitude to all the members of Organizing Committee of AIMTDR 2018 Conference and also to authors, reviewers, sponsors, volunteers, etc., for their wholehearted support and active participation. Our special thanks to Mr. P. Kaniappan, Managing Director, WABCO India Ltd, Chennai, who kindly agreed to act as President of National Advisory Committee (NAC) of the AIMTDR 2018 Conference. We also express our sincere thanks to Chairman Dr. S. Gowri, Honorary Professor, and Co-chairman Dr. P. Hariharan, Professor, Department of Manufacturing Engineering, Anna University, Chennai, for their wholehearted support. We would like to express our sincere thanks to Research Scholars Mr. K. R. Sunilkumar, Mr. U. Goutham, Mr. V. Mohankumar and Mr. R. Prabhu and also UG/PG students of the Department of Manufacturing Engineering, Anna University, for their contributions in the preparation of this volume. High-quality papers have been selected after peer review by technical experts. We hope you find the papers included in the Proceedings of AIMTDR 2018 Conference are interesting and thought-provoking. We also like to express our gratitude for the support provided by WABCO India Ltd., Chennai; Kistler Instruments India Pvt. Ltd., Chennai; AMETEK Instruments India Pvt. Ltd., Bengaluru; Central Manufacturing Technology Institute, Government of India, Bengaluru; Defence Research and Development Organisation, Government of India, New Delhi; and Ceeyes Engineering Industries Pvt Ltd., Trichy.

ix

x

Foreword

Finally, we would like to express our gratitude to the National Advisory Committee (NAC) members of AIMTDR 2018 Conference for providing the necessary guidance and support. Uday Shanker Dixit Vice-President, National Advisory Committee AIMTDR, Guwahati, India

Preface

All India Manufacturing Technology, Design and Research (AIMTDR) Conference is considered globally as one of the most prestigious conferences held once in two years. It was started in 1967 at national level at Jadavpur University, Kolkata, India, and achieved the international status in the year 2006. It was organized by various prestigious institutions such as Jadavpur University, IIT Bombay, IIT Madras, CMTI Bangalore, PSG iTech, IIT Kanpur, CMERI, IIT Delhi, NIT Warangal, IIT Kharagpur, BITS Ranchi, VIT Vellore, IIT Roorkee, Andhra University, IIT Guwahati and College of Engineering Pune. The recent edition of the AIMTDR Conference, 7th International and 28th All India Manufacturing Technology, Design and Research (AIMTDR) Conference 2018, was jointly organized by the Departments of Manufacturing Engineering, Mechanical Engineering and Production Technology during 13–15 December 2018 at College of Engineering Guindy, Anna University, Chennai, India, with the theme ‘Make in India—Global Vision’. A major focus was given on recent developments and innovations in the field of manufacturing technology and design through keynote lectures. About 550 participants registered for the conference. During the conference, researchers from academia and industries presented their findings and exchanged ideas related to manufacturing technology and design. Of the 750 papers received initially, 330 papers were finally selected after rigorous review process for publication. Selected papers from the conference are being published by Springer in the series Lecture Notes on Multidisciplinary Industrial Engineering in five volumes, namely Volume 1—Additive Manufacturing and Joining, Volume 2—Forming, Machining and Automation, Volume 3—Unconventional Machining and Composites, Volume 4—Micro and Nano Manufacturing and Surface Engineering and Volume 5—Simulation and Product Design and Development. Chennai, India May 2018

M. S. Shunmugam M. Kanthababu

xi

Contents

Part I 1

2

Forming

Effect of Heat Treatment on Formability of AA6082 by Single Point Incremental Forming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S. Maharajan, D. Ravindran and S. Rajakarunakaran

3

Forming Behavior of AA5052-H32 and AA6061-T6 During Single Point Incremental Forming . . . . . . . . . . . . . . . . . . . . . . . . . M. M. Ghadmode, R. R. Pawar and B. U. Sonawane

17

3

Pull-Out Forming: Experiments and Process Simulation . . . . . . . . S. Kumar

4

Failure Prediction and Forming Behavior of AA5754 Sheets at Warm Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sudhy S. Panicker, Kaushik Bandyopadhyay and Sushanta Kumar Panda

5

6

7

8

29

53

Magnetic Pulse Forming and Punching of Al Tubes—A Novel Technique for Forming and Perforation of Tubes . . . . . . . . . . . . . . Sagar Pawar, Sachin D. Kore and Arup Nandy

67

Experimental Investigation on the Forming of AA 5052-H32 Sheet Using a Rigid-Body-Based Impact in a Shock Tube . . . . . . . S. K. Barik, R. Ganesh Narayanan and N. Sahoo

79

An Experimental Study on Single-Point Incremental Forming of AA5083 Sheet Using Response Surface Methodology . . . . . . . . . Gautam Kumar, Saurabh, Maharshi Roshan, Kumar Nandan and Kuntal Maji

91

Study and Establishment of Manufacturing Process of Molybdenum Liners Using Warm Flow Forming Process . . . . . . 105 P. S. S. R. K. Prasad, Navneet Verma, Narendra Kumar and K. M. Rajan xiii

xiv

Contents

Part II 9

Machining

Influence of Surgical Drill Geometry on Drilling Performance of Cortical and Trabecular Bone . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Ramesh Kuppuswamy and Brett Christie-Taylor

10 Study of Cutting Temperature and Chip Formation in Drilling of AA6351–B4C Composite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 S. Thirumalai Kumaran, G. S. Samy, M. Uthayakumar and Tae Jo Ko 11 A Study of Parameters Affecting Cutting Forces in Minimum Quantity Lubrication-Assisted Cross-Peripheral Grinding of Alumina Ceramic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 A. V. Manu, V. G. Ladeesh and R. Manu 12 Condition Monitoring of Abrasive Waterjet Milling Using Acoustic Emission and Cutting Force Signals . . . . . . . . . . . . 153 U. Goutham, M. Kanthababu, S. Gowri, K. R. Sunilkumar, M. Mathanraj, J. John Rozario Jegaraj and R. Balasubramanian 13 A Novel Small Quantity Lubrication Method to Assess Grindability of Inconel 718 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Sirsendu Mahata, Manas Bhattacharyya, Bijoy Mandal and Santanu Das 14 Performance of Carbon Nanotubes Based Cutting Oil on Turning of AISI 1040 Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 M. Amrita, P. Yogesh Chandra, P. Venkata Ramana, U. Shyam Sai and Chatti Sreeram 15 Investigations on the Influence of Serration Parameters on Cutting Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 P. Bari, P. Wahi and M. Law 16 Effect of Minimum Quantity Lubrication on Tool Wear and Surface Integrity During Hard Turning of EN31 Steel . . . . . . 205 Jitendra Kumar Verma, Gaurav Bartarya and Jitendra Bhaskar 17 Temperature Profiling of Microwave–Metal Discharge Plasma Channel Using Image Processing Technique . . . . . . . . . . . . . . . . . . 219 Anurag Singh and Apurbba Kumar Sharma 18 Applicability of CaF2 Solid Lubricant-Assisted Minimum Quantity Lubrication in Turning for Sustainable Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Mayur A. Makhesana, K. M. Patel and Anand S. Patel

Contents

xv

19 Cryogenic Machining of AZ31B Magnesium Alloy for Bio-implant Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Vaibhav Tibrewal, Kalpit Dak, Aundhe Himanshu, Hema Kumar, P. Kuppan and A. S. S. Balan 20 Experimental Investigation on Machining Parameters of Hastelloy C276 Under Different Cryogenic Environment . . . . . . 253 S. Vignesh and U. Mohammed Iqbal 21 Machining of EN-31 Steel and Experimental Analysis of Various Process Parameters Using Minimum Quantity Lubrication . . . . . . 269 Ashutosh Saini and S. K. S. Yadav 22 Effect of Tool Material on Trepanning of CFRP Composites . . . . . 283 B. R. Jayasuriya, A. Harsha Vardhan and V. Krishnaraj 23 Effect of Air Delivery Pressure and Flow Rate on Surface Integrity in Minimum Quantity Cooling Lubrication Grinding of Inconel 718 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295 Anirban Naskar, Amit Choudhary, Biddu Bhushan Singh and S. Paul 24 Effect of Different Geometric Texture Shapes on Wettability and Machining Performance Evaluation Under Dry and MQL Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 Sarvesh Kumar Mishra, Sudarsan Ghosh and Sivanandam Aravindan 25 Evaluation of Surface Morphology of Yttria-Stabilized Zirconia with the Progress of Wheel Wear in High-Speed Grinding . . . . . . . 315 Amit Choudhary, Anirban Naskar and S. Paul 26 Grindability and Surface Integrity of Nickel-Based Cast Superalloy IN-738 by Vitrified Alumina Wheel . . . . . . . . . . . . . . . . 325 Srinivasa Rao Nandam, A. Venugopal Rao, Amol A. Gokhale and Suhas S. Joshi 27 Simulating the Effect of Microstructure in Metal Sliding and Cutting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339 A. S. Vandana and Narayan K. Sundaram 28 What Do Chip Morphologies Tell Us About the Cutting Process? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349 Koushik Viswanathan, Anirudh Udupa, Dinakar Sagapuram and James B. Mann 29 Simultaneous Optimization of Milling Process Responses for Nano-Finishing of AISI-4340 Steel Through Sustainable Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361 Muhammed Muaz and Sounak Kumar Choudhury

xvi

Contents

30 Assessment of Cutting Tool Reliability During Turning Considering Effects of Cutting Parameters and Machining Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 Gaddafee Mohamad and Satish Chinchanikar 31 An Experimental Investigation on Productivity and Product Quality During Thin-Wall Machining of Aluminum Alloy 2024-T351 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385 G. Bolar and S. N. Joshi 32 An Approach of Minimizing Energy Consumption in the Machining System Using Job Sequences Varying Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 Md. Shahzar Jawaid, S. C. Srivastava and S. Datta 33 Comparative Study on the Performance of Different Drill Bits for Drilling CFRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407 N. S. Sowjanya, V. G. Ladeesh, R. Manu and Jose Mathew 34 A Cyber-Physical System Improves the Quality of Machining in CNC Milling Machine—A Case Study . . . . . . . . . . . . . . . . . . . . 421 Ganesh Kumar Nithyanandam, Saravana Kumar Sellappan and Selvaraj Ponnumuthu 35 Challenges in Machining of Silica–Silica Cone for Aerospace Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431 Ashish Tewari, T. Srinivasulu, B. Hari Prasad and A. P. Dash 36 Optimization of Cutting Parameters for Hard Turning of WC–Co–Ni–Cr (15% Binder) Mill Rolls on CNC Lathe with Polycrystalline Diamond . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441 Mahesh J. Hunakunti, Vaishali Jagannath, Ramesh S. Rao, S. Srinivas, S. Shyamsundar and D. Ashokkumar 37 Multi-response Optimization of End Milling on Al6061–Sicp Metal Matrix Composite–Hybrid GRA-PCA Approach . . . . . . . . . 451 B. Ravi Sankar and P. Umamaheswarrao 38 Experimental Investigation on Dewaxed Tungsten Carbide-Based Self-lubricant Cutting Tool Material . . . . . . . . . . . . 461 A. Muthuraja 39 An Experimental Investigation on Horizontal Surface Grinding of Mild Steel Using Different Lubricating Oils . . . . . . . . . . . . . . . . 471 Soutrik Bose and Nabankur Mandal 40 Comparison Between Advanced Cutting Tools to Achieve a Better Cutting Condition for the Machining of Aluminium . . . . . 481 S. K. Pattnaik, M. Behera, S. Padhi and S. K. Sarangi

Contents

xvii

41 Investigation on Machining Responses during Hard Turning of AISI D2 Steel under Dry, Wet and Nano-based MQL Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495 Vaibhav Chandra, Sudarsan Ghosh and P. Venkateshwara Rao 42 Effect of Machining Parameters on Surface Integrity in End Milling of Inconel 625 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505 Ramesh Rajguru and Hari Vasudevan 43 Tool Wear Behavior in Milling of Hardened Custom 465 Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517 V. Prasath, V. Krishnaraj, J. Kanchana and B. Geetha Priyadharshini 44 Experimental Study on Machining of EN24 Using Minimum Quantity Lubrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527 Gaurav Tyagi, J. Bhaskar, S. K. Singhal and G. Bartarya 45 Investigative Study of Temperature Produced During Turning Operation Using MQL and Solid Lubricants . . . . . . . . . . . . . . . . . 539 Anand S. Patel, Mayur A. Makhesana and K. M. Patel 46 Effect of Dressing Infeed on Alumina Wheel During Grinding Ti–6Al–4V Under Varying Depth of Cut . . . . . . . . . . . . . . . . . . . . 551 Manish Mukhopadhyay, Souvik Chatterjee, Pranab Kumar Kundu and Santanu Das 47 Experimental Evaluation of Surface Roughness, Dimensional Accuracy, and MRR in Cylindrical Grinding of EN 24 Steel . . . . . 561 Pankaj V. Mohire and Raju S. Pawade Part III

Automation

48 Bio-inspired Knowledge Representation Framework for Decision Making in Product Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573 Varun Tiwari, Prashant Kumar Jain and Puneet Tandon 49 Heuristic Algorithmic Approach for Automatic Generation of Pin Layout for Robotic Unloading of Sheet Metal Parts . . . . . . . 587 A. Ramesh Babu 50 Voxel-Based Strategy for Efficient CNC Machining . . . . . . . . . . . . 601 A. Kukreja, H. D. Mane, M. Dhanda and S. S. Pande 51 Effect of Geometrical and Process Parameters on Utilization of Sheet Material in Plasma and Laser Cutting Processes . . . . . . . 613 N. Venkatesh, S. Sabari Sriram, V. Satish Chandran and A. Ramesh Babu

xviii

Contents

52 Development of Manufacturability Indices for Prismatic Parts . . . . 625 Manish Kumar Gupta, Pramod Kumar Jain and Abinash Kumar Swain 53 A Cyber-Physical System Architecture for Smart Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 637 Jitin Malhotra, Faiz Iqbal, Ashish Kumar Sahu and Sunil Jha 54 Measurement of Bores Using Scanning Mode of Articulated Arm Coordinate Measuring Machines . . . . . . . . . . . . . . . . . . . . . . . . . . . 649 Ashik Suresh and P. B. Dhanish 55 A Method for Evaluation of Simple Torus Surfaces . . . . . . . . . . . . 659 T. S. R. Murthy 56 An Artefact-Based Continues Performance Verification of Coordinate Measuring Machine . . . . . . . . . . . . . . . . . . . . . . . . . 671 Goitom Tesfay and Rega Rajendra 57 Development of Welding Fixture for Rocket Motor Casing Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 Venkateswarlu Chepuru, P. Kiran and B. Hari Prasad 58 Generation of Sequence of Machining Operations Through Visualization of End Product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689 G. V. S. S. Sharma, P. Srinivasa Rao and B. Surendra Babu 59 Monitoring the Dynamics and Tracking of a Vehicle Using Internet of Things (IoT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699 Sri Harsha Dorapudi, Vivek Varma Buddharaju, V. V. Vimal Varma and R. Ramesh 60 Automated Production of Medical Screws Using Titanium Bar on Indigenous Sliding Headstock Automat . . . . . . . . . . . . . . . . . . . 709 Manohar Bulbule, Naveen Hosamani and S. R. Chandramouli 61 Process Mechanization and Automation for Hybrid TIG MAG Arc Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723 Onkar S. Sahasrabudhe and D. N. Raut 62 Two-Sided Assembly Line Balancing—A Company Case Study Solved by Exact Solution Approach . . . . . . . . . . . . . . . . . . . . . . . . 733 Ashish Yadav and Sunil Agrawal

About the Editors

M. S. Shunmugam is a Professor (Emeritus) in the Manufacturing Engineering Section in the Department of Mechanical Engineering, Indian Institute of Technology (IIT) Madras. After receiving his PhD in Mechanical Engineering from IIT Madras in 1976, he has worked in IIT Bombay (from 1977 to 1980) and in IIT Madras from 1980 onwards. He was a visiting faculty member at Michigan Technological University during 1989-1991 and was a member in the board of governors of IIT Madras during 2012-2013. Dr. Shunmugam’s research interests include metrology, machine tools, manufacturing, gears, micro-machining and computer applications in manufacturing. He has published about 130 peer-reviewed international journal papers, 15 peer-reviewed national journal papers, 75 international conferences and about 80 national conferences. M. Kanthababu is a Professor in the Department of Manufacturing Engineering in Anna University, Chennai, India and the Director of the Centre for Intellectual Property Right and Trade Marks in Anna University. He has completed his MS in Mechanical engineering and PhD in Advanced Manufacturing Technology from IIT Madras. Prof. Kanthababu’s research interests include manufacturing technology, composite materials and machining, and automation in manufacturing. He has published more than 30 peer reviewed international journal papers and 2 books, and holds one patent.

xix

Part I

Forming

Chapter 1

Effect of Heat Treatment on Formability of AA6082 by Single Point Incremental Forming S. Maharajan , D. Ravindran and S. Rajakarunakaran

Abstract This paper investigates the formability of annealed and un-annealed AA6082 sheets by using single point incremental forming (SPIF) process. Lightweight alloys represent key materials for the future of the manufacturing industry, and their use in the emerging fields, such as automobile and aerospace engineering, is a continual topic for researchers of all over the world. It is a reliable area where huge opportunities available for rapid prototyping. In this experimental work, the sheet metal is deformed into the required shape by a hemispherical tool in the vertical CNC milling machine. The straight groove tests are carried out by varying the forming process parameters like feed, speed and step depth. Four experiments are conducted with two sets of different forming process parameters for annealed and un-annealed sheet metals. The formability, forming time and maximum step depth are examined for each experiment. It is found that better formability is achieved in annealed AA6082 aluminium sheet than un-annealed AA6082 aluminium sheet and also observed that step depth plays a major role in forming time and formability. Keywords Single point incremental forming · AA6082 aluminium alloy · Formability · Coordinate measuring machine

1.1 Introduction Incremental forming is a forefront sheet metal forming technology enabling the formation of the desired shape by passing a sheet metal through a series of small incremental plastic deformation. A CNC vertical milling machine traces the movement of forming tool path. A special die is not required for this process. A fixture is enough to hold the sheet metal in all the four sides, and highly sophisticated machines are not required to carry out the forming operation. In this machining operation, forming S. Maharajan (B) · S. Rajakarunakaran Department of Mechanical Engineering, Ramco Institute of Technology, Rajapalayam, India e-mail: [email protected] D. Ravindran Department of Mechanical Engineering, National Engineering College, Kovilpatti, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_1

3

4

S. Maharajan et al.

tool is fitted in the tool holder instead of milling cutter. The CNC vertical machine can carry out the entire operation with its CNC programming. This process is only suitable for low-series production. Manufacturing of a large variety of complicated shapes, which are difficult to form using other forming technology, can be easily formed using incremental forming method. Formability can be enhanced by heat treatment process in the incremental forming technology which in turn will be used for prototyping process. Incremental forming plays a vital role in automotive industry and biomedical applications where flexibility is needed to manufacture the sheet metal with high accuracy and dimensions [1]. By increasing feed rate, formability and surface roughness are increased which simultaneously reduces the forming time. The friction between the tool and sheet can be reduced by freely rotating the ball end tool which is supported with proper lubrication in all the directions [2]. The straight groove test is a suitable method for finding formability. The ball ended forming tool is used for increasing formability with less feed rate by lowering the friction between tool ball end and sheet metal [3]. Feed rate, vertical step down, rotating speed of the tool and lubrication are the major process parameters. Lubrication plays a major role in enhancing the formability. Feed rate is the least factor for maximizing formability and minimizing the surface roughness [4]. An increasing trend in temperature and a reducing trend in surface roughness are observed when tool rotates in both clockwise and anticlockwise directions. Surface roughness reduces insignificantly as the rotating speed of the tool is increased. Simultaneously, the forming forces decrease significantly as the rotating speed of the tool is increased [5]. Formability will increase as the temperature varies from 100 to 250 °C. Progressive forming provides high inclination angle and exceeds forming limits in the incremental forming technology [6]. Heat generation between sheet and tool increases as the rotating speed of the tool increases. Sometimes, such heat is beneficial in enhancing the formability, and at the same time, the surface damage should be assessed due to the friction between tool and surface of the sheet. Formability can be improved with non-hemispherical tool. Step depth plays a major role in enhancing formability. By reducing step depth, formability can be increased [7]. Tool diameter is the significant factor in increasing wall angle, formability and surface roughness. Formability can be increased by reducing the tool diameter and feed rate. By properly controlling the tool diameter, surface roughness can be reduced [8]. Surface finish is significantly good in curved path when compared to straight path of the tool. Hence, higher the contact between tool and workpiece, greater will be the surface roughness. Smooth curve tool path enhances formability which, in turn, increases the flow of material [9]. Forming forces are reduced by reducing vertical step depth. Surface roughness is reduced by controlling the step depth. Thickness distribution is uniform when wall angle is greater than 65° [10]. Forming force could be reduced by increasing the grain size of the sheet. Suitable heat treatment temperature should be maintained to control the grain size [11]. In the hybrid optimization technique, it was found that feed rate is the most important factor, followed by vertical depth and the tool diameter [12]. Appropriate heat treatment process provides sheet material with homogeneous and fine microstructure. Heat treatment process increases the strength of the component. Various experiments conducted reveal that the formability can be significantly

1 Effect of Heat Treatment on Formability of AA6082 …

5

increased for heat-treated sheet material in the tube hydroforming of Cu alloy and Al alloy materials [13]. On comparing stretch bending with single point incremental forming, it was found that SPIF formability enhancement is more than stretch bending operation. Forming force increases as the tool diameter increases. This is because contact area is more in large tool diameter. At the same time, higher step down is possible with higher tool diameter which reduces the forming time [14]. Since the localization of strain can be inhibited in incremental forming technique, forming limit is similar for both flattened sheet metal and uniform sheet metal. So, this process is suitable for cold recycling process of sheet metal waste [15]. The microstructure study is important in understanding the formable behaviour of Al sheet material in any forming technology. The necking ductility can be improved with the combination of coarse and fine grain microstructure. Aluminium alloy microstructure can be refined due to the presence of Mg [16]. The expressions used to determine major strains and minor strains are given as follows, as developed by Narayanasamy and Sathiya Narayanan [17]: Major strain (e1) = ln(Major diameter of the ellipse/Original diameter of the circle) Minor strain (e2) = ln(Minor diameter of the ellipse/Original diameter of the circle) Fully recrystallized microstructure with good formability can be obtained at 350 °C annealing temperature for Al 1145 alloy sheet. Heat treatment of aluminium alloy enhances the formability of sheet metal [18]. The forming limit curve in incremental forming is quite different from that in conventional forming. It appears to be a straight line with a negative slope in the positive regions of the minor strains in the forming limit diagram [19]. Most of the researchers have carried out experiments for un-annealed sheet metal by varying process parameters like feed, speed and step depth and found the formability. From the literature survey, it is evident that annealed sheet metal formability work in the SPIF has not been carried out so far. In this work, a new aluminium alloy AA6082 is used for which annealing process has been carried out. Response parameters such as formability, forming time and maximum step depth are experimentally determined and discussed. From the experimental work, it is clear that the forming time had substantially reduced in the annealed sheet metal when compared to un-annealed sheet metal. Generally, the annealing process is carried out to refine crystal structure, improve ductility and relieve internal stresses. In addition to this, it is found that the annealing heat treatment process also helps to improve the formability of the sheet metal and achieve maximum step depth at the time of fracture of the sheet metals.

6

S. Maharajan et al.

1.2 Experimental In this work, AA6082 aluminium alloy sheet (Cu—0.06%, Mg—0.97%, Si—1.09%, Fe—0.22%, Mn—0.46%, Cr—0.21%, Al—remaining) was deformed to the desired shape by using incremental forming for straight groove test. AA6061 aluminium alloy sheet has been replaced with AA6082 aluminium alloy sheet in many applications, since AA6082 gives higher strength. The addition of a large amount of manganese controls the grain structure which in turn results in a stronger alloy. The sheet metal thickness was 0.5 mm and was cut into dimensions of 200 × 200 mm. The laser marked 5-mm grid circle was printed on the two sheet metals, and another two sheet metals were printed with 2-mm grid circle to facilitate strain measurement after and before deformation. High-carbon and high-chromium steel tool is used. It has high wear and abrasion resistant properties. It is heat treatable and will offer hardness in the range 55–62 HRC and is machinable in the annealed condition. The D2 tool was used for this research. This tool was made with hemispherical end. Cold-work tool steels include the high-carbon, high-chromium steels or group D steels. These steels are designated as group D steels and consist of D2, D3, D4, D5 and D7 steels. These steels contain 1.5–2.35% of carbon and 12% of chromium with elastic modulus of 210 GPa. Tool dip diameter is 5 mm, and the length of the tool is 100 mm. The D2 tool steel picture is shown in Fig. 1.1. AA6082 sheet metal with 5-mm grid circle forming is shown in Fig. 1.2. Grid circle diameter and deformed grid circle dimensions after forming process were measured in coordinate measuring machine (CMM) in Ramco Institute of Technology, Rajapalayam. A CNC vertical milling machine specification is Travel, x-axis: 810 mm, y-axis: 510 mm and z-axis: 510 mm; Maximum Spindle speed, 6000 rpm; and Maximum Feed, 1000 mm/min. The incremental forming operation is carried out in CNC vertical milling machine at Precision Automation in Trichy as shown in Fig. 1.3. The input forming process parameters were spindle speed, feed and step depth. Favourable outputs were formability, forming time and maximum step depth achieved at the time of fracture. Four experiments were carried out for two different conditions with three input process parameters for annealed and un-annealed sheets. Response parameters Fig. 1.1 High-carbon and high-chromium tool steel (D2 tool steel)

Fig. 1.2 AA6082 sheet metal with 5-mm grid circle

1 Effect of Heat Treatment on Formability of AA6082 …

7

Fig. 1.3 Incremental forming process in CNC vertical milling machine set-up

Table 1.1 Experiments carried out for two different conditions with three input process parameters (annealing and un-annealing AA6082 aluminium sheets) Exp. No.

AA6082 alloy sheet condition

Spindle speed (rpm)

Feed (mm/min)

Step depth (mm)

1

Un-annealing

1200

100

0.2

2

Un-annealing

1400

150

0.4

3

Annealing

1200

100

0.2

4

Annealing

1400

150

0.4

were found through experiments and compared between annealed and un-annealed sheet metal. Most importantly, formability was calculated for each experiment and compared annealed sheet with un-annealed sheet. The results were discussed. Four experiments details are given in Table 1.1.

1.3 Results and Discussions 1.3.1 Analysis of Formability of the Un-annealed AA6082 Sheet Metal Un-annealed Aluminium Alloy 6082 Sheet 1 (First Experiment). First experiment was carried out for un-annealed sheet metal with the input forming process parameters such as spindle speed of 1200 rpm, feed of 100 mm/min and step depth of 0.2 mm. The time taken for forming operation was 23 min 10 s. Maximum depth of forming was achieved with 11.4 mm at the time of fracture. The deformed sheet metal is shown in Fig. 1.4. The original grid circle diameter is 5 mm. After incremental forming operation, the elliptical shape grid major axis length and minor axis length were measured by using CMM. The coordinate Measuring Machine (CMM) is shown in Fig. 1.5.

8

S. Maharajan et al.

Fig. 1.4 AA6082 deformed sheet metal after forming operation

Fig. 1.5 Measuring enlarged grid circle major and minor axis length using CMM

After forming operation, the experimental values were tabulated. The formability was calculated by using the formula as developed by Narayanasamy and Sathiya Narayanan [17]: Major strain (e1) = ln(Major diameter of the ellipse/Original diameter of the circle) (1.1) Minor strain (e2) = ln(Minor diameter of the ellipse/Original diameter of the circle) (1.2) As per modified Cockroft–Latham criterion [20], Formability = Major strain (e1) + Minor strain (e2)

(1.3)

The formability values are calculated and tabulated in the below-mentioned Table 1.2. The formability limit curve (FLC) was plotted for major strain versus minor strain for the un-annealed AA6082 sheet 1 for the first experiment. The FLC is shown in Fig. 1.6. Un-annealed Aluminium Alloy 6082 Sheet 2 (Second Experiment). Second experiment was carried out for un-annealed sheet metal with the input forming process parameters such as spindle speed of 1400 rpm, feed of 150 mm/min and step depth of 0.4 mm. The time taken for forming operation was 7 min 36 s. Maximum depth of forming was achieved with 10.8 mm at the time of fracture. The original grid circle diameter is 5 mm. After incremental forming operation, the elliptical shape grid major axis length and minor axis length were measured by using CMM. The formability values are calculated and tabulated in Table 1.3.

1 Effect of Heat Treatment on Formability of AA6082 …

9

Table 1.2 Formability analysis for un-annealed sheet 1 with step depth of 0.2 mm Major circle dia (mm)

Minor circle dia (mm)

Major strain

Minor strain

Formability

12.0422

4.3427

0.878979

−0.1409

0.738037

12.0242

4.0731

0.877483

−0.2050

0.672449

12.0015

4.2314

0.875594

−0.1669

0.708689

12.0059

4.3481

0.87596

−0.1397

0.736261

11.9978

4.0527

0.875285

−0.2100

0.66523

11.9084

4.0735

0.867806

−0.2049

0.662871

12.0042

4.0308

0.875819

−0.2154

0.660346

5.2053

4.9846

0.040239

−0.0030

0.037154

Average value

0.610129

Fig. 1.6 Formability limit curve for un-annealed sheet 1 (experiment 1) Table 1.3 Formability analysis for un-annealed sheet 2 with step depth of 0.4 mm Major circle dia (mm)

Minor circle dia (mm)

Major strain

Minor strain

Formability

8.1247

4.9716

0.485471

−0.0057

0.479775

8.1103

4.9732

0.483697

−0.00537

0.478323

8.1468

4.8999

0.488187

−0.02022

0.467964

8.2001

4.8536

0.494708

−0.02972

0.464991

8.0108

4.9756

0.471353

−0.00489

0.466461

8.3096

4.9782

0.507974

−0.00437

0.503604

8.4132

4.9756

0.520364

−0.00489

0.515472

8.2132

4.9816

0.496305

−0.00369

0.492618

Average value

0.483651

10

S. Maharajan et al.

Fig. 1.7 Formability limit curve for un-annealed sheet 2 (experiment 2)

The FLC was plotted for major strain versus minor strain for the un-annealed AA6082 sheet 2 for the second experiment. The FLC is shown in Fig. 1.7. From the above results, it was confirmed that the formability was decreased with increasing step depth from 0.2 to 0.4 mm. At the same time, forming time was decreased with increasing the step depth 0.4 mm from 0.2 mm, spindle speed of 1400 rpm from 1000 rpm and feed rate of 150 mm/min from 100 mm/min. Maximum amount of depth could be achieved in 0.2 mm step depth with feed rate of 100 mm/min and spindle speed of 1000 rpm in the first experiment when compared to second experiment process parameters, such as 0.4 mm step depth, 150 mm/min and spindle speed of 1400 rpm.

1.4 Analysis of Formability of the Annealed AA6082 Sheet Metal Annealed Aluminium Alloy 6082 Sheet 3 (Third Experiment). AA6082 aluminium alloy sheet metal was heated to 400 °C for 2 hours in electric arc furnace and the AA6082 sheet was subsequently cooled to room temperature in the furnace itself for 2 hours. Thus, AA6082 was annealed. Annealing process increased the ductility. Third experiment was carried out for annealed sheet with input forming process parameters such as spindle speed of 1200 rpm, feed rate of 100 mm/min and step depth of 0.2 mm. The time taken for forming operation was 9 min 35 s. Maximum depth of forming was achieved with 14 mm at the time of fracture. The original circle diameter was 2 mm for annealed sheet. After incremental forming operation, the

1 Effect of Heat Treatment on Formability of AA6082 …

11

Table 1.4 Formability analysis for annealed sheet 3 with step depth of 0.2 mm Major circle dia (mm)

Minor circle dia (mm)

Major strain

Minor strain

Formability

4.8333

2.1344

0.882382

0.065038

0.947421

4.8786

2.1098

0.891711

0.053446

0.945157

4.8361

2.1567

0.882961

0.075432

0.958393

4.8685

2.1309

0.889639

0.063397

0.953036

4.8982

2.1873

0.895721

0.089521

0.985242

4.8594

2.1897

0.887768

0.090617

0.978385

4.8051

2.1389

0.876531

0.067144

0.943675

4.8754

2.1534

0.891055

0.073901

0.964956

Average value

0.959533

elliptical shape grid major axis length and minor axis length was measured by using CMM. The formability values are calculated and tabulated in the below-mentioned Table 1.4. The FLC was plotted for major versus minor strain for the annealed AA6082 sheet 3 for the third experiment. The FLC is shown in Fig. 1.8. From the above results, the forming time was reduced to 9 min to 35 s for annealed sheet when compared to un-annealed sheet whose forming time was 23 min 10 s for the same input forming process parameter. Maximum depth of forming was increased up to 14 mm for annealed sheet when compared to un-annealed sheet whose maximum depth of forming was recorded as 11.4 mm. The formability was increased to 0.959533 for annealed sheet when compared to un-annealed sheet formability, recorded as 0.61013. Fig. 1.8 Formability limit curve for annealed sheet 3 (experiment 3)

12

S. Maharajan et al.

Table 1.5 Formability analysis for annealed sheet 4 with step depth of 0.4 mm Major circle dia (mm)

Minor circle dia (mm)

Major strain

Minor strain

Formability

4.1237

2.1005

0.723604

0.049028

0.772632

4.1678

2.1045

0.734241

0.050931

0.785172

4.1136

2.1056

0.721151

0.051453

0.772604

4.1568

2.1567

0.731598

0.075432

0.807031

4.1983

2.1344

0.741533

0.065038

0.806571

4.1459

2.1098

0.728973

0.053446

0.782419

4.1005

2.1567

0.717962

0.075432

0.793394

4.1547

2.1309

0.731093

0.063397

0.794491

Average value

0.789289

Annealed Aluminium Alloy 6082 Sheet 4 (Fourth Experiment). Fourth experiment was carried out for annealed sheet metal with the input forming process parameters such as spindle speed of 1400 rpm, feed of 150 mm/min and step depth of 0.4 mm. The time taken for forming operation was 7 min 35 s. Maximum depth of forming was achieved with 13.4 mm at the time of fracture. The formability values are calculated and tabulated in the below-mentioned Table 1.5. The FLC was plotted for major versus minor strain for the annealed AA6082 sheet 4 for the fourth experiment. The FLC is shown in Fig. 1.9. The results shown in Table 1.6 reveal that annealing is applied to AA6082 sheet to stimulate softening, refine its crystal structure, strengthen its bending properties and consistency in the plastic deformation and reduce the chances of cracking during the forming process. Annealing process also improves the ductility and toughness of the sheet metal. Ductility is the extent to which a material can undergo plastic Fig. 1.9 Formability limit curve for annealed sheet 4 (experiment 4)

AA6082 alloy sheet condition

Un-annealing

Un-annealing

Annealing

Annealing

Exp. No.

1

2

3

4

Table 1.6 Results

1400

1200

1400

1200

Spindle speed (rpm)

150

100

150

100

Feed (mm/min)

0.4

0.2

0.4

0.2

Step depth (mm)

0.789289

0.959533

0.483651

0.610129

Formability

7 min 35 s

9 min 35 s

7 min 36 s

23 min 10 s

Forming time

13.4

14.0

10.8

11.4

Max step depth (mm)

1 Effect of Heat Treatment on Formability of AA6082 … 13

14

S. Maharajan et al.

deformation, that is, the extent to which a material can be plastically deformed before fracture. Hence, the sheet metal would undergo a large amount of strain hardening due to improved ductility and toughness. Plastic deformation is uniformly distributed throughout a particular section of the material during the forming process. The results show that better formability could be achieved with the annealing process.

1.5 Conclusion In this research, two un-annealed AA6082 sheet metals and two annealed AA6082 sheet metals were deformed with a straight groove test in the single point incremental forming process by CNC vertical milling machine. The following conclusions were made while conducting the experiments by varying input forming process parameters such as speed, feed and step depth for annealed and un-annealed conditions. 1. Formability was significantly increased with reducing step depth. 2. Forming time was substantially reduced in the annealed sheet metal when compared to un-annealed sheet metal. 3. Formability was greatly increased with annealed sheet metal when compared to un-annealed sheet metal. 4. Step depth plays a major role in the formability. Since both annealed and unannealed sheet formability is enhanced with reducing the step depth to the level of 0.2 mm. Among all the experiments, better formability can be obtained in the annealed AA6082 sheet metal with input forming process parameter such as spindle speed of 1200 rpm, feed of 100 mm/min and step depth of 0.2 mm. Maximum depth of forming was achieved with 14 mm which was highest value achieved among all experiments, and forming time was 9 min 35 s which was less when compared to other forming experiments. Acknowledgements The authors would like to thank Ramco Institute of Technology, Rajapalayam, and National Engineering College, Kovilpatti, for providing excellent infrastructure facilities to carry out experiments and take measurements in coordinate measuring machine and also appreciate our students to extend their support to do the experimental works.

References 1. Emmens, W.C., Sebastiani, G., Van den Boogaard, A.H.: The technology of incremental sheet forming—a brief review of the history. J. Mater. Process. Technol. 210, 981–997 (2010) 2. Pandivelan, C., Jeevanantham, A.K.: Multi objective optimization of single point incremental sheet forming of AA5052 using Taguchi based grey relational analysis coupled with principal component analysis. Int. J. Precis. Eng. Manuf. 15, 2309–2316 (2014)

1 Effect of Heat Treatment on Formability of AA6082 …

15

3. Kim, Y.H., Park, J.J.: Effect of process parameters on formability in incremental forming of sheet metal. J. Mater. Process. Technol. 130–131, 42–46 (2002) 4. Baruah, A., Pandivelan, C., Jeevanantham, A.K.: Optimization of AA5052 in incremental sheet forming using grey relational analysis. Measurement 106, 95–100 (2017) 5. Durante, M., Formisano, A., Langella, A., Memola Capece Minutolo, F.: The influence of tool rotation on an incremental forming process. J. Mater. Process. Technol. 209, 4621–4626 (2009) 6. Ji, Y.H., Park, J.J.: Formability of magnesium AZ31 sheet in the incremental forming at warm temperature. J. Mater. Process. Technol. 201, 354–358 (2008) 7. McAnulty, T., Jeswiet, J., Doolan, M.: Formability in single point incremental forming: a comparative analysis of the state of the art. CIRP J. Manuf. Sci. Technol. 16, 43–54 (2017) 8. Shanmuganatan, S.P., Senthil Kumar, V.S.: Modeling of incremental forming process parameters of Al3003(O) by response surface methodology. Procedia Eng. 97, 346–356 (2014) 9. Mugendirana, V., Gnanavelbabu, A. Comparison of FLD and thickness distribution on AA5052 aluminium alloy formed parts by incremental forming process. Procedia Eng, 97, 1983–1990 (2014) 10. Malwad, D.S., Nandedkar, V.M.: Deformation mechanism analysis of single point incremental sheet Metal forming. Procedia Mater. Sci. 6, 1505–1510 (2014) 11. Shrivastava, P., Tandon, P. Investigation of effect of grain size of forming forces in Single Point Incremental Sheet Forming. Procedia Manuf. 2, 41–45 (2015) 12. Raju, C., Sathiya Narayanan, C.: Application of hybrid optimization technique in a multiple sheet single point incremental forming process. Measurement 78, 296–308 (2016) 13. Fatemi, A., Morovvati, M.R., Biglari, F.R.: The effect of tube material, microstructure, and heat treatment on process responses of tube hydroforming without axial force. Int. J. Adv. Manuf. Technol. (2009) 14. Centeno, G., Bagudanch, I., Martínez-Donaire, A.J., García-Romeu, M.L., Vallellano, C.: Critical analysis of necking and fracture limit strains and forming forces in single point incremental forming. Mater. Des. 63, 20–29 (2014) 15. Takan, H., Kitazawa, K., Goto, T.: Incremental forming of nonuniform sheet metal: possibility of cold recycling process of sheet metal waste. Int. J. Mach. Tools Manuf. 48, 477–482 (2008) 16. Chandra Sekhar, K., Narayanasamy, R., Velmanirajan, K.: Experimental investigations on microstructure and formability of cryorolled AA5052 sheets. Mater. Des., 53, 1064–1070 (2014) 17. Narayanasamy, R., Sathiya Narayanan, C.: Some aspects on fracture limit diagram developed for different steel sheets. Mater. Sci. Eng. A 417, 197–224 (2006) 18. Velmanirajan, K., Syed Abu Thaheer, A., Narayanasamy, R., Madhavan, R., Suwas, S.: Effect of annealing temperature in Al 1145 alloy sheets on formability, void coalescence, and texture analysis. J. Mater. Eng. Perform. 22, 1091–1107 (2013) 19. Shim, M.-S., Park, J.-J.: Formability of aluminium sheet in incremental forming. J. Mater. Process. Technol. 113, 654–658 (2001) 20. Martins, P.A.F., Bay, N., Skjoedt, M., Silva, M.B.: Theory of single point incremental forming. CIRP Ann. Manuf. Technol. 57, 247–252 (2008)

Chapter 2

Forming Behavior of AA5052-H32 and AA6061-T6 During Single Point Incremental Forming M. M. Ghadmode , R. R. Pawar

and B. U. Sonawane

Abstract In present work, an attempt was made to study the forming behavior of aluminum alloy AA5052-H32 and AA6061-T6 during a forming process called “single point incremental forming (SPIF).” Forming behavior of 1.5 mm thick material in terms of forming depth, forming limit strains and thickness reduction are studied by varying wall angles from 40° to 70°. SPIF was carried on a VMC-PVM 40, the HSS tool of diameter 10 mm having spherical end and the lubricant 10W30 was used. Tool paths for various samples are generated using ARTCAM software. Results reveal that the formability of AA5052-H32 is better than that of AA6061-T6. It is also observed that the limiting wall angle for forming AA5052-H32 is 70°, whereas the same is 50° for AA6061-T6. Keywords Incremental forming · SPIF · Forming limit diagram · Mylar tape

2.1 Introduction In today’s manufacturing scenario, much research is carried out to reduce the product development time at its initial stages. One such process is single point incremental forming (SPIF), which forms the sheet metal incrementally without the use of complex tooling. Thus, single point incremental forming process can be used for prototyping various symmetric and asymmetric components. Jeswiet et al. [1] had noted that forming limit of sheet metal is increased in SPIF process than those in conventional-sheet metal stamping process. Thus, gaining the advantages of forming the sheets with local deformation, an increase in formability of the sheets can be achieved. The advancements in incremental sheet forming have made easy to produce various sheet metal parts by generating tool paths for different shapes and sizes without investing time to develop new die. The SPIF process parameters such as

M. M. Ghadmode (B) · R. R. Pawar · B. U. Sonawane Department of Production Engineering & Industrial Management, College of Engineering Pune, Pune 411005, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_2

17

18

M. M. Ghadmode et al.

speed (rpm), feed (mm/min), step size (mm), and tool diameter (mm) are influencing the formability, surface roughness, and thinning of sheets. The process parameters have the significant effect on the surface finish of the part. Bhattacharya and Maneesh [2] showed that the surface roughness increases with increase in tool diameter and step depth, but decreased up to certain wall angle and beyond that angle it increased with increase in incremental depth. Malwad and Nandedkar [3] conducted the multi-angle test and stated that more thickness reduction can be achieved at larger wall angle, but uniform thickness distribution was achieved at wall angles less than 65°. Shojaeefard et al. [4] studied effect of tool diameter, vertical step, feed, and speed on waviness of the component formed by SPIF and concluded that the use of higher tool diameter and minimum vertical step leads to lesser waviness. Many researchers have studied effect of wall angle over surface roughness. But very few research is available to find the forming limit strains. Most of researcher’s have used screen printing to mark circular grids on backside of the sheet before forming. In this research work, we have engraved circular grids by laser engraving on backside of each sheet, which helped us to plot forming limit strains and thickness strains. The focus was kept on finding out the forming limit strains by varying angle from 40° to 70° for aluminum alloy AA5052-H32 and AA6061-T6 having 1.5 mm thickness during single point incremental forming.

2.2 Experimental Procedure The experimental setup, as shown in Fig. 2.1, consists of a SPIF fixture, which was mounted on a VMC-PVM 40, the HSS tool of diameter 10 mm having spherical end and the lubricant 10W30 was used. This fixture, which is used to clamp the sheet during incremental forming, is having one backing plate with circular hole of 170 mm diameter, a clamping plate, 4 rods, and clamping screws. To get easy material flow, 6 mm fillet radius is also provided at the hole of backing plate. Tool path is generated using ARTCAM software for forming conical shape having 130 mm bottom diameter and depth was kept to its maximum extent for the forming process.

SPIF Fixture to clamp the blank Aluminum alloy Blank

Fig. 2.1 Experimental setup of SPIF

Spherical End Tool mounted on VMC

2 Forming Behavior of AA5052-H32 and AA6061-T6 …

19

2.2.1 Blank Material The blank materials are aluminum alloys AA5052-H32 and AA6061-T6 (both 1.5 mm thick), which are commercially utilized in automobile sector. The mechanical properties and chemical composition are as given in Tables 2.1 and 2.2, respectively. A set of experiment was carried out wherein the process parameters such as spindle speed (n), step size (z), tool diameter (dt), and feed (ƒ), which were kept constant and the wall angles were varied at 40°, 45°, 50°, 55°, 60°, 65°, and 70°. These experiments are done twice to take care of variations and the average of the output is calculated to plot major strain, minor strain, and thickness strain (Table 2.3). The blanks were cut in 176 mm × 176 mm size, which were further marked by laser engraving as shown in Fig. 2.2. Most of researcher’s have used screen printing to mark circular grids on backside of the sheet before forming. In experimental techniques, we have engraved circular grids by laser engraving on backside of each sheet which helped us to plot forming limit strains and thickness strains. As a standard practice, diameter of circle to be kept up to 3 mm, but for economical point of view we have selected circular grids of 5 mm diameter, touching each other are marked over the surface area of 160 mm × 160 mm on one side of the blanks. The power of laser was kept to 100 W and a hair-line thickness was kept for the circles engraved by laser engraving machine. These circles get elongated into ellipse after SPIF process. The conical shape was chosen for the experiment with bottom diameter of 130 mm. Table 2.1 Blank material (mechanical properties) Property

AA5052-H32

AA6061-T6

Tensile strength (MPa)

240

368

Yield strength (MPa)

193

279

Total elongation (%)

10.4

10

Table 2.2 Chemical compositions (%) of blank material Element

Si

Fe

Cu

Mn

AA5052-H32

0.120

0.300

0.005

0.095

AA6061-T6

0.64

0.34

0.25

1.038

Element

Mg

Cr

Zn

Ti

AA5052-H32

2.463

0.191

0.001

0.015

AA6061-T6

1.03

0.193

0.004

0.023

Table 2.3 Parameters kept constant during experiment Control factor

Speed (rpm)

Step size (mm)

Tool dia (mm)

Feed (ƒ) (mm/min)

Level

1000

0.2

10

1000

20

M. M. Ghadmode et al.

Numbers are marked on in rolling direction (RD)

Fig. 2.2 Circular marking on blank

Fig. 2.3 SPIF tool (φ10 mm)

For the continuous downward motion of the tool during incremental forming, a spiral tool path was generated using Autodesk ARTCAM software (Fig. 2.3).

2.2.2 Strain Measurement During single point incremental forming, usually stretching takes place and all the circles, which come under the forming area get elongated due to this stretching effect. Circles having initial diameter of 5 mm were elongated and the shape of the circle became elliptical shape after SPIF process. As the material is isotropic in nature, and hence material got stretched uniformly in all directions. The measurement was carried for these elongated circles along the center line and along diagonal of the sheet (as marked in Fig. 2.2). For measurement of major and minor diameter of these ellipses formed after incremental forming, a Mylar tape was used [5] as shown in Fig. 2.4. For the correctness of the measurement, a magnifying glass (3x zoom) was used. The major strain and minor strain was obtained by using the following equation as suggested by Kumar [5], L n ε= Lo

L n major or minor diameter of ellipse L o original diameter of the circle.

Ln dL = ln L Lo

(1.1)

2 Forming Behavior of AA5052-H32 and AA6061-T6 …

21

Fig. 2.4 Mylar Tape and magnifying glass

The strains were plotted as major strain on vertical axis and minor strain on horizontal axis. It was followed for each wall angle for both materials AA5052-H32 and AA6061-T6 having 1.5 mm thickness.

2.3 Results and Discussion Formability can be measured in terms of forming limit strains, thickness distribution (uniform thickness reduction), and forming depth achieved. The following results are obtained from the experiments carried out.

2.3.1 Formability Measured in Terms Forming Strains Figure 2.5a–c shows the conical-shaped components formed by SPIF process from materials mentioned above for wall angles 40°, 45°, 50°, 55°, 60°, 65°, and 70°. It clearly observed from Fig. 2.5a that AA5052-H32 has been successfully formed using SPIF up to 65° and no cracks are observed. A crack is developed in the sample formed with wall angle 70°. It can be seen from Fig. 2.5b that cracks are developed in AA6061-T6 in samples having wall angles 50° and above. Enlarged view of crack developed in failed components 50° (AA6061-T6) and 70° (AA5052-H32) can be seen in Fig. 2.5c. Figure 2.6a–g shows the comparison between strains developed along centerline of the sheet after forming the sheets using SPIF process. It can be seen from Fig. 2.6a, b that both the materials are formed successfully with no crack developed for wall

22

M. M. Ghadmode et al.

65°

60°

40° 45°

60°

65°

70°

55°

55°

40°

50°

(a) AA5052-H32 samples

50°

45°

50°

(b) AA6061-T6 samples

70°

(c) Failed component 500 (AA6061-T6) & 700 (AA5052-H32) Fig. 2.5 a–c Conical components ranging from 40° to 70° formed by SPIF

angles 40° and 45°. From angle 50° onwards, in material AA6061-T6 cracks started developing at major strain value of 0.52 (for 50°) and goes on decreasing this value to 0.44 (for 55°), 0.40 (for 60°), and 0.40 (for 65°) as wall angle increases. Whereas, the corresponding values for material AA5052-H32 are 0.46 (for 50°), 0.60 (for 55°), 0.74 (for 60°), and 0.98 (for 65°) and cracks are not developed in AA5052-H32. In other words, with increase of wall angle, more straining of the component takes place, but material AA6061-T6 fails early than material AA5052-H32 (even though the amount of major strain is more as compared with that of AA6061-T6). It is also observed that, the component is getting stretched uniformly in all the region of sheet, only the region in contact with tool radius is stretched more and failure occurs in this region only. For the above reason, it is necessary to construct the thickness distribution of the sheets. Thickness can be obtained from the data collected on major strain and minor strain.

2 Forming Behavior of AA5052-H32 and AA6061-T6 …

0.25

40°_5052 40°_6061

0.20

23

45°_5052

0.35

45°_6061

0.30

Major Strain

Major Strain

0.25 0.15 0.10 0.05

-0.20

-0.10

0.00 0.00 0.10 Minor Strain

0.20

0.00 -0.20

-0.10

0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 -0.10 0.00 0.10 Minor Strain

-0.20

0.50

Major Strain

Major Strain

0.60 0.40 0.30 0.20

(e) Wall angle 60ᵒ

1.00 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 -0.10 0.00

65°_5052 65°_6061

0.70

0.00 -0.10 0.00 0.10 Minor Strain

0.20

(d) Wall angle 55ᵒ

0.10 -0.20

0.20

Major Strain

Major Strain

0.20

0.80

60°_6061

0.00 0.10 Minor Strain

55°_5052 55°_6061

(c) Wall angle 50° 60°_5052

0.10

(b) Wall angle 45ᵒ

50°_5052 50°_6061

-0.20

0.15

0.05

(a) Wall angle 40ᵒ 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 -0.10 0.00 0.10 Minor Strain

0.20

0.20

-0.20

0.10

Minor Strain (f) Wall angle 65ᵒ

Fig. 2.6 a–g Strain measurement along centerline for AA5052-H32 versus AA6061-T6

0.20

24

M. M. Ghadmode et al.

Major Strain

70°_5052 70°_6061

-0.20

-0.10

0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 0.00 0.10 Minor Strain

0.20

(g) Wall angle 70ᵒ

Fig. 2.6 (continued)

2.3.2 Formability Measured in Terms Thickness Distribution Figure 2.7a–f shows the thickness distribution compared with the desired thickness reduction by sine law [6]. Figure 2.7a, b exhibits the results of thickness distribution for wall angle 40° and 45°, which clearly indicates that the thickness distribution is similar for both materials, and values are near to sine law of thinning. The peak of the curve at center indicates that tool has not reached to that point, and hence forming is not carried out at that location, and hence no thickness reduction is observed. Figure 2.7c–g indicates that there is a local thinning occurred in material AA6061T6 and hence material failed. Whereas, AA5052-H32 material is stretched fully following sine law of thinning and it got failed at 70° wall angle (please refer Fig. 2.7g) due to local thinning just similar to the local thinning occurred in AA6061-T6 at 50° wall angle and onwards.

2.3.3 Formability Measured in Terms of Forming Depth Forming depth of each formed conical cup is measured by digital height gauge LH600E (Mitutoyo Make). Three random readings have been taken for each component and the average height of the same are shown in Table 2.4 and Fig. 2.8. It can be clearly seen from Fig. 2.8 that, for material AA5052-H32, forming depth increases as wall angle increases up to 65° and reduces for 70°. Whereas, for material AA6061-T6, forming depth increase up to 45° and starts reducing for wall angle 50° and onwards. It can be concluded from the above point that formability of AA6061-T6, in terms of forming depth, is less than that of material AA5052-H32.

Sheet Thickness in mm

2 Forming Behavior of AA5052-H32 and AA6061-T6 …

1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5

Sine Law_40°

40°_5052

25

40°_6061

32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Node points along center line

Sheet Thickness in mm

(a) Wall angle 40ᵒ

1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5

Sine Law_45°

45°_5052

45°_6061

32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Node points along center line

Sheet Thickness in mm

(b) Wall angle 45ᵒ

1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5

Sine Law_50°

50°_5052

50°_6061

32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Node Points along center line (c) Wall angle 50ᵒ Fig. 2.7 a–g Thickness distribution along centerline of formed blanks

26

M. M. Ghadmode et al.

Sheet Thickness in mm

(c) Wall angle 50ᵒ

55°_5052

Sine Law_55°

1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5

55°_6061

32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Node Points along center line

Sheet Thickness in mm

(d) Wall angle 55ᵒ

Sine Law_60°

1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5

60°_5052

60°_6061

32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Node Points along Center Line

Sheet Thickness in mm

(e) Wall angle 600

Sine Law_65°

1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5

65°_5052

65°_6061

32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Node Points along Center Line (f) Wall angle 65°

Fig. 2.7 (continued)

Sheet Thickness in mm

2 Forming Behavior of AA5052-H32 and AA6061-T6 …

70°_6061

70°_5052

Sine Law_70°

1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4

27

32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Node Points along Center Line

Fig. 2.7 (continued) Table 2.4 Forming depth (average) for both materials Wall Angle (°)

AA5052-H32_1.5t (mm)

AA6061-T6_1.5t (mm)

40

48.72

49.55

45

54.79

52.08

50

64.23

12.77

70.43

15.85

84.34

12.81

65

115.47

11.54

70

16.02

12.95

Forming depth (in mm)

55 60

120 110 100 90 80 70 60 50 40 30 20 10 0 35

5052_1.5t

6061_1.5t 115.47 84.34

64.23 48.72

49.55

70.43

54.79

52.08 15.85

12.81

11.54

16.02

12.77

40

45

12.95

50

55

60

Wall angle (in degrees) Fig. 2.8 Forming depth versus wall angle

65

70

75

28

M. M. Ghadmode et al.

2.4 Conclusions Formability of AA6061-T6 and AA5052-H32 are studied by varying wall angle which are 40°, 45°, 50°, 55°, 60°, 65°, and 70° by keeping all other parameters are kept constant, and it can be concluded from the results obtained by experiments carried out that, the limiting angles for AA6061-T6 is 50° and it was 70° for AA5052-H32. More forming depth, i.e., up to 115.47 mm can be achieved for material AA5052-H32 with a wall angle of 65°, whereas material AA6061-T6 is formed only 11.54 mm at wall angle of 65°. In other words, the formability of material AA5052-H32 is better as compared with material AA6061-T6. Acknowledgements We are thankful to the Department of Production Engineering and Industrial Management, College of Engineering Pune, for their technical and financial support for this research work.

References 1. Jeswiet, J., Micari, F., Hirt, G., Bramley, A., Duflou, J., Allwood, J.: Asymmetric single point incremental forming of sheet metal. CIRP Ann. Manuf. Technol. 54, 88–114 (2005) 2. Bhattacharya, A., Maneesh, K.: Formability and surface finish studies in single point incremental forming. J. Manuf. Sci. Eng. 133, 061020-1–061020-8 (2011) 3. Malwad, D.S., Nandedkar, V.M.: Deformation mechanism analysis of single point incremental sheet metal forming. Procedia Mater. Sci. 6, 1505–1510 (2014) 4. Shojaeefard, M.H., Khalkhali, A., Shahbaz, S.: Analysis and optimization of the surface waviness in the single-point incremental sheet metal forming. Proc. IMechE Part E: J. Process Mech. Eng. IMech 0(0), 1–7 (2018) 5. Kumar, R.: Analysis of major strains and minor strains in sheet metal forming. Int. J. Appl. Innov. Eng. Manag. 2, 194–198 (2013) 6. Oleksik, V.: Influence of geometrical parameters, wall angle and part shape on thickness reduction of single point incremental forming. Procedia Eng. 81, 2280–2285 (2014)

Chapter 3

Pull-Out Forming: Experiments and Process Simulation S. Kumar

Abstract Pull-out forming is a forming process/technology wherein a tubular work material is pulled out from a hole opening at the surface of a tube to produce a product that is used to join with another tube for a high-pressure application. This forming process results in good accuracy, surface finish, better mechanical and metallurgical properties at a lower unit cost. A simulation methodology has been proposed for the pull-out forming process including design validation and experiments. The experimental set-up is fabricated to carry out several test runs to validate the simulation. The FE simulation study of pull-out process has been done at room temperature on tubular shape workpiece of pure aluminium metal. The result is used to investigate the effective stress distribution, maximum deforming load and the wall thickness of pull-out under various process parameter conditions (speed, friction, no. of pull-out, etc.). Furthermore, the pull-out forming process has been optimized in order to get the uniform thickness and defect free product. Keywords Pull-out forming · Computer aided engineering · Simulation · FEM

3.1 Introduction Pull-out forming is a forming process based on pulling out of a given tubular workpiece to produce a ducted product for conveying gaseous and liquid fluids under high pressure with good accuracy, surface finish, better mechanical and metallurgical properties at a lower unit cost. It is formed by superplastic forming, by which other matching parts can be attached to produce a fluid-tight system. The aim of this forming process is to provide an improved tubular part having an integral pull-out formed part having an acceptable degree of thin-out at the rim of the pull-out to facilitate connection of ducts or other tubular members to the tubular in an assembly. S. Kumar (B) Department of Mechanical Engineering, IIT BHU, Varanasi 221005, India e-mail: [email protected]

© Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_3

29

30

S. Kumar

Sanders [1] proposed an apparatus for superplastic forming of tubing by pullouts, in which a pull-die having a cross section larger than the hole and about equal to the desired internal cross section of the tubular protrusion is used. There are about ten patents filed during 1931–1995 based on cold pull-out forming. Taylor [2], Keller [3] and Latham [4] filled patents in 1950, 1931 and 1970, respectively, on different materials using pulling methods of making pipe fittings using orthogonal arrangements, whereas Yoshida et al. [5] proposed a method of producing multiple walls by cold pulling method. Javorik [6] proposed a method by expanding a tubular member using pulling force. Nakamura [7] proposed a method for manufacturing a pipe by projection method. Krips et al. [8] proposed a method of securing parts using hollow member by pulling them. Bumgarner [9] proposed a method called as isostatic bulge forming. Schafer [10] proposed an apparatus for shaping a hollow body hydraulically. Rathke et al. [11] proposed a process of electromagnetically forming a tubular workpiece by applying an electromagnetic force provided by an energized work coil to the workpiece radially of its longitudinal axis and by simultaneously applying an axial compressive force to the workpiece. Fu et al. [12] given a method for evaluation of forming system design through CAE simulation and conducted the study on integrated simulation framework, design index and system design index. Dixit [13] analysed large plastic deformation using finite element method and formulated the problem using incremental updated Lagrangian method. DEFORM [14] is a finite element method (FEM) based simulation system designed for forming and heat process analysis used by metal forming and related industries. Kau [15] developed a remote quick CAE system. This system can be accessed through Word Wide Web, which can be used to quickly generate an optimized sculpture die surface and estimated the forming load on die. Gunesekera [16] analysed dies using CAE packages. He proposed CAE methods to analyze different types of dies. Mehar [17] developed a methodology for systematic design of metal forming (hot metal extrusion) via CAE simulation. He articulated the concept of large plastic deformation in metal forming system design. Mehar [18] proposed a quantitative design index and systematic approach for the evaluation of the designed dies based on the deformation load (on work and load acting on die). The efficiency and validity of the proposed CAE process technology have been verified using DEFORM-3D FEM simulation package. Schulz [19] proposed a methodology to make a tubular part. The method uses gas pressure to super plastically form a portion of a side wall of an end-sealed tube, heated to a superplastic temperature in a die, into a side pocket of the die to be formed by the pull-out method. Okada et al. [20] proposed an apparatus for manufacturing a T-joint, consists of a tube portion and a collar portion, by using burring techniques. This was also made using the pull-out forming method.

3 Pull-Out Forming: Experiments and Process Simulation

31

3.2 CAE Simulation Procedure To conduct a CAE simulation, it needs to represent the metal forming system in terms of geometry content, i.e., the deforming system, the models (deformation work) and tooling (tooling assembly). The CAD model should be in suitable formats (.stl). It also requires a physical model, mathematical model and numerical model. The physical model idealizes the real engineering problems and abstracts them to comply with certain physical theory with assumptions. The mathematical model specifies the mathematical equations such as the differential equations in FEM analysis the physical model. It also details the boundary and initial conditions and constraints. The numerical model describes the element types, mesh density and solution parameters. The solution parameters further provide detailed calculation tolerances, error bounds, iteration specifications and convergence criteria. After the simulation, the calculated results need to be analyzed and evaluated. If the results and solutions are not satisfactory, the suggested changes and modifications for the metal forming system in terms of part design, tooling design, process configuration and material selection can be made for next round simulation. The process is iterative until all the system design requirements and specifications are satisfied. Numerical Simulation Framework A numerical simulation of metal forming system, include, process and tooling deformation. It provides an approach for simultaneous modelling of deformation behaviour, metallurgical phenomena and thermal phenomena in a metal forming processes including interaction and interplay of physical behaviour. The systematic design of the metal forming system needs a computer-based engineering simulation approach, to simulate the forming process and tooling deformation simultaneously. Figure 3.1 depicts simulation and analysis framework for metal forming system design for pull-out forming including die structure and stress analysis. In the pullout forming process simulation, there are three main issues to be addressed, viz. metallurgical phenomena modelling, mechanical behaviour modelling and thermal phenomena modelling. In the simulations and analysis, dynamic physical and geometrical boundary conditions in between the simulation of workpiece deformation and die structure and stress analysis need to be established in each simulation step like physical boundary conditions, mechanical boundary conditions (BCs) and thermal BCs. An interpolation of the state variables (stress, strain, pressure or temperature) from one simulation is used to the others simulation to determine the boundary conditions and analysis scenario. In the deforming process simulation, on the other hand, the BCs information from die structure and deformation analysis are also needed to be interpolated for the establishment of the BCs in work deformation simulation. The optimization process optimizes the process condition, etc.

32

S. Kumar

Fig. 3.1 Simulation frame work [12, 21]

3.3 Design of Tubular Pull-Out Process Process design of pull-out forming consists of two main steps: (1) Design of tooling and (2) Process parameter design. Design of Toolings Design of tooling consists of the design of all the tolling set-up required for pull-out forming process. The design of tooling consists of the following steps: (a) Design of die: A die has been designed in which work (tube) part is to be inserted for pull-out process, die is openable and mainly consist of two part, upper part and lower part. It is designed in two parts, so that after forming process it can be removed. A circular hole, vertically upward symmetric to the axis, has been made in upper part of the die, this hole diameter depends on the thickness of the work part and the ball diameter say pulling ball. Hole diameter is taken as the addition of the ball diameter and two times of the thickness of the work part. Material of the die is taken harder than the material of the work for pull-out. (b) Design of ball: A pulling ball has been designed for this process. The diameter of the ball depends on the inner diameter of the pull-out product, diameter is

3 Pull-Out Forming: Experiments and Process Simulation

33

taken equal to the inner diameter of the pull-out product. The hardness of the ball material is taken more than that of work part, so that it can bear the load of pulling the work material. (c) Design of cut-out: This is the most important design part in pull-out process design, shape of the lip and height of pull-out depends on the cut-out size and shape. The shape of the cut-out is taken as ellipse, during the process of pulling when load comes to the work material applied by ball, then the minor axis of the ellipse will compresses as a result after pulling, the lip size of pull-out will be circular. Cut-out size will be smaller than that of ball diameter, for higher length of the pull-out, cut-out size should be as small as possible. The recommended ratio of major to minor of cut-out is 1.5. Load required for pulling will increase as the cut-out size will decrease. After the pull process lip of the pull-outs may have cracks, this is not because of the wrong design of cut-out, this is because of the hardness of the work material, it can be illuminated by heat treatment (annealing) of the work (in which cut-out is made) before going for the pull-out forming. (d) Design of rod: pulling rod is designed for this process. Rod diameter and the material used for rod is the main consideration for designing the pulling rod, diameter of the rod is taken smaller than the cut-out size made in the work and hardness of the rod material should be very high so that it can bear the pulling load. For example, if work material is ‘AL’, rod material should be taken EN-9 as the hardness of the EN-9 material is high. Process Parameter Design Process parameter design is related from the pulling velocity and value of coefficient of friction taken in pull-out forming process. Load required for pull-out forming should be as less possible and for the surface finishing of inner wall of the pullout, pulling velocity and coefficient of friction value should be small. The improper value of these two can leads the high load and affect the surface finish of inner wall of the pull-out. Pulling velocity and coefficient of friction is further taken as different scenarios in evaluation criterion.

3.3.1 Systematic Pull-Out Process Design and Evaluation Earlier, layman experience and hit-and-trials methods have been used for designing metal forming processes. But it needs a more scientific analysis and calculation to make the product cheaper. Pull-out forming system requires the design of the process, metal formed part, tooling and material selection as well as configuration. That is why a scientifically systematic methodology is needed for evaluating pull-out forming designs methodology [12].

34

S. Kumar

3.3.2 Evaluation Criteria for Pull-Out Process Design It covers the following six issues: (a) Deformation load for deforming body The deformation load is one of the key parameters for determining the forming process and tooling design and it should be as small as possible. Deformation load is determined by material properties, process configuration and tooling design. (b) Load acting on die During the process of forming, hot work comes directly in contact with the die. At high temperature, the die wall would be deformed, if it suffers high temperature that may cause changing of the die profile. So the load acting on the die should be as small as possible. This lays a key role in evaluating the pull-out process design requirement. (c) Temperature of the die Due to heat transfer between die and work, the die suffers somewhat more temperature than initials. Moreover, the energy consumed in overcoming the friction between die and work converts into heat. If sticking friction occurs then more amount of energy is needed to overcome the friction, so more heat energy would be generated. More heat leads to a more softening effect of the die and leads to more deviation from die profile. (d) Effective stress Resultant of all the force acting at a point is called effective stress. If effective stress acting at a point exceeds a limit, the component may fail. More is the range of effective stress acting on all point of a component more will be the variation in die profile. (e) Damage In the metal forming process, the deforming body is subjected to large plastic deformation and the material may reach a limit of ductility and any further deformation can lead to ductile fracture. The fracture could happen when the damage factor reaches a critical value. (f) Strain rate Material properties largely depend on the strain rate. More is the strain rate, more will be the strain hardening, and therefore high strain rate is undesirable.

3 Pull-Out Forming: Experiments and Process Simulation

35

3.3.3 Design Index To numerically represent the design evaluation criteria and to see how this contributes to the entire system design and evaluation, a design index (DI) is defined as: 

   Pi − Pmax + Pmin /2   DI = k Pmax − Pmin /2

(3.1)

DI is the design index, Pi is the output value of a specific design parameter for ith design scenario. Pmax is the maximum output value of the design parameter in all the design scenarios, Pmin is the minimum output value of the design parameter, k is a characteristic factor, which is equal to + (plus) when the bigger the output value, the better the designed system, while it is equal to − (minus) if the smaller the output value, the better the design system. Therefore, the characteristic factor is − (minus) for the deformation load output. Besides, DI varies between +1 and −1. For a given design parameter, it is the best design scenario when it has a value of +1 and the worst case when its value is −1.

3.3.4 System Evaluation Criterion The deformation load on the work body, damage factor and the effective stress of tooling are considered as the three criteria to judge the pull-out forming process. The tooling here refers to the main working components in tooling/die assembly. They are the work, the die insert, etc. To evaluate the metal forming system design, a system level evaluation criterion [12] is used on the basis of the above-defined three basic criteria as (a) design indexes of deformation load (DIDL ) (b) damage factor (DIDF ) and (c) effective stress (DIES ). The system design index (DIsystem ) is thus represented as:  (DIDL + DIDF + DIES j )   DIsystem = (3.2) 2 + Nj N j is the number of die components which the effective stress needs to be considered in the system design evaluation such as work, die, etc. Accordingly, the system design index may have a value between +1 and −1. When the value of DIsystem is +1, the designed system is the best solution for all the design scenarios, and it is the worst when it is −1.

36

S. Kumar

3.4 Implementation and Case Study Cold forming system design for the tubular shape part is conducted to illustrate how the developed methodology helps the design of a metal forming system. The main concern in this design is the metal formed part, pulling rod and die design. The geometry of the part and die assembly is shown in Figs. 3.2, 3.3, 3.4 and 3.5. For conducting this case study, three types of hole, elliptical, elliptical with different ratio of major to minor axis and loft had been created in the tubular-shaped part. To conduct the above case study, different pull-out hole profile [elliptical (Fig. 3.2a), elliptical with different ratio of major to minor axis (Fig. 3.2b), and loft (Fig. 3.2c)] is generated. The models are generated by the commercial AutoCAD system and converted into suitable data formats to run into the CAE simulation system. (a) Model creation All the models (die, workpiece/work, pull rod and pulling ball) used for CAE experiments are made in AutoCAD. After the creation of models, all are placed carefully in their respective positions. After the successful creation of the model, it needs to be transferred to the CAE package (DEFORM 3D) for analysis. The model is exported from AutoCAD through

Fig. 3.2 Dimension of hole in work used for case 2 (a), case 2 (b), case 2 (c)

3 Pull-Out Forming: Experiments and Process Simulation

Fig. 3.3 Generalized dimension of work and die used in analysis

(ALL DIMENSIONS ARE IN MM) Fig. 3.4 Dimensions of pulling (rod and ball) for pull-out forming

37

38

S. Kumar

Fig. 3.5 Final shape obtained, case 1 scenario(E)-double ball

stereolithography (*.stl) formats. Total surface is divided into small facets. Facets have three sides. In the file, the detail information about facets is present. In the pull-out forming process, workpiece and die to plays an important role. The total force required, maximum temperature generated, etc. Figures 3.3 and 3.4 show the dimension of the workpiece, die, pulling rod and pulling ball used for the analysis purpose. (b) Material model Pulling rod and pulling ball are considered as a rigid body while the work is assumed to be plastic. The work material is AL-2014(250C-480C). For work, the material properties used are: E = 68900 (MPa), γ = 0.33, thermal expansion = 23(10−6 /°C), thermal conductivity = 192 (W/mK), heat capacity = 2.433 and emissivity = 0.7. Die has been taken as a rigid body. (c) Friction model In pull-out forming process, the coulomb friction model has been adopted and these are represented as: f = 0.2, f = 0.6, f = 0.8 (d) Mesh model In this case work is discredited into 165000 and 150000 messes, respectively. With these models and inputs of simulation constraints and controls, the simulation for each design scenario is executed. The result obtained from the CAE simulations is tabulated in Tables 3.1, 3.2, 3.3, 3.4, 3.5 and 3.6 for all the cases so that evaluation criteria can be applied on these data.

−1

0.706

+0.048

+1

+0.016

−0.655

0 703

−0.237

316

0

−0.297

Damage factor

DIDF

Effective stress (MPa)

DIES

DIsystem

315

269

370

DIDL

−0.66

0

316

−0.99

0.789

−1

380

B B1

B1

B2

A

Scenarios

Deformation load (KN)

Parameter

Case 1 with double ball

+0.072

+1

315

−1

0.793

+0.217

255

B2

Table 3.1 CAE simulation result for case 1 with double ball

C

−0.367

0

316

−0.585

0.743

−0.517

366

B1

−0.176

0

316

−0.487

0.763

−0.043

258

B2

D

−0.597

0

316

−1

0.790

−0.793

374

B1

E**

0 −0.655

−1 −0.710

316

+1

0.563

+0.965

323

B1

317

−1

0.793

−0.130

259

B2

+1

+1

315

+1

0.676

+1

246

B2

F

+0.596

0

316

+0.788

0.587

+1

322

B1

+0.5866

0

316

+0.760

0.590

+1

246

B2

3 Pull-Out Forming: Experiments and Process Simulation 39

40

S. Kumar

Table 3.2 CAE simulation result for case 1 with single ball Case 1 with single ball Parameter

Scenarios A

Deformation load (KN)

E

C**

D

E

F

B1

B1

B1

B1

B1

B1

246

249

212

248

251

212

DIDL

−0.743

−0.8974

+1

−0.8461

−1

+1

Damage factor

0.625

0.551

0.518

0.577

0.586

0.527

DIDF

−1

+0.3831

+1

−0.1028

−0.2710

+0.8317

Effective stress (MPa)

317

318

318

318

318

319

DIES

+1

0

0

0

0

−1

DIsystem

−0.247

−0.171

+0.666

−0.316

−0.423

+0.277

For optimizing a pull-out forming process, effective strain, effective strain rate, temperature do not play an important role, therefore, these are excluded from the actual evaluation process. Damage is a deciding factor in case of cold forming. In cold forming effective strain, effective strain rate and temperature have no active role. Therefore, load on workpiece and pulling rod are main evaluating parameters.

3.4.1 Discussion and Analysis The maximum value of load, damage and effective stress obtained, based on the CAE simulation run by DEFORM 3D are tabulated in Tables 3.1, 3.2, 3.3, 3.4, 3.5 and 3.6 for design index (DI) calculation and for system evaluation. Design index for each case and corresponding system design indexes are calculated based on Eqs. (3.1) and (3.2) for different design scenarios are tabulated in Tables 3.1, 3.2, 3.3, 3.4, 3.5 and 3.6. The final part shape at which the part dimensions met is presented in Fig. 3.5 for case 1 scenario (E) with double ball and Fig. 3.6 for case 1 scenario (C) with single ball. From Table 3.1, case 1 (double ball), comparing scenarios A to F, scenario E has less value of deformation load, damage and effective stress from the rest of the scenarios. From the system perspective, it is found that the best design scenario is scenario E as the deformation load, damage and effective stress value for this scenario have the best combination and the system design index is also the best. From Table 3.3, case 2 (double ball), comparing scenarios A to F, scenario C has less value of deformation load and damage from the rest of the scenarios. However, the value of effective stress is less in scenario A. From the system perspective, the design index is shown in Table 3.3. Based on the distribution of design index and the above analysis, it is found that the best design scenario is scenario C.

−0.54

0.820

−0.08

+1

+0.126

−0.71

0.812

−0.62

316

0

−0.443

Damage factor

DIDF

Effective stress (MPa)

DIES

DIsystem

314

224

349

DIDL

−0.543

0

316

−0.74

0.831

−0.89

354

B B1

B1

B2

A

Scenarios

Deformation load (KN)

Parameter

Case 2 with double ball

Table 3.3 CAE simulation result for case 2 with double ball

−0.523

+0.33

315

−1

1.03

−0.90

228

B2

C**

+0.650

0

316

+0.952

0.575

+1

301

B1

D

0 −0.470

−0.33 +0.523

316

−0.66

0.818

−0.75

350

B1

316

+0.90

0.593

+1

207

B2

−0.35

−0.33

316

−0.11

0.826

−0.63

225

B2

E

−0.663

0

316

−0.99

0.869

−1

357

B1

−0.56

−0.33

316

−0.35

0.883

−1

229

B2

F

+0.615

0

316

+0.99

0.568

+0.857

305

B1

+0.30

−1

317

+1

0.571

+0.90

208

B2

3 Pull-Out Forming: Experiments and Process Simulation 41

42

S. Kumar

Table 3.4 CAE simulation result for case 2 with single ball Case 2 with single ball Parameter

Scenarios A

Deformation load (KN)

B

C**

D

E

F

B1

B1

B1

B1

B1

B1

217

222

191

218

222

192

DIDL

−0.6774

−1

+1

−0.7419

−1

+0.9354

Damage factor

0.394

0.379

0.260

0.381

0.379

0.258

DIDF

−1

−0.7794

+0.9705

−0.8088

−0.7794

+1

Effective stress (MPa)

318

318

318

319

318

318

DIES

+1

+1

+1

−1

+1

+1

DIsystem

−0.225

−0.259

+0.990

−0.850

−0.259

+0.978

From Table 3.5, case 3 (double ball), comparing scenarios A to F, scenario F has less value of deformation load, damage and effective stress from rest of the scenarios. From the system perspective, it is found that the best design scenario is scenario F. In case 1, with a single ball, comparing scenarios A to F, scenario C has less value of damage from the rest of the scenarios. However, the value of effective stress is less in scenario A and deformation load is the same in scenario C and F. From the system perspective, the design index is shown in Table 3.2. Based on the distribution of design index and the above analysis, it is found that the best design scenario is scenario C. From Table 3.4, case 2 (single ball), comparing scenarios A to F, scenario C has less value of deformation load, from the rest of the scenarios. However, the value of damage is less in scenario F and the value of effective stress is the same in all scenarios. From the system perspective, the design index is shown in Table 3.4. Based on the distribution of design index and the above analysis, it is found that the best design scenario is scenario C. From Table 3.6, case 3 (single ball), comparing scenarios A to F, scenario C has less value of deformation load and damage, from rest of the scenarios. However, the value of effective stress is less in scenario A and B. From the system perspective, the design index is shown in Table 3.6. Based on the distribution of design index and the above analysis, it is found that the best design scenario is scenario C. Now it has been decided which scenario is the best scenario in all the cases depending on the system design index for a single ball as well as the double ball. Here, our aim to find which scenario of a particular case has the best result in pullout the shape. On analysing the final shape of all the scenarios in all the three cases, scenario E of case 1 has the best shape in case of the double ball and scenario C of case 1 has the best shape in case of single ball pull-out process. In case 1, cut-out size is the smallest in all three cases. It has been seen that if the cut-out size is less, the pull-out size will be more. In case of the double ball, scenario E of case 1 and In case of single ball scenario C of

241

−0.89

0.788

+0.66

314

−0.98

−0.403

348

−0.95

0.737

−0.50

316

−0.99

−0.087

DIDL

Damage factor

DIDF

Effective stress (MPa)

DIES

DIsystem

−0.963

−0.99

316

−0.91

0.802

−0.99

354

B B1

B1

B2

A

Scenarios

Deformation load (KN)

Parameter

Case 3 with double ball

Table 3.5 CAE simulation result for case 3 with double ball

−0.736

−0.99

315

−0.25

0.916

−0.97

250

B2

C

−0.526

−0.99

316

+0.12

0.637

−0.71

309

B1

−0.246

−0.99

315

+1

0.742

−0.75

224

B2

D

−0.83

−0.99

316

−0.54

0.743

−0.96

350

B1

−0.68

−1

316

−0.16

0.904

−0.89

241

B2

E

−1

−1

317

−1

0.815

−1

355

B1

−0.75

−1

316

−0.26

0.918

−1

253

B2

F**

+1

+1

45.7

+1

0.498

+1

31.2

B1

+0.333

+1

45.2

−1

1.02

+1

13.7

B2

3 Pull-Out Forming: Experiments and Process Simulation 43

44

S. Kumar

Table 3.6 CAE simulation result for case 3 with single ball Case 3 with single ball Parameter

Scenarios A

Load(KN)

B

C**

D

E

F

B1

B1

B1

B1

B1

B1

226

230

195

229

234

198

DIDL

−0.5897

−0.7948

+1

−0.7435

−1

+0.846

Damage factor

0.629

0.696

0.313

0.615

0.671

0.286

DIDF

−0.6731

−1

+0.8682

−0.6048

-0.878

+1

Effective Stress(MPa)

316

316

319

318

317

319

DIES

+1

+1

−1

−0.333

+0.333

−1

DIsystem

−0.087

−0.264

+0.289

−0.560

−0.515

+0.282

Fig. 3.6 Final shape obtained, case 1 scenario(C)single ball

case 1 has the best shape, because sidewall thickness of the pull-out is uniform and thin-out around the lip is very less from rest of the cases. From the analysis of the final shape of the pull-out, it is also found that the best shape can be found at the lowest pulling velocity and at the minimum coefficient of friction. From the design perspective, based on the design index, the best design is case 1 as the final shape of pull-out and design index value for the case is the best.

3 Pull-Out Forming: Experiments and Process Simulation

45

3.5 Experimental Verification Pull-out forming of commercial grade aluminium (2014) is carried out to find out the load required at cold condition at the similar (pulling speed, temperature, etc.) conditions. • Specimen: Commercial grade aluminium (2014). Work length = 73 mm, outer diameter = 31 mm, inner diameter varies as 27, 25, 23, 21 mm. • Lubricant: Grease mixed with graphite powder (20%) is used in the present work for pull-out forming of commercial grade aluminium (2014). • Velocity: Pulling velocity is less than 0.5 mm/s, because machine is operated manually for the pull-out forming experiment. • Temperature: Experiment is done at room temperature, so the whole set-up of experiment was at room temperature. Pull-out forming has been conducted on a 1000 kN vertical milling machine as shown in Fig. 3.7. The pulling tool is installed with a calibrated load-cell. For each pull-out forming run, the die, bolt, table, ram, tool head and the specimen are first cleaned properly. The die, pulling rod and pulling ball are lubricated with grease mixed with graphite powder (20%) for aluminium workpiece. The die is carefully placed in the recess of the die holder, to ensure that the die is symmetrically placed with respect to the recess. Pull-out forming is conducted at an average table feed speed since table speed (vertically up and down) is manually controlled till the desired

Fig. 3.7 Photographs of the pull-out forming: various stages used in experiment

46

S. Kumar

Table 3.7 Load obtained: Pull-out CAE simulation and experimental conditions for different samples Sample No.

Sample thickness (in mm) (annealedAN/unannealed-UA)

Cut-out shape in billet (sample)

Simulation load (in KN)

Experimental load (in KN)

Remarks

1

2 (UA)

Elliptical

5.91

6.01

4

3 (A)

Elliptical

24.10

24.76

Good

5

4 (A)

Elliptical

31.70

32.10

Good

6

5 (A)

Elliptical

42.10

42.45

Good

7

4 (A)

Circular

40.30

40.79

Good

Cracked

Fig. 3.8 Pull-out product at the end of experiment (Sample-1)

length of pull-out is obtained. Load versus displacement readings were found from the simulation of a similar condition are shown below. Several experiments have been conducted with samples having different-different thicknesses. Table 3.7 shows experimental set-up photographs in the pull-out process for different samples under similar process conditions. Figure 3.8, 3.9, 3.10, 3.11, 3.12, 3.13, 3.14, 3.15, 3.16, 3.17, 3.18 and 3.19 shows the workpiece after pull-out and experimental (load cell)/simulation (DEFORM) reading of load required under (pull speed = 0.5 mm/s, workpiece temp. 200 °C) under similar experimental-CAE simulation conditions. Figure 3.20 shows the initial workpiece and pull-out CAE result at end of 80th run under pull speed = 0.5 mm/s, workpiece temp. 200 °C. In sample-1 (without annealed), a crack on the lip of the pull-out was observed, therefore, after testing sample-1, other samples are heat treated (annealing) at 4000 °C for an hour and then cooled. It can be observed that samples tested after heat treatments do not have any cracks on the lip of the pull-outs.

3 Pull-Out Forming: Experiments and Process Simulation

47

Fig. 3.9 Pull-out load versus stroke (CAE simulation—similar experimental condition (Sample-1)

Fig. 3.10 Pull-out product at the end of experiment (Sample-2)

Fig. 3.11 Pull-out product at the end of experiment (Sample-3)

3.6 Conclusions The work is based on experiments and simulation to come-out with a design criterion in pull-out forming. Pull-out forming can be done by making circular cut-out also, but in this case, the load required is more. On the basis of this, and pull-out forming process can be evaluated within a short time and can be used to design the die, pulling

48

S. Kumar

Fig. 3.12 Pull-out product at the end of experiment (Sample-4)

Fig. 3.13 Pull-out load versus stroke (CAE simulation and experimental condition for (Sample-4)

Fig. 3.14 Pull-out product at the end of experiment (Sample-5)

rod and pulling ball, etc. It is, therefore, very clear that experimental and theoretical result (CAE simulation) agrees closely with each other under annealed workpieces. It is also seen that more is the cut-out size the less pull-out length is obtained in forming. For better pull-out, a shape and cut-out size should be smaller and in proper ratio (major to minor axis ratio) of the cut-outs. It is found that pulling load is less in the case of elliptical cut-outs. Since the pull-out length of the shape is optimized by experiments and CAE simulation, this indicates that pull-out length of any tubular

3 Pull-Out Forming: Experiments and Process Simulation

49

Fig. 3.15 Pull-out load versus stroke (CAE simulation and experimental condition for (Sample-5)

Fig. 3.16 Pull-out product at the end of experiment (Sample-6)

Fig. 3.17 Pull-out load versus stroke (CAE simulation and experimental condition for (Sample-6)

workpiece can be optimized. It is also concluded that the pull-out process can be used for other shape outlets also.

50

S. Kumar

Fig. 3.18 Pull-out product at the end of experiment (Sample-7)

Fig. 3.19 Pull-out load versus stroke (CAE simulation and experimental condition for (Sample-7) Fig. 3.20 a initial workpiece, b pull-out CAE (end of 80th run) (pull speed = 0.5 mm/s, workpiece temp. 20 °C)

3 Pull-Out Forming: Experiments and Process Simulation

51

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

13. 14. 15. 16. 17. 18. 19. 20. 21.

Sanders, D.G: Superplastic forming of tubing pull-outs. U. S. Patent No: 6430812 (1998) Taylor, J.H.: Method for making pipe T’S. U. S. Patent No: 1911653 (1931) Keller, W.F.: Method of making pipe fittings. U. S. Patent No: 2507859 (1950) Latham, R.J.: Method of making stainless steel and like tubes and fittings with branches. U. S. Patent No: 3535909 (1970) Yoshida, T., Matsui, S., Atsuta, T.: Method of producing multiple wall. U. S. Patent No: 4449281 (1984) Javorik, L.: Method for expanding a Tubular member. U. S. Patent No: 4590655 (1986) Nakamura, M.: Method for manufacturing A pipe with projections. U. S. Patent No: 4840053 (1989) Krips, H., Podhorsky, M.: Method of securing parts to a hollow member. U. S. Patent No: 4875270 (1989) Bumgarner, J.R.: Method of isostatic bulge forming. U. S. Patent No: 5419171 (1995) Schafer, A.W.: Apparatus for hydraulically shaping a hollow body. U. S. Patent No: 5435163 (1995) Rathke, J., Burger, E.V., Peterson, E.M., Horan, C.J.: Electromagnetically forming a tubular work piece. U. S. Patent No: 5826320 (1998) Fu, M.W., Yong, M.S., Tong, K.K., Muramatsu, T.: A methodology for evaluation of metal forming system design and performance via CAE simulation. Int. J. Prod. Res. 44(6), 1075– 1092 (2006) Dixit, P.M.: Finite element formulation for large deformation elasto-plastic problems (3-D). QIP Sponsored Course, IITK, January 7–12 (2002) DEFORM 3D manual: Analysis and design tool for forming. Scientific-forming Technologies Corporation (2003) Kao, Y.C.: Development of a remote quick CAE system on sculptured metal extrusion die surface. J. Mater. Process. Technol. 140, 116–128 (2003) Gunesekera, J.S.: CAD/CAM of dies. Ellis Horwood Limited Mehar, K.K.: CAE simulation of extrusion process and its experimental verification. J. Model. Simul. Des. Manuf. 1, 01–10 (2010) Mehar, K.K.: Computer aided engineering & optimization of extrusion process using simulation. M. Tech. Thesis, Institute of Technology, B.H.U, Varanasi, India (2005) Schulz, D.W.: Tool for sealing superplastic tube. U. S. Patent No: 5649439 (1997) Okada, K., Asao, H., Yonemura, H.: T-joint manufacturing apparatus. U. S. Patent No: 4676088 (1985) Forouhandeh, F., Kumar, S., Ojha, S.N., Prakash, T.O.: Modeling of sheet hydroforming of CP Titanium for semispherical cup shape product. J. Model. Simul. Des. Manuf. 3(1–2), 147–156 (2012)

Chapter 4

Failure Prediction and Forming Behavior of AA5754 Sheets at Warm Temperature Sudhy S. Panicker , Kaushik Bandyopadhyay and Sushanta Kumar Panda

Abstract The prediction of the warm forming behavior of aluminum alloy sheets is vital in the automotive industry to conceive and improvise the tool design to fabricate intricate thin-walled lightweight components without premature thinning or splitting. In this work, Marciniak–Kuczynski-based-forming limit diagram (MK-FLD) of AA5754-H22 sheet at 30 and 200 °C was successfully developed incorporating Hill48 anisotropic yield theory and Swift strain rate sensitive hardening law. Drastic improvement of FLD was noticed at 200 °C as the FLD0 was 1.35 times higher compared to that at 30 °C. The developed FE models accurately captured the fracture location, maximum thinning, and strain localization region. The suitability of MK-FLD and FE simulations was demonstrated using the experimental data available from the literature. The strain path followed by the failure node during the deformation of Hasek, biaxial, and deep drawing samples was observed to follow tension–compression, tension–tension, and plane strain modes, respectively. Keywords Warm forming · AA5754 · Forming limit diagram · FE modeling

4.1 Introduction The severity in emission rules and enduring demand for improved fuel economy compelled the automotive industries to manufacture aluminum intensified lightweight auto-body components [1]. Aluminum alloys emerged as an alternative for generally used low-carbon steel as they possess high specific strength, excellent corrosion resistance, and good recyclability. However, the low-formability of aluminum sheets still

S. S. Panicker · S. K. Panda (B) Department of Mechanical Engineering, Indian Institute of Technology Kharagpur, Kharagpur 721302, India e-mail: [email protected] K. Bandyopadhyay Department of Materials Science and Engineering, Korea University, Anam-Dong, Seongbuk-gu, Seoul 136-701, South Korea © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_4

53

54

S. S. Panicker et al.

acts as a hindrance for their extensive application. Warm forming is a promising technology that significantly enhances the formability of aluminum alloy sheets. Li and Ghosh [2] described that the drastic improvement in ductility of AA5754, AA5182, and AA6111 sheets with an improved post-uniform elongation at 200–250 °C was the result of the exponential rise in strain rate sensitivity. Dramatic enhancement of limiting strains of the alloys mentioned above was observed at 250 °C and found to be comparable with the forming limit diagram (FLD) of low-carbon steel sheets [3]. Panicker et al. [4] demonstrated a drastic reduction of drawing force during deep drawing of AA6082-T6, AA5052-H32, and AA2014-T6 sheets at 350 °C. Numerical modeling plays a vital role in conceiving new tool design and reduces the number of trial experiments which boosts economic benefits. The accuracy in numerical modeling of sheet metal deformation process is ensured by incorporating proper yielding criterion, strain rate sensitivity, and forming limit strains. Therefore, there is a great interest in developing analytical approaches to analyze the strain localization behavior and to predict the FLD. Abedrabbo et al. [5], while developing a coupled thermomechanical numerical model, used Marciniak–Kuczynski (MK) model integrating anisotropic Barlat2000-2d yield theory and Voce strain rate-dependent hardening function to accurately predict the forming behavior of AA5754-O and AA5182-O sheets at temperatures ranging from 25 to 260 °C. More recently, Bandyopadhyay et al. [6] used the MK-based analytical FLD as the failure criteria to predict formability of tailor welded steel blanks. It was observed that no open literature provided the analytical prediction of failure and deformation behavior of AA5754-H22 sheets at 200 °C using MK-FLD based on Hill48 anisotropy yield theory and coupled with strain rate sensitivity. The primary aim of the present work is to predict the deformation behavior of 1.0-mm-thickautomotive grade AA5754-H22 sheet under both stretch forming and deep drawing modes at a warm isothermal temperature of 200 °C. Also, the FE models of both the warm forming processes were developed using commercially available FE code LS-DYNA-971 incorporating MK damage model based on Hill48 yield theory and Swift strain rate sensitive hardening law. The warm forming behavior of the sheet metal was successfully assessed in terms of deformed part depth, failure locations, thickness, and strain distributions.

4.2 Finite Element Model Figure 4.1a, b, respectively, illustrates the quarter-symmetric finite element (FE) model developed to perform the present stretch forming and deep drawing simulations. The hemispherical punch of φ 50 mm was used for stretch forming, whereas a flat-bottomed cylindrical punch of φ 54.8 mm was used for deep drawing process. For the stretch forming model, an analytical drawbead was designed and locked to the binder at a radial distance of 36.0 mm. The corner radius of the deep drawing die and flat-bottomed punch was 8.0 mm. Drawbead was absent in the deep drawing setup, and the sheet metal was allowed to draw into the die cavity under the constant

4 Failure Prediction and Forming Behavior of AA5754 …

55

Fig. 4.1 FE models developed for warm forming simulations along with the blank dimensions used for: a stretch forming and b deep drawing processes

blank holding force of 1.5 kN. All the tooling was assumed as rigid bodies. The FE models of both these processes were developed according to the warm forming setup dimensions as detailed in Panicker et al. [7]. Two blank geometries used for stretch forming simulations were: (i) Hasek specimen with 20 mm width at the middle (40 mm recess radius trimmed out of 100 mm circular blank) and (ii) 100 mm × 100 mm square blank (designated as biaxial specimen). The circular blank diameter selected for warm deep drawing was 110 mm. The blank was discretized with 1 mm × 1 mm mesh size assigning four-node Belytschko–Lin–Tsay shell elements. The requisite blank properties of the AA5754 material was derived from the uniaxial tensile testing data. Also, the symmetric boundary conditions were applied in the YZ and XZ nodes, and coefficient of friction assumed at the die–blank interface was 0.23.

4.2.1 Hill48 Anisotropy Yield Theory The quadratic Hill48 anisotropic model for plane stress condition is shown in Eq. (4.1). Hill48 is one of the extensively used yield criteria in both research and industrial applications due to the simplicity in deriving yield coefficients and its direct physical significance. The yield coefficients F, G, H, and N could be calculated through direct relationships with Lankford anisotropy parameters (R values) as given in Eq. (4.2). As the sheet metal deformation was considered to be under plane stress condition, the knowledge of yield strength (σ0 , σ45 , σ90 ) and R0 , R45 , and R90 were sufficient to determine the yield locus [6].

56

S. S. Panicker et al.

Fig. 4.2 a True stress–strain response and b Hill48 yield locus at 30 and 200 °C 2 2 = H (σx x − σ yy )2 + Gσx2x + Fσ yy + 2N σx2y − 1 = 0

(4.1)

H 1 N H = R0 ; − = R45 ; and = R90 G F+G 2 F

(4.2)

The R values and flow stress responses of AA5754 material at 30 and 200 °C were obtained by carrying out uniaxial tensile tests using the sample prepared according to the geometry mentioned in the inset of Fig. 4.2a [8]. The standard procedures to evaluate R values at elevated temperatures are detailed elsewhere [7, 9]. The flow stress was assumed to follow Swift hardening law incorporating the strain rate sensitivity of the material, described as σ = K (ε + ε0 )n ε˙ m . The parameters K, n, and m, respectively, represent the strength coefficient, strain hardening coefficient, and strain rate sensitivity, and these values evaluated at 30 and 200 °C are provided in Table 4.1. The R values were found to increase at 200 °C. This increment was in agreement with those reported in previous literature [5, 9]. Globally, an increase of R value represents improved drawability due to enhanced resistance to thinning. The proper incorporation of strain rate sensitivity was crucial as an exponential increment of m value was observed at 200 °C [10]. Figure 4.2a shows the true stress–true strain plot of the AA5754 material evaluated at 30 and 200 °C under the initial strain rate of 0.001 s−1 . The compatibility of flow stress predicted using Swift strain rate sensitive hardening law is also shown in the figure. The Hill48 anisotropy yield locus plotted at 30 and 200 °C are depicted in Fig. 4.2b. The yield locus at 200 °C was found to shrink due to the reduction of uniaxial yield strength.

K (MPa)

378.60

297.11

Temp. (°C)

30

200

0.146

0.297

n 0.054

0.007

m 1.128

0.709

R0 1.003

0.556

R45 1.171

0.798

R90 0.453

0.519

F

0.469

0.585

G

Table 4.1 Hardening coefficients, Lankford anisotropy parameters and Hill48 coefficients of AA5754 sheets at 30 and 200 °C H 0.530

0.415

N 1.386

0.167

4 Failure Prediction and Forming Behavior of AA5754 … 57

58

S. S. Panicker et al.

4.2.2 Failure Criteria In the present study, the FLD evaluated through MK model (referred as MK-FLD) based on Hill48 yield theory and Swift hardening law was utilized as the failure criteria. This analytical model assumes a local zone of thickness imperfection on infinite sheet metal in order to simulate the preexistent defects. As shown in Fig. 4.3, the inhomogeneity in thickness was modeled as a groove with a slightly lower thickness compared to the nominal thickness of the sheet. The geometric imperfection factor was defined as: f =

tb = f 0 exp(ε3b − ε3a ) ta

(4.3)

where the initial imperfection factor, f 0 = t0b /t0a . Now, the plastic strain increment corresponding to each strain path is assigned to the homogeneous region. The strain increment in the groove region was solved using force equilibrium and strain consistency condition. Localized necking was considered to happen when the non-uniform plastic flow develops and confines so that the ratio between effective plastic strain increment in the imperfection zone and the nominal sheet thickness goes beyond a critical value [11]. The strain developed in the homogenous section was considered as the limiting strain of the material. The initial imperfection factor, f 0 , was assumed as 0.998. The experimental limiting strains were evaluated by performing limiting dome height (LDH) tests using blanks of varying width geometry. The same stretch forming setup as mentioned in Sect. 4.2 was used to carry out LDH tests. The hemispherical punch stretches the AA5754 sheet and tests were stopped at the onset of visible necking on the deformed sample. The circular grid analysis (CGA) was performed to obtain the necking strains. Further, these necking strains were plotted on true major vs. minor strain space to obtain the limiting strain curve. The detailed procedures of LDH tests and CGA to evaluate the experimental limiting strains are discussed in Panicker et al. [7]. Fig. 4.3 Geometrical representation of inhomogeneity assumed in MK theory

4 Failure Prediction and Forming Behavior of AA5754 …

59

4.3 Results and Discussion 4.3.1 Failure Strains and Evolution of Strain Path The MK-FLD of AA5754 sheet developed at 30 and 200 °C are shown in Fig. 4.4. The limiting strains were found to improve drastically at 200 °C. As a result, the FLD0 (where the FLD intersected plane strain path) at 200 °C was approximately 1.35 times higher compared to that at 30 °C. The developed MK-FLD at both these temperatures was closely matching with the experimentally obtained forming limit strains. Figure 4.5 demonstrates that fracture on the component happened when the strain state surpassed the analytical MK-FLD. The FE predicted failure region on the component appeared in red in the post-processing window. The strain paths followed by each sample during the warm forming processes were assessed from the FE simulations. In order to plot the strain path of each deformed specimen, a failure node was identified on the respective blanks, namely, N1 for Hasek, N2 for biaxial and N3 for deep drawing blank. The strain evolution of the identified nodes plotted on the major vs. minor strain space is also depicted in Fig. 4.5. It was observed that the tool and blank geometry showed predominant influence on the evolution of the strain path. During LDH tests, the material flow was restricted by the drawbead and binder combination. Even though hemispherical punch was used for both Hasek and biaxial specimen, their deformation mode was different due to the difference in the specimen geometry. The strain path followed by lower width Hasek specimen was along the tension–compression regime, whereas biaxial specimen followed tension–tension mode. However, plane strain mode prevailed

Fig. 4.4 MK-FLDs evaluated at 30 and 200 °C

60

S. S. Panicker et al.

Fig. 4.5 MK-FLD and failure strain paths during the stretch forming and deep drawing processes at 200 °C

during warm deep drawing, which was attributed to the flat-bottomed shape of the punch. Here, instead of stretching, the material was allowed to flow into the die cavity, and the sheet metal underwent bending and subsequent unbending.

4.3.2 Validation with Experimental Results Part Depth. The experimental and FE predicted part depths obtained after stretch forming of Hasek and biaxial specimens, and deep drawn blank of AA5754 at 30 and 200 °C are given in Table 4.2. The stretch forming at 200 °C resulted in 26.83 and 29.04% improvement in part depth for Hasek and biaxial specimens, respectively, when compared to that at 30 °C. This significant improvement of part depth was attributed to the drastic increase in the formability limits of AA5754 at 200 °C. However, the improvement in part depth was only 4.10% during warm deep drawing. Figure 4.6 demonstrates the experimental and FE predicted part depths of warm Table 4.2 Comparison of experimental and FE predicted part depths obtained for different samples deformed at 30 and 200 °C Specimen geometry

Part depth at 30 °C

Part depth at 200 °C

Predicted

Experiment [7]

Predicted

Experiment [7]

Hasek

14.93

13.79

20.70

17.49

Biaxial

15.95

14.77

21.89

19.06

Deep drawing

13.61

13.17

14.04

13.71

4 Failure Prediction and Forming Behavior of AA5754 …

61

Fig. 4.6 Experimental and FE predicted part depth obtained from a Hasek, b biaxial, and c deep drawing blank at 200 °C

formed components at the initiation of necking/failure. It should be noted that the experimental part depth data were from Panicker et al. [7]. The predicted part depths at 200 °C were within the relative error range of 18.35, 14.84, and 2.40% for Hasek, biaxial, and deep drawing cases, respectively. However, the failure regions were accurately predicted. As per the geometry of the tooling and blank, the shape and size of the final component also varied. The dome-shaped component was formed during stretch forming while flat-bottomed hat-shaped component was obtained during deep drawing. The failure of the component occurred at the tangential location of the separation of punch and sheet metal. The materials at this juncture were prone to withstand the punch load, and the moment when the punch load exceeded the load-bearing capacity, localized necking or fracture was initiated. Thickness Distribution. Figure 4.7 provides the comparative demonstration of experimental and FE predicted thickness distributions for all the three deformation modes at 200 °C. The experimental thickness data of isothermal warm deep drawing of AA5754 was taken from Panicker and Panda [12]. The developed FE model accurately captured the thickness distribution and maximum thinning location. The position of maximum thinning was identified as the region of rupture. It was observed that the thinning of the samples was gradual during stretch forming. The minimum thickness location of biaxial specimen was approximately at 16.0 mm curvilinear distance from the pole (Fig. 4.7a), whereas it was at about 12.0 mm for the Hasek specimen (Fig. 4.7b). This deviation was attributed to the differences in specimen geometry. The biaxial sample was able to withstand more deformation due to the increased area of contact of the sheet with the hemispherical punch. The lower contact area of Hasek specimen caused relatively lower contact friction and reduced the load-bearing capacity. Hence, the location of maximum thinning was moved inward for Hasek specimen. However, as observed from Fig. 4.7c, the warm deep

62

S. S. Panicker et al.

Fig. 4.7 Thickness distributions of a biaxial, b Hasek, and c deep drawn components deformed at 200 °C

drawing resulted in sudden thinning at approximately 26.0 mm from the center of the cup, i.e. at the cup corner region. Strain Distribution Profile. The encouraging ductility possessed by the material at 200 °C was mainly due to improved strain rate sensitivity (Table 4.1), which enhanced the redistribution of strains to delay the strain localization. It was observed from Fig. 4.8a, b that the strain distribution profile was well-developed and evenly distributed during the elevated temperature stretch forming of both biaxial and Hasek specimens. The positive values of both major and minor strains indicated tension– tension mode for the biaxial specimen. The unsupported region of Hasek specimen underwent lateral draw-in, and hence the major strains evolved were positive and minor strains were negative. Therefore, the deformation of this lower width specimen was along the tension–compression region. An abrupt rise in the major strains in Fig. 4.8c indicated sudden strain localization occurred during deep drawing. Analogous to the thickness distribution profile, this failure location was identified as the location of peak major strain. FE predicted failure locations were also identical to that observed in the experiment. The experimental strain distribution data of Hasek and biaxial samples were obtained from Panicker et al. [7], and that of warm deep drawing conditions were from Panicker and Panda [12].

4 Failure Prediction and Forming Behavior of AA5754 …

63

Fig. 4.8 Strain distribution profiles of a biaxial, b Hasek, and c deep drawn components deformed at 200 °C

The location of sudden necking/failure during isothermal warm deep drawing was consistently located at the corner of the hat-shaped component [12]. The major strain at this region was approximately 4.9 times higher compared those of the adjacent nodes. The materials at punch corner region underwent bending and unbending under predominant radial tensile load. Even though an increase of R values and strain rate sensitivity aided to a minor improvement in deep drawing part depth, the comparable strength state (see Fig. 4.2b) and thermal softening initiated early fracture at the punch corner region. Hence, the rapid evolution of thinning and strain localization was observed along the plane strain condition. Therefore, it was concluded that significant improvement of formability could not be expected during warm deep drawing by maintaining all the tooling at the same elevated temperature conditions.

64

S. S. Panicker et al.

4.4 Conclusions In the present work, analytical MK-based FLD was successfully developed incorporating Hill48 anisotropy yield theory and Swift strain rate sensitive hardening law. The developed MK-FLD was utilized to predict the formability limits and deformation behavior of AA5754-H22 sheets at 200 °C. The influence of specimen and tooling geometry on the warm forming behavior has been discussed, and the following major conclusions were derived. • Drastic improvement in formability limits of the AA5754 sheet material was observed at 200 °C. The FLD0 at 200 °C was approximately 1.35 times higher than that at 30 °C. The improvement in part depths of Hasek, biaxial, and deep drawing specimens at elevated temperature were 26.83, 29.04, and 4.10%, respectively. • The deformation using hemispherical punch showed uniform strain distribution with delayed strain localization, whereas sudden necking was observed at the corner of the warm deep drawn component. • The failure strain paths of lower width Hasek specimen was along the tension– compression region, the biaxial specimen was along the tension–tension region, and the deep drawing blank was along the plane strain condition. Also, the FE simulations accurately captured the maximum thinning and strain localization zones. • The simultaneous impact of thermal softening and improved strain redistribution property of the material at 200 °C resulted in higher formability of stretch formed components. However, the improvement in part depth was very negligible during isothermal warm deep drawing of the AA5754 sheet.

References 1. Mallick, P.K. (ed.): Materials, Design and Manufacturing for Lightweight Vehicles. Elsevier (2010) 2. Li, D., Ghosh, A.: Tensile deformation behavior of aluminum alloys at warm forming temperatures. Mater. Sci. Eng., A 352(1–2), 279–286 (2003). https://doi.org/10.1016/S09215093(02)00915-2 3. Li, D., Ghosh, A.K.: Biaxial warm forming behavior of aluminum sheet alloys. J. Mater. Process. Technol. 145, 281–293 (2002). https://doi.org/10.1016/j.jmatprotec.2003.07.003 4. Panicker, S.S., Prasad, K.S., Basak, S., Panda, S.K.: Constitutive behavior and deep drawability of three aluminum alloys under different temperatures and deformation speeds. J. Mater. Eng. Perform. 26(8), 3954–3969 (2017). https://doi.org/10.1007/s11665-017-2837-x 5. Abedrabbo, N., Pourboghrat, F., Carsley, J.: Forming of AA5182-O and AA5754-O at elevated temperatures using coupled thermo-mechanical finite element models. Int. J. Plast 23(5), 841– 875 (2007). https://doi.org/10.1016/j.ijplas.2006.10.005 6. Bandyopadhyay, K., Lee, M.-G., Panda, S.K., Saha, P., Lee, J.: Formability assessment and failure prediction of laser welded dual phase steel blanks using anisotropic plastic properties. Int. J. Mech. Sci. 126, 203–221 (2016). https://doi.org/10.1016/j.ijmecsci.2017.03.022

4 Failure Prediction and Forming Behavior of AA5754 …

65

7. Panicker, S.S., Singh, H.G., Panda, S.K., Dashwood, R.: Characterization of tensile properties, limiting strains, and deep drawing behavior of AA5754-H22 sheet at elevated temperature. J. Mater. Eng. Perform. 24(11), 4267–4282 (2015). https://doi.org/10.1007/s11665-015-1740-6 8. Dhara, S., Basak, S., Panda, S.K., Hazra, S., Shollock, B., Dashwood, R.: Formability analysis of pre-strained AA5754-O sheet metal using Yld96 plasticity theory: Role of amount and direction of uni-axial pre-strain. J. Manuf. Processes 24, 270–282 (2016). https://doi.org/ 10.1016/j.jmapro.2016.09.014 9. Abedrabbo, N., Pourboghrat, F., Carsley, J.: Forming of aluminum alloys at elevated temperatures—Part 1: Material characterization. Int. J. Plast. 22(2), 314–341 (2006). https://doi.org/ 10.1016/j.ijplas.2005.03.005 10. Panicker, S.S., Panda, S.K.: Formability analysis of AA5754 alloy at warm condition: appraisal of strain rate sensitive index. Mater. Today Proc. 2(4–5), 1996–2004 (2015). https://doi.org/ 10.1016/j.matpr.2015.07.169 11. Dasappa, P., Inal, K., Mishra, R.: The effects of anisotropic yield functions and their material parameters on prediction of forming limit diagrams. Int. J. Solids Struct. 49(25), 3528–3550 (2012). https://doi.org/10.1016/j.ijsolstr.2012.04.021 12. Panicker, S.S., Panda, S.K.: Improvement in material flow during nonisothermal warm deep drawing of nonheat treatable aluminum alloy sheets. J. Manuf. Sci. Eng. 139(3), 031013 (2016). https://doi.org/10.1115/1.4034594

Chapter 5

Magnetic Pulse Forming and Punching of Al Tubes—A Novel Technique for Forming and Perforation of Tubes Sagar Pawar , Sachin D. Kore

and Arup Nandy

Abstract Magnetic pulse forming is a high-speed forming process in which electromagnetic force is used to deform the workpieces. The magnetic pulse forming and punching are a method of punching holes in the tube. The force required for forming and punching is due to electromagnetic interaction between coil and workpiece. In this work, simultaneous forming and punching have been studied experimentally. For this study, two different punches having different geometry, die which is split into four different parts and coil are designed and manufactured. The detailed study of the effect of punch geometry on the punching has been carried out. The result shows that the concave punches are more suitable for the punching as complete slug separation is possible as well as the punched hole diameter is also close to the required diameter of the hole. Keywords Electromagnetic forming · Perforation · Magnetic pulse forming · Punching

5.1 Introduction The electromagnetic or magnetic pulse forming is a high-strain forming technique. It uses magnetic fields to create Lorentz forces to deform the workpiece. In electromagnetic forming technology, workpieces are preferably of a high conductive material. The electromagnetic forming process occurs within microseconds, and because of the large forces, the workpiece undergoes high acceleration. The magnetic pulse forming S. Pawar (B) · A. Nandy Department of Mechanical Engineering, Indian Institute of Technology Guwahati, Guwahati 781039, India e-mail: [email protected] A. Nandy e-mail: [email protected] S. D. Kore School of Mechanical Sciences, Indian Institute of Technology Goa, Ponda, Goa 403401, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_5

67

68

S. Pawar et al.

and punching also called as electromagnetic forming and perforation (EMFP) which combines the forming and perforation operations. Perforation is a common practice for metal across many industries. Perforated tubes are not only used in automobiles but also in the machinery where the exhaust system is present. Perforated tubes are used in safety guards, liquid filtering agricultural strainers, filter elements and oil rig filtration. The tubes perforated by EM forming and perforation process can be used in automobile muffler. In conventional punching of tubes, the metal sheet is perforated by conventional practices, then rolled and welded into tubes, which is time consuming. In magnetic pulse forming and punching, only one step is needed. The arrangement of tool coil and the workpiece defines the process type. In the experiment of electromagnetic tube expansion coil is placed inside the tube while for compression, the tube is placed inside the coil. In magnetic pulse forming and punching, the coil is placed inside the tube for the expansion of the tube and punches are placed around the tube. With this arrangement, punching happens during expansion. A large amount of work has been done on electromagnetic forming of tubes and magnetic pulse welding. Psyk et al. [1] studied the electromagnetic forming and identified that different applications of electromagnetic forming could be achieved by different geometries and arrangement of coils. Bay et al. [2] studied the electromagnetic forming process, material behaviour and computational modelling. Electromagnetic forming experiments were performed for free forming and cavity fill operations by Oliveira et al. [3]. The electromagnetic field generated and deformation pattern obtained during electromagnetic expansion of tube have been studied by Ghatule and Kore [4]. They concluded that maximum electromagnetic field is generated at the middle of the coil. To concentrate the magnetic field in the required area, field shapers are used in electromagnetic forming. Suzuki et al. [5] did experiments to study the effect of different geometries of field shaper in electromagnetic tube bulging operation. Gotoh et al. [6] compared the highstrain rate shearing process with quasi-static strain rate process for the shearing of commercial aluminium, by using the shearing speed of 10 m/s. They concluded that at the high-strain rate process, the material flow was concentrated in a narrow band of the region known as an adiabatic shear band. Because of the adiabatic shear band, the appearance of the sheared-off edge is smooth and, in case of low-speed shearing, the appearance of the sheared-off edge looks dull and fibrous. Many researchers have studied the high-speed impact of a projectile on the sheet. Børvik et al. [7] carried out some experiments with the help of gas gun projectile having three different nose shapes to perforate the 12-mm-thick sheet plate at high velocity. They concluded that shape of nose plays a very significant role in failure pattern as well as in energy absorption mechanism. It was observed that conical and hemispherical projectile penetrates the plate by pushing the material while blunt projectile caused failure by plugging. Rusinek et al. [8] studied the failure process of mild steel sheet subjected to the hemispherical projectile. Petalling phenomenon was observed when the projectile is lubricated, while for dry condition radial necks were observed which reduces the formation of petals.

5 Magnetic Pulse Forming and Punching of Al Tubes—A Novel …

69

The impacted target workpiece materials may fail by different modes like spalling, scabbing, plugging, petalling and fragmentation [9]. Fracture results when the primary stress value is more than ultimate compressive strength, and it occurs in low strength, low density targets. Plugging develops nearly cylindrical slug as the same of projectile shape and size. In this case, fracture appears due to large heat produced by the moving slug. In shear deformation, heat is generated, and it is restricted in narrow region, which decreases the strength of the material. The decrease in strength reduces the stability, and it is called an adiabatic shearing process. Plugging is mostly found if blunt projectile penetrates thin, hard plates. Petalling is produced by high radial and circumferential tensile stresses. The detail study and investigation in the literature survey shows that the electromagnetic forming and perforation of tubes had not been studied. Even for the perforation of the tubes, no direct method is available in the literature besides some punching machine which uses a die and punches for the punching operation. Some of the materials are not easy to perforate, and perforation of these types of material with conventional perforation process is difficult. In case of high-strain rate perforation, the kinetic energy of the projectiles is far more than that required to penetrate the workpiece, so for localized perforation, structural deformation surrounding the hole is less for high-strain rate perforation as compared to low-strain rate processes. In this research study, a novel technique is developed for the combined forming and punching of tubes.

5.2 Experimental Methodology 5.2.1 Working Principle The equivalent circuit diagram for the set-up of magnetic pulse forming of the tube is shown in Fig. 5.1. In this circuit diagram, capacitor (C), inductance (L) and resistor (R) are in series. The magnetic forming and punching set-up consists of capacitor bank, punches, a coil and die as shown in Fig. 5.2. During the process, a high capacity

Fig. 5.1 Equivalent circuit diagram for magnetic pulse forming

70

S. Pawar et al.

Fig. 5.2 Schematic representation of magnetic pulse forming and punching of tube set-up

voltage (V ) is supplied to the capacitor bank for a small duration of time. After the charging of capacitor bank, the energy stored (E c ) in the capacitor bank is released through the coil, which is given by Eq. (5.1), E c (t) =

1 C V (t)2 2

(5.1)

where E is the electric field intensity (V/m). The current flowing through the coil generates a magnetic field; this generated magnetic field induces an eddy current in the workpiece which establishes another magnetic field in the conductive workpiece. These two magnetic fields oppose each other, and the resultant force is used for deforming the workpiece. The relations between changing electric and magnetic fields are described by Maxwell’s equations, and these governing equations are given by Eqs. (5.2)–(5.5),  · E = ρ ∇ εo

(5.2)

 · B = 0 ∇

(5.3)

  × E = − ∂ B ∇ ∂t

(5.4)

  × H = Jfree + ∂ D ∇ ∂t

(5.5)

5 Magnetic Pulse Forming and Punching of Al Tubes—A Novel …

71

where ρ is charge density (C/m3 ), εo is permittivity (F/m), B is magnetic flux density (Wb/m2 ), H is magnetic field intensity (A/m), and J is volume current density (A/m2 ). According to Lenz law, the direction of induced current is in such a way that the magnetic field produced by the induced current opposes the original magnetic field. The magnetic fields generate Lorentz force (F), given by Eq. (5.6), F = J × B

(5.6)

The current passes through the coil is given by Eq. (5.7),  I =U

C −βt e sin(ωt) L

 (5.7)

where U is the electric potential, β is the damping factor, and ω is the frequency. The magnetic field generated increases with the increase in discharge coil current, given by Eq. (5.8), B=

μN I l

(5.8)

where μ is the permeability, N is the number of turns of the coil, and l is the length of the coil.

5.2.2 Experimental Set-up The experimental set-up consists of die made of cast iron, coil made up of 16 mm2 Cu cable, punches made up of 5 mm dia. SS Allen screws and 100 m long Al 6061-T6 tubes having an outer diameter 50 mm and thickness 1.5 mm. Two types of punches, namely pointed and concave punches, are manufactured, and they are fitted in the die. Punch geometries and manufactured punches are shown in Fig. 5.3. The coil is connected to the capacitor bank by high voltage switch; it is placed inside the tube, and the tube is placed inside the die. Nylon rod is used to support the coil. Punches are placed around the tube and fitted in the die as shown in Fig. 5.4b. The length of the die is 100 mm, and holes are drilled on the periphery of the die. For easy removal of the die after punching, die is divided into four parts as shown in Fig. 5.4a. Figure 5.5 shows the different parts of the experimental set-up.

72

S. Pawar et al.

Fig. 5.3 a Pointed punch and b concave punch

Fig. 5.4 a Die front VIEW and b punch fitted in die [10]

Fig. 5.5 Experimental set-up

5.3 Results and Discussions Experiments are carried out on heat-treated Al 6061-T6 tubes. To increase the workability of Al 6061-T6 tube, annealing is done. Annealing is a heat treatment process where a metal is heated to recrystallization temperature and then permitted to cool slowly. The workability of the metal can be restored by producing crystal having a new set of slip planes. This process softens the metal so that the cutting and forming operations on the metal can be done easily.

5 Magnetic Pulse Forming and Punching of Al Tubes—A Novel …

73

The samples are heat treated in an oven for 3 h at the temperature of 4120 °C. Then, the heating is stopped, and the samples are allowed to cool down slowly in the oven itself. Experiments are conducted at different discharge energies. At low discharge energies, deformation is not observed. After 6.0 kJ discharge energies for both the pointed as well as concave punches, maximum deformation is observed, and clear perforation is possible. Discharge energies are varied from 2.9 to 6.2 kJ as shown in Table 5.1. From the experiment, it is observed that with the increase in discharge energy, the diameter of tube increases. The maximum deformed diameter of the tube was found to be 56 mm.

5.3.1 Experiment with Pointed Punch Figure 5.6 shows the tubes after the experiments with pointed punch. Discharge energies are varied from 4.1 to 6.2 kJ. It can be observed that clear perforation is Table 5.1 Operating conditions

Discharge energy (%)

Discharge energy (kJ)

55

2.9

8.0

80

60

3.4

8.7

80

65

4.1

9.5

80

70

4.7

10.2

90

75

5.4

11.0

94

80

6.2

11.7

104

Fig. 5.6 Tubes after the experiment with pointed punch punching

Voltage (kV)

Current (kA)

74

S. Pawar et al.

Fig. 5.7 Petal formation in case of pointed punch punching

possible at 6.2 kJ discharge energy. The average diameter of hole obtained at 6.2 kJ discharge energy is 3 mm. Separation of the slug from the hole is not observed in case of pointed punch punching. The diameter of tube after the deformation is 56 mm. Cross section of hole in case of pointed punch punching shows that petalling phenomenon as shown in Fig. 5.7. The alloy in front of punch is pushed out latterly with the onset of failure in terms of circumferential and radial necks. This causes a petalling phenomenon in case of pointed punch punching.

5.3.2 Experiments with Concave Punch Al tubes after the experiments with concave punches are shown in Fig. 5.8. The average diameter of hole obtained at 6.2 kJ discharge energy is 4 mm, which is greater than the pointed punch arrangement. In concave punch arrangement, complete removal of sheared slug takes place, and the separated slug is shown in Fig. 5.9. The cross section of hole shows clear punching in case of concave punch as shown in Fig. 5.10. In comparison with a pointed punch, concave punch shows better results as the diameter of the perforated holes is more, and the complete removal of slug takes place. The diameter of hole perforated by concave punch is near to the required diameter.

5 Magnetic Pulse Forming and Punching of Al Tubes—A Novel …

75

Fig. 5.8 Tubes after the experiment with concave punch punching

Fig. 5.9 Slug separated from the hole

5.4 Conclusions The experiments carried out in this research paper show how the electromagnetic forces can be used for the combinations of operation, i.e. for the forming and punching of the tubes simultaneously. The cross-sectional study of hole shows petalling

76

S. Pawar et al.

Fig. 5.10 Clear perforation in case of concave punch

phenomenon in case of pointed punch while a clear punching is possible with concave punch. The detailed study of the effect of geometry of punches on the punching shows that the concave punches are better than the pointed punches as the clean slug separation is possible as well as the diameter of the hole punched is also closer to the required diameter of hole.

References 1. Psyk, V., Risch, D., Kinsey, B.L., Tekkaya, A.E., Kleiner, M.: Electromagnetic forming—a review. J. Mater. Process. Technol. 211(5), 787–829 (2011) 2. Bay, F., Jeanson, A.C., Zapata, J.A.: Electromagnetic forming processes: material behaviour and computational modelling. Procedia Eng. 81, 793–800 (2014) 3. Oliveira, D.A., Worswick, M.J., Finn, M., Newman, D.: Electromagnetic forming of aluminum alloy sheet: free-form and cavity fill experiments and model. J. Mater. Process. Technol. 170(1), 350–362 (2005) 4. Ghatule, P., Kore, S.D.: Coupled 3D finite element modeling of electromagnetic free expansion of Al tube. Int. J. Adv. Mater. Manuf. Charact. 3(1), 95–98 (2013) 5. Suzuki, H., Murata, M., Negishi, H.: The effect of a field shaper in electromagnetic tube bulging. J. Mech. Work. Technol. 15(2), 229–240 (1987) 6. Gotoh, M., Yamashita, M.: Study of high-rate shearing of commercially pure aluminum sheet. J. Mater. Process. Technol. 110, 253–264 (2001) 7. Børvik, T., Langseth, M., Hopperstad, O.S., Malo, K.A.: Perforation of 12 mm thick steel plates by 20 mm diameter projectiles with flat, hemispherical and conical noses: part I: experimental study. Int. J. Impact Eng. 27(1), 19–35 (2002) 8. Rusinek, A., Rodríguez-Martínez, J.A., Zaera, R., Klepaczko, J.R., Arias, A., Sauvelet, C.: Experimental and numerical study on the perforation process of mild steel sheets subjected to perpendicular impact by hemispherical projectiles. Int. J. Impact Eng. 36(4), 565–587 (2009)

5 Magnetic Pulse Forming and Punching of Al Tubes—A Novel …

77

9. Backman, M.E., Goldsmith, W.: The mechanics of penetration of projectiles into targets. Int. J. Eng. Sci. 16(1), 1–99 (1978) 10. Kore, S.: A system for simultaneously electromagnetic shape forming and perforation on electrically conductive work piece and a method there of. Patent application number 201831020436

Chapter 6

Experimental Investigation on the Forming of AA 5052-H32 Sheet Using a Rigid-Body-Based Impact in a Shock Tube S. K. Barik , R. Ganesh Narayanan

and N. Sahoo

Abstract In the present study, a high-velocity sheet-forming experiment has been performed by using a hemispherical end nylon striker inside a shock tube. The striker attains a high velocity after its interaction with the shock wave inside the shock tube and impacts the sheet metal mounted at the end flange which tries to deform it at a high velocity. Three different velocity conditions have been attempted to track the progress in sheet metal forming. The formability of the material at different velocities is measured by the various parameters such as bulge height, effective strain distribution and the limit of strain at the necking region. A comparative analysis has been performed with a quasi-static punch stretching experiment to analyse the effectiveness of the high-velocity forming. From the results, it is observed that the sheet is deformed uniformly without strain localization and the limiting strains are increased by 50–60% after the high-velocity forming. Keywords High-velocity forming · Shock tube · Striker · Inertial effect · Limiting strain

6.1 Introduction There is a significant rise in demand for lightweight alloys in the aerospace and automobile industries which enables an opportunity in the area of research to reduce the fuel consumption and greenhouse emission. In the present days, lightweight materials of higher strength such as aluminium and magnesium alloys are the key alternatives to reduce the weight of the vehicle in order to improve the overall efficiency [1]. From a review from Joost [2], it has been observed that the fuel efficiency of a vehicle can be improved by 6–8% for each 10% reduction in weight which encases a new direction to achieve the fuel economy and greenhouse emission targets. Friedlyander

S. K. Barik · R. Ganesh Narayanan (B) · N. Sahoo Department of Mechanical Engineering, Indian Institute of Technology Guwahati, Guwahat, Assam 781039, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_6

79

80

S. K. Barik et al.

et al. [3] have concluded that weight savings of 20–25% can be achieved by replacing the steel surface body of the automobile with the lightweight materials. However among all the lightweight materials, aluminium alloys are generally preferable due to their ease of availability, higher strength to weight ratio, increase in resistance to corrosion and better recycling potential [4]. Despite several advantages, the noticeable drawbacks such as inferior formability at room temperature, higher potential of wrinkling during conventional forming and higher initial cost of the aluminium alloys than the traditionally used steel make more expensive and difficult to use them in mass production in any structural application [5]. However, the formability of aluminium alloys can be enhanced either by forming through an elevated temperature or by forming at higher strain rates as aluminium alloys are higher sensitive to rate of loading at room temperature condition when it reaches a threshold limit. The strain rate has a significant effect on the material properties. Due to this reason, determining strain-dependent mechanical properties accurately is one of the great issues during material properties analysis. Traditionally, quasi-static tensile test and split Hopkinson Pressure Bar (SHPB) test are the two standard experiments have been widely used to extract the plastic properties of the materials at lower and higher strain rate, respectively [6, 7], but the major limitation in these experiments is obtaining the results only at a uniaxial direction which is not wise to use the same results in the prediction of material properties during multi-axial forming. For that reason, quasi-static bulge testing technology has been introduced for the measurement of multi-axial mechanical properties [8, 9]. In order to study the strainrate-dependent properties, various modifications have been performed on the basic bulge testing device, but in a limitation, several complexities have been observed in the experimental set-up during the experiment. It has also been observed that if the forming process is performed by any high energy rate forming (HERF) processes such as electrohydraulic forming, electromagnetic forming and explosive forming which develop a high strain rate loading condition can be used as the devices to extract rate-dependent material properties [10–12]. These high-velocity forming processes also help to increase the forming limit of the materials significantly because of the most influencing factor called inertial stabilization [11, 12] which results in an increase in limiting strain by 45–50% when compared with the conventional forming processes. Besides several advantages in all the above HERF processes, the most common limitations are high capital cost, complexity in instrumentation and difficulties in handling. In order to minimize the above limitations, a shock tube has been used in various studies to study the dynamic response of the thin metallic sheets subjected to varying levels of shock loading [13, 14]. Stoffel [13] performed experimental studies on steel and aluminium plates and modelled a mathematical relation by considering elastic– plastic behaviour, kinematic hardening and strain rate sensitivity. Justusson et al. [14] used a shock tube to extract the bi-axial rate-dependent properties of aluminium where the out of plane deflection and strain fields were recorded by using the DIC technique. Looking into the application of the shock tube, the present work involves in the application of a hemispherical end nylon striker which is driven by the assistance of

6 Experimental Investigation on the Forming of AA 5052-H32 Sheet …

81

Table 6.1 Mechanical properties of AA5052-H32 sheet Material

RD

σ ys (MPa)

σ u (MPa)

n

K (MPa)

eu (%)

et (%)

r

AA5052H32 (1 mm)

0°

161 ± 2

223 ± 3

0.19

350 ± 3

7.8 ± 1.4

8.5 ± 1.2

0.63

45°

177 ± 3

236 ± 3

0.18

366 ± 3

7.6 ± 0.8

8.5 ± 0.8

0.85

90°

152 ± 2

214 ± 2

0.18

378 ± 4

7.7 ± 1.2

8.4 ± 1.1

0.85

Table 6.2 Chemical composition of AA5052-H32 sheet (Weight percentage) Material AA5052-H32 (1 mm)

Mg

Cu

Si

Fe

Mn

Cr

Al

2.68

0.10

0.92

0.31

0.13

0.32

Balance

the shock wave. The striker moves at a high velocity after receiving energy from the shock wave and deforms the sheet material at a higher velocity which is kept at the end of the shock tube. It depicts a kind of material stretching process used to analyse the material-forming behaviour in multi-axial direction, but at a higher strain rate.

6.2 Experimental Procedure 6.2.1 Material Details and Mechanical Properties In the current experimental investigation, AA5052-H32 alloy of 1 mm thickness is considered for the analysis. In order to characterize the material properties, tensile testing has been performed in accordance with ASTM-E8. The plastic strain ratio (r) across 0°, 45° and 90° to the rolling direction has been obtained as per ASTM-E517. Both the tests are conducted in a 200 kN Universal Testing Machine with a crosshead speed of 1 mm/min. All the tests are conducted thrice to ensure repeatability. The mechanical properties of the tested material along the three directions are listed in Table 6.1. The chemical composition of the base material is obtained by EDX analysis and the elemental distribution is shown in Table 6.2.

6.2.2 Shock Tube Experiments In the present work, a striker has been driven by the help of the shock wave inside the shock tube. The shock tube is a long rigid cylinder which is divided into a high-pressure-driver section and a low-pressure-driven section. Both the sections are separated by a diaphragm made of a metal sheet or by layers of thin Mylar sheets. When the pressure difference across both the driver and driven section reaches a critical value, the rapid rupture of the diaphragm generates a shock wave inside the

82

S. K. Barik et al.

Fig. 6.1 Shock tube facility used for experiments

tube and it propagates along the driven section with a Mach number greater than one. The propagation of the shock wave generates a high-pressure and high-temperature gas behind it. When the shock wave travels down the tube and imparts on an end wall, it is reflected back by producing a higher pressure and temperature gas behind the reflected shock wave. This sudden rise and propagation of the pressure can be used to drive a rigid body at a high velocity by keeping it inside the shock tube. The rigid body or the striker can be used as a punch which hits the specimen kept at the end of the tube and deform it at a high velocity. In the present work, a shock tube has been used with an overall length of 4 m, consisting of a 2-m driver section and a 2 m of driven section as shown in Fig. 6.1. The inner diameter of the tube is 55 mm with a thickness of 11 mm. At the end of the shock tube, two flanges having a male and a female groove of inner diameter 185 mm are installed in order to mount a circular specimen. Both the end flanges have a free area to deform of diameter 105 mm. A striker made of nylon rod is used in the experiment which has a hemispherical end with a diameter 54.7 mm and an overall length of 95 mm. A digital pressure gauge is installed in the driver section to measure the bursting pressure. Two pressure transducers (PCB Piezotronic; USA; Model 113B22) of sensitivity 14.62 mV/bar have been mounted in the driven section at a distance of 385 and 885 mm from the end of the tube to measure the incident shock pressure, reflected shock pressure as well as the shock Mach number during the experiment. One more pressure transducer is mounted at the middle of an end flange to measure the pressure generated when the striker approaches towards the end flange to hit the specimen as shown in Fig. 6.2. The signals from pressure transducers are sent through a signal conditioner (PCB Piezotronics Model 482C) before getting recorded by the Yokogawa DLM2022 digital oscilloscope. In order to measure the striker velocity, an IR emitter and a photodiode receiver are mounted in the shock tube facing towards each other. Both the IR sensors are connected to a circuit with an LM358 (voltage comparator). A DC voltage of 5 V is supplied to the circuit for powering the sensors and the output terminal from the circuit is connected to an oscilloscope. When the striker passes through the sensors,

6 Experimental Investigation on the Forming of AA 5052-H32 Sheet …

83

Fig. 6.2 Schematic of the shock tube with instrumentation

an interruption occurs while receiving the IR rays from the emitter to the photodiode which results in a sudden drop in voltage and it continues until the striker passes through the sensors. Once the striker passes, the photodiode again receives the signal from the IR emitter. Both the sensors are mounted at a distance of 300 mm from the position of the striker.

6.2.3 Strain Distribution Strain distribution on the sample during deformation is also an important indicator of formability of a sheet. Higher uniformity in strain distribution results in higher limiting strains that lead to better formability. In order to measure the effective strain distribution after the sheet metal forming, circular grids are printed on the surface of the samples by chemical etching method. The average diameter of each circle is 2.63 ± 0.01 mm. Since the material is anisotropic, Hill’s 1948 yield function is used for the calculation of effective strain. The limit of deformation has been set by the velocity of the striker because it is not possible to stop the test exactly at the necking at such high strain rates. The limiting strain generated during this process is compared with a conventional punch stretching tests in which the samples are deformed up to the initiation of necking.

84

S. K. Barik et al.

Table 6.3 Average pressures obtained from experiments Avg. Burst pressure, bar (psi)

Avg. Incident Shock pressure, bar (psi)

Avg. Reflected shock pressure, bar (psi)

Avg. Pressure at End flange bar (psi)

Avg. Shock Mach number (M s )

6.08 ± 0.3 (88.32 ± 5)

2.41 ± 0.1 (34.95 ± 1.45)

5.4 ± 0.1 (78.32 ± 1.45)

18.32 ± 0.1 (265.7 ± 1.45)

1.46 ± 0.01

6.74 ± 0.3 (97.75 ± 5)

2.49 ± 0.2 (36.11 ± 2.9)

5.71 ± 0.2 (82.81 ± 2.9)

25.09 ± 0.2 (363.9 ± 2.9)

1.49 ± 0.01

7.45 ± 0.3 (108.05 ± 5)

2.66 ± 0.2 (38.58 ± 2.9)

6.21 ± 0.2 (90.06 ± 2.9)

40.25 ± 0.3 (583.77 ± 4.5)

1.52 ± 0.01

6.3 Results and Discussion 6.3.1 Calibration of Shock Tube Pressures The shock tube experiments with the striker are conducted thrice to ensure the repeatability in bursting pressure and the Mach number generated inside the shock tube. The average incident pressure, reflected pressure and the pressure at the end flange as well as the shock Mach number (M s ) for different experimental conditions are shown in Table 6.3. The pressure signal sensed by the pressure transducers is illustrated in Fig. 6.3. The shock wave reflects back after hitting the striker which helps to drive the striker forward at a high velocity. From the pressure signals, it is noticed that the rise in pressure at the end flange is quite high due to the rapid motion of the striker which compresses the air in front of it. The pressure at the end flange monotonically increases with the higher bursting pressure. This sharp rise in pressure may also contribute to the metal forming process which results in more uniformity and enhancement in formability.

6.3.2 Determination of the Striker Velocity When the striker passes through the IR Sensors, there is an interruption in receiving the IR rays from the IR emitter which results in a sharp drop in voltage. When the striker crosses the sensors, again the voltage rises to its original value. Figure 6.4 illustrated the voltage output obtained from the IR sensor for the case of an experiment at Mach number (M s ): 1.46 which shows a sharp drop in voltage signal for a lapse of time when the striker passes through the sensors. The velocity of the striker is obtained by dividing the drop time with the length of the striker. Table 6.4 shows the velocity of the striker for all the set of experiments.

6 Experimental Investigation on the Forming of AA 5052-H32 Sheet …

Fig. 6.3 Pressure time history at bursting pressure of a 6.08 bar b 6.74 bar and c 7.45 bar Fig. 6.4 Signal obtained from the IR sensor for an experiment at Mach number (M s ): 1.46

85

86 Table 6.4 Measured velocity by IR sensor for the set of experiments

S. K. Barik et al. Avg. Shock Mach number (Ms )

Time (Ms )

Velocity of the striker (Ms )

1.46 ± 0.01

2.82 ± 0.08

33.62 ± 0.87

1.49 ± 0.01

2.29 ± 0.06

41.48 ± 1.06

1.52 ± 0.01

1.90 ± 0.06

49.79 ± 1.33

6.3.3 Free Forming Experiments Several experiments are conducted at different loading conditions to analyse the forming behaviour of the 1-mm-thick aluminium sheets, but only three different velocity conditions are taken into consideration for the better analysis. The results obtained from the high-velocity experiments are compared with an experiment performed quasi-statically by using a conventional hydraulic punch-stretching device. The peak deflection at different velocity conditions is measured by a vertical heightmeasuring device, and the results are depicted in Fig. 6.5. From the results, it is observed that with the rise in velocity of forming, the material is stretched more uniformly and it results in higher deflection without strain localization, whereas for the case of quasi-static forming, due to the necking, the material cannot deform further and it results in lower dome height. Generally, during a low rate of plastic deformation in materials, the local velocity at any local site varies simultaneously. When the velocity gradient between the two regions reached a critical value, the necking phenomenon gradually initiates and it restricts further material deformation which results in failure at that particular location [15]. When the material stretching is performed at high velocity, the added acceleration minimizes the velocity gradient by producing an inertial force on the material. The inertial forces produce additional tensile stress in the region outside Fig. 6.5 Peak deflection at different velocity loading conditions

6 Experimental Investigation on the Forming of AA 5052-H32 Sheet …

87

the necking region which helps the material to deform further before it fails due to necking [11, 15].

6.3.4 Strain Distribution Analysis The effective strain distribution along the rolling direction (X-axis) and transverse direction (Y-axis) for 1-mm-thickness sheets are shown in Fig. 6.6. As reported in the literature [12], in a conventional punch stretching, due to the frictional constraint between the punch and the workpiece, the major and minor strains are distributed nonuniformly in the pole region which results in higher strain gradient in the frictional zone and it restricts the material flow by necking. But from the results obtained in Fig. 6.6, it is clearly illustrated that the strain uniformly increases from the base region to the pole region and a single peak is observed near to the pole which confirms the absence of frictional constraint between the striker and the specimen and a biaxial mode of stretching is observed in both the directions. During all the trials, necking starts near the pole region, but always appears along the X-axis (RD) which results in less variation in the strain at different velocity conditions as shown in Fig. 6.6a, because the material stretching is restricted due to the initiation of necking. But, along the Y-axis (TD) due to the further stretching of the material, the effective strain value increases even higher at higher velocity as illustrated in Fig. 6.6b. The deformed samples at different velocities are shown in Fig. 6.7. At velocity 41.48 m/s, the necking is initiated but when the material is stretched at even higher velocity such as 49.79 m/s, the crack is initiated in the necking region without further material stretching.

Fig. 6.6 Effective strain distribution along a X-axis and b Y-axis at different velocities for 1-mm thickness sheet

88

S. K. Barik et al.

Fig. 6.7 Deformed sheets of AA 5052-H32 at different velocity conditions

6.3.5 Formability Limits In order to analyse the increment in the forming limits during the high-velocity deformation studies in shock tube experiments, the sheet material is stretched up to failure. The distribution of failure strains for both high-speed forming and quasi-static forming is shown in Fig. 6.8, from which it is notified that for the case of 1-mmthickness sheet, all the points during the high-velocity forming lie approximately in the 18.3–44.5% (Major strain) and 5.5–18.25%. (Minor strain) strain range. However, for the case of quasi-static condition, the strain limits lie in the range of 3.8–10.4% (Major strain) and 2.2–6.8% (Minor strain), respectively. A significant improvement in the formability is observed during this comparative analysis for the case of highvelocity forming. The most dominating phenomenon during the high-velocity forming is the inertial stabilization [11, 15]. The inertial forces during the material stretching help the material to avoid sharp local velocity gradient in between the two adjacent points on the material surface which helps in further stretching without strain localization. Apart from this, due to the generation of higher dislocation density in the material during the high-velocity forming results in higher tendency of multi-slip motion of dislocation, and helps in higher plasticity and strength of the material which is absent in quasi-static loading cases [12]. Fig. 6.8 Limit strains comparison between quasi-static test and shock tube test

6 Experimental Investigation on the Forming of AA 5052-H32 Sheet …

89

6.4 Conclusions A high-velocity forming experiment has been conducted by using a shock tube to understand the material-forming behaviour. From the experimental analysis, the following conclusions are derived. • Formability of AA 5052-H32 sheets of 1-mm thickness is found to be significantly higher than the conventional quasi-static mode of forming. From the experimentally determined forming limits, it is concluded that the limit strains are increased by nearly 50–60% during the biaxial mode of forming. The significant improvement in strain limit occurs due to the effect of inertial stabilization. • Due to the frictional constraint between the punch and the sheet in quasi-static metal forming, necking occurs at the region where the velocity gradient is more. This results in lower the bulge height when compared to the experiments performed at a high velocity where the material is stretched uniformly without strain localization resulting in larger dome height. • The strain distribution across both the X-axis and Y-axis during the high-velocity forming confirms that the frictional constraint is very less between the striker and the sheet which results in a uniform bi-axial stretching and the deformation is large near the pole region.

References 1. Lutsey, N.: Review of technical literature and trends related to automobile mass reduction technology. Inst. Transp. Stud. 1(3), 1–40 (2010). http://doi.org/10.1016/j.ijhydene.2015.03.004 2. Joost, W.J.: Reducing vehicle weight and improving US energy efficiency using integrated computational materials engineering. 64(9), 1032–1038 (2012). https://doi.org/10.1007/s11837012-0424-z 3. Fridlyander, I.N., Sister, V.G., Grushko, O.E., Berstenev, V.V., Sheveleva, L.M., Ivanova, L.A.: Aluminum alloys: promising materials in the automotive industry. Met. Sci. Heat Treat. 44(9– 10), 365–370 (2002). https://doi.org/10.1023/A:1021901715578 4. Smerd, R., Winkler, S., Salisbury, C., Worswick, M., Lloyd, D., Finn, M.: High strain rate tensile testing of automotive aluminium alloy sheet. Int. J. Impact Eng 32(1–4), 541–560 (2006). https://doi.org/10.1016/j.ijimpeng.2005.04.013 5. Guo, W.G., Zhang, X.Q., Su, J., Su, Y., Zeng, Z.Y., Shao, X.J.: The characteristics of plastic flow and a physically-based model for 3003 Al-Mn alloy upon a wide range of strain rates and temperatures. Eur. J. Mech. A/Solids 30(1), 54–62 (2011). https://doi.org/10.1016/j.euromechsol. 2010.09.001 6. Singh, N.K., Cadoni, E., Singha, M.K., Gupta, N.K.: Dynamic tensile behaviour of multi phase high yield strength steel. Mater. Des. Elsevier Ltd 32(10), 5091–5098 (2011). https://doi.org/ 10.1016/j.matdes.2011.06.027 7. Yu, H., Guo, Y., Lai, X.: Rate-dependent behaviour and constitutive model of DP600 steel at strain rate from 10–4 to 103s–1. Mater. Des. Elsevier Ltd 30(7), 2501–2505 (2009). https:// doi.org/10.1016/j.matdes.2008.10.001 8. Hill, R.C.: A theory of the plastic bulging of a metal diaphragm by lateral pressure. The London, Edinburgh, and Dublin Philos. Mag. J. Sci. 41(322), 1133–1142 (1950). https://doi. org/10.1080/14786445008561154

90

S. K. Barik et al.

9. Oliveira, D.A., Worswick, M.J., Finn, M., Newman, D.: Electromagnetic forming of aluminium alloy sheet: gree-form and cavity fill experiments and model. J. Mater. Process. Technol. 170(1– 2), 350–362 (2005). https://doi.org/10.1016/j.jmatprotec.2005.04.118 10. Dariani, B.M., Liaghat, G.H., Gerdooei, M.: Experimental investigation of sheet metal formability under various strain rates. Proc. Institution Mech. Eng. Part B J. Eng. Manuf. 223(6), 703–712 (2009). https://doi.org/10.1243/09544054JEM1430 11. Balanethiram, V.S., Daehn, G.S.: Enhanced formability of interstitial free iron at high strain rates. Scripta Metallurgicaet Materialia 27(12), 1783–1788. https://doi.org/10.1016/0956716x(92)90019-B 12. Ahmed, M., Kumar, D.R., Nabi, M.: Enhancement of formability of AA5052 alloy sheets by electrohydraulic forming process. J. Mater. Eng. Perform. 26(1), 439–452 (2017). https://doi. org/10.1007/s11665-016-2446-0 13. Stoffel, M.: Evolution of plastic zones in dynamically loaded plates using different elasticviscoplastic laws. Int. J. Solids Struct. 41(24–25), 6813–6830 (2004). https://doi.org/10.1016/ j.ijsolstr.2004.05.060 14. Justusson, B., Pankow, M., Heinrich, C., Rudolph, M., Waas, A.M.: Use of a shock tube to determine the bi-axial yield of an aluminium alloy under high rates. Int. J. Impact Eng 58, 55–65 (2013). https://doi.org/10.1016/j.ijimpeng.2013.01.012 15. Ghosh, A.K., Hecker, S.S.: Failure in thin sheets stretched over rigid punches. Metall. Trans. A 6(5), 1065–1074 (1975). https://doi.org/10.1007/BF02661361

Chapter 7

An Experimental Study on Single-Point Incremental Forming of AA5083 Sheet Using Response Surface Methodology Gautam Kumar, Saurabh, Maharshi Roshan, Kumar Nandan and Kuntal Maji Abstract This article presents an experimental study on the effects of different process parameters on the formability of AA5083 aluminum alloy sheet in single-point incremental forming (SPIF) using response surface methodology (RSM). Tool diameter, incremental depth, and feed rate were considered as inputs, and forming wall angle and average thickness reduction or thinning of the deformed sheet were taken as outputs in forming a square pyramid shape by SPIF. RSM models were developed for the considered outputs based on the experimental data collected following the central composite design of experiment method. Forming wall angle was increased with the increase in incremental depth, and optimum values were obtained for both tool diameter and feed rate for maximum wall angle. The average deformed sheet thickness was decreased with the increase in tool diameter and feed rate, and it increased with the incremental depth. Maximum forming angle and minimum thinning were obtained for the combination of higher feed rate, lower incremental depth, and higher tool diameter. Keywords Aluminum alloy sheet · Single-point incremental forming · Response surface methodology · Optimization

7.1 Introduction Single-point incremental forming (SPIF) is a low-cost die-less flexible sheet metalforming process suitable for small batch production and rapid prototyping compared to the conventional sheet metal-forming processes which require high investment for tooling. The demand for lightweight materials such as sheet metals of aluminum alloy, especially in automobile and aerospace industries for saving energy and environmental aspects, has increased. AA5083 alloy has the highest strength among 5000 series and has very good formability. Sheet metal parts made of aluminum alloy is widely used in car manufacturing, ship construction, pressure vessels, and other G. Kumar (B) · Saurabh · M. Roshan · K. Nandan · K. Maji Department of Mechanical Engineering, National Institute of Technology, Patna 800005, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_7

91

92

G. Kumar et al.

Fig. 7.1 Schematic diagram of increment forming

vehicle bodies, storage tank, fuselage and wings of modern aircraft, etc. Incremental forming has better formability than that of conventional sheet metal forming due to localized and through-thickness shearing deformations. Investigating the effects of different process parameters for identifying optimal processing conditions is important for effective applications of the SPIF process. Figure 7.1 shows the schematic diagram of incremental forming. Blank holder is used to holding the sheet properly. In Fig. 7.1, ti and tf are the initial sheet thickness and deformed sheet thickness, respectively, and α represents the wall angle of the deformed sheet. Some research works on the effects of different process parameters in SPIF carried out in the past are discussed below. Jeswiet et al. [1] investigated single-point incremental forming of aluminum alloy 3003 sheet of different thicknesses to maximum forming angles. It was revealed that large extent of plastic deformation could be achieved, and stress–strain behavior could be characterized for plane strain condition using compression test similar to Ford test. Martins et al. [2] proposed an analytical model to predict formability limits in single-point incremental forming considering different modes of deformation using membrane analysis. The proposed model could explain the effects of different parameters and the enhanced formability of the process. Bhattacharya et al. [3] investigated the effects of different process parameters on maximum formable angle and roughness of deformed sheet surface incrementally formed using experiments based on the design of experiment methods. Tool diameter and incremental stepdown values were observed to have negative effect on the maximum forming angle. Maximum formable angle was found less for thinner sheets. There was no significant effect observed of feed rate on the formable angle. Larger tool nose diameter was found to have deteriorating effects on the surface finish of the formed sheet. Kurra et al. [4] investigated the formability and sheet thickness distributions of extra deep drawing steel through incremental forming of different wall shapes like conical, circular, elliptical, and exponential varying wall angle using both experiments and numerical simulations. Finite element simulation results were found to be in good agreement compared to the experimental data. Finite element analysis of single-point incremental forming of magnesium alloy sheet was carried out by Senthil et al. [5] to

7 An Experimental Study on Single-Point Incremental Forming …

93

estimate stress–strain and thickness distributions. The numerically predicted were in good correlation with the experimental results. Single-point incremental forming of annealed aluminum alloy sheet of grade 3003 was studied by Shanmuganatan et al. [6] though conducting experiments based on design of experiments and response surface models. It was found that the average thickness of deformed sheet increases with the decrease in tool diameter, increase in feed rate and increase in step depth and smaller tool diameter increases the formability. Empirical modeling of surface roughness was developed by Kurra et al. [7] using different artificial intelligence techniques like neural network, support vector, and genetic programming by considering different processing conditions by varying the nose diameter of forming tool, step height, feed rate, wall angle, and lubricating conditions. Different surface roughness measures were analyzed and optimal processing conditions were determined to obtain the maximum surface finish. Kurra et al. [8] presented multi-stage incremental forming of steel sheets to get steeper wall angles deformed sheet, experimentally. The cylindrical, square, and spherical cups shape was formed using multi-stage incremental forming, and effect of different process parameters on it was discussed. Gupta et al. [9] investigated the effects of feed rate and tool rotation speed on heat generation in incrementally formed AA5754-H32 alloy sheet. Temperatures were found to be increased with the increase of movement between tool and workpiece. It was observed that formability increased due to heat generated by increasing tool rotation speed. Raju et al. [10] investigated a single-point incremental forming for multiple commercially pure aluminum sheets. The author found that the major true strain was higher on topmost sheet and decreases from the top to bottom sheet. Mostafanezhad et al. [11] studied the deformation of Al 1050 sheet in incremental forming done by utilizing two tools simultaneously through Box-Behnken design of experiments and response surface methodology. Parametric effects were investigated considering parameters such as forming angle, tool diameter, sheet thickness, and incremental step height on the maximum resultant force taken as output responses. It was found that the sheet thinning was most influenced by wall angle and forming force was affected by initial thickness and step down. The objective of this paper is to study the effects of the process parameters on the average wall angle and sheet thinning in SPIF using experiments and through the response surface methodology. The optimal parameters for obtaining maximum wall angle and minimum thickness reduction of the deformed sheet are also to be determined by the desirability-based response optimization method.

7.2 Experimental Setup and Procedure The experimental work on SPIF was carried out on a three-axis CNC vertical milling machine (Model No.: BMV 35 T12, Make: BFW, India) with tooling consisting of a holding plate, a backing plate, base plates, and SPIF tools as shown in Fig. 7.2a. Experiments were conducted on SPIF of Al-alloy (AA5083) sheet of 1.0-mm thickness having the composition as shown in Table 7.1. Aluminum alloy sheet samples

94

(a)

G. Kumar et al.

(b) Tool holder Forming Tool AA5083 Blank Clamped Fixture

Fig. 7.2 Experimental setup of SPIF a CNC milling machine b tooling

of required size were cut in a wire EDM machine and then the samples were clamped in the fabricated fixture as shown in Fig. 7.2b. Experiments were performed following the CCD method. The ranges of process parameters, i.e., tool diameter, step height, and feed rate for conducting SPIF of the AA5083 sheet as given in Table 7.2. Tool rotational speed was kept constant at 150 rpm. Fifteen sets of experiments (23 + 2 × 3 + 1 = 15) according to the CCD design of experiments method for three input factors. Each set of experiment was repeated three times. The fabricated forming tool and deformed sheet metal of square pyramid shape are shown in Fig. 7.3a, b, respectively. The deformed sheet metal was measured in a coordinate measuring machine (Model No: DCC 112-102 CMM, Make: Phoenix), and the forming wall angle was calculated using the triangulation method as shown in Fig. 7.4. Experimental data collected based on the design of experiments method is given in the following table. Average of three values for each set of experiments is reported in Table 7.3. Response surface modeling and optimization of the wall angle and sheet thinning was carried out to study the effects of different process parameters such as, tool diameter, step height and feed rate on the responses, namely wall angle and sheet thinning as discussed in the following section.

7.3 Response Surface Modeling and Optimization Response surface methodology (RSM) is a collection of mathematical and statistical techniques which are very much useful for modeling and optimization of responses or outputs of interest of any manufacturing process or system that are influenced by several independent variables or inputs. The responses or the outputs can be expressed in terms of the measurable independent variables or input parameters as given in Eq. (7.1).

Si

0.4

Element

%wt present

0.4

Fe

Table 7.1 Chemical composition of AA5083 0.1

Cu 0.4–1.0

Mn 0.25

Zn 0.15

Ti

0.05–0.25

Cr

4.0–4.9

Mg

Balance

Al

7 An Experimental Study on Single-Point Incremental Forming … 95

96

G. Kumar et al.

Table 7.2 Input parameters with levels Sl. no

Parameters

Minimum value

Mid value

Maximum Value

1.

Tool diameter (mm)

6

8

10

2.

Step height (mm)

0.25

0.5

0.75

3.

Feed rate (mm/min)

60

80

100

(a)

(b)

Φ 10 mm

Φ 8 mm

Φ 6 mm

Fig. 7.3 a Forming tool and b deformed sheet

Fig. 7.4 Coordinate measuring machine with the deformed sheet

Probe bar Deformed Sheet

y = f (x1 , x2 , . . . xn ) + ε

(7.1)

where x1 , x2 , . . . xn are the independent variables, and ε is the observed error or noise of the response y. RSM determines a suitable approximation functional relationship between the response and the independent variables based on the experimental data using the least square method of error minimization. Usually, a quadratic or secondorder polynomial within the ranges of the independent variables considered is used to obtain an approximate model based on the experimental data collected from the process or system. Multiple regression analysis is used for developing the empirical

7 An Experimental Study on Single-Point Incremental Forming …

97

Table 7.3 Experimental data on SPIF of AA5083 sheet based on CCD method Sl. no.

Inputs

Outputs

Tool diameter d (mm)

Step height (mm)

Feed rate (mm/min)

Average wall angle (°)

Average thinning (mm)

1

6

0.25

60

21.4

0.06

2

10

0.25

60

16.1

0.06

3

6

0.25

100

49.6

0.44

4

10

0.25

100

48.8

0.31

5

6

0.75

60

18.3

0.07

6

10

0.75

60

22.4

0.04

7

6

0.75

100

54.2

0.43

8

10

0.75

100

46.7

0.32

9

6

0.50

80

45.2

0.28

10

10

0.50

80

40.0

0.24

11

8

0.50

60

21.6

0.06

12

8

0.50

100

51.8

0.43

13

8

0.25

80

55.5

0.29

14

8

0.75

80

61.8

0.31

15

8

0.50

80

61.0

0.35

models required in the RSM, as given in Eq. (7.2). y = β0 +

k 

βi xi +

i=1

k 

βii xi2 +

i=1

k−1  k 

βij xi xj ,

(7.2)

i=1 j=2

where β values are called the regression coefficients. The method of least square is used to determine the regression coefficients of Eq. (7.2). The response optimization was carried out using the desirability function method. This method works through the transformation of the response variable f (xi ) to a desirability value D, where, 0 ≤ D ≤ 1 and xi are the input variables. The desirability value increases as the response increases in case of maximization. A desirability function D(f (xi )) lying within a range of (0, 1) for any function f (xi ) can be produced using the following transformation as the given Eq. (7.3).

D(f (xi )) =

⎧ ⎪ ⎨ ⎪ ⎩

0,  if f (xi ) ≤ fmin r , if fmin ≤ f (xi ) ≤ fmax

f −fmin fmax −fmin

1,

if

(7.3)

f (xi ) ≥ fmax

where f min , f max, and r, are denoting minimum and maximum values of the function, and weight value varying between 0.1 and 10 signifies the shape of the desirability

98

G. Kumar et al.

function. A target value and an allowable minimum response value are required to set to maximize a response. The desirability values for the response above the target value and below the minimum acceptable value are taken as one and zero, respectively. The closer the response to the target, the closer the desirability is to one. To determine the optimal solution, desirability value is maximized using a reduced gradient algorithm. Similarly, the desirability function for minimization of the response value may also be calculated as given in Eq. (7.4).

D(f (xi )) =

⎧ ⎪ ⎨ ⎪ ⎩

1,  if f (xi ) ≤ fmin r , if fmin ≤ f (xi ) ≤ fmax

f −fmin fmax −fmin

0,

if

(7.4)

f (xi ) ≥ fmax ,

7.4 Results and Discussion A second-order (quadratic) model was developed for the response, i.e., wall angle using the RSM based on the collected experimental data based on the CCD method. The analysis was carried out using the Minitab 17 software. The following expressions were obtained for the wall angle, α in terms of input variables. α = −407.7 + 41 d − 94.8 h + 7.33 f − 2.518 d 2 + 95.6 h2 − 0.03993 f 2 + 0.68 dh − 0.0222 df − 0.018 hf

(7.5)

The R2 value of the developed model of the response bending angle was found to be equal to 0.967, which indicates the adequate ability of the model is to make further predictions. The ANOVA test was carried out as given in Table 7.4. Here, the terms, such as DF, Seq SS, Adj SS, Adj MS, F, and P stand for the degrees of freedom, sequential sum of square error, adjusted sum of square error, adjusted mean Table 7.4 Results of ANOVA for forming wall angle (α) in SPIF Source

DF

Adj SS

Adj MS

F

P

Regression

9

11183.1

1242.57

112.81

0.000

Linear

3

6977.4

2325.81

211.15

0.000

Square

3

4182.4

1394.12

126.57

0.000

0.71

0.555

139.54

0.000

Interaction

3

23.3

7.77

35

385.5

11.01

Lack-of-fit

5

369.6

73.92

Pure error

30

15.9

0.53

Total

44

11568.6

Residual error

7 An Experimental Study on Single-Point Incremental Forming …

99

Fig. 7.5 Variation of forming wall angle for different combinations of input variables

square error, F value, and P value test statistic, respectively. P values of all of the terms in the developed model were found to be significant for 95% confidence level. The variation of the forming wall angle with the different input process parameters, i.e., tool diameter, step height, and feed rate are shown in Fig. 7.5.

100

G. Kumar et al.

Table 7.5 Results of ANOVA for thinning (t) in SPIF Source

Adj SS

Adj MS

F

P

Regression

DF 9

0.923897

0.102655

170.98

0.000

Linear

3

0.843657

0.281219

468.38

0.000

Square

3

0.064494

0.021498

35.81

0.000

Interaction

3

0.015746

0.005249

8.74

0.000

9.50

0.000

Residual error

35

0.021014

0.000600

Lack-of-fit

5

0.012881

0.002576

Pure error

30

0.008133

0.000271

Total

44

0.944911

The wall angle increased first and then decreased with the increase in tool diameter. This happened due to initial increase in deformation for higher tool diameter, and further increase in tool diameter leads to higher spring back for larger tool-job contact area which reduces the wall angle. Wall angle increased with the increase in incremental depth or step height due to more stretching and shearing deformation. The thinning or reduction in thickness of the deformed sheet was expressed through the Eq. (7.6) in terms of the input process parameters considered. t = −0.996 + 0.1776 d + 1.934 h + 0.00011 f − 0.01014 d 2 − 0.889 h2 + 0.000003 f 2 − 0.0508 dh − 0.000073 df + 0.00025 hf

(7.6)

The R2 value for the response surface model of thickening was found 0.978 which shows the goodness-of-fit of the model to the experimental data. The ANOVA test for the thickening model was carried out as shown in Table 7.5. As the probability values (P) for the combined linear, square, and interaction terms were found to be less than 0.05, they were found to have significant contributions toward the response thickening. Surface plots for the thinning with the input variables are shown in Fig. 7.6. The thinning was increased with the increase in step height or incremental depth, and it decreased with the increase in tool diameter. However, the feed rate was found to have insignificant effect on the thickness reduction in SPIF of AA5083 sheet within the operating window considered. The reduction of the thickness of deformed sheet or thinning decreased with the increase in tool diameter due to the less deformation for larger contact area. The sheet thinning increased with the increase in incremental depth because of higher shearing deformation. The feed rate was found to have insignificant effect on thickness reduction of sheet in the considered range of parameters. The optimal solutions for maximum wall angle and minimum thickness reduction of the deformed sheet in SPIF were obtained as 51° and 0.22 mm, respectively. The lower and target values for wall angle were set as 17.2° and 61.9°, respectively, and 0.03 and 0.46 mm were given as the target and upper values for the reduction in sheet

7 An Experimental Study on Single-Point Incremental Forming …

Fig. 7.6 Variation of thinning for different combinations of input variables in SPIF

101

102

G. Kumar et al.

thickness in SPIF. The corresponding optimal input parameters were found as a tool diameter of 8.4 mm, a step height of 0.4 mm and a feed rate of 100 mm/min within the operating ranges of the parameters taken for a maximum composite desirability value of 0.6422. Therefore, the maximum formability obtained for higher feed rate, lower incremental depth and medium-sized tool diameter with respect to the operating conditions considered.

7.5 Conclusions Experimental investigations were carried out to study the effects of process parameters such as tool diameter, incremental depth or step height, and feed rate on the forming wall angle and reduction in thickness of the deformed sheet in SPIF of Al-alloy sheet. The effect of three different process parameters, namely tool diameter, feed rate, and incremental depth was studied on the wall angle and the average thickness of deformed sheet or thinning in single-point incremental forming through response surface methodology. The wall angle increased first and then decreased with the increase in tool diameter. This happened due to initial increase in deformation for higher tool diameter, and further increase in tool diameter leads to higher spring back for larger tool-job contact area which reduces the wall angle. Wall angle increased with the incremental depth or step height due to more stretching and shearing deformation. The thinning was increased with the increase in step height or incremental depth, and it decreased with the increase in tool diameter. However, the feed rate was found to have insignificant effect on the thickness reduction in SPIF of AA5083 sheet within the operating window considered. Maximum forming angle for minimum thinning was determined using desirability-based response optimization. Maximum formable angle with minimum thinning was obtained for higher feed rate and tool diameter, and lower incremental depth or step height. Acknowledgements The financial support of the Department of Science and Technology (DST), Science and Engineering Research Board (SERB), India for funding this research work under Early Career Research (ECR) award with project File No.-ECR/2016/001134 is gratefully acknowledged.

References 1. Jeswiet, J., Hagan, E., Szekeres, A.: Forming parameters for incremental forming of aluminium alloy sheet metal. IMechE Part B J. Eng. Manuf. 216, 1367–1371 (2002) 2. Martins, P.A.F., Bay, N., Skjoedt, M., Silva, M.B.: Theory of single point incremental forming. CIRP Ann. Manuf. Technol. 57, 247–252 (2008) 3. Bhattacharya, A., Maneesh, K., Reddy, N.V., Cao, J.: Formability and surface finish studies in single point incremental forming. J. Manuf. Sci. Eng. 133, 061020-1 (2011) 4. Kurra, S., Regalla, S.P.: Experimental and numerical studies on formability of extra-deep drawing steel in incremental sheet metal forming. J. Mater. Res. Technol. 3(2), 158–171 (2014)

7 An Experimental Study on Single-Point Incremental Forming …

103

5. Senthil, R., Gnanavelbabu, A.: Numerical analysis on formability of AZ61A magnesium alloy by incremental forming. Procedia Eng. 97, 1975–1982 (2014) 6. Shanmuganatan, S.P., Kumar, V.S.S.: Modeling of incremental forming process parameters of Al 3003 (O) by response surface methodology. Procedia Eng. 97, 346–356 (2014) 7. Kurra, S., Rahman, N.H., Regalla, S.P., Gupta, A.M.: Modeling and optimization of surface roughness in single point incremental forming process. J. Mater. Res. Technol. 4(3), 304–313 (2015) 8. Kurra, S., Nasih, H.R., Jasti, N.V.K., Maheshwar, D.: Experimental studies in multi stage incremental forming of steel sheets. Mater. Today: Proc. 4, 4116–4122 (2017) 9. Gupta, P., Jeswiet, J.: Observations on heat generated in single point incremental forming. Procedia Eng. 183, 161–167 (2017) 10. Raju, C., Haloi, N., Narayanan, C.S.: Strain distribution and failure mode in single point incremental forming (SPIF) of multiple commercially pure aluminum sheets. J. Manuf. Process. 30, 328–335 (2017) 11. Mostafanezhad, H., Menghari, H.G., Esmaeili, S., Shirkharkolaee, E.M.: Optimization of twopoint incremental forming process of AA1050 through response surface methodology. Measurement 127, 21–28 (2018)

Chapter 8

Study and Establishment of Manufacturing Process of Molybdenum Liners Using Warm Flow Forming Process P. S. S. R. K. Prasad , Navneet Verma , Narendra Kumar and K. M. Rajan Abstract Liners used in shaped charge help in achieving deeper penetration in target material. The penetration potential is directly proportional to square root of liner material density and maximum jet length. Studies reveal that molybdenum finds great potential for use as liners owing to its high sound speed and relatively high material density. The high sound speed helps in achieving high-velocity coherent jet tips which are important to attain higher jet length. The process of precise manufacturing of molybdenum liners is challenging. In this paper, warm flow forming has been attempted for the same, as molybdenum is difficult to form by cold working. Effect of warm working of the molybdenum material is studied. Design of blank and preform along with Press tools for making pre-form, rollers, mandrels, and tailstock support has been brought out. The process parameters such as speed and feed used in the process are explained in brief. Keywords Warm flow forming · Molybdenum liner · Shaped charge

8.1 Introduction Shaped charge plays an important role in the design of Armament systems. These are used in the warhead to defeat/damage the intended target using focused energy along the charge axis such that the target is rendered incapable of performing its intended function. Shaped charge is ideally an explosive device with a hollow cavity at one end and a detonator in the other. In this, a metallic liner in the shape of hemisphere or cone is surrounded by a high explosive charge and then encased in a suitable material such as steel or aluminum which provides confinement during detonation. Upon detonation, the liner material is deformed at a very high strain rate and forms high-velocity jet which penetrates target plates [1, 2]. P. S. S. R. K. Prasad (B) · N. Verma · N. Kumar · K. M. Rajan Armament Research and Development Establishment, Pashan, Pune, Maharashtra 411021, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_8

105

106

P. S. S. R. K. Prasad et al.

A number of factors guide the effectiveness of shaped charges. Most important of them being liner design, materials used, microstructure, and the manufacturing process. Of these, liner material is the key influencing factor, as the penetration potential is linked to ductility and density of liner material. However, this may not be necessary in all the cases as many other parameters also play crucial roles. Copper, having FCC crystal structure, has been used extensively as liner material over the years owing to its reasonable density and excellent ductility. On the other hand, certain body centered cubic (BCC) crystal structure materials like tantalum and molybdenum have shown great promise to be utilized for the purpose even though the BCC structure suggests that these materials are less than optimally ductile. As discussed above, the penetration potential is also dependant on jet length which in turn is determined by the jet tip velocity and particulation time. The jet tip velocity is further defined by bulk sound speed of liner material. In order to obtain a solid coherent jet, the collision velocity, responsible for jet formation, must be smaller than the liner material bulk speed of sound. Thus, a material with high bulk sound speed can have higher collision velocity and consequently higher jet tip velocity [1]. The maximum jet tip velocity that can be achieved is roughly about 2.34 times the bulk sound speed of the liner material [3].

8.2 Molybdenum as Liner Material The development of molybdenum-lined shaped charges is relatively less explored. However, the high sound speed of molybdenum (5.124 km/s vs. 3.94 km/s for copper) coupled with high material density (10.2 g/cc vs. 8.93 g/cc for copper) makes it an interesting choice for liner material. It has been reported that the jet velocities obtained using molybdenum have been experimentally found to be greater than 12 km/s which is about 25% higher than that obtained using copper. Though molybdenum produces very high jet velocity, in order to have increased jet ductility, processing of basic raw material is very important. The material should have high purity and the finest grain size with near equi-axed grains [4–6].

8.3 Characteristics of Molybdenum Material Molybdenum has low toughness and ductility at low temperature and therefore, in order to have improved formability, it must be heated above room temperature prior to forming. Also, during forming, the job experiences high variable strain rates and tri-axial stresses; hence, molybdenum is usually formed in hot/warm conditions in order to decrease the probability of brittle fracture. Molybdenum heats and cools more rapidly due to its high thermal conductivity and low specific heat, so maintaining of temperature during flow forming is critical. Molybdenum readily oxidizes when heated in air at temperature above 450 °C, the rate of oxidation is lower at low

8 Study and Establishment of Manufacturing Process of Molybdenum …

107

temperature because of formation of an adherent oxide layer. This oxidation can be prevented by working at temperature below 450 °C or by working in vacuum or in a controlled non-oxidizing atmosphere [7, 8].

8.4 Manufacturing of Molybdenum Liner 8.4.1 Manufacturing Processes for Liner Liner can be manufactured by several manufacturing processes, and these include machining, press working, cold forging, warm forging, hot forging, electroforming, flow forming, etc. Each of these methods has their relative advantages and disadvantages, and affects the final performance of liner. In the present work, flow forming process is chosen so as to achieve high dimensional consistency and also desired grain flow pattern.

8.4.2 Intricacies in Liner Manufacturing by Flow Forming Achievement of dimensional accuracy in liner manufacturing is very critical as it directly affects the performance of shape charge. Some of the intricacies involved are maintaining of thickness variation, run out, perpendicularity of base with liner axis, concentricity of cone, surface finish, etc. These intricacies have to be overcome by accurate mounting of flow forming mandrel and rollers, suitable clamping of blank and correct center support, proper selection of process parameters, etc.

8.4.3 Warm Flow Forming of Molybdenum Liner To carry out warm Flow Forming, various methods that can be employed are preheating of blanks in furnace, high induction heating, heating by hot air in chamber, flame heating, laser heating, heating by heated roller, and friction heating [9].

108

P. S. S. R. K. Prasad et al.

Fig. 8.1 Molybdenum cone

Table 8.1 Design parameters of the cone

Material

99.95% Molybdenum

Thickness difference

≤ 0.04 ⊥ axis

Hardness of liner

230–250 HV

8.5 Experimental Setup The details of liner and tooling are explained below.

8.5.1 Salient Features of the Component Figure 8.1 shows the actual configuration of the molybdenum cone and Table 8.1 highlights the basic design parameters.

8.5.2 Machine The machine used is a 2-roller CNC flow forming machine capable of manufacturing axially symmetric rotating hollow components, e.g., cones, ogive, etc.

8 Study and Establishment of Manufacturing Process of Molybdenum …

109

8.5.3 Tooling Used Main toolings for flow forming are mandrel, rollers, and tail stock. During flow forming, the toolings are subjected to severe service conditions, and therefore, design and manufacture of the tooling must be robust enough to withstand the forming stresses. Mandrel is a rotating element which supports the job during flow forming process. It must have same shape as that of the final component and should be hard with high fatigue strength to resist wear and fatigue loading. Geometrical accuracy and surface finish of the mandrel are important to obtain desired run out and thickness variation. Design of mandrel is shown in Fig. 8.2, and the key design parameters are listed in Table 8.2. Rollers in flow forming are power-driven and used to apply force on the blank as it gets formed over the mandrel. The geometry of the roller has a significant effect on the power consumed and accuracy of the job. They also encounter extreme loading conditions during forming. The complete configuration of roller and its interaction with blank is explained in Figs. 8.3 and 8.4 and the main design features are listed in Table 8.3. Nose radius and roller incident angle play an important role in achieving desired thickness and surface finish. It also helps in avoiding unwanted thickening of work material ahead of forming region. Tailstock Support Hydraulically-operated tailstock support is used to provide sustained pressure to grip the blank on the mandrel before the start of flow forming operation. The tip of the tailstock support should be optimally designed to ensure Fig. 8.2 Mandrel for flow forming

68°

R 7.0

* All Dimensions in mm

Table 8.2 Design parameters of mandrel

Material

ASTM A681—D2

Hardness

50–60HRC

Surface finish

0.8 µm

110

P. S. S. R. K. Prasad et al.

* All Dimensions in mm

Fig. 8.3 Roller design Fig. 8.4 Roller and blank interaction

Table 8.3 Roller configuration and incident angle

Material

Tool Steel ASTM A681—D2

Hardness

58–62 HRC

Roller diameter and Nose radius

300 mm; R3

Roller incident angle

5°

8 Study and Establishment of Manufacturing Process of Molybdenum …

111

that the start of roller and blank interaction is as close as possible to the tip of cone in order to minimize undeformed zone at the tip.

8.6 Experimental Process Manufacturing of liner was evolved through iterative process involving different pre-form designs, pre-heating and in-process heating of blank, and by variations in different flow forming parameters. Following sequence of activities were undertaken to carry out the experiments for manufacturing of molybdenum liner.

8.6.1 Preparation of Blank The raw material used is obtained after thermo mechanical processing (TMP), a process utilized to improve the room temperature ductility by optimizing microstructure. In flow forming process, axial thickness of the formed component is always equal to the original thickness of the blank. This follows the sine law and as such diameter of the finished component is equal to the diameter of the starting blank [10]. The thickness of the blank is derived from sine rule: t = T Sin α

(8.1)

where, α = Half angle of cone = 34° t = Final thickness of component = 1.2 mm T = Blank thickness = 1.2/sin 34° ≈ 2.0 Also, Diameter of the blank = 75 mm

8.6.2 Pre-Form Design A pre-form was made by creating a dimple of depth 6 mm with included angle 68° in blank sheet by press working, as shown in Fig. 8.5. This facilitates holding of blank on mandrel using tail stock by applying appropriate pressure. Pre-form operation was carried out in warm condition. Figure 8.6 shows the pre-formed blanks for flow forming.

112

P. S. S. R. K. Prasad et al.

Fig. 8.5 Pre-form nosing operation

Fig. 8.6 Pre-formed molybdenum blanks

8.6.3 Flow Forming Operation Pre-formed blank soaked in a furnace at 400 °C for 30 min was mounted on the mandrel with the help of tailstock support. Flow forming operation was then carried out while continuously heating the workpiece using oxy-acetylene flame in order to avoid heat loss, as shown in Fig. 8.7. The cone just after finishing of flow forming operation is shown in Fig. 8.8. Another important aspect in flow forming is selection of process parameters, i.e., mandrel speed and roller feed. After carrying out several experimental trials, following parameters were finalized Speed of the mandrel 150 rpm Feed 15 mm/min Gap between roller and mandrel 1.2 mm

8 Study and Establishment of Manufacturing Process of Molybdenum …

113

Fig. 8.7 Flow forming with oxy-acetylene flame heating

Fig. 8.8 Flow formed cone on the mandrel

8.6.4 Final Finishing Operations The post-forming finishing operations involving tip and length machining were performed on CNC lathe. Holding of semi-finished cone on the machine is critical and as such vacuum mandrel was used to avoid damage of finished surface as shown in Fig. 8.9.

8.7 Analysis and Observation The cones manufactured by flow forming were subjected to dimensional and metallurgical analysis. The comparative study of metallurgical reports of raw material and flow-formed cones were carried out.

114

P. S. S. R. K. Prasad et al.

Fig. 8.9 Vacuum mandrel for final finishing

8.7.1 Dimensional and Metallurgical Analysis The thickness of the cones were measured immediately after forming, i.e., before finishing operations on three different planes from nose to base. As evident from Fig. 8.10, thickness is slightly above the specified values at nose and base portions of the cone. This is attributed to difference in surface velocity along the length of the cone due to change in contact diameter from nose end to the base. Microstructure analysis for grain size and their orientation was carried out preand post-flow forming using optical microscope at 500X. As shown in Fig. 8.11, it was observed that the blank material obtained by thermo mechanical process have smaller grain which gets elongated after flow forming. The elongation of grains is more predominant in longitudinal direction due to rollers pressure and movement along the axis as compared to transverse direction.

Fig. 8.10 Thickness variation in molybdenum liners

8 Study and Establishment of Manufacturing Process of Molybdenum …

Longitudinal

115

Transverse

(a) Blanks obtained by TMP

Longitudinal

Transverse

(b) Finished Liner after Flow Forming Fig. 8.11 Microstructure of blank and flow formed liner at 500X

8.8 Conclusions In the present work, flow forming process is chosen to manufacture the liners in order to achieve high dimensional consistency and desired grain flow pattern. As molybdenum is difficult to form at room temperature, warm flow forming is carried out. The design aspects of tooling and effects of process parameters are covered in brief. Dimensional and metallurgical analysis of flow formed cone has also been presented. It has been observed that longitudinal elongation of grains occur due to roller pressure and directional flow of material during flow forming. With this process, molybdenum liners, as shown in Fig. 8.12, can be manufactured meeting all the design parameters. However, better ways of in-process heating can be explored to attain homogeneous heating to achieve better liner properties. Fig. 8.12 Flow formed molybdenum cone

116

P. S. S. R. K. Prasad et al.

References 1. Walter, W.P., Zukas, J.A.: Fundamentals of Shaped Charges. Wiley, New York (1989) 2. Poole, C.: Penetration of a Shaped Charge. Ph.D. Thesis, Corpus Christi College, University of Oxford (2005) 3. Held, Manfred: Liners for shaped charges. J. Battlefield Technol. 4(3), 1–6 (2001) 4. Lichtenberger, A., Verstraete, N., SaIignon, D., Daumas, M.T., Collard, J.: Shaped charges with Molybdenum liner. In: 16th International Symposium on Ballistics San Francisco, C.A., USA, 23–28 September 1996 5. Baker, E.L., Daniels, A., Pham, J., Vuong, T., DeFisher, S., Jet break-up characterization of Molybdenum shaped charge liners. US Army, Armament Research, Development and Engineering Center 6. Baker, E.L., Voorhis, G.P., Campbell, R., Choi, C.S.: Development of Molybdenum shaped charge liners producing high ductility jets. In: Proceedings of 14th International Symposium on Baliistics, Quebec, Canada (1993) 7. Gupta, C.K.: Extractive Metallurgy of Molybdenum. CRC Press (1992) 8. Campbell, F.C.: Elements of Metallurgy and Engineering Alloys. ASM International (2008) 9. Zhan, M., Yang, H., Guo, J., Wang, X.: Review on hot spinning for difficult-to-deform lightweight metals. Trans. Nonferrous Metals Soc. China 25, 1732–1743 (2015) 10. Roy, P.K.: An investigation on evaluation and optimisation of Process Parameters in Flow Forming. Ph.D. Thesis, University of Pune, India (1997)

Part II

Machining

Chapter 9

Influence of Surgical Drill Geometry on Drilling Performance of Cortical and Trabecular Bone Ramesh Kuppuswamy

and Brett Christie-Taylor

Abstract Numerous orthopedic operations involve inserting screws and wires to attach plates, help align, and immobilize fractured bones, so that they may heal correctly. These operations require drilling into the bone so that screws can be used in the operation. Furthermore, the surgical tools require a large degree of flexibility to access the surgical sites in the complex shaped human joints. The flexibility feature on the surgical drills has an adverse effect on the capacity to transmit high bone drilling forces. Furthermore, prevention of necrosis or cell death is crucial for surgical interventions such as bone marrow stimulation, bone marrow donor recruitment, bone biopsies, or removal of osteosynthesis materials. The more efficient the drilling operation, the smaller the impact on the bone. This results in a healthier bone which enables implants to support higher loads with a greater chance of success. This research unveils the influence of surgical drill geometry, specifically the effect that the web dimensions have on drilling performance of cortical and trabecular bone. The drilling performance was analyzed through study of process parameters: thrust force, temperature rise, and surface finish of the drilled hole. A series of experiments were carried out on both Poly-methyl Methacrylate (PMMA) and bovine bone test samples. Keywords Surgical drill · Cortical and trabecular bone · Web · Gashing

9.1 Introduction Minimally invasive surgical techniques have tremendous benefits for patients, but it offers huge demands on tool design. Despite technological advancements, minimally invasive techniques still require advanced surgical skills since sight and maneuverability of surgical instruments are reduced and hampered. The orthopedic operations involve inserting screws and wires into bones to attach plates that help align and immobilize fractures, so that the bones may heal correctly. In order for the operation R. Kuppuswamy (B) · B. Christie-Taylor Department of Mechanical Engineering, University of Cape Town, Cape Town, South Africa e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_9

119

120

R. Kuppuswamy and B. Christie-Taylor

to be a success, the plates need to be securely attached to the bone. The greater the force that these implants can withstand before they are removed, the higher the probability of the surgery being successful. In addition, the healthier the bone remains after surgery, the greater the success of the surgical endeavor. In order to screw into the bone, pilot holes need to be drilled before the screws can be set. The more efficiently that these holes are drilled and the smaller the impact that they have, the higher the chance of surgical success. There are numerous ways that drilling operations can affect the surrounding bone. The first of which is the heat that is generated during the drilling operation. Drilling operations generate a large amount of heat, which can be very detrimental to the bone strength. When the bone’s temperature is raised above 47 °C, thermal necrosis begins to occur [1]. Bone necrosis is the death of the surrounding bone tissue, and this can lead to microcracks and eventual bone collapse, which severely compromise the strength of the bone. This can lead to an implant failing due to the bone’s lowered strength. Thrust force is also an important measure of drilling performance. A low axial thrust force induces smaller strain on the surrounding bone, and hence, a more precise surgery can be accomplished. A lower thrust force may also lead to a better surface finish with fewer microcracks generated during the drilling process. This is important as an increase in the severity of these cracks can lead to a decrease in the stiffness and elastic modulus of the bone. Since implants into the bone rely on the mechanical strength of the bone in order to stay secure, weakening the bone may lead to a higher number of implant failures. By optimizing the drill efficiency, the impact on the surrounding bone is lessened and the chance of a successful implant is increased. This research is centered on drill geometry development for drilling cortical bone as it makes up nearly 80% of skeletal mass and has a dense outer surface and primarily responsible for supporting the body. When researching this topic, it was found that most past research was largely centered on drilling of cortical bone which is much denser than trabecular bone and is the main load-bearing element in our bodies. It forms an outer shell through which the trabecular bone runs. The past drilling research was also attempted on caprine femur bones, the middle diaphysis of porcine bones and the bovine bones and acrylic substance Poly-methyl Methacrylate (PMMA) as the properties closely resemble the human bone as described in Table 9.1 [2–4]. The literature survey suggests contradictory views on bone removal from the animal and storage to maintain its properties. Past drilling experiments were conducted on: fresh bones which are less than 2 h old; bone that was stored in deep freezer at around −40 °C; and bone specimen frozen at −20 °C for up to 4 weeks. However, the findings suggest no significant change in properties to those tested upon immediately [5]. The orthopedic drilling operations make use of cannulated and non-cannulated drill bits of size between 2.5 and 3.5 mm. Non-cannulated drill bits are the most common type of drill used in surgery, and cannulated drill bits are used in procedures where the orientation of the hole in three dimensions is critical for the success of the operation [6]. Although the surgical drill geometry features such as helix angle, drill point angle, clearance angle, and web thickness are important, past research tabulates the surgical drill performance behavior for the drill shape and helix angle [7]. A surgical drill processing tests have confirmed that higher drill speed resulted

9 Influence of Surgical Drill Geometry on Drilling Performance … Table 9.1 Properties’ comparison of human, bovine and organic bone

121

Properties

Human bone

Bovine bone

PMMA

Thermal conductivity W

0.16–0.34

0.15–0.35

0.15–0.4

Specific heat  

1140–2370

1300

1400

Thermal Diffusivity  2

0.44–0.56 × 10−6

0.3 × 10−6

0.11 × 10−6

1900

1800

1400

Tensile strength (MPa)

115

115

83

Shear modulus (GPa)

3.5

3.5

3

mK

J kg(K)

m sec



Density

kg m3



in the least increase in temperature [8]. Another study suggests that the increase in drill speed results in a decreased thrust force which leads to a smaller increase in the amount of heat generated [2]. However, another past experimental research on surgical drilling reveals that when the surgical drill speed on bone was increased from 1155 to 11,300 RPM, a temperature increase was observed, whereas when the surgical drill speed was increased from 27,000 to 97,000 RPM, a decrease in temperature was noted [9]. Avoiding the surgical drill breakage and optimizing the bone drilling performance have a positive impact in robotically assisted surgery [10]. All the past researches suggest that very little research exists that looks at the effect of web thinning in relation to bone drilling although it is expected that gashing and web thinning would have a large impact on the overall thrust force. Therefore, this research was attempted on understanding the effect of web thinning on the surgical drill performance.

9.2 Experimental Set-up To begin to understand the effect that web size has on surgical drilling performance, a series of experiments were planned. These were carried out on a high-speed machining center (RFM600). The tests were run using two different types of material. The first set of tests was run into PMMA specimens (30 × 30 × 15) to get an understanding of the expected behavior, and the second set of experiments was run into bovine tibia bone samples. The second run of tests was carried out on bovine tibia bone which was stored in a deep freeze until they were needed. For both tests, factors that

122

R. Kuppuswamy and B. Christie-Taylor

were investigated were the axial thrust, temperature rise, and drill surface quality. The test samples were mounted on the dynamometer to capture the drilling thrust signatures. The dynamometer was connected to a charge amplifier before the signal was sent to a Dewe-43, 8 channel DAQ box. This was connected to a computer, and the data was analyzed using Dewesoft-7-SE software. The data loggers for both the thermocouples and the dynamometer were triggered and stopped manually. Two experiments were run for each web size, and the readings averaged to give a final result. The experimental set-up can be seen in the following images. The diameter of the drill was set as 3 mm as this is a common size of drill used in orthopedic surgery. The web of a drill bit was thinned using a computer-controlled tool and cutter grinder. The S-type grinding technique was used to grind the webs to the correct size. Using a #600 grit silicon carbide grinding wheel, gashing was done for thinning the web of the selected surgical drill bit. After each pass, the size of the web was checked using a Leica optical microscope. The web feature of each surgical drill bit is shown in Fig. 9.1. Figure 9.2 shows the experimental set-up for surgical drilling process on PMMA and bovine tivia bone samples. Table 9.2 enumerates the cutting conditions used for the surgical drilling process.

9.3 Results and Discussion 9.3.1 Axial Forces Figure 9.3 shows the behavior of axial thrust with various web size while drilling the PMMA material at conditions: drilling speed: 28.3 m/min; feed rate: 180 mm/min; and web length 0.224–0.885 mm. The results suggest that when the web length was reducing from 0.885 to 0.224 mm, the axial thrust was reduced from 18.71– 19.12 N to 11.83–12.52 N, and the results were analyzed. When looking at the results from the experiments done into PMMA and bovine tibia bone, both show a linear relationship between the web size and the axial thrust force with a R2 value of 0.8825 and 0.8676, respectively. This agrees with the theoretical model which shows a linear relationship between web size and thrust force. The surgical drilling process is equated to a summation of processes: (i) axial extrusion process facilitated by the web (chisel edge) and (ii) cutting process facilitated by the lip of the cutting edges. At the downward feed of the surgical drill, a compressive force is applied at the web to facilitate a material flow from the test workpiece. When analyzing the theoretical model, one can see that when the web dimensions decrease, the length of the cutting lip increases, causing the component of the thrust force attributable to cutting to increase. However, while the cutting force increases, the component of the thrust force that is due to extrusion decreases, and this is shown to have a much greater impact on the size of the thrust force. The amount of extrusion is dependent on the web length, making the size of the web a very important factor

9 Influence of Surgical Drill Geometry on Drilling Performance …

123

Web-length = 0.885 mm; Drill diameter= φ3 mm

Web-length = 0.612 mm; Drill diameter= φ3 mm

Web-length = 0.552 mm; Drill diame-ter= φ3 mm

Web-length = 0.429 mm; Drill diameter= φ3 mm

Web-length = 0.224 mm; Drill diameter= φ3 mm

Fig. 9.1 Surgical drill bits used with different web length after the gashing process

124

R. Kuppuswamy and B. Christie-Taylor

Fig. 9.2 Experimental set-up and data acquisition arrangement for the drilling experiments on PMMA and bovine tibia bone

Table 9.2 Surgical drilling conditions for the PMMA and bone Experimental conditions

Values

Drill diameter (mm)

3

Surgical drill cutting speed (m/min)

28.3

Feed rate (mm/min)

120–180

Web length (mm)

0.885, 0.612, 0.552, 0.429, 0.224

in determining the overall thrust force. The theoretical model and experimental data follow the same trend. The extrusion area (AE) induced by the chisel edge is given as (see Fig. 9.4) AE = w · l where l = web length and w = width of the web

(9.1)

9 Influence of Surgical Drill Geometry on Drilling Performance …

125

F Experimental values for PMMA

25

25

20

20

15

15

10

10 cutting ratio (r) = 0.3 & shear angle = 17.5 deg cutting ratio (r) = 0.5 & shear angle = 30 deg

5

5

cutting ratio (r) = 0.75 & shear angle = 41 deg cutting ratio (r) = 1 & shear angle = 60 deg

0

0

0.2

0.4

0.6

0.8

1

0

a)

30

Surgical Drill Axial Thrust (F (N)

30

(Experimental Values for PMMA & Bone)

a)

a

values for PMMA)

Surgical Drill Axial Thrust (F (N) (Theretical

a

F Experimental values for Bone

Web length, l (mm)

Fig. 9.3 Axial forces both experimental and theoretical values for various cutting conditions on PMMA and bovine tibia bone

Fig. 9.4 Surgical drill geometry details

126

R. Kuppuswamy and B. Christie-Taylor

The applied stress to initiate the extrusion flow of material from the workpieces has to overcome both compressive stress and shear stress as given below. σa · w · l = σc · w · l + 2wlτ y

(9.2)

where σ a is the applied stress, σ c is the compressive stress, and τ y is the shear yield stress. As the surgical drill penetrates the bone material, the process of chip formation initiates when the maximum shear stress reaches a limiting value which is the shear yield stress. Applying the von Mises yield criterion, the shear yield stress is equated for the bone arterial given below [11] σy τy = √ 3

(9.3)

Equations (9.2) and (9.3) are combined and rewritten using the term applied load on the web, F w (as given below) σy Fw = σc · w · l + 2w · l √ 3

(9.4)

A merchant circle diagram was applied on the lip of the surgical drill (see Fig. 9.4), and hence F c is given as Fc =

Fs Cos(λ − α) Cos(φ + λ − α)

(9.5)

where λ = surgical drill-bone interface friction (0.5), φ = hear angle, α = rake angle (10°), and F s = shear force. The shear force F s is given as Fs =

wl t1 τs Sin(φ)

(9.6)

where wl = length of lip and t 1 = axial feed/revolution (0.06 mm/rev). The cutting ratio (r) was assumed from 0.3 to 1.0 and rake angle α = 10° and using Eq. (9.7), the shear angle (φ) is computed as r=

Sin(φ) Cos(φ − α)

(9.7)

Applying the properties of PMMA such as yield stress: 52–71 MPa, tensile strength: 47–70 MPa, shear stress: 25–62 MPa, cutting ratio (r): 0.2–1 mm, and lip angle of the surgical drill = 120°, F c was computed as

9 Influence of Surgical Drill Geometry on Drilling Performance …

Fc =

wl t1 τs Cos(λ − α) Sin(φ) Cos(φ + λ − α)

127

(9.8)

The axial force F a is summation of axial extrusion force facilitated by the web length and cutting force facilitated by the lip of the cutting edges and given as Fa = Fc + Fw

(9.9)

Applying the cutting conditions and physical properties applicable for surgical drilling of PMMA, the axial force was computed and expressed against the varying web length (see Fig. 9.5). Fa =

σy wl t1 τs Cos(λ − α) + σc · w · l + 2w · l √ Sin(φ) Cos(φ + λ − α) 3

(9.10)

The axial force values were computed using Eq. 9.10 and the material properties and compared with the experimental values. The results are shown in Fig. 9.3. The close conformity between the theoretical and experimental values for bovine tibia bone validates the theoretical analysis.

9.3.2 Drilling Temperature The temperature rise has an adverse effect on surgical drilling, and hence, the drilling process was analyzed for thermal damage. These readings were obtained during the same set of experiments that determined the axial force. An IR-750 infrared thermometer was used to obtain the temperature readings. The temperature of the bone was taken before the specimen was drilled and the hole temperature immediately after drilling. Table 9.3 lists down the maximum temperature measured while performing the surgical drilling operation on PMMA and bovine tibia bone. When analyzing the temperature results, there is also a direct relationship between the web and the temperature rise. However, the correlation is not as strong with a R2 value of 0.8493 for PMMA and 0.5624 for bone. This agrees with the results for the axial thrust which also decreases with web size. A lower thrust force means that the frictional force along the cutting lips will be smaller causing a decrease in the temperature rise. The lower correlation seen is due to the variation that is seen in the results. The friction between the drill bit and the hole walls generates a lot of heat, and the longer the contact, the greater the temperature rise. This is also the reason for the difference in the magnitude of the temperature rise between the bone and PMMA tests. The PMMA tests were done through a 15-mm sample, while the average bone thickness was around 7 mm. The increased drill time led to an increased temperature rise and higher recorded temperatures even though the drilling forces were very similar. This suggests that, even though the web size impacts the temperature rise, both the friction

128

R. Kuppuswamy and B. Christie-Taylor

Fig. 9.5 Surface texture of the drilled surface of a bovine tibia bone using various chisel-edged surgical drills at conditions: drill diameter (mm): φ3; surgical drill cutting speed (m/min):28.3; feed rate (mm/min): 122; and web length (mm): 0.885, 0.612, 0.552, 0.429, and 0.224

9 Influence of Surgical Drill Geometry on Drilling Performance …

129

Table 9.3 Temperature measured while performing the surgical drilling operation on PMMA and bone Web (mm)

Test no

Maximum temperature for PMMA (°C)

Maximum temperature for bone (°C)

0.885

1

41.83

25

2

44.71

24.3

1

37.82

24

2

36.89

24.5

1

34.57

22.7

2

33.2

23.1

1

30.4

21.9

2

29.77

22.3

0.612 0.552 0.224

between the hole wall and the drill bit and the overall drill time play a more important role in the heat generated during drilling.

9.3.3 Surface Quality The effect of web size on surgical drilling of bovine tibia bone was also studied through the measurement of the surface quality of the drilled hole. This was done by looking at the surface of the drilled hole using a scanning electron microscope. Images were obtained at X1500 magnification levels. Figure 9.5 shows the images that explain the surface quality of the drilled holes for the bovine tibia bones. Clearly, the quantity and severity of the cracks along the surface increased as the size of the web increased. The cause behind this increase is the higher thrust force that is seen with increasing web dimensions. A higher applied force means that, as the drill moves through the material, there is greater flex and wobble in the drill bit. This results in greater cracks and machining marks along the drilled surface of the bone. This is important as an increase in the severity of these cracks can lead to a decrease in the stiffness and elastic modulus of the bone. Since any implants into the bone rely on the mechanical strength of the bone to stay secure, weakening the strength of the bone may lead to a higher number of implant failures. The results from the SEM images show that the surface quality can be improved by decreasing the web length as this leads to less microcracks and in turn a stronger surface. It was initially assumed that PMMA would make a good substitute for bone due to the similar properties that both substances exhibit. After looking at both the force and temperature results, this was proven to be true. Initially, experiments were carried out into PMMA to get an understanding of what the results in bone could be expected to be. The reasoning for this was that PMMA is a uniform substance whose characteristics are well known and whose behavior is predictable, while bone, being an organic substance, may exhibit different properties in different samples. When

130

R. Kuppuswamy and B. Christie-Taylor

looking at the thrust force that was obtained for both sets of experiments, the results closely follow each other. PMMA had a slightly higher force experienced for each web size which can be explained by the different conditions that the experiments were conducted under. The linear relationship between web dimensions and thrust force was also observed for both PMMA and bovine tibia bone as well as the relationship between the web size and temperature rise. The difference in the magnitude of the temperature rise between the bone and PMMA samples can be explained by the different measurement techniques between the two experiments, the size of the samples, and the machining centers that were used. The conclusion of this is that the results that were obtained during testing into PMMA are a valid way of gaining an understanding of how the geometry will affect drilling performance into bone.

9.4 Conclusions The main aim of this project was to determine the effect that web thinning had on drilling performance into bovine tibia bone to allow for more successful orthopedic procedures. Through both theoretical analysis and experimentation, it was shown that web dimensions had a large impact on drilling performance, and in this way, the objectives of the report were met. It was found that there was a direct relationship between the size of the web and the axial thrust force. From this, one can conclude that, to minimize the thrust force, the size of the web should be reduced. A reduced axial thrust would allow the surgeon to be more precise during operations which would increase the chance of a success. It was also found that, by reducing the web size, the temperature rise during drilling could be reduced. This helps to eliminate the occurrence of thermal necrosis which is a significant factor in implant failure. Finally, by reducing the thrust force and temperature rise, the quantity and severity of the microcracks on the surface of the drilled material are reduced. This allows for the drilled hole to have greater surface quality and strength and further increases the chance of a successful operation. From these results, one can conclude that, to increase the drilling performance and improve the outcome of orthopedic procedures, the web dimensions of the surgical drill should be minimized. This is significant as improvements in operative techniques in orthopedic procedures will reduce both recovery time and repeat surgery as well as decrease surgical costs. Acknowledgements The authors wish to thank Dr. A. Devlig, University of Cape Town, Medical school, Cape Town, South Africa, toward the supply of bovine tibia bone samples and for the insight on surgical drill-related issues while drilling the bones and tissues. Further thanks go to Mrs. Penny Louw for her assistance with the optical microscope and her help with surface roughness testing; Mrs. Miranda Waldron for her assistance with the scanning electron microscope; and Mr. Horst Emrich for his assistance with the milling machine. This project was supported by fund NRF grant: Incentive Funding for Rated Researchers (IPRR)—South Africa through Reference: IFR150204113619 and Grant No: 96066.

9 Influence of Surgical Drill Geometry on Drilling Performance …

131

References 1. Sener, B., Guhan, D., Bahar, G., Ergun, K., Imad, S.: Effect drilling speeds of Irrigation temperature on heat control in vitro at different drilling depths. Clin. Oral Implants Res. 20, 294–298 (2009) 2. Mali, V., Warhatkar, H., Pawade, R.: Assessment of Cutting Forces and Temperature in Bone Drilling. In: All India Manufacturing Technology, Design and Research Conference, Pune, (2016) 3. Gupta, V., Pandey, P., Silberschmidt, V.: Rotary ultrasonic bone drilling: improved pull out strength and reduced damage. Med. Eng. Phys. 41, 1–8 (2017) 4. Augustin, G., Davila, S., Toma, U., Tomislav, S., Danko, B., Slaven, B.: Temperature changes during cortical bone drilling with a newly designed step drill and an internally cooled drill. Int. Orthop. 36, 1449–1456 (2012) 5. Elias, S.D., Carl, H.: Factors affecting the determination of the physical properties of femoral cortical bone. Acta Orthop. Scand. 37(1), 29–48 (1966) 6. Bertollo, N., Walsh, W.R.: Drilling of Bone: practicality, Limitations and Complications associated with Surgical Drill-Bits, Intech (2011) 7. Karmani, S., Lam, F.: The design and function of surgical drills and K-wires. Curr. Orthopaedics 18, 484–490 (2004) 8. Mohamed., Carl, W., Norman., Sherif, T.: Heat generation during implant drilling: the significance of motor speed. Oral Maxillofac Surg. 60, 1160–1169 (2002) 9. Kalidindi, V.: Optimisation of drill design and coolant systems during dental implant surgery. University of Kentucky master’s Thesis (2004) 10. Adili, A.: Robot-assisted orthopaedic surgery. Semin. Laparoscopic Surg. 11(2), 89–98 (2004) 11. Kapakjian, S., Steven, R.S.: Book. Manufacturing Processes for Engineering Materials, 5th edn. Pearson Education, Singapore (2009)

Chapter 10

Study of Cutting Temperature and Chip Formation in Drilling of AA6351–B4C Composite S. Thirumalai Kumaran , G. S. Samy , M. Uthayakumar and Tae Jo Ko Abstract The present study investigates the temperature rise in the cutting zone and its influences on chip formation behavior in high-speed drilling. The novelty of the study is to prepare a neutron-absorbing B4 C reinforced AA6351 composite subjected to the application of nuclear fuel transportation and storage. The stir casting method is used to prepare the composite. The TiN-coated solid carbide drill tool with different point angles of 118° and 135° is used to machine the composite. The experimental results show that the spindle speed majorly influences on temperature rise rather than feed rate for both the point angle tool. The maximum temperature value of 150.3 °C is noted for 135° point angle at higher spindle speed and feed rate. The chip formation with respect to different input conditions is analyzed through scanning electron microscopy and thermographic images. It is evident that the mass of chip extraction is increased with respect to an increase in spindle speed. The discontinuous and low pitch conical chips reduced the chip clogging and improved the drilling quality. Keywords Drilling · MMCs · Point angle · Temperature · Chip profile

10.1 Introduction The high-speed drilling of ceramic reinforced aluminum-based composite is an acute process to achieve desired geometrical and dimensional shape for any structural applications [1]. Moreover, the suitable fabrication process improves the physical properties of the composite. The conventional stir casting is widely used to fabricate such composite [2]. The secondary process is involved to attain closer tolerance value S. Thirumalai Kumaran · G. S. Samy · M. Uthayakumar Faculty of Mechanical Engineering, Kalasalingam Academy of Research and Education, Krishnankoil, Tamil Nadu 626126, India T. J. Ko (B) School of Mechanical Engineering, Yeungnam University, 214-1 Dae-dong, Gyeongsan-si, Gyeongsangbuk-do 712749, South Korea e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_10

133

134

S. Thirumalai Kumaran et al.

in the final product. Drilling is a foremost process to make holes which are used to construct the metal structure [3]. The temperature rise in the tool–work interface zone is a focal area to minimize the tool wear and surface roughness [4]. It is difficult to measure the tool–work interface temperature since the tool cutting edge is not visible during drilling. Researchers used several methods to measure the temperature of the cutting zone [5–7]. Uçak and Çiçek [8] experimentally investigated the influence of cutting parameter on the temperature and quality of drill hole. They used thermocouple to measure the temperature rise during drilling. They also reported that the cryogenic cooling reduced the heat generation in cutting zone and offered better surface quality. Taskesen and Kutukde [9] analyzed the drilling temperature of B4 C reinforced aluminum composite by using pyrometer. The results indicated that the temperature of the cutting zone increases with an increase in spindle speed. The increase in reinforcement majorly affected the cutting tool temperature and caused severe tool wear. Bhowmick et al. [10] used the infrared thermometer to measure the cutting zone temperature during drilling. In addition, thermography technique was also used to measure the tool, workpiece surface, and chip temperature. The results revealed that the MQL drill reduced the thermal softening of work material. Lower built-up-edge formation on the cutting edge reduced the thrust force and torque. The chip morphology with respect to temperature rise was effectively investigated by Samy et al. [11]. Ekkard Brinksmeier et al. [12] investigate the chip extraction during drilling of Ti6Al4V and the drill hole surface temperature was measured by infrared camera. They concluded that lower the feed rate offers good chip extraction. The chip accumulation in drill hole increased the temperature of the cutting zone. The present study experimentally reports the effect of temperature on chip formation behavior with respect to different input parameters. Infrared thermal image camera is used to measure the cutting zone temperature during the drilling of AA6351B4C composite. Initially, the temperature rise is examined with a different point angle, spindle speed, and feed rate. Finally, the chip formation profile is analyzed through scanning electron microscopy (SEM) and thermographic images.

10.2 Materials and Methods 10.2.1 Material Preparation and Properties The AA6351 with 10 wt% of B4 C MMC is fabricated through stir casting method. The base metal of AA6351 (Al–97.8%, Si–1%, Mn–0.6%, Mg–0.6%) has a high structural strength and boron carbide has a neutron-absorbing characteristic. B4C is preheated and dispersed into the melt with constant stirring action to achieve uniform particle distribution. The composite melt is then poured into a circular die [13]. The geometrical shape of composite is achieved through fine turning process. The tensile behavior of the prepared composite is tested as per ASTM B557M-10, and the results

10 Study of Cutting Temperature and Chip Formation in Drilling …

135

Fig. 10.1 Tensile behavior of the composite

are shown in Fig. 10.1. The uniform dispersion of B4C particles is ensured from the energy-dispersive spectroscopic image shown in Fig. 10.2. The prepared composite is used for nuclear fuel transportation and storage application.

10.2.2 Experimental Procedure The LMW JV 55 vertical drilling machine (Amal Jyothi College of Engineering, Kanjirappally, Kerala) is used to perform the experimental study under dry conditions. The drilling machine has a maximum spindle speed of 6000 rpm, feed rate of 10 m/min, and 7.5 kW spindle power [14]. The level of experiments is formatted by using Taguchi’s L18 orthogonal array. The drilling study is carried out with 118° and 135° point angle tool, spindle speed of 3000, 4000 and 5000 rpm, feed rate of 20, 40 and 60 mm/s. The drill hole is made by using monolayer TiN-coated solid carbide 8 mm diameter tool. The experimental setup and the measurement methods are shown in Fig. 10.3. The non-contact method of temperature measurement is done by using FLIR E60 infrared thermal image camera. The camera is fixed 1 m apart from the cutting zone and the surface temperature of the drill tool and machined surface is measured continuously during machining of the prepared composite. The captured thermographic image is analyzed by using FLIR Tools software. The emissivity of both the tool and the workpiece is found based on the reference to the black body. Finally, the emissivity of TiN-coated drill tool and the workpiece is determined to be 0.89 and

136

Fig. 10.2 EDS spectrum and elemental mapping

Fig. 10.3 Experimental setup

S. Thirumalai Kumaran et al.

10 Study of Cutting Temperature and Chip Formation in Drilling …

137

0.66, respectively. At the end of every consecutive machining, the extracted chips are collected to analyze the chip morphology.

10.3 Results and Discussion The reduction of heat generation in high-speed machining is an essential factor to minimize the tool wear and surface roughness. The proper selection of input parameter helps to reduce the cutting temperature during drilling of AA6351–B4 C composite. In addition to that, the chip formation behavior is majorly affected by temperature rise in the cutting zone. Figure 10.4a–f depicts the raise in temperature along the tool–workpiece interface zone during drilling of AA6351–B4 C composite at spindle speed of 3000 rpm and feed rate of 20 mm/s. The machining is considered as three stages. Fig. 10.4a depicts the first stage of machining since tool enters the workpiece and starts machining. In second stage, it is clearly evident that the temperature increases with an increase in the depth of hole penetration as shown in Fig. 10.4b–e [15, 16]. In third stage after finishing the hole, the temperature is rapidly dropped as shown in Fig. 10.4f. From the thermographic image, the cutting zone temperature and the maximum temperature are reported to examine the effects of input conditions. Furthermore, it is clearly evident that the temperature increases with an increase in the depth of hole penetration. Finally, the heat is distributed along the surface of the workpiece and through chip extraction. Figure 10.5a, b shows the influences of cutting parameter on the temperature rise for 118° and 135° point angle. The spindle speed has a major influence on the heat generation compared with feed rate. The increase in spindle speed increases the temperature from 71.9 °C to 140.5 °C. In the case of 118° point angle, the temperature rise with respect to the spindle speed is increased from 53.8 to 78.7%. Similarly, the temperature rise is increased by 50% with consequent increase in feed rate. In the case of 135° point angle, the heat generation is comparatively high. This is due to the high frictional contact between the tool and workpiece. The temperature rise at medium spindle speed is almost similar to that of 118° point angle. The maximum temperature of 150.3 °C is noted at higher spindle speed and feed rate. The heat generation in the cutting zone majorly influences the chip extraction during drilling. The increase in spindle speed accelerated the deformation of work material due to the increase in temperature at the tool–work interface. Fig. 10.6a–c shows the different chip profile at 118° point angle of the drill tool. The thermographic image witnesses the temperature distribution in the cutting zone, and it clearly shows the increase in mass quantity of chip extraction with increasing spindle speed. At low spindle speed of 3000 rpm, it is noted that the close pitch with continuous curling chips is produced due to low shearing strain [17]. Furthermore, the pitch of formed chip increases due to the continuous plastic deformation and angular stretch. At high spindle speed, the curling angle is furthermore reduced by the increase in angular stretch and generates concentric long continuous chips. These sharp curling chips disturb the extraction and increase the temperature.

138

S. Thirumalai Kumaran et al.

(a)

(d)

(b)

(e)

(c)

(f)

Fig. 10.4 a–f Temperature rise (a) first stage b–e second stage and f third stage

This leads to the degradation of tool life and surface quality. In the case of 135° point angle, discontinuous chips are formed at lower spindle speed which is due to low plastic deformation. At 5000 rpm, ribbon-shaped chips are formed which perturb the chip extraction in the flute causing high friction and temperature. Further, the regression model in Eq. (10.1) is generated using Minitab which predicts the future values, where C1, C2, and C3 represent the point angle, spindle speed, and feed rate, respectively.

10 Study of Cutting Temperature and Chip Formation in Drilling …

139

Fig. 10.5 a–b Temperature rise with respect to spindle speed and feed rate a 118° and b 135° point angle

Temp = 64−0.69 C1−0.0099 C2 + 2.52 C3 − 0.000000 C2 ∗ C2 −0.00062 C3 ∗ C3 + 0.000318 C1 ∗ C2−0.0115 C1 ∗ C3−0.000138 C2 ∗ C3 (10.1)

10.4 Conclusions The drilling of AA6351-B4 C composite is investigated to analyze the temperature rise and chip morphology. The following are the conclusions that are drawn: • The spindle speed majorly influenced the temperature rise irrespective to the point angle of tool. The temperature range of 71.9–140.5 °C was observed at 118° point

140

S. Thirumalai Kumaran et al.

(a) 3000 rpm

(b) 4000 rpm

(c) 5000 rpm Fig. 10.6 a–c Chip formation with 118° point angle tool

angle. The maximum temperature of 150.3 °C was measured at high spindle speed and feed rate for 135° point angle tool. • The increase in spindle speed increased the angular stretch of continuous chips and the pitch of consecutive curling (118° point angle tool). For 135° point angle, the wide contact of the cutting edge reduced the strain rate of the discontinuous chips and controlled the temperature. • The sharp curling and ribbon shape chips disturbed the chip extraction and increased the temperature at the cutting zone. This may cause the reduction of tool life and surface quality.

10 Study of Cutting Temperature and Chip Formation in Drilling …

141

Acknowledgements This work was supported by the Technology Innovation Program (or Industrial Strategic Technology Development Program (10067064, Development of tungsten carbide and PCD tools for high-quality finishing of high-precision parts and molds) funded By the Ministry of Trade, Industry & Energy (MOTIE, Korea).

References 1. Thirumalai Kumaran, S., Uthayakumar, M., Slota, A., Zajac, J.: Application of Grey relational analysis in high speed machining of AA (6351)-SiC-B4 C hybrid composite. Int. J. Mater. Prod. Technol. 51, 17–31 (2015) 2. Kalaiselvan, K., Murugan, N., Parameswaran, S.: Production and characterization of AA6061– B4 C stir cast composite. Mater. Des. 32, 4004–4009 (2011) 3. El-Sonbaty, I., Khashaba, U.A., Machaly, T.: Factors affecting the machinability of GFR/epoxy composites. Compos. Struct. 63, 329–338 (2004) 4. Matsumura, T., Hori, I., Shirakashi, T.: Analysis of cutting temperature in drilling process. Int.J. Mater. Form. 3, 499–502 (2010) 5. Bono, Matthew, Ni, Jun: A method for measuring the temperature distribution along the cutting edges of a drill. J. Manuf. Sci. Eng. 124, 921–923 (2002) 6. Ueda, T., Nozaki, R., Hosokawa, A.: Temperature measurement of cutting edge in drilling— effect of oil mist-. CIRP Ann. Manuf. Technol. 56, 93–96 (2007) 7. Cardoso, L., Teixeira, R., Henrique, C.: Contribution to dynamic characteristics of the cutting temperature in the drilling process considering one dimension heat flow. Appl. Therm. Eng. 3, 3806–3813 (2011) 8. Uçak, N., Çiçek, A.: The effects of cutting conditions on cutting temperature and hole quality in drilling of Inconel 718 using solid carbide drills. J. Manufact. Process. 31, 662–673 (2018) 9. Taskesen, A., Kutukde, K.: Non-contact measurement and multi-objective analysis of drilling temperature when drilling B4 C reinforced aluminum composites. Trans. Nonferrous Met. Soc. China 25, 271–283 (2015) 10. Bhowmick, S., Lukitsch, M.J., Alpas, A.T.: Dry and minimum quantity lubrication drilling of cast magnesium alloy (AM60). Int. J. Mach. Tools Manuf 50, 444–457 (2010) 11. Samy, G.S., Thirumalai Kumaran, S.: Measurement and analysis of temperature, thrust force and surface roughness in drilling of AA (6351)-B4 C composite. Measurement 103, 1–9 (2017) 12. Brinksmeier, E., Pecat, O., Rentsch, R.: Quantitative analysis of chip extraction in drilling of Ti6 Al4 V. CIRP Ann. Manuf. Technol. 64, 93–96 (2015) 13. Thirumalai Kumaran, S., Uthayakumar, M.: Investigation on the machining studies of AA6351SiC-B4 C hybrid metal matrix composites. Int. J. Mach. Mach. Mater. 15, 28–30 (2014) 14. Samy, G.S., Thirumalai Kumaran, S., Uthayakumar, M.: An analysis of end milling performance on B4 C particle reinforced aluminum composite. J. Australian Ceram. Soc. 53, 373–383 (2017) 15. Ozcelik, B., Bagci, E.: Investigation of the effect of drilling conditions on the twist drill temperature during step-by-step and continuous dry drilling. Mater. Des. 27, 446–454 (2006) 16. Tai, B.L., Stephenson, D.A., Shih, A.J: Workpiece temperature during deep-hole drilling of cast iron using high air pressure minimum quantity lubrication. J. Manuf. Sci. Eng. 135, 031019031019-7 (2013) 17. Kumar, M.S., Prabukarthi, A., Krishnaraj, V: Study on tool wear and chip formation during drilling carbon fiber reinforced polymer (CFRP)/titanium alloy (Ti6Al4V) stacks. Procedia Eng. 64, 582–592 (2013)

Chapter 11

A Study of Parameters Affecting Cutting Forces in Minimum Quantity Lubrication-Assisted Cross-Peripheral Grinding of Alumina Ceramic A. V. Manu , V. G. Ladeesh

and R. Manu

Abstract Ceramic materials have a widespread application in optical, biomedical, and aerospace field because of the high mechanical strength, excellent corrosion, and wear resistance. But due to the high hardness and low-fracture toughness of ceramics, it is a great challenge to make damage-free and precise ceramic parts with conventional methods of machining. Cross-peripheral grinding (CPG) is an advanced method of machining where a rotating hollow diamond core tool with diamond grits impregnated or electroplated at the lateral and end face of the bottom portion promotes grinding action for material removal. In this study, minimum quantity lubrication (MQL)-assisted CPG experiments are carried out to find the effect of process parameters on the cutting force. Box-Behnken design (BBD) was selected to generate the experimental runs. Analysis of variance is performed, and the significant parameters which affect cutting force are found out. Keywords Cross-peripheral grinding · Minimum quantity lubrication · Cutting force · Ceramics machining · Analysis of variance

11.1 Introduction Ceramics are attractive for many applications because of their superior properties such as resistance to chemical degradation, low density, wear resistance, and high strength at elevated temperatures. Due to this sufficient mechanical strength and outstanding thermostability, ceramics find application in aerospace, optical, and biochemical field. But it is a challenge to make damage-free precision ceramic parts with conventional methods of machining due to their low-fracture toughness and high hardness. In this scenario, Cross-peripheral grinding (CPG) can be suggested using which precision machining can be done. Cross-peripheral grinding (CPG) is an advanced method of machining which can be used for the precise machining of A. V. Manu (B) · V. G. Ladeesh · R. Manu Manufacturing Technology Section, Department of Mechanical Engineering, National Institute of Technology Calicut, Calicut, Kerala 673601, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_11

143

144

A. V. Manu et al.

Fig. 11.1 Illustration of CPG

brittle ceramics. In CPG, a rotating hollow diamond core tool (refer Fig. 11.1) with diamond grits impregnated at the lateral and end face of the bottom portion promotes the grinding action for material removal. Cutting force is an important factor which affects the surface quality of the machined workpiece. Higher cutting forces induce chipping and subsurface cracks, and hence the cutting force plays an important role in the service life. For precision machining of ceramics, care should be taken that the force value cannot increase beyond a critical limit as it may lead to the formation of median cracks. In the brittle mode, material removal takes place by the propagation of cracks as seen in the indentation fracture mechanism. During unloading, the lateral cracks are generated, and hence these cracks cause fragmentation of the affected region or the plastic zone and they get chipped off. This cracks the bond between atoms and hence then material removal is made easy. More material removal takes place in brittle mode as the applied force is more and also the depth of indentation is high. But chance for failure of the workpiece is high as there is the formation of median cracks which causes strength degradation. Median cracks are those cracks which are formed in the direction perpendicular to the chip flow. CPG is a process where the material removal takes place by repeated scratching action of the diamond grains. Hence the total force acting will be the combined effect of the active number of diamond grits. One of the major factors in the grinding system of engineering ceramics is the coolant-lubricant application. The proper application of lubricant is essential for carrying out cross-peripheral grinding of ceramics. The chips generated must be removed from the grinding zone in order to reduce the friction and to enable the proper grinding mechanism. The lubricant supplied must penetrate the tool-workpiece contact zone for the effective removal of chips and to absorb the heat generated at the contact zone. Fluid flow through the grinding zone was analyzed by Guo et al. [1]. Their studies suggest that conventional cutting fluid application may not adequately penetrate the wheelworkpiece contact zone. The overall performance of CPG is heavily dependent on the mode of application of cutting fluid as it is a significant factor affecting the cutting

11 A Study of Parameters Affecting Cutting Forces …

145

Fig. 11.2 Loaded tool

force which is evident from the studies of Emami et al. [2]. In MQL machining, an oil-air spray called oil mist is delivered into the tool-work contact zone. MQL not only reduces the consumption of cutting fluids in machining but can also provide better lubrication. The number of diamond grits taking part in the grinding operation depends not only on the orientation of diamond grits on the periphery of the tool but also the loading of the tool which plays a serious role, affecting the cutting force. Loading is caused by the deposition of ceramic powder on the tool (Fig. 11.2) which increases friction and hinders the smooth grinding operation. Loading of the tool causes masking of the diamond grits preventing the smooth material removal. Emami et al. [3] conducted studies on minimum quantity lubrication in grinding process of zirconia engineering ceramics. He found that in MQL grinding, a better surface texture containing lower amount of surface defects were obtained. Also, there was a better lubricant penetration to the grinding contact zone. The mist not only gave a lubrication effect but also removed the zirconia dust efficiently. From the studies of Park et al. [4], it was identified that nozzle pressure played an important role in the size of lubricant droplets and the effectiveness to carry the lubricant into the machining zone. The effect of nozzle distance was also considered which was identified as an important factor affecting the ability of lubricant to carry away the grains formed. Their investigation pointed out that higher nozzle pressure provides more but small droplets. When nozzle distance is increased fewer droplets are deposited on the surface. The influence of lubricant mist parameters on grinding force and surface quality was closely analyzed by Tawakoli et al. [5]. They identified nozzle orientation and air pressure as the significant factors affecting the quality of the surface machined. They also reported that nozzle location is an important factor affecting the performance of MQL oil mist. Emami et al. [6] investigated the effects of liquid atomization and delivery parameters of minimum quantity lubrication on the grinding process of alumina engineering ceramics. Lubricant impinging velocity played a serious role as it affects lubrication efficiency. The impinging velocity must be more than the wheel rotation speed in order to break the air barrier which is

146

A. V. Manu et al.

formed around the tool. Optimal MQL delivery parameters were found out. Efficient lubrication using MQL spray can decrease the challenges existing in engineering ceramics grinding process by reducing grinding forces and surface roughness.

11.2 MQL-Assisted CPG From the literature review, it was clear that the mode of application of coolant played a significant role in the performance of cross-peripheral grinding. The advantage of MQL is that the consumption of lubricant is limited, and there is effective cooling of the tool-workpiece contact zone. Another advantage of atomized lubricant delivery is that the lubricant can penetrate the air barrier which is formed due to the highspeed rotation of the tool. In flood lubricant system, this penetration of the lubricant to the contact zone cannot be expected, moreover, the lubricant and the alumina dust from machining forms a paste which increases the friction at the contact zone while machining under the assistance of flood coolant system. Dry machining is not recommended for ceramics as the high temperature generated causes the failure of the tool. Hence MQL is the technique which is the best option available for the lubrication during CPG using diamond core tool.

11.2.1 MQL System Minimum quantity lubrication technique (MQL) is highly advantageous compared to other methods of lubrication. In this technique, metered quantity of lubricant is supplied in minute quantity hence the problem of coolant disposal which is associated with other methods of lubrication is solved. In the lubrication system (Fig. 11.3), the lubricant is atomized to 4–6 µm particle sizes; they form a mist and are coalesced at the point of application to lubricate the two surfaces under friction. The spraying action occurs in the jet spray nozzle block where the lubricant is atomized and until then it is kept separate in order to provide maximum cooling efficiency. Further, as Fig. 11.3 KENKO mist lubricant system

11 A Study of Parameters Affecting Cutting Forces …

147

the coolant expands, it gets more capacity to absorb heat from the source (tool-job interface) and whereas here along with lubricant, air is applied this helps out to blow chips during cutting operations which further avoid heat generation. This lubricant can be controlled through the controlling knob which is placed in the nozzle block.

11.2.2 Identification of Process Parameters Speed, feed, and depth of cut are the principal elements of machining. A schematic representation of nozzle-tool arrangement is shown in Fig. 11.4. The depth of cut is the distance that the tool bit moves into the workpiece from its surface. Feed rate is the distance by which the tool advances across the workpiece per minute. In CPG, depth of cut, feed, and the spindle speed determine the surface finish, power requirements, and material removal rate, and hence it is important to study their effects on the cutting force while performing CPG. From the literature, it was clear that the velocity of the lubricant mist and the orientation of the nozzle play an important role in the cutting force generated. For a specific MQL and nozzle design, the gas velocity at the nozzle exit varies only with the inlet gas pressure. The variation in the inlet gas pressure affects the gas flow rate and correspondingly gas velocity. Therefore, inlet pressure is identified as an important parameter which affects the efficiency of cutting. The nozzle orientation is also important as the impinging angle plays a crucial role in carrying away the alumina dust. Fig. 11.4 Tool-nozzle arrangement

148

A. V. Manu et al.

Fig. 11.5 Experimental setup

11.3 Experimental Setup and Procedure 11.3.1 Experimental Setup To study the effect of process parameters on cutting force in the feed direction, CPG is conducted where slots are machined on the workpiece. Experiments are conducted on CNC Vertical Machining Center (Agni BM45 TC24 4-axis VMC). The experimental setup (Fig. 11.5) consists of a numerical control machining system, a cutting force data acquisition system, and a minimum quantity lubrication system. The outer diameter of the diamond core tool used is 3.07 mm. The mesh size of diamond grits ranges between 50 and 60 µm. The maximum spindle speed of the machine is 6000 rpm, and a feed rate of 1–10,000 mm/min can be provided. The cutting force data is acquired using Kistler 9257B dynamometer and then processed with the aid of Dynoware software.

11.3.2 Design of Experiments Box-Behnken design (BBD) is selected for generating the experimental runs. BoxBehnken designs are a class of rotatable second-order designs based on three-level incomplete factorial designs. The advantage of BBD is that it has no combinations for which all factors are at their highest or lowest levels and hence avoiding the experiments which are to be performed under extreme conditions. Five factors at three levels are selected (Table 11.1), and hence a total of 46 experimental runs has to be conducted (40 base runs and six center points).

11 A Study of Parameters Affecting Cutting Forces …

149

Table 11.1 Process parameters Parameters

Unit

Level 1

Level 2

Level 3

Speed

rpm

500

1500

2500

Feed

mm/min

2

4

6

Depth of cut (DoC)

mm

0.4

0.6

0.8

Impinging angle

degree

15

30

45

Inlet pressure

kg/cm2

0.8

1

1.2

11.4 Results and Discussion 11.4.1 ANOVA for Cutting Force Analysis of variance is used to determine the significant factors and its contribution to response. Table 11.2 shows the ANOVA for cutting force. The regression model for the cutting force is obtained as Force = −1.73 − 0.007130 Speed + 14.82 Feed + 28.60 DoC + 1.115 Impinging Angle − 8.42 Inlet pressure − 1.633 Feed × Feed − 0.01560 Impinging Angle × Impinging Angle. Backward elimination method is used to get the regression model. It is seen that the main effects of depth of cut, speed, feed, impinging angle, and inlet pressure, and the squared effects of feed and impinging are significant. This can be confirmed from the P-values given in Table 11.2. From the P-value of lack of fit, it can be confirmed that the lack-of-fit is insignificant. Table 11.2 ANOVA for cutting force Source

Dof

Adj SS

Adj MS

F-value

P-value

Model

7

2196.94

313.848

21.47

0

Linear

5

1694.76

338.952

23.18

0

Speed

1

813.48

813.476

55.64

0

Feed

1

196.76

196.757

13.46

0

DoC

1

523.64

523.643

35.82

0

Impg. angle

1

115.54

115.541

7.9

0

Inlet pressure

1

45.34

45.343

3.1

0.046

Square

2

502.18

251.088

17.17

0

Feed × Feed

1

435.44

435.466

29.78

0

Impg. angle × Impg. angle

1

125.6

125.599

8.59

0.006

Error

38

555.56

14.62

Lack of fit

33

520.1

15.761

2.22

0.19

Pure error

5

35.46

7.091

Total

45

2752.49

150

A. V. Manu et al.

Fig. 11.6 Main effects plot for cutting force

11.4.2 Main Effects Plot Main effects plot (refer Fig. 11.6) is used to study the effect of process parameters on cutting force. It is seen that the cutting force value decreases as the spindle speed increases. But analyzing the cutting force vs feed and cutting force vs depth of cut graphs, it is seen that the cutting force value increases with increase in feed and depth of cut values. This may be because the number of active diamond grits increases with an increase in feed and depth of cut. Hence the mean cutting force also increases. With the increase in inlet pressure, it is seen that the cutting force decreases, this is because the efficiency with which the chips are removed increases and hence there is no loading of the tool. Analyzing the effects of the impinging angle, it is seen that the impinging angle of 15° gave the minimum cutting force value which confirms that the effectiveness of flushing away the alumina dust as well as the lubrication efficiency was maximum at this impinging angle. Thus we can confirm that nozzle orientation plays an important role in the performance of cross-peripheral grinding with the aid of minimum quantity lubrication.

11.4.3 Tool Surface Morphology While performing CPG with the aid of flood cooling system, the disadvantage spotted was the loading of the tool. The alumina dust formed a paste with the coolant, and it formed a coating on the tool which prevented the efficient grinding of alumina by the diamond grits. The usage of MQL shows that there is no loading (Fig. 11.7) of the tool happening during machining. The diamond grits are free of alumina dust (Fig. 11.8), this is because of the gas which penetrates the air jacket surrounding the rotating tool and flushing away the alumina dust at the contact zone.

11 A Study of Parameters Affecting Cutting Forces …

151

Fig. 11.7 SEM image of tool after machining

Fig. 11.8 Diamond grits free of alumina dust

11.5 Conclusions Cross-peripheral grinding of alumina is performed with the assistance of minimum quantity lubrication. It is seen that the process parameters like speed, feed, depth of cut, impinging angle, and inlet pressure are significant factors affecting cutting force. Usage of MQL can be suggested as a possible solution to prevent loading of the tools. From the experiments, it is seen that the nozzle orientation plays an important role in flushing away the dust, which enhances the process. Further studies need to be conducted regarding the tool wear and the process performance.

152

A. V. Manu et al.

References 1. Guo, C., Malkin, S.: Analysis of fluid flow through grinding zone. ASME J. Eng. Ind. 114, 427–434 (1992) 2. Emami, M., Sahegi, M.H., Sarhan, A.A.D., Hasani, F.: Investigating the minimum quantity lubrication in grinding of Al2 O3 engineering ceramic. J. Clean. Prod. 66, 632–643 (2014) 3. Emami, M., Sadegi, M.H., Sarhan, A.: Minimum quantity lubrication in grinding process of zirconia engineering ceramic. Int. J. Min., Metall. Mech. Eng. 1, 4052–4060 (2013) 4. Park, K.H., Olortegui-Yume, J., Yoon, M.C., Kwon, P.: A study on droplets and their distribution for minimum quantity lubrication (MQL). Int. J. Mach. Tools Manuf. 50, 824–833 (2010) 5. Tawakoli, T., Hadad, M.J., Sadeghi, M.H.: Influence of oil mist parameters on minimum quantity lubrication—MQL grinding process. Int. J. Mach. Tools Manuf 50, 521–531 (2010) 6. Emami, M., Sadeghi, M.H., Ahmed, A.H.: Investigating the effects of liquid atomization and delivery parameters of minimum quantity lubrication on the grinding process of alumina engineering ceramic. J. Manuf. Process. 15, 74–388 (2013)

Chapter 12

Condition Monitoring of Abrasive Waterjet Milling Using Acoustic Emission and Cutting Force Signals U. Goutham , M. Kanthababu , S. Gowri, K. R. Sunilkumar , M. Mathanraj , J. John Rozario Jegaraj and R. Balasubramanian Abstract Abrasive waterjet machining (AWJM) is widely used in the industries with the distinct advantages of no thermal distortion, high flexibility, small cutting forces, etc. In this work, the influence of abrasive waterjet milling (AWJ milling) process parameters such as waterjet pressure (P), traverse rate (TR), abrasive flow rate (AFR) and standoff distance (SOD) on the depth of cut (DOC), material removal rate (MRR) and surface roughness (Ra ) were studied. An attempt has also been carried out to monitor the AWJ milling using two sensors namely, acoustic emission (AE) and cutting force dynamometer. The acquired AE sensor signal is analysed in the time- and frequency domains, whereas the cutting force signal is analysed in the time domain. The result indicated that AE and cutting force signals are influenced by the changes in the input process parameters conditions. Keywords Abrasive waterjet milling · Condition monitoring · Acoustic emission sensor · Cutting force dynamometer · Signal processing

12.1 Introduction AWJM is one of the promising unconventional machining methods employed for cutting difficult-to-cut materials. Apart from regular cutting, it is also used for turning, threading, slotting, milling, etc. If the depth of cut is controlled during AWJM, then it is known as AWJ milling. During AWJ milling, the high-pressure waterjet along with abrasives is not allowed to pass all the way to cut through the entire depth of the workpiece (not through cut). Therefore, controllable parameters to achieve the U. Goutham · M. Kanthababu (B) · S. Gowri · K. R. Sunilkumar · M. Mathanraj Department of Manufacturing Engineering, College of Engineering Guindy, Anna University, Chennai 600025, India e-mail: [email protected] J. John Rozario Jegaraj Defence Research and Development Laboratory, Hyderabad, India R. Balasubramanian Precision Engineering Division, Bhabha Atomic Research Center, Mumbai, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_12

153

154

U. Goutham et al.

desired depth of cut (DOC), material removal rate (MRR) and surface roughness (Ra ) have to be studied. In order to achieve the required DOC, MRR and Ra monitoring of the AWJ milling is also essential.

12.2 Literature Review Literature review related to the influence of AWJ milling process parameters is briefly presented here. The use of abrasive waterjet for milling was first reported by Hashish [1]. Hashish attempted to control the depth of the cut of isogrid structures using a mask and rotary table [2]. Ojmertz [3] introduced discretised AWJ milling to eliminate the problems in the manipulator dynamics. Some researchers have investigated the influence of standoff distance concerning MRR in AWJ milling [1, 4, 5]. They have reported that MRR is relatively low at higher standoff distance. Haghbin et al. [6] compared the performance of submerged and unsubmerged AWJ milling in 316L stainless steel and 6061-T6 aluminium with a different nozzle angle and standoff distance. They have observed that submerged AWJ milling resulted in narrow features than that of the unsubmerged AWJ milling. Momber and Kovacevic [7] found a linear decrease in the depth of cut with an increase in the standoff distance. Srinivasu et al. [8] found that the jet axial velocity decreased as the jet diverged with increasing standoff distance. Fowler et al. [9] have observed that the MRR is high at the lowest traverse rate and decreases rapidly with increase in the traverse rate. Shipway et al. [10] have also observed a similar trend. Hashish [1] suggested that a traverse rate of more than 1000 mm/min is recommended to achieve surface uniformity. Few authors have observed that the traverse rate has a significant influence on the surface roughness [5, 9]. Pal and Tandon [11] have investigated the machinability index and mechanical properties of the various materials such as Al 6061 alloy, Al 2024, brass 353, Ti6Al4V, AISI 304 and tool steel in AWJ milling. They have found that for low machinability indexed materials, lower traverse rate is required to produce pockets. Hashish [12] investigated the feasibility of AWJ milling in gamma titanium aluminide by using a mask. He has observed that there is a decrease in the depth of cut as the step over increases. Fowler et al. [13] investigated the effects of AWJ milling process parameters on the surface characteristics in titanium alloy (Ti6Al4V) with different abrasives, namely aluminium oxide, garnet, glass beads and steel shot. They have found that the ratio between the hardness of the workpiece and the abrasives is more important than particle shape. Also, it is found that MRR and Ra increase with an increase in particle hardness. Kanthababu et al. [14] observed that the step-over distance and the traverse rate are having a significant influence on the depth of cut during AWJ milling of Titanium alloy (Ti6Al4V). Literature review related to monitoring of AWJ milling is briefly presented here. Previous researchers have reported the utilisation of AE sensors for monitoring abrasive waterjet milling [13, 14]. Rabani et al. [15] introduced a concept called transfer rate of energy (TRE) to demonstrate the possibility of constant jet penetration by

12 Condition Monitoring of Abrasive Waterjet Milling …

155

adjusting the jet feed velocity. Axinte and Kong [16] have monitored the energy transferred through the workpiece during AWJ milling by integrated energy-based monitoring method. Kovacevic et al. [17] monitored the depth of cut by measuring the normal force acting on the workpiece and correlated with the depth of cut. From the literature review, it is observed that few works have been carried to study the influence of AWJ milling process parameters. There are only limited studies that have reported considering the influence of AE and cutting force signals during AWJ milling. Therefore, in this work, condition monitoring of AWJ milling is carried out using the signals acquired from the AE and cutting force dynamometer.

12.3 Experimental Set-up Experiments were carried out in stainless steel 304 (SS 304) plate of thickness 6 mm. AWJ milling is carried for a pocket-size of 20 mm × 10 mm using a AWJM system (Make: OMAX Corporation, Model: 2626). The photograph of the experimental work is shown in Fig. 12.1. AWJ milling input process parameters considered were waterjet pressure (P), traverse rate (TR), abrasive flow rate (AFR) and standoff distance (SOD). These process parameters were chosen at five different levels in order to study the influence of each process parameter on the responses such as depth of cut (DOC), material removal rate (MRR) and surface roughness (Ra ) (Table 12.1). The photograph of the AWJ milled pockets is shown in Fig. 12.2. The condition monitoring studies were carried out using two different sensors, namely AE sensor and cutting force dynamometer. AE signals were acquired using AE sensor (Make: Kistler; Model: 8152) with a sampling frequency of 2 MHz. The AE sensor was fixed on the top of the workpiece with a thin layer of water-resistant silica gel, which acts as a couplant, to avoid air gap between workpiece and sensors. A compatible piezotron coupler (Make: Kistler; Model: 5125B) was used to amplify the AE signals obtained from the AE sensor. After that, the amplified AE signals were converted into digital signals using the 16-bit multi-channel analogue-to-digital (A/D) conversion card (Make: National instruments, Model: PCI-6133). A Threeaxis (x–y–z) strain-gauge-based cutting force dynamometer (Make: Micro Mech Instruments) was placed below the workpiece to measure the normal force acting Fig. 12.1 Experimental set-up

Workpiece AE sensor Cutting Force dynamome-

156 Table 12.1 AWJ milling process parameters

U. Goutham et al. Abrasive material

Garnet

Abrasive mesh size

#85

Orifice material

Sapphire

Orifice diameter (mm)

0.25

Focusing nozzle length (mm)

101.6

Focusing nozzle diameter (mm)

0.75

Step over (mm)

0.3

Jet impact angle

90°

Waterjet pressure (MPa)

150, 165, 180, 195, 210

Traverse rate (mm/min)

1300, 1500, 1700, 1900, 2100

Abrasive flow rate (kg/min)

0.22, 0.32, 0.42, 0.52, 0.62

Standoff distance (mm)

13, 14.5, 16, 17.5, 19

Fig. 12.2 AWJ milling of stainless steel 304

on the workpiece. The cutting force signals converted into digital signals by data acquisition system (Make: National Instruments, Model: 9219). The sampling rate of cutting force signal is 100 Hz. Signal processing was carried out on the data collected from the AE sensor in time-domain and frequency domain using MATLAB R2010a. The first 65,536 data points from the AE sensor signals were used for the time-domain analysis, whereas the first 8192 data points were used for the frequency domain analysis. In the case of cutting force, only time-domain analysis was carried out.

12 Condition Monitoring of Abrasive Waterjet Milling …

157

12.4 Results and Discussion The following section deals with the influence of various AWJ milling process parameters such as waterjet pressure (P), traverse rate (TR), abrasive flow rate (AFR) and standoff distance (SOD) on the depth of cut (DOC), material removal rate (MRR) and surface roughness (Ra ). It also deals with the analysis of AE signals in timeand frequency domains with respect to variation in the process parameters. In timedomain analysis, root mean square (RMS) values were obtained. For the frequency domain analysis, the fast Fourier transform (FFT) technique was used to get the power spectra of the sensor signal.

12.4.1 Influence of Waterjet Pressure Figure 12.3 indicates the influence of waterjet pressure (P) on the depth of cut (DOC) and material removal rate (MRR). It is observed that depth of cut (DOC) and material removal rate (MRR) linearly increases with increase in the waterjet pressure. This indicates that as the waterjet pressure increases, the kinetic energy of the jet along with abrasive particles also increases and results in a higher DOC and hence higher MRR is observed. Figure 12.3 also shows that as the waterjet pressure increases, the surface roughness (Ra ) is also increasing. The increase in the waterjet pressure increases the abrasive particles fragmentation, which leads to decrease in the cutting ability of the jet. Therefore, the surface roughness (Ra ) values increase with increase in the waterjet pressure. Fig. 12.3 Pressure versus output response

158

U. Goutham et al.

Fig. 12.4 Traverse rate versus output response

12.4.2 Influence of Traverse Rate Figure 12.4 indicates the influence of traverse rate on the depth of cut (DOC) and material removal rate (MRR). It is observed that depth of cut (DOC) and material removal rate (MRR) linearly decrease with the increase in the traverse rate. The increase in the traverse rate results in the decreased jet exposure time on the workpiece, which leads to the reduction of the number of impact particles. Thus, there is a decrease in depth of cut (DOC) and material removal rate (MRR) with an increase in traverse rate. Figure 12.4 also illustrates the effect of traverse rate on surface roughness (Ra ). From Fig. 12.4, it is observed that as the traverse rate increases, decreasing trend of the surface roughness (Ra ) is observed.

12.4.3 Influence of Abrasive Flow Rate Figure 12.5 indicates the influence of abrasive flow rate (AFR) on the depth of cut (DOC) and material removal rate (MRR). From Fig. 12.5, it is observed that depth of cut (DOC) and material removal rate (MRR) increase up to 0.42 kg/min, afterwards, a decreasing tendency has been observed. In the case of surface roughness (Ra ), the surface roughness decreases up to 0.42 kg/min, after that increasing trend has been observed. As the abrasive flow rate increases, higher number of abrasive particles will be involved in the AWJ milling process. Therefore, the waterjet is distributed to a very high number of abrasive particles and leads to decrease in the kinetic energy of the jet, and as a result low DOC, low MRR and higher surface roughness (Ra ) are observed.

12 Condition Monitoring of Abrasive Waterjet Milling …

159

Fig. 12.5 AFR versus output response

12.4.4 Influence of Standoff Distance (SOD) The influence of standoff distance (SOD) on output parameters as shown in Fig. 12.6. From Fig. 12.6, it is observed that depth of cut (DOC) and material removal rate (MRR) decrease with an increase in the standoff distance (SOD). Figure 12.6 also shows that the surface roughness (Ra ) linearly decreases as the standoff distance (SOD) increases. This is due to the fact that as the standoff distance (SOD) increases, the jet loses its cohesiveness and leads to the decrease in the depth of cut (DOC), material removal rate (MRR) and surface roughness (Ra ). Fig. 12.6 SOD versus output response

160

U. Goutham et al.

12.4.5 Analysis of AE Signal Figure 12.7 shows that the time-domain analysis of the AERMS signals with the variation in the input process parameters. From the time-domain analysis (Fig. 12.7), it is observed that increasing trend of AERMS values is observed with increase in the pressure and decreasing trend is observed with increase in the traverse rate and standoff distance (SOD). In the case of abrasive flow rate, it decreases up to 0.42 kg/min and then it is found to be increasing. Figures 12.8, 12.9, 12.10 and 12.11 show the PSD plot of the AE sensor signals with respect to variation in the AWJ milling input process parameters. From Figs. 12.8, 12.9, 12.10 and 12.11, it is found that the dominant frequency range of all AE signal is found to be between 30 and 50 kHz. Figure 12.8 shows that the peak amplitude of the dominant frequency is increasing with respect to increasing in the waterjet pressure. Figure 12.9 shows that the peak amplitude (PSD) of the dominant

Fig. 12.7 Input parameters versus AERMS (V)

PSD (V2/Hz)

0.2 0.15 0.1 0.05 0 0

210

50

195

100

180

150

165 200

150

Frequency (kHz)

Fig. 12.8 Power spectra of the AE signal (waterjet pressure)

Pressure (MPa)

12 Condition Monitoring of Abrasive Waterjet Milling …

161

PSD (V2/Hz)

0.2 0.15 0.1 0.05 0 0 0.62

50

0.52

100

0.42

150

0.32 200

0.22

Abrasive flow rate (kg/min)

Frequency (kHz)

Fig. 12.9 Power spectra of the AE signal (abrasive flow rate)

PSD (V2/Hz)

0.1

0.05

0 0 2100

50

1900

100

1700

150

1500 200

1300

Frequency (kHz)

Traverse rate (mm/min)

Fig. 12.10 Power spectra of the AE signal (traverse rate)

frequency range. It is found to be decreasing with respect to the increase in the abrasive flow rate up to 0.42 kg/min after that it is found to be increasing. Figures 12.10 and 12.11 show the peak amplitude (PSD) of dominant frequency, and it is found to be decreasing with respect to increase in traverse rate and standoff distance (SOD).

12.4.6 Analysis of Cutting Force Signal Figure 12.12 shows the influence of waterjet input parameters on the normal cutting force. From Fig. 12.12, it is observed that the normal cutting force increases with an

162

U. Goutham et al.

PSD (V2/Hz)

0.2 0.15 0.1 0.05 0 0 19

50

17.5

100

16

150

14.5 200

13

Frequency (kHz)

Stand off distance (mm)

Fig. 12.11 Power spectra of the AE signal (standoff distance)

Fig. 12.12 Input parameters versus cutting force signal

increase in the water pressure. While with an increase in the traverse rate, the normal force acting on the workpiece is found to be decreased. However, in the case of abrasive flow rate, it increases with increase in the abrasive flow rate. As increasing in the standoff distance, the normal force acting on workpiece decreased. Kovacevic et al. [18] have observed the similar results.

12 Condition Monitoring of Abrasive Waterjet Milling …

163

12.5 Conclusions In this work, condition monitoring of AWJ milling using sensors such as AE and cutting force dynamometer was carried during AWJ milling of SS 304. The influencing of input parameters such as waterjet pressure (P), traverse rate (TR), abrasive flow rate (AFR) and standoff distance (SOD) was studied on the output responses such as depth of cut (DOC), material removal rate (MRR) and surface roughness (Ra ) and analysed. The acquired sensor signals were analysed in the time-domain and frequency domain with respect to input process parameters. Based on the experimentation and analysis, the following conclusions were drawn. The depth of cut (DOC) and material removal rate (MRR) increase with increase in pressure and decrease with increase in traverse rate and standoff distance. Whereas in the case of surface roughness (Ra ), it increases with waterjet pressure and decreases with traverse rate and standoff distance. DOC and MRR are increasing with AFR up to 0.42 kg/min, whereas with further increase in the AFR, they follow the decreasing trend. However, in the case of surface roughness, it follows opposite trend as that of DOC and MRR. From the time-domain analysis, it is observed that AERMS has good response with input process parameters. In case of frequency domain analysis, the dominant frequency of AE signals is found to be between 30 and 50 kHz. Cutting force sensor signal also follows trends with varying input process parameters. The acoustic emission signals and cutting force signals can be considered as a potential source for online monitoring of AWJ milling. Acknowledgements The authors would like to acknowledge the Anna University, Chennai, for providing the research fund under the scheme of Anna Centenary Research Fellowship (CFR/ACRF/2017/6). Also, the authors would like to acknowledge the financial support provided under Special Assistance Programme (SAP) by the University Grants Commission (UGC), New Delhi, India. (UGC Ref. No. F.3-41/2012 (SAPII) dated 01.11.2012).

References 1. Hashish, M.: Milling with abrasive waterjets: a preliminary investigation. In: Proceedings of 4th US Waterjet Conference, pp. 1–10. Berkeley, CA (1987) 2. Hashish, M.: Controlled-depth milling of isogrid structures with AWJs. J. Manuf. Sci. Eng. 120(1), 21–27 (1998) 3. Ojmertz, K.M.C.: Abrasive waterjet milling: an experimental investigation. In: Proceedings of 7th American Water Jet Conference, pp. 777–791. Seattle, USA (1993) 4. Laurinat, A, Louis, H, Meier-Wiechert, G.: A model for milling with abrasive water jets. In: Proceedings of 7th American Water Jet Conference, pp. 119–139. Seattle, WA (1993) 5. Ojmertz, C.: A Study on Abrasive Waterjet Milling. Chalmers University of Technology (1997) 6. Haghbin, N., Spelt, J.K., Papini, M.: Abrasive waterjet micro-machining of channels in metals: comparison between machining in air and submerged in water. Int. J. Mach. Tools Manuf. 88, 108–117 (2015). https://doi.org/10.1016/j.ijmachtools.2014.09.012 7. Momber, A.W., Kovacevic, R.: Principles of Abrasive Water Jet Machining. Springer Science & Business Media (2012)

164

U. Goutham et al.

8. Srinivasu, D.S., Axinte, D.A., Shipway, P.H., Folkes, J.: Influence of kinematic operating parameters on kerf geometry in abrasive waterjet machining of silicon carbide ceramics. Int. J. Mach. Tools Manuf. 49(14), 1077–1088 (2009). https://doi.org/10.1016/j.ijmachtools.2009. 07.007 9. Fowler, G., Shipway, P.H., Pashby, I.R.: Abrasive water-jet controlled depth milling of Ti6Al4V alloy–an investigation of the role of jet–workpiece traverse speed and abrasive grit size on the characteristics of the milled material. J. Mater. Process. Technol. 161(3), 407–414 (2005). https://doi.org/10.1016/j.jmatprotec.2004.07.069 10. Shipway, P.H., Fowler, G., Pashby, I.R.: Characteristics of the surface of a titanium alloy following milling with abrasive waterjets. Wear 258(1–4), 123–132 (2005). https://doi.org/10. 1016/j.wear.2004.04.005 11. Pal, V.K., Tandon, P.: Identification of the role of machinability and milling depth on machining time in controlled depth milling using abrasive water jet. The Int. J. Adv. Manuf. Technol. 66(5– 8), 877–881 (2013). https://doi.org/10.1007/s00170-012-4373-z 12. Hashish, M.: AWJ milling of gamma titanium aluminide. J. Manuf. Sci. Eng. 132(4), 041005– 041005-9 (2010). https://doi.org/10.1115/1.4001663 13. Fowler, G., Pashby, I.R., Shipway, P.H.: The effect of particle hardness and shape when abrasive water jet milling titanium alloy Ti6Al4V. Wear 266(7–8), 613–620 (2009). https://doi.org/10. 1016/j.wear.2008.06.013 14. Kanthababu, M., Ram, R.M., Emannuel, N.P., Gokul, R., Rammohan, R.: Experimental investigations on pocket milling of titanium alloy using abrasive water jet machining. FME Trans. 44(2), 133–138 (2016). https://doi.org/10.5937/fmet1602133K 15. Rabani, A., Marinescu, I., Axinte, D.: Acoustic emission energy transfer rate: a method for monitoring abrasive waterjet milling. Int. J. Mach. Tools Manuf. 61, 80–89 (2012). https://doi. org/10.1016/j.ijmachtools.2012.05.012 16. Axinte, D.A., Kong, M.C.: An integrated monitoring method to supervise waterjet machining. CIRP Ann. 58(1), 303–306 (2009). https://doi.org/10.1016/j.cirp.2009.03.022 17. Kovacevic, R.: Monitoring the depth of abrasive waterjet penetration. Int. J. Mach. Tools Manuf. 32(5), 725–736 (1992). https://doi.org/10.1016/0890-6955(92)90026-D 18. Kovacevic, R., Mohan, R., Zhang, Y.M.: Cutting force dynamics as a tool for surface profile monitoring in AWJ. J. Eng. Ind. 117(3), 340–350 (1995). https://doi.org/10.1115/1.2804339

Chapter 13

A Novel Small Quantity Lubrication Method to Assess Grindability of Inconel 718 Sirsendu Mahata , Manas Bhattacharyya , Bijoy Mandal and Santanu Das Abstract Inconel 718 is a nickel-based superalloy, widely used in high-temperature applications due to its high fatigue endurance and ability to retain mechanical and chemical properties at an elevated temperature. But it is an extremely difficult task to machine the material due to its high hardness and low thermal conductivity. As a consequence, grinding of Inconel 718 also poses severe problems such as higher force requirement and high heat generation. Surface quality, thereby, deteriorates and wheel life is also shortened. Use of conventional cutting fluids further increases the net manufacturing cost. In the present work, a novel method of application of a cheap detergent–water solution with the help of a micropump is highlighted and a comparative study is made with conventional cutting fluids in terms of force requirements, surface quality, specific energy, grinding ratio and chip form observations. Keywords Grinding · Inconel · Detergent water · Micropump · Surface roughness

13.1 Introduction Inconel 718 is one of the most widely used materials among the family of nickel-based superalloys, having vast applications in aerospace industries and high-temperature applications due to its high hot strength and good anti-corrosion properties. But high coefficient of thermal expansion, less thermal conductivity and high hardness at an elevated temperature restrict its machinability due to high force requirement and intense heat generation that ultimately result in poor surface quality of the product [1]. One of the most common techniques adapted for controlling grinding zone temperature is by flooding the zone with cutting fluid. However, the stiff air layer around the wheel formed due to centrifugal action restricts penetration of the fluid inside the grinding zone, thereby reducing the effectiveness of the method [2]. A meagre amount of fluid that reaches the grinding zone may be subjected to film S. Mahata · M. Bhattacharyya · B. Mandal · S. Das (B) Department of Mechanical Engineering, Kalyani Government Engineering College, Kalyani, Nadia, West Bengal 741235, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_13

165

166

S. Mahata et al.

boiling, which further increases the heat generated at the wheel–workpiece interface [3]. It is a well-known fact that lubricating properties of cutting fluids are substantially diminished if interface temperature reaches a certain limit [4]. To address the issue of stiff air, rexine-pasted wheel was used to enhance penetration of grinding fluid into the grinding zone for effective control of temperature. A pneumatic barrier was also employed along with a compound nozzle in order to break the stiff air layer for better fluid penetration. A compound nozzle along with a pneumatic barrier was used to break the stiff air layer for improved fluid penetration [5, 6]. Force and energy requirements along with surface roughness were found to reduce significantly using this technique. In flood cooling technique, there is a significant wastage of cutting fluid. In order to reduce the quantity of cutting fluid in grinding, minimum quantity lubrication (MQL) or small quantity lubrication (SQL) techniques have been given utmost importance of late. In one such work conducted by Babic et al. [7], a small amount of water was injected into compressed air to form a mist jet whereby better cooling effects were found to achieve than conventional cooling systems. Small quantity lubrication (SQL) technique, considered as an effective ‘green’ cooling technology, is accomplished by applying a quite small amount of fluid which can significantly reduce the environmental hazard along with the volume of fluid required [8]. A layer of grease applied on the surface of the specimen while grinding titanium has shown fairly good grindability [9]. Tawakoli et al. reported [10] that MQL technique, which is a near dry grinding method, can improve grinding performance in respect of grinding force, wheel wear and surface roughness, in comparison to dry grinding. A group of researchers used SQL method with less harmful alkaline soap water as coolant, on titanium grade 1 alloy [11] and on Inconel 718 alloy [12]. The results obtained are quite encouraging for both with regard to force requirement as well as surface finish. Mukhopadhyay et al., in another work [13], also found alkaline soap water to be quite effective in reducing grinding force and surface roughness when applied in SQL mode. Similar encouraging results were obtained by Mahata et al. while grinding AISI4340 steel using biodegradable edible soya bean oil, applied drop by drop onto the workpiece surface [14]. In the present work, a new SQL technique has been attempted, where cheap detergent solution is delivered at the wheel–work interface in the form of a microjet, with the help of a small pump. The observed results in terms of force requirements, surface quality, specific energy, grinding ratio and chip forms have been compared with those achieved using conventional cutting fluid.

13.2 Experimental Procedure In the present investigation, surface grinding of Inconel 718 was performed on a surface grinding machine of HMT made using alumina wheel, as shown in Fig. 13.1. Details of the equipment, tools and workpieces used are listed below in Table 13.1 and the chemical composition of Inconel 718 is given in Table 13.2.

13 A Novel Small Quantity Lubrication Method …

167

Fig. 13.1 Surface grinding machine used Table 13.1 Equipment details Surface grinding machine

Make: HMT Praga Division, India, Model: 452 P Infeed resolution: 1 µm Main motor power: 1.5 kW Maximum spindle speed: 2800 rpm

Grinding wheel

Make: Carborundum Universal Limited Type: disc, Dimension: F200 mm × 20 mm × F31.75 mm Specification: AA46/54K5V8

Workpiece

Material: Inconel 718 Dimension: 126 mm × 60 mm × 6 mm, Hardness: 85 HRB

Force indicator

Make: Sushma Grinding Dynamometer, Bengaluru Model: SA 115 Range: 0.1–100 kg Resolution: 0.1 kg

Force dynamometer

Make: Sushma Grinding Dynamometer, Bengaluru Model: SA 116 Range: 0.1–100 kg Resolution: 0.1 kg

Wheel dresser

Make: Solar, India Specification: 0.5 carat single point diamond tip Dressing depth: 20 µm Feed: 1.9 m/min

Table 13.2 Chemical composition of Inconel 718 alloy (in wt%) obtained by spectrometry Ni

Fe

Cr

Nb

Mo

Ti

Co

Al

Others

54.9

18.6

17.1

4.6

2.8

0.8

0.6

0.3

0.3

168

S. Mahata et al.

Fig. 13.2 Microjet with micropump arrangement

In this experimental work, grinding of Inconel 718 is performed in four different environmental conditions, viz. dry, flood cooling method, cooling using micropump with blassocut oil and cooling with detergent–water solution using micropump. Initially, dry grinding was performed for 20 up-grinding passes at 10 µm infeed. Then, flood cooling was employed for the next 20 passes, using blassocut oil as the cutting fluid. The cutting oil was delivered in the contact zone through a nozzle having an outer diameter of 6 mm with a flow rate of 210 ml/s. In the next set of experiments, a specially designed micronozzle (diameter 1.23 mm) as shown in Fig. 13.2 was used to supply the blassocut oil in the form of a jet which penetrated deep inside the grinding zone. A constant flow rate of 205 ml/min was maintained by a microsized pump of only 10 W power requirement. Finally, the same micronozzle was used to deliver the detergent solution, made using detergent (Sunlight) mixed with water in the ratio of 1:20 by weight, which was used as a cutting fluid. The dilution pH and the density of the detergent are 10.1 and 1.15 kg/l, respectively. A constant flow rate of 205 ml/min was maintained. Due to high velocity of the jet, it is likely that fluid breaks the stiff air layer easily and enters the grinding zone. The used coolant was filtered by the filter attached to the exit port and was disposed into the tank containing the micropump.

13.3 Results and Discussion Figure 13.3a, b represents the comparison of tangential force F t and normal force F n for four different environmental conditions, viz. dry, flood cooling application of blassocut oil, microjet delivery of blassocut oil and microjet delivery of detergent– water solution. From the graphs, it can be seen that tangential force is always lower than normal force which is quite natural. This may be mainly because of large negative rake angle of the grits which cause rubbing and ploughing to be the dominant form of

13 A Novel Small Quantity Lubrication Method … 0

(a)

Ft(N) Dry

10

169 0

20

Ft(N) Flood cooling

10

20

Ft(N) Detergent

Ft(N) Microjet blassocut

25

Grinding Force (N)

20

15

10

5 0

10

20

0

10

20

No. of passes 0

(b)

Fn(N)Dry

10

20

Fn(N) Flood cooling

0 Fn(N) Microjet blassocut

10

20

Fn(N) Detergent

Grinding Force (N)

100 80 60 40 20 0 0

10

20

0

10

20

No. of passes

Fig. 13.3 Comparison of a tangential force F t and b normal force F n for 20 passes under different conditions of environment

material removal rather than shearing. Comparing tangential and normal force under different environmental conditions, it is observed that during dry grinding, the force requirement is always high compared to any other conditions of cooling. Under dry condition, the grits develop wear flats rapidly due to intense friction and rubbing at the grit–workpiece interface in the absence of any cooling and lubrication, thereby increasing the force. On the other hand, minimum forces are observed in case of microjet delivery of detergent water. This may be because of better penetration of alkaline detergent water achieved due to the high velocity of liquid jet, which could easily break through the stiff air layer surrounding the grinding wheel and reach deep

170

S. Mahata et al.

inside the grinding zone. Forces are further reduced due to sufficient retention of grit sharpness caused by better cooling and lubrication effect provided by the coolant. Average surface roughness (Ra ) and maximum surface roughness (Rz ) were measured at six different locations by tracing the stylus of a portable surface roughness tester across the surface of the workpiece perpendicular to the grinding direction. Mean value of the average surface roughness and maximum surface roughness under different environmental conditions was calculated, and the same is shown graphically in Fig. 13.4. The highest average surface roughness is obtained in case of dry grinding obviously due to the absence of lubrication, while lowest surface roughness is observed using detergent water and blassocut oil delivered by microjet pump. This is due to better penetration achieved owing to higher jet velocity, along with the cooling and lubricating effect of the detergent as well as the oil used that reduces friction and temperature in the grinding zone. Quality of ground surface is not solely assessed by roughness but also by certain other surface defects which are given in Table 13.3. Few surface cracks are observed in dry and flood cooling condition. In dry grinding, high temperature produced due to absence of coolant makes the surface susceptible to defects in the form of microholes and cracks. Chip redeposition has been found to occur in microjet using blassocut

Fig. 13.4 Comparison of surface roughness of Inconel 718 after 20 passes under different environmental conditions

Table 13.3 Surface quality and defects under different environmental conditions Dry

Flood cooling

Microjet with blassocut oil

Microjet with detergent water

Surface burn

No

No

No

No

Chip redeposition

No

No

Yes

No

Surface cracks

Yes

Yes

No

No

Burr formation

Yes

Yes

Yes

No

13 A Novel Small Quantity Lubrication Method …

171

oil, while burr formation takes place in nearly all environmental conditions except with detergent–water solution. In this experimental work, grinding chips are collected during the 17th pass onwards for all the four environmental conditions. Collected chips are then observed under a stereo microscope and images obtained are shown in Fig. 13.5a through Fig. 13.5d. From the images, it can be seen that chips obtained in dry grinding are larger in size than other conditions and are of blocky type. The absence of coolant causes high heat generation that results in melting and subsequent amalgamation of loaded chips in the intergrit spaces to form larger chunks of blocky chips. On the other hand, in flood cooling, chip size has been found to reduce considerably, some curled chips are observed indicating favourable shearing action. Blassocut oil when delivered in the form of microjet gives a combination of both shearing curled/leafy type as well as blocky type chips. Flowing type chips are mostly observed when grinding Inconel 718 with microjet delivery of detergent–water solution. Such flowing type chips indicate smooth cutting action along with high surface quality of the ground workpiece, achieved due to better cooling and lubrication of the grinding zone. Figure 13.6a through Fig. 13.6d shows the ground surface characteristics when grinding Inconel 718 at 10 µm infeed under different environmental conditions. During dry grinding, absence of the coolant results in the formation of small microholes which may be caused by removal of material from the surface. Flood cooling using blassocut oil also results in some surface defects like cracks, probably due to improper penetration of cutting oil into the grinding zone, causing less effective lubrication. On the other hand, when the same oil is delivered in the form of microjet, no visible surface cracks are detected, although some chip redeposition occurs on the ground surface. Finally, the application of detergent water in the form of a microjet is found to render a smooth ground surface, devoid of any visible defects. Ground surface obtained in this condition shows long and uniform grinding (lay) marks indicating better performance than the other conditions. Grinding ratio is defined as the ratio of the work material removal rate to the wheel material removal rate. Hence, higher value of grinding ratio usually indicates better grindability. From Fig. 13.7, it is evident that grinding ratio obtained is the lowest in dry grinding than in other conditions. The reason behind this is the absence of coolant (which provides lubrication in the contact zone as well) which significantly increases the grinding force, and consequently wheel material removal also increases. Grinding ratio achieved is better in case of flood cooling and microjet delivery of blassocut oil than dry grinding. But the highest grinding ratio is obtained in case of grinding using micropump with detergent water compared to the other cooling techniques. This may be due to better penetration of cutting fluid into the grinding zone because of high velocity of the jet providing much better cooling and lubrication effect. Specific energy is the energy required per unit volume of material removed. It is quite high in grinding compared to other machining operations. It depends primarily on the tangential component of grinding force (F t ) and is calculated after every five passes for different environmental conditions using the following formula:

172

S. Mahata et al.

Fig. 13.5 Chip forms observed under a dry condition, b flood cooling, c microjet with blassocut oil and d microjet with detergent–water solution

13 A Novel Small Quantity Lubrication Method …

173

Fig. 13.6 Typical ground surface morphology observed under a dry condition, b flood cooling, c microjet with blassocut oil and d microjet with detergent–water solution

Fig. 13.7 Grinding ratio obtained after 20 passes under different environmental conditions

174

S. Mahata et al.

Fig. 13.8 Comparison of specific energy after every five passes under different environmental conditions

     Specific grinding energy = (Ft × Vwheel )/ b × d × Vworkpiece J/mm3 where Ft V wheel b d V workpiece

Tangential force (N) Velocity of grinding wheel (m/s) Width of workpiece (6 mm) Actual depth of material removed (µm) Table feed (m/s).

Variation of specific energy under different environmental conditions is shown in Fig. 13.8. From the graphs, it is seen that the overall energy requirement for first five passes is lower due to smaller infeed and elastic property of the wheel-work system. Specific energy requirement is found to be significantly high in case of dry grinding, which indicates higher rubbing and ploughing taking place in this condition due to the absence of lubrication. Grinding with microjet delivery of detergent water exhibited lowest specific energy as is evident from Fig. 13.8. It may be due to better cooling and lubrication effect of detergent–water solution, which may have penetrated well into the contact zone due to higher jet velocity, reducing friction and rubbing and consequently the tangential forces as well.

13.4 Conclusions A comparative study of grindability of Inconel 718 using alumina grinding wheel was made under four different environmental conditions at 10 µm infeed. Major conclusions are enlisted as under. 1. Grinding with microjet delivery of detergent–water solution exhibited the least grinding forces may be due to effective lubrication achieved by the high-speed jet of detergent–water solution in comparison with the other processes.

13 A Novel Small Quantity Lubrication Method …

175

2. Lowest surface roughness values are obtained in case of grinding with microjet delivery of detergent–water solution in the work. Favourable results are obtained under the same condition with respect to ground surface quality which is almost free from any major surface defects. 3. Highest G-ratio and lowest specific energy requirement are obtained in case of grinding with microjet delivery of detergent–water solution, which, therefore, indicates a good method of fluid delivery. 4. Microjet fluid delivery technique is favourable from economic point of view due to quite low power consumption (10 W only) of the micropump used. With regard to health and safety of machine tools and operator, the effort to find an alternative machining technique has been directed towards dry or near dry grinding. Detergent used in the experiment being safe and non-toxic is not at all hazardous to health and hygiene, and thereby creates a pollution-free environment, as against the toxic mineral oils commonly used as cutting fluid. Grinding with microjet delivery of detergent–water solution may be recommended as it is found to reduce grinding forces as well as produces better surface finish of the product.

References 1. Shokrani, A., Dhokia, V., Neman, S.T.: Environmentally conscious machining of difficult-tomachine materials with regard to cutting fluids. Int. J. Mach. Tools Manuf 57, 83–101 (2012) 2. Brinksmeier, E., Heinzel, C., Wittmann, M.: Friction, cooling and lubrication in grinding. CIRP Ann., Manuf. Technol. 48(2), 581–598 (1999) 3. Howes, T.D., Toenshoff, H.K., Heuer, W.: Environmental aspects of grinding fluids. Ann. CIRP 40(2), 623–629 (1991) 4. Cholakov, G.S., Guest, T.L., Rowe, G.W.: Lubricating properties of grinding fluids: comparison of fluids in surface grinding experiments. Lubr. Eng. 4812, 155–163 (1992) 5. Mandal, B., Singh, R., Das, S., Banerjee, S.: Improving grinding performance by controlling air flow around a grinding wheel. Int. J. Mach. Tools Manuf 51(9), 670–676 (2011) 6. Mandal, B., Das, G.C., Das, S., Banerjee, S.: Improving grinding fluid delivery using pneumatic barrier and compound nozzle. Prod. Eng.-Res. Dev. 8(1–2), 187–193 (2014) 7. Babic, D., Murray, D.B., Torrance, A.A.: Mist jet cooling of grinding processes. Int. J. Mach. Tools Manuf. 45(10), 1171–1177 (2005) 8. Sharma, S.V., Singh, G.R., Sorby, K.: A review on minimum quantity lubrication for machining processes. Mater. Manuf. Process. 30(8), 935–953 (2015) 9. Das, M., Mandal, B., Das, S.: An experimental investigation on grindability of titanium grade I under different environmental conditions. Manuf. Technol. Today 14(2), 3–10 (2015) 10. Tawakoli, T., Hadad, M.J., Sadeghi, M.H., Daneshi, A., Stöckert, S., Rasifard, A.: An experimental investigation of the effects of workpiece and grinding parameters on minimum quantity lubrication-MQL grinding. Int. J. Mach. Tools Manuf. 49(12–13), 924–932 (2009) 11. Kundu, A., Mukhopadhyay, M., Mahata, S., Banerjee, A., Mandal, B., Das, S.: Grinding Titanium grade 1 alloy with an alumina wheel using soap water. Procedia Manuf. 20, 338–343 (2018) 12. Mahata, S., Kundu, A., Mukhopadhyay, M., Banerjee, A., Mandal, B., Das, S.: Exploring grindability of Inconel 718 using small quantity cooling and lubrication technique. Mater. Today: Proc. 5(2), 4523–4530 (2018)

176

S. Mahata et al.

13. Mukhopadhyay, M., Kundu, P.K., Das, S.: Experimental investigation on enhancing grindability using alkaline based fluid for grinding Ti–6Al–4V. Mater. Manuf. Process. (2018). https://doi. org/10.1080/10426914.2018.1476759 14. Mahata, S., Samanta, A., Roy, J., Mandal, B., Das, S.: Influence of minimum quantity lubrication on grinding performance of annealed AISI 4340 steel. Indian Sci. Cruiser 31(2), 17–26 (2017)

Chapter 14

Performance of Carbon Nanotubes Based Cutting Oil on Turning of AISI 1040 Steel M. Amrita , P. Yogesh Chandra , P. Venkata Ramana , U. Shyam Sai and Chatti Sreeram

Abstract Manufacturing industries use a large quantity of cutting oils during production, which after usage are disposed of to environment, causing an imbalance in the ecosystem. In the view of reducing the utilization of cutting oil, sustainable methods of manufacturing are being explored. Minimum quantity lubrication (MQL) is a method where a minute quantity of coolant or lubricant is supplied precisely at the cutting zone, thereby providing necessary cooling and lubrication. To enhance the effectiveness of MQL application, base fluid with enhanced properties is to be used. Carbon nanotubes (CNTs) have excellent thermal conductivity as well as good lubricating properties. In the present work, CNTs are dispersed in base cutting fluid and applied in machining using minimum quantity lubrication. The effectiveness of CNT based cutting fluid on turning of AISI 1040 steel is evaluated at different cutting velocities and feeds. Response parameters like cutting temperature, cutting forces, and surface roughness are determined, and performance is compared with dry machining and MQL application of the regular cutting fluid. Chip morphology is also studied. Keywords Carbon nanotubes · Temperature · Machining forces · Surface roughness · Chip morphology

M. Amrita (B) · U. Shyam Sai · C. Sreeram Department of Mechanical Engineering, Gayatri Vidya Parishad College of Engineering (A), Visakhapatnam, India e-mail: [email protected] P. Yogesh Chandra System Engineer Trainee, TCS, Bangalore, India P. Venkata Ramana Tata Steel Processing and Distribution Ltd., Kalinganagar, Orissa, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_14

177

178

M. Amrita et al.

14.1 Introduction Flood machining is widely used in manufacturing industries where a large quantity of cutting fluid, used as a coolant is applied during the machining process. Annual consumption of coolant by a large automobile processing unit is approximately 2.28 million liters/year [1]. Annual coolant consumption in Germany is approximately 75,500 tons [1] and in Japan is 10,000 kl. These coolants have to be disposed of to the environment after usage. Most of the countries have classified cutting fluids under “hazardous wastes” [2] due to their impact on the environment and ecology. Disposal cost of coolant in Japan is ¥ 2250/l [3]. This has increased the percentage of the amount to be spent by manufacturing industries for coolant treatment and disposal. This led to the exploration of sustainable methods of manufacturing which are environment-friendly, economic and employee-friendly like the use of cryogenic tools, self-lubricating tools, and minimum quantity lubrication. In minimum quantity lubrication (MQL), a small quantity of cutting fluid of order 50–500 ml/h is utilized which is approximately three to four times lesser than the quantity used in flood machining [4]. Nanofluids are suspensions of nanoparticles like graphene, graphite, carbon nanotubes, metallic colloids of gold, silver, and copper in base fluids like ethylene glycol, water, etc. Dispersion of nanoparticles in the base fluid can augment the thermal and physical properties of the fluids as the thermal conductivity of nanoparticles is higher than the base fluid [5]. Therefore, nanoparticle enhanced cutting fluids have recently attracted the attention of researchers. Proper selection and use of nanoparticles can improve the performance of base cutting fluid by reducing the cutting temperature and cutting forces. Mia et al. [6] suggested the implementation of sustainable methods in a reduction in usage of cutting fluids, i.e., dry, minimum quantity lubrication spray system and solid lubricant with compressed air cooling system which would result in the reduction in coolant cost, operating cost, recycling cost and disposal and promote cleaner production. Behera et al. [7] carried out a comparative study of a high-pressure jet, cryogenic machining, and MQL with regular fluid and nanofluid cutting environments during machining of Inconel 718 superalloy. The parameters such as cutting force, flank wear, surface finish, surface defects, surface topography, and residual stresses were studied. Su et al. [8] investigated the effect of nanofluid MQL with vegetable-based oil and ester oil for turning of AISI 1040 steel and the properties such as viscosity, surface tension, wettability, and thermal conductivity is measured and found that graphite-LB2000 was the optimal base oil for graphite oil-based nanofluid MQL machining as it showed significant reduction in cutting force and temperature. Sharma et al. [9] tried to understand the mechanism in working of different nanoparticles mixed in fluids and their effects on different machining operations. High nanoparticle concentration yielded better surface finish and more lubrication due to rolling/sliding/filming and surface enhancement effect like mending and polishing of nanoparticles compared to dry machining and conventional cutting fluid machining. Prasad and Srikant [10] studied the performance of nanofluids at different weight percentages of nanographite powder, i.e., 0.1, 0.3, and 0.5% by weight

14 Performance of Carbon Nanotubes Based Cutting Oil …

179

at different flow rates and was found that an increase in percentage inclusion of the nanographite leads to better performance of the fluids in terms of properties and machining responses like cutting forces, temperatures, surface roughness, and tool wear. 0.3wt% nanographite particle inclusions at 15 ml/min flow rate were found to be the best combination after experimentation. Sayuti et al. [11] used a mist of SiO2 nanolubrication applied by air pressure in turning of hardened steel AISI 4140 and optimum SiO2 nanolubrication parameters to achieve correct lubrication conditions for the lowest tool wear, and best surface quality was investigated. It was found that nanolubricant concentration, nozzle angle, and air carrier pressure affect the lubrication properties. Srikant et al. [12] reviewed the available literature and examined nanofluids as potential candidates for minimum quantity lubrication which tells the prominence of MQL and the requirement of lubrication facilities of fluid used in this context. It also states that nanofluids are potential sources of such properties. Vereschaka et al. [13] showed that contact processes at cutting can be controlled by varying the structure, composition, and properties of coatings. Nanodispersed ceramic multilayered coatings on ceramic tool improved the machining quality and significantly increased its wear resistance. Sharma et al. [14] reviewed many research works on the application of different types of nanoparticle-dispersed cutting fluids in different machining processes. Nanofluids are widely being investigated for their effectiveness in application of MQL machining. Different nanoparticles with good thermal conductivity like Al2 O3 , CuO, and nanoparticles with excellent lubricating properties like MoS2 , boric acid, CF2, nanographite have being tested. Carbon nanotubes (CNTs) have an excellent thermal conductivity of order 2000 W/mK and also have excellent lubricating properties. Use of Arabic gum for dispersing CNT was not seen by authors in many papers studied. In the present work, CNT based cutting fluid formed using Arabic gum as a surfactant was applied to the machining of steels. To evaluate its thermal properties, the temperature was measured during machining, and to evaluate lubricating properties, cutting forces and surface roughness of the finished surface were measured. Present work aims to find the effectiveness of CNT dispersed cutting fluid applied as MQL in machining operations. Small-scale industries usually have machines with low horsepower. As per Boeing team in St. Louis, Missouri [15], a shop’s lower-rpm machining centers can realize much more of their potential productivity while working at higher feeds. Since most of the small-scale industries have 2–3 hp machining centers, the present work focuses on making use of maximum speed available in such machines and working at higher feeds to have appreciable metal removal rate. Under such conditions, the effectiveness of MQL application of CNT based cutting fluid over the MQL application of conventional cutting fluid and dry machining has been investigated.

180

M. Amrita et al.

14.2 Experimentation 14.2.1 Preparation of Nanofluids Nanofluids are prepared by dispersing carbon nanotubes in water-based cutting oil in three different trails at different proportions and weight percentages. Arabic gum is used as the surfactant. The use of it may reduce the thermal conductivity of carbon nanotubes dispersed cutting fluid, but it has to be used to improve the dispersion stability, as carbon nanotubes are hydrophobic in nature. In the first trial, 0.3% w/w of CNT is dispersed in water without adding any surfactant and is sonicated for 40 min in a bath sonicator. In the second trial, CNT is dispersed in water at 0.3% w/w and surfactant (Arabic gum) is added to it, at 0.1% w/w and sonicated for 40 min in a bath sonicator. Later, cutting oil is added to it such that the ratio of cutting fluid and water is 1:20. In the third trial, the concentration of surfactant is increased and strong method of sonication, i.e., probe sonicator, is used. CNT is dispersed in water at 0.3% w/w and surfactant (Arabic gum) added to it at 0.2% w/w and is sonicated for 40 min in an ultrasonic probe sonicator, and then, cutting oil is added to it in ratio 1:20.

14.2.2 Machining Tests To test the efficiency of CNT based cutting fluid in machining, it is supplied at a constant flow rate using MQL system. Air pressure is set at 2 kg/cm2 , and cutting fluid flow rate is maintained at 3 ml/min. Turning operation is performed at different cutting velocities of 44, 76, and 84 m/min and feed rates of 0.193, 0.304, and 0.386 mm/rev, keeping the depth of cut as constant at 0.25 mm. Cutting forces are measured by using strain gauge dynamometer, cutting temperatures are determined using IR thermometer Extech 42570, and Mitutoyo Surftest SJ301 is employed to measure the surface roughness. Emissivity for measuring cutting temperature is set as 0.31. Machining conditions are shown in Table 14.1, and the experimental setup is shown in Fig. 14.1. Table 14.1 Machining conditions

Cutting velocity (v) (m/min)

44

76

85

Feed (f) (mm/rev)

0.193

0.305

0.386

Depth of cut (mm)

0.25

Cooling conditions

Dry

Soluble oil (MQL)

CNT (MQL)

14 Performance of Carbon Nanotubes Based Cutting Oil …

181

Fig. 14.1 Experimental setup

14.3 Results and Discussions 14.3.1 Dispersion Stability of Nanofluids In the first trial, the carbon nanotubes dispersed in water at 0.3% w/w without any surfactant were not stable even for 2 h and got precipitated as shown in Fig. 14.2a. In the second trial, the carbon nanotubes dispersed in water at 0.3% w/w with the surfactant Arabic gum added to it at 0.1% w/w were stable after one day as shown in Fig. 14.2b. But the addition of cutting oil caused the dispersion to be unstable as shown in Fig. 14.2c. In the third trial, the carbon nanotubes were dispersed in water at 0.3% w/w with the surfactant Arabic gum added to it at 0.2% w/w and were sonicated for 40 min using ultrasonic probe sonicator. It was found that the solution was stable even after 48 h as shown in Fig. 14.2d. Addition of cutting oil also caused the dispersion to be stable even after 48 h as shown in Fig. 14.2e, and so, the experimentation was performed at these proportions of the mixture.

14.3.2 Machining Performance Cutting Forces. The components of resultant forces generated during machining, i.e., cutting force (F c ), feed force (F f ), and thrust forces (F t ), are recorded at different machining conditions and shown in Table 14.2. Lowest forces are shown in bold. Among all three components of machining forces, i.e., cutting forces (F c ), feed force (F t ), and thrust force (F t ), CNT based cutting fluid showed the best performance in ≈80% of the cases. Feed force is found to be the lowest compared to thrust and cutting force. Forces obtained with MQL application of soluble oil are found to be

182

M. Amrita et al.

Fig. 14.2 Dispersion stability of nanofluids

less than dry machining in almost all cases. This is due to the cooling and lubricating properties of the cutting fluid used as a coolant. Water and concentrate cutting fluid in ratio 20:1 have provided cooling and lubrication, even when applied in minute quantities. It could significantly remove the heat generated at the cutting zone as well as reduce the friction compared to dry machining. This could be achieved by reducing the usage of coolant by approximately 80% as compared to flood cooling. Addition of CNT in soluble oil has further decreased the cutting forces in most of the cases. This can be attributed to the presence of carbon nanotubes, which may have provided better cooling and lubrication. The cylindrical structure of CNT may have provided rolling action, thereby reducing cutting forces. Figure 14.3 shows the variation of all components of cutting forces with feed rates at different cutting velocities. It was found that at all cutting velocities, forces decreased with feed and then increased for dry machining and MQL application of conventional cutting fluid. But with CNT dispersed cutting fluid, forces initially increased and then became stable with a further increase in feed. At higher velocities, i.e., at 76 and 85 m/min, CNT dispersed cutting fluid performed well in reducing cutting forces. This may be due to the lubricating nature of CNT dispersed in cutting fluid. Figure 14.4 shows the variation of all components of cutting forces with cutting velocities at different feed rates. It was found that at all feed rates, forces increased with speed and then decreased. This trend is very clear at higher velocities. The low force at higher velocities may be because at high cutting velocities, higher

255

0.386

207

191

225

247

208

240

221

215

251

189

181

147

229

184

165

223

219

161

Numbers in bold represents lowest forces obtained in each machining condition

213

259

0.193

85

331

0.386

0.305

247

247

76

0.193

0.305

294

0.386

272 210

44

0.193

26

18

23

43

38

50

45

53

45

F f (N) Dry

CNT

Dry

Soluble oil

F c (N)

0.305

Cutting velocity (m/min)

Feed (mm/rev)

Table 14.2 Cutting forces generated during machining

14

13

14

10

21

14

28

39

28

Soluble oil

10

6

3

11

15

11

4

40

13

CNT

F t (N)

278

205

174

351

216

266

294

225

294

Dry

217

183

153

236

197

130

227

179

227

Soluble oil

184

153

186

193

181

201

168

124

135

CNT

14 Performance of Carbon Nanotubes Based Cutting Oil … 183

184

M. Amrita et al.

Fig. 14.3 Variation of cutting force (F c ), feed force (F f ), and thrust force (F t ) with feed rates at different cutting velocities

Fig. 14.4 Variation of all components of force with cutting velocities at varying feed rates

temperatures are generated which cause plastic deformation of metals, which may have caused easy shearing action of metal chips from the workpiece. Cutting Temperatures. The tooltip temperature is detected by the infrared thermometer. The variations of cutting temperature with respect to time at different cutting velocities and feed rates are plotted in Fig. 14.5. At cutting velocity of 44 m/min and feed of 0.193 mm/rev, the cutting temperature was found to be least for CNT dispersed cutting oil for more than half of the machining time, and soluble oil machining has recorded lower temperature during the final stage of machining. Dry machining has recorded the maximum temperature during machining. At a cutting velocity of 44 m/min and feed of 0.305 and 0.386 mm/rev, the cutting temperature was found to be higher with dry machining followed by CNT dispersed cutting oil, and minimum temperature was observed with soluble oil machining. A similar trend was observed at medium cutting velocity as well as at high cutting velocity and also at varying feed rates. This shows that the application of conventional cutting fluid as MQL has provided sufficient cooling causing a decrease in temperature generated at the cutting zone. With the dispersion of CNT in cutting fluid, it was expected that the thermal

14 Performance of Carbon Nanotubes Based Cutting Oil …

185

Fig. 14.5 Variation of cutting temperature with respect to time at different cutting velocities and feed rate

conductivity of the cutting fluid would increase, due to the high thermal conductivity of CNTs and would reduce the cutting temperature compared to the conventional cutting fluid. But the results of cutting temperature show that CNT based cutting fluid could reduce the temperature compared to dry machining, but not to the extent provided by conventional cutting fluid. This may be due to the addition of a surfactant which is essential for stable dispersion of CNT in cutting fluid. Xie et al. [16] reported that the addition of surfactant in nanofluids improves the stability of CNT nanofluids, but its excess addition may hinder this improvement. Surface Roughness. The surface roughness is determined using Mitutoyo Surftest SJ301, and the mean roughness (Ra) is recorded and presented in Table 14.3. For all cutting velocities and feeds, the mean roughness (Ra) was found to be maximum with dry machining. With the addition of soluble oil as MQL, surface roughness was found to decrease. Application of soluble oil as MQL caused 1– 34% reduction in surface roughness with respect to dry machining. This shows that cutting oil has provided some lubrication which has caused a good finish of the workpiece. This is in accordance with the results obtained in cutting forces which showed reduction due to the application of coolant. With the application of CNT based

186

M. Amrita et al.

Table 14.3 Variation of surface roughness for all cutting conditions Cutting velocity (m/min) 44

76

85

Feed (mm/rev)

Surface roughness (µm)

% decrease w.r.t dry machining

Dry

Soluble oil

CNT

Soluble oil

0.193

4.228

3.996

2.892

5.49

31.60

0.305

6.899

6.484

4.142

6.02

39.96

CNT

0.386

5.498

5.228

4.817

4.91

12.39

0.193

6.478

6.416

6.044

0.96

6.70

0.305

6.599

6.467

6.323

2.00

4.18

0.386

7.226

6.159

6.044

14.77

16.36

0.193

3.482

3.181

2.31

8.64

33.66

0.305

4.007

3.523

3.194

12.08

20.29

0.386

6.278

4.167

4.08

33.63

35.01

cutting fluid, surface roughness was found to further reduce at all considered cases. The percentage decrease in surface roughness is found to be more with CNT dispersed cutting oil compared to conventional soluble oil with respect to dry machining. The percentage decrease in surface roughness with CNT dispersed cutting oil was found to be 4–39% with respect to dry machining. A similar trend has been seen with cutting forces. Reduction in surface roughness as well as cutting forces with CNT dispersed cutting oil reveals the lubricating nature due to the addition of CNTs in cutting oil. Chip Morphology. Table 14.4 shows the chips obtained in every trail under the three machining conditions. Blue colored chips indicate the high-temperature generation and low cooling rate. Golden colored chips indicate the medium rate of cooling. Table 14.4 Chip morphology

14 Performance of Carbon Nanotubes Based Cutting Oil …

187

Silver colored chips indicate the fast rate of cooling. Chip morphology also shows better cooling of cutting zone with MQL application compared to dry machining. Chips from soluble oil and CNT dispersed cutting oils showed an almost similar rate of cooling. Low speed and high feed, i.e., at v = 44 m/min and f = 0.305 mm/rev and 0.386 mm/rev chips obtained are of a discontinuous type. Blue colored chips were obtained with dry machining. This shows high temperature generated at the cutting zone, which is taken by the chips giving it a blue color. With MQL application of cutting oil as well as CNT dispersed cutting oil, chips obtained were of golden color, showing the reduced temperature at the cutting zone. This could be possible only by effective heat removal by the coolants from the cutting zone. Colors of chips at all cutting velocities and feed rates show that chips obtained from dry machining show higher temperature compared to that obtained from MQL application of conventional as well as CNT based coolant.

14.4 Conclusions Performance of carbon nanotubes based cutting oil was studied by comparing its application with dry machining and machining with soluble oil as cutting oil. These were the conclusions drawn from the present work. 1. It was observed that there is better dispersion stability for CNT dispersed cutting oil at a weight percentage of 0.3% w/w of CNT and 0.2% w/w of Arabic gum using probe sonicator. 2. The cutting forces are found to be lesser for CNT dispersed cutting oil followed by soluble oil and dry machining which indicates that CNT has provided better lubrication than that of soluble oil and dry machining. 3. The cutting temperature that was recorded was found to be higher under dry machining conditions. For the soluble oil and CNT dispersed cutting oil, the cutting temperature was either lower for soluble oil or almost equal for both of them. Addition of Arabic gum to CNT dispersed cutting oil might have reduced the thermal conductivity of CNT resulting in no significant change in cutting temperature of CNT dispersed cutting oil as compared to that of the soluble oil or higher temperature under CNT dispersed cutting oil machining conditions. 4. The surface roughness was found to be better for CNT dispersed cutting oil followed by soluble oil and dry machining due to better lubrication provided by CNT. 5. Chip morphology showed the high temperature with dry machining and an almost similar rate of cooling from soluble oil and CNT dispersed cutting oil.

188

M. Amrita et al.

References 1. Klocke, F., Eisenblaetter, G.: Dry cutting. Ann. CIRP 46(2), 519–526 (1997) 2. Astakhov, V.P.: Machining: Fundamentals and Recent Advances, Springer, Chap. 7., Ecological Machining: Near-Dry Machining (2008) 3. Feng, F.S., Hattori, M: Process Cost and Information Modeling for Dry Machining. NIST Publication (2000) 4. Ali, S.M., Dhar, N.R., Dey. S.K.: Effect of minimum quantity lubrication (MQL) on cutting performance in turning medium carbon steel by uncoated carbide insert at different speed-feed combinations. Adv. Prod. Eng. Manag. 6(3) (2011) 5. Wang, X.Q., Mujumdar, A.S.: Heat transfer characteristics of nanofluids: a review. Int. J. Therm. Sci. 46(1), 1–19 (2007) 6. Mia, M., Gupta, M.K., Singh, G., Królczyk, G., Pimenov, D.Y.: An approach to cleaner production for machining hardened steel using different cooling-lubrication conditions. J. Clean. Prod. 187, 1069–1081 (2018) 7. Behera, B.C., Alemayehu, H., Ghosh, S., Rao, P.V.: A comparative study of recent lubri-coolant strategies for turning of Ni-based superalloy. J. Manuf. Process. 30, 541–552 (2017) 8. Su, Y., Gong, L., Li, B., Liu, Z., Chen, D.: Performance evaluation of nanofluid MQL with vegetable-based oil and ester oil as base fluids in turning. The Int. J. Adv. Manuf. Technol. 83(9–12), 2083–2089 (2016) 9. Sharma, A.K., Tiwari, A.K., Dixit, A.R.: Mechanism of nanoparticles functioning and effects in machining processes: a review. Mater. Today: Proc. 2(4–5), 3539–3544 (2015) 10. Prasad, M.M.S., Srikant, R.R.: Performance evaluation of nano graphite inclusions in cutting fluids with MQL technique in turning of AISI 1040 steel. Int. J. Res. Eng. Technol. 2(11), 381–393 (2013) 11. Sayuti, M., Sarhan, A.A., Salem, F.: Novel uses of SiO2 nano-lubrication system in hard turning process of hardened steel AISI 4140 for less tool wear, surface roughness and oil consumption. J. Clean. Prod. 67, 265–276 (2014) 12. Srikant, R.R., Prasad, M.M.S., Amrita, M., Sitaramaraju, A.V., Krishna, P.V: Nanofluids as a potential solution for minimum quantity lubrication: a review. Proc. Inst. Mech. Eng., Part B: J. Eng. Manuf. 228(1), 3–20 (2014) 13. Vereschaka, A.S., Grigoriev, S.N., Tabakov, V.P., Sotova, E.S., Vereschaka, A.A., Kulikov, M.Y. Improving the efficiency of the cutting tool made of ceramic when machining hardened steel by applying nano-dispersed multi-layered coatings. In: Key Engineering Materials, vol. 581, pp. 68–73. Trans Tech Publications (2014) 14. Sharma, A.K., Tiwari, A.K., Dixit, A.R.: Improved machining performance with nanoparticle enriched cutting fluids under minimum quantity lubrication (MQL) technique: a review. Mater. Today: Proc. 2(4–5), 3545–3551 (2015) 15. https://www.mmsonline.com/articles/when-spindle-speed-is-a-constraint 16. Xie, H., Yu, W., Li, Y., Chen, L.: Discussion on the thermal conductivity enhancement of nanofluids. Nanoscale Res. Lett. 6(1), 124 (2011)

Chapter 15

Investigations on the Influence of Serration Parameters on Cutting Forces P. Bari , P. Wahi

and M. Law

Abstract Serrated cutters, with their complex geometries, reduce cutting forces, and hence find use in the high-productivity milling of difficult-to-cut materials. Variations in serration profiles and parameters such as amplitudes, wavelengths and phase shifts influence the degree to which cutting forces can be reduced. In an effort to develop guidelines to preferentially design serrations to reduce cutting forces and improve productivity, this paper investigates the influence of serration parameters on cutting forces. A geometric force model explaining the mechanisms contributing to reducing forces is presented, and validated experimentally. Sensitivity analysis is then carried out on the validated model to understand the influence of serration parameters. It is observed that for a sinusoidal serration profile, for defined cutting conditions, the phase shift of the serrations between successive teeth contributes more to reducing forces than the serration amplitude and/or the wavelength. These results are instructive for the design of better serrated cutters. Keywords Serrated cutters · Serration parameters · Chip thickness · Cutting forces · Sensitivity analysis

15.1 Introduction Manufacturing of parts made from aluminum, nickel, and titanium based alloys used in the aerospace industries, often involves 80–90% of the bulk material being removed through machining. Meeting productivity goals for such parts requires aggressive cutting conditions with high depths-of-cuts at high speeds, which, due to the difficult-to-cut nature of these materials, remains challenging. Furthermore, increased process forces due to high-depths-of-cuts may lead to process-induced unstable chatter vibrations that may damage the tool, the part, and the machine. There is a need thus for strategies that reduce forces, avoid machining instabilities, P. Bari · P. Wahi · M. Law (B) Machine Tool Dynamics Laboratory, Department of Mechanical Engineering, Indian Institute of Technology Kanpur, Kanpur 208016, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_15

189

190

P. Bari et al.

Fig. 15.1 a Regular end mill, b serrated end mill

and that do not compromise productivity goals. Cutting with serrated cutters offers one such potential solution. Serrated cutters—an example of which is shown and contrasted with a regular end mill in Fig. 15.1, have complex geometries that can preferentially reduce process forces. Serrations can be sinusoidal, circular, trapezoidal, or even of the half-circular type, with each of these profiles having further variations in their amplitude, wavelength and phase shift of the serrations between successive teeth. Serrations change and reduce the local engagements of the tool with the part, and also reduce the contact length between the tool and part. These factors, combined with the complex geometries that contribute to producing a non-uniform and continuously varying chip thickness, help reduce the process forces. Because of their ability to reduce process forces, serrated cutters have found favour in the aerospace industries. However, there appears to be no method to their designs, with all cutting tool makers preferring different profiles and serration parameters. Furthermore, even though these serrated cutters have been around for decades, focused research to make them better has not entirely informed the designs of new cutters. Some classical early work on straight fluted serrated end mills was presented by Tlusty et al. [1], in which it was shown that due to the reduced contact between the tool and the workpiece on account of serrations, cutting forces reduce. Later, Campomanes [2] presented a detailed mechanics based model that included the influence of helix with sinusoidal serrations, and used an approximate chip thickness model [3]. Merdol and Altintas [4], and Dombovari et al. [5] reported modelling of cylindrical and tapered serrated end mills with a generalized representation of serration profiles using cubic-splines. A generalized mathematical modelling approach with arbitrary tool geometry was also proposed by Kaymakci et al. [6], wherein variation of local tool angles along serrations was included. These studies [1–6] focused more on presenting models for serrated cutters, and less on the influence of serration parameters on cutting forces. Some recent work was reported on optimizing serration profiles by Koca and Budak [7], and by Tehranizadeh and Budak [8], however they did so only for a fixed cutting parameter set.

15 Investigations on the Influence of Serration …

191

In an effort to develop guidelines to preferentially design serrations to reduce cutting forces and improve productivity, this paper investigates the influence of serration parameters on cutting forces for two different operating conditions. We report results only for serrations of the sinusoidal kind. At first, a geometric model for serrated cutters is presented in Sect. 15.2 to help explain the mechanisms that contribute to reducing forces. Following which, the force model, explained in Sect. 15.3 is validated experimentally in Sect. 15.4. Sensitivity analysis is then carried out in Sect. 15.5 on the validated model to understand the influence of serration parameters. This is followed by the main conclusions that instruct the design of improved serrated cutters.

15.2 Geometric Model of Serrated Cutter This section describes the geometry of serrated cutters that lead to a preferential reduction in cutting forces. At first, local geometries are defined, followed by describing a generalized method to model serrations, followed by regenerative delays caused by serrations. We conclude this section with discussions on variations in local chip thickness due to serrations. Models described in this section are based on the classic work done by Merdol and Altintas [4], and by Dombovari et al. [5].

15.2.1 Modelling of Serration Profile A schematic, and a cross-sectional view of a sinusoidal serrated cutter is shown in Fig. 15.2. xyz coordinate frame is attached to the tool as shown. The cutter can have N number of flutes (teeth), but as an example, only three (i th , (i + 1)th and (i + l)th ) flutes are shown in Fig. 15.2. As there is a wavy surface along the flute of the serrated cutter, the local radius changes along the flute and the height. The local radius for the i th flute at the height z is defined as: Ri (z) =

D − Ri (z), 2

(15.1)

wherein D is shank diameter of the cutter, and Ri (z) is the variation in local radius for the sinusoidal serration profile:  2π z − ψi , Ri (z) = A − A sin λ cos η 

(15.2)

wherein A is the serration amplitude which is half of peak to peak serration height; λ is the wavelength; η is the helix angle; and ψi is the phase shift.

192

P. Bari et al.

Fig. 15.2 a Geometry of the serrated cutter, b cross-sectional view at height z. Figure is adapted and modified from [5]

The local radius for a representative four-fluted sinusoidal serrated cutter, with parameters as measured and listed in Table 15.1, is shown in Fig. 15.3 wherein radius varies with a peak value of 8 mm for an amplitude of 0.22 mm, and a wavelength of 1.62 mm. Table 15.1 Parameters of the serrated tool under consideration Serration profile

Serration amplitude

Wavelength

Shank diameter

Initial phase shift

Helix angle

Rake angle

No of flutes

Cutter type

Sinusoidal

A= 0.22 mm

λ = 1.62 mm

D = 16 mm

ψ= [107 199 236 326] degree

η = 20 degree

10 degree

N =4

Cylindrical Serrated end mill

Fig. 15.3 Variation of local radius along height

8

Radius[mm]

7.9 7.8 7.7 7.6 flute 1

7.5

0

0.5

flute 2

flute 3

1

j*dz[mm]

flute 4

1.5

2

15 Investigations on the Influence of Serration …

193

There is phase shift in the serration profile of each flute since the serrations start differently for different flutes. This creates a variation of local radius along the flute and height. The angular phase shift at the starting is given by: ψi =

i−1 

ϕ p,k

(15.3)

k=1

wherein ϕ p,k is the pitch angle between the k th and (k + 1)th flute, and ψ1 = 0. The angular position for the i th flute at height z, measured from the y axis in a clockwise direction, called the instantaneous radial immersion angle, is calculated as follows: ϕi (z, t) = t +

i−1 

ϕ p,k −

k=1

2z tan η , D

(15.4)

wherein  is the clockwise spindle speed (rad/s). For serrated cutters, there is also an axial immersion angle that is the angle between the z axis of the cutter and the normal vector n i (z) to the local flute tangent, as shown in the Fig. 15.2a. It is expressed by the derivative of the local radius for the i th flute at the height z as follows: cot κi (z) =

dRi (z) . dz

(15.5)

As discussed above, the local radius, the instantaneous radial immersion angle, and the axial immersion angle, all vary along height of the serrated cutter, and their variation is described next, in Sect. 15.2.2.

15.2.2 Variation of Tool Geometry Since geometry changes along the height, the serrated cutter is considered to be an assembly of small differential axial disc elements, and the variation in geometry is described across each element. Local tool angles such as the rake angle (γn,i (z)), the axial immersion angle (κi (z)), and the helix angle (ηi (z)) are calculated in the rake face coordinate frame—shown in Fig. 15.4. This rake face coordinate frame is defined in each element relative to tool reference coordinate frame that is fixed at the bottom end of the tool, using a transformation as in [6]. All local angles vary due to a change in the direction of the rake face frame across each element. As a representative example, variation of axial immersion angle along the serration profile is shown in Fig. 15.4c. In addition to the complex variations in the local geometries, cutting with serrated cutters can also lead to missed-cut effects due to delays between the cut surfaces being generated, as discussed next in Sects. 15.2.3 and 15.2.4.

194

P. Bari et al.

(b)

(a)

(c)

Fig. 15.4 a Schematic of complex serrated end mill, b rake face coordinate frame, c variation of axial immersion angle

15.2.3 Regenerative Multiple Delays Between the Serrated Flutes Due to the serration profile and changes in the local radius across the height of the cutter, it is possible that the cut surface generated previously, was made either by the same flute in the previous pass or by other flutes in same pass. This leads to multiple delays, where delay is the time elapsed between formation of the current surface being generated and the previous surface generated. If the current cut surface is made by the i th flute at time t, and the previous cut surface was made by (i + l)th flute at time t − τi,l (z, t), then at the same angular position one can rewrite the instantaneous radial immersion angle as (Ref. Fig. 15.2b):   ϕi (z, t) = ϕi+l z, t − τi,l (z, t) ,

(15.6)

wherein τi,l (z, t), the multiple delay term in serrated cutters is expressed by the ratio of effective pitch angle and spindle speed as follow: 1  ϕi,l (z, t) = ϕ p,(i+k)   k=1 l−1

τi,l (z, t) =

mod N (z, t).

(15.7)

This delay is used in the estimation of the local chip thickness as discussed next in Sect. 15.2.4.

15 Investigations on the Influence of Serration …

195

15.2.4 Local Chip Thickness Elemental geometric static chip thickness is defined as the local distance between previous and current cut surface in the direction of normal vector n i (z) of the flute as: N   h stg,ie (z, t) = min (Ri (z) − Ri+l (z)) + f i,l (z, t) sin φi (z, t) sin κi (z), l=1

(15.8)

wherein f i,l (z, t) is the corresponding feed motion during τi,l (z, t) and Ri (z) is calculated from Eqs. (15.1 and 15.2). This however, does not signify the actual physical chip thickness. Elemental physical static chip thickness is calculated by multiplying two screening functions as: h ist (z, t) = gi (z, t)h g,ie (z, t)

(15.9)

where gi (z, t) = gri,i (z, t)gh,i (z, t), wherein the screening function due to radial immersion is: gri,i (z, t) =

1 ifϕen ≤ (ϕi (z, t) mod 2π ) ≤ ϕex 0 otherwise

(15.10)

and the screening function due to missed cut effect is: gh,i (z, t) =

1 ifh stg,ie (z, t) ≥ 0 . 0 otherwise

(15.11)

When the geometric chip thickness becomes negative due to the inclusion of variation of local radius in chip thickness (Eq. 15.8) then those corresponding flute(s) do not cut the surface. This is called the missed cut effect which causes non-uniform chip thickness, and can also be understood from Fig. 15.5a. Variation in the chip thickness profile for the four-fluted sinusoidal serrated cutter under consideration with cutting conditions as in Case 1, listed in Table 15.2, is shown in Fig. 15.5b. Figure 15.5b shows that for regular end mill all four flutes are always in cutting condition. It also shows that at the tool tip of serrated cutter only the 3rd flute is sharing the chip load and the others are in rest i.e. no cutting. But at z = 2 mm (from the bottom) of the serrated cutter, only the 1st flute is sharing the load. This irregular chip distribution of serrated cutters produces periodic cutting forces with non-uniform amplitudes, which contributes to a reduction in instantaneous cutting forces, as explained in Sect. 15.3.

196

P. Bari et al. Regular Endmill

(b) 0.02 (a)

0.01

Chip Thickness[mm]

0

Serrated end mill at tool tip (z=0mm) 0.1

flute 1

flute 2

flute 3

flute 4

0.05 0

Serrated end mill at z=2 mm

0.1 0.05 0 0.05

0.055

0.06

0.065

0.07

0.075

time[sec]

Fig. 15.5 a Possible missed-cut effect of the serrated tool. Figure is adapted and modified from [5], b chip thickness distributions for four fluted regular and serrated end mill

Table 15.2 Operating conditions Case no

Depth of cut (mm)

Spindle speed (rpm)

Feed rate (mm/rev/tooth)

Radial immersion

1

2

2

6

8000

0.02

100% (slotting)

10,000

0.1

50% down milling

15.3 Force Model for Serrated Cutter All differential elemental forces are calculated in each element of the discretized serrated tool. The cutting forces for the i th flute at a height z are found in the radialtangential-and-axial, i.e. rta directions as shown in Fig. 15.6. Due to variation of the local tool geometry along the serrations, and along the axis of the tool, this rta frame changes its orientation—see Fig. 15.6b. Differential forces evaluated in the rta frame are given by:   dF r t a,i (z, t) = K c h ist (z, t) + K e

dz gi (z, t) sin κi (z)

(15.12)

 T   K c = K c h ist (z, t), vc,i (z), γn,i (z) = K rc K tc K ac

(15.13)

wherein primary cutting force coefficient vector is:

and the edge cutting force coefficient vector is:

15 Investigations on the Influence of Serration …

197

Fig. 15.6 a Differential forces acting on an infinitesimal dz axial segment on the i th edge, b variation of cutting force along flute and height. Figure is modified from [5]

 T   K e = K e vc,i (z), γn,i (z) = K re K te K ae

(15.14)

wherein v is the cutting speed, and shows that in addition to being dependent on the local rake angle and the chip thickness, the coefficients also depend on the speed. Differential forces evaluated in the rta frame are transformed to the fixed machine coordinate frame (xyz) using the orthogonal to oblique transformation method [9]. Transformed differential forces in the xyz frame are given by: dF x yz,i (z, t) = T xr,i (z, t)dF r t a,i (z, t)

(15.15)

wherein the force transformation matrix is given by ⎤ − sin ϕi sinκi − cos ϕi − sin ϕi cosκi T xr,i (z, t) = ⎣ − cos ϕi sinκi sin ϕi − cos ϕi cosκi ⎦ cosκi 0 −sinκi ⎡

(15.16)

with ϕi := ϕi (z, t), κi := κi (z). The total lumped cutting force vector acting on the cutting tool in the x, y and z directions is calculated by integrating the differential force vector along flute and summing the contribution of all flutes:

198

P. Bari et al.

(a)

(b) 140 120

Force [N]

100 80 60 40

At steady state

20 Resultant Force (Serrated)

0 10.62

10.625

10.63

10.635

Resultant Force (Regular)

10.64

10.645

10.65

time[sec]

Fig. 15.7 a Schematic of apparent axial depth of cut for a four fluted serrated cutter. Figure is modified from [4], b comparison of cutting forces between serrated and regular end mill



F x yz (t) = Fx Fy Fz

T

N  

ap

=

dF x yz,i (z, t)

(15.17)

i=1 0

Due to the serrations, only a very small portion of the flute is actually in contact with the workpiece. The apparent axial edge contact length reduces as shown in Fig. 15.7a (a1 + a2 + a3 becomes a/N ). The highlighted white portion in Fig. 15.7a is the out of cut portion (negative contact), and the shaded grey portion is the actual cutting portion (positive contact). Hence in white portions, cutting forces are zero and less than regular end mills, whereas in the grey portions, cutting forces are more than regular end mills. When the white portions dominate over grey portions, as they do with serrated cutters, overall cutting forces reduce. This reduction in forces is calculated by substituting h ist (z, t) from Eq. (15.9) into Eq. (15.12) and finally solving Eq. (15.17) which is demonstrated through simulation results in Fig. 15.7b, which shows the resultant forces in the x–y plane for a regular end mill as well as for a serrated cutter with cutting conditions as in Case 1—listed in Table 15.2. The force model presented here is validated experimentally next in Sect. 15.4.

15.4 Experimental Validation of Cutting Forces The experimental setup to measure and validate the force model discussed in Sect. 15.3, is shown in Fig. 15.8. A three axis vertical milling machine was instrumented with a three component table top dynamometer, on which the workpiece was directly mounted. An Aluminium alloy, Al7075 was cut with cutting conditions as in Case 1—listed in Table 15.2. Serrated cutter parameters are as in Table 15.1.

15 Investigations on the Influence of Serration …

199

Fig. 15.8 Experimental setup

Al7075 was specifically chosen, since the geometry dependent cutting force coefficients (from Eq. 15.13 and 15.14) for the corresponding tool and workpiece (Al7075) are available for this material, as given in [4, 5]. The comparison between predicted and experimental resultant cutting forces in the x-y plane is shown in Fig. 15.9. While the mean value for measured and modelled forces match reasonably well, the measurements do not reproduce the intermittency of the modelled forces. This could potentially be attributed to the low pass filter of the control unit of dynamometer. The cut-off frequency of the built in filter within the control unit at 200  Hz is too  Hz = 533.3 Hz as in Case low with respect to tooth passing frequency = 8000×4 60 1—listed in Table 15.2. Due to low pass filter actual signal is bypassed and local peaks are not observed in the simulated force profile. Treating the force model to be validated, sensitivity of forces to changes in serration parameters is discussed in Sect. 15.5.

200

P. Bari et al. 100 Resultant Force (Simulated)

Resultant Force (Experimental)

Force [N]

80

60

40

20 10.62

10.625

10.63

10.635

10.64

10.645

10.65

time[sec] Fig. 15.9 Comparison of measured and predicted forces for serrated end mill

15.5 Sensitivity of Cutting Forces to Changes in Serration Parameters Sensitivity analysis is carried out by increasing (doubling) and decreasing (halving) the serration amplitude and the wavelength, in turn, while keeping all other parameters fixed. Sensitivity to changes in the phase shift are also discussed. All analysis is done for two sets of cutting conditions as listed in Table 15.2. For all results, model predicted resultant forces in the x–y plane for the original serration parameters are compared with forces obtained with a change in parameters.

15.5.1 Influence of a Change in Serration Amplitude Keeping other parameters fixed, amplitude of the serrations are changed (increased and decreased) to investigate its influence on the predicted resultant cutting force shown in Fig. 15.10. For cutting with parameters of Case 1 (Fig. 15.10a), it is observed that the maximum resultant force corresponding to original amplitude (0.22 mm), double amplitude (0.44 mm) and half amplitude (0.11 mm) are ~87 N, ~90 N (~4% increase) and ~88 N (~1.3% increase). For Case 2 (Fig. 15.10b), corresponding forces are ~662 N, ~677 N (2.3% increase) and ~617 N (~7% decrease). As is clearly evident, the resultant force does not appear to be very sensitive to a change in the serration amplitude, for cutting with either of the cutting parameter set.

15 Investigations on the Influence of Serration … Fig. 15.10 Influence of a change in amplitude of serrations on resultant force a Case 1 b Case 2

100

201

(a) Original amplitude

double amplitude

half amplitude

Force [N]

80

60

40

20 10.62

800

10.625

10.63

10.635

10.64

10.645

10.65

time[sec]

(b) Original Amplitude

Double Amplitude

Half Amplitude

Force [N]

600

400

200

0 10.62

10.625

10.63

10.635

10.64

time[sec]

15.5.2 Influence of a Change in Serration Wavelength Influence of a change in serration wavelength as compared to the original wavelength can be understood by the comparisons of the predicted resultant cutting forces in Fig. 15.11. For Case 1 it is observed that the maximum resultant force corresponding to original wavelength (1.62 mm), double wavelength (3.24 mm) and half wavelength (0.81 mm) are ~87 N, ~106 N (~22% increase) and ~85 N (~2% decrease). For Case 2 corresponding forces are ~661 N, ~658 N (no appreciable change) and ~681 N (~3% increase). It may hence be concluded that for specific combinations of cutting parameters, doubling of the serration wavelength, may adversely increase the cutting forces, pointing to the fact that wavelength might be an important parameter while designing serrated cutters.

202

P. Bari et al.

Fig. 15.11 Influence of a change in wavelength on resultant force a Case 1 b Case 2

120

(a) Original wavelength

double wavelength

half wavelength

Force [N]

100

80

60

40

20 10.62

800

10.625

10.63

10.635

10.64

10.645

10.65

time[sec]

(b) Original Wavelength

Double Wavelength

Half Wavelength

Force [N]

600

400

200

0 10.62

10.625

10.63

10.635

10.64

time[sec]

15.5.3 Influence of a Change in Serration Phase Shift Influence of a change in serration phase shift as compared to the original phase shift can be understood by the comparisons of the predicted resultant cutting forces in Fig. 15.12. For cutting with the parameters of Case 1 (Table 15.2) it is observed that the maximum resultant force corresponding to the original phase shift ([107 199 236 326] degree), regular phase shift ([0 90 180 270] degree) and other phase shift ([270 90 180 0] degree) are ~87 N, ~75 N (~14% decrease) and ~72 N (~17% decrease). For Case 2, corresponding forces are ~661 N, ~557 N (~16% decrease) and ~661 N (~0.1% decrease). From this analysis it may be concluded that for specific combinations of cutting parameters, a change in phase shift can preferentially reduce cutting forces.

15 Investigations on the Influence of Serration … Fig. 15.12 Influence of a change in phase shift on resultant force a Case 1 b Case 2

203

(a) 100 Original phase shift

Regular phase shift

Other phase shift

Force [N]

80

60

40

20 10.62

10.625

10.63

10.635

10.64

10.645

10.65

time[sec]

(b) 800 Original phase shift

Regular phase shift

Other phase shift

Force [N]

600

400

200

0 10.62

10.625

10.63

10.635

10.64

time[sec]

From the sensitivity analysis, it is evident that forces appear to be consistently more sensitive to a change in phase shift as compared to a change in serration amplitude and/or wavelength—thus suggesting that phase shift is also an important factor to be considered while designing serrated cutters that could potentially result in improved reduction of cutting forces.

15.6 Conclusions This paper discussed the influence of serration parameters on cutting forces. For a serrated cutter with a sinusoidal serration pattern, an experimentally validated force model was subject to sensitivity analysis to explore the role of serration amplitude, wavelength, and the phase shift of the serrations between successive teeth. Sensitivity

204

P. Bari et al.

analysis was carried out with two different sets of cutting conditions. For the cases investigated, the amplitude of the serration was found to be an insignificant design parameter as we get maximum ~7% decrease in resultant force. It was also observed that the phase shift and the serration wavelength can significantly influence cutting forces, with forces either increasing (maximum ~22% increase due to wavelength) or reducing (maximum ~17% decrease due to phase shift) depending on the combination of cutting and serration parameters. These results clearly motivate the need for a well-designed optimization study that will include the influence of different types of serration profiles and their parameters for a wide range of cutting conditions—which forms part of the planned future work, with the ultimate aim of this study being to inform the design of next-generation of serrated cutters that preferentially reduce cutting forces.

References 1. Tlusty, J., et al.: Use of special milling cutters against chatter. In: Proceedings of NAMRC 11, SME, 1983 2. Campomanes, M.L.: Kinematics and dynamics of milling with roughing tools. In: Metal Cutting and High Speed Machining (2002) 3. Altintas, Y., Ber, A.: Manufacturing Automation: Metal Cutting Mechanics, Machine Tool Vibrations, and CNC Design (2001) 4. Merdol, S.D., Altintas, Y.: Mechanics and dynamics of serrated cylindrical and tapered end mills. J. Manuf. Sci. Eng. (2004) 5. Dombovari, Z., et al.: The effect of serration on mechanics and stability of milling cutters. Int. J. Mach. Tools Manuf. (2010) 6. Kaymakci, M., et al.: Unified cutting force model for turning, boring, drilling and milling operations. Int. J. Mach. Tools Manuf. (2012) 7. Koca, R., Budak, E.: Optimization of serrated end mills for reduced cutting energy and higher stability. Procedia CIRP (2013) 8. Tehranizadeh F., Budak, E.: Design of serrated end mills for improved productivity. Procedia CIRP (2017) 9. Budak, E., et al.: Prediction of milling force coefficients from orthogonal cutting data. J. Manuf. Sci. Eng. (1996)

Chapter 16

Effect of Minimum Quantity Lubrication on Tool Wear and Surface Integrity During Hard Turning of EN31 Steel Jitendra Kumar Verma , Gaurav Bartarya

and Jitendra Bhaskar

Abstract Use of cutting fluids during machining process possesses a significant threat to environmental as well as operator’s health. Application of minimum quantity lubrication (MQL) to the machining operation significantly reduces the volume of cutting fluid along with reduction in machining forces. MQL influences the surface integrity of the machined workpiece too. The present work examines the effect of MQL application on surface integrity issues like surface roughness and subsurface hardness along with the tool wear, during turning of hardened EN 31 steel using a coated tungsten carbide insert. It could be observed that cutting in MQL conditions provided a better surface finish and low flank wear when compared with machining without MQL. The study of the microhardness profile in the subsurface region also showed that the workpiece turned with MQL had a crust with lower hardness than a workpiece turned without MQL. This would result in higher fatigue life of the hard-turned component. Keywords Finish hard turning · Minimum quantity lubrication · Surface roughness · Flank wear · Microhardness profile

16.1 Introduction Hard turning is a process for finishing external surface of a rotational part of hardened steel with hardness more than 45 HRC using conventional machining approach. Apart from surface finish requirement, such components require excellent surface integrity for better fatigue life. Surface roughness is an important performance parameter for hard turning process [1], which is severely affected by the flank wear present on the cutting tool. Surface roughness adversely affects friction properties, fatigue strength, and corrosion resistance of the machined components. In actual practice, J. K. Verma · J. Bhaskar Harcourt Butler Technical University, Kanpur, UP 208002, India G. Bartarya (B) Indian Institute of Technology, Bhubaneswar, Odisha 752050, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_16

205

206

J. K. Verma et al.

surface roughness is affected by many factors such as cutting tool variables, cutting conditions of process, and workpiece variables. Tool variable includes tool material, tool geometry features such as rake angle, nose radius, etc. Cutting condition is specified by speed, feed, and depth of cut used in the process. Workpiece variables depend on mechanical and chemical properties of workpiece material [2]. Tool wear (Fig. 16.1) also affects other surface integrity issues like variation of microhardness along the cross-section and tensile residual stresses on the machine surface. Tungsten carbide tool inserts are highly popular for the machining of various hard grade metals and alloys such as various grades of steels, graphite, and white cast iron, etc. Past few decades have seen a great deal of advancement in the performance of these cutting tools. To improve the surface condition such as tool wear characteristics and surface hardness, cutting tools are often coated by different coating material such as TiC, TiN, titanium aluminum nitride and titanium carbon nitride using chemical vapor deposition (CVD) process [3]. Coating also helps to provide better lubrication at the chip/tool and tool/workpiece interfaces thereby reducing friction and temperatures at the cutting edge. Although these coating materials save insert from wearing out fast, they give away easily in tiring machining conditions such as prevailing during hard turning. A major issue in finish hard turning is sharp hardness variation in the subsurface zone that leads to brittleness of the crust material with respect to the core material [4]. This happens due to diffusion of carbide particles from the core toward crust due to temperature gradient. The tool wear during the machining increases the temperature in tool–workpiece interaction zone leading to microstructure change and hardness variations. Mostly, hard turning is done in dry conditions. Minimum quantity lubrication method, which is known to reduce tool flank wear rate, may be a worthy solution for the problems like tool wear, surface roughness, and subsurface microhardness variations [5]. The problems ranging from environmental issues to health, as well as manufacturing cost call for minimal use of cutting fluids during machining wherever possible. Fig. 16.1 Tool flank wear on a coated tungsten carbide tool

16 Effect of Minimum Quantity Lubrication on Tool Wear …

207

Minimal quantity lubrication (MQL) comprises of a setup to spray a small volume of cutting fluid (approximately 10–100 ml/h) using compressed air toward the machining zone, thus providing an alternative cooling method to do away with these hazards [6]. In many machining processes, MQL is the key to achieve near-dry machining successfully. For machining components under near-dry machining conditions, MQL needs to be understood as a system. It comprises many individual components like MQL media, feeding technology, parameter settings, tools, and machine tools; they all mutually affect the operational performance [7]. All elements and subsystems in a minimum quantity lubrication system need careful designing and coordination among them to achieve the expected results. The function of MQL is to reduce friction and adhesion between the tool, the chip, and the workpiece. Due to this, the amount of frictional heat is also reduced leading to reduction in the tool and the working temperature which helps in arresting the tool wear and workpiece microstructure changes. MQL application has proved to be effective over the conventional flooding method of lubrication [8]. The performance of cutting tools was evaluated through surface finish and tool life obtained with MQL cutting conditions and compared with the performance of these tools in dry and wet machining conditions. Cutting with MQL provided better machining performance over that without MQL. MQL has been applied to milling operations too [9]. Optimum conditions for higher surface finish, lower cutting force, and higher tool life were analyzed for semi-finish milling of Monel 400 metal using a single-insert end-milling cutter of 10 mm diameter. A cutting speed of 125 m/min, feed of 0.020 mm/tooth, and a depth of cut of 0.35 mm were recommended for producing best surface finish within the cutting parameter range selected using MQL. For lowest cutting forces, a cutting speed of 125 m/min, feed of 0.010 mm/tooth, and a depth of cut of 0.20 mm were recommended. Under MQL application, a cutting speed of 75 m/min, feed of 0.015 mm/tooth, and 0.35 mm depth of cut were found to be giving best tool life for the chosen cutting parameter range. Variation in subsurface hardness across the section of a hard-turned component has also been observed [10]. It remains maximum at subsurface, and then a softer inner region was formed which was followed by a harder core beneath. The white layer produced was around 15–25% harder than the core material. The use of MQL may be helpful in mitigating this issue. Researchers [11] have discussed many issues in hard turning which limits its applicability in industry. It was observed that white layer, tensile surface residual stresses, and subsurface hardness are the critical process performance issues that need to be carefully worked upon. An experimental study [12] showed that the nose wear on the tool increased with cutting time. A higher tool life could be recorded when machining with MQL. This is due to reduction in overall machining temperature which allows the cutting edge of the tool to remain in required shape and form for longer time than that in dry cutting. The conclusions from above research works formed a basis to study the effect of use of MQL on tool flank wear and surface finish. Tool wear, which severely affects

208

J. K. Verma et al.

the subsurface integrity, if it can be contained through use of MQL, it would be of great benefit for component’s fatigue performance as well as for environment.

16.2 Experimentation Details The work material selected for the study is EN-31 alloy steel. This steel has excellent hardenability and is basically known as bearing steel as it is primarily used in bearing production such as ball and roller bearings. It is also used to make camshafts, dies and punches, spinning tools, axels and gears, driving pinion and other link components used in automotive industries, energy products, and many other general mechanical engineering applications. A round bar of 30 mm diameter and approx. 150 mm length was chosen for the turning experiments. The chosen steel was hardened to approx. 60 HRC, using standard hardening procedure. The conditions prevailing during heat treatment are provided in Table 16.1. Cutting inserts employed for hard turning should have high compressive strength, high hot hardness, and a high resistance to abrasive wear along with chemical stability at elevated temperatures. Popular tool materials are tungsten carbide, ceramic, and cubic boron nitride (CBN). Tungsten carbide is a low-cost tool material but has very low tool life when machining hard materials. On the other hand, CBN exhibits best tool life for hard turning operation but is an expensive material. It was decided to use low-cost coated tungsten carbide inserts to prove their utility in hard turning under MQL conditions by achieving lower tool wear (and hence an enhanced tool life). A chamfered edge type cemented carbide insert (Type: CNMG 120408S-61) was selected to machine hardened EN31 steel bar as chamfered edge tools have a better tool life under hard turning conditions. A tool holder (Seco make) CCLNR2525M12 was employed. The turning of workpiece in dry and MQL conditions was conducted on micromatic CNC lathe (Simple turn 5075). Figure 16.2 shows the schematic diagram of the experimental setup. The range of cutting parameters, selected for the experiments, is provided in Table 16.2. Box–Behnken method was used to design the experiments and deciding about the number of experiments. This method proposes a three-level design for fitting the response surface. This design is formed using combination of 3k factorials along with incomplete block designs. It reduces total experimentation cost by reducing the quantity of workpiece and cutting edges required, as 15 sets for the present work. Table 16.1 Heat treatment procedure of EN31 steel

Elements

Objectives

Temperature for hardening

800–850 °C

Medium for quenching

Oil

Temperature for tempering

180–225 °C

Rockwell hardness

60–64 HRC

16 Effect of Minimum Quantity Lubrication on Tool Wear …

209

Fig. 16.2 Schematic diagram of the setup

Lubricant Tank

Valve

Gravity feed

Compressor

Valve

Nozzle

Work

Tool

feed

Table 16.2 Range of machining parameters selected

Machining parameters

Low level

Middle level

High level

Cutting speed ‘v’ (in m/min)

120

150

180

Feed ‘f ’ (in mm/rev)

0.08

0.12

0.16

Depth of cut ‘d’ (in mm)

0.1

0.15

0.2

Taylor Hobson made Surtronic 3P instrument was used to measure the surface finish produced. Three readings for surface roughness were taken on a selected location near the start of cut. Then the average of the three data points was taken as Ra value. This gives the surface finish produced when insert has just been pressed in action. Microhardness test was done using a microhardness tester. The sample was cut from the machined bar from the end where the machining started. A USB connected Dino-lite optical microscope, having magnification from 20× to 230×, with Dino-capture 2.0 software was used to measure the wear generated on the tool flank after each cutting test of equal volume removal.

16.3 Results and Discussions 16.3.1 Effect of MQL on Surface Finish Hard turning is essentially a finish machining process where the quality of the surface finish remains the most vital outcome of the process. Therefore, surface roughness

210

J. K. Verma et al.

was recorded to determine the finishing performance for hard turning under both the cutting conditions, i.e., with MQL and with dry cutting (without MQL). Figure 16.3 provides a better visualization of variation of surface roughness when machining with and without MQL. It was observed that surface roughness of machined part produced with MQL was better than that of the part machined without MQL. Thus, it may be concluded that use of MQL for hard turning of steel is certainly advantageous for finish purpose. The experimental results were also analyzed using response surface methodology through DataFit 9, a statistical analysis software. Regression analysis was used to develop a relationship between cutting parameters and the surface roughness of the work. The regression equation for surface roughness when machining was performed without MQL is given below Ra = 0.000128v2 + 150.26f 2 − 79.83d 2 − 0.002083f v − 81.25fd + 0.0816d v − 0.0390v − 20.12f + 19.55d + 2.67 (R2 : 77.25%)

(16.1)

The regression equation for surface roughness when machining was performed with MQL takes form as Ra = 0.000127v2 + 209.26f 2 − 98.16d 2 − 0.0062f v − 56.25fd + 0.065d v − 0.03511v − 37.56f + 25.85d + 2.57 (R2 : 96.89%)

(16.2)

Fig. 16.3 Comparison between surface roughnesses produced with MQL and without MQL

16 Effect of Minimum Quantity Lubrication on Tool Wear …

211

The higher R2 value indicates the better fit of statistical model with the data. The surface finish variation found with MQL for various tests followed a smoother pattern whereas the data for machining without MQL contained high variability. So, with MQL, the surface finish also becomes easily predictable. Figure 16.4a, b shows the variation in surface finish with cutting speed and feed for a constant 0.1 mm depth of cut using the regression formulae for machining without and with MQL, respectively. Surface roughness first starts decreasing with increase in feed for both cutting conditions. This behavior may be caused due to plowing of material present for small feeds, which results into rougher surface; as the feed increases, the plowing decreases thereby giving a better surface. As the feed is increased further, the surface roughness produced also starts increasing again. This is a general trend in conventional machining due to high feed marks and dragging of tool along the feed direction. This also may be due to localized thermal softening of workpiece region near cutting zone that results in tearing of the softer metal. Increasing cutting speed from 120 to 180 m/min also shows deterioration of surface finish which might be due to localized thermal softening of workpiece due to high heat accumulation at higher cutting speeds. The similar results were observed for other depth of cuts of 0.15 and 0.2 mm

16.3.2 Effect of MQL on Tool Flank Wear High wear on the tool causes higher machining forces and also higher surface roughness. It, thereby, reduces the usable tool life due to surface quality issues. Tool life depends upon the amount of tool wear occurred during machining. Therefore, it is imperative to find out the relation of cutting parameters to the tool flank wear. It enables to choose favorable cutting conditions to keep flank wear in control, thereby producing superior surface integrity. The equal volume of material was removed from each turning experiment to compare the flank wear occurred on the insert. Figure 16.5 compares flank wear produced during machining with and without applying MQL. The graph shows that for all the experimental runs, flank wear generated is smaller for machining with MQL than that generated during machining in dry condition. A regression analysis to study the dependency of flank wear on the prevailing cutting conditions was made. The regression equation for flank wear (VB) when machining was performed without MQL can be given as below: VB = 0.00005967v 2 − 5.338f 2 − 17.61d 2 − 0.00429f v + 17.375fd − 0.04833d v − 0.00286v + 6.30f + 10.037d − 0.6564 (R2 : 73.12%)

(16.3)

The regression equation for flank wear, when machining was performed with MQL, is given as: VB = 0.00008467v2 + 47.005f 2 − 25.0166d 2 − 0.00677f v − 5.7499fd

212

J. K. Verma et al.

(a) Without MQL

(b) With MQL Fig. 16.4 Surface roughness variation during machining with depth of cut = 0.1 mm

16 Effect of Minimum Quantity Lubrication on Tool Wear …

213

Fig. 16.5 Flank wear generated during hard turning with MQL and without MQL (dry)

− 0.02933d v − 0.01187v + 1.228f + 12.199d + 0.1090 (R2 : 96.89%) (16.4) The coefficient of multiple determinations R2 for Eq. (16.3) is 73.12% and coefficient of multiple determinations R2 for Eq. (16.4) is 96.89%. A higher R2 value is indicative of a better fit of the regression model with the data. The response surfaces are shown in Fig. 16.6a, b for cutting in dry and MQL conditions, respectively. The effect of cutting speed and feed on the flank wear at 0.1 mm constant depth of cut can be observed from the graphs. In dry hard turning, wear increases with increase in both speed and feed. But with use of MQL, flank wear values decrease for the same cutting conditions. At high feed of 0.16 mm/rev, the wear increases with cutting speed for dry conditions whereas for cutting with MQL, the trend is reverse it shows that with MQL, the tool gets cooled efficiently so thermal softening of tool is reduced, also at high cutting speeds, most of the heat produced during machining is carried away by the chips, so the thermal softening of the tool reduces considerably, further resulting in low tool wear values. Similar reason can be cited for the reduction in tool wear with feed at high level of cutting speed (180 m/min). For other depth of cuts of 0.15 and 0.2 mm, similar results were observed. The results show that with proper selection of cutting parameters and application of MQL, coated carbide tools might be used effectively machining hard metals, thereby reducing overall cost of the process by acting as worthy replacement of CBN and ceramic inserts.

214

J. K. Verma et al.

(a) Without MQL

(b) With MQL Fig. 16.6 Tool flank wear variation for machining with 0.1 mm depth of cut

16 Effect of Minimum Quantity Lubrication on Tool Wear …

215

16.3.3 Effect of MQL on Microhardness Variation Hardness variations across the cross-section of the workpiece were measured using Vickers microhardness tester. Multiple indents were taken at several different points. Microhardness was measured with 1 kgf weight beneath the top surface at 30 µm difference. Measurements were made at five positions. The microhardness tests were performed in the subsurface zone up to nearly 150 µm depth at various locations. Figure 16.7a, b shows the variation in hardness across the section when machined at vc = 120 m/min, f = 0.12 mm/rev, and d = 0.1 mm with dry (without MQL) condition and using MQL, respectively. It was observed that when machining was Fig. 16.7 Microhardness in subsurface region at vc = 120 m/min, f = 0.12 mm/rev, d = 0.1 mm

(a)Without MQL

(b) With MQL

216

J. K. Verma et al.

performed in dry condition, microhardness varied from machine surface to toward center of disk sample from 638 to 532 HV in the subsurface zone (Fig. 16.7a). On the contrary, when machining was performed with MQL, microhardness value varied from 575 to 503 HV (Fig. 16.7b). Hence, it is clear that using MQL to hard turning reduced the subsurface hardness up to 10% and also the overall hardness profile shifts to lower hardness levels. For the cutting tests at 120 m/min cutting speed, 0.12 mm feed, and maximum cutting depth of 0.2 mm, for machining without MQL, hardness profile from the circumference toward center of disk at 30 µm difference was found to be reducing from a high hardness value of 681 HV to a fairy low hardness value of 513 HV (Fig. 16.8a). On the contrary, when machining was performed with MQL, the hardness value varied from 583 HV near the machine surface to 526 HV (Fig. 16.8b) toward core. The surface hardness got reduced by around 15%. Hence, it is again evident that minimum quantity lubrication is extremely helpful in producing gentle hardness profile on a hard-turned steel component, which varies within a smaller range as compared to machining in dry condition. Also, on comparing graphs from Figs. 16.7 and 16.8, it could be seen that with an increase in the prevailing depth of cut value from 0.1 to 0.2 mm, the hardness value just below machined surface increased from 638 to 681 HV. It shows that the subsurface hardness value increases with an increment in prevailing depth of cut. It might be due to more chip load at higher depth of cuts, which results into more work hardening of the material culminating into greater hardness levels. Whereas, previous works done have shown the effect of MQL on surface finish and tool wear but they did not test the subsurface changes occurred due to use of MQL. The work shows that the not only surface finish gets better and wear reduces with use of MQL but also the subsurface hardness reduces thereby making the finished surface less brittle; hence, making it less susceptible to fatigue failure.

16.4 Conclusions Present work explores the scope of higher tool life of coated carbide tool inserts in terms of lower tool wear along with improvement in surface finish during finish hard turning of EN 31 steel in MQL environment. The following conclusions were made • It was observed that hard turning of steel with MQL provided lower surface roughness as compared to that achieved during hard turning without using MQL. • Tool experienced lesser wear while turning with MQL. • The subsurface microhardness profile also showed lower hardness values in the subsurface zone with MQL than that without using MQL. So, in general, it could be observed that minimum quantity lubrication performed impressively in enhancing the performance characteristics of hard turning process as it improved surface finish, reduced wear and also arrested the variation in subsurface hardness across the section and helped in creating a crust with low hardness. The work

16 Effect of Minimum Quantity Lubrication on Tool Wear …

217

Fig. 16.8 Microhardness in subsurface region at vc = 120 m/min, f = 0.12 mm/rev, d = 0.2 mm

(a)Without MQL

(b) With MQL

also showed that the coated tungsten carbide inserts performed well for hard turning with MQL. It could be concluded that carbide inserts might become economical alternative to the costly tool materials like ceramic and CBN if right combination of cutting parameters are chosen. Acknowledgements Optical microscope facility available in Manufacturing Science Laboratory of IIT Kanpur was used to measure the tool wear values.

218

J. K. Verma et al.

References 1. Heisel, U., Lul, Z.M.: Investigation of cooling and lubricating liquids. Prod. Eng. III, 23–26 (1993) 2. Chaudhari, R.G., Hashimoto, F.: Process controls for surface integrity generated by hard turning. In: 3rd CIRP Conference on Surface Integrity (CIRP CSI), vol. 45, pp. 15–18 (2016) 3. Das, S.R., Dhupal, D., Kumar, A.: Experimental investigation on cutting force and surface roughness in machining of hardened AISI 52100 steel using CBN tool. In: 5th International & 26th All India Manufacturing Technology, Design and Research Conference, pp. 336-1–336-6, IIT Guwahati (2014) 4. Boubekri, N., Shaikh, V.: Minimum quantity lubrication (MQL) in machining: benefits and drawbacks. J. Ind. Intell. Inf. 3(3), 205–209 (2015) 5. Vikram, K.C.H.R., Ramamoorthy, B.: Performance of coated tools during hard turning under minimum fluid application. J. Mater. Process. Technol. 185, 210–216 (2007) 6. Bensouilah, H., Aouici, H., Meddour, I., Yallese, M.A., Mabrouki, T., Girardin, F.: Performance of coated and uncoated mixed ceramic tools in hard turning process. Measurement 82, 1–18 (2016) 7. Dhar, N.R., Islam, M.W., Islam, S., Mithu, M.A.H.: The influence of minimum quantity of lubrication (MQL) on cutting temperature, chip and dimensional accuracy in turning AISI1040 steel. J. Mater. Process. Technol. 171(1), 93–99 (2006) 8. Kamata, Y., Obikawa, T.: High speed MQL finish-turning of Inconel 718 with different coated tools. J. Mater. Process. Technol. 192–193, 281–286 (2007) 9. Soman, A., Shanbhag, V., Kuppan, P.S.M.: Influence of MQL application on surface roughness, cutting force and tool wear for milling of Monel 400. Int. J. Mech. Prod. Eng. (IJMPE) 3(5), 2320–2092 (2015) 10. Bartarya, G., Choudhury, S.K.: Effect of tool wear on white layer thickness and subsurface hardness on hard turned EN31 steel. In: 5th International & 26th All India Manufacturing Technology, Design and Research Conference, pp. 854-1–854-6, IIT Guwahati (2014) 11. Bartarya, G., Choudhury, S.K.: State of the art in hard turning. Int. J. Mach. Tools Manuf. 53(1), 1–14 (2012) 12. Chinchanikar, S., Choudhury, S.K.: Hard turning using HiPIMS-coated carbide tools: wear behaviour under dry and minimum quantity lubrication (MQL). Measurement 55, 536–548 (2014)

Chapter 17

Temperature Profiling of Microwave–Metal Discharge Plasma Channel Using Image Processing Technique Anurag Singh

and Apurbba Kumar Sharma

Abstract This work pertains to determination of temperature profile of microwave– metal (MW–m) discharge plasma channel in a domestic microwave applicator of 2.45 GHz through image processing. Plasma channel was generated between a pointed thoriated tungsten cylindrical tool (diameter: 1 mm) and a thin (thickness: 0.6 mm) stainless steel (SS) sheet workpiece (tool-work gap ≈ 1–2 mm) placed on a refractory base for microwave drilling. Interaction of incoming microwaves with pointed tool generated plasma channel between tool and thin SS sheet workpiece. Images of plasma channel were obtained during microwave drilling and later processed using MATLAB to determine temperature profile of the plasma channel. An elaborate MATLAB code was written to obtain RGB (red, green, and blue) values of each pixel and subsequently fed these RGB values into a set of equations of wavelength range (to obtain wavelength of each pixel) and ultimately fed wavelength into Wien’s displacement law to determine temperature profile of entire plasma channel. The results indicate that the temperature of plasma channel was beyond 4100 K in the inner zone; however, observed temperature reached up to 3800 K in the immediate outer zone of the plasma channel. The experimental results revealed that the temperature of plasma channel formed between tool and workpiece were sufficient to melt and ablate the material from the targeted zone. The produced hole in the sheet was of maximum diameter 1.07 mm with entry and exit circularity approximately 85%. Results indicate considerable consistency between experimental and theoretical observation with respect to temperature profile of the plasma channel. Keywords Microwave drilling · Microwave–metal (MW–m) discharge plasma channel · Image processing

A. Singh (B) · A. K. Sharma Mechanical and Industrial Engineering Department, Indian Institute of Technology Roorkee, Roorkee 247667, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_17

219

220

A. Singh and A. K. Sharma

17.1 Introduction Microwave drilling (MWD) is a nascent non-traditional thermal energy-based process that utilizes microwave energy, directly or indirectly, to drill engineering materials. It was first attempted by Jerby and Dikhtiar [1] using a metallic concentrator. A metallic rod was used to concentrate incoming microwave energy for localized softening and subsequent insertion the same concentrator to shape boundary of the hole in various non-conductive materials [1–7]. While Jerby et al. went on to expand understanding of their own microwave drilling method, several other researchers looked for different method employing microwave hybrid heating (MHH) and microwave– metal discharge plasma channel to drill various materials, even few metallic materials, which was not possible with earlier method [8–15]. Microwave–metal discharge is basically an electrical discharge, generated by interaction of microwaves with metallic objects containing sharp tips, corners or other discontinuities. It is one of the most complex and inadequately understood microwave-material interaction phenomena. Limited, though motivating literatures are available dealing with thermal characterization of this discharge phenomenon and its associated effects, viz. heating, plasma and photocatalytic effect [16]. Heating effect was inferred remarkable (temperature > 1000 °C) from the melting of iron by Wang et al. [17] through an indirect calorimetric method. They also recognized several factors, viz. microwave power, irradiation time, metal amount, and atmosphere, influencing discharge intensity and consequent heating effect. Wang et al. [18] was the first to experimentally characterize microwave–metal discharge through spectrum acquisition and analysis. They reported that discharges are mainly concentrated in visible region along with some peaks in ultraviolet region. Many studies explored potential of this phenomenon in applications such as microwave-assisted pyrolysis, pollutant removal, synthesis, etc. [19–24], but none of them reported temperature profiling of microwave–metal discharge plasma channel. Image processing techniques are computer-based simple, versatile, and inexpensive tools used to enhance pictorial information of raw images [25]. One of the most common usages of image processing is facial matching for criminal identification employed by law enforcement agencies of developed countries [26]. It also find application in the realm of remote sensing, military reconnaissance, film industry, materials science, and many more [25]. Image processing techniques are suitable to characterize events/objects that are difficult-to-reach (e.g., celestial bodies’ temperature), handle (e.g., combustion flame temperature), hazardous, transient (discharge, ignition, and combustion analysis), miniscule (phase/grain size), etc., in nature. At present, contact type and non-contact type of instruments are mostly used for temperature measurement. Both are limited to temperature measurement of a particular strategic spot. In order to achieve temperature of all spots in case of contact type of instruments, numerous thermocouple can be used, which will make entire temperature measurement system very costly and high-maintenance. In case of noncontact type of instruments, requirement of scanning the entire target area will make

17 Temperature Profiling of Microwave–Metal Discharge …

221

the process slower. Infrared thermal camera can be used to obtain temperature distribution of the entire area, but they are extremely expensive. Thus, image processing of consumer-grade video camera provides a non-contact type cheaper alternative to determine temperature distribution merely by image acquisition and further processing of the image [27]. There are various cases where image processing techniques were used to determine fossil fuel flame temperature [28–31]. This work describes a non-contact type method to determine temperature profile of microwave–metal (MW–m) discharge plasma channel through image acquisition and processing. Temperature profiling will help to recognize temperature zones of MW–m discharge plasma channel. This temperature data can then be used for minimizing the heat-affected zone (HAZ), and hence in obtaining better machining accuracy during microwave drilling. Further, temperature profiling may help in optimizing the microwave power input for drilling different materials.

17.2 Experimental Setup and Image Acquisition The experimental setup for generating microwave–metal discharge plasma channel to carry out microwave drilling was prepared inside a domestic microwave applicator (Make: LG, Model: MS2021CW), as shown in schematic diagram of Fig. 17.1. It consists of a metallic holder fixed at the top of cavity to hold a pin vice, which ultimately holds the cylindrical pointed tool of diameter 1 mm. The cylindrical pointed tool, made of thoriated tungsten, is intentionally pointed at the tip. When the microwave power is switched on, interaction of incident microwaves with pointed end of the tool alters the distribution of free electrons such that a intensified electric field zone is formed at the tip of the tool. This intensified electric field is sufficient to cause electrical breakdown of surrounding air medium inside the microwave applicator. Electrical breakdown of air generates a plasma channel in the vicinity of the tool tip.

Fig. 17.1 Microwave drilling experimental setup along with image acquisition and image processing

222

A. Singh and A. K. Sharma

Microwave drilling experiments at 700 W of microwave power were performed for duration of 60–70 s inside the closed microwave applicator cavity in dark condition. Dark condition was maintained by completely covering the oven lamp in order to avoid radiation of light from light source other than plasma channel. Avoidance of light radiation from light source other than plasma channel is vital in order to ensure that recording device only captures color intensities generated due to plasma channel. A consumer-grade video camera (display resolution: 1920 × 1080) was used to record all the microwave drilling experiments. Video camera was held against the microwave applicator’s window door screen to record the experiments. Suitable images from the video recording of experiments were obtained. Later the images were properly cropped to obtain the images of plasma channel of resolution 50 pixels × 50 pixels.

17.3 Image Processing and Discussion Every digital image is made up of finite number of minute elements called pixel. Each pixel represents a specific color depending on their respective RGB values. The color of each pixel corresponds to intensity of light photon striking from the source. Earlier obtained images of plasma channel were transferred to MATLAB code, which successively executed following three functions: (a) obtained RGB values of each pixel of cropped plasma channel image, (b) fed RGB values into a set of equations of wavelength range [32], which are function of RGB values, to find wavelength of each pixel, and (c) Finally, fed wavelength of each pixel into following mathematical equation of Wien’s displacement law to convert pixel wavelength into pixel temperature of the plasma channel: λmax × T = 2898 µm K

(17.1)

where λmax is pixel wavelength obtained by image processing technique and T is pixel temperature of plasma channel image radiating at wavelength of λmax . As mentioned earlier, microwave drilling experiments were performed inside the closed microwave applicator cavity in dark condition and video recordings were performed through the microwave applicator’s window door screen. Window door screens are generally made of a perforated metallic sheet interposed between two (inner and outer) transparent sheets. They help to minimize microwave leakage but allow passage of visible radiation, so that we can safely observe the processing inside the cavity. Prior to recording of microwave drilling experiments and the plasma channel in domestic microwave applicator, imaging in microwave applicator and subsequent processing of candle flames were done with same consumer-grade video camera (image resolution: 16 MP) to determine shielding effect of window door

17 Temperature Profiling of Microwave–Metal Discharge …

223

screen on attenuation of light radiation and corresponding temperature measurement. Imaging of candle flame was done with open door and closed door in dark condition and later images were processed according to sequence mentioned in previous paragraph to determine their respective temperature profile (Fig. 17.2). Candle flame images with open door were found to be more intense than images with closed door. This observation implies that window door screen has some effect on light radiation attenuation. However, image processing of both candle flame images revealed that temperature attenuation due to window door screen is within 5%, which can be tolerated in high-temperature analysis. After establishing the tolerable attenuation in temperature due to window door screen, same recording/imaging device was used to record (with closed door) the plasma channel generated during microwave drilling inside the same multi-mode microwave applicator. A photographic view of microwave drilling setup is shown in Fig. 17.3a. Processing of plasma channel image (Fig. 17.3b) revealed a typical temperature profile, as shown in Fig. 17.3c. Temperature zones of plasma channel

Fig. 17.2 a Candle flame image with open door; b temperature profile of open door candle flame; c candle flame image with closed door; d temperature profile of closed door candle flame

224

A. Singh and A. K. Sharma

Fig. 17.3 a Photographic view of experimental setup; b image of plasma channel; c temperature profile of plasma channel

are similar to that obtained for candle flame. Temperature zones consist of two distinct zones (inner and outer) separated by a narrow boundary zone, as shown in Fig. 17.3c. The temperature profile shows maximum temperature beyond 4100 K in the inner zone of plasma channel. Temperature drops slightly to the range of 4050–4100 K toward boundary of plasma channel, whereas it drops rather steeply in the outer zone of plasma channel, reaching up to 3800 K. Steep drop of plasma channel temperature in outer zone can be attributed to heat transfer to surrounding environment. Temperature profile of plasma channel obtained from image processing was compared with the thermal state of the tool just after the experimental trials. It was observed that the high melting point (>3673 K) thoriated tungsten tool got significantly heated up to red-hot condition and consequently it got slightly worn out. This observation confirms that tool gets heated above the temperature of 3673 K, which is fairly close to the temperature of inner zone of the plasma channel (>4100 K) estimated through image processing technique. Therefore, it can be said that the experimental and theoretical observations show considerable consistency with respect to temperature of the inner zone of the plasma channel. Because of high temperature of MW–m discharge plasma channel, it successfully drilled through holes in thin SS sheet kept below the tool. Removal of material from the SS sheet is attributed to melting (≈1700 K, melting temperature of SS) and/or ablation (≈3100 K, evaporation temperature of SS) by high-temperature plasma channel. Partial ablation of SS sheet around the hole periphery confirms that temperature of boundary zone of plasma channel went beyond 3100 K, which is in agreement to temperature of boundary zone (4050–4100 K) indicated by image processing technique. Figure 17.4 shows typical profile of entry side of drilled hole in SS sheet, which contains considerable HAZ. Entry and exit circularity of holes obtained are above 85% and overcut is up to 0.07 mm. Thermal damage around the holes presents the scope for further improvement of quality of drilled holes.

17 Temperature Profiling of Microwave–Metal Discharge …

225

Fig. 17.4 Stereoscopic image (20X) of typical hole drilled in SS sheet

17.4 Conclusions Temperature profiling of microwave–metal discharge plasma channel generated inside multi-mode microwave applicator was carried out. Simultaneously, feasibility of microwave drilling of metallic materials was confirmed. Following are the major conclusions that can be drawn from this work: i.

Results obtained through image processing technique are found considerably consistent with the experimental observations. ii. Image processing technique based on RGB color model and Wien’s displacement law may prove to be a simple, inexpensive, fast, and non-contact tool for temperature measurement of inaccessible target in various other applications. iii. Shielding effect of window door screen can be eliminated by in situ recording of the microwave drilling experiments, which, in turn, would enhance accuracy of temperature profiling by image processing technique. iv. Besides qualitative verification of image processing results, other non-contact type of temperature measuring instrument can be employed to quantitatively verify the results of image processing technique.

References 1. Jerby, E., Dikhtiar, V.: U.S. Patent No. 6,114,676. U.S. Patent and Trademark Office, Washington, DC (2000)

226

A. Singh and A. K. Sharma

2. Jerby, E., Dikhtyar, V., Aktushev, O., Grosglick, U.: The microwave drill. Science 298(5593), 587–589 (2002) 3. Jerby, E., Dikhtyar, V., Aktushev, O.: Microwave drill for ceramics. Am. Ceram. Soc. Bull. 82(1), 35–37 (2003) 4. Jerby, E., Aktushev, O., Dikhtyar, V., Livshits, P., Anaton, A., Yacoby, T., et al.: Microwave drill applications for concrete, glass and silicon. In: 4th World Congress Microwave & RadioFrequency Applications, pp. 7–12, Austin (2004) 5. Jerby, E., Thompson, A.M.: Microwave drilling of ceramic thermal-barrier coatings. J. Am. Ceram. Soc. 87(2), 308–310 (2004) 6. Jerby, E., Dikhtyar, V.: Drilling into hard non-conductive materials by localized microwave radiation. In: Advances in Microwave and Radio Frequency Processing, pp. 687–694, Berlin (2006) 7. Eshet, Y., Mann, R.R., Anaton, A., Yacoby, T., Gefen, A., Jerby, E.: Microwave drilling of bones. IEEE Trans. Biomed. Eng. 53(6), 1174–1182 (2006) 8. Titto, J.G., Sharma, A.K., Kumar, P.: A feasibility study on drilling of metals through microwave heating. i-Manager’s J. Mech. Eng. 2(2), 1 9. George, T.J., Sharma, A.K., Kumar, P., Kumar, R., Das, S.: Microwave drilling: future possibilities and challenges based on experimental studies. In: International Conference on Emerging Trends in Manufacturing Technology, Kerala (2012) 10. Lautre, N.K., Sharma, A.K., Kumar, P.: Distortions in hole and tool during microwave drilling of perspex in a customized applicator. In: Microwave Symposium, pp. 1–3, Tampa (2014) 11. Lautre, N.K., Sharma, A.K., Kumar, P., Das, S.: Performance of different drill bits in microwave assisted drilling. Int. J. Mech. Eng. Robot. Res. 1(1), 22–29 (2014) 12. Lautre, N.K., Sharma, A.K., Kumar, P., Das, S.: Microwave drilling with Litz wire using a domestic applicator. Bonfring Int. J. Ind. Eng. Manage. Sci. 4(3), 125–131 (2014) 13. Lautre, N.K., Sharma, A.K., Kumar, P., Das, S.: A photoelasticity approach for characterization of defects in microwave drilling of soda lime glass. J. Mater. Process. Technol. 225, 151–161 (2015) 14. Lautre, N.K., Sharma, A.K., Pradeep, K., Das, S.: A simulation approach to material removal in microwave drilling of soda lime glass at 2.45 GHz. Appl. Phys. A 120(4), 1261–1274 (2015) 15. Lautre, N.K., Sharma, A.K., Das, S., Kumar, P.: On crack control strategy in near-field microwave drilling of soda lime glass using precursors. J. Therm. Sci. Eng. Appl. 7(4), 041001 (2015) 16. Sun, J., Wang, W., Yue, Q., Ma, C., Zhang, J., Zhao, X., Song, Z.: Review on microwave– metal discharges and their applications in energy and industrial processes. Appl. Energy 175, 141–157 (2016) 17. Wang, W., Liu, Z., Sun, J., Ma, Q., Ma, C., Zhang, Y.: Experimental study on the heating effects of microwave discharge caused by metals. AIChE J. 58(12), 3852–3857 (2012) 18. Wang, W., Fu, L., Sun, J., Grimes, S., Mao, Y., Zhao, X., Song, Z.: Experimental study of microwave-induced discharge and mechanism analysis based on spectrum acquisition. IEEE Trans. Plasma Sci. 45(8), 2235–2242 (2017) 19. Hussain, Z., Khan, K.M., Hussain, K.: Microwave–metal interaction pyrolysis of polystyrene. J. Anal. Appl. Pyrol. 89(1), 39–43 (2010) 20. Sun, J., Wang, W., Liu, Z., Ma, Q., Zhao, C., Ma, C.: Kinetic study of the pyrolysis of waste printed circuit boards subject to conventional and microwave heating. Energies 5(9), 3295–3306 (2012) 21. Sun, J., Wang, W.L., Ma, C.Y., Yue, Q.Y.: Study on the promotion effect of microwave-metal discharge on the microwave pyrolysis of electronic waste. Adv. Mater. Res. 1088, 843–847 (2015) 22. Al-Wakeel, H.B., Karim, Z.A., Al-Kayiem, H.H.: Optimizing electro-thermo Helds for soot oxidation using microwave heating and metal. IOP Conf. Ser. Mater. Sci. Eng. 78(1), 012018 (2015) 23. Whittaker, A.G., Mingos, D.M.P.: Arcing and other microwave characteristics of metal powders in liquid systems. J. Chem. Soc. 9, 1521–1526 (2000)

17 Temperature Profiling of Microwave–Metal Discharge …

227

24. Chen, W., Gutmann, B., Kappe, C.O.: Characterization of microwave-induced electric discharge phenomena in metal-solvent mixtures. Chem. Open 1(1), 39–48 (2012) 25. Chitradevi, B., Srimathi, P.: An overview on image processing techniques. Int. J. Innovative Res. Comput. Commun. Eng. 2(11), 6466–6472 (2014) 26. Abdullah, N.A., Saidi, M.J., Rahman, N.H.A., Wen, C.C., Hamid, I.R.A.: Face recognition for criminal identification: an implementation of principal component analysis for face recognition. AIP Conf. Proc. 1891(1), 020002 (2017) 27. Panditrao, A.M., Rege, P.P.: Estimation of the temperature of heat sources by digital photography and image processing. In: Instrumentation and Measurement Technology Conference, pp. 218–223, Singapore (2009) 28. Shimoda, M., Sugano, A., Kimura, T., Watanabe, Y., Ishiyama, K.: Prediction method of unburnt carbon for coal fired utility boiler using image processing technique of combustion flame. IEEE Trans. Energy Convers. 5(4), 640–645 (1990) 29. Jiang, F., Liu, S., Lu, G., Yan, Y., Wang, H., Song, Y., et al.: Experimental study on measurement of flame temperature distribution using the two-color method. J. Therm. Sci. 11(4), 378–382 (2002) 30. Lu, G., Yan, Y., Riley, G., Bheemul, H.C.: Concurrent measurement of temperature and soot concentration of pulverized coal flames. IEEE Trans. Instrum. Meas. 51(5), 990–995 (2002) 31. Jiang, Z.W., Luo, Z.X., Zhou, H.C.: A simple measurement method of temperature and emissivity of coal-fired flames from visible radiation image and its application in a CFB boiler furnace. Fuel 88(6), 980–987 (2009) 32. Kodal, A., Majumder, S., Deka, U.: Measurement of plasma parameters using digital image processing technique. In: International Symposium on Devices MEMS, Intelligent Systems & Communication, pp. 22–27, India (2011)

Chapter 18

Applicability of CaF2 Solid Lubricant-Assisted Minimum Quantity Lubrication in Turning for Sustainable Manufacturing Mayur A. Makhesana , K. M. Patel

and Anand S. Patel

Abstract The conventional cutting fluids have a significant share in the total production cost. It also causes environmental hazards and health damages if not managed properly. Minimization or elimination by some other by some efficient means is of current research focus. Through the experiments, this research work demonstrates the effectiveness and performance of solid lubricant-assisted minimum quantity lubrication in turning. Micron-sized calcium fluoride as solid lubricant is mixed with SAE 40 to prepare lubricant mixture. The output responses are analyzed by varying concentration of the solid lubricant in SAE 40 oil and the flow rate of the mixture. The output responses are measured in form of surface quality, chip-tool interface temperature, and tool flank wear. The comparative study of experimental results with dry, wet cooling, MQL, and MQSL has shown some motivating trends. Process performance is improved with the application of solid lubricant-assisted machining in form of the reduction in surface roughness, tool wear, and chip-tool interface temperature. From the results, the use of CaF2 as a solid lubricant with MQL can be considered as an environment-friendly and cost-effective alternative for lubrication. The results can be used for metal cutting industries, opening the possibilities of developing sustainable manufacturing practices. Keywords MQL · Flank wear · Chip-tool interface temperature · Flow rate

18.1 Introduction and Literature Review Machining is considered as the most versatile manufacturing process able to produce precise shape, size, and surface quality of work part. The relative motion between cutting tool and workpiece causes plastic deformation of material leading to the generation of heat during machining. Machining of ferrous materials results in faster tool wear due to higher machining temperature and also causes tool failure and poor M. A. Makhesana (B) · K. M. Patel · A. S. Patel Mechanical Engineering Department, Institute of Technology, Nirma University, Ahmedabad, Gujarat 382481, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_18

229

230

M. A. Makhesana et al.

surface integrity [1]. The blunt tool also results in excessive power consumption while machining. Hence, it is essential to reduce the heat generated in machining to improve surface finish and longer tool life. Conventional cutting fluid has been applied to machining area to reduce the friction coefficient between tool-work contact and therefore minimizing the heat generated. These conventional metal cutting fluids mostly have a negative environmental effects, and they cannot be considered for sustainable manufacturing. Besides the environmental risks, due to its poor waste disposal, they cause severe health hazard even within the factory [2]. Looking at the breakdown of total production cost, the cost of cutting fluids takes a larger part in the same. And hence, the use of these cutting fluid cannot be justified on an economic point of machining [3]. The review of literature has provided an insight into the emerging and effective practices for the development of clean and green production. So that the machining can be costeffective and environment-friendly [4]. In recent times, researchers have been working on minimum quantity lubrication and use of solid lubricants in machining as an alternative to cutting fluids. The minimum quantity lubrication takes a very small quantity of lubricant oil converted in form of aerosols with the use of compressed and applied to the machining area with a small opening nozzle [5]. While machining with higher cutting speeds and feed rate, friction model is developed and MQL is used as lubrication strategy. It is concluded that MQL is an effective option as compared to wet cooling [6]. A new lubrication approach, in form of combined cooling method by pre-cooling the work part by application of MQL is proposed. The result demonstrated the improvement in process performance proving it to be a cost-effective alternative in machining [7]. The lowest value of surface roughness is reported while machining with different ranges of the depth of cut in turning. The reduction of cutting forces is also reported with the use of MQL as compared to dry and wet machining [8]. In another work, improved machining performance is reported with the application of MQL in form of reduction of chip-tool interface temperature and cutting forces [9]. Researchers have proposed and applied solid lubricant in machining as an approach of lubrication. Graphite, calcium fluoride (CaF2 ), molybdenum disulfide (MoS2 ), and boric acid (H3 BO3 ) can be used as solid lubricants. Solid lubricants can be applied in powder form or by mixing it with cutting fluid [10]. Graphite and MoS2 are applied in machining, and its effect on surface finish and cutting forces was analyzed. The results of solid lubricant-assisted lubricants were compared with flood cooling strategy. Results revealed the effectiveness of MoS2 -assisted machining due to its better adhesion tendency as compared graphite [11]. Researchers have been working in the direction of the application of solid lubricants in various machining process leading to a clean and sustainable machining approach [12, 13]. From the current state of research, it can be summarized that there is a need to focus efforts in the area of application of solid lubricants in machining. The concept of solid lubricant mixed with MQL can increase process performance and a reduction in overall machining cost. The work is done either by using the solid lubricant in powder form or by using MQL with cutting fluid. Also, the previous study focused to analyze the effects of solid lubricants without considering any parameter’s effects

18 Applicability of CaF2 Solid Lubricant-Assisted …

231

related to lubrication environment. Few studies are available to assess the effects of various concentrations of solid lubricants and lubricant flow rate in MQSL. Hence, the present work experimentally investigates the performance of MQSL by varying concentration of solid lubricant and the flow rate of lubricant in turning process at selected machining parameters.

18.2 Experimental Details The turning experiments were carried out on HMT made lathe machine under dry, wet, MQL, and MQSL environments. The fabricated minimum quantity lubrication system is mounted on lathe machine as shown in Fig. 18.1, containing main parts as lubricant reservoir, spray nozzles, air pressure controls, and controlling knobs. The main parts of the system are connected by pipes, cables, and required controls and fittings. The compressed air is supplied from the air compressor at required air pressure. The solid lubricant flow is controlled by governing the flow of compressed air. The lubricant mixture leaves nozzle with high pressure in form of aerosol and diverted to machining area. The workpieces in form of round bar with dimensions Ø50 × 200 mm were used for turning operations. Many researchers had worked to find the proper weight ratio of solid lubricant in heterogeneous mixture for the better surface finish and less temperature generation. Vamsi Krishna and Rao [14] used 5, 10, 15, 20, 30, and 40% solid lubricant by weight with SAE 40 oil for testing cutting fluids lubricating and cooling properties. Hence for experiments, 10 and 20% of solid lubricant is mixed with SAE 40 oil and the same is compared with that of dry, wet, and MQL machining. The experiments are performed to study the effect of different solid lubricant concentrations and different

Fig. 18.1 Setup of MQSL mounted on lathe machine

232

M. A. Makhesana et al.

Table 18.1 Experimental conditions used in machining Work material and dimensions

AISI 1040 steel (Fe = 97.7%, C = 0.385%, Si = 0.390%, Mn = 0.991%, P = 0.0128%, S = 0.019%) Ø50 × 200 mm

Cutting conditions

Cutting speed = 130 m/min Feed = 0.2 mm/rev Depth of cut = 0.5 mm

Cutting tool and geometry

CNMG1204085 TN4000 clearance angle (α) = 0°, nose radius = 0.8 mm

Surface roughness tester

Surftest SJ-210

Chip-tool interface temperature Flank wear

Tool-work thermocouple Microscope with image analysis software

Machining conditions

Dry, wet, MQL, MQSL machining with 10 and 20% CaF2 concentration Lubricant flow rates—250 and 500 ml/h

Air pressure

0.3 MPa

Lubricants

Wet cutting environment (cutting fluid) Soluble oil—Servo cut oil ratio—1:20 in water MQL—SAE 40 oil MQSL—Calcium fluoride powder mixed in SAE 40 oil Nozzle tip diameter—1 mm, Nozzle distance from machining zone—15 mm

flow rates of lubricant on surface roughness, chip-tool interface temperature, and flank wear. The details of experimentation are shown in Table 18.1. The optimized value of process parameters is obtained by grey relational analysis for the minimum quantity solid lubrication environment and same has been used for experiments. The turning operation as a part of experiments was carried out by using coated carbide insert of TN4000 grade. The insert with multi-layer coating of CVD—TiNTiCN-Al2 O3 and with higher cobalt content has good toughness which permits heavy depths of cut and interrupted cuts. The insert designation as per ISO code is CNMG1204085 TN4000. The inserts were rigidly mounted on a tool holder with PCLNR2020K12 I 7D designation, as per ISO Code of Widex make.

18.2.1 Measurements Mitutoyo made surface roughness tester is used to measure value of surface roughness after machining. Surface roughness is measured at three different locations and at two different lengths of workpiece, and average value is considered for comparison of results. New cutting tool insert is used for each machining conditions. Measurement of flank wear is carried out using portable microscope equipped with image

18 Applicability of CaF2 Solid Lubricant-Assisted …

233

analysis software. The chip-tool interface temperature is measured by the tool-work thermocouple arrangement developed and mounted on the machine. The calibration of tool-work thermocouple was done by the means of furnace at known temperature. Continuous chip obtained by the machining of workpiece material and tungsten carbide rod is used to form the hot junction. Hot junction of thermocouple was inserted in a metal plate with K type calibrated thermocouple. Metal plate attached with K type thermocouple and hot junction of tool-work thermocouple was kept in a furnace, temperature reading was obtained by K type thermocouple corresponding to emf generated by tool-work thermocouple. For each value of emf, average value of temperature was taken and given an input to the GraphPad software for obtaining the thermoelectric relationship.

18.3 Results and Discussion 18.3.1 Flank Wear Analysis While machining, tool wear is unavoidable due to direct rubbing action between tool and workpiece surface. From the comparison results shown in Fig. 18.2, with the concept of MQSL, tool life is improved as compared to dry and MQL environments. Figure 18.2 shows the maximum flank wear of tool against cutting time for various cutting conditions. The rapid increase of flank wear is observed in dry cutting immediately after 4 min of machining indicating the entry of tool into high wear rate zone. For wet cooling, MQL, and MQSL conditions, the high wear rate zone started approximately after 12 min of machining. From the comparison, it is clear that wet cooling and MQSL with 10% concentration of CaF2 has increased tool life

Fig. 18.2 Comparison of flank wear in various machining environments

234

M. A. Makhesana et al.

by lowering flank wear. The results are in agreement with the fact that in MQL and MQSL, the heat transfer takes place mainly by evaporation making it more efficient than convective heat transfer as in flood machining [14]. Similar trend of results in form of longer tool life and improvement of surface finish is reported with the use of MQL and solid lubricants [15]. The longer tool life with the use of the solid lubricant with 10% concentration is due to its ability to retain in interface area and creating a thin lubricating film. While 20% of concentration has reduced, tool life due to reduction in thermal conductivity of lubricant mixture resulted in higher temperatures leading to larger flank wear. For selected tool life as measured as 0.35 mm of flank wear, effective tool life achieved in dry machining is 9 min, whereas for other cutting conditions it is reported more than 16 min. Figure 18.3 shows the comparison of images of tool insert captured after dry, wet, MQL, and MQSL machining environment. Nose wear is also observed in all inserts.

Fig. 18.3 Images of cutting tool insert in various machining environments, a dry, b wet, c MQL, d MQSL

18 Applicability of CaF2 Solid Lubricant-Assisted …

235

Fig. 18.4 Variations of chip-tool interface temperature in various machining environments

18.3.2 Analysis of Chip-Tool Interface Temperature The heat generated during machining is the result of plastic deformation due to rubbing action of the tool with respect to the workpiece surface. It is very important in machining to control the amount of heat generated as it really affects the rate of tool wear and surface integrity of workpiece [16]. From the comparative analysis of results shown in Fig. 18.4, it is clear that the highest chip-tool temperature is observed in case dry machining followed by MQL, wet cooling, and MQSL environments. The cooling action of MQSL is almost same as that of wet cooling which shows excellent lubricating capabilities of solid lubricants. Another reason is the application of compressed air supplied in MQSL creating an excellent cooling effect by taking heat away from the machining area. Reduction in temperature leads toward the reduction in flank wear and surface roughness (Fig. 18.2 and 18.5).

18.3.3 Analysis of Surface Roughness Figure 18.5 shows the comparison of average surface roughness values for selected cutting conditions. It is as expected that due to rapid tool wear observed immediately after 4 min of machining, highest value of surface roughness is reported in dry machining. While for other cutting environments, no major change in values of surface roughness is observed. The results of surface roughness in case of MQSL approaches indicate the effectiveness of solid lubricants as compared to wet cooling. However, 10% concentration of solid lubricant has lowered the Ra value as compared to that of 20% concentration. The same can be linked with flank wear comparison in terms of lower flank wear with 10% concentration retaining longer tool life that 20% concentration of CaF2 . The better lubricating action of solid lubricant reduces

236

M. A. Makhesana et al.

Fig. 18.5 Variations of surface roughness in various machining environments

the frictional forces resulting in reduction in temperature and tool wear and leading to improved surface finish [17]. Work is reported with the application of MoS2 and graphite with oil at concentration (20% wt) in machining of Ti-6Al-4V alloy. Reduced values reported of cutting forces and resulted in improved surface finish with solid lubricants as compared to dry and wet cooling. Molybdenum disulfide (MoS2 ) has improved machining performance in form of lower tool wear and surface roughness as compared to graphite [18].

18.3.4 Effects of Different Lubricant Flow Rates on Surface Roughness From the experimental results compared with different concentrations of solid lubricant, it is concluded that 10% of CaF2 has improved machining performance by lowering tool flank wear and surface roughness. Further, it is decided to compare the effects of various flow rates of lubricant in MQL and MQSL environments by keeping 10% concentration of solid lubricant fir MQSL. The flow rate in MQL and MQSL approach is controlled by flow regulating knob of MQL system. Figure 18.6 compares the effect of different flow rates by which lubricant mixture supplied in MQL and MQSL. It can be seen that lower flow rates in case of MQL and MQSL have lowered the value of surface roughness than that of wet cooling wherein the flow rate is few liters per hour of machining. Hence, it is clear that if the lubricant can be supplied in optimize and effective way, it can improve the machining performance. MQSL lubrication with 500 ml/h has proven to be better due to the lubrication properties of solid lubricant added in mixture than MQL without solid lubricant.

18 Applicability of CaF2 Solid Lubricant-Assisted …

237

Fig. 18.6 Variations of surface roughness with various flow rates of lubricants

18.4 Conclusion Minimum quantity lubrication added with solid lubricant is applied in turning operation and compared with the results available from dry, wet and MQL machining environments. Due to inherent lubricating properties of solid lubricant, it has improved surface finish as compared to other cutting conditions. Tool life is improved while application of MQSL approach, as solid lubricant can penetrate effectively between tool-work interfaces resulting in better lubrication and cooling action performed by compressed air. From the selected combinations, MQSL with 10% concentration of CaF2 has improved process performance. It can be also concluded from the comparison of different flow rates of lubricant that instead of applying large quantities of cutting fluids, an optimum flow rate of the same can improve the process performance. Hence, the use of solid lubricant coupled with the concept of minimum quantity lubrication can be a feasible alternative leading toward sustainable manufacturing practices. Analysis and measurement of thermo-physical properties of lubricant mixture are not considered in the present study. Advanced CFD modeling techniques can be employed to model mist-based cooling techniques to predict the temperature distributions on the cutting tool under different lubricating conditions. In future, different combinations of solid lubricant particle size and air pressures should be examined for higher cutting speeds and feed rates. The similar study can also be conducted for different types of coolants, and their respective convective heat transfer coefficients can be analyzed in form of its effect on output responses.

238

M. A. Makhesana et al.

References 1. Bruni, C., Gabrielli, F.F., Simoncini, M.: Effect of the lubrication-cooling technique, insert technology and machine bed material on the workpart surface finish and tool wear in finish turning of AISI 420B. Int. J. Mach. Tools Manuf. 46(12–13), 1547–1554 (2006) 2. Howes, T.D., Toenshoff, H.K., Heur, W.: Environmental aspects of grinding fluids. Ann. CIRP 40(2), 623–630 (1991) 3. Brinksmeier, E., Walter, A., Janssen, R., Diersen, P.: Aspects of cooling lubrication reduction in machining advanced materials. Proc. Inst. Mech. Eng. 213(8), 769–778 (1999) 4. Ghosh, S., Rao, P.V.: Application of sustainable techniques in metal cutting for enhanced machinability: a review. J. Clean. Prod. 100, 17–34 (2015) 5. Varadharajan, A.S., Philip, P.K., Ramamoorthy, B.: Investigations on hard turning with minimal pulsed jet of cutting fluid. In: Proceedings of the International Seminar on Manufacturing Technology beyond 2000, Bangalore, India, November 17–19, pp. 173–179 (1999) 6. Banerjee, N., Sharma, A.: Identification of a friction model for minimum quantity lubrication machining. J. Clean. Prod. 83, 437–443 (2014) 7. Shokoohi, Y., Khosrojerdi, E., Shiadhi, B.H.R.: Machining and ecological effects of a new developed cutting fluid in combination with different cooling techniques on turning operation. J. Clean. Prod. 94, 330–339 (2015) 8. Hadad, M., Sadeghi, B.: Minimum quantity lubrication-MQL turning of AISI4140 steel alloy. J. Clean. Prod. 54, 332–343 (2013) 9. Saini, A., Dhiman, S., Sharma, R., Setia, S.: Experimental estimation and optimization of process parameters under minimum quantity lubrication and dry turning of AISI-4340 with different carbide inserts. J. Mech. Sci. Technol. 28(6), 2307–2318 (2014) 10. Reddy, N.S.K., Rao, P.V.: Experimental investigation to study the effect of solid lubricants on cutting forces and surface quality in end milling. Int. J. Mach. Tools Manuf. 46(2), 189–198 (2006) 11. Singh, D., Rao, P.V.: Performance improvement of hard turning with solid lubricants. Int. J. Adv. Manuf. Technol. 38(5–6), 529–535 (2008) 12. Reddy, N.S.K., Rao, P.V.: Performance improvement of end milling using graphite as a solid lubricant. Mater. Manuf. Processes 20(4), 673–686 (2005) 13. Mukhopadhyay, D., Banerjee, S., Reddy, N.S.K.: Investigation to study the applicability of solid lubricant in turning AISI 1040 steel. Trans. ASME J. Manuf. Sci. Eng. 129(3), 520–526 (2006) 14. Rao, D.N., Vamsi Krishna, P.: The influence of solid lubricant particle size on machining parameters in turning. Int. J. Mach. Tools Manuf. 48(1), 107–111 (2008) 15. Philip, P.K., Varadharajan, A.S., Ramamoorthy, B.: Influences of cutting fluid composition and delivery variables on performance in hard turning using minimal fluid in pulsed jet form. J. Inst. Eng. India 82(1), 68–72 (2001) 16. Du, G.C., Chen, Y., Wei, Z.Z.: Effects of solid lubricants on hard turning. In: Proceedings of the 2nd International Conference on Electronic and Mechanical Engineering and Information Technology, pp. 1147–1149 (2012) 17. Lawal, S.A., Choudhury, I.A., Nukman, Y.: A critical assessment of lubrication techniques in machining processes: a case for minimum quantity lubrication using vegetable oil-based lubricant. J. Clean. Prod. 41, 210–221 (2013) 18. Moura, R.R., Da Silva, M.B., Machado, A.R., Sales, W.F.: The effect of application of cutting fluid with solid lubricant in suspension during cutting of Ti-6Al-4V alloy. Wear 332–333, 762–771 (2015)

Chapter 19

Cryogenic Machining of AZ31B Magnesium Alloy for Bio-implant Applications Vaibhav Tibrewal, Kalpit Dak, Aundhe Himanshu, Hema Kumar, P. Kuppan and A. S. S. Balan Abstract Magnesium and its alloys are slowly entering into the field of bio-implants as a substitute to currently used materials because of their mechanical properties and physiological benefits. However, the magnesium alloys corrode much before than the bone is fully healed because of their high corrosion rate in physiological environment of body. In this experiment, AZ31B magnesium alloy has been subjected to turning operation under dry and cryogenic environment. This research is an attempt to study the effects of cutting speed and feed rate on forces, surface roughness, temperature and microstructure. Furthermore, a comparative study is done on the effects of machining environment on these factors. The results show that a combination of high cutting speed and low feed rate with cryogenic environment gives the best surface finish. Keywords Bio-implants · Surface roughness · Turning · Cryogenic environment · Chip morphology

19.1 Introduction Biomedical industries currently use titanium alloys as implants because of its high corrosion resistance and high strength to weight ratio [1]. Cobalt-chromium alloys and stainless steels are some other materials which found in their application implants. However, their mechanical properties do not match with that of the bone tissues. This mechanical mismatch between bone and metallic implants results in stress shielding. This phenomenon occurs when the implant carries the bulk of the V. Tibrewal · K. Dak · A. Himanshu · H. Kumar · P. Kuppan School of Mechanical Engineering, Vellore Institute of Technology, Vellore 632014, India A. S. S. Balan (B) Centre for Innovative Manufacturing Research, Vellore Institute of Technology, Vellore 632014, India e-mail: [email protected]

© Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_19

239

240

V. Tibrewal et al.

load and surrounding bone tissue that experiences a reduced loading stress. The reduced loading stress ultimately leads to bone resorption [2]. The major drawback of these materials is that they are non-degradable in the physiological environment, which means that a second surgery will be needed in order to remove them [3]. Over a period of time, due to slow degradation of these alloys, they release cobalt, chromium and nickel ions into the body which are toxic in nature for the tissues. These ions are also released because of mechanical wear and corrosion of the materials [2]. Thus, though these alloys are used in large scale, a substitute is needed to overcome their drawbacks. Magnesium and its alloys are widely used in automotive and aerospace industries [4]. But due to its properties, magnesium and its alloys have recently found their use in biomedical applications. The biodegradable nature of magnesium alloys in physiological environment and similar elastic strength to that of bone tissue makes it suitable as an implant material [5]. In addition to these, density of magnesium alloys that are similar to bone (magnesium and its alloys have around 1.74 g/cm3 while bone have 1.8–2.4 g cm3 ) does not release toxic ions into the body [2] and can resist various biomechanical forces [3]. The major drawback of using magnesium alloys is that their rate of degradation is very high. Also, they lose their integrity when subjected to high chloride environments and at high pH level of 7.4–7.6 [6]. The degradation via the corrosion process rapidly produces hydrogen gas, thus not giving the bones enough time to heal leading to deterioration of the mechanical integrity of the implant much before the complete healing [7]. The various types of corrosion a magnesium implant can go through are galvanic corrosion (if metals of dissimilar electrochemical properties are used), pitting corrosion, crevice corrosion, fretting corrosion (can be considered physiologically beneficial because the metallic ions can be absorbed by surrounding tissues or can be dissolved and excreted via kidneys), stress corrosion, and corrosion fatigue [2]. To overcome this problem, the corrosion resistance can be enhanced by various ways such as by implementing a reasonable design, cathodic protection, alloy medication, surface modification, and coating [4]. In this paper, we performed turning operation on magnesium AZ31B alloy to modify its surface properties. Surface quality is an important factor in manufacturing of products. Surface integrity contributes to mechanical strength, hardness, fatigue life, and corrosion and wear resistance [8]. Machining operation subjected to certain conditions can impart compressive residual stresses to the material and refine the surface grains which in turn enhances the corrosion resistance. Also, it has been observed that the basal plane (0001) of AZ31B magnesium alloy offers more resistance towards corrosion compared to any other basal plane [9]. One of the major factors affecting the machining operation is the way of cooling. The performance and productivity of machining operations are widely influenced by wear formation and heat generated [10]. Water-based coolants cannot be used on magnesium and its alloys because magnesium on reacting with water releases hydrogen gas. Oil-based coolants produce perilous gases leading to explosion and also affect human health [11]. In this paper, we performed the turning operation by subjecting the material to both dry and cryogenic environments. Liquid nitrogen has been chosen as the coolant because of its abundance, and its odourless, colourless, tasteless and non-toxic nature [12].

19 Cryogenic Machining of AZ31B Magnesium Alloy …

241

The present work demonstrates the comparison between dry and cryogenic machining and their effects on forces, surface roughness, temperature and grain refinement by varying the cutting speed and feed rate.

19.2 Experimental Set-up The chemical composition of AZ31B, a magnesium alloy is shown in Table 19.1. The microstructure of the base sample is shown in Fig. 19.1. The geometry of material was of a cylindrical rod having a diameter of 20 mm and length 250 mm. The cutting tool used was uncoated carbide tool. The geometry of cutting tool is presented in Fig. 19.2. The properties of AZ31B alloy and the specifications of the tool are listed in Table 19.2. The various parameters considered for the turning operation are listed in Table 19.3. The operations were performed for both dry and cryogenic environment. A CNC lathe (Ace Micromatic make) was used to perform the turning operations. Machining length was taken as 15 mm. Mean forces were measured using a piezoelectric dynamometer (Kistler make). To measure the temperature, an infrared camera (FLIR) was used and the emissivity of the material was set to 0.18. For creating the cryogenic environment, liquid nitrogen was supplied along the relief side of the tool at a pressure of 2.5 bar. The average surface roughness of the machined surface was measured as per ASTM G171-03 standard using a surface profilometer Table 19.1 Chemical composition of AZ31B alloy % Al

% Zn

% Mn

% Si

% Mg

3.2

1.02

0.20

0.10

Rem

Fig. 19.1 Microstructure of AZ31B alloy

242

V. Tibrewal et al.

Fig. 19.2 Geometry of the tool

Table 19.2 Properties of AZ31B and specifications of the cutting tool Property

Value

Tensile strength

260 MPa

Ultimate bearing strength

385 MPa

Poisson’s ratio

0.35

Density

1.77 g/cm3

Thermal expansion co-efficient (0–100 °C)

26 µm/m °C

Thermal conductivity

96 W/m K

Emissivity

0.18

Tool specification

Value

Tool

CNMG 120408 MP

RE (Corner radius)

0.794 mm

LE (Cutting edge effective length)

12.096 mm

IC (Inscribed circle diameter)

12.7 mm

S (Insert thickness)

4.763 mm

(MAHR make) by taking a transverse length of 4 mm and setting the stylus speed to 0.5 mm/s. The tool wear was measured using a digital microscope (Dino-Lite, Taiwan make). For obtaining the microstructure, the workpiece was cut in the shape of an extruded semicircle using wire EDM. Initial polishing was carried out on silicon carbide papers of grit size 240–1500. After this, diamond polishing was done on a TexMat 1000 Pad for about 5–6 min. It was then cleaned with ethanol. The etchant used was aceticpicral solution. The etchant was applied on the surface for about 3–4 s and then cleaned with ethanol. The microstructure was observed using an optical microscope. The experimental set-up is shown in Fig. 19.3.

Feed rate (mm/rev)

0.10

0.15

0.20

0.10

0.15

0.20

0.10

0.15

0.20

Cutting speed (m/min)

125

125

125

155

155

155

185

185

185

Table 19.3 Machining parameters

0.5

0.5

0.5

0.5

0.5

0.5

0.5

0.5

0.5

DOC (mm)

73

30.5

29.8

69.7

47.9

33.5

64

51.9

42.9

85

37.8

40.2

81.5

45.2

46.8

72

60.9

56.7

65

33.8

23

61.7

35.48

29

47

47

32.02

69

35

28.1

64

44

36.2

56

48.2

40.8

Cryo

Dry

Dry

Cryo

Thrust force (N)

Cutting force (N)

0.84

0.51

0.22

0.97

0.914

0.25

1.51

1.06

0.314

Dry

0.72

0.763

0.21

0.87

0.791

0.226

1.168

0.834

0.29

Cryo

Surface roughness Ra (µm)

369

342

309

329

253

296

293

208

277

Dry

126

109

151

121

106

141

84

82

80

Cryo

Temperature (°C)

19 Cryogenic Machining of AZ31B Magnesium Alloy … 243

244

V. Tibrewal et al.

Fig. 19.3 Experimental set-up

19.3 Results and Discussion 19.3.1 Cutting Force and Thrust Force The variation of mean cutting force and thrust force with respect to feed rate, cutting speed and machining environment has been shown in Fig. 19.4. It was observed that, both the cutting and thrust forces increased with respect to the feed rate because of strain-hardening effect [11]. Also, the forces decreased with increase in cutting speed at lower feed rate. This can be attributed to the fact that with increase in speed, the temperature between cutting tool and the workpiece increases which softens the material and aids in easy removal of material and thus decreasing the forces [13]. However for a higher feed rate of 0.2 mm/rev, the forces increased with increase in

Fig. 19.4 Variation of a cutting force and b thrust force with feed and speed for dry and cryogenic environment

19 Cryogenic Machining of AZ31B Magnesium Alloy …

245

cutting speed. This is because at these conditions, strain hardening dominates over the thermal softening effect. This can be also inferred from the chips. The chips formed at this feed rate were discontinuous (except for cutting speed of 125 m/min) indicating the brittleness of the material [9]. The chips for 0.2 feed rate are shown in Fig. 19.9. It was observed that higher forces were generated for cryogenic machining compared to dry machining. This is because liquid nitrogen rapidly cools the surface which hardens the material resulting in low impact toughness, and a large force is required to remove the material [14]. Also, the friction is reduced at the tool– workpiece interface because of cushioning effect in case of cryogenic machining [11]. A cushion is formed as the liquid nitrogen expands as it is expelled out of the nozzle and penetrates into the cutting zone in gaseous form, thereby reducing the contact friction between the tool and the chips [15].

19.3.2 Surface Roughness The variation of surface roughness with respect to feed rate, cutting speed and machining environment has been shown in Fig. 19.5. With increase in feed rate, the surface roughness (Ra ) increases. This is because the path travelled by the tool during turning operation is a helical one; thus, there is a small portion of material left uncut. The amount of this uncut portion increases as the feed is increased. Hence, lower feed rates give improved surface finish [16]. As the cutting speed increases, the roughness was found to be decreasing. This may be because higher cutting speed produces less cutting forces, thereby reducing the tool chatter and hence enhancing the surface finish. Moreover, less built-up edges

Fig. 19.5 Variation of surface roughness with feed and cutting speed for dry and cryogenic environment

246

V. Tibrewal et al.

are formed at higher cutting speed which also promotes better surface quality [17, 18]. Also, it was observed that the surface roughness for cryogenic machining was 10–20% less than for dry machining. This is because of less flank wear due to cushioning effect of liquid nitrogen. Also, the heat generated the friction is much less in cryogenic machining [10, 19]. The best surface finish was found for higher cutting speed and low feed rate in both dry and cryogenic environment [20].

19.3.3 Temperature Figure 19.7 shows the variation of maximum temperature with respect to feed rate, cutting speed and machining environment. It is observed that the temperature increases with increase in the feed rate. The same trend is seen with cutting speed also. The main reason being that as these parameters increases, the material removal rate increases, thus increasing the friction and so give an increasing temperature trend [17]. Under cryogenic environment, the maximum temperature reached was found to be 60–70% less than when put under dry environment. This is because of the intense cooling action of the liquid nitrogen which dominates over the heat generated due to deformation and friction. Moreover, the friction at tool–workpiece interface is also reduced because of the cushioning effect of liquid of nitrogen [10, 15, 17]. The FLIR images for different machining conditions are shown in Fig. 19.6.

Fig. 19.6 FLIR images of different machining conditions

19 Cryogenic Machining of AZ31B Magnesium Alloy …

247

Fig. 19.7 Variation of temperature with feed and speed for dry and cryogenic environment

19.3.4 Microstructure The microstructure of the base material is shown in Fig. 19.1, and machined workpiece are shown in Fig. 19.8. It was observed that significant grain refinement occurred when machined under cryogenic environment when compared to dry machining environment. This is because the higher forces were generated during cryogenic machining which induced severe plastic deformation at the surface and sub-surface level thereby modifying the microstructure to a greater depth [21]. The thickness of layer increased from 31 µm (speed 125 m/min, feed 0.1 mm/rev) to 39 µm (speed 185 m/min, feed 0.2 mm/rev) in case of cryogenic environment. The trend was similar for dry environment. The grains formed due to cryogenic machining were smaller than that formed due to dry machining. This is because the smaller grains leads to the higher value of hardness as per the Hall–Petch relationship [22]. Moreover, high-strain rate leads to finer grains because of dynamic recrystallization [1, 23]. Higher temperatures near the shear zones also affect the grain refinement. In cryogenic machining, smaller grains are generated because of cold flow [24]. The variation of hardness with various machining conditions is shown in Fig. 19.8.

19.3.5 Chip Morphology The chips obtained during the machining of AZ31B alloy are shown in Fig. 19.9. The continuous nature of chips faded as the feed rate increases. Similar trend was observed with variation in cutting speed. The chips were discontinuous at feed of 0.2 mm/rev except for dry machining at 125 m/min. This is because at higher feed rate, the strain rate increases inducing brittleness and thus length of continuous chips

248

V. Tibrewal et al.

Fig. 19.8 Microstructure of AZ31B alloy after machining under various conditions

Fig. 19.9 Chips obtained for various machining parameters

19 Cryogenic Machining of AZ31B Magnesium Alloy …

249

decreases. The continuity of chips also decreased with an increase in cutting speed. The chips obtained during dry and cryogenic machining were comparable in length. However, the radius of the curvature of chips was smaller for those produced during cryogenic machining [25, 26].

19.4 Conclusion Turning operation on AZ31B magnesium alloy was successfully performed under various conditions. The influence of cutting speed, feed rate and machining environment on parameters such as forces, surface roughness, temperature, microstructure, chip morphology, corrosion rate, hardness and tool wear have been studied and are listed below: • Cutting force and thrust force increased with an increasing feed rate and decreased with an increasing cutting speed. Change in trend was observed for 0.2 mm/rev feed rate where the forces increased with an increase in cutting speed. Machining under cryogenic environment induced larger forces compared to machining under dry environment. • Surface roughness increased with an increase in feed rate and decreased with an increase in cutting speed. Best surface finish was observed for low feed and high cutting speed. Cryogenic machining gave much better surface finish compared to dry machining. • Temperature increased with both increase in feed rate and cutting speed. The temperature recorded for cryogenic machining was 60–70% less than that recorded for dry machining. • Microstructure of initial and machined alloy was taken. Significant grain refinement was seen for cryogenic machining compared to dry machining. Also, the thickness of the affected layer increased when the machining conditions was changed from 125 m/min speed, 0.1 mm/rev feed to 185 m/min speed, 0.2 mm/rev feed. • The continuous nature of the chips decreased with an increase in both feed rate and cutting speed. The chips were discontinuous for all cutting speed at 0.2 mm/rev feed rate (except for 125 m/min in dry machining environment).

References 1. Chen, Q., Shu, D., Hu, C., Zhao, Z., Yuan, B.: Grain refinement in an as-cast AZ61 magnesium alloy processed by multi-axial forging under the multitemperature processing procedure. Mater. Sci. Eng. A 541, 98–104 (2012). https://doi.org/10.1016/j.msea.2012.02.009 2. Poinern, G.E.J., Brundavanam, S., Fawcett, D.: Biomedical magnesium alloys: a review of material properties, surface modifications and potential as a biodegradable orthopaedic implant. Am. J. Biomed. Eng. 2(6), 218–240 (2012). https://doi.org/10.5923/j.ajbe.20120206.02

250

V. Tibrewal et al.

3. Radha, R., Sreekanth, D.: Insight of magnesium alloys and composites for orthopaedic implant applications—a review. J. Magnes. Alloys 5, 286–312 (2017). https://doi.org/10.1016/j.jma. 2017.08.003 4. Song, G.L.: Corrosion behavior and prevention strategies for magnesium (Mg) alloys. General Motors Corporation, USA. https://doi.org/10.1533/9780857098962.1.3 5. Zhang, E., Yang, L., Xu, J., Chen, H.: Microstructure, mechanical properties and bio-corrosion properties of Mg-Si(-Ca,Zn) alloy for biomedical application. Acta Biomater. 6, 1756–1762 (2010). https://doi.org/10.1016/j.actbio.2009.11.024 6. Chen, Y., Xu, Z., Smith, C., Sankar, J.: Recent advances on the development of magnesium alloys for biodegradable implants. Acta Biomater. 10, 4561–4573 (2014). https://doi.org/10. 1016/j.actbio.2014.07.005 7. Uddin, M.S., Rosman, H., Hall, C., Murphy, P.: Enhancing the corrosion resistance of biodegradable Mg-based alloy by machining-induced surface integrity: influence of machining parameters on surface roughness and hardness. Int. J. Adv. Manuf. Technol. 90, 2095–2108 (2017). https://doi.org/10.1007/s00170-016-9536-x 8. Nasr, M.N.A., Outeiro, J.C.: Sensitivity analysis of cryogenic cooling on machining of magnesium alloy AZ31B-O. Procedia CIRP 31, 264–269 (2015). https://doi.org/10.1016/j.procir. 2015.03.030 9. Pu, Z.: Cryogenic machining and burnishing of AZ31B magnesium alloy for enhanced surface integrity and functional performance. Thesis and dissertations—Mechanical engineering. 5 (2012). http://uknowledge.uky.edu/me_etds/5 10. Yildiz, Y., Nalbant, M.: A review of cryogenic cooling in machining processes. Int. J. Mach. Tools Manuf. 48, 947–964 (2008). https://doi.org/10.1016/j.ijmachtools.2008.01.008 11. Dinesh, S., Senthilkumar, V., Asokan, P., Arulkirubakaran, D.: Effect of cryogenic cooling on machinability and surface quality of bio-degradable ZK60 Mg alloy. Mater. Des. 87, 1030–1036 (2015). https://doi.org/10.1016/j.matdes.2015.08.099 12. Pu, Z., Outeiro, J.C., Batista, A.C., Dillon Jr., O.W., Puleo, D.A., Jawahir, I.S.: Surface integrity in dry and cryogenic machining of AZ31B Mg alloy with varying cutting edge radius tools. Procedia Eng. 19, 282–287 (2011). https://doi.org/10.1016/j.proeng.2011.11.113 13. Shi, K., Zhang, D., Ren, J., Yao, C., Huang, X.: Effect of cutting parameters on machinability characteristics in milling of magnesium alloy with carbide tool. Adv. Mech. Eng. 8(1), 1–9 (2016). https://doi.org/10.1177/1687814016628392 14. Magadum, S., Arun Kumar, S., Yoganath, V.G., Srinivasa, C.K., GuruMurthy, T.: Evaluation of tool life and cutting forces in cryogenic machining of hardened steel. Procedia Mater. Sci. 5, 2542–2549 (2014). https://doi.org/10.1016/j.mspro.2014.07.506 15. Dilip Jerold, B., Pradeep Kumar, M.: Experimental comparison of carbon-dioxide and liquid nitrogen cryogenic coolants in turning of AISI 1045 steel. Cryogenics 52, 569–574 (2012). https://doi.org/10.1016/j.cryogenics.2012.07.009 16. Malleswara Rao, J.N., Sumalatha, M., Kesava Rao, V.V.S., Anurupa, V., Srivalli, G.: Variation of surface roughness with feed rate on mild steel components produced by CNC lathe. Int. Res. J. Eng. Technol. 3(06) (2016) 17. Danish, M., Ginta, T. L., Habib, K., Carou, D., Rani, A. M. A., Saha, B. B.: Thermal analysis during turning of AZ31 magnesium alloy under dry and cryogenic conditions. Int. J. Adv. Manuf. Technol. 91, 2855–2868 (2017). https://doi.org/10.1007/s00170-016-9893-5 18. Pu, Z., Song, G.-L., Yang, S., Outeiro, J.C., Dillon Jr., O.W., Puleo, D.A., Jawahir, I.S.: Grain refined and basal textured surface produced by burnishing for improved corrosion performance of AZ31B Mg alloy. Corros. Sci. 57, 192–201 (2012). https://doi.org/10.1016/j.corsci.2011. 12.018 19. Al-Dolaimy, K.A.: Effect of cutting parameters on surface roughness in turning operations. Al-Qadisiyah J. Eng. Sci. 9(4) (2016) 20. Paul, S., Dhar, N.R., Chattopadhyay, A.B.: Beneficial effects of cryogenic cooling over dry and wet machining on tool wear and surface finish in turning AISI 1060 steel. J. Mater. Process. Technol. 116, 44–48 (2001). https://doi.org/10.1016/S0924-0136(01)00839-1

19 Cryogenic Machining of AZ31B Magnesium Alloy …

251

21. Viswanathan, R., Ramesh, S., Subburam, V.: Measurement and optimization of performance characteristics in turning of Mg alloy under dry and MQL conditions. Measurement (2018). https://doi.org/10.1016/j.measurement.2018.02.018 22. Pu, Z., Dillon Jr., O.W., Jawahir, I.S., Puleo, D.A.: Microstructural changes of AZ31 magnesium alloys induced by cryogenic machining and its influence on corrosion resistance in simulated body fluid for biomedical applications. In: Proceedings of the ASME 2010 International Manufacturing Science and Engineering Conference MSEC2010, October 12–15, 2010, Erie, Pennsylvania, USA. https://doi.org/10.1115/msec2010-34234 23. Rotella, G., Umbrello, D.: Finite element modeling of microstructural changes in dry and cryogenic cutting of Ti6Al4V alloy. CIRP Ann. Manuf. Technol. (2014). https://doi.org/10. 1016/j.cirp.2014.03.074 24. Swaminathan, S., Shankar, M.R., Lee, S., Hwang, J., King, A.H., Kezar, R.F., Rao, B.C., Brown, T.L., Chandrasekar, S., Compton, W.D., Trumble, K.P.: Large strain deformation and ultra-fine grained materials by machining. Mater. Sci. Eng. A 410–411, 358–363 (2005). https://doi.org/ 10.1016/j.msea.2005.08.139 25. Aramcharoen, A.: Influence of cryogenic cooling on tool wear and chip formation in turning of titanium alloy. Preocedia CIRP 46, 83–86 (2016). https://doi.org/10.1016/j.procir.2016.03.184 26. Bermingham, M.J., Palanisamy, S., Kent, D., Dargusch, M.S.: A comparison of cryogenic and high pressure emulsion cooling technologies on tool life and chip morphology in Ti-6Al-4V cutting. J. Mater. Process. Technol. 212, 752–765 (2012). https://doi.org/10.1016/j.jmatprotec. 2011.10.027

Chapter 20

Experimental Investigation on Machining Parameters of Hastelloy C276 Under Different Cryogenic Environment S. Vignesh

and U. Mohammed Iqbal

Abstract Materials with high strength and light in weight are in demand in highprecision manufacturing. Machining these materials becomes tedious in extreme cutting conditions where cutting fluids sometimes fail. Hence an alternate method of cooling is needed to replace the conventional coolant. This paper presents the effects of cryogenic coolants in turning of Hastelloy C276. This alloy finds its major applications in sophisticated applications like nuclear reactors, pressure vessels and heat exchangers. The present approach was carried out to develop a comparative study between two cryogenic coolants (liquid nitrogen and carbon dioxide). Experiments were performed using Taguchi L9 orthogonal technique, and the parameters considered were speed, feed rate and depth of cut. Better output in terms of surface finish was observed in liquid nitrogen when compared to carbon dioxide. Keywords Hastelloy C276 · Cryogenic coolants · Surface roughness

20.1 Introduction Cryogenic coolants are benign to manufacturing sector as it provides a safe, sustainable and non-toxic environment to workers. In addition to improved cooling, it also extends its wide hand improvising tool life and surface texture. Elimination of secondary cleaning process after machining is also one of the major advantages in cryogenic machining [1]. Several comparisons were made during recent years. Pereira et al. [2] compared the performance of CO2 as cryogenic coolant in hard turning with dry machining and found the improvement in tool life of 60%. Hong et al. [3] have found that the tool life during cryogenic turning of Ti–6Al–4V has increased S. Vignesh Department of Mechanical Engineering, Faculty of Engineering and Technology, SRM Institute of Science and Technology, Vadapalani Campus, Chennai 600026, India U. Mohammed Iqbal (B) Department of Mechanical Engineering, Faculty of Engineering and Technology, SRM Institute of Science and Technology, Kattankulathur, Chennai 603203, India e-mail: [email protected]; [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_20

253

254

S. Vignesh and U. Mohammed Iqbal

by five times when compared to other cooling approaches. Umbrella et al. [4] experimentally found that cryogenic turning of AISI 52100 showed greater improvement in surface integrity. Ghosh et al. [5] and Ezugwu et al. [6] have studied the performance of cryogenic coolants in machining and reported that cryogenic coolants show improvement in heat removal from cutting zone and also reduction of thermal softening in cutting tools. They also suggest that use of liquid nitrogen gas (LN2 ) provides better cooling and lubrication since it absorbs heat and evaporates into the atmosphere and also it forms a layer between the chip and tool face that provides a lubrication effect. Hastelloy C276 finds applications in metal processing, steam turbine power plant, reciprocating engine parts, heat treating equipment due to its ability to withstand chemical reactions and providing easiness in forming and welding. Due to its high level of hardness, retention of strength at higher temperatures and low-thermal conductivity machining becomes difficult [7–9]. Numerous studies have been conducted to identify the optimal cutting conditions and environment for machining Hastelloy C276. Major literature suggests that cryogenic cooling is the advanced and recent method in providing better cooling and lubrication on the machining region. Surface roughness of the material determines the surface integrity of the material which is much dependent on feed rate, cutting speed and depth of cut. Kaynak et al. [10] have suggested that very few studies have been carried out in studying the effects of cryogenics on parameters like surface roughness, material removal rate (MRR) and tool wear. They have also found that LN2 as cryogenic coolant in machining super alloys has shown major improvement in providing better surface integrity. This study focusses on comparing the performance of two cryogenic cooling environments (viz. CO2 and LN2 ) and determining the optimal turning conditions in terms of surface roughness for Inconel C-276.

20.2 Experimentation 20.2.1 Work Material and Tool Hastelloy C-276 with length of 44 mm and diameter of 16 mm was selected as a work material for this study. The chemical compositions of the chosen material were taken using OES-Foundry Master-Pro as given in Table 20.1. The turning insert used here is Tagutech-make metal working carbide insert (model-TNMG 160404MT TT5030) with cutting diameter of 0.4 mm and coated with tungsten by CVD process which provides moderate hardness and lesser material adhesion.

C

0.006

Elements

%

0.2

Mg 0.07

Si 0.001

S

Table 20.1 Chemical composition of Hastelloy C-276 in wt% 0.015

K 14.7

Cr 17.6

V

0.2

Co

6.1

Fe

3.3

W

Bal

Ni

20 Experimental Investigation on Machining Parameters … 255

256

S. Vignesh and U. Mohammed Iqbal

Fig. 20.1 Jobber XL turning machine

Table 20.2 Selected parameters and their levels Parameter

Unit

Levels −1

0

1

Spindle speed

m/min

60

65

70

Feed rate

mm/rev

0.04

0.06

0.08

Depth of cut

mm

0.5

0.75

1.25

20.2.2 Experimental Details The CNC turning machine used here is Jobber XL model of GEEDEE Weiler-make with maximum spindle speed of 4000 rpm as shown in Fig. 20.1. Edwin et al. [11] have studied the importance of TNMG tools and carbide tools in machining process. Patil et al. [12] have suggested that speed, feed rate and depth of cut as most influencing parameters during CNC machining. The same process parameters were considered, and levels were varied based on machine specifications and property of the material. The selected levels for the study are shown in Table 20.2. A cryogenic transfer system shown in Fig. 20.2a, b was designed with an insulated copper tube connected to the cylinder along with ON/OFF control valve with maximum flow rate of 5 lit/min and operating pressure of 0.7 bar. Surface roughness was measured using Surfcom 1400G perthometer.

20.2.3 Design of Experiments Considering the levels of machining parameters, Taguchi technique is utilized to assure minimum variability in mean and also to minimize the number of experimental

20 Experimental Investigation on Machining Parameters …

(a)

257

(b)

Fig. 20.2 a Cryogenic transfer device, b Transfer device attached to the cryo-cylinder with flow control valve

runs. The L9 orthogonal array was adopted here with eight degrees of freedom to potentially evaluate the factors in least trials. The structured orthogonal L9 array is shown in Table 20.3. The quality responses are optimized by signal-to-noise ratio (S/N) which is the ratio between expected signal and unexpected noise. It improves the quality performance by reducing the variance and refining the measured mean. The following are the S/N ratio prescriptions given in Eqs. (20.1)–(20.3). (i)

Smaller-the-better: n = −10 log10 [Mean of sum of squares of {measured data − ideal data}] (20.1)

This is chosen for undesirable characteristics for which ideal value is zero. (ii) Larger-the-better: n = −10 log10 [Mean of sum of squares of reciprocal of measured data] (20.2)

Table 20.3 Design of L9 orthogonal array using Taguchi technique

S. No.

Speed (m/min)

Feed rate (mm/rev)

Depth of Cut (mm)

1

60

0.04

0.5

2

60

0.06

0.75

3

60

0.08

1.25

4

65

0.04

0.75

5

65

0.06

1.25

6

65

0.08

0.5

7

70

0.04

1.25

8

70

0.06

0.5

9

70

0.08

0.75

258

S. Vignesh and U. Mohammed Iqbal

(iii) Nominal-the-best: n=

10 log10 square of mean variance

(20.3)

In this study, the considered response parameter (i.e., surface roughness) has to be minimized hence (i) smaller is better characteristics is taken.

20.3 Results and Discussion The experimental data were recorded, and the recorded data were analyzed further for its anomaly and reported below.

20.3.1 Surface Roughness The measured surface roughness for the experimental runs is shown in Table 20.4. The surface roughness was measured using ISO 1302 with cutoff length of 0.25 mm, sampling length of 0.8 mm, least count 1 µm and average of three readings was taken. Mia et al. [13] have elaborated the importance of Taguchi technique and finding the cumulative distribution function (CDF) for an optimization problem. With aid of Minitab, CDF for surface roughness was derived to determine the distribution function of the considered input parameters, and ANOVA table was calculated for analyzing the variance of the distribution and to identify the most influencing parameter in this study. Figures 20.3 and 20.4 show the cumulative distribution function for surface roughness ‘Ra ’ in LN2 and CO2 environment. The stepped lines are connected to show the Table 20.4 Measured Ra for LN2 and CO2 environment Exp. No.

Speed N (m/min)

Feed rate f (mm/rev)

Depth of cut DC (mm)

Ra (µm) LN2

Ra (µm) CO2

1

60

0.04

0.5

1.7621

1.1196

2

60

0.06

0.75

1.1748

0.5802

3

60

0.08

1.25

1.3281

0.6048

4

65

0.04

0.75

0.9221

0.5565

5

65

0.06

1.25

0.612

0.7475

6

65

0.08

0.5

0.7915

0.2868

7

70

0.04

1.25

0.6627

0.6451

8

70

0.06

0.5

1.579

2.3157

9

70

0.08

0.75

0.5972

0.4142

20 Experimental Investigation on Machining Parameters …

259

Empirical CDF for Ra(LNG) Normal

100

Mean StDev N

1.050 0.4373 9

Mean StDev N

0.7504 0.6028 9

Percent

80 60 40 20 0

0.0

0.5

1.0

1.5

2.0

Avg. Ra LNG cumulative distribution function for Ra(LNG)

Fig. 20.3 Cumulative distribution function for Ra in LN2

Empirical CDF for Ra(CO2) Normal 100

Percent

80 60 40 20 0

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Avg. Ra CO2 cumulative distribution function for Ra(LNG)

Fig. 20.4 Cumulative distribution for Ra in CO2

fitness of the experimental data, while the curve represents the normal distribution of data taken for the study. From the graph is also clear that the distributive function of ‘Ra ’ in LN2 is better than that of CO2 which means the fitness of data for LN2 is higher when compared to CO2 . The better fitness in empirical distribution function for LN2 cryo-coolant than CO2 is due to the better fluidic properties and heat absorption of the nitrogen. This was well agreed with the results of Jawahir et al. [1] and Ghosh et al. [5] as they have also found that LN2 provides better performance in terms of cooling and lubrication.

260

S. Vignesh and U. Mohammed Iqbal

CDF comparison for CO2 and LNG Normal

Variable Avg. Ra CO2 Avg. Ra LNG

100

Percent

80

Mean 0.7504 1.050

60

StDev 0.6028 0.4373

N 9 9

40 20 0 -1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Data

Fig. 20.5 Combined CDF plot for Ra in LN2 and CO2

The empirically combined (for LN2 and CO2 ) CDF for surface roughness (Fig. 20.5) reveals that the normal distribution considered for the study fits the experimental results very well and also helps in estimating the percentile of surface roughness. The mean surface roughness and standard deviation are shown in Table 20.5. It is noticed from Table 20.5 that around 80% of data falls below 1 µm for both LN2 and CO2 which implies the higher capability of cryo-coolants in producing better surface finish due to their reachability to the machining areas. The result goes well with the previous research work conducted by Pereira et al. [2] and Ghosh et al. [5]. The coefficient of variation (C v ) is 41.6% and 80.33% for LN2 and CO2 , respectively. This implies a minor variation of obtained data. This variation is attributable to the difference in surface roughness values obtained in different cryogenic conditions. As suggested by Kanase et al. [8], the chosen input parameters in this study are proved to be significant as there is a minor variation in the obtained Ra . The response table for S/N ratios for surface roughness is listed in Tables 20.6 and 20.7 for LN2 and CO2 , respectively, which evaluates the contribution of factor. Table 20.6 reveals that the cutting speed is the most influencing parameter followed by depth of cut and feed rate. This shows by increasing the cutting speed there is a direct influence in surface roughness. The increasing trend of surface roughness during machining was found when the cutting speed was increased this may be attributed due to the build-up edge (BUE) formation during machining with LN2 as cryogenic coolant. Kramar et al. [14] have explored the cutting parameters during Table 20.5 CDF parameters for normal distribution

LN2

CO2

Mean

1.050 µm

0.7504 µm

Std. Dev

0.4373 µm

0.6028 µm

N

9

9

Cv %

41.6

80.33

20 Experimental Investigation on Machining Parameters … Table 20.6 Response table for Ra in LN2

Table 20.7 Response table for Ra in CO2

261

Levels

Cutting speed (m/min)

Feed rate (mm/rev)

Depth of cut (mm)

1

−2.9282

−0.2142

−2.3193

2

2.3335

−0.4009

1.2609

3

1.3277

1.3480

1.7913

Delta

5.2616

1.7489

4.1106

Rank

1

3

2

Levels

Cutting speed (m/min)

Feed rate (mm/rev)

Depth of cut (mm)

1

2.70498

2.63897

0.85784

2

6.15563

−0.01247

5.82498

3

1.38987

7.62399

3.56767

Delta

4.76576

7.63647

4.96714

Rank

3

1

2

machining hard material, and their result shows the surface integrity gets altered during at high speeds. Dilip et al. [15] have already reported the influence of cutting speed when machining in similar LN2 condition. Table 20.7 shows the S/N response for Ra in CO2 condition is studied. Here, the most influencing parameters were found to be feed rate followed by depth of cut and cutting speed. The results obtained by Hong et al. [3] suggest that though LN2 offers better surface finish in terms of surface roughness and cooling at lower cutting speeds, CO2 coolant shows better performance at higher cutting speeds and moderate feed rates due to absence of build-up Edge (BUE) formation during turning. The contribution Table 20.6 reveals the concurrence of the results obtained with the above statement. Further, study on CO2 reveals that depth of cut also plays a role in machining which goes well with the above result as it is second most influencing parameter. As suggested by Ghosh et al. [5], this may be attributed to the tool material and tool signature taken into study which can be carried out in future research. From both the observations, it is clearly visible that in cryogenic machining all the considered parameters have equal contribution in attaining better surface finish with cutting speed and feed rate as primary influencing parameter and depth of cut as the secondary influencing parameter. Figures 20.6 and 20.7 are the graphical representations (main effect plots) of changes in performance characteristics with the variation in machining parameters. From the main effects, plot for S/N ratios of average Ra (LN2 ) and (CO2 ) peak points are chosen as the optimum levels of machining parameters as suggested by Mia et al. [13].

262

S. Vignesh and U. Mohammed Iqbal

Main Effects Plot for SN ratios Data Means

3

CUTTING SPEED m/min

Depth of cut

Feed rate mm/rev

Mean of SN ratios

2 1 0 -1 -2 -3 60

65

70

0.04

0.06

0.08

0.50

0.75

1.25

Signal-to-noise: Smaller is better

Fig. 20.6 Main effects plot for avg Ra (LN2 )

Main Effects Plot for SN ratios Data Means

8

CUTTING SPEED m/min

Depth of cut

Feed rate mm/rev

Mean of SN ratios

7 6 5 4 3 2 1 0 60

65

70

0.04

Signal-to-noise: Smaller is better

Fig. 20.7 Main effects plot for avg Ra (CO2 )

0.06

0.08

0.50

0.75

1.25

20 Experimental Investigation on Machining Parameters …

263

Table 20.8 Optimum machining parameters Cutting environment

Cutting speed (m/min)

Feed (mm/rev)

Depth of cut (mm)

Ra (µm)

LN2

65

0.08

1.25

0.926

CO2

65

0.08

0.75

1.125

Table 20.9 Confirmation experiment for optimized levels

Cutting environment

Cutting speed (m/min)

Feed rate (mm/rev)

Depth of cut (mm)

LN2

67

0.08

1.10

CO2

63

0.08

0.89

The peak points observed for LN2 cryo-coolant are 65 m/min, 0.08 mm/rev and 1.25 mm. Similarly, for CO2 coolant, it is observed as 65 m/min, 0.08 mm/rev and 0.75 mm. The main effect plots reveal that the cutting speed and feed rate are influencing in similar fashion irrespective of environment with depth of cut influencing lesser when compared to other two parameters. As higher speeds are influential in tribological behavior of the material during machining, the above the result can be attributed to this reason. This was in convergence with the findings done by Umbrello et al. [4] and Dilip et al. [15] where they suggest the depth of cut to be the least influent in cryogenic study.

20.3.2 Optimum Machining Parameters From the responses, the optimum set of machining parameters was found as given in Table 20.8 such as level three of depth of cut, level two of spindle speed, level three of feed rate for LN2 , level two of depth of cut, level two of spindle speed and level three of feed rate for CO2 . Confirmation experiments were conducted to obtain the optimum surface roughness, and the readings obtained are shown in Table 20.9. It reveals that there is a major similarity between the predicted and experimented machining levels with 5% deviation.

20.3.3 Analysis of Variance for Ra In orthogonal experiment, ANOVA is used to calculate the response magnitude in percentage and also show the variance of the experimental analysis. Tables 20.10 and 20.11 evaluate the contribution of factors. The P-magnitude of the control factors

264

S. Vignesh and U. Mohammed Iqbal

Table 20.10 ANOVA for Ra in LN2 environment Source

DF

Adj SS

Adj MS

F-value

P-value

Cutting speed (m/min)

2

0.669

0.334

32.54

0.722

Feed rate (mm/rev)

2

0.093

0.046

21.36

0.237

Depth of cut

2

0.503

0.251

29.91

0.144

Error

2

0.263

0.131

Total

8

1.529

Table 20.11 Model summary for Ra (LN2 )

S

R-sq.

R-sq. (adj)

R-sq. (pred.)

0.362863

82.78%

71.14%

72.00%

declares the statistical significance to the confidence level of 95%. The F-value of the cutting speed, possessing a value of 32.54 establishes itself as the most significant factor, while other factors have insignificant effects. The obtained results converge with findings done by Pereira et al. [2] in predicting minimum surface roughness. The P-value infers that all the variables have statistical significance (as P < 0.05). For LN2 environment, cutting speed influences 73% of the process followed by depth of cut and feed rate. The factor significance implies that the considered factors were influencing the current study. This is very well agreed with the previous works done by Ezugwu et al. [6] and Stoic et al. [7] as they suggest the key machining parameters such as speed, feed rate and depth of cut to be most influencing when performing machining under constrained conditions. Similarly, an ANOVA table was evaluated for LN2 environment which is shown in Tables 20.12 and 20.13. The F-value of the feed rate, possessing a value of 65.40 establishes itself as the most significant factor, while other factors have insignificant effects. Revelations by Table 20.12 ANOVA for Ra in CO2 environment Source

DF

Adj SS

Adj MS

F-value

P-value

Cutting speed (m/min)

2

0.5376

0.2688

25.82

0.459

Feed rate (mm/rev)

2

0.9160

0.4580

65.40

0.526

Depth of cut

2

0.8765

0.4382

47.34

0.427

Error

2

0.6536

0.3268

Total

8

2.9836

Table 20.13 Model summary for Ra in CO2 environment

S

R-sq.

R-sq. (adj)

R-sq. (pred.)

0.52863

82.78%

71.14%

72.00%

20 Experimental Investigation on Machining Parameters …

265

Cryogenic coolant sprayed through nozzle

Fig. 20.8 Cryo-coolant sprayed to the machining area through nozzle

Dilip et al. [15] in suggesting the advantage of CO2 in resisting the formation of BUE at higher speeds may be attributed to the above result as cutting speed is least influencing in CO2 machining. The percentage contributions for CO2 environment are feed rate with more than 60% contribution followed by depth of cut and cutting speed. In ANOVA, R-value is found to be more than 70% which indicates the significance of chosen parameters and their levels in this study. The cryo-coolant device attached to a nozzle is shown in Fig. 20.8. The designed device showed better performance during machining. Nozzle diameter, angle of inclination and coolant pressure are few parameters that can be taken into consideration for future research.

20.4 Conclusions The results of experimental analyses and Taguchi optimization conducted to analyze the effects of cryo-coolants on surface roughness in turning of Hastelloy C 276 are presented below (i)

For LN2 environment: • The most influencing factor for surface roughness is cutting speed and its contribution to the surface roughness of material is 72.5%. • To obtain minimum surface roughness, the parameters obtained for best quality are cutting speed = 65 m/min feed rate = 0.08 mm/rev depth of cut = 1.25 mm.

(ii)

For CO2 environment: • The most influencing factor for surface roughness is feed rate and its contribution to the surface roughness of material is 52%.

266

S. Vignesh and U. Mohammed Iqbal

• To obtain minimum surface roughness, the cutting speed of 65 m/min, feed rate of 0.08 mm/rev and depth of cut of 0.75 mm are found to be a better combination. The surface roughness obtained was 0.926, 1.125 µm for LN2 and CO2 , respectively. (iv) Better material removal rate and significant increase in tool life were observed during cryogenic machining. (v) Comparing the liquid nitrogen and carbon dioxide cryo-coolants, LN2 gives better surface finish. (vi) Overall, cryo-coolants were found to be efficient and environment friendly when compared to conventional machining coolants. (vii.) Further, research can be tried on other machining process like milling and grinding. (iii)

References 1. Jawahir, I.S., Puleob, D.A., Schoopa, J.: Cryogenic machining of biomedical implant materials for improved functional performance, life and sustainability. Sci. Dir. Proc. CIRP 46, 7–14 (2016) 2. Pereira, O., Rodriguez, A., Barrerio, J.: Cryogenic and minimum quantity lubrication for an eco-efficiency turning of AISI 304. J. Clean. Prod. 139, 440–449 (2016) 3. Hong, S.Y., Markus, I., Jeong, W.: New cooling approach and tool life improvement in cryogenic machining of titanium alloy Ti–6Al–4V. Int. J. Mach. Tools Manuf. 41, 2245–2260 (2001) 4. Umbrello, D., Micari, F., Jawahir, I.S.: The effects of cryogenic cooling on surface integrity in hard machining: a comparison with dry machining. CIRP ANNALS 61(1), 103–106 (2012) 5. Ghosh R, Zurechi Z, Frey JH (2003) Cryogenic machining with brittle tools and effects on tool life. In: ASME 2003 International Mechanical Engineering Congress and Exposition, pp 853–865 6. Ezugwu, E.O.: Key improvements in the machining of difficult-to-cut aerospace superalloys. Int. J. Mach. Tools Manuf 45, 1353–1367 (2005) 7. Stoic, A., Lucic, M., Kopac, J.: Evaluation of the stability during hard turning. J. Mech. Eng. 52(11), 723–737 (2006) 8. Tanaji, K., Jada, D.B.: Enhancement of surface finish for CNC turning cutting parameters by using Taguchi method. Indian J. Res. 3(5), 88–91 (2013) 9. Shokrani, A., Dhokia, V., Newman, S.T.: Investigation of the effects of cryogenic machining on surface integrity in CNC end milling of Ti–6Al–4 V titanium alloy. J. Manuf. Process. 21, 172–179 (2016) 10. Kaynak, Y., Lu, T., Jawahir, I.: Cryogenic machining-induced surface integrity: a review and comparison with dry, MQL, and flood-cooled machining. Machin. Sci. Technol. 18(2), 149–198 (2014) 11. Paul, N.E.E., Marimuthu, P., Venkatesh Babu, R.: Machining parameter setting for facing En8 steel with TNMG insert. Am. Int. J. Res. Sci. Technol. Eng. Math. 3(1), 87–92 (2013) 12. Nagnath, P.S., Pimpal Gaonkar, D.S., Laxman Rao, A.S.: Optimization of process parameters in CNC turning machine. In: 10th IRAJn International Conference, 27th October 2013, Tirupati, India. ISBN: 978-93-82702-368 (2013) 13. Mia, M., Dhar, N.R.: Optimization of surface roughness and cutting temperature in highpressure coolant-assisted hard turning using Taguchi Method. Int. J. Manuf. Technol. https:// doi.org/10.1007/s00170-016-8810-2 (2017)

20 Experimental Investigation on Machining Parameters …

267

14. Kramar, D., Kopac, J.: The high performance manufacturing aspect of hard-to-machine materials. J. Adv. Prod. Eng. Manag. (APEM) 1–2, 3–14 (2009) 15. Dilip Jerold, B., Pradeep Kumar, M.: Experimental comparison of carbon-dioxide and liquid nitrogen cryogenic coolants in turning of AISI 1045 steel. Cryogenics 52, 569–574 (2012)

Chapter 21

Machining of EN-31 Steel and Experimental Analysis of Various Process Parameters Using Minimum Quantity Lubrication Ashutosh Saini

and S. K. S. Yadav

Abstract The paper presents an experimental study of surface roughness and tool wear in turning operation of EN-31 alloy steel with ceramic tool under different cutting parameters. The investigation shows impacts of three distinctive cutting parameters in particular cutting velocity, feed rate, and depth of cut influencing surface roughness just as tool wear amid turning task of EN-31 alloy steel. The exploratory setup comprises of three diverse cutting rate, feed rate, and depth of cut for turning of EN-31. EN-31 is chosen as the work material in view of its more application. EN-31 has been utilized for ball and roller bearing, turning instruments, beading rolls, punches, and dies, and by its character, it has high opposition nature against wear and can be utilized for parts abstract to serious abrasion wear or high surface loading. From exploratory investigation, it is seen that turning of hardened steel with minimum quantity lubrication (MQL) gives better surface roughness and tool wear when contrasted turning without MQL. The outcomes demostrated that when MQL is connected in machining procedures can diminish the cutting temperature and gives lubrication between tool and workpiece. These lead to longer tool life and enhanced surface quality. Keywords MQL · Turning · Steel · Tool wear and surface roughness

21.1 Introduction Machining is an essential process of finishing by which jobs are produced to desired dimensions and surface finish by gradually removing the excess material from the preformed blank in the form of chips with the help of cutting tools moved past the work surface. Machining is a removal process. It consists of the shaping of a part through removal of material. The material of cutting tool is harder than the workpiece being formed. Hard turning is known machining process with cutting inserts made A. Saini · S. K. S. Yadav (B) Department of Mechanical Engineering, Harcourt Butler Technical University, Kanpur 208002, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_21

269

270

A. Saini and S. K. S. Yadav

from either cubic boron nitride (CBN) or ceramic. The fundamental variables which are mostly used in the machining processes are cutting speed or velocity [V ], ft/min or in/sec, feed [f ], in/rev or mm/rev, and depth of cut [d], in. or mm [1]. These three variables have a major effect on material removal rate [MRR], which has a main role in determining the power requirements. In hard turning process, minimum quantity lubrication has shown useful effect in reduction of temperature and lubrication to tool and workpiece. With the help of this method, tool life and surface roughness quality can be improved. MQL is a method which sprays little amount of cutting fluid to the cutting zone area with the help of compressed air. The two mixtures give the combination of mist air and utilize it between the surface of tool and workpiece. In this manner, it can be seen that MQL is environment friendly if properly utilized [2]. To improve the surface from microlevel to nanolevel, nanofluid is useful in the machining process [3].

21.1.1 Tool Wear Tool wear describes the gradual failure of cutting tools due to regular operation. There are various mechanisms of wear, e.g., abrasion wear, adhesion wear, and diffusion wear. Abrasion and adhesion are primarily responsible for the flank wear, whereas diffusion may play an important role in the development of crater wear at high speed. Tool wear is the unavoidable cause tool failure in the machining process. Tool life is the duration of actual cutting time after which the tool is no longer usable. The extent of tool wear has a strong effect on dimensional accuracy and surface finish obtained. According to the cutting, tool wear is classified as: Flank wear, crater wear, and corner wear.

21.1.2 Selection of Cutting Tool Since the machining operation is basically a deformation process of the work material through the application of a force by the cutting tool, the stability of the geometric form (form stability) of the tool is a key factor. One basic point is that the tool must be harder than the work material. The materials commonly used for making the cutting tools are high carbon steel, high-speed steel (HSS), cemented carbide, and ceramic. Cutting tools utilized for hard turning require high hardness, high compressive strength, and high resistance to abrasive wear, thermal resistance, and chemical stability at elevated temperatures. Different cutting inserts are available, for example, cubic boron nitride (CBN), ceramic, and cermet. Carbides have very low tool life though CBN displays maximum tool life for hard turning operation. Ceramics are non-metallic and inorganic solids. Ceramics having high melting points so they are heat resistant, great hardness and strength and have low electrical and thermal conductivity, so they are good insulators. They are hard-wearing, chemical inertness

21 Machining of EN-31 Steel and Experimental Analysis of Various …

271

so they are unreactive with other chemicals. So, ceramic is chosen as an inserts in the present work, to approve their utility in hard turning under MQL conditions.

21.1.3 Minimum Quantity Lubrication Cutting fluid have been utilized effectively in machining to broaden device life of the cutting apparatus; however, the issues identified with condition, well-being, and assembling cost recommend reducing their utilization at whatever point conceivable. Least amount fluid (MQL), which splashes little measure of cutting liquid (in the scope of roughly 10–100 mL/h) to the cutting zone region with the assistance of packed air, was an option for this. Vegetable oils are considered for use in MQL because of their great oil and high-weight execution. This examination demonstrates the execution of fired cutting apparatuses in terms of tool life under MQL with stream rate of 50 mL/h utilizing castor oil as the cutting liquid. Minimum quantity lubrication is a recent method introduced in machining to obtain safe, environmental, and economic advantages by decreasing the utilization of coolant lubricant fluids in metal cutting. Accordingly, utilizing MQL strategy brings in a remarkable reduction of machining costs by decreasing the amount of lubricant utilized as a part of machining. In MQL, a little amount of oil stream rate (Up to 100 mL/h) is used. Most importantly, it is imperative to blend the air and the lubricant to persuade the blend to be spread on the cutting surface. Two distinctive blending methods can be used (i) blending inside the nozzle and (ii) blending outside the nozzle. In the first technique, the pressurized air and the cutting fluid are combined inside the nozzle by a blending device. The mixing oil is gotten by the mixture of oil and compressed air, while a negligible cooling activity is accomplished by the pressurized air that accomplishes the cutting surface. A few points of interest are determined after applying this method. Fog and perilous vapors are decreased and the blend setting is very easy to control [4]. In hard turning process, we used servo cut oil as the lubrication. A compact compressor of limit 50 L was associated to inlet of flow control unit by pipe of 8 mm diameter. In flow control unit, a regulating valve was utilized to control the amount of compressed air according to the requirement. Another pipe was associated with the outlet of flow control unit. The other end of this pipe was fixed to nozzle. A pressure gauge was associated just before nozzle keeping in mind to check pressure of compressed air at the exit. Cutting fluid was put in a bottle which was associated with a nozzle. Angle of nozzle was kept at 90° with the axis of workpiece. The proportion of oil (servo cut S water-based cutting oil) to water was kept 1:20. In dry cutting, material was turned without utilization of any cutting liquid. While in MQL, cutting material was turned with the utilization of cutting fluid.

272

A. Saini and S. K. S. Yadav

21.2 Literature Review Gowd et al. [1] shown that EN-31 is picked as the work material in light of its wide application as material for ball and roller bearing, turning devices, beading moves, and punches and dies, and by its character, it has high protection nature against wear and can be utilized for parts subjected to extreme scraped area, and wear or high surface stacking. Chaudhari et al. [2] said that the estimation of flank wear at which ordinary power begins expanding quickly may rely upon work and cutting device materials and cutting conditions. Patole and Kulkarni [3] described that the point of this exploration work is centered on improvement of process parameters under minimum quantity lubrication (MQL) utilizing nanoliquid in turning of AISI 4340.

21.3 Experimental Procedure The machining of hardened steel was performed in CNC lathe machine. EN31 hardened steel was turned under two conditions which are dry cutting and MQL cutting. In dry cutting, material was turned without utilization of any cutting liquid. While in MQL, cutting material was turned with the utilization of cutting fluid.

21.3.1 Hard Turning with MQL After the entire setup in CNC lathe machine, hardened steel was machined utilizing MQL. The compressed air which was obtained utilizing electric power air compressor was adjusted to 4 bar pressure with the help of pressure regulator and was kept constant throughout the experiment. A cutting fluid was also supplied simultaneously to the nozzle. Cutting fluid and compressed air were mixed inside a nozzle and form a mist solution at the exit. This atomized solution was sprayed precisely between tool and workpiece during turning.

21.3.2 Surface Integrity Test Surface roughness This test was performed to check the roughness (Ra ) of finished surface. There are two methods for numerical evaluation of surface roughness: Center line average (CLA) The center line average method is defined as the average value of the ordinates between the surface and the mean line, measured on both side of it. According to

21 Machining of EN-31 Steel and Experimental Analysis of Various …

273

Indian Standards, the surface finish is measured in terms of “CLA” values and it is denoted by Ra . CLA value or Ra =

a + b + c + d ··· + n n

where a, b, c, d… are the ordinates measured on both sides of the mean line and n is the number of ordinates. Root mean square (RMS) The root mean square average method is defined as the square root of the arithmetic mean of the squares of the ordinates. Mathematically,  RMS value Rq =

a 2 + b2 + c2 + d 2 · · · + n 2 n

where a, b, c, d…. are the measured ordinates and n is the number of ordinates. Roughness was measured for each workpiece at three different points. Then average was calculated of three points and recorded as Ra value. Range and cut-off set in Surtronic 3P instrument during measurement are 99.99 µm and 0.8 mm.

21.3.3 Tool Wear Measurement Tool wear reduces the tool life, because tools are costly it was necessary to find less amount of tool wear occurred at flank surface and rake surface. Wear occur at flank surface is called flank wear. A USB-connected Dino-lite optical microscope with Dino-capture 2.0 software was used to measure the tool wear. It has the freedom to magnify from 20× to 230×.

21.3.4 Design of Experiment Number of experiments to be performed was chosen by the assistance of Box– Behnken method. Box–Behnken design of experiments was considered because it gives lesser no. of experiments when compared to Taguchi’s method or a full factorial design. It reduced the total experimentation cost by decreasing the quantity of workpiece, and cutting inserts required 15 for the work. This strategy proposed three-level designs for fitting response surfaces. These designs were formed by joining 3 k factorials with incomplete block designs. It can be seen that the Box–Behnken design is a spherical design with all points lying on a sphere’s radius. Likewise, the Box–Behnken design does not contain any point at the vertices of the cubic region made by the upper and lower limits for every variable. Each factor is set at one of

274 Table 21.1 Cutting parameter of Box–Behnken design

A. Saini and S. K. S. Yadav Run

Velocity (m/min)

Feed (mm/rev)

Depth of cut (mm)

1

100

0.06

0.2

2

100

0.16

0.2

3

160

0.06

0.2

4

160

0.16

0.2

5

100

0.11

0.1

6

100

0.11

0.3

7

160

0.11

0.1

8

160

0.11

0.3

9

130

0.06

0.1

10

130

0.06

0.3

11

130

0.16

0.1

12

130

0.16

0.3

13

130

0.11

0.2

14

130

0.11

0.2

15

130

0.11

0.2

the three equally spaced values generally coded as −1, 0, and 1 at least three levels are required for following good [5]. −1: low-level parameter, 0: medium-level parameter, and +1: high-level parameter. Above all, parameters relating to the central point (0, 0, 0) are repeated twice to set up that the experimental data are inside the normal scattering and repeatability is guaranteed. Box–Behnken designs require less treatment combinations, in issues, including 3 or 4 factors. The experiments were designed by considering three cutting parameters, i.e., speed, feed, and depth of cut. Low-level parameters were taken as − 1, medium-level parameters were taken as 0, and high-level parameters were taken as +1. The ranges of speed, feed, and depth of cut were chosen based on lathe machine ability, tool manufacture criteria, and surface finish parameters as per the literature. Now considering velocity as X 1, feed as X 2, and depth of cut as X 3, the following Table 21.1 shows cutting parameters at which experiments will be performed.

21.4 Results and Discussion The hard turning experiments were executed according to the experimental plan. The desired surface finish produced and the flank wear of the tool were measured of the samples (EN-31 workpiece) turned at high speed. The results have been analyzed by regression analysis and response surface approach. We see the impact of the cutting

21 Machining of EN-31 Steel and Experimental Analysis of Various …

275

parameters, i.e., speed, feed, and depth of cut on the surface roughness and flank wear, have been analyzed for hard turning with and without MQL.

21.4.1 Effect of Various Parameters on Surface Roughness Since hard turning is a finishing procedure, the quality of the surface finish remains the most expected result of the procedure. Consequently, to decide the surface finish, the roughness was measured for both cutting conditions, i.e., with and without MQL. Table 21.2 presents experimental result of surface roughness (Ra ) for different combination of cutting conditions. From Table 21.2, it was watched that surface roughness of machined part produced with MQL was superior to the machined part without MQL. In this work, experimental results were used for modeling using response surface roughness methodology [RSM]. Regression analysis was applied to study the impact of cutting conditions on the surface roughness work [6]. A relation between desired output and independent input was accomplished. The relation is represented by: Ra = f (V, f, d) Table 21.2 Experimental results of surface roughness without and with MQL Run

Velocity (m/min)

Feed (mm/rev)

Depth of cut (mm)

Ra without MQL (µm)

Ra with MQ (µm)

Decrease in Ra (%)

1

100

0.06

0.20

0.81

0.61

24.69

2

100

0.16

0.20

0.87

0.62

28.73

3

160

0.06

0.20

0.96

0.81

15.62

4

160

0.16

0.20

0.86

0.75

12.79

5

100

0.11

0.10

0.56

0.29

48.21

6

100

0.11

0.30

0.75

0.46

38.66

7

160

0.11

0.10

0.85

0.68

20.00

8

160

0.11

0.30

0.78

0.68

12.82

9

130

0.06

0.10

0.79

0.59

25.31

10

130

0.06

0.30

0.72

0.47

34.72

11

130

0.16

0.10

0.92

0.66

28.26

12

130

0.16

0.30

0.75

0.71

5.33

13

130

0.11

0.20

0.71

0.48

32.39

14

130

0.11

0.20

0.85

0.6

29.41

15

130

0.11

0.20

0.83

0.51

38.55

276

A. Saini and S. K. S. Yadav

where Ra is the desired response and f is the response function. The response surface equation in second order for three variables is given by: Ra = a11 v2 + a22 f 2 + a33 d 2 + a12 f v + a23 f d + a13 dv + a1 v + a2 f + a3 d + a0 where Ra is desired response and a0, a1 , a2 , a3 , a11 , a22 , a33 , a12 , a13 , a23 are coefficients. Regression analysis was utilized to characterize desired response. On solving and allocating coefficients, a surface roughness equation was formed for both MQL and without MQL conditions. Surface roughness equation when machining was performed without MQL: Ra = 0.0000101852 × speed × speed + 27.6667 × feed × feed − 7.08333 × doc × doc − 0.0266667 × feed × speed − 5 × feed × doc − 0.0216667 × doc × speed + 0.00653519 × speed + 6.05 × doc − 1.32 × feed − 0.2866 Surface roughness equation when machining was performed with MQL: Ra = 0.0000 486111 × speed × speed + 49.5 × feed × feed − 4.625 × doc × doc − 0.0116667 × feed × speed + 8.5 × feed × doc − 0.0141667 − doc × speed − 0.00460556 × speed + 2.88167 × doc − 10.4233 × feed + 0.811644 Figure 21.1a, b shows that influence of cutting speed and feed on roughness during turning with MQL and without MQL. Figure 21.1a indicate that as the increase in cutting speed the surface roughness increases and with increasing the feed surface roughness also increases. Figure 21.1b shows that as the cutting speed and feed increases, the surface roughness increases with lesser value than turning process without MQL.

Fig. 21.1 Influence of cutting speed and feed on surface roughness while turning at 0.1 mm depth of cut a without MQL and b with MQL

21 Machining of EN-31 Steel and Experimental Analysis of Various …

277

Fig. 21.2 Influence of cutting speed and feed on surface roughness while turning at 0.2 mm depth of cut a without MQL and b with MQL

Figure 21.2 a, b shows that influence of cutting speed and feed on roughness during turning with MQL and without MQL. Figure 21.2 a shows that as the cutting speed and feed increases, the surface roughness increases. Further, increase in depth of cut the effect of feed on turning will be more than effect of cutting speed. Figure 21.2b shows that as the cutting speed and feed increases, the surface roughness less increases compared to without MQL. As the depth of cut increases, the effect of feed will be more on surface roughness compared to the effect of cutting speed. It means the effect of feed on 0.2 mm depth of cut is more on surface roughness compared to velocity on 0.1 mm depth of cut. From Fig. 21.2b, it is observed that as the cutting speed and feed increase, the surface roughness increases less compared to without MQL. Figure 21.3a, b show that influence of cutting speed and feed on roughness during turning with MQL and without MQL. Figure 21.3a shows that as the cutting speed increases, the surface roughness increases and as the feed increases, the surface roughness also increases. As the depth of cut increases from 0.2 to 0.3, the effect of feed will be more on surface

Fig. 21.3 Influence of cutting speed and feed on surface roughness while turning at 0.3 mm depth of cut a without MQL and b with MQL

278

A. Saini and S. K. S. Yadav

roughness compared to the effect of cutting speed. It is also observed that when the cutting speed and feed increases, the surface roughness increases less compared to without MQL while turning of workpiece.

21.4.2 Effect of Various Parameters on Tool Wear Tool life depends upon the amount of tool wear occurred during machining. As the wear increases, tool life decreases. The equal volume of material was removed from each turning experiment to compare the flank wear occurred on the insert. Table 21.3 presents experimental result of tool wear (VB) for various combinations of cutting condition. The regression analysis of the flank wear data for various cutting conditions was made. The regression equation for flank wear when machining was performed without MQL is as below: Fw = 0.0000161111 × speed × speed − 5.8 × feed × feed − 17.15 × doc × doc − 0.0446667 × feed × speed − 11.6 × feed × doc − 0.0281667 × doc × speed + 0.00799944 × speed + 11.0202 × doc + 12.8077 × feed − 1.52364

Table 21.3 Experimental results for tool wear without and with MQL Run

Velocity (m/min)

Feed (mm/rev)

Depth of cut (mm)

Flank wear without MQL (mm)

Flank wear with MQL (mm)

Decrease in flank wear (%)

1

100

0.06

0.20

0.721

0.716

0.69

2

100

0.16

0.20

1.261

0.656

47.97

3

160

0.06

0.20

0.911

0.648

28.86

4

160

0.16

0.20

1.183

0.621

47.50

5

100

0.11

0.10

0.77

0.32

58.44

6

100

0.11

0.30

0.813

0.789

2.95

7

160

0.11

0.10

1.08

0.791

26.75

8

160

0.11

0.30

0.785

0.653

16.81

9

130

0.06

0.10

0.73

0.543

25.61

10

130

0.06

0.30

0.661

0.61

7.71

11

130

0.16

0.10

1.121

0.657

41.39

12

130

0.16

0.30

0.82

0.817

0.36

13

130

0.11

0.20

0.813

0.382

53.01

14

130

0.11

0.20

0.85

0.485

42.94

15

130

0.11

0.20

1.394

0.783

43.83

21 Machining of EN-31 Steel and Experimental Analysis of Various …

279

Flank wear equation when machining was performed with MQL is: Fw = 0.0000509722 × speed × speed + 25.75 × feed × feed + 4.2375 × doc × doc + 0.0055 × feed × speed + 4.65 × feed × doc − 0.0505833 × doc × speed − 0.00277444 × speed + 5.06683 × doc 6.725 × feed + 0.428772 Figure 21.4a, b shows the influence of cutting speed and feed on roughness during turning with MQL and without MQL. Figure 21.4a depicts that with the increase in cutting speed, the tool wear increases and similarly with increase in feed, tool wear increases. From Fig. 21.4b, it is understood that as the cutting speed increases, the tool wear increases less and as the feed increases, tool wear increases less as compared to without MQL. Figure 21.5a, b shows the influence of cutting speed and feed on roughness during turning with MQL and without MQL. Figure 21.5 a depicts that as the increase in cutting speed and feed the tool wear more increases as compare to 0.1 mm of depth

Fig. 21.4 Influence of cutting speed and feed on tool wear while turning at 0.1 mm depth of cut. a Without MQL and b with MQL

Fig. 21.5 Influence of cutting speed and feed on tool wear while turning at 0.2 mm depth of cut. a Without MQL and b with MQL

280

A. Saini and S. K. S. Yadav

Fig. 21.6 Influence of cutting speed and feed on tool wear while turning at 0.3 mm depth of cut. a Without MQL and b with MQL

of cut. As the depth of cut increases from 0.1 to 0.2 mm, the effect of feed will be more on tool wear as compared to the effect of velocity. Figure 21.5b shows that with increase in cutting speed and feed, tool wear increases less compared to turning process without MQL. Figure 21.6a, b shows the influence of cutting speed and feed on roughness during turning with MQL and without MQL. From Fig. 21.6a, it can be understood that on increasing the cutting speed beyond a certain limit, tool wear begins decreasing. This was happened due to thermal softening and straining. Increment in surface temperature of workpiece makes it soft. In this manner, turning of soften material applies a lesser tool force which results about diminishing of tool wear. On increasing feed, tool wear also gets decreasing and this may again be attributed to thermal softening of the workpiece. From Fig. 21.6b, it can be concluded that by increasing the cutting speed and feed, tool wear increases less as compared to turning process without MQL.

21.5 Conclusions Hard turning on CNC Lathe machine have been performed, the results and investigation of finish hard turning of EN 31 steel with ceramic inserts with and without Minimum Quantity Lubrication have been done. After experiments it is found that cutting speed and feed rate have the most significant effect over tool wear rate followed by depth of cut. Surface roughness increases with increase of feed rate and depth of cut but decreased with cutting speed. The feed rate has the most dominant effect on surface roughness. Hence, smaller values of feed rate and depth of cut must be selected in order to achieve better surface finish during steel turning operation.

21 Machining of EN-31 Steel and Experimental Analysis of Various …

281

Following conclusions could be made. 1. Surface roughness increases with the increase of cutting speed and feed rate but depth of cut affects the surface roughness in negligible manner. The feed rate has the most dominant effect on surface roughness. Hence, smaller values of feed rate and depth of cut must be selected in order to achieve better surface finish during steel turning operation. 2. On increasing cutting speed, surface roughness increases because of thermal softening of the workpiece material. The reason could be given that as the cutting speed increases, surface temperature of workpiece increases which makes it soft. This is called thermal softening. Turning of a softer steel gives rougher surface than hardened steel. 3. On increasing the feed, surface roughness increases for both MQL and without MQL cases. Tool wear will be less when turning of hardened steel done with MQL than without MQL. 4. On increasing the feed, the tool wear increases for both MQL and non-MQL cases. 5. Surface finishes improved mainly due to reduction of wear and damage at the tool tip by the application of MQL. 6. On the basis of experiment, conclude that surface roughness improved 48.21% at 100 m/min velocity, 0.11 mm/rev feed, and 0.10 mm depth of cut in experiment no. 5. But in case of similar experiment no. 13, 14, and 15, surface roughnesses will be different due to hardening of workpiece material. 7. There are best improved tool wear 58.44% at 100 m/min velocity, 0.11 mm/rev feed and 0.10 mm depth of cut in experiment no. 5. But in case of similar experiment no. 13, 14, and 15, tool wear will be different due to hardening of workpiece material. 8. Turning of hardened steel with MQL provides better surface roughness such as decrease in surface roughness 48.21% in experiment no. 5 when compared turning without MQL. In this process, decrease in cutting temperature is the main reason to improve surface roughness with MQL compared to without MQL because MQL provides the benefits mainly by decreasing the cutting temperature, which improves the chip tool interaction and maintains sharpness of the cutting edge. This technique decreases friction between chip and the cutting face of the tool which leads to better surface finish and tool life.

References 1. Gowd, G.H., Goud, M.V., Theja, K.D., Reddy, M.G.: Optimal selection of machining parameters in CNC turning process of EN-31 using intelligent hybrid decision making tools. GCMM 97, 125–133 (2014) 2. Chaudhari, R.G., Hashimoto, F.: Process controls for surface integrity generated by hard turning. 3rd CIRP Conf. Surf. Integr. 45, 15–18 (2016)

282

A. Saini and S. K. S. Yadav

3. Patole, P.B., Kulkarni, V.V.: Optimization of process parameters based on surface roughness and cutting force in MQL turning of AISI 4340 using nano fluid. PMME 5, 104–112 (2016) 4. Boubekri, N., Shaikh, V.: Minimum quantity lubrication (MQL) in machining: Benefits and drawbacks. J. Ind. Intell. Inf. 3, 205–209 (2015) 5. Manohar, M., Joseph, J., Selvaraj, T., Sivakumar, D.: Application of Box Behnken design to optimize the parameters for turning inconel 718 using coated carbide tools. Int. J. Sci. Eng. Res. 4, 620–642 (2013) 6. Abhang, L., Hameedullah, M.: Simultaneous optimization of multiple quality characteristics in turning EN-31 Steel. ICMPC 2, 2640–2647 (2015)

Chapter 22

Effect of Tool Material on Trepanning of CFRP Composites B. R. Jayasuriya , A. Harsha Vardhan

and V. Krishnaraj

Abstract The primary objective of this work is to compare the effect of tool material during trepanning of woven CFRP composites. Composites are commonly used, especially, for their lightweight, stiffness, and strength. Trepanning is a machining operation especially performed for creating an effective hole in CFRP composites. The experiment is performed with two trepanning tools [polycrystalline diamond (PCD) and solid carbide tool (SC)] at different feed and spindle speed. The measured parameters such as thrust force and torque of the tool are measured using dynamometer, and the acoustic emission signal is obtained using AE sensor for analysis. The acoustic emission (AE) indicates the crack propagation occurred during trepanning of CFRP composites. The voltage obtained is high when the shearing of last ply occurs (post-delamination occurs). It is found that the PCD tool material has higher magnitude of thrust force compared to carbide tool material. The delamination factor obtained indicates that the increasing feed increases the delamination factor. The delamination factor is less with the usage of PCD tool at low feed compared to the solid carbide tool. It is found that, the delamination factor can be controlled by using suitable tool material and machining parameters such as spindle speed and feed. Thus, the amount of delamination is found to be small with PCD tool compared to the solid carbide and less feed and high spindle speed is recommended for good dimensional accuracy of hole. Keywords Carbon fiber-reinforced composite · Trepanning · Delamination · Acoustic emission

B. R. Jayasuriya · A. H. Vardhan · V. Krishnaraj (B) Department of Production Engineering, PSG College of Technology, Coimbatore, Tamil Nadu 641004, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_22

283

284

B. R. Jayasuriya et al.

22.1 Introduction The composite material has greater advantage because of its remarkable properties such as light, strong, corrosion resistance, high fatigue strength, and faster assembly and the composite materials are formed by choosing appropriate matrix and reinforcement. Composites can be molded into complex shapes. Fiber-reinforced polymer matrix composites have been used in military, aerospace, and automobile industries due to their remarkable mechanical properties. The binding polymer is often thermoset resin such as epoxy, but other thermoset or thermoplastic polymers, such as polyester, vinyl ester or nylon, are used. Trepanning is a hole-making process in which the outer side (i.e., circumference) of the hole is only removed. Hence, trepanning process leaves a core in the work material and has better dimensional and geometrical accuracy in machining. The total volume of the material removed by the trepanning tool is less, leading to a reduction in the cutting forces and power. Mathew et al. observed that the last few fibers are finely trimmed by the trepanning tool which did not occur while with the use of other tools [1]. The crack propagation begins at the critical feed rate which in turn increases the thrust force is mathematically derived by Mathew et al. [2]. Nagaraja et al. studied cutting parameters such as feed and spindle speed is responsible for the occurrence f delamination which increases delamination factor. The feed rate is observed to be main factor to increase delamination, thrust force, and torque [3]. Tsao et al. observed that the delamination and the accuracy of the hole are determined by the feed and drill diameter. The candle stick drill and saw drill lead to a smaller delamination factor than twist drill [4]. The study by Nagaraja et al. reveals that delamination factor increases with increase in drill diameter and feed rate and decreases with increase in spindle speed. The experimental results reported by Nagaraja et al. result that TiN-coated solid carbide drills are preferable to uncoated solid carbide drills, for achieving minimum drilling delamination [5]. Kanagaraju et al. analyzed the parameters such as thrust force and delamination factor depends on the drill point angle and the feed rate [6]. Increase in feed rate has greater influence in delamination than increase in cutting speed. The use of pilot hole has resulted in delamination reduction which is analyzed by Marques et al. [7]. Velu et al. observed that medium spindle speed with low feed gives better hole quality characteristics for medium-sized diameter drill. It is also observed that the feed rate is found to be the most influencing factor than compared to speed and drill diameter [8]. Chadha et al. identified that the delamination (damage) increases with increase in the feed and the spindle speed. The damage also decreased with increase in the number of layers in a composite [9]. Durão et al. concluded that the feed rate is the influential parameter to increase the delamination [10]. The critical thrust forces at propagation of the delamination zone due to drilling of the infinite FRP/metallic strip are derived by Kim et al. [11]. The study performed by Shetty et al. [12] observed that the spindle speed, feed, coating made on the tool, and pint angle of the drill tools have major influence compared to the drill diameter. The results analyzed by Satyanarayana et al. reveal that the minimum delamination

22 Effect of Tool Material on Trepanning of CFRP Composites

285

damage is obtained at higher spindle speed, lower drill diameter, and lower feed rate [13, 14]. Nafiz concluded that it is possible to decrease delamination factor to minimum with varying feed rate in hole exit of CFRP composites [15]. Ragunath et al. [16] observed that the delamination factor is high at the exit point of the drill bit and less in the entry position the because of push delamination at the exit. Tate et al. [17] delamination factor is influenced by parameters such as feed, point angle, chisel edge width, and speed. This paper focuses on the effect of tool material on thrust force and torque along with delamination factor by considering two different tool materials. The AE RMS value is obtained during trepanning and analyzed for identifying the crack propagation of the last ply of the composites.

22.2 Experimental Design The T300 woven CFRP laminate is made by prepreg tape (compression molding process) supplied by Hexcel Composites. The thickness of the ply is 0.25 mm, and the laminate of 2 mm is obtained by laying eight plies one over the other and cured using autoclave. The volume fraction of the CFRP is 54%. The trepanning tools (PCD and solid carbide) used for the study are ground accurately in tool and cutter grinder and its dimensions are measured using profile projector and coordinate measuring machine (CMM). Experimentation is made using Makino vertical machining center—Model S33 (VMC) along with acoustic emission sensor and drill dynamometer. Further, the measurement of delamination in the holes is taken using Dino-Lite digital microscope and further analysis is performed and results are generated. The trepanning tools used in the machining process are shown in Fig. 22.1. The diagrammatical representation of the measuring setup made in CMM is shown in Fig. 22.2. The parameters measured using both optical profile projector and CMM machine are listed in Table 22.1. The schematic diagram of the experimental setup is shown in Fig. 22.3. At first, the workpiece is clamped and mounted on the drill dynamometer which is placed below the spindle of VMC. The dynamometer is then connected to amplifier and finally to the PC. The AE sensor is fixed to one end of the workpiece in order to Fig. 22.1 Trepanning tool used for the study

286

B. R. Jayasuriya et al.

Fig. 22.2 Measuring dimensions in CMM

Table 22.1 Trepanning tool parameters

Parameter

Value

Axial rake angle

0°

End clearance angle

8°

Side clearance angle

5°40

Radial rake angle

3°

Outer diameter

12 mm

Inner diameter

6 mm

absorb the AE signal. The data obtained during machining are processed and stored using drill dynamometer software and DEWE software. The experimental design made for the machining is presented in Table 22.2.

22.3 Experimental Procedure The CNC code for drilling through holes is generated in the vertical milling center and edited for different machining parameters using FANUC control panel. During trepanning operation, the thrust force and torque data are recorded using software interface with the Syscon tool dynamometer. The AE RMS value is also recorded using DEWE software. The change in magnitude of torque and thrust force is measured every time using drill dynamometer when there is change in magnitude of 0.1 N and the data obtained are stored and then converted into graph using PC. The AE sensor with an accuracy of 0.0001 V is used acquire the AE signal. The workpiece, AE sensor. and the dynamometer setup in the machine are shown in Fig. 22.4.

22 Effect of Tool Material on Trepanning of CFRP Composites

287

Fig. 22.3 Diagrammatic representation the experimental setup Table 22.2 Design of experiment and predicted results S. No.

Tool material

Spindle speed (rpm)

1

PCD

1000

Feed (mm/rev)

Feed rate (mm/min)

AE RMS (V)

Del. factora

Thrust force (N)

Torque (Nm)

0.02

20

0.628

1.274

174.2

0.31

2

0.04

40

0.694

1.301

202.5

0.36

3

0.02

60

0.752

1.345

215.5

0.37

4

0.04

80

0.783

1.376

229.5

0.38

0.02

20

0.595

1.293

147.2

0.23

6

0.04

40

0.771

1.306

184.1

0.26

7

0.02

60

1.597

1.333

196.8

0.36

0.04

80

1.671

1.365

215.5

0.38

0.02

20

0.474

1.303

101.6

0.14

10

0.04

40

0.527

1.314

119

0.20

11

0.06

60

2.049

1.329

130.6

0.22

12

0.08

80

2.150

1.361

135

0.25

5

PCD

2000

8 9

13

Carbide

Carbide

1000

0.02

20

2.356

1.287

91.4

0.17

14

0.04

40

2.526

1.294

108

0.18

15

0.06

60

2.728

1.358

121.5

0.21

16

0.08

80

1.484

1.345

129.5

0.24

a Delamination factor

2000

288

B. R. Jayasuriya et al.

Fig. 22.4 Diagrammatic representation the experimental setup

22.4 Result and Discussion 22.4.1 Analysis of Fiber Pull Out The data obtained from the AE software and the dynamometer are analyzed below. One of the damaging mechanisms observed during trepanning operation is fiber pull out which is a failure mechanism occurs due to force unbalance which appears during machining of composites and crack propagation. This takes place after debonding of fiber and fracture of fiber which takes place away from the crack plane during machining. The fiber pulls out observed with the use of PCD tool is shown in Fig. 22.5. It is also found that increase in feed increases the amount of delamination. The amount of delamination obtained with solid carbide tool is shown in Fig. 22.6. The amount of delamination is found to be higher in the hole made with solid carbide tool compared to PCD tool. Fig. 22.5 Comparison of delamination with the use of PCD tool at 1000 rpm

22 Effect of Tool Material on Trepanning of CFRP Composites

289

Fig. 22.6 Comparison of delamination with the use of solid carbide tool at 1000 rpm

It is also found that with the use of PCD tool, the fiber is finely trimmed, and hence, the amount of delamination is less, whereas the fiber is not fully trimmed with the use of carbide tool.

22.4.2 Analysis of Thrust Force and Torque on Tool Material The tool material has a greater impact on the thrust force and torque. The graphical representation of the tool material on thrust force is shown in Fig. 22.7. The magnitude of thrust force decreases with the use of solid carbide tool material and increases with PCD tool material which increases amount of delamination. This change in the magnitude of thrust force is due to the material properties such as hardness of the tool material. Thus, the change in the hardness of tool material accordingly changes the magnitude of thrust force and torque. The spindle speed affects the magnitude of thrust force because when spindle speed is high, material removal rate is higher; hence, the temperature of the fiber increases which in turn softens the fiber. Therefore, the fiber easily gets cut. The magnitude of thrust force is less when spindle speed is high which reduces the amount of delamination. Increase in feed increases the thrust force in linear manner. The graphical representation of the tool material on thrust force and torque is shown in Fig. 22.8. Increase in feed increases the magnitude of torque. The magnitude 270

Thust force (N)

Fig. 22.7 Effect of thrust force on tool material

220

PCD tool at 1000 rpm PCD tool at 2000 rpm Carbide tool at 1000 rpm Carbide tool at 2000 rpm

170 120 70 0.01

0.02

0.03

0.04

Feed (mm/rev)

0.05

0.06

290

B. R. Jayasuriya et al. 0.4

Fig. 22.8 Effect of torque on tool material Torque (Nm)

0.35 0.3 0.25 0.2

PCD tool at 1000 rpm PCD tool at 2000 rpm Carbide tool at 1000 rpm Carbide tool at 2000 rpm

0.15 0.1 0.01

0.02

0.03

0.04

0.05

0.06

Feed (mm/rev)

of torque is higher with the use of PCD tool than compared to solid carbide tool. The spindle speed has a significant impact on the torque. When spindle speed increases, the magnitude of torque decreases. Hence, lower feed and higher spindle speed are preferable for obtaining better quality hole. When feed increases, the magnitude of thrust force and torque increases. Hence, the increases in magnitude of thrust force leads to higher rate of delamination in the machining process.

22.4.3 Effect of Delamination Factor The delamination factor is that the ratio of maximum diameter of delamination zone (Dmax ) to the original diameter of the hole (D). Generally, the delamination factor is greater than 1 because of the external parameters. The measuring method of delamination factor is shown in Fig. 22.9. Fd =

Fig. 22.9 Measurement of delamination factor

Dmax D

22 Effect of Tool Material on Trepanning of CFRP Composites 245 225

Thrust force (N)

Fig. 22.10 Analysis of thrust force on delamination factor

291

205 185

PCD tool at 1000 rpm PCD tool at 2000 rpm Carbide tool at 1000 rpm Carbide tool at 2000 rpm

165 145 125 105 85 1.25

1.27

1.29

1.31

1.33

1.35

1.37

1.39

Delamination factor

The graphical representation of the delamination factor with thrust force is shown in Fig. 22.10. It is found that the delamination factor increases with increase in thrust force. Increase in thrust force increases the crack propagation which in turn increases the amount of delamination in CFRP composites. Hence, less magnitude of delamination factor is preferable for better quality hole. The effect of tool material on the delamination is shown in Fig. 22.11. The delamination factor is less with the use of PCD tool and high with solid carbide tool at low feed. Since the fiber is trimmed with the use of PCD tool, the delamination factor is less at low feed compared to carbide tool. The delamination factor is high for PCD tool compared to solid carbide tool at higher feed. This is due to the difference in hardness of PCD and solid carbide tool. The graphical representation of the feed on the delamination factor is shown in Fig. 22.11. The graph indicates that the increase in feed increases the delamination factor. When the feed increases, the thrust force and torque increase which initiates the crack propagation in the ply of the CFRP composites. This promotes the increase in amount of delamination. Hence, less feed is recommended in order to reduce the amount of delamination. The graphical representation of the spindle speed on the delamination factor is shown in Fig. 22.11. It is found that the increase in spindle speed decreases the delamination factor. When the spindle speed is high, the cutting speed is also high, so that temperature of the fibers increases and it becomes soft in nature. Hence, the fiber is easily machined rather than the crack propagation in the fiber. Hence, the delamination is less at higher spindle speed. 1.38

Delamination factor

Fig. 22.11 Effect of torque on tool material

1.36 1.34 1.32 PCD tool at 1000 rpm PCD tool at 2000 rpm Carbide tool at 1000 rpm Carbide tool at 2000 rpm

1.3 1.28 1.26 0.015

0.02

0.025

0.03

0.035

0.04

Feed (mm/rev)

0.045

0.05

0.055

292

B. R. Jayasuriya et al. 0.8

signal at which machining occurs

AC RMS (V)

0.7 0.6 0.5

AE RMS Delamination occur

0.4 0.3 0.2 0.1 4.0786 4.2516 4.4246 4.5976 4.7706 4.9436 5.1166 5.2896 5.4626 5.6356 5.8086 5.9816 6.1546 6.3276 6.5006 6.6736 6.8466 7.0196 7.1926 7.3656 7.5386 7.7116 7.8846 8.0576 8.2306

0

Time (s)

Fig. 22.12 AE signal obtained at a feed of 0.04 mm/rev and speed of 1000 rpm

22.4.4 Effect of AE RMS It is found that increase in thrust force increases the AE RMS value, and this is because when the thrust force is higher, the feed is high; hence, the amount of strain energy exhibited during trepanning operation is also high in magnitude. Increase in spindle speed increases the cutting speed which in turn increases the AE RMS magnitude. It is found that the increase in delamination factor, increases the AE RMS magnitude, because when the amount of delamination (fibre pull out), the release of the energy will be higher due to the crack propagation which occur in the plies of CFRP composites. The graphical representation of the AE signal obtained during the trepanning of the CFRP composites with PCD tool is shown in Fig. 22.12. It is found that the maximum voltage is obtained when the trepanning operation begins and gradually reduces. A maximum value of voltage is deducted at the end of machining which indicates the delamination or the crack propagation which is happened in the last ply of the CFRP composite.

22.5 Conclusions Experiment of trepanning on CFRP composite is carried out at various feed and spindle speed with drill dynamometer and AE setup. It is concluded that the tool material has a greater impact on the delamination of composites because of the change in tool material properties, the thrust force and torque also varies, which in turn increases the magnitude of delamination factor and AE RMS signal. The delamination factor is higher in magnitude when the feed is high and low at higher spindle speed. The AE RMS value obtained indicates the crack propagation and the time of occurrence. The crack propagation is occurring when the thrust force exceeds

22 Effect of Tool Material on Trepanning of CFRP Composites

293

the critical value. The AE RMS value can be related to the delamination factor in order to find the thrust force by relating it, and this may lead to usage of AE rather than drill dynamometer for measurement. Therefore, appropriate tool material, low feed, and high spindle speed are used to obtain a better quality hole and the major defect of machining of composites is the fiber pull-out which is less with the use of trepanning tool.

References 1. Mathew, J., Ramakrishnan, N., Naik, N.K.: Investigations into the effect of geometry of a trepanning tool on thrust and torque during drilling of GFRP composites. J. Mater. Process. Technol. 91, 1–11 (1999) 2. Mathew, J., Ramakrishnan, N., Naik, N.K.: Trepanning on unidirectional composites: delamination studies. Composites: Part A 30, 951–959 (1999) 3. Shetty, N., Herbert, M.A., Shetty, D.S., Shetty, R., Shivamurthy, B.: Effect of process parameters on delamination, thrust force and torque in drilling of carbon fiber epoxy composite. Res. J. Recent Sci. 2(8), 47–51 (2013). ISSN 2277-2502 4. Tsao, C.C., Hocheng, H.: Taguchi analysis of delamination associated with various drill bits in drilling of composite material. Int. J. Mach. Tools Manuf 44, 1085–1090 (2004) 5. Herbert, M.A., Shetty, D., Vijay, G.S., Shetty, R.: Evaluation of drilling induced delamination of carbon fiber reinforced polymer composite using solid carbide drills. Eur. Sci. J. 10, 15 (2014). ISSN: 1857-7881 (Print) e-ISSN 1857-7431 6. Kanagaraju, T., Alfred Tennyson, M., Anoop, A.G., Balachandran, A., Edison Power Singh, P. (2016) Influence of delamination factor in drilling of carbon fibre reinforced composite. Int. J. Adv. Res. Trends Eng. Technol. (IJARTET) 3(19) 7. Marques, A.T., Durão, L.M., Magalhães, A.G., Silva, J.F., Tavares, J.M.R.: Delamination analysis of carbon fibre reinforced laminates. In: 16th International Conference on Composite Materials 8. Velu, G.K., Shanmugasundaram, S.M., Velu, C.: Delamination analysis using digital image processing by IMAGE J and LabVIEW for drilling on GFRP composite laminates. Int. J. Res. Appl. Sci. Eng. Technol. 3(IX), 342 (2015) 9. Chadha, V., Gupta, S., Singari, R.M.: Optimization of cutting parameters on delamination using Taguchi method during drilling of GFRP composites. In: Proceedings of the International Multi Conference of Engineers and Computer Scientists 2017, vol. II, IMECS (2017), 15–17 Mar 2017 10. Miguel, P.D., Manuel, R.S., Albuquerque, V.H., Marques, S., Andrade, N.G.: Drilling damage in composite material. Materials 7, 3802–3819 (2014) 11. Kim, G.W., Lee, K.Y.: Critical thrust force at propagation of delamination zone due to drilling of FRP/metallic strips. Compos. Struct. 69, 137–141 (2005) 12. Shetty, D., Rajat, A., Shetty, G., Patil, P.: The effect of machining parameters on drilling induced delamination of carbon fiber reinforced polymer composite. Int. J. Appl. Eng. Res. 12(19), 8286–8293 (2017). ISSN 0973-4562 13. Satyanarayana, K.V., Krishna Prasad, D.V.V.: Optimization of drilling parameters for delamination factor in of hybrid fiber reinforced polymers. IJISET—Int. J. Innov. Sci. Eng. Technol. 3(6), 353 (2016) 14. Shetty, N., Herbert, M.A., Shetty, D.S., Shetty, R. and Murthy, B.R.N.: Investigation into the effects of process parameters on delamination during drilling of BD-CFRP composite using Taguchi design of experiments and response surface methodology. Int. J. Mech. Eng. Technol. (IJMET)

294

B. R. Jayasuriya et al.

15. Ya¸sar, N., Günay, M.: The influences of varying feed rate on hole quality and force in drilling CFRP composite. J. Sci. GU J. Sci. 30(3), 39–50 (2017) 16. Ragunath, S., Velmurugan, C., Kannan, T.: A review of influential parameters in drilling delamination on fiber reinforced polymer composites. Int. J. ChemTech. Res. 10(7), 298–303 (2017) 17. Tate, G.S., Shaikh, A.M., Mane, V.S.: Experimental and analytical investigation of drilling on GFRC material for enhancement of drilling quality: a review. IOSR J. Mech. Civil Eng. (IOSR-JMCE), 55–60

Chapter 23

Effect of Air Delivery Pressure and Flow Rate on Surface Integrity in Minimum Quantity Cooling Lubrication Grinding of Inconel 718 Anirban Naskar , Amit Choudhary , Biddu Bhushan Singh and S. Paul Abstract Minimum Quantity Cooling Lubrication (MQCL) seems to be an effective technique to improve the surface integrity of the ground surface. The performance of MQCL may be affected by different MQCL parameters in combination with lubrication ability and viscosity of the grinding fluid. Therefore, in the present study, Inconel 718 has been ground with neat oil (typical viscosity of 27–42 cSt) and water-based nanofluids (typical viscosity of 0.66 cSt) at different air delivery pressure and fluid flow rate to reveal their effect on the integrity of the ground surface. Ground surface morphology and surface residual stress have been assessed to represent the surface integrity. No visible effect of MQCL parameters on the ground surface morphology and surface residual stress have been observed for all the cutting fluids used. Neat oil produced better ground surface morphology and much lower residual stress as compared to nanofluids throughout the experimental domain. Keywords Nickel-based alloy · MQCL grinding · Surface integrity

23.1 Introduction Nickel-based alloys are one of the most widely used alloys in the aerospace industry. This is primarily because they maintain their strength up to very high temperature, corrosion resistance and fatigue strength [1]. These alloys are used in the making of aeroengine parts, e.g. combustion chamber, turbine blade, low- and high-pressure turbine section, etc. [2]. Around 50% of an aeroengine was made from nickel-based superalloys among which 70% was comprised of Inconel 718 [3]. But, to use these alloys, it needs to undergo machining process, e.g. grinding to maintain final dimensional tolerance and finishing where it finds its difficulty. While layer formation,

A. Naskar (B) · A. Choudhary · B. B. Singh · S. Paul Machine Tool and Machining Laboratory, Department of Mechanical Engineering, Indian Institute of Technology Kharagpur, Kharagpur, West Bengal 721302, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_23

295

296

A. Naskar et al.

surface burn, tensile residual stress, etc., are the major problems that occur while grinding such alloys [4, 5]. Various researchers have done grindability study of Inconel 718 in flood cooling mode with different grinding wheels. cBN wheel of different porosity was attempted by Cai and Rowe [6] in flood cooling mode to study the effect of porosity on the grindability of Inconel 718. It was observed that high-porosity wheel produced better surface due to less wheel loading and proper fluid delivery. cBN wheel was also used by Patil [7] in his work with flood cooling and it was found to be appropriate for Inconel 718 grinding. Two superabrasive wheels (diamond and cBN) and one conventional wheel (alumina) were used in a single work to compare the performance of the wheels in flood cooling [8]. Grinding temperature, force, roughness and grinding ratio were combined as a single performance index to evaluate the grindability of Inconel 718. The diamond wheel was found to be the best among all the wheels. In another study, surface residual stress was found to be highly tensile while grinding Inconel 718 with single alumina (SA) wheel [9]. Residual stress was varied in the range of 700–1000 MPa depending on the downfeed. Limitation of flood cooling comes from the environmental point of view. Therefore, the elimination of flood cooling is a must requirement for green manufacturing. But, dry machining/grinding leads to poor product quality due to excessive temperature generation. Minimum quantity cooling lubrication is a technique that seems to be efficient which can maintain the product quality while using the very small amount of fluid (mL/h) as compared to flood cooling (L/min) [10]. Dry grinding of Inconel 718 was performed by Sinha et al. [11] using conventional grinding wheels (silicon carbide and alumina). Performance of alumina wheel was found to be better than SiC wheel in terms of grinding forces, the coefficient of friction which led to better surface integrity. An attempt was made to optimize the minimum quantity cooling lubrication (MQCL) parameters [12]. Air delivery pressure of 8 bar along with 150 mL/h fluid flow rate was found to be optimum while grinding Inconel 718 using alumina wheel. Nozzle placement was observed to be an important factor that affected the grinding forces significantly. The intermediate and bottom position of the nozzle produced lower grinding forces as compared to top position. Nanofluid was also applied in MQCL mode to improve the grindability of Inconel 718 [13]. Performance of Ag and ZnO nanofluid was found to be better than dry and MQCL soluble oil which was attributed due to better wettability and lubricity of nanofluids. The review of the past literature indicates that some studies have been undertaken on the effect of MQCL flow parameters in nickel-based alloy grinding. However, those studies dealt with single grinding fluid of fixed viscosity. The fluid of different viscosity may work effectively at different MQCL operating parameters. In addition, in the previous studies, only force and roughness were used as an evaluating factor. Our earlier research [14] indicated that during grinding of Inconel 718 with singlelayer electro-plated cBN wheel, flood cooling and neat oil (applied using MQL technique) provided minimum tensile residual stress on the ground surface. Aqueous nanofluids could not lower the tensile residual stress to that extent. But the study was limited to narrow domain of MQL parameters.

23 Effect of Air Delivery Pressure and Flow Rate on Surface …

297

Therefore, in the present study, an attempt has been made to observe the effect of MQCL parameters on the surface integrity of Inconel 718 while using neat oil [typically 27–42 cSt)] and two water-based nanofluids (typically 0.66 cSt) which are having widely different viscosity. Surface residual stress along with ground surface morphology has been studied to represent the surface integrity of the ground product. Nickel-based superalloys are rather difficult-to-grind materials [4] yielding rapid wheel wear as compared to the grinding of ferrous alloys with cBN wheels [15]. Thus, in the present work, the effect of wheel wear has also been investigated on different grindability characteristics particularly with an emphasis on residual stress.

23.2 Experimental Details Grinding was undertaken in plunge surface up-grinding mode using Cosmos EARTH-6030 NC grinding machine. Inconel 718, having a dimension of 8 mm × 8 mm × 25 mm, was used as workpiece material. Single-layer electro-plated superabrasive cBN wheel of grit size 151 µm (ABN 900 grit of Element 6) and a diameter of 200 mm was employed for the experiments. For each experiment, ten passes were carried. To study the effect of wheel wear on the surface integrity, a worn wheel has been used and the surface residual has been compared with the new wheel. Details of grinding parameters, MQCL parameters and grinding environment are listed in Table 23.1. Water-based WS2 and hBN nanofluid of 0.1% volume concentration were used for the experiments. Sodium dodecyl sulphate (SDS) surfactant was used to increase the stability of the nanofluids. Distiled water was used as the dispersing medium. Details about the composition of the nanofluids are provided in Table 23.2. The mixture of distiled water, SDS and nanoparticles were sonicated in ultrasonic probe sonicator for 1 h to obtain a homogenous mixture. Table 23.1 Grinding and MQCL details

Wheel speed

30 m/s

Table speed

2 m/min

Downfeed

10 µm

Environment

Neat oil (HP—Trimofin 23), hBN nanofluid and WS2 nanofluid

Air delivery pressure

3, 4.5 and 6 bar

Fluid flow rate

100, 250 and 400 mL/h

Horizontal stand-off

20 mm

Vertical stand-off distance

15 mm

Nozzle angle inclined to horizontal

10°

298

A. Naskar et al.

Table 23.2 Composition of the nanofluids used Density (g/cm3 )

S. No.

Particle

% volume concentration

Mass of nanoparticle in 100 mL of water (in g)

Mass of surfactant in 100 mL of water (in g)

1

WS2

7.5

0.1

0.7500

0.3750

2

hBN

2.29

0.1

0.2290

0.1145

Grinding forces were measured both in tangential and normal direction using piezoelectric dynamometer (Kistler–9257B). Scanning electron microscope (ZEISS SEM-EVO18) was used to study the ground surface morphology. Residual stress measurement was performed using X-ray diffraction technique (PANalytical EMPYREAN). Cr source was employed for the measurement using the (220) crystallographic plane which had a nominal 2θ value of 129.5°. Five different sin2 ψ was used that varied from 0 to 0.5 both in a positive and negative direction to get the residual stress.

23.3 Results and Discussion 23.3.1 Ground Surface Morphology Figure 23.1 shows the variation of ground surface morphology at different air delivery pressure under different grinding environments. No significant effect of air delivery pressure on the ground surface morphology has been observed for any cooling lubrication environment. hBN and WS2 nanofluids show the presence of surface irregularities at all the air delivery pressures. It seems that the ploughing effect is very much present on these surfaces leading to the generation of surface irregularities. However, in the case of neat oil, clear ground surface has been observed at all the pressures. Grinding marks have been found to be much clear for neat oil as compared to WS2 and hBN nanofluid. Surface irregularities, e.g. ploughing have been found to be almost absent for neat oil. For neat oil also, no clear distinction on surface morphology could be made among different air delivery pressures. Similar observations can also be made while studying the effect of cutting fluid flow rate on surface morphology as shown in Fig. 23.2. Fluid flow rate could not affect the surface morphology for both neat oil and water-based nanofluids. For hBN and WS2 nanofluids, it shows surface with full of irregularities while neat oil produced clean ground surface. Above discussed results can be directly correlated with the lubricating nature of the grinding fluids. The force ratio (F t /F n ) usually indicates the lubricating nature of the particular cutting fluid. Lower the force ratio, better the lubricating nature [16]. Figure 23.3 shows the grinding force ratio (F t /F n ) of different grinding fluids.

23 Effect of Air Delivery Pressure and Flow Rate on Surface …

4.5 bar

6 bar

WS 2

Neat oil

hBN

3 bar

299

Fig. 23.1 Ground surface morphology of different cooling lubricants at different air delivery pressures at a constant fluid flow rate of 250 mL/h while using the new wheel

The force ratio for neat oil is lower than hBN and WS2 nanofluid which implies that neat oil provided better lubrication than nanofluids. It seems that better lubrication provided less ploughing and more cutting action. As an outcome, neat oil could produce better surface than hBN and WS2 nanofluids.

23.3.2 Surface Residual Stress Measured surface residual stress at different air delivery pressure and fluid flow rate for different cutting fluids is shown in Fig. 23.4 for both new wheel and worn wheel. No clear trend has been observed on the effect of air delivery pressure and fluid flow rate on residual stress. For hBN nanofluid, all the parameters show comparable results only when the air delivery pressure increased to 6 bar and flow rate increased to 400 mL/h it shows highest residual stress for both the new and worn wheel. At higher flow rate, agglomeration of hBN nanoparticles may be the cause of

300

A. Naskar et al.

250 ml/hr

400 ml/hr

WS2

Neat oil

hBN

100 m//hr

Fig. 23.2 Ground surface morphology of different cooling lubricants at different fluid flow rates at a constant air delivery pressure of 4.5 bar while using the new wheel

force ratio (Ft/Fn)

0.50

Neat oil hBN WS2

fluid flow rate 250 ml/hr

0.45 0.40 0.35 0.30

(b)

0.50

force ratio (Ft/Fn)

(a)

Neat oil hBN WS2

air delivery pressure 4.5 bar

0.45 0.40 0.35 0.30

3.0

3.5

4.0

4.5

5.0

5.5

air delivery pressure (bar)

6.0

100

200

300

400

500

flow rate (ml/hr)

Fig. 23.3 Tangential to normal force ratio under different cooling lubrication conditions at different a delivery pressure and b flow rates while using new wheel

23 Effect of Air Delivery Pressure and Flow Rate on Surface … 600

residual stress (MPa)

500

worn wheel new wheel

hBN

400 300 200 100 0 3 bar

4.5 bar

6 bar

100 ml/hr 250 ml/hr 400 ml/hr

4.5 bar

250 ml/hr 600

residual stress (MPa)

500

worn wheel new wheel

WS2

400 300 200 100 0 3 bar

4.5 bar

6 bar

100 ml/hr 250 ml/hr 400 ml/hr

4.5 bar

250 ml/hr 600 500

residual stress (MPa)

Fig. 23.4 Variation of residual stress with flow rate and air delivery pressure for nanofluids and neat oil

301

worn wheel new wheel

Neat oil

400 300 200 100 0 3 bar

4.5 bar

250 ml/hr

6 bar

100 ml/hr 250 ml/hr 400 ml/hr

4.5 bar

302

A. Naskar et al.

higher residual stress generation. WS2 nanofluid shows comparable residual stress at all the parameters while using the worn wheel. With the new grinding wheel at 400 mL/h flow rate, it shows lowest residual stress. Much lower residual stress has been observed while grinding with neat oil at all the air delivery pressure and fluid flow rate. These results are well aligned with the ground surface morphology that has been discussed in the previous section. It seems that better lubrication ability of the neat oil promoted more cutting action, which resulted in lower residual stress. Another interesting observation has been made on the effect of wheel wear on residual stress. For hBN and WS2 nanofluids, worn wheel produced more tensile residual stress whereas, for neat oil, worn wheel generated less tensile residual stress (Fig. 23.4) which is beneficial for product life. With the increase in wheel wear, both tangential and normal force increases. Higher tangential force leads to higher grinding zone temperature which results in more tensile stress (or less compressive) and higher normal force indicates more mechanical action which results in more compressive (or less tensile) residual stress. As with the increase in wheel wear both the forces increase, the state of residual stress would depend on the dominant factor. Therefore, to explain the difference in residual stress between the new and worn wheel for different cutting fluids, a relative force ratio (RFR) has been proposed as shown below: Relative Force Ratio (RFR) =

Fn worn wheel −Fn new wheel Fn new wheel Ft worn wheel −Ft new wheel Ft new wheel

This RFR represents the relative increase in normal force to relative increase in tangential force due to wheel wear. Figure 23.5 shows the calculated ratio (RFR) for all the cutting fluids. It can be clearly seen that for neat oil, the ratio is more than 1 at all the parameters whereas for nanofluids it is mostly less than one. It indicates that an increase in normal force for neat oil is more than the tangential force which Fig. 23.5 Variation of ratio RFR with different MQCL parameters for different cutting fluids

hBN WS2 Neat oil

1.6 1.4 1.2

RFR

1.0 0.8 0.6 0.4 0.2 0.0

3.0 bar

4.5 bar

250 ml/hr

6 bar

100 ml/hr 250 ml/hr 400 ml/hr

4.5 bar

23 Effect of Air Delivery Pressure and Flow Rate on Surface …

303

resulted in lowering the residual stress for the worn wheel as compared to the new wheel. For nanofluids, increase in tangential force is more than the normal force which resulted in higher tensile residual stress.

23.4 Conclusions In this work, an attempt has been made to study the effect of different MQCL parameters on the surface integrity of Inconel 718 while grinding with superabrasive cBN wheel. Water-based nanofluids and neat oil have been used as cutting fluid. No clear trend on the effect of MQCL parameters on the ground surface morphology was observed. Severe ploughing was observed on the surfaces those were ground with hBN and WS2 nanofluid. However, neat oil produced a clean ground surface with very little or no ploughing. Similar results were obtained for residual stress. No straightforward conclusion could be drawn on the effect of MQCL parameters on residual stress. Neat produced much lower tensile residual stress as compared to hBN and WS2 nanofluid. An interesting observation was made on the effect of wheel wear on residual stress. For neat oil, residual stress was found to be less tensile as compared to new wheel whereas, for hBN and WS2 nanofluids, the residual stress became more tensile for the worn wheel as compared to the new wheel.

References 1. Thakur, A., Gangopadhyay, S.: State-of-the-art in surface integrity in machining of nickel-based super alloys. Int. J. Mach. Tools Manuf. 100, 25–54 (2016) 2. Ulutan, D., Ozel, T.: Machining induced surface integrity in titanium and nickel alloys: a review. Int. J. Mach. Tools Manuf. 51(3), 250–280 (2011) 3. Schafrik, R., Sprague, R.: Gas turbine materials. Adv. Mater. Process. 5, 29–34 (2004) 4. Xu, X.P., Yu, Y.Q., Xu, H.J.: Effect of grinding temperatures on the surface integrity of a nickel-based superalloy. J. Mater. Process. Technol. 129(1–3), 359–363 (2002) 5. Österle, W., Li, P.X.: Mechanical and thermal response of a nickel-base superalloy upon grinding with high removal rates. Mater. Sci. Eng., A 238(2), 357–366 (1997) 6. Cai, R., Rowe, W.B.: Assessment of vitrified CBN wheels for precision grinding. Int. J. Mach. Tools Manuf 44(12–13), 1391–1402 (2004) 7. Patil, D.V.: Grindability study of Inconel 718 using monolayer galvanically bonded monolayer cBN wheel. (Ph.D. Thesis), IIT Kharagpur (2006) 8. Liu, Q., Chen, X., Gindy, N.: Assessment of Al2 O3 and superabrasive wheels in nickel-based alloy grinding. Int. J. Adv. Manuf. Technol. 33(9–10), 940–951 (2007) 9. Yao, C.F., Jin, Q.C., Huang, X.C., Wu, D.X., Ren, J.X., Zhang, D.H.: Research on surface integrity of grinding Inconel 718. Int. J. Adv. Manuf. Technol. 65(5–8), 1019–1030 (2013) 10. Attanasio, A., Gelfi, M., Giardini, C., Remino, C.: Minimal quantity lubrication in turning: effect on tool wear. Wear 260(3), 333–338 (2006) 11. Sinha, M.K., Setti, D., Ghosh, S., Rao, P.V.: An investigation on surface burn during grinding of Inconel 718. J. Manuf. Process. 21, 124–133 (2016)

304

A. Naskar et al.

12. Sinha, M.K., Setti, D., Ghosh, S., Rao, P.V.: An alternate method for optimisation of minimum quantity lubrication parameters in surface grinding. Int. J. Mach. Mach. Mater. 18(5–6), 586– 605 (2016) 13. Sinha, M.K., Madarkar, R., Ghosh, S., Rao, P.V.: Application of eco-friendly nanofluids during grinding of Inconel 718 through small quantity lubrication. J. Clean. Prod. 141, 1359–1375 (2017) 14. Naskar, A., Singh, B.B., Choudhary, A., Paul, S.: Effect of different grinding fluids applied in minimum quantity cooling-lubrication mode on surface integrity in cBN grinding of Inconel 718. J. Manuf. Process. 36C, 44–50 (2018) 15. Upadhyaya, R.P., Malkin, S.: Thermal aspects of grinding with electroplated CBN wheels. J. Manuf. Sci. Eng. 126(1), 107–114 (2004) 16. García, E., Sánchez, J.A., Méresse, D., Pombo, I., Dubar, L.: Complementary tribometers for the analysis of contact phenomena in grinding. J. Mater. Process. Technol. 214(9), 1787–1797 (2014)

Chapter 24

Effect of Different Geometric Texture Shapes on Wettability and Machining Performance Evaluation Under Dry and MQL Environments Sarvesh Kumar Mishra , Sudarsan Ghosh and Sivanandam Aravindan Abstract The present study deals with application of tool and lubricant-based strategies for titanium machining. Laser surface texturing is used to generate patterns of different shapes over the rake face of tungsten carbide cutting tools. Contact angle (CA) at the plain and textured tools is measured with DI water, oil, and MQL fluid in 1:10 oil-to-water ratio. The contact angle values for textured surfaces increase compared to plain surface, and cutting fluid droplets do not spread over the textured surfaces. Further machining experiments are conducted at constant cutting conditions, and maximum 10% reduction in cutting force is obtained for chevron-shaped textures. The textured surfaces under MQL environment offer the combined action of convective heat transfer and fluid retention and release upon requirement. This helps in reducing cutting force, thrust force, and abrasive wear zone over the secondary zone owing to improved lubrication mechanism due to textures. Keywords Textured tools · MQL · Wettability · Contact angle · Cutting forces · Ti6al4v machining

24.1 Introduction Titanium alloys fall in the category of highest used materials in aerospace, defense, automotive, marine, and biomedical applications. The most commonly used among these alloys is Ti6Al4V (grade 5 titanium). Ti6Al4V is extensively used in lowtemperature sections of aerospace and automotive applications for both rotating and non-rotating components. Components like fasteners for low- and high-temperature engine segments, disks, spacers, seals, compressor blades in low-temperature section, and structural parts are some of the common applications. High specific strength, fatigue performance, and low thermal conductivity are among the properties that make it commercial workhorse for aerospace industries. In 5th Annual Titanium S. K. Mishra (B) · S. Ghosh · S. Aravindan Department of Mechanical Engineering, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_24

305

306

S. K. Mishra et al.

Europe 2017 Conference, Netherlands, it was reported that the global demand for titanium alloys will reach 16,000 metric tons in defense application alone by the year 2021 [1]. These forecasts confirm the increasing consumption and rising higher demands of Ti6Al4V alloys in the near future. The machining of Ti6Al4V alloys is a bit challenging and it is considered as low machinability rating material due to poor thermal properties, chemical reactivity, and high strength. The major challenges involved are high cutting forces, built-up edge formation, cratering of the cutting tool, and high thermal stresses on machined surface. To mitigate these machining-induced problems, different tooling and lubricoolant strategies have been proposed in past. Textured tools are considered as a tool-based strategy incorporating tribological advantages for improved machining performances of cutting tools. The patterns/textures/structured hierarchy created on tool surface offer reduction in cutting forces [2, 3], friction [4, 5], and contact length [6]. Sometimes the mechanism of chip formation at rake surface is also changed due to texture-induced interfacial micro-cutting (IMP-µC) or derivative cutting [7]. Different types of geometric-shaped textures have been tried in past based on trial and error approaches to understand the effect of texture shape. Under dry cutting condition, the effect of texture shape and depth of texture has been found negligible in case of machining Ti6Al4V alloy [6]; however, the texture area density affects the performance of the textured tools. In their study, Sugihara and Enomoto [8] performed the comparative analysis of effect of different texture shapes for machining of medium carbon steel. Micro-dimples (closed shape) and micro-grooves (open shape) were fabricated on cutting tool surface and used under dry, wet, and paste lubricated conditions. The performance of closed textures under flood environment exhibited superior performance compared to other shapes. Minimum quantity lubrication (MQL) is a cost-effective and eco-friendly coolant supply strategy using very less amount of coolant compared to flood cooling. Considering the two different approaches for tool-based and coolant-based methods, the basis is formulated for the present experimental study. The study is an attempt to understand the behavior of different texture shapes with cutting fluids and its base components for adsorption phenomenon and liquid–solid–air interaction at room temperature. Further, the machining tests for textured tools were performed for dry and MQL environments and the results were compared with that of plain cutting tools.

24.2 Experimental Details Nanosecond pulsed Nd:YAG laser is used to create different geometric shapes on tool rake face (WC/Co: CNMA120408) as mentioned in our previous studies [6, 7]. The fabricated shapes are circular (C), rectangular (R), triangular (T), and chevron (Ch) (Fig. 24.1). The selection of geometric shape for texturing is based on closed shape textures (circular, square, and triangular as non-connected single texture units) and combination of open and closed shape textures (connected array of chevron-shaped

24 Effect of Different Geometric Texture Shapes …

307

Fig. 24.1 Laser textured tools with different geometric shapes

single texture units). The combination is kept to investigate the effect of texture shapes under both lubricated and dry conditions. The study aims to compare the effect of various closed shape texture on machining responses. Also, the results are compared with a new geometric shape (chevron shape) combining the advantage of both open and closed nature of textures with respect to lubricant retention. The textured tools are tested for contact angle measurement with goniometer (OCA: Dataphysics Goniometer, Germany) using 10 µL liquid volume at tool temperature. To simulate the wetting behavior on different texture shapes, the CAs are measured with DI water, sunflower oil, and oil-to-water mix fluids separately. Water, oil, and oil in water mix (1:10 ratio added with SDBS in 1:10 oil proportion) are used to check hydrophobic/hydrophilic and oleophobic/oleophilic nature of textured surfaces. Machining experiments were performed on Leadwell CNC turning center for cylindrical Ti6Al4V bar of 70 mm diameter with prepared cutting tools. The cutting conditions are kept constant to vc = 110 m/min, f = 0.15 mm/rev, and a p = 1 mm and replicated four times each for each tool conditions. Cutting forces were measured by three-component dynamometer (Kistler: 9129AA, Switzerland), and worn surface was analyzed by optical microscope (Zeta 20, Zeta Instruments, USA). Sunflower oil is mixed in water with 1:10 ratio to prepare MQL fluid. The liquid is ultrasonicated

308

S. K. Mishra et al.

Fig. 24.2 a MQL setup/jet impingement tool and b droplets over the surface

and mixed in magnetic stirrer for 150 min and 30 min, respectively. MQL delivery system (Fig. 24.2a) forces the fluid at the tooltip with pressure and flow rate at 5 bar and 150 mL/h, respectively.

24.3 Results and Discussion 24.3.1 Variation in Static Contact Angle for Different Fluids on Textured WC/Co MQL jet impinges on the tool surface and further breaks into smaller droplets upon impact (Fig. 24.2b). The accelerated droplets enter into the secondary zone and may stop into the zone due to friction. The zone has a number of droplets into stagnant condition at the end of droplet motion. The droplet’s interaction on the tool surface can be simulated using contact angle measurement techniques. Water, oil, and MQL fluid are used to measure contact angle with polished WC/Co surfaces. The results are shown in Fig. 24.3a–c for the different cases. In the case of water– WC/Co interaction, textured surfaces have shown increased CA values compared to plain surface. The increasing contact angle is due to entrapment of air inside the texture pockets and water is unable to interact with the texture boundary inside. Compared to other textures, the chevron textured tools have lower CA value. The pattern generated in the continuous single chevron units has helped in spreading the droplet, and reduction in CA is achieved. The textured surfaces show improved hydrophobic nature compared to plain polished surfaces, and it could be further helpful to understand the interaction of MQL fluid with tool surface. To check the oleophobic/oleophilic nature of generated textures, sunflower oil is used. The CA values for textured surfaces increase compared to plain surface. The textures offer the lower wetting and spreading of liquid over the surface compared to plain surface and thus the smearing of liquid droplets is reduced. Chevron- and triangular-shaped textures exhibit the lower value of CA compared to other texture

24 Effect of Different Geometric Texture Shapes …

309

Fig. 24.3 Variation of contact angle for different texture shapes with a water droplet, b oil, and c oil in water (1:10 ratio)

shapes. The least value for chevron-shaped texture suggests the texture guided flow of liquid over the surface and thus increasing wettability compared to other texture shapes. The single unit of textures with continuous channeling increases the wettability of chevron texture compared to other textures. The observation seems interesting as the oil can seep inside the textures and may cause a guided oil flow inside channels formed by textures. Figure 24.3c depicts the contact angle variation for cutting fluid (oil in water) on different surfaces. The highest CA value for chevron-shaped textures is achieved. The nature of surfaces (triangular- and chevron-shaped textures) is being less non-wetting

310

S. K. Mishra et al.

and less non-spreading compared to the water (hydrophobic) and oil (oleophilic). The measured contact angle and the corresponding response in machining tests are evaluated for machining tests further in the text.

24.3.2 Main Cutting Forces from Machining Tests Cutting forces for plain and textured tools under dry and MQL environments are depicted in Fig. 24.4. Compared to plain cutting tools, circular and rectangular textures have shown a nominal decrease in main cutting force under dry condition. The forces for triangular- and chevron-shaped textured tools have increased corresponding to plain tools under dry cutting. While comparing the forces under MQL conditions, better response of textured tools is achieved. The reduction in forces for triangular- and chevron-shaped textures is higher compared to circular- and rectangular-shaped textures. The lubricant retention in case of triangular textures improved due to constricted geometry of shape that suggests a possible explanation for maximum reduction in cutting forces (∼ 10%). Higher CA values achieved in triangular and chevron textures suggest that the lubricant is retained over the secondary zone. In case of highly wetting conditions (low CA values), chances for vaporization increases as the lubricant droplets enter into secondary zone and lubricating film formation is decreased. High contact angle promotes the thick fluid film formation and hence improved lubrication can be suggested. Improved lubrication and reduced friction can also be correlated with thrust forces achieved in machining (Sect. 24.3.3). High reduction in cutting force under MQL environment for chevron textured tools can been considered due to combined effect of open and closed nature of textures. The fluid entering in chevron-shaped channels resides over the chevron unit continuously over the length of generated channel. The fluid under continuous interaction with chip underside in this case helps in reducing the heat and physico-chemical interaction with chip underside hence resulting in improved lubrication and reduced forces. 350

Main Cutting Force (N)

Fig. 24.4 Variation of cutting force for plain and textured tools

Dry

MQL

300

250

200

P

C

R

T

Fabricated textured shapes

Ch

24 Effect of Different Geometric Texture Shapes …

311

24.3.3 Thrust Forces Under dry cutting, textured tools does not show any great improvement in thrust force reduction as the chips embed into the textured spaces. The embedded chips increase the frictional stresses, temperature, and adhesion of work material. The primary reason for high chip embedding inside textures are high-temperature-induced undercutting as mentioned elsewhere [6]. The usefulness of textures to reduce friction and thrust forces can be obtained only if there remains a separating medium between chip underside and rake face. The separating medium here being cutting fluid or any organometallic product at tool–chip interaction zone will be helpful in reducing the friction due to formation of low-shear-strength film . The retention of cutting fluids is increased due to increased contact angle and thus improving the chances for better lubrication. Under MQL, the maximum reduction in thrust force is achieved for triangular texture (18.4%) when compared to dry cutting. Comparing the thrust forces for different texture shapes under MQL by maximum reduction in thrust force is 10.8% for chevron-shaped tools.

24.3.4 Analysis of Rake Surface for Plain and Textured Tools The tool rake surface analysis for plain and textured tools under MQL environment is shown in Figs. 24.5 and 24.6, respectively. The wear surfaces show the generation of distinct zones over the tool after machining. In case of plain cutting tool, built-up layer formation and abrasion wear are highly dominant as depicted in marked zone (Fig. 24.5). The abrasion marks are severe near the cutting edge and it extends over the rake surface at some distance away from cutting edge. The extension of abrasion mark over the rake surface at longer distance suggests lack of lubricant interaction with sliding chips. The plain tools are unable to facilitate lubricant entry into the secondary zone and thus chances for liquid droplets to enter 200

Thrust Force (N)

Fig. 24.5 Variation of thrust force for plain and textured tools

Dry

MQL

150

100

50

P

C

R

T

Fabricated texture shapes

Ch

312

S. K. Mishra et al.

Fig. 24.6 Rake surface analysis for plain cutting tool under MQL

the capillary zone are reduced. This makes plain tools unable to reduce friction and the generated frictional stresses in the machining zone cause heavy abrasion and intense rubbing/chipping at cutting edge. The analysis of rake surface for textured tools with different geometric shapes shows significant reduction in abrasion zone. The abrasion wear is still present in circular and rectangular textured tools but reduced compared to plain cutting tools. The observations show that the shift in the abrasion wear zone is more toward the cutting edge. The results suggest the lubricants in case of textured tools are interacting in tool–chip contact zone to a greater extent. The ability of lubricants to interact within secondary zone in case of textured tools depends on two factors. The first being the MQL droplets that get vaporized into the secondary zone and form a vapor layer. The layer improves the cooling inside the capillary zone and at the rake face due to convective heat transfer. The second factor is entrapment of fluid into the texture units and consequent supply into the secondary interface upon chip sliding. With textured tools under MQL condition, the combined action of these two factors helps in reducing the friction and abrasive wear over the tool surface. The better results are achieved in case of triangular and chevron textured tools. Abrasion wear and resulting tool deterioration are reduced in the case of triangular and chevron textured tool. It is attributed to lubricant retention and formation of lubricating film at the texture tool–chip interface (Fig. 24.7).

24 Effect of Different Geometric Texture Shapes …

313

Fig. 24.7 Rake surface analysis for differently textured tools

24.4 Conclusions The present study investigates the effect of different geometric texture shape on adsorption phenomenon, hydrophobic/hydrophilic, and oleophobic/oleophilic nature of surfaces and liquids. The interaction with plain and textured tool surfaces with different liquids is used to understand the machining results. Cutting force, thrust force, and abrasion over the rake face are considered as machining output and the results are discussed. Following conclusions can be drawn for the presented investigation: (a) Textured surfaces have higher CA values for water, oil, and oil in water liquids. The value ranges from hydrophobic to oleophilic and hydrophilic to less hydrophilic depending upon surface textures and liquids. (b) Maximum reduction in cutting force and thrust force are 10 and 11%, respectively, for chevron textured tools under MQL environment. Both chevron and triangular textures perform better than circular and rectangular textures. (c) The constricted geometry of triangular textures help in retaining the lubricant over the surface, whereas combination of open and close nature of texture shape offers better performance to chevron-shaped textures. (d) Abrasion marks for textured tools shifts toward the cutting edge and remains nearly absent at rake face for chevron-shaped tools. It confirms the ability of these geometric-shaped textures to offer the effective lubrication in case of chevron textured tools.

314

S. K. Mishra et al.

References 1. Leckerc, A., Gourmelen, J., Chastang, J-F., Sandrine, P., Niedhammer, I., Lanoe, J-L.: Annual Report, International Titanium Association, vol. 82 (2009) 2. Xing, Y., Deng, J., Zhao, J., Zhang, G., Zhang, K.: Cutting performance and wear mechanism of nanoscale and microscale textured Al2 O3 /TiC ceramic tools in dry cutting of hardened steel. Int. J. Refract. Met. Hard. Mater. 43, 46–58 (2014) 3. Kawasegi, N., Sugimori, H., Morimoto, H., Morita, N., Hori, I.: Development of cutting tools with microscale and nanoscale textures to improve frictional behavior. Precis. Eng. 33, 248–254 (2009) 4. Koshy, P., Tovey, J.: Performance of electrical discharge textured cutting tools. CIRP Ann— Manuf. Technol. 60, 153–156 (2011) 5. Jianxin, D., Ze, W., Yunsong, L., Ting, Q., Jie, C.: Performance of carbide tools with textured rake-face filled with solid lubricants in dry cutting processes. Int. J. Refract. Met. Hard. Mater. 30, 164–172 (2012) 6. Mishra, S.K., Ghosh, S., Aravindan, S.: 3D finite element investigations on textured tools with different geometrical shapes for dry machining of titanium alloys. Int. J. Mech. Sci. 141, 424–449 (2018) 7. Mishra, S.K., Ghosh, S., Aravindan, S.: Characterization and machining performance of lasertextured chevron shaped tools coated with AlTiN and AlCrN coatings. Surf. Coat. Technol. 334, 344–356 (2018) 8. Sugihara, T., Enomoto, T.: Performance of cutting tools with dimple textured surfaces: a comparative study of different texture patterns. Precis. Eng. 49, 52–60 (2017)

Chapter 25

Evaluation of Surface Morphology of Yttria-Stabilized Zirconia with the Progress of Wheel Wear in High-Speed Grinding Amit Choudhary , Anirban Naskar

and S. Paul

Abstract High-speed grinding of advanced technical ceramics reduces the amount of brittle fracture, and surface and subsurface damage. Despite several advantages, high grinding temperature encountered in high-speed grinding aggravates wheel wear, which severely alters the wheel profile and reduces its useful life. This article aims to study the effect of wheel wear on ground surface characteristics of yttriastabilized zirconia (YSZ) at high grinding speed of 160 m/s using single-layer electroplated diamond grinding wheel under flood cooling environment. Grinding forces were monitored, and surface morphology and topography were investigated. Initially, grit pullout and grit flattening increased the grinding forces and force ratio (normal to tangential grinding force) and reduced the fracture on the ground surface. Later, the incidents of grit fracture reduced the grinding force ratio and increased the fracture on the ground surface. The surface roughness, on the other hand, gradually decreased with the progress of wheel wear. Keywords Wheel wear · High-speed grinding · Ceramics · Zirconia · Electroplated diamond wheel

25.1 Introduction Advanced technical ceramics such as alumina, zirconia, silicon nitride, and silicon carbide are considered as futuristic materials because of their unique properties. They are highly potential materials for various advanced engineering applications, but their widespread use is limited due to high manufacturing cost and defects generated during their processing [1]. As ceramics are hard materials, final dimensional tolerances are achieved using diamond abrasive processes. Less productivity owing to poor machining characteristics and the high cost of diamond abrasives constitute a significant amount of cost enhancement for ceramic components. Since the quality A. Choudhary (B) · A. Naskar · S. Paul Machine Tool and Machining Laboratory, Department of Mechanical Engineering, Indian Institute of Technology Kharagpur, Kharagpur, West Bengal 721302, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_25

315

316

A. Choudhary et al.

of product and the economics of the process are intensely affected by the condition of the grinding wheel, wheel wear studies are essential to find out optimum sampling period for grinding process. The wheel wear in metallic grinding has been studied and understood to a good extent [2]. The typical wheel wear consists of initial high wear, followed by a gradual steady wear, and then rapidly accelerating wear indicating the end of useful life of the wheel. Three main wheel wear mechanisms have been identified, viz. (i) attrition wear which leads to the development of wear flats due to gradual rubbing of abrasive grits over the work material, (ii) grit fracture, and (iii) bond fracture [2–4]. The typical wheel wear behavior, similar to metallic grinding, has also been observed in lowspeed creep feed grinding of advanced ceramics [5]. However, the main drawback of low-speed grinding of advanced ceramics is the generation of surface and subsurface defects owing to severe micro-brittle fracture [6], which impairs the functionality of the ceramic components. As an alternative to improve the grinding of advanced ceramics, high-speed grinding has been utilized by various researchers [7, 8]. At higher wheel speeds, the scale of grinding gets reduced [9]. In grinding of brittle materials like ceramics, with the reduction of grinding scale, the energy required for brittle fracture becomes considerably higher than the energy required for plastic deformation [10], which reduces the probability of occurrence of micro-brittle fracture and associated defects in the ground components. Yin and Huang [7] have reported reduced grinding forces and micro-brittle fracture for various advanced ceramics in high-speed grinding using the resin-bonded diamond grinding wheel. Besides the advantages, high wheel speed in grinding adversely affects the wheel wear. High wheel speed employed in grinding leads to high grinding temperature [11], which aggravates the wheel wear. Hwang et al. [12] have studied the wear of single-layer electroplated diamond grinding wheel in high-speed grinding of silicon nitride. They reported a reduction in wheel life owing to increased sliding length at higher grinding speeds. The grinding forces increased progressively due to attrition wear of the diamond grits. Huang [13] also indicated greater sliding length to be the main reason for rapid wheel wear in the high-speed deep grinding of yttria-stabilized zirconia (YSZ) using resin-bonded diamond grinding wheel. Moreover, a high temperature close to 800 °C generated in the high-speed grinding of zirconia [14] might also aggravate the wheel wear. The literature present above shows the advantages of using high wheel speeds in ceramic grinding. The main limitation in employing high-speed grinding is the aggravated wheel wear and its effect on ground surface characteristics, which has not been given adequate attention in the available literature. Single-layer electroplated grinding wheels are particularly favored in high-speed grinding because of safety issues [1]. Despite good performance [9], the main limitation of electroplated wheels in grinding is their useful life as they cannot be dressed and used again. Therefore, this article aims to study the wear of single-layer electroplated diamond grinding wheel and its effect on the ground surface characteristics of YSZ in high-speed grinding under conventional flood cooling environment. The grinding forces were monitored with the progress of wheel wear. Ground surface morphology and surface roughness were investigated at different wheel wear stages.

25 Evaluation of Surface Morphology of Yttria-Stabilized …

317

25.2 Materials and Methods High-speed grinding of YSZ was carried out on Log-O-Matic LFS 2 CNC 6540 surface grinder in up-grinding mode with single-layer diamond grinding wheel under flood cooling environment. Coolant supply pressure and flow rate were 10 bar and 20 L/min, respectively. Two different sizes of YSZ samples were used, (i) sample A—50 mm × 25 mm × 7.5 mm (grinding face: 50 mm × 7.5 mm) and (ii) sample B—20 mm × 6 mm × 7 mm (grinding face: 20 mm × 7 mm). Some important mechanical and thermal properties of the YSZ are given in Table 25.1 Altogether, 100 passes were given on sample A. Five number of samples B were prepared, one with new wheel and then after every 25 passes on sample A, to investigate the effect of wheel wear on ground surface morphology and topography as the wheel wear progressed. To prepare sample B for morphological and topographical studies, ten passes were given on each of the samples B. The experimental setup is shown in Fig. 25.1 and complete experimental details are given in Table 25.2. A PDA878 grade of diamond grits (manufactured by Element Six) having an average grit size of 151 µm was chosen for single-layer diamond grinding wheel. The selected grade was having mostly blocky grits. Figure 25.2 shows an electron microscopic image of loose diamond grits. The work material was mounted on piezoelectric crystal dynamometer (Kistler 9254) to measure tangential and normal grinding forces. The dynamometer was connected to a digital oscilloscope (Agilent Technologies, DSO-X-2014A) via two single-channel charge amplifiers (Kistler type 5051). Average forces of last five passes out of ten passes given on sample B are reported. Ground surface morphology was observed under SEM (Zeiss EVO 18 research). The surface roughness of the ground samples was measured perpendicular to the grinding direction using a 3D tactile surface profilometer “Form Talysurf 50” of Taylor Hobson make. The evaluation length of 4 mm with the cut-off length of 0.8 mm was used for surface roughness Table 25.1 Nominal mechanical and thermal properties of YSZ Density

Hardness (vickers)

Elastic modulus

Fracture toughness

Thermal conductivity

6.02 g/cm3

13 GPa

200 GPa

8 MPa/m0.5

3 W/m K

grinding wheel

coolant nozzle dynamometer

cleaning nozzle workpiece digital oscilloscope

Fig. 25.1 Experimental setup

charge amplifiers

318

A. Choudhary et al.

Table 25.2 Experimental conditions Machine tool

Log-O-Matic LFS 2 CNC 6540 high-speed surface grinding machine

Grinding wheel

Single-layer electroplated diamond wheel Wheel diameter 125 mm Wheel width 10 mm Grit diameter 151 µm

Work material

Yttria-stabilized zirconia Sample A: 50 mm × 25 mm × 7.5 mm Sample B: 20 mm × 6 mm × 7 mm

Grinding parameters

Grinding speed 160 m/s Downfeed 100 µm Table speed 500 mm/min

Environment

Conventional flood cooling (water soluble oil) • 5% concentration soluble oil (veedol—solucut super)

Fig. 25.2 SEM image of diamond grits (PDA878) used for grinding wheel

100 µm

measurements. Five linear roughness profiles, each 0.5 mm away, were measured for each sample and average roughness value of those five profiles is then reported. The grinding wheel was observed under a stereo zoom optical microscope (Zeiss Discovery V20 fitted with AxioCam ICc 5 camera module) before and after grinding to explore the wheel wear mechanism. Post-grinding, before observing the wheel under the zoom microscope, the wheel was cleaned for 30 min in an ultrasonic bath filled with 2-propanol.

25.3 Results and Discussion Figure 25.3 shows the variation of specific normal force (Fn , Fig. 25.3a), specific tangential force (Ft , Fig. 25.3b) and the force ratio (Fn /Ft , and Fig. 25.3c) for sample B with the progress of wheel wear. Specific normal force increased steeply

25 Evaluation of Surface Morphology of Yttria-Stabilized … (a) specific normal force ( , N/mm)

2

319

(b) specific tangential force ( , N/mm)

0.3

8

(c) force ratio (

)

6 0.2

1

0

new 25 50 75 100 wheel passes passes passes passes

4

0.1

100 new 25 50 75 wheel passes passes passes passes

2

new 25 50 100 75 wheel passes passes passes passes

Fig. 25.3 Variation of grinding forces and force ratio with the progress of wheel wear

till 25 passes, then increased gradually till 50 passes, and remained constant till 75 passes. After 75 passes, a reduction in specific normal force can be observed till 100 passes. On the contrary, specific tangential force continued increasing till 100 passes. Consequently, the ratio of normal to tangential force first increased up to 50 passes and then decreased up to 100 passes. The onset of wheel wear gets reflected by the change in grinding forces and the force ratio (Fn /Ft ). The normal grinding force is a measure of ease of penetration of the grinding grit into the work material, whereas the tangential grinding force is a measure of energy consumption. Figure 25.4 shows the optical image of the grinding wheel before and after grinding. The evidence of grit pullout (grits 1, 2, and 3 in Fig. 25.4a), grit flattening (grits 4, 5, 6, 7, and 8 in Fig. 25.4a), and grit fracture (grits 7 and 8 in Fig. 25.4a) can be observed in Fig. 25.4. In a single-layered electroplated grinding wheel, the grits are held mechanically with the bond material. During electroplating, some of the grits are not bonded properly and come out of the (a) before grinding

(b) after grinding

4 1 3

2

5

8

grinding direc on

6 7 8

7 grit pull-out;

grit flattening;

grit fracture

Fig. 25.4 Optical images of grinding wheel showing different wheel wear events

320

A. Choudhary et al.

bond matrix in starting few grinding passes. Because of this, active grits available for chip formation (C) decrease and maximum undeformed chip thickness (h m ) increases as per Eq. (1) [9].  hm =

3 vw C tan α vs



a ds

 21 (25.1)

where C is the active cutting point density, α is the semi-included grit angle, vw is the work speed, vs is the wheel speed, a is the wheel downfeed, and ds is the wheel diameter. As the h m is increased, both the grinding forces, tangential and normal, increase [12]. Grits 1, 2, and 3 in Fig. 25.4a represent a similar case of improperly bonded grits which got removed within 10–15 passes of the grinding. Because of this, a rapid rise in both the grinding forces can be observed till 25 passes in Fig. 25.3a, b. Along with grit pullout (grits 1, 2, and 3), instances of grit flattening were also observed (grit 4, 5, 6, 7, and 8). Grit flattening takes place due to attrition wear which progressively increases the grinding forces [3]. With the development of grit flattening, it becomes difficult for a grit to penetrate into the work material, which results in more rubbing. Consequently, the force ratio (Fn /Ft ) increases, as the change in normal force is more as compared to the change in tangential grinding force. The rise in the force ratio (Fn /Ft ) in Fig. 25.3c till 50 passes is attributed to predominated grit flattening. After 50 passes, high grinding temperature and high grinding forces, both the normal and the tangential, must have introduced sufficient amount of thermomechanical stresses in the grit leading to grit fracture. From Fig. 25.4, it can be observed that grit fracture has taken place on the same grits 7 and 8 where the grits flats were there. Typically for friable grits, the grit fracture occurs across the grinding direction which gives rise to multiple cutting points and thereby increasing the cutting point density virtually. For such a case, h m reduces as per Eq. (25.1). With a reduction in h m , secondary rubbing increases which increase the force ratio (Fn /Ft ). On the contrary, if the grit fracture occurs along the grinding direction, observed here for grit 7 and 8 in Fig. 25.4a, no change in h m takes place. However, reduction in wear flat area eases the grit penetration into the work material, which reduces the normal grinding force. A reduction in the wear flat area facilitates the cutting action of the grit, which increases the tangential grinding force. The similar trend of reduction in normal grinding force and increase in tangential cutting force has also been observed in Fig. 25.3a, b. As a result, the force ratio (Fn /Ft ) decreased, which is also evident from Fig. 25.3c. The large fluctuation in forces at 50 passes may be attributed to the transient state of grits undergoing fracture due to the development of wear flats and high thermomechanical stresses. Figure 25.5 shows the morphological features of samples B prepared at different wheel wear conditions. New wheel (Fig. 25.5a) yielded mild fracture and a peaky surface having high crests and deep troughs. Fracture and peakedness of the surface got reduced as the wheel was used progressively till 50 passes (Fig. 25.5b, c). Further grinding with the wheel beyond 50 passes provided severely fractured ground

25 Evaluation of Surface Morphology of Yttria-Stabilized …

321

Fig. 25.5 Morphology of samples prepared at different wheel wear conditions

surfaces, Fig. 25.5d, e after 75 and 100 passes, respectively. The morphological features observed in Fig. 25.5 were found in good agreement with the sharpness of the grits on the grinding wheel. New wheel having sharp grits induced mild fracture and yielded a leptokurtic surface having high crests and deep troughs as can be seen in Fig. 25.5a. With the development of wear flats on the grinding grits up to 50 passes, the fracture got arrested and is clearly depicted in Fig. 25.5b, c. Grit fracture after 50 passes due to high thermomechanical stresses once again sharpened the grits and yielded fractured surface after 75 and 100 passes as can be seen in Fig. 25.5d, e, respectively. The events of grit flattening and grit fracture have been identified in Fig. 25.4. Arithmetic surface roughness (Ra) for samples B ground at different wheel wear conditions is shown in Fig. 25.6. It can be observed that new wheel provided the highest surface roughness values. This is attributed to high protruding improperly bonded diamond grits on the new wheel, which provided high crests and deep troughs (Fig. 25.5a). Grits 1, 2, and 3 identified in Fig. 25.4a belong to this category. Sample B prepared after 25 passes showed a rapid fall in surface roughness approximately by 30% and is attributed to the removal of loosely held grits and simultaneous development of wear flats on the diamond grits. At 50 passes, the surface roughness decreases by 11% owing to the progress of wear flats. The grit fracture beyond 50 passes yielded severely fractured YSZ surface with further reduction in surface roughness. It can be observed from Fig. 25.5c, d that large fractured patches removed the crests on the ground surface, resulting in lower surface roughness. A reduction in surface roughness due to fracture has been previously explained by the authors in detail [15].

322

A. Choudhary et al.

Fig. 25.6 Surface roughness of samples prepared at different wheel wear conditions

arithmetic roughness (Ra, µm)

4

3

2

1

0

new wheel

25 passes

50 passes

75 passes

100 passes

25.4 Conclusions Effect of wheel wear on ground surface characteristics in high-speed grinding of YSZ under conventional flood cooling environment was studied. Following conclusions were made: Wheel wear mechanism consisted of grit pullout, grit flattening, and grit fracture. Initially, grinding forces (normal and tangential) and force ratio (normal to tangential) increased owing to grit pullout and grit flattening. Later, because of grit fracture, tangential force increased and normal force decreased, resulting in a reduction in the force ratio. New grinding wheel yielded mild fracture and a leptokurtic surface having high crests and deep troughs. With the progress of grit flattening, surface fracture got reduced. Later, the occurrence of grit fracture yielded a severely fractured surface. New wheel provided the highest surface roughness owing to the presence of over protruding loosely held grits. Initially, surface roughness decreased rapidly. Thereafter, it decreased gradually throughout the progress of wheel wear. Acknowledgements The authors gratefully acknowledge the funding support received from: (i) ARDB, MoD, Government of India (Sanction No. ARDB/01/2031772/M/I, dated—August 10, 2015) and (ii) DST, FIST, Government of India (Sanction No. SR/FST/ET-II-003/2000, dated— May 20, 2002).

25 Evaluation of Surface Morphology of Yttria-Stabilized …

323

References 1. Hwang, T.W., Evans, C.J., Whitenton, E.P., Malkin, S.: High speed grinding of silicon nitride with electroplated diamond wheels, part 1: wear and wheel life. J. Manuf. Sci. Eng. 122, 32–41 (2000) 2. Stetiu, G., Lal, G.K.: Wear of grinding wheels. Wear 30, 229–236 (1974) 3. Malkin, S., Cook, N.H.: The wear of grinding wheels, part 1—attritious wear. J. Eng. Ind. 93, 1120–1128 (1971) 4. Malkin, S., Cook, N.H.: The wear of grinding wheels, part 2—fracture wear. J. Eng. Ind. 93, 1129–1133 (1971) 5. Liao, T.W., Li, K., McSpadden Jr., S.B.: Wear mechanisms of diamond abrasives during transition and steady stages in creep-feed grinding of structural ceramics. Wear 242, 28–37 (2000) 6. Zhang, B., Zheng, X.L., Tokura, H., Yoshikawa, M.: Grinding induced damage in ceramics. J. Mater. Process. Technol. 132, 353–364 (2003) 7. Yin, L., Huang, H.: Ceramic response to high speed grinding. Mach. Sci. Technol. 8(1), 21–37 (2004) 8. Ramesh, K., Yeo, S.H., Gowri, S., Zhou, L.: Experimental evaluation of super high-speed grinding of advanced ceramics. Int. J. Adv. Manuf. Technol. 17, 87–92 (2001) 9. Malkin, S., Guo, C.: Grinding technology: theory and applications of machining with abrasives. Industrial Press, New York (2008) 10. Bifano, T.G., Dow, T.A., Scattergood, R.O.: Ductile-regime grinding: a new technology for machining brittle materials. J. Eng. Ind. 113, 184–189 (1991) 11. Klocke, F., Verlemann, E., Schippers, C.: High-speed grinding of ceramics. In: Jahanmir, S., Ramulu, M., Koshy, P. (eds.) Machining of ceramics and composites, pp. 119–138. Marcel Dekker, New York (1999) 12. Hwang, T.W., Evans, C.J., Malkin, S.: An investigation of high speed grinding with electroplated diamond wheels. Annal. ClRP 49(1), 245–248 (2000) 13. Huang, H.: Machining characteristics and surface integrity of yttria stabilized tetragonal zirconia in high speed deep grinding. Mater. Sci. Eng., A 345, 155–163 (2003) 14. Xie, G.Z., Huang, H.: An experimental investigation of temperature in high speed deep grinding of partially stabilized zirconia. Int. J. Mach. Tools Manuf. 48, 1562–1568 (2008) 15. Choudhary, A., Naskar, A., Paul, S.: Effect of minimum quantity lubrication on surface integrity in high-speed grinding of sintered alumina using single layered diamond grinding wheel. Ceramics International, article in press (2018)

Chapter 26

Grindability and Surface Integrity of Nickel-Based Cast Superalloy IN-738 by Vitrified Alumina Wheel Srinivasa Rao Nandam, A. Venugopal Rao, Amol A. Gokhale and Suhas S. Joshi Abstract IN-738 is a Ni-based superalloy widely used for manufacture of industrial gas turbine blades and vanes in vacuum investment cast and precipitation hardened condition. The castings so produced undergo certain machining operations to achieve fitment accuracies in the engine assemblies. However, these alloys are categorized as difficult-to-cut materials due to complex composition, the presence of hard refractory elements, high strength, and hot hardness. Precision grinding operations on these alloys are generally carried out by the special machineries to achieve higher grindability and better surface integrity. The surface integrity plays a vital role in the functional performance of turbomachinery components. In the present work, an experimental study has been carried on an IN-738 LC superalloy having polycrystalline grain structure by a conventional surface grinding process using a low-cost and higher performance vitrified alumina wheel. Grindability of the alloy in terms of grinding ratio, specific grinding energy and surface integrity aspects such as surface roughness, surface hardening and microstructural changes was evaluated. The grinding swarf was also analyzed to evaluate the deformation cum failure mechanisms involved during the grinding process. Keywords Superalloys · Vitrified alumina wheel · Grindability · Surface integrity · Grinding swarf

26.1 Introduction Precipitation hardened nickel-based superalloys are widely used in the manufacture of hot section components (gas turbine blades and vanes) of modern industrial power generation gas turbines and aerospace engines due to their excellent high-temperature creep–rupture strength and hot corrosion resistance [1]. IN-738 alloy is a vacuum S. R. Nandam (B) · A. V. Rao Defence Metallurgical Research Laboratory, DRDO, Hyderabad 500058, India e-mail: [email protected] A. A. Gokhale · S. S. Joshi Department of Mechanical Engineering, IIT Bombay, Mumbai 400076, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_26

325

326

S. R. Nandam et al.

melted and cast alloy which exhibits superior tensile and elevated temperature stress rupture properties comparable to those of the widely used alloy IN-713C along with substantially better sulfidation resistance [2]. The microstructure of IN738 is highly complex. The main alloying elements chromium, cobalt, and the refractory elements such as molybdenum, tungsten, tantalum contribute to solid solution strengthening of the face-centered cubic gamma (γ) phase, i.e., the austenitic nickel matrix. The alloy is strengthened by precipitation hardening due to mainly Ni3 (Al/Ti), a gamma prime phase (γ ) and other phases such as MC and M23 C6 carbides from tantalumand chromium-based, respectively [3]. The rich amount of chromium also forms a protective layer of chromium oxide (Cr2 O3 ) during service which enhances the corrosion resistance. Machining operations are essential in precision components to achieve stringent dimensional and geometrical accuracies, and high surface finish. Cutting forces and heat are generated in metal cutting as a result of plastic deformation of the layer being machined [4]. The significantly higher cutting forces and heat generated during machining are due to the high shear strength, fracture strain, hot hardness, and low thermal conductivity of nickel-based superalloys. These conditions result in rapid tool wear and severely affect the dimensional accuracy and surface integrity of the machined surfaces. Many researchers found that the metallurgical characteristics of the Ni-based superalloys like high strength matrix, hard intermetallics, hard abrasive particles, and the high work hardening effect are the primary reasons for the poor machinability [5, 6]. Further, the machinability of cast nickel-based superalloys is worse when compared with wrought alloys due to the presence of more refractory elements and the non-uniform grain structure of the former material. Grinding process is generally used for machining of hard and difficult-to-cut materials to obtain the required dimensional and geometrical accuracies. The grindability of any engineering material is described by the grinding forces, grinding temperature, specific grinding energy, and grinding ratio etc. [7]. The high temperatures at the grinding zone cause possible thermal damage to the workpiece during grinding with abrasive grinding wheels. Hence, more attention has been paid to improve surface quality of aerospace alloys by reducing excessive heat and cutting forces during the grinding process [8]. The most commonly adopted techniques for controlling the above for grinding of cast nickel-based super alloys is creep feed grinding technique [9, 10]. The large depth of cut, long arc of cut, and very slow workpiece speed (feed) are the prominent features of the creep feed grinding process. But the creep feed grinding process is found to be one of the highly expensive processes due to the requirement of heavy duty machine tools, specially designed grinding wheels and high pressure coolant system. Therefore, affordability by small- and medium-scale industries is a major concern. The surface integrity (SI) of a machined component is the sum of all the elements that describe the conditions existing on or at the surface. The term was coined initially by Field and Kahles in the year 1964 [11]. The SI mainly consists of surface topography and surface metallurgy. The former describes the surface roughness or lay or texture of the outermost layer of the component’s surface and the latter describes the nature of the altered layers below the surface with respect to the base

26 Grindability and Surface Integrity of Nickel-Based Cast …

327

or matrix material. The usual SI problems that exist after machining heat resistant superalloys as reported in the literature are surface texturing, cavities, cracking, metallurgical transformation, plastic deformation, increased microhardness, increased surface roughness, formation of tensile residual stresses and heat affected zone, recast or redeposited material, change of physical properties, etc. SI of the component is of prime concern especially in aerospace applications as the failures are primarily caused due to fatigue, creep, and stress corrosion etc. [12–15]. The cause and effect of process parameters on SI in machining were shown in Fig. 26.1. The published literature on grindability and surface integrity aspects of nickelbased cast superalloys is scarce in the open literature. To understanding the mechanisms involved in grinding of Ni-based cast superalloy and to evaluate the performance of widely used grinding wheels, an experimental study was undertaken in the present work. A polycrystalline cast IN-738 LC alloy was used for this purpose. Grinding experiments were conducted using a conventional surface grinding machine with a vitrified alumina wheel. The grindability and surface integrity aspects were evaluated systematically.

Fig. 26.1 Effect of machining parameters on SI

328

S. R. Nandam et al.

26.2 Experimental Setup 26.2.1 Work Material A workpiece of 75 mm × 40 mm × 15 mm size was extracted by wire EDM process from a vacuum investment cast, polycrystalline structure component of IN-738 LC alloy. Rough EDM surfaces and heat affected zones of workpiece were removed by machining all the surfaces to a depth of 1 mm by milling process with multiple depths of cuts under flood coolant. The nominal chemical composition of the alloy by weight percentage is shown in Table 26.1 [16]. The physical and mechanical properties of the alloy published by INCO, alloy developers [17], are shown in Tables 26.2 and 26.3, respectively. The lower percentage of elongation indicates poor ductility (deformation) of the alloy.

26.2.2 Machine Tool The grinding experiments were performed by a horizontal spindle type, precision conventional surface grinding machines of model PRAGA 452P from M/s HMT machine tools limited. The power rating of the main spindle of the machine was 1.5 kW, and the maximum running rotational speed was 2800 rpm.

26.2.3 Grinding Wheel Vitrified aluminum oxide wheels are readily available in the open market with low cost and are regularly used for grinding high-strength structural steels. The grinding wheel used in these experiments was a conventional toolroom type, white color grinding wheel of specification A60K5V8 from M/s CUMI. The dimensions of the wheel were 200 mm of outer diameter and 20 mm of width with 31.75 mm of inner diameter for wheel mounting on the spindle. The specification of the grinding wheel is shown in Table 26.4.

26.2.4 Cutting Environment The coolant used in the experiments was soluble cutting oil (neat oil), Servocut S product from M/s IOCL. The coolant was prepared with cutting oil of 5% concentration in water to form stable emulsion. Grinding experiments were conducted under flood coolant environment with flow rate of 50 l/min and 3 bar pressure.

Co

8.59

C

0.11

16.08

Cr

1.75

Mo 2.67

W 1.75

Ta

Table 26.1 Chemical composition of IN-738 LC in weight percentage 0.9

Nb 3.43

Al 3.38

Ti 0.03

Cu

0.05

Zr

0.5

Fe

0.18

Si

Bal

Ni

26 Grindability and Surface Integrity of Nickel-Based Cast … 329

330 Table 26.2 Physical properties of the alloy

S. R. Nandam et al. Property Density

Table 26.3 Mechanical properties of the alloy

Table 26.4 Specification of A60K5V8 grinding wheel [18]

(kg/m3 )

Value at 20 °C

Value at 1093 °C

8110

–

Dynamic modulus of elasticity (GPa)

201

201

Thermal conductivity (W/m/K)

11.8

27.3

Specific heat (J/Kg/°C)

418.4

711.3

Melting range (°C)

1230–1315

–

Coefficient of thermal expansion

6.45 × 10 E−6/°C

8.85 × 10 E−6

Poisson’s ratio

0.28

0.30

Property

Value

0.2% yield strength

896 MPa

Tensile strength

1034 MPa

% Elongation

7

% Reduction of area

9

Impact properties at room temperature

76 J

Nomenclature

Description

Type

A

Abrasive

Aluminum oxide

60

Grit number

Medium

K

Grade (hardness)

Medium

5

Structure

Dense

V

Bond material

Vitrified

8

Abrasive

Aluminum oxide

26.2.5 Dynamometry Various mechanisms such as rubbing, plowing, shearing, fracturing contribute to grinding forces during grinding process [19]. Piezoelectric-based precision cutting force dynamometer, model 9265B from M/s Kistler, Switzerland, was mounted on the magnetic table of surface grinding machine. A multicomponent (eight channels) charge amplifier model 5080A was used for amplifying analog signal of piezoelectric sensors (electrical charge). A data acquisition system, digital-to-analog converter (DAC) model 5697A was used to convert this amplified charge into analog data of cutting forces. A personal computer-based software Dynoware was used for analysis

26 Grindability and Surface Integrity of Nickel-Based Cast … Table 26.5 Grinding parameters

Parameter

331 Value

Length of stroke table

118 mm

Number of strokes per minute

40

Table traverse speed

4.7 m/min

Table cross-feed

0.7 mm/stroke

Table cross-speed

28 mm/min

Spindle rotational speed

2800 rpm

In-feed (depth of cut)

0.05 mm

of the cutting force data in terms of the maximum and minimum cutting forces in perpendicular x- and z-axes.

26.3 Experimentation The wheel was trued and dressed by a single point diamond point dresser mounted on the table and positioned below the wheel. The angle of the dresser was 7.5° to the vertical plane, which was parallel to the front surface of the grinding wheel. The dressing operation was carried out with a depth of 0.05 mm under manual feed in two passes. The grinding parameters used during the experiments are shown in Table 26.5.

26.4 Results and Discussion 26.4.1 Grinding Swarf The grinding swarf was collected from the surface table immediately after the grinding experiments. The fragmented particles of alumina and metallic chips of various configurations were observed under 50× magnification by a binocular optical digital microscope of model DM6000 M from M/s Leica, Germany. The fragmented particle of alumina is shown in Fig. 26.2a. The size of the majority of the particles was found to be 300–400 μm. Figure 26.2b shows a curved slice type chip from the work material of size around 800–900 μm. The signs of plastic deformation of the chips indicate ductile fracture from gamma phase matrix. The curvature of the chip might be due to the restriction to the free flow of the chips by the abrasives or the bond material. Tso [20] has also observed similar slice type of chips when grinding Inconel 718 by CBN grinding wheel in surface grinding operation. Figure 26.2c shows irregular shaped fragments of chips of around 200 μm. These might be brittle

332

S. R. Nandam et al.

Fig. 26.2 a Fragments of alumina, b metallic chip, c fragments of intermetallic chips

fragments of intermetallic phase, gamma prime precipitates, which are found to be existed around 40% of volume fraction in the alloy.

26.4.2 Grindability Material Removal Rate (MRR): The MRR was calculated based on weight loss method. The sample was weighed before and after grinding experiments by an electronic precision weight balances of 0.001 g least count. The readings recorded were 452.6 and 451.5 g, respectively. The calculated weight difference of 1.11 g indicates the material removal during the grinding process. The product of surface area of the work material and depth of cut gives the volume of material removed. The calculated volume of material removal was 150 mm3 . The corresponding weight of the material can be removed is 1.22 g. It was a bit higher than the actual value of 1.11 gm. The actual depth of cut calculated from the experimental material removal was 0.045 mm, which is very near the set value of 0.05 mm. The deviation of depth of cut value of 0.005 mm (10%) might be due to deflection of machine spindle due to thrust force from the work material and progressive wear out of the grinding wheel at higher values of in-feed. Grinding Forces: The tangential cutting force (Fx) along the grinding direction and normal cutting force (Fz) parallel to depth of cut during the grinding process were measured. The maximum tangential cutting force was 26.98 N, and the maximum normal cutting force was 51.27 N. The signatures of cutting forces recorded are shown in Fig. 26.3a, b. It is observed that the cutting forces increase during the grinding process. Similar trend was observed by many researchers during machining of superalloys due to work hardening effect. But, in the present study, the calculated hardness values at machined surface shown in Table 26.6 indicate a lower surface hardness than bulk material. The increase in cutting forces might be due to loading and rubbing of the grinding wheel while progressing of grinding process. Grinding Ratio: Grinding ratio is defined as the ratio of the volume of work material removed and the volume of wear of the grinding wheel. The initial and

26 Grindability and Surface Integrity of Nickel-Based Cast …

333

Fig. 26.3 a Tangential cutting force (Fx). b Normal cutting force (Fz)

Table 26.6 Readings of surface hardness of ground surface Reading

1

2

3

Bulk

R1

344.0

347

349

359

R2

349.0

360

351

351

R3

351.0

364

352

353

Avg.

348.0

357.0

350.7

354.3

the final diameters of the grinding wheel after dressing, as measured by a precision micrometer of 0.01 mm least count, were 199.95 and 199.30 mm, respectively. The calculated number of strokes required to machine the width of workpiece of 40 mm under a cross-feed of 0.7 mm per stroke was 58 strokes. The calculated time to complete the 58 strokes with table speed of 40 strokes per minute was 1.45 min. The calculated volume of wear of the grinding wheel was 4391 mm3 . The volume rate of wear of grinding wheel per minute was 3028 mm3 . The calculated value of the volume of material removal of work material per minute was 9440 mm3 . Therefore, the calculated grinding ratio was 3.2. The general surface grinding process indicates the value of 50 for easy to grind materials and three for difficult to grind materials [7] due to reasonably higher wear of grinding wheel and lower material removal of work material. Specific Grinding Energy: The specific grinding energy (J/mm3 ) is defined as the energy required for removal of unit volume of material from the work material. The multiple of cutting force and the spindle speed gives the power consumption of the machine tool during grinding process. The calculated spindle speed was 29.2 m/s for 199.3 mm diameter grinding wheel at wheel rotational speed of 2800 revolutions per minute. The calculated power consumption was 1497 W for the maximum normal cutting force of 51.27 N. This value was very near (99.8%) to the value of spindle power of the machine tool (1500 W). The calculated specific grinding energy was 19 J/mm3 . Similar value of 27 J/mm3 was observed by Ravuri et al. [21] in surface grinding of Inconel 718 by a standard resin bonded alumina wheel (A80K6B) of higher width (25 mm).

334

S. R. Nandam et al.

26.4.3 Surface Integrity Aspects The Surface Roughness: The surface roughness was measured by a 2 μm conisphere-shaped diamond stylus with touch probe roughness measurement instrument of Form Talysurf Intra model from M/s Taylor Hobson, UK. The measurement was carried out on the ground surface at three different locations along and across the length of the workpiece. The sampling length selected was 10 mm, and the cutoff length selected was 0.8 mm with a spacing of 0.5 mm as per ISO 3274 standard. The average value of the surface roughness (Ra) achieved was 0.1 μm along the length and 0.6 μm across the length of workpiece. The achieved surface roughness value was much below the value of 1.0 μm of investigations by Ding et al. [9] and Caggiano and Teti [22] with CBN grinding wheels. The improvement of surface roughness might be due to the rubbing and polishing effect of the alumina abrasives. Surface Hardness: The hardness values were measured on the machined surface at three locations with 20 kg load at three different positions by a standard Vickers hardness measuring machine, model VRS 270 from M/s Affri, Italy, with a diamond pyramid indenter. The average value was found to be 351.9 VHN. This value was very near to the surface hardness of bulk material of 354.3 VHN. The measured hardness values of machined surface and bulk surface were shown in Table 26.6. Many of the researchers have observed the increase of surface hardness on the machined surface due to surface hardening or work hardening effect during machining process [23, 24]. But few researchers have observed decrease in surface microhardness on nickelbased superalloy due to the effect of thermal softening [25, 26]. The minor decrease of surface hardness (around 1–2%) in this study might be either due to above thermal softening effect or microstructural modification of surface of work material. Microstructural Evaluation: Two samples of 20 mm × 20 mm size were cut by WEDM process perpendicular to the machined surface from the work component. These samples were mounted in Bakelite mold under hot pressing, and care has been taken to orient machined surface as the top surface for one sample and un-ground surface as the top surface of the latter one. Metallography sample preparation was carried out by standard method using SiC bonded papers of 200–1200 grit size. These samples were cleaned by acetone and smooth cloth. Further polishing was done by lapping process with diamond paste of various sizes start from 9 to 1 μm. The cleaned sample surface was etched under standard Kalling’s 2 reagent waterbased solution by dip etching process. These samples were examined under optical microscope under various magnifications are shown in Fig. 26.4. Figure 26.4a, c, e are optical images of un-ground sample under 12.5×, 200× and 500× magnifications, respectively, while Fig. 26.4b, d, and f are of the machined surfaces at comparable magnifications. It was observed from Fig. 26.4b that the dendrite structure has certain modifications when compared with un-ground surface. Zhou [27] has observed surface deformation when IN-718 was turned under high-speed machining with SiC whisker reinforced alumina cutting tool. Zhang et al. [28] have observed recrystallized grains and formation of γ–γ eutectics when annealed and cellular recrystallised structure of a nickel-based cast superalloy heat treated below γ solvus

26 Grindability and Surface Integrity of Nickel-Based Cast …

335

Fig. 26.4 a Un-ground surface at 12.5×, b ground surface at 12.5×, c un-ground surface at 200 ×, d ground surface at 200×, e un-ground surface at 500×, f ground surface at 500×

336

S. R. Nandam et al.

temperatures. The γ solvus temperatures for IN-738 alloy found to be in the range of 1160–1175 °C. The similar re-crystallizations were also observed (Fig. 26.4d, f) in the present study. This was due to the thermomechanical effect on the machined surface by grinding process at much lower γ solvus temperatures of the alloy.

26.5 Conclusions The specific grinding energy of IN-738 alloy evaluated by a conventional surface grinding process with a toolroom purpose vitrified alumina grinding wheel was 19 J/mm3 . The overall grinding efficiency achieved was 99.8%. The lower value of grinding ratio 3.2 was due to reasonably higher wear of alumina grinding wheel. The wear of abrasives also attributed for formation of new and sharp cutting edges for successive grinding operations. The dimensional accuracy of 0.005 mm was achieved on the machined surface. The average surface roughness value of Ra achieved on the machined surface is 0.1 μm along the length. The rubbing and polishing actions of grinding wheel were predominant to achieve higher surface finish. The ductile and brittle fractures have found to be existed in the work material. The nickel phase matrix has contributed for ductile fracture, and where intermetallic precipitates of the gamma prime phase have contributed for brittle fracture during grinding process. Though any notable increase in surface hardness was not observed on the machined surface, minor microstructural changes were observed. As a consequence, the vitrified alumina grinding wheels can be proficiently used for simpler geometries to save machining time of costlier creep feed grinding process in the manufacture of turbomachinery components. Acknowledgements The authors would like to give their deep gratitude to Defence R&D Organization for financial support and to Dr. Vikas Kumar, Director, DMRL for giving opportunity to carry out the experimental work. Authors are thankful to officers and staff of Mechanical Engineering Group, Metallography Section of DMRL for their kind cooperation. Authors also thankful to Directional Solidification Groups of DMRL for providing work material.

References 1. Pospisilova, S., Podrabsky, T., Stransky, K., Dobrovska, J.: Heterogenity of Inconel 713 LC and Inconel 738 LC, TMT 2006, pp 249–252. Spain (2006) 2. Egbewnde, A.T., Buckson, R.A., Ojo, O.A.: Analysis of laser beam weldability of Inconel 738 superalloy. Mater. Charact. 61, 569–574 (2010) 3. Geddes, B., Leon, H., Huang, X.: Superalloys, Alloying and Performance. ASM International, Ohio (2010) 4. Shaw, M.C.: Metal Cutting Principles. Oxford Science Publications, USA (1984) 5. Komanduri, R., Schroeder, T.A.: On shear instability in machining a nickel iron base superalloy. Trans. ASME, J. Eng. Ind. 108, 93–100 (1986) 6. Arunachalam, R., Mannan, M.A.: Machinability of nickle-based high temperature alloys. J. Mach. Sci. Technol. 4(1), 127–168 (2000)

26 Grindability and Surface Integrity of Nickel-Based Cast …

337

7. Boothroyd, G., Knight, W.A.: Fundamentals of Machining and Machine Tools, 3rd edn. CRC Press, Taylor & Francis Group, England (2013) 8. Aspinwall, D.K., Soo, S.L., Curtis, D.T., Mantle, A.L.: Profiled superabrasive grinding wheels for the machining of a nickel-based superalloy. Ann. CIRP 56/1, 335–338 (2007) 9. Ding, W., Xu, J., Chen, Z., Su, H., Fu, Y.: Grindability and surface integrity of cast nickel-based superalloy in creep feed grinding with CBN abrasive wheels. Chin. J. Aeronaut. 23, 501–510 (2010) 10. Cameron, A., Bauer, R., Warkentin, A.: An investigation of the effects of wheel-cleaning parameters in creep-feed grinding. Int. J. Mach. Tools Manuf. 50, 126–130 (2010) 11. Guy, B., Tishler, N.: Introduction to Surface Integrity, Pamplet 1, TM70-974, GE, Aircraft Engine Group, Cincinnati, USA (1970) 12. Arunachalam, R.M., Mannan, M.A., Spowage, C.: Surface integrity when machining age hardened IN 718 with coated cutting tools. Int. J. Mach. Tools Manuf. 1481–1491 (2004) 13. Saoubi, R.M., Outeiro, J.C., Chandrasekaran, H., Dillon Jr., O.W., Jawahir, I.S.: A review of surface integrity in machining and its impact on functional performance and life of machined products. Int. J. Sustain. Manuf. 1(1/2), 203–236 (2008) 14. Pawade, R.S., Suhas, S., Joshi, P.K.: Brahmankar, Effect of machining parameters and cutting edge geometry on surface integrity of high-speed turned IN718. Int. J. Mach. Tools Manuf. 48, 15–28 (2008) 15. Thakur, A., Gangopadhyay, S.: State-of-the-art in surface integrity in machining of nickel-based superalloys. Int. J. Mach. Tools Manuf. 100, 25–54 (2016) 16. Vadayar, S., Rani, S.D.: Hot corrosion behaviour of nickel based superalloy. Int. J. Appl. Res. Mech. Eng. (IJARME) 3(1), 2231–5950 (2013) 17. Alloy IN-738 Technical Data, pp. 1–11. The International Nickel Company, Inc., New York 18. Technical Catalogue of Vitrified Tool Room Grinding Wheels, Carborundum Universal Limited, Chennai. https://www.cumi-murugappa.com/abrasives 19. Sinha, M.K., Ghosh, S., Paruchuri, V.R.: Modelling of specific grinding energy for IN718 superalloy. Proc. IMechE Part B: J. Eng. Manuf. 1–18 (2017) 20. Tso, P.-L.: An investigation of chip types in grinding. J. Mater. Process. Technol. 53, 521–532 (1995) 21. Ravuri, B.P., Goriparthi, B.K., Revuru, R.S., Anne, V.G.: Investigations on grinding of IN718 using newly developed graphene nano platelets impregnated grinding wheels, pp. 381-1–381-6. AIMTDR 2014, IIT Guwahati 22. Caggiano, A., Teti, R.: CBN grinding performance improvement in aircraft engine components manufacture. Procedia CIRP 9, 109–114 (2013) 23. Sharman, A.R.C., Hughes, J.I., Ridgway, K.: Workpiece surface integrity and tool life issues when turning Inconel 718™ nickel based superalloy. Mach. Sci. Technol. 8(3), 399–414 (2004) 24. Devillez, A., Lecoz, G., Dominiak, S., Dudzinski, D.: Dry machining of IN718 work piece surface integrity. J. Mater. Process. Technol. 1590–1598 (2011) 25. Yao, C.F., Jin, Q.C., Huang, X.C., Wu, D.X., Ren, J.X., Zhang, D.H.: Research on surface integrity of grinding Inconel718. Int. J. Adv. Manuf. Technol. 65, 1019–1030 (2013) 26. Gupta, U., Hithesh, K., Nandam S.R., Appa Rao, G.: Performance evaluation of coated carbide tools in high speed turning of advanced P/M nickel based superalloy. J. Mater. Sci. Surf. Eng. 5(6), 661–665 (2017) 27. Zhou, J.: Identification of subsurface deformation in machining of IN718. Appl. Mech. Mater. 117–119, 1681–1688 (2012) 28. Zhang, B., Cao, X., Liu, D., Liu, X.: Surface recrystallization of single crystal nickel-based superalloy. Trans. Nonferrous Met. Soc. China 23, 1286–1292 (2013)

Chapter 27

Simulating the Effect of Microstructure in Metal Sliding and Cutting A. S. Vandana

and Narayan K. Sundaram

Abstract Recent experimental discoveries at the mesoscale (100 µm–few mm) in ductile metal sliding and cutting regimes have shown that workpiece surfaces undergo folding by self-contact. The folds subsequently form crack-like features and damage the slid surface. Streaklines of plastic flow associated with surface folding are highly undulating or sinuous. Sinuous flow results in large cutting forces and poor surface finish among other consequences in metal cutting. Thus, understanding the causal mechanism behind surface fold formation and sinuous flow is necessary. It is found from the experiments that microstructure-related property inhomogeneity is a major cause for these phenomena. The aim of this study is to capture surface folding in metal sliding regime in a minimal way and to replicate sinuous flow in the metal cutting regime by introducing microstructure-related property inhomogeneity in the workpiece. In the sliding regime, a square-shaped inhomogeneity is introduced in the workpiece, which is plastically softer or harder than the surrounding substrate. A ‘pseudograin model’ is employed in metal cutting regime, wherein the workpiece surface is divided into a set of ‘grains’ with each grain type plastically different from the others. These models, despite their simplicity, successfully capture all the experimentally observed characteristics of surface folding and sinuous flow in sliding and cutting, thereby establishing the microstructural origin of such complex plastic flows. Keywords Sliding · Cutting · Metals · Plasticity

27.1 Introduction Metal interfaces in sliding-cutting regimes often involve large strain plastic deformation with formation of surface folds in sliding [1] and highly sinuous flow in cutting [2]. Intuitively, ductile metals should be easier to cut, but it is observed that such metals are difficult to machine with large cutting forces [3], thick, wrinkled A. S. Vandana · N. K. Sundaram (B) Indian Institute of Science, Bangalore 560012, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_27

339

340

A. S. Vandana and N. K. Sundaram

chips and poor surface finish [4]. Recent situ experiments [2, 6] at the mesoscale have shown that sinuous flow is the cause for this paradoxical behavior rather than non-uniform deformation and cracking [5]. Experiments have also pointed out that the inhomogeneity associated with metal microstructure is a causal factor for such surface folding and sinuous flow [1, 6]. This paper explores the effect of incorporation of inhomogeneity in both metal sliding and cutting regimes. In metal sliding, we have developed a minimal model to capture surface fold formation. A simple pseudograin model is employed in the metal cutting regime. Lagrangian FE framework is used for both the regimes. While it would seem that direct crystal plasticity finite elements (CPFE) [7] would be the ideal way to incorporate microstructure in metal workpieces, this is not feasible at the high plastic strains involved in these phenomena [8]. In cutting, other microstructurebased models do exist [9, 10], but these have only explored the incipient stages of deformation. It is interesting to note that these simple models capture all experimentally observed features in both the sliding and cutting regimes.

27.2 Procedure ABAQUS Explicit is used for all the simulations. Simulations are done using a pure Lagrangian approach. Figure 27.1 shows the FE schematic for the sliding regime, and Fig. 27.2 shows the pseudograin model for the metal cutting regime. α is the rake angle (correspondingly, θ is the incidence angle); V is the speed at which the specimen/workpiece is pushed against the wedge/tool; h0 is the interaction depth. The yield stress ratio between the grains and the bulk of the specimen is denoted by r y . A yield stress ratio (r y) of less than 1 denotes a soft grain while a value greater than 1 signifies a hard grain. Contact between wedge/tool and specimen/workpiece is modeled using a master-slave contact algorithm. Contact constraints are enforced kinematically using Lagrange multipliers. The friction model applied is a capped Coulomb model with interfacial friction coefficient μ. Low values of friction coefficients (up to μ = 0.2) are used to avoid element distortion. Fig. 27.1 Schematic of FE model used in metal sliding regime

27 Simulating the Effect of Microstructure in Metal Sliding …

341

Fig. 27.2 FE simulation setup and pseudograin model employed in metal cutting regime

27.2.1 Procedure Overview for Metal Sliding Regime As can be seen from Fig. 27.1, the otherwise homogeneous workpiece consists of a square-shaped inhomogeneity. The size of the inhomogeneity/pseudograin is denoted by D, and the depth at which it is located is denoted by d. Values of simulation parameters used are shown in Table 27.1. The hard wedge is modeled as an elastic body with E = 206.8 GPa and ν = 0.29. The specimen used is Al-1100 and is modeled using von Mises isochoric plasticity. The elastic properties for Al-1100 are E = 69 GPa and ν = 0.33. For a detailed procedure, see [11]. Table 27.1 Simulation parameters used in metal sliding regime

Parameter

Values used in simulation

α

−65°, −75° (θ = 25°, 15°)

V

5 mm/s

h0

75 µm

D

380, 250, 130, 50 µm

d

0, 11, 32, 53, 74 µm

μ

0.1, 0.2

342 Table 27.2 Simulation parameters used in metal cutting regime

A. S. Vandana and N. K. Sundaram Parameter

Values used in simulation

α

0°, 5°, 45°

V

5 mm/s

h0

125 µm

D

100, 50, 25 µm

μ

0.1, 0.2

27.2.2 Procedure Overview for Metal Cutting Regime The pseudograin model involves partitioning the workpiece into a set of grains using Voronoi tessellation. Each grain type is similar in elastic properties but has different plastic flow properties (can be softer or harder). The average grain size is denoted by D. As a Lagrangian framework is used, a sacrificial layer is employed to enforce material separation. This separation layer is denoted as S. A thin homogeneous layer (see H in Fig. 27.1) with uniform plastic flow properties is provided between the subdivided layer and separation layer. Such a layer is essential to prevent direct interaction of subdivided grain layer and separation layer. The cutting tool is modeled as a rigid body. Workpiece material is either OFHC Cu or Al-1100 and is modeled using von Mises isochoric plasticity. The elastic properties of OFHC Cu are E = 112 GPa and ν = 0.34. Corresponding values for Al-1100 are E = 69 GPa and ν = 0.33. The other simulation parameters are shown in Table 27.2. Details of the simulation procedure are given in [12] .

27.3 Results and Discussion—Metal Sliding Regime 27.3.1 Interaction of a Wedge with a Homogeneous Specimen Figure 27.3 shows the result obtained with a homogeneous specimen. As can be seen, there is no surface fold formation. The streaklines of flow are laminar and steady (Fig. 27.3b). These results agree with the classical prow model of Challen and Oxley [13].

27.3.2 Folding with a Soft Grain In contrast to the homogeneous specimen, inhomogeneous specimens show surface folding. A softer grain on nearing wedge face forms a bump (Fig. 27.4a). The bump forms a self-contact/fold subsequently (Fig. 27.4b). As the fold traverses the wedge face, the fold grows in length (Fig. 27.4c). On exiting the wedge tip, the fold forms

27 Simulating the Effect of Microstructure in Metal Sliding …

343

Fig. 27.3 Laminar, steady flow in a homogeneous specimen (θ = 25°, μ = 0.2)

Fig. 27.4 Surface fold formation and development with a softer grain (θ = 25°, r y = 0.85, D = 380 µm, μ = 0.2)

a crack-like feature, damaging the surface (Fig. 27.4d). This is in perfect correlation with experimental observations [1, 11].

27.3.3 Folding with a Harder Grain A hard grain also showed surface folding (Fig. 27.5). The grain forms a depression on nearing the wedge face unlike the bump in case of the specimen with softer grain. The depression forms a fold with the fold growing in length as it traverses the wedge and finally exiting the wedge tip as a crack-like feature similar to the case of the specimen with softer grain.

344

A. S. Vandana and N. K. Sundaram

Fig. 27.5 Surface fold formation with a harder grain (θ = 25°, r y = 1.15, D = 380 µm, μ = 0.2)

27.3.4 Implications of the Minimal Model Metal sliding regime is traditionally modeled using the prow model of Challen and Oxley [13]. Such a model is homogeneous, thus failing to capture surface folding. Even homogeneous FE fails to capture surface folding. Here, the minimal model extends homogeneous FE with the introduction of a grain, making the model capable of capturing folding and slid surface damage in a single pass. Thus, this model can impact various surface generation processes because of its ability to accurately predict surface damage due to folding and the formation of crack-like features.

27.3.5 Other Types of Inhomogeneities The grain in the minimal model used for our simulations is perfectly bonded to the workpiece. The bond strength can be varied from zero to perfectly bonded case to see fold-type flow patterns in a broader range of alloys. The zero-bond strength case defines a loosely adhered particle in the workpiece. Such a model is a potential model for a near-surface inclusion.

27.4 Results and Discussion—Metal Cutting Regime Unlike the metal sliding regime, chips are formed in metal cutting. Ductile metals at low-speed cutting produce thick chips with mushroom-like features on the backside. These chips fall under the Type-I category in Nakayama’s [5] classification.

27.4.1 Sinuous Flow Development Figure 27.6 shows a snapshot from an OFHC Cu cutting simulation at t = 0.97 s. The chip root is demarcated by the circular arrow. At the chip root, a bump nucleates

27 Simulating the Effect of Microstructure in Metal Sliding …

345

Fig. 27.6 Sinuous flow in OFHC Cu. The background color shows equivalent plastic strain (α = 0°, V = 5 mm/s, h0 = 125 mm, D = 100 mm)

(as indicated by the green arrow). The bump is pushed against the slowed down chip material to form a self-contact (indicated by purple arrow). The self-contact, which defines one edge of a fold, grows in length and subsequently opens up. This process repeats, giving the characteristic mushroom-like appearance of Type-I chips. It can be also seen that the strain field is highly heterogeneous. This simulation also shows the importance of including microstructure in cutting simulations for capturing sinuous flow.

27.4.2 Grain Deformation in Sinuous Flow Figure 27.7 shows two snapshots from the same simulation as that of Fig. 27.6. Purple and red arrows in panel (a) show two undeformed neighboring grains. Panel (b) shows the same grains when they are in the chip. The grains undergo significant rotation and stretching. The aspect ratio for the final shape of the grains is about nine. This is another characteristic of sinuous flow.

Fig. 27.7 Two snapshots from an OFHC Cu cutting simulation showing grain deformation in sinuous flow

346

A. S. Vandana and N. K. Sundaram

27.4.3 Type-I Chips and Sinuous Flow Type-I chips which are produced in annealed metal cutting result in large cutting forces, large chip thicknesses making such metals difficult to cut. These chips are of low surface finish and quality. Therefore, it is necessary to understand the mechanisms behind Type-I chip formation. The present study shows how sinuous flow results in Type-I chip formation. Understanding sinuous flow is thus the key for controlling Type-I chip formation.

27.4.4 Implications of Pseudograin Model in Metal Cutting Clearly, this simple pseudograin model captures sinuous flow and its associated characteristics. From both sliding-cutting regime simulation results, it can be seen that some form of microstructure-related inhomogeneity model is necessary to capture such complicated folds and associated sinuous flow. It should also be noted that sinuous flow can occur even at low friction and low cutting speeds as the simulations involved such conditions. Thus, like metal sliding regime, here, too complicated models like CPFE are not necessary.

27.5 Conclusion The simulations undertaken in the present study establish microstructure-related inhomogeneity as a cause for surface fold formation in metal sliding and sinuous flow in metal cutting. A minimal model which involves a square-shaped grain embedded in the specimen is able to capture surface fold formation in metal sliding; in its absence, the specimen does not fold. A softer grain produces a bump on workpiece surface, subsequently forming a self-contact/fold and a crack-like feature on exiting the wedge tip, while a harder grain produces a depression instead of a bump. Sinuous flow in metal cutting is reproduced by means of a pseudograin model. A homogeneous layer is introduced to prevent direct interaction between subdivided layer and separation layer. Despite their simplicity, both models successfully capture experimentally observed features of surface fold formation and sinuous flow. More details on experimental comparison, parametric effects and implications of these models are given in [11, 12] and [14]. Acknowledgements This work is supported in part by SERB Grant EMR/2017/002621 from the Department of Science and Technology (DST), Government of India.

27 Simulating the Effect of Microstructure in Metal Sliding …

347

References 1. Sundaram, N.K., Guo, Y., Chandrasekar, S.: Mesoscale folding, instability, and disruption of laminar flow in metal surfaces. Phys. Rev. Lett. 109(10), 106001 (2012) 2. Yeung, H., Viswanathan, K., Compton, W.D., Chandrasekar, S.: Sinuous flow in metals. Proc. Natl. Acad. Sci. 112(32), 9828–9832 (2015) 3. Williams, J.E., Smart, E.F., Milner, D.R.: Metallurgy of machining. Pt. 1. Basic considerations and the cutting of pure metals. Metallurgia 81, 3–10 (1970) 4. Shaw, M.C.: Metal Cutting Principles. Oxford University Press, New York (2005) 5. Nakayama, K.: The formation of saw-toothed chip in metal cutting. In: Proceedings of the International Conference on Production Engineering, Tokyo, vol. 1, pp. 572–577 (1974) 6. Yeung, H., Viswanathan, K., Udupa, A., Mahato, A., Chandrasekar, S.: Sinuous flow in cutting of metals. Phys. Rev. Appl. 8(5), 054044 (2017) 7. Roters, F., Eisenlohr, P., Hantcherli, L., Tjahjanto, D.D., Bieler, T.R., Raabe, D.: Overview of constitutive laws, kinematics, homogenization and multiscale methods in crystal plasticity finite-element modeling: theory, experiments, applications. Acta Mater. 58(4), 1152–1211 (2010) 8. Hansen, N.: Cold deformation microstructures. Mater. Sci. Technol. 6(11), 1039–1047 (1990) 9. Chuzhoy, L., DeVor, R., Kapoor, S., Beaudoin, A., Bammann, D.: Machining simulation of ductile iron and its constituents, Part 1: Estimation of material model parameters and their validation. J. Manuf. Sci. E. T. ASME 125(2), 181–191 (2003) 10. Simoneau, A., Ng, E., Elbestawi, M.: The effect of microstructure on chip formation and surface defects in microscale, mesoscale, and macroscale cutting of steel. CIRP Ann. - Manuf. Technol. 55(1), 97–102 (2006) 11. Vandana, A.S., Sundaram, N.K.: Interaction of a sliding wedge with a metallic substrate containing a single inhomogeneity. Tribol. Lett. 65(4), 124 (2017) 12. Vandana, A.S., Sundaram, N.K.: Simulation of sinuous flow in metal cutting. Tribol. Lett. 66(3), 94 (2018) 13. Challen, J., Oxley, P.: An explanation of the different regimes of friction and wear using asperity deformation models. Wear 53(2), 229–243 (1979) 14. Sundaram, N.K., Mahato, A., Guo, Y., Viswanathan, K., Chandrasekar, S.: Folding in metal polycrystals: microstructural origins and mechanics. Acta Mater. 140, 67–78 (2017)

Chapter 28

What Do Chip Morphologies Tell Us About the Cutting Process? Koushik Viswanathan, Anirudh Udupa, Dinakar Sagapuram and James B. Mann

Abstract Chip morphologies have been documented, classified and characterized for decades within the metal cutting community. Based on these studies, it is generally appreciated that chip morphologies reflect the underlying cutting process. However, specific details of plastic flow processes that are responsible for chip shape, roughness and morphology have hitherto remained largely unclear. In this work, we show a unique correspondence between mesoscale (10–100 μm) plastic flow modes and resulting in chip morphologies. Using high-speed in situ observations and image analysis, we demonstrate three distinct plastic flow modes—all non-homogeneous and unsteady—that are unlike the usually assumed homogeneous shear plane/zone models for metal cutting. These flow modes have their own unique initiation and evolution criteria, as reflected in chip strains and material flow patterns. The latter are computed directly using image correlation techniques. The results show that all prior chip classifications can be collapsed into four primary chip types. Flow transitions between these types are also demonstrated using simple material-independent and non-metallurgical techniques. Keywords Metal cutting · Plasticity · Deformation processing

K. Viswanathan (B) Department of Mechanical Engineering, Indian Institute of Science, Bangalore, India e-mail: [email protected] A. Udupa Center for Materials Processing and Tribology, Purdue University, West Lafayette, USA D. Sagapuram Department of Industrial and Systems Engineering, Texas A&M University, College Station, USA J. B. Mann Department of Mechanical Engineering, University of West Florida, Pensacola, USA © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_28

349

350

K. Viswanathan et al.

28.1 Introduction The process of chip formation in metal cutting involves large strain deformation with effective plastic strains ranging between 1 and 10. In plane-strain (2D) orthogonal cutting, this process has been approximated as occurring via localized shear on a shear plane [1–3]. The basis for this model is the assumption that material flow in the vicinity of the tool is smooth and homogeneous, or in other words, laminar. That this laminar flow assumption is not universally valid has also been recognized for some time. The most common examples occur in high strength and commercially important metals such as Ti and Ni [4–6]. Shear band flow is highly non-homogeneous, and consequently non-laminar, typified by periodic narrow bands of high strain distribution (type 4, Fig. 28.1). These bands represent regions of intense local strain and are separated by large blocks of relatively unstrained material [7–9]. The most commonly accepted mechanism for shear banding is catastrophic shear. Here, localization occurs as a result of continued strain hardening competing with local temperature-induced material softening. In addition to these two chip morphologies, two other chip types (types 1, 3, Fig. 28.1) have been identified based on post-mortem optical observations [10, 11]. However, the precise reason for their formation, as well as the details of their evolution, has hitherto been unclear. It is only using high resolution in situ techniques that these questions are presently being answered [12, 13]. Together, the chip types shown schematically in Fig. 28.1 represent four distinct types that are observed in metal cutting, with some intermediate cases (e.g. characteristics of both types 1, 2) also possible. The unconstrained nature of plastic flow accompanying the chip formation process implies that observed morphologies are naturally reflective of the details of underlying plastic flow. This paper shows that there exists a unique correspondence

Fig. 28.1 Schematic showing typical cutting operation (centre) with four distinct chip morphologies (labelled types 1–4)

28 What Do Chip Morphologies Tell Us About the Cutting Process?

351

between operative plastic flow modes and resulting in chip morphologies. The flow modes are characterized on the mesoscale (10–100 μm) using in situ and ex situ techniques.

28.2 Experimental Configuration Plane-strain cutting, utilizing a linear planing configuration (Fig. 28.2), was used to study the chip formation mechanics and surface plastic flow in low-speed cutting, with V 0 = 1 mm/s, h0 = 100 μm. Out-of-plane deformation was constrained by a glass block clamped against the side of the workpiece. The deformation was observed, in situ, and through the clamped glass block, using a high-speed camera (pco.dimax) at resolution ~1.5 μm. The recorded digital images were then analysed using particle image velocimetry (PIV) to obtain local material velocity and deformation fields. These were then used to virtually compute streaklines to visualize the flow [14], and see Fig. 28.2. A range of workpiece materials was studied: annealed OFHC Cu (68 HV) and half-hard OFHC Cu (120 HV), representing highly workable (ductile) and partially hardened metals of moderate workability, respectively; and Ti-6Al-4V (346 HV) and Inconel 718 (458 HV) representing systems that readily flow localize. High-speed steel tools (Mo Max, McMaster) were used, all with rake angle γ = 0°. Shear band flow, primarily a high-speed cutting phenomenon, was analysed in 2D (rotary) cutting (0.5–10 m/s) [9, 15]. For material displacements under these conditions (strain rates 104 –105 /s), a micro-marker technique was used to map local material displacements. This involved scribing parallel markers on the workpiece surface prior to cutting and measuring their corresponding shape/position after deformation using an SEM. By comparing the initial undeformed marker shapes with the final deformed shapes, local material displacement and strain field information could be calculated [15].

Fig. 28.2 Experimental configuration showing plane-strain orthogonal cutting and associated in situ imaging framework. Images are recorded using a high-speed camera attached to an optical microscope. The digital images are processed to obtain deformation history (strain, strain rate)

352

K. Viswanathan et al.

28.3 Results The observations have revealed a unique correspondence between operative mesoscale plastic flow modes and resulting in chip morphologies. Each of the chip types (1–4) shown in Fig. 28.1 is now individually discussed.

28.3.1 Type 2 (Laminar Flow) Laminar flow, or smooth homogeneous shear deformation, is considered the ‘norm’ in metal cutting practice. This type of flow (corresponding to chip type 2) forms the basis for the shear zone/plane models that are widely used to describe the mechanics of cutting [1–3]. Figure 28.2 shows flow in the deformation zone in cutting halfhard Cu, as revealed by the computed virtual streaklines (overlaid). Laminar flow is reflected in both the streaklines (smooth, uniform) and the strain field (homogeneous, not shown). The corresponding strain rate field (not shown) is very reminiscent of classical shear zone models, with a well-defined zone of concentrated shear. Consequently, the chip thickness ratio λ = hc /h0 ~ 2 with the corresponding cutting force ~400 N. The strain in the chip is quite uniform, with effective plastic strain between 1.5 and 2, and comparable to that (~1.5) estimated using the shear plane model [1]. Very minor deviations from this idealized model are observed (e.g. small-scale features/protrusions on the back surface), and the chip is clearly of type 2 (Fig. 28.1).

28.3.2 Type 1 (Sinuous Flow) While laminar flow was observed when cutting pre-strained or hardened materials (such as hardened Cu), a vast majority of highly strain hardening and soft materials (e.g. annealed Cu, Al or even stainless steels) do not show a type 2 chip. Instead, the corresponding chip is of type 1 (Fig. 28.1) and consists of characteristic mushroomshaped features on the chip back surface. Correspondingly, in situ analysis revealed the underlying flow mode to be fundamentally different from laminar flow, and see Fig. 28.4. The wavy streaklines overlaid on the image reflect large-amplitude material folding and significant local rotation. This type of deformation is referred to as sinuous flow [16]. The chip consists primarily of such folds stacked on top of each, resulting in very large thickness, λ ~ 12. When compared to laminar flow, this λ is consistently 6–8 times larger. Folding, and consequently sinuous flow, is not just a transient process. The entire chip from the beginning of the cut to the end, beyond so-called steady state, is comprised of a series of folds stacked on top of each other. This is clearly reflected in the streaklines in Fig. 28.4.

28 What Do Chip Morphologies Tell Us About the Cutting Process?

353

Folding is triggered by a plastic buckling instability ahead of the tool, and its occurrence is not governed by lubrication at the chip–tool interface [13, 17]. In fact, folding and sinuous flow occurred irrespective of the presence/absence of a cutting lubricant (Mobil 1 5w-30). The corresponding strain field is also reflective of the highly non-homogeneous nature of the plastic flow (Fig. 28.4). The chip consists of alternating regions of high and low strains, coincident with the folds themselves. The strain alternates between 4 and ~10 along the chip length. This inhomogeneity arises due to the flow: Repeated material folding leads to alternate regions of high and low strain, similar to the plastic buckling/bending of a beam [17]. Naturally, as this strain field indicates, sinuous flow is accompanied by severely redundant deformation. The cutting force (for annealed Cu) was ~1600 N, nearly four times that when cutting half-hard Cu (Fig. 28.3). Similarly, the representative volume averaged strain in the chip (7.95) was five times that in the laminar flow chip. A characteristic feature of sinuous flow, and one that has been often observed using purely post-mortem analyses, is the occurrence of mushroom-like structures on the backside of the chip [Fig. 28.4 (right)]. This is characteristic of the type 1 chip and results from the folding process as revealed by our in situ observations. These features can hence be used to uniquely identify the occurrence of sinuous flow from ex situ observations. The mushroom morphology has been interpreted in the past [10] as arising from cracking on the chip back surface. However, the present observations clearly show that it is caused by sinuous flow/folding, the latter also explaining the very large chip thickness. Furthermore, the workpiece surface created via sinuous flow always showed significant surface degradation (comprised of cracks and tears) as well as poor surface finish, in comparison with laminar flow. Importantly, a layer Fig. 28.3 Typical in situ image showing laminar flow in cutting of hardened Cu. The strain field (background colour) and streaklines are digitally computed using image correlation techniques

354

K. Viswanathan et al.

Fig. 28.4 Sinuous flow and formation of type 1 chip. Image on the left shows highly heterogeneous strains underlying sinuous flow, with high strains concentrated in folded regions. Consequently, the chip free surface is characterized by mushroom-shaped features (right)

of residual strain (of thickness ~ h0 , strain > 3) was also formed on the sub-surface. This may be compared with an equivalent surface formed via laminar flow (strain ~ 0.3).

28.3.3 Type 3 (Segmented Flow) In some highly strain hardened metals (e.g. α-brass), cutting occurs via the formation of neither a type 1 nor type 2 chip, but instead proceeds in the following manner [18]. Deformation is largely homogeneous during incipient stages leading to strain accumulation in the chip. At a critical value of accumulated strain, a crack is nucleated on the chip free surface and propagates towards the tool tip. The distance of propagation is a function of the rake angle γ , the degree of pre-hardening, and the cutting velocity. This crack nucleation and growth occur periodically, leading to a segmented type of chip (type 3, Fig. 28.1). The strain field in this type 3 chip is again non-homogeneous, and see Fig. 28.5 (left). The arrow at Q1 shows a single crack that has propagated towards the tool tip. Q2 shows another incipient crack nucleation event in an adjacent segment. It is clear from the strain field that crack nucleation starts at a certain value of accumulated strain. That this type of chip is a result of pseudo-brittle fracture is also evident from observations in the width direction; see Fig. 28.5 (right). The crack (red arrows) runs through the entire width of the workpiece and is not arrested at grain boundaries or other heterogeneities such as small folds (yellow arrows). In comparison with type 1 (sinuous flow), cutting forces with the type 3 chip are much lower, comparable with that a type 2 chip (laminar flow). Correspondingly, the mean chip thickness ratio λ ~ 3. It is also noteworthy that the cracks are oriented roughly parallel to an imaginary shear plane running from the tool tip to the workpiece free surface.

28 What Do Chip Morphologies Tell Us About the Cutting Process?

355

Fig. 28.5 Segmented flow, with periodic crack initiation on the chip free surface. Strain field (left) is again heterogeneous, with strain concentrated close to crack initiation points. The SEM image (right) shows fracture planes bounding each segment

28.3.4 Type 4 (Shear Banding) Shear banding has been studied for a very long time, beginning with the pioneering experiments of Tresca. This flow mode is characterized by severe flow localization resulting from adiabatic shear [6–9, 19, 20] and results in a type 4 chip. While shear banding represents a more extreme strain heterogeneity compared to sinuous flow (type 1) or segmented flow (type 2). A recent study [9], using in situ imaging, has shown that shear banding arises from a two-stage process: an initiation phase involving the development of a weak material zone and a subsequent evolution phase where strain imposition occurs. In the initiation stage, the material ‘softens’ locally, establishing a weak interface/plane extending from the tool tip to workpiece free surface. Most of the localized straining then occurs in the evolution stage, where the chip material slides along this weak interface (the shear band itself). In this regard, even though the morphologies of type 3 and 4 chips might appear similar at first glance, the underlying mechanics is very different. Type 3 chips (segmented flow) form via periodic fracture propagation from the workpiece free surface while type 4 chips are formed via the two-stage shear banding process. Figure 28.6 (top row) shows a typical shear-banded chip in high-speed cutting experiments with Ti-6Al-4V. Intense shear (strain > 20) is evident from this optical micrograph image: An initially equiaxed grain has been sheared into two contiguous segments (A and A ) by an intersecting shear band. The grain boundary crossing the band is highlighted in the image (white dotted line) for reference. Similar high shear strains have been reported in a few instances elsewhere [4, 5, 21, 22]. Notice that no fracture is observed within the band itself. Strain distribution in the vicinity of the shear band interface was obtained by tracking the local curvature of pre-inscribed micro-markers in the chip (see Sect. 28.2). The assumption that loading occurs by simple shear enabled the computation of shear strains at these locations along the band interface. Strain profiles on both sides of the

356

K. Viswanathan et al.

Fig. 28.6 Properties of shear banding flow. (Top row) Type 4 chip in cutting of CP Ti, V 0 = 3 m/s. Note that displacements are continuous across the band (as seen by the continuous grain boundary). (Bottom row) Strain field in the vicinity of a single band shows very high values (strain ~ 9) close to the band with very little deformation away from it

interface were computed in this manner for more than ten adjacent markers and a 2D strain map obtained by linear interpolation, and see Fig. 28.6 (bottom row). The strain distribution in segment A is shown superimposed on the raw SEM image. The severe strains and flow localization are clearly seen in the figure. The strain reaches zero very quickly with increasing distance (y) from the interface, and this is shown along four different markers (white dashed lines in strain image, bottom row, right). From these profiles, it is clear that the strain in the vicinity of the band can reach very large values (ε ~ 20), with data in each adjacent segment identical; see Fig. 28.6 (bottom row).

28 What Do Chip Morphologies Tell Us About the Cutting Process?

357

28.3.5 Transitions Between Flow Modes Given the observations linking distinct flow modes uniquely to resulting chip morphologies, the following two questions are now pertinent: In a specific cutting operation, which flow mode is the most preferred? Is it possible to transition from any given flow mode (chip type) to this desirable flow mode without any metallurgical alterations to the workpiece? To answer the first question, we note that type 1 chip (sinuous flow), with its highly redundant deformation, is almost always concomitant with the occurrence of surface defects. Thus, cutting via sinuous flow not only involves large cutting forces, but also results in very poor surface finish. Of all four flow modes, it is perhaps the least desirable from a cutting point of view. At the other end of the spectrum, the type 2 chip (laminar flow) is ideal for it involves low cutting forces and very good surface finish. Unfortunately, this flow mode appears to be more the exception than the norm in several cutting operations and with many materials. The second question, namely can we transition from a type 1 (not preferred) to a type 2 chip, is answered by considering the mechanics of sinuous flow and folding. In earlier work [12, 13], we have shown that folding and the type 1 chip arise due to plastic buckling on the material free surface. Buckling may be suppressed, and hence, sinuous flow was avoided by three different techniques. The first (see Fig. 28.7) involves the use of a constraint ahead of, and parallel with, the cutting tool. This effectively imposes a net compressive force on the workpiece free surface, thereby preventing buckling and resulting in laminar flow. The second [12] is pre-hardening the material, a method that is known to result in laminar flow. Finally, a change in rake angle γ (more positive) also suppresses buckling [13], folding and hence sinuous flow. In all three cases, sinuous flow gives way to laminar flow and a type 1 chip is changed to type 2. Other examples of flow transitions have also been studied, in particular the transition from type 1 to type 3 via the use of mechanochemical effects. Details of these transitions are discussed elsewhere [12, 23]. Fig. 28.7 Transitions between flow modes may be effected in various ways. This figure shows one such route to go from sinuous (type 1) to laminar (type 2) chip via constraint application

358

K. Viswanathan et al.

28.4 Implications and Conclusions The observations establish a direct and one-to-one correspondence between operative plastic flow modes and observed chip morphologies. These results raise points of interest and questions that have consequences for the foundations of metal cutting and large strain deformation phenomena in general. (a) The occurrence of the type 1 chip, due to plastic buckling, material folding and sinuous flow, in low-speed cutting of annealed pure metals, poses obvious questions with regard to predictive models. Only recently have finite element (FE) methods been developed to predict these types of flows [24, 25]. These models are necessarily phenomenological since physics-based simulations (e.g. using crystal plasticity) are incapable of handling the large strains inherent to cutting. (b) Sinuous flow may be expected to occur in a variety of material systems, based on the factors that promoting plastic buckling (e.g. workability, strain hardening). This is confirmed by ex situ chip morphology observations in material systems such as pure metals, Ni alloys, stainless steels, mild steel and even ductile polymers. (c) The different plastic flow modes reported here are strongly dependent on the deformation state of the workpiece. Based on the results presented here and elsewhere on flow transitions, it is quite possible that all four flow modes can be reproduced possibly in the same rudimentary material system (such as Cu) by varying the initial material state and cutting velocities. (d) It is clear from the different types of flow modes and the material instabilities that govern them (e.g. buckling, fracture, localization) that cutting is, perhaps, the most difficult plastic deformation process to model. (e) In the light of the four chip types and corresponding flow modes described in this work, another question worth pondering is: Do other unsteady flows exist, corresponding to hitherto undocumented chip types? Given the one-toone correspondence reported in this work, careful multi-scale analyses of chip morphologies could hold the key to uncovering as-yet unobserved flow modes in metal cutting. In closure, we note that the shear plane/zone model, proposed nearly 75 years ago for cutting of highly ductile metals, certainly warrants a closer revisit. The fact that chip formation can occur via different material instabilities with the resulting deformation bearing little resemblance to the shear zone model is quite illuminating. Together, these observations strongly suggest that chip morphology is fundamentally determined by the stability of underlying plastic flow.

28 What Do Chip Morphologies Tell Us About the Cutting Process?

359

References 1. Merchant, M.E.: Mechanics of the metal cutting process. I. Orthogonal cutting and a type 2 chip. J. Appl. Phys. 16(5), 267–275 (1945) 2. Piispanen, V.: Theory of formation of metal chips. J. Appl. Phys. 19(10), 876–881 (1948) 3. Okushima, K., Hitomi, K.: On the Cutting Mechanism for Soft Metals. Memoirs of the Faculty of Engineering, Kyoto University, vol. 19, pp. 135–166 (1957) 4. Zener, C.: The micro-mechanism of fracture. In: Fracturing of Metals. ASM, pp. 3–31 (1948) 5. Kravz-Tarnavskii, V.P.: Zeitschift der Russischen Metallurgischen (A peculiar band discovered in steel), vol. 3, pp. 162–167 (1928) (in Russian). See translation at Dodd, B., Walley, S.M., Yang, R., Nesterenko, V.F.: Major steps in the discovery of adiabatic shear bands. Metall. Mater. Trans. A 46(10), 4454–4458 (2015) 6. Recht, R.F.: Catastrophic thermoplastic shear. J. Appl. Mech. 31(2), 189–193 (1964) 7. Davies, M.A., Burns, T.J., Evans, C.J.: On the dynamics of chip formation in machining hard metals. CIRP Ann. Manuf. Technol. 46(1), 25–30 (1997) 8. Komanduri, R., Von Turkovich, B.F.: New observations on the mechanism of chip formation when machining titanium alloys. Wear 69(2), 179–188 (1981) 9. Sagapuram, D., Viswanathan, K., Mahato, A., Sundaram, N.K., M’Saoubi, R., Trumble, K.P., Chandrasekar, S.: Geometric flow control of shear bands by suppression of viscous sliding. Proc. Roy. Soc. A Math. Phys. Eng. Sci. 472, 20160167 (2016) 10. Nakayama, K.: The formation of saw-toothed chip in metal cutting. In: Proceedings of the International Conference on Production Engineering, Tokyo, vol. 1, pp. 572–577 (1974) 11. Davies, M.A., Chou, Y., Evans, C.J.: On chip morphology, tool wear and cutting mechanics in finish hard turning. CIRP Ann. Manuf. Technol. 45(1), 77–82 (1996) 12. Viswanathan, K., Udupa, A., Yeung, H., Sagapuram, D., Mann, J.B., Saei, M., Chandrasekar, S.: On the stability of plastic flow in cutting of metals. CIRP Ann. 66(1), 69–72 (2017) 13. Udupa, A., Viswanathan, K., Ho, Y., Chandrasekar, S.: The cutting of metals via plastic buckling. Proc. R. Soc. A Math. Phys. Eng. Sci. 473(2202), 20160863 (2017) 14. Viswanathan, K., Mahato, A., Yeung, H., Chandrasekar, S.: Surface phenomena revealed by in situ imaging: studies from adhesion, wear and cutting. Surf. Topogr. Metrol. Prop. 5(1), 014002 (2017) 15. Sagapuram, D., Viswanathan, K.: Viscous shear banding in cutting of metals. J. Manuf. Sci. Eng. 140(11), 111004 (2018) 16. Yeung, H., Viswanathan, K., Compton, W.D., Chandrasekar, S.: Sinuous flow in metals. Proc. Natl. Acad. Sci. 112(32), 9828–9832 (2015) 17. Shanley, F.R.: Inelastic column theory. J. Aeronaut. Sci. 14(5), 261–268 (1947) 18. Guo, Y., Compton, W.D., Chandrasekar, S.: In situ analysis of flow dynamics and deformation fields in cutting and sliding of metals. Proc. R. Soc. A Math. Phys. Eng. Sci. 471(2178), 20150194 (2015) 19. Gente, A., Hoffmeister, H.W., Evans, C.J.: Chip formation in machining Ti6Al4V at extremely high cutting speeds. CIRP Ann. Manuf. Technol. 50(1), 49–52 (2001) 20. Shaw, M.C., Vyas, A.: Chip formation in the machining of hardened steel. CIRP Ann. Manuf. Technol. 42(1), 29–33 (1993) 21. Timothy, S.P., Hutchings, I.M.: The structure of adiabatic shear bands in a titanium alloy. Acta Metall. 33(4), 667–676 (1985) 22. Wingrove, A.L.: The influence of projectile geometry on adiabatic shear and target failure. Metall. Trans. 4(8), 1829–1833 (1973) 23. Udupa, A., Viswanathan, K., Saei, M., Mann, J.B., Chandrasekar, S.: Material-independent mechanochemical effect in the deformation of highly-strain-hardening metals. Phys. Rev. Appl. 10(1), 014009 (2018) 24. Sundaram, N.K., Mahato, A., Guo, Y., Viswanathan, K., Chandrasekar, S.: Folding in metal polycrystals: microstructural origins and mechanics. Acta Mater. 140, 67–78 (2017) 25. Vandana, A.S., Sundaram, N.K.: Simulation of sinuous flow in metal cutting. Tribol. Lett. 66(3), 94 (2018)

Chapter 29

Simultaneous Optimization of Milling Process Responses for Nano-Finishing of AISI-4340 Steel Through Sustainable Production Muhammed Muaz

and Sounak Kumar Choudhury

Abstract A modified Taguchi-grey relational analysis method has been used in this research work to optimize the flat end milling process considering chip compression ratio (Q) and workpiece surface roughness (Ra ) simultaneously. Both of the output responses are of equal importance. Therefore, optimization of the process considering both of them at the same time is more practical than optimizing the process considering only a single response at a time. The effectiveness of dry cutting has been experimentally investigated as compared to flooded lubrication condition giving equal weights to the two important process responses simultaneously. From the findings of the analysis and the experimental results, it is recommended to perform flat end milling operation on AISI 4340 steel in dry cutting condition at high speed and low feed rate. Flooded lubrication technique is not feasible for flat end milling of this steel. Performing dry machining on the recommended cutting parameters will lead to cleaner and sustainable production which aims at reducing/omitting waste and making the process environment-friendly. The order of significance of the factors based on the analysis, in sequence, is spindle speed, feed rate, and machining environment. Analysis of variance (ANOVA) performed on grey relational grades confirms the order of significance. Keywords Modified Taguchi-grey relational analysis · Multi-objective optimization · Nano-finishing · Sustainable production · Flat end-milling

29.1 Introduction Flat end milling operation with indexable inserts is widely used in metal cutting industry. However, distortion of workpiece surface quality, high forces, large chip compression ratio, and abundant usage of hazardous metal cutting fluids during milling are among the major concerns. These problems are directly associated with M. Muaz (B) · S. K. Choudhury Department of Mechanical Engineering, Indian Institute of Technology Kanpur, Kanpur 208016, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_29

361

362

M. Muaz and S. K. Choudhury

cutting parameters and lubricating conditions. An unambiguous and deep-seated understanding of the optimum levels of input parameters and their order of significance is necessary for the improvement of the process and for making the whole process environmentally sustainable. Optimization of the cutting process is very complex because it includes continuous as well as discrete input parameters along with non-correlated output responses. Traditional techniques like Taguchi method gives only local optimum solution rather than a global one [1]. Integrating Taguchi method with GRA (TGRA) is a promising technique which provides a single optimum condition by optimizing the grey relational grades (GRG) obtained by combining all output responses. Introducing robustness in TGRA made this technique very useful in improving the quality of the product as well as the cutting process by selecting and optimizing cutting parameters and lubricating conditions carefully. Industrial activities are responsible for the release of air pollutants into the environment causing detrimental effects on the environment and the human health. Dry machining is believed to result in poor machining performance like high forces and deteriorated surface finish. Therefore, lubricant usage in bulk is a common practice in the metal cutting industry. Flooded lubrication technique is responsible for high cost and environmental/health issues. Diseases related to dermatology and respiration are common to the machinists working in the environment highly polluted with the fluids fumes. Improper disposal of the waste fluid pollutes water and soil too. Asthma patients, through current medical investigations, were found to be seriously affected by environmental factors like particle sizes and air-suspended metals [2]. Therefore, it is desirable to eradicate the hazardous cutting fluids completely from metal cutting industry as far as environmental issues are concerned [3]. From environmental point of view, dry cutting is the best solution because of no problem of hazardous cutting fluid disposal, danger to operator’s health, and pollution to water and air. However, suitable measures have to be adopted to compensate for the primary functions of the cooling lubricant. One method is to optimize the whole problem considering the important output responses simultaneously. Multi-objective optimization of manufacturing processes has been considered by various researchers through improved/integrated statistical techniques [4–6]. Integrating Taguchi method with GRA is also an effective way of optimization. Julong provided a detailed introduction to the grey system theory. According to him, a grey system means some of the information about the system is unknown [7]. A complex multiple responses optimization problem can be simplified by calculating a single optimum grey relational grade by using grey relational analysis [1, 8]. Most of the researchers, who used Taguchi-based grey relational analysis, did not use S/N ratios of the responses for GRA determination, although it is a fundamental concept in Taguchi analysis. For example, electric discharge machining was optimized by Lin and Lin with multi-performance characteristics using an orthogonal array with GRA and showed effective improvement in the process performance through this approach [9]. Lin applied Taguchi method with GRA without using S/N ratios for optimization of turning operation considering tool life, surface roughness, and cutting force as performance characteristics [10]. Taguchi method was used with grey relational analysis to optimize the thin-film sputtering process with multiple quality

29 Simultaneous Optimization of Milling Process Responses …

363

characteristics in color filter manufacturing by Chiang and Hsieh [11]. They also considered actual responses to perform GRA and did not use S/N ratios for GRA. End milling of Inconel 718 super alloy using Taguchi-based grey relational analysis was optimized by Maiyar et al. [12] without considering S/N ratios. Sarıkaya et al. used Taguchi and GRA for multi-response (vibration signals, cutting force, and surface roughness) optimization of milling of AISI 1050 steel [13]. Sarikaya and Güllü performed Taguchi-based grey relational analysis to optimize cutting velocity, flow rate for MQL, and cutting fluids for MQL in turning difficult-to-cut alloy Haynes 25 considering Ra , flank wear, and MRR simultaneously [14]. Both of the groups [13, 14] did not apply GRA on the S/N ratios of the responses. They used S/N ratios of grey relational grades for further analysis. To summarize, generally, GRA was not performed on S/N ratios of the responses in TGRA, although it makes the optimization technique more robust by reducing the variation in the response values due to noise factors. Even if the optimization of milling operation has been considered elsewhere, no study has been systematically performed for optimization of the finishing process through flat end milling operation considering the two equally important responses simultaneously aiming to circumvent hazardous metal cutting fluids completely. In this study, flat end milling operation performed on AISI 4340 steel at dry and flooded lubricating conditions and different speed-feed combinations by using multi-layered TiCN/Al2 O3 /TiN CVD coated tungsten carbide inserts was investigated experimentally and statistically. An attempt has been made to understand whether the use of flooded lubrication can be circumvented effectively during flat end milling of the steel by using a newer generation coated carbide tool performing optimization of the whole process intelligently. Optimization of the flat end milling process was done through the modified TGRA to achieve minimum surface roughness and lower chip compression ratio simultaneously. As a contrast to other studies, in the present work, grey relational analysis using equal weights for both of the responses was performed on the S/N ratios of the response values. Results of this study can be utilized by all manufacturing industries where milling process is used for machining steels. Prudent utilization of the results would impact the environment positively. The result obtained in the present work was one of its kinds when both of the two important output responses were optimized simultaneously giving equal weights to them as well as performing grey relational analysis on their S/N ratios. Henceforth, the materials and the experimental techniques used were described, in Sect. 29.2 while results were discussed through graphical analysis first, and then optimization was performed in Sect. 29.3. Finally, conclusions were given in Sect. 29.4.

29.2 Materials and Methods Flat end milling tests were performed on EMCO Concept Mill 250 Machining Center (Fig. 29.1) with a maximum spindle speed of 10,000 rpm and 7 kW power using two

364

M. Muaz and S. K. Choudhury

Fig. 29.1 Work plan

indexable multi-layer TiCN/Al2O3/TiN CVD coated tungsten carbide inserts as a tool and AISI 4340 steel as workpiece material in the form of a block (250 mm × 100 mm × 32 mm). Tooling parameters were kept fixed throughout the experiments. Milling cutter has 90° approach angle and 16 mm diameter. The tool holder (F 4042.W16.016.Z02.08) and the inserts (ADMT 080304 R-D56) were made by Walter Tools. Input parameters and their levels were selected according to the manufacturer’s recommendation, machine capabilities, and literature survey (Table 29.1). Spindle speed was taken in RPM and feed rate in mm/min because in the CNC milling machine used, the speed and feed rate could be given only in RPM and mm/min, respectively. L18 (21 × 32 ) orthogonal array with three factors and 18 runs was selected based on Taguchi design of experiments method using MINITAB 18. The depth of cut in all experiments was taken as 0.5 mm. The surface roughness was measured by using SURFTEST SJ-210. The instrument used for measuring the surface roughness gave values in μm. Therefore, the unit of Ra taken in this Table 29.1 Input variables and their levels

Parameters Cond.

Spindle speed (RPM)

Feed rate (mm/min)

Levels

A

B

C

1

Dry

2500

500

2

Flooded

5000

1000

7500

1500

3

29 Simultaneous Optimization of Milling Process Responses …

365

study was μm. As the results obtained were less than 1 μm, the term nano-finishing was used. Deformed chip thickness was measured by observing the chips under a digital microscope (Model: AM7013MZT4 Dino-Lite Premier) with a maximum magnification of 230× and the resolution of 5 Megapixel. Tool flank wear was also observed under the same microscope. The tool wear criterion was taken as 0.2 mm for flank wear [15]. No significant tool wear was observed during the experiments because highly wear resistant multicoated milling inserts were used. The cutting fluid used was an emulsion of cutting oil, ServoCut S (Make: Indian Oil Corporation Limited, India), dispersed homogeneously in water in the ratio of 1:20. ServoCut S meets the specification; IS: 1115-1986 (Reaffirmed 1996) as a performance standard. The hardness of the workpiece was tested by Rockwell type Macromet I hardness tester, make: Buehler Ltd. USA.

29.3 Results and Discussion 29.3.1 Graphical Analysis The effects of lubricating condition on surface roughness (Ra ) and chip compression ratio (Q) are discussed first in this section through graphical analysis. It is clear from Fig. 29.2 that at most of the cutting conditions, Ra was less for dry condition as compared to flooded lubrication. This is attributed to the hardness of the cooled down chips that reduced the surface finish during flooded lubrication. Losing heat

Fig. 29.2 Surface roughness at all machining conditions

366

M. Muaz and S. K. Choudhury

Chip Compression Coefficient

to the cutting fluid, the chips were suddenly cooled down and got hardened due to quenching. These chips made scratches on the finished surface of the workpiece which was softer than the quenched chips. No significant built up edge and tool wear were developed because of low depth of cut taken for finishing operation on relatively softer steel. Chip compression ratio was analyzed to address the technological differences between dry and wet machining. Large chip compression ratio indicates more deformation of workpiece material which in turn depicts low process efficiency. Therefore, low ratio is required. Figure 29.3 shows that the ratio is always lower in case of flooded lubrication condition as compared to dry cutting. The rationale for low ratios in case of flooded lubrication is the cooling effect of the emulsion applied during the cutting. High-specific heat of the emulsion made it possible to extract the heat suddenly from the cutting zone. Therefore, the deformation of the chip due to heating was reduced, and hence the low ratio was obtained. As far as the effects of feed and speed on the ratio were concerned, the ratio decreased with increase in spindle speed and increased with increase in feed rate. It is clear from the above discussion that wet machining was found to be better if only chip compression ratio was considered, whereas dry machining was found to be better at most of the speed-feed combinations when only workpiece surface roughness was considered. When both of the response parameters were considered simultaneously, complex relationships exist between the selected machining conditions and the output responses. Therefore, concrete conclusions could not be drawn by merely observing the patterns followed by the responses with respect to the machining

Fig. 29.3 Chip compression ratio at all machining conditions

29 Simultaneous Optimization of Milling Process Responses …

367

conditions. Hence rigorous statistical analysis and single as well as multi-objective optimization were performed in the following sections.

29.3.2 Optimization First of all Taguchi analysis was done in a conventional manner to determine the optimum levels of input parameters considering one response at a time. Secondly, modified grey relational analysis-assisted Taguchi method was performed to determine the optimum experimental conditions considering both responses simultaneously. Mathematical calculations were carried out using Microsoft Excel and MINITAB 18. Taguchi analysis was first performed considering chip compression ratio and surface roughness one by one as output response. As both of the responses must be at their lowest levels in an ideal condition, smaller-is-better criterion was used to calculate S/N ratios by using Eq. 29.1. S/N = −10 log10

   1 yi2 n

(29.1)

where n is the number of measurements in a trial/row, and yi is the measured value. The optimum level of a parameter is the one having largest value of S/N ratio. Considering only Q, the optimum parameters are high spindle speed (7500 rpm), low feed rate (500 mm/min), and flooded lubrication (Table 29.2) while considering only Ra ; optimum parameters are dry cutting environment, high spindle speed (7500 rpm), and low feed rate (500 mm/min) (Table 29.2). Rank shown in response table and Fvalue in ANOVA table depict the significance of a parameter among three. From Tables 29.2 and 29.3, spindle speed was found to be the most significant factor. Hybrid Taguchi-Grey Relational Analysis-Multi-objective optimization. In this approach, the problem of optimizing two output responses is converted into a single objective problem. This technique is neither a purely Taguchi analysis technique nor GRA technique but a combination of the two. Conventional Taguchi method Table 29.2 S/N response table for Q and Ra Q

Ra

Level

A

B

C

Level

A

B

C

1

−4.106

−4.731

−3.801

1

6.988

2.918

6.882

2

−3.882

−3.877

−3.966

2

5.653

6.595

6.477

−3.259

−4.100

9.449

5.664

Delta

0.224

1.472

0.298

Delta

1.335

6.531

1.158

Rank

3

1

2

Rank

2

1

3

3

3

0.226

6.555

0.268

0.252

7.301

B

C

Er

Tl

SS

A

Q

17.0

12.0

2.0

2.0

1.0

Dof

0.429

0.021

0.134

3.277

0.226

MSS

Table 29.3 ANOVA table for Q and Ra

6.392

156.29

10.776

F

Sig.

0.987

1.000

0.993

P

0.013

0.000

0.007

155.085

18.194

4.242

128.639

4.010

SS

Ra

17.0

12.0

2.00

2.00

1.00

Dof

9.123

1.516

2.121

64.320

4.010

MSS

1.399

42.423

2.645

F

Sig.

0.716

1.000

0.870

P

0.28

0.00

0.13

368 M. Muaz and S. K. Choudhury

29 Simultaneous Optimization of Milling Process Responses …

369

can be applied to only a single response at a time and gives different results for different output parameters. GRA gives a particular grade to each experimental run and describes which run gives better results. GRA alone neither determines the significance of the factors nor gives optimum levels of parameters other than the one used in experimental runs. However, in cases where more than one responses are to be optimized, and the significance of the factors are to be determined simultaneously, the hybrid Taguchi-grey relational analysis is a very efficient technique. In this study, Taguchi method was first used for the design of experiments, and the L18 orthogonal array was selected. S/N ratios are then determined for all output responses. S/N ratio is the transformed form of loss function which is a measure of the deviation of a performance characteristic from its desired value. The use of S/N ratio makes this technique insensitive to the variation of the noise factors. Then grey relational analysis (GRA) was performed to combine the S/N ratios of both of the output responses for each experimental run with assigning equal weights to both of the responses. Again, Taguchi method was applied to grey relational grades to determine the optimum levels and the order of significance of input parameters. Problem formulation In mathematical terms, the problem for this study can be described as: “Minimize: f (Q, Ra ),” subject to independent decision variables as: cutting environment, A; spindle speed, B (rpm); feed rate, C (mm/min). The range of input decision variables should be: A, Dry/Flooded; B, 2500 ≤ 5000 ≤ 7500; C, 500 ≤ 1000 ≤ 1500 The problem is then converted to a single optimization problem as: “Maximize GRG; 0 < GRG ≤ 1,” subjected to the machining parameters in the range described earlier. Optimization Methodology Step-by-step procedure for the method used in this study is described in this section. L18 orthogonal array was chosen using MINITAB 18 software for the selected input parameters and their levels. Following steps have been applied to grey relational analysis-assisted Taguchi method. Step 1: Recording experimental values of the responses Experimental values of surface roughness and chip compression ratio were recorded in triplets, and average of the three readings were plotted in Figs. 29.2 and 29.3. Step 2: Calculation of S/N ratios for each response value S/N ratios were calculated for both of the responses separately for each experimental run by using Eq. (29.1). Smaller-the-better criterion was used for S/N ratios of both of the responses. S/N ratios are tabulated in Table 29.4.

0.282 0.156 0.000 0.585 0.546 0.481 0.871 0.808 0.787 0.355 0.282 0.233 0.746 0.664 0.611 1.000 0.935 0.913

−4.591

−4.868

−5.212

−3.922

−4.009

−4.151

−3.292

−3.431

−3.477

−4.429

−4.591

−4.697

−3.568

−3.747

−3.864

−3.007

−3.151

−3.198

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

Mean GRG, Grmean = 0.562

3

3

3

2

2

2

1

1

1

3

3

3

2

2

2

1

1

3

2

1

3

2

1

3

2

1

3

2

1

3

2

1

3

2

1

1

1

1

x i (1)

Q

S/N

C

Exp no.

B

A

L18 array

Table 29.4 Taguchi-grey relational analysis

0.087

0.065

0.000

0.389

0.336

0.254

0.767

0.718

0.645

0.213

0.192

0.129

0.519

0.454

0.415

1.000

0.844

0.718

i (1)

0.852

0.885

1.000

0.563

0.598

0.663

0.395

0.410

0.437

0.701

0.722

0.795

0.491

0.524

0.546

0.333

0.372

0.410

γi (1)

7.258

8.558

9.415

5.713

7.676

5.147

2.237

3.096

1.778

8.822

11.049

11.592

6.260

6.343

8.430

3.692

2.140

4.567

S/N

Ra

0.550

0.704

0.768

0.393

0.600

0.336

0.040

0.132

0.000

0.711

0.943

1.000

0.451

0.468

0.669

0.194

0.031

0.274

x i (2)

0.450

0.296

0.232

0.607

0.400

0.664

0.960

0.868

1.000

0.289

0.057

0.000

0.549

0.532

0.331

0.806

0.969

0.726

i (2)

0.526

0.628

0.683

0.452

0.556

0.430

0.343

0.366

0.333

0.634

0.898

1.000

0.476

0.484

0.602

0.383

0.340

0.408

γi (2)

0.689

0.756

0.842

0.507

0.577

0.546

0.369

0.388

0.385

0.668

0.810

0.897

0.484

0.504

0.574

0.358

0.356

0.409

Gr

5

4

2

10

7

9

16

14

15

6

3

1

12

11

8

17

18

13

Rank

Equal weights

370 M. Muaz and S. K. Choudhury

29 Simultaneous Optimization of Milling Process Responses …

371

Step 3: Normalization of S/N ratios S/N ratios are normalized such that the original values are transformed into a comparable sequence. Larger values of S/N ratios are always better. Therefore, the sequence of S/N ratios is normalized by using the following equation Eq. (29.1) which is used for larger-the-better criterion: xi (k) =

yi (k) − min(yi (k)) max(yi (k)) − min(yi (k))

(29.2)

The normalized sequences for S/N ratio of Q and Ra are shown in Table 29.4. yi (k) is S/N ratio for kth response of ith experimental run while max (yi (k)) and min (yi (k)) are the maximum and minimum values of S/N ratios, respectively, among all experimental runs for kth response. Step 4: Evaluation of grey relational coefficients (γ ) and grey relational grades (Gr) γ is calculated using Eqs. (29.3) and (29.4) while Gr is calculated using Eqs. (29.5) and (29.6) [1]. γi (k) =

min + ζ ∗ max i (k) + ζ ∗ max

(29.3)

ζ is distinguishing coefficient, ζ ∈ (0, 1] where ζ = 0.5; min = 0; max = 1 and i (k) = |xo (k)−xi (k)|

(29.4)

xi (k) is the normalized values of S/N ratios and x o (k) = 1. Gr =

n 

w(k) ∗ γ (xo (k), xi (k))

(29.5)

k=1 n 

w(k) = 1

(29.6)

k=1

Equal weights are given to both of the output responses as both are equally important for the process. Therefore, w(1) = 0.5 and w(2) = 0.5. The calculated values of γ and Gr are tabulated in Table 29.4. Step 5: Analysis of GRG Response table for GRG was generated (Table 29.5) according to Taguchi method by calculating average values of GRGs for each factor level. A larger value of GRG indicates the optimum level of a factor. Ranks shown in the table describe the order of significance of the factors. Finally, ANOVA was performed to confirm the order of significance as shown in Table 29.6. This gives optimum conditions as A1B3C1 (i.e., dry cutting environment, spindle speed of 7500 rpm, and feed rate of 500 mm/min).

372

M. Muaz and S. K. Choudhury

Table 29.5 Response table for GRG A

B

C

Level 1

0.5623

0.3775

0.6089

Level 2

0.5622

0.5320

0.5653

Level 3

0.7771

0.512

Delta

0.0001

0.3996

0.0965

Rank

3

1

2

Table 29.6 ANOVA table for GRA SS

Dof

MSS

F

Sig.

P-value

A

0.000000040

1

0.000

0.000

0.00366

0.99634

B

0.487

2

0.244

135.188

1.00000

0.00000

7.777

0.99318

0.00682

C

0.028

2

0.014

Error

0.022

12

0.002

Total

0.537

17

0.032

Rank 1 from response table for GRG (Table 29.5) and highest F-value from ANOVA for GRG (Table 29.6) confirms that the spindle speed is the most significant factor followed by feed rate and cutting condition. Very low P-values for factors B and C confirm that spindle speed and feed rate have significant effects on minimizing Ra and Q simultaneously. Although factor A has significance to 75.9% confidence interval, the dry condition should be recommended compared to the flooded condition as flooded lubrication is uneconomical as well as detrimental to the environment. Step 6: Prediction of GRG and confirmatory experiment The experimental confirmation test is the final step. As flooded lubrication is acom mon practice in the industry, initial parameters are set at A2B3C1. GRG Gr was calculated using Eq. (29.7) for initial parameter setting and for optimum levels of input parameters selected from Table 29.5. Experiments were performed under specified milling conditions for confirmation, and observed and predicted values are tabulated in Table 29.7. An improvement in GRG from initial parameter setting to 

Table 29.7 GRG for initial and optimum conditions Initial parameter setting

Optimum condition Prediction

Experiment

A1B3C1

A1B3C1

Level

A2B3C1

S/N (r)

−3.007

−3.197

9.422

11.828

S/N (Ra )   GRG Gr 

0.4242

0.8238

0.904

29 Simultaneous Optimization of Milling Process Responses …

373

the predicted one, depicted from Table 29.7, suggests that dry cutting is better than flooded lubricating condition. 

Gr = Grmean +

p   Gr j − Grmean

(29.7)

j=1

where Gr j : mean GRG at a particular level, p: number of the parameters and Grmean : total mean GRG. Hence, when overall performance of the finish milling of steel is considered, dry cutting is coming out to be the more suitable condition than that of the flooded condition.

29.4 Conclusions This study contributes to a new comprehension of machining performance of AISI 4340 steel in flat end milling with coated carbide tools across a useful region of process operability. It targets the investigation of the effectiveness of dry cutting as compared to flooded lubrication condition giving equal weights to two important process responses simultaneously. In this work, TGRA technique is modified, utilized, and got better results. An attempt has been made to know whether the use of flooded lubrication can be circumvented effectively during flat end milling of AISI 4340 steel by using a newer generation coated carbide tool at intelligently optimized machining conditions. Since there is a need for multi-response optimization to improve today’s manufacturing processes, this paper proposes some conceptual improvements in grey relational analysis-assisted Taguchi method. As a contrast to other studies, in the present work, grey relational analysis using equal weights for two responses was performed on the S/N ratios of the response values that made this technique insensitive to the variation of the noise factor. Limitation of this study is that the results are valid only for the selected tool-work combination. The conclusions drawn are summarized as follows: • Dry cutting is found out to be more suitable than flooded lubricating condition when the overall performance of the finish milling of steel is considered. • The optimum machining condition recommended from single response problem is A2B3C1 as far as only Q is concerned. However, from single response problem for Ra as well as from multiple responses GRA-assisted Taguchi analysis, optimum machining condition is same as A1B3C1. Therefore, it is recommended to perform flat end milling operation on AISI 4340 steel in dry cutting condition at high speed (7500 rpm) and low feed rate (500 mm/min). • Hybrid Taguchi-grey relational analysis reveals that spindle speed is the most significant factor affecting the output responses followed by feed rate and then machining environment. ANOVA table also confirms the order of significance.

374

M. Muaz and S. K. Choudhury

• GRA-assisted Taguchi method is an efficient statistical technique to optimize flat end milling operation considering Q and Ra simultaneously. • Flooded lubrication technique is not feasible for flat end milling of AISI 4340 steel. Dry machining should be used at optimum machining conditions to make the manufacturing process environmentally green and economically cheap.

References 1. Rao, V.R.: Advanced Modeling and Optimization of Manufacturing Processes: International Research and Development. Springer-Verlag, London Limited, London (2011) 2. Veremchuk, L.V., Tsarouhas, K., Vitkina, T.I., Mineeva, E.E., Gvozdenko, T.A., Antonyuk, M.V., Rakitskii, V.N., Sidletskaya, K.A., Tsatsakis, A.M., Golokhvast, K.S.: Impact evaluation of environmental factors on respiratory function of asthma patients living in urban territory. Environ. Pollut. 235, 489–496 (2018) 3. Choudhury, S.K., Muaz, M.: Natural oils as green lubricants in machining processes. Ref. Modul. Mater. Sci. Mater. Eng. (2018) (Elsevier) 4. Kalsi, N.S., Sehgal, R., Sharma, V.S.: Multi-objective optimization using grey relational Taguchi analysis in machining. Int. J. Organ. Collect. Intell. 6(4), 45–64 (2016) 5. Subramaniam, S.T.M., Thangarasu, S.K.: Multi-response milling process optimization using the Taguchi method coupled to grey relational analysis. Mater. Test. 58(5), 462–470 (2016) 6. Balraj, U.S., Krishna, A.G.: Multi-objective optimization of EDM process parameters using Taguchi method, principal component analysis and grey relational analysis. Int. J. Manuf. Mater. Mech. Eng. 4(2), 29–46 (2014) 7. Julong, D.: Introduction to grey system theory. J. Grey Syst. 1, 1–24 (1989) 8. Abhang, L.B., Hameedullah, M.: Determination of optimum parameters for multi-performance characteristics in turning by using grey relational analysis. Int. J. Adv. Manuf. Technol. 63, 13–24 (2012) 9. Lin, J.L., Lin, C.L.: The use of the orthogonal array with grey relational analysis to optimize the electrical discharge machining process with multiple performance characteristics. Int. J. Mach. Tools Manuf. 42(2), 237–244 (2002) 10. Lin, C.L.: Use of the Taguchi method and grey relational analysis to optimize turning operations with multiple performance characteristics. Mater. Manuf. Process. 19(2), 209–220 (2004) 11. Chiang, Y.M., Hsieh, H.H.: The use of the Taguchi method with grey relational analysis to optimize the thin-film sputtering process with multiple quality characteristic in color filter manufacturing. Comput. Ind. Eng. 56(2), 648–661 (2009) 12. Maiyar, L.M., Ramanujam, R., Venkatesan, K., Jerald, J.: Optimization of machining parameters for end milling of Inconel 718 super alloy using Taguchi based grey relational analysis. Procedia Eng. 64, 1276–1282 (2013) 13. Sarikaya, M., Yilmaz, V., Dilipak, H.: Modeling and multi-response optimization of milling characteristics based on Taguchi and gray relational analysis. Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 230(6), 1049–1065 (2016) 14. Sarikaya, M., Güllü, A.: Multi-response optimization of minimum quantity lubrication parameters using Taguchi-based grey relational analysis in turning of difficult-to-cut alloy Haynes 25. J. Clean. Prod. 91, 347–357 (2015) 15. Kitagawa, T., Kubo, A., Maekawa, K.: Temperature and wear of cutting tools in high-speed machining of Inconel 718 and Ti-6Al-6V-2Sn. Wear 202(2), 142–148 (1997)

Chapter 30

Assessment of Cutting Tool Reliability During Turning Considering Effects of Cutting Parameters and Machining Time Gaddafee Mohamad

and Satish Chinchanikar

Abstract In the era of sustainable manufacturing, assessment of cutting tool reliability during machining is extremely valuable as machining with a worn-out tool severely affects the overall machined surface quality and dimensional accuracy. On the other hand, replacing the reliable tool well before its life affects the machining economy. With this view, in the present work, empirical model to assess cutting tool reliability is developed considering the effects of cutting parameters and machining time during turning. Experiments were carried out on AISI 4340 steel (35 HRC) using multi-layer TiAlN coated carbide tool. The unknown model coefficients were obtained by minimizing the least squares error between experimental and predicted values of cutting tool reliability. The predicted values of cutting tool reliability varying with machining time found in good agreement with experimental values of reliability indicated that the developed model could be used to assess cutting tool reliability for the given tool and work material combinations and within the domain of the cutting parameters selected. It has been observed that cutting speed followed by machining time are the prominent factors by which cutting tool reliability is mostly affected. However, depth of cut has been observed to have less effect on cutting tool reliability in comparison to feed. Keywords Reliability · Turning · Tool wear · Empirical model · Cutting parameters · Machining time

Nomenclature AISI ISO HRC PVD

American Iron and Steel Institute International Organization for Standardization Rockwell hardness measured on the C scale Physical vapor deposition

G. Mohamad (B) · S. Chinchanikar Department of Mechanical Engineering, Vishwakarma Institute of Information Technology, Pune 411048, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_30

375

376

TiAlN T F nw G γ0 A Cs R VB VBcritical V F D

G. Mohamad and S. Chinchanikar

Titanium aluminum nitride Time, min Normal force on the tool flank Limit state function Rake angle of the tool Clearance angle of the tool Side cutting edge angle of the tool Reliability function Actual flank wear, mm Critical flank wear, mm Cutting speed, m/min Feed rate, mm/rev Depth of cut, mm

30.1 Introduction The rapid enhancement in the cutting tool technology enables metal cutting processes at higher production rate using higher values of cutting parameters. However, selection of cutting parameters, especially cutting speed mainly depends on the cutting tool material, tool geometry, type of machining operation and machining conditions (dry/wet machining, etc.). The assessment of cutting tool reliability during machining is highly important as machining with unreliable tool (worn-out tool) affects dimensional accuracy, surface finish and often leads to scraping of the machined component. In case of catastrophic failure of the cutting tool, the entire production line comes to halt and may severely damage the machine tools [1]. Several researchers have made significant efforts to evaluate the cutting tool reliability using experimental-based and mathematical models. A group of researchers developed predictive models to know the progress of tool wear with machining time. Chinchanikar et al. [2] evaluated flank wear rate considering abrasion, adhesion, and diffusion wear mechanisms. Their study concludes abrasion as the prominent wear mechanism while machining harder steel and adhesion while machining softer steel. Further, their study indicates that cutting tool is more reliable up to a flank wear value of 0.2 mm as catastrophic failure or chipping off of the cutting edges occurred beyond the flank wear value of 0.2 mm. Dawson and Kurfess [3] also observed abrasion and adhesion as dominant wear mechanisms. From the study of Huang and Liang [4], it can be concluded that cutting tools are reliable up to flank wear value of 0.2 mm during turning hardened steel. Several attempts have been made by the researchers to correlate cutting tool reliability (tool wear) with the cutting force. Attempts have been made to model the flank wear considering effect of cutting force on tool wear [5–7]. Attempts have been also made to predict reliability of cutting tools in terms of tool life considering the

30 Assessment of Cutting Tool Reliability During Turning …

377

effects of cutting parameters and type of cutting tool material. Dureja [8] obtained flank and crater wear of CBN tool by developing statistical models in terms of cutting parameters and work material hardness. Klim et al. [9] developed a model to predict cutting tool reliability using the experimental values of tool flank wear varying with machining time at different cutting conditions. Carlson and Strand [10] predicted the tool life using the expanded Taylor’s tool life equation. The model to predict the cutting tool reliability can be developed in the form of expanded Taylor’s tool life equation. Wang et al. [11] derived a reliability subordinate disappointment rate to evaluate the cutting tool reliability. A proportional hazards model was also employed by the researchers to achieve the cutting tool reliability. Although several efforts have been made to evaluate the cutting tool reliability by modeling the tool wear, very few of them have assessed the cutting tool reliability and its variation with machining time considering the effects of cutting parameters while turning using coated carbide tool. In the present era of green manufacturing, assessment of cutting tool reliability and its deterioration with machining time will have paramount importance as the machined surface quality and dimensional accuracy will greatly affect using an unreliable tool (severely worn tool). With this view, in this work, assessment of cutting tool reliability is carried out during turning of AISI 4340 steel at various cutting conditions. Reliability of cutting tool varying with machining time is evaluated at various cutting conditions to figure out the behavior of cutting tool reliability at different stages of tool wear. A digital microscope was used to measure the growth of flank wear after regular interval of length of machining.

30.2 Reliability Model During machining a fresh tool wears out with machining time and hence, reliability of a cutting tool; in terms of tool life with better-machined surface quality, deteriorates with time. Cutting tool reliability can be obtained by obtaining the values of limiting state function (denoted as G). The limiting state function is obtained by subtracting actual flank wear (VBactual ) from critical flank wear value (VBcritical ). Critical flank wear value is set considering the tool life criteria. At the start of the experiment, the reliability of the cutting tool (expressed in the scale of 0–1) is one as the cutting tool insert is new and flank wear is considered as zero (no wear when the insert is new). However, the cutting tool reliability decreases as tool wears out with machining time. Archard [12] showed the relationship between the wear rate and the normal force on the tool flank. The normal force on the tool flank varying with the progress in flank wear can be expressed in terms of cutting parameters and machining time [13] as: Fnw = K V a1 f a2 d a3 t a4

(30.1)

The volume of worn-out tool material (dV ), during a time interval (t) can be written as [14]:

378

G. Mohamad and S. Chinchanikar

dV =

d tan VB dVB (1 − tan γ0 tan α)cos Cs

(30.2)

where γ 0 is the rake angle of the tool and α is the clearance angle of the tool. C s is the side cutting edge angle of the tool, and d is the depth of cut. By combining Eqs. (30.1), (30.2) and after integrating and substituting the initial conditions as VB = 0 at t = 0, the simplified equation for VB can be obtained. The detailed derivation can be referred from [13]. Once the value of VB (value of actual flank wear) is obtained, the reliability of the cutting tool can be obtained by finding the limit state function value ‘G’. The cutting tool reliability can be expressed as, R=

VBcritical − VBactual G = VBcritical VBcritical

(30.3)

In general, the reliability of the cutting tool can be expressed as: R = K V a1 f a2 d a3 t a4

(30.4)

The unknown coefficients, exponents of V, f, d and t and constant K can be determined by minimizing the least squares error between experimental and predicted cutting tool reliability values at various time instants for various cutting conditions.

30.3 Experimental Details Experiments were carried out to validate and calibrate the proposed reliability model. The developed experimental-based model was used to investigate the effect of cutting parameters, namely, cutting speed, feed, and depth of cut and machining time on the cutting tool reliability. Dry turning operations were performed using PVD-applied TiAlN carbide tool. The tool was rigidly mounted on a right-hand style tool holder having an approach angle of 75°, a rake angle of −6°, side cutting edge angle of 15°, and clearance angle of 6°.

30.3.1 Experimental Procedure Central rotatable composite design (CCD) test matrix was employed to design the experiments [15]. Initially, experiments were carried out using a constant depth of cut of 0.8 mm. Some additional experiments were also carried out to investigate the response of depth of cut on the cutting tool reliability (Table 30.1). Digital microscope was used to measure the flank wear and hence to obtain the cutting tool reliability varying with machining time.

30 Assessment of Cutting Tool Reliability During Turning … Table 30.1 Experimental matrix

379

Experiment number

Cutting speed (m/min)

Feed (mm/rev)

Depth of cut (mm)

1

265

0.275

0.8

2

200

0.2

0.8

3

300

0.2

0.8

4

142

0.275

0.8

5

200

0.3

0.8

6

100

0.2

0.8

7

200

0.1

0.8

8

142

0.125

0.8

9

265

0.125

0.8

10

142

0.2

0.5

11

200

0.2

1.5

12

265

0.2

1.5

13

142

0.2

1.5

30.4 Results and Discussion The exponents of V, f, d and t and constant K, in Eq. (30.4), were obtained by minimizing the least squares error between experimental and predicted reliability values corresponding to different time instants at various cutting conditions (Table 30.1). Reliability of the cutting tool obtained varying with time at different cutting conditions was used to calibrate the model. However, while calculating reliability of the cutting tool, critical value of tool wear is considered as 0.2 mm or the occurrence of the catastrophic failure. After obtaining the unknown coefficients, the equation to predict cutting tool reliability can be expressed as: 1  cos Cs (1 − tan γ0 tan α) 2 −0.40 −0.24 −0.18 −0.38 Vc f d t R = 0.003103 tan α

(30.4)

After substituting the tool geometry, the final empirical equation to predict cutting tool reliability is expressed as: R = 4.252 V −0.4 f −0.24 d −0.18 t −0.38

(30.5)

The plot between experimental and predicted reliability values obtained using Eq. (30.5) varying with time at different cutting conditions is shown in Fig. 30.1. A correlation coefficient value obtained between experimental and predicted reliability is 0.84 which shows that the developed equation can be used to assess the cutting tool

380

G. Mohamad and S. Chinchanikar

Fig. 30.1 Experimental and predicted cutting tool reliability

reliability for the said workpiece-tool pair within the range of the cutting parameters selected in the present study in turning of AISI 4340 steel. With a view to have better understanding of effect of cutting parameters on cutting tool reliability, reliability of the cutting tool was computed (Eq. 30.5) by varying one input parameter at a time and keeping the other parameters constant. Plots showing effect of cutting speed on cutting tool reliability varying with cutting time, feed and depth of cut are plotted and shown in Fig. 30.2a, b. From the plots, it can be seen that the cutting speed and machining time dominantly affect the cutting tool reliability in comparison to feed and depth of cut which can be also confirmed from the higher values of exponents obtained for cutting speed and time from Eq. (30.5). The results obtained from the present study are in line with the available literature indicating that the cutting tool reliability is a function of cutting parameters, especially the cutting speed and machining time. Moreover, it also depends on the workpiece-tool combination and cutting environment. 3-D plots showing variation of cutting tool reliability with cutting parameters and machining time are plotted to understand the interaction effect. The surface plots of cutting tool reliability varying with cutting parameters, either speed and feed (Fig. 30.3a) or speed and depth of cut (Fig. 30.3b) or feed and depth of cut (Fig. 30.3c), confirm that the cutting speed has the prominent effect on the cutting tool reliability followed by the feed and depth of cut. Surface plots showing variation of cutting tool reliability with cutting speed and machining time (Fig. 30.4a), feed and machining time (Fig. 30.4b), and depth of cut and machining time (Fig. 30.4c) are plotted. From the above plots, it can be seen that the cutting speed and the machining time are the prominent factors affecting cutting tool reliability. Feed and depth of cut observed to be lower effect on cutting tool reliability. Further, it can be seen that cutting tool reliability decreases drastically in the first phase of machining. However, this effect can be seen as more prominent at higher cutting speed in comparison to feed and depth of cut.

30 Assessment of Cutting Tool Reliability During Turning …

(a) Machining time

381

(b) Feed

(c) Depth of cut Fig. 30.2 Effect of cutting speed on cutting tool reliability

30.5 Conclusions In the present work, empirical model to assess the cutting tool reliability (PVDTiAlN coated carbide tool) during turning of AISI 4340 steel has been developed. The unknown coefficients of the model were determined by minimizing the least squares error between experimental and predicted values of cutting tool reliability. It has been observed that the predicted cutting tool reliability values are in good agreement with the experimental values. Surface plots plotted for cutting tool reliability reveal that cutting speed followed by machining time as the prominent factors which influence the cutting tool reliability. Feed and depth of cut observed to be less prominent to affect the cutting tool reliability. Further, it has been observed that cutting tool reliability decreases drastically in first phase of machining. However, this effect can be seen as more prominent at higher cutting speed in comparison to feed and depth of cut. The good agreement between the predicted and experimental values of cutting tool reliability showed that the developed model could be used effectively to assess

382

G. Mohamad and S. Chinchanikar

Fig. 30.3 Surface plots of cutting tool reliability considering interaction effect of a cutting speed and feed b cutting speed and depth of cut c feed and depth of cut

30 Assessment of Cutting Tool Reliability During Turning …

383

Fig. 30.4 Surface plots of cutting tool reliability considering interaction effect of a cutting speed and machining time b feed and machining time c depth of cut and machining time

384

G. Mohamad and S. Chinchanikar

the cutting tool reliability for the given tool and work material combination and within the domain of the cutting parameters selected in the present study.

References 1. Chinchanikar, S., Choudhury, S.K.: Machining of hardened steel—experimental investigations, performance modeling and cooling techniques: a review. Int. J. Mach. Tools Manuf. 89, 95–109 (2015) 2. Chinchanikar, S., Choudhury, S.K.: Predictive modeling for flank wear progression of coated carbide tool in turning hardened steel under practical machining conditions. Int. J. Adv. Manuf. Technol. 76, 1185–1201 (2015) 3. Dawson, T.G., Kurfess, T.R.: Modeling the progression of flank wear on uncoated and ceramiccoated polycrystalline cubic boron nitride tools in hard turning. J. Manuf. Sci. Eng.: ASME 128(1), 104–109 (2006) 4. Huang, Y., Liang, S.Y.: Modeling of CBN tool flank wear progression in finish hard turning. ASME J. Manuf. Sci Eng. 126(1), 98–106 (2004) 5. Chinchanikar, S., Choudhury, S.K.: Cutting force modeling considering tool wear effect during turning of hardened AISI 4340 alloy steel using multi-layer TiCN/Al2O3/TiN coated carbide tools. Int. J. Adv. Manuf. Technol. 83(9), 1759–1762 (2016) 6. Chinchanikar, S., Choudhury, S.K.: Characteristics of wear, force and their inter-relationship: in-process monitoring of tool within different phases of the tool life. Procedia Mater. Sci. 5, 1424–1433 (2014) 7. Chinchanikar, S., Choudhury, S.K.: Investigations on machinability aspects of hardened AISI 4340 steel at different levels of hardness using coated carbide tools. Int. J. Refract. Met. Hard Mater. 38, 124–133 (2013) 8. Dureja, J.S.: Optimisation of tool wear during hard turning of AISI-H11 steel using TiN coated CBN-L tool. Int. J. Mach. Mach. Mater. 12(1/2), 37–53 (2012) 9. Ennajimi, Z.E., Klim, M., Balazinski, C., Fortin, C.: Cutting tool reliability analysis for variable feed milling of 17-4PH stainless steel. Wear 195, 206–213 (1996) 10. Stran, F., Carlson, T.E.: A statistical model for prediction of tool life as a basis for economical optimization of the cutting process. Ann. CIRP 41(1), 79–82 11. Wang, K.S., Lin, W.S., Hsu, F.S.: A new approach for determining the reliability of a cutting tool. Int. J. Adv. Manuf. Technol. 17, 705–709 (2001) 12. Archard, J.F.: Contact and rubbing of flat surfaces. J. Appl. Phys. 24(1), 981–988 (1953). Bhattacharya, A.: Metal Cutting Theory and Practice, New Central Book Agency, Calcutta (2000) 13. Chinchanikar, S., Choudhury, S.K. Modeling and evaluation of flank wear progression of coated carbide tools in turning hardened steels. Int. J Manuf. Technol. Manag. 27(1/2/3) (2013) 14. Bhattacharya, A., Ham, I.: Analysis of tool wear—part 1: theoretical models of flank wear. ASME J. Eng. Ind. 91(3), 790–796 (1969) 15. Cochran, W.G., Cox, G.M.: Experimental Designs. Wiley, New York (1957)

Chapter 31

An Experimental Investigation on Productivity and Product Quality During Thin-Wall Machining of Aluminum Alloy 2024-T351 G. Bolar

and S. N. Joshi

Abstract Thin-wall parts have a wide range of applications in aerospace and electronics industry due to their high strength–weight ratio. However, thin-wall parts easily deflect under the generated machining forces, thereby, resulting in loss of product quality, productivity, and form accuracy. Therefore, it is very essential to select proper process parameters. In the present work, an experimental study has been carried out on the influence of productivity parameters on milling force, surface roughness, and dimensional accuracy. It was noted that increase in material removal rate significantly influences the product quality and form accuracy. Higher material removal rate results in chatter vibration which deteriorated the surface finish. This study provides important and valuable guidelines to manufacture thin-wall parts. Keywords Thin-wall machining · Surface finish · Wall deflection · Material removal rate · Chatter · Milling force

31.1 Introduction Components having very thin cross sections are used in aircraft, automobile, mold making, and electronic industries. But machining such components is a challenging task. During the machining process, more than 90% of the material is removed by machining. As a result, the part during the final stages of the machining will have very low rigidity. Therefore, the part deflects and deforms, leading to loss of dimensional accuracy. Also, due to the low rigidity, the thin wall vibrates leading to poor surface finish. Research works on various aspects of thin-wall machining have been reported. G. Bolar (B) Department of Mechanical and Manufacturing Engineering, Manipal Institute of Technology, MAHE, Manipal 576104, India e-mail: [email protected] S. N. Joshi Department of Mechanical Engineering, Indian Institute of Technology Guwahati, Guwahati 781039, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_31

385

386

G. Bolar and S. N. Joshi

Wu et al. [1] compared the use of finite element method (FEM) and finite difference method (FDM) for predicting the thin-wall deformations. It was noted that both the models could predict the deformation values within 10% of the experimental results. The results also confirmed that FDM is a highly reliable and suitable method for measuring deformation. Gao et al. [2] worked on high-speed milling of curved thin walls made of difficult to machine titanium alloy. The influence of two tool path strategies, feed rate, spindle speed, and axial depth of cut on machining quality in terms of wall deformation was analyzed. Cutting force showed an increasing tendency with the increase in feed rate and axial cut depth while a decreasing trend was noted with increase in spindle speed. Finally, the compensation method developed was found to be effective in reducing the deformation by 53%. Wang et al. [3] used FEM and particle swarm optimization (PSO) simultaneously to optimize the fixture and cutting parameters. Bolar and Joshi [4] developed a FEM model to simulate the machining of thin-wall part and predicted the cutting forces, wall deflection, chip morphology, and cutting temperature accurately. Si-meng et al. [5] proposed a novel way to predict errors during thin-wall machining using finite element method (FEM). By taking into account the deflection of the workpiece/tool system and the springback deformation of the workpiece. Wang et al. [6] proposed an optimization algorithm to reduce the thin-wall deformation. Han et al. [7] carried out experiments to determine the deflection of thin-wall workpiece by varying the workpiece material, thickness, length, and height. It was reported that the wall deflection was significantly influenced by thickness and workpiece material type. Bolar et al. [8] investigated the machining approaches along with feed rate and number of teeth on machining forces, surface roughness, and deflection while machining C-shaped thin-wall parts. It was observed that an approach which uses both concave + convex machining produced superior thin-wall components. Qu et al. [9] studied the stability of thin-wall machining operation using 3-D stability lobe diagram. Stability lobe was developed considering the spindle speed, tool position, and axial cut depth. It was concluded that feed rate had minimal influence on chatter during thin-wall machining, but it influenced the machined surface quality. From the reported literature, it was noted that, while machining the thin-wall parts, maintaining precise dimensional accuracy and superior finish is a difficult task. It is essential to employ the proper process parameters to obtain process performance. Therefore, the present study investigates the influence of process parameters viz. feed per tooth and axial cut depth on milling force, surface finish, and product accuracy during thin-wall machining process and the implication on increasing the material removal rate (MRR).

31.2 Experimental Details The objective of the work is to analyze the implication of increasing the material removal rate (MRR) on milling force, surface finish, and dimensional accuracy of the thin-wall part. It has to be noted that, for end milling operation, the MRR can be expressed as:

31 An Experimental Investigation on Productivity and Product …

387

Fig. 31.1 a Experimental fixture, b Workpiece shape and dimensions, c End mill cutter

MRR = ad · rd · f t · z · n

(31.1)

where ad denotes the axial cut depth, r d is the radial cut depth, f t the feed per tooth, z is the number of cutting teeth, and n the spindle speed. MRR can be maximized by optimizing feed per tooth, axial and radial cut depth [10]. In the present work, experiments were planned and conducted by varying feed per tooth and axial cut depth. Experiments were performed using the setup shown in Fig. 31.1a. Aluminum alloy 2024-T351 workpiece was used to carry out the experiments. The geometry and dimensions of the workpiece are shown in Fig. 31.1b. Workpiece was machined using 12 mm carbide end mills having four flutes as shown in Fig. 31.1c. A three-axis vertical machining center was used to conduct the experiments. The generated milling forces were measured using a piezoelectric sensor-based dynamometer (Kistler 9272B) along with a charge amplifier (Kistler 5070A). The surfaces of the machined parts were examined for measuring the surface finish using a surface profilometer (Taylor Hobson Talysurf CCI lite). In-process deflection of the thin wall was measured using a linear variable differential transformer (Solartron AX/5/S). For conducting the experiments, the spindle speed was fixed at 3500 rev/min and radial depth of cut was fixed at 1.25 mm. The process parameter combinations used in the present study are listed in Table 31.1.

388 Table 31.1 Process parameter combinations

G. Bolar and S. N. Joshi Expt. No.

Feed/tooth (mm/z)

Axial depth of cut (mm)

1

0.02

12

2

0.04

12

3

0.06

12

4

0.02

24

5

0.04

24

6

0.06

24

31.3 Results and Discussion 31.3.1 Milling Force In thin-wall machining operation, generated milling forces are an important measure of performance, since it influence the surface quality and form accuracy of the machined parts. The influences of feed per tooth and axial cut depth on milling force components were analyzed. Figure 31.2a shows the variation of milling force components (F x , F y , and F z ) with feed per tooth (f z ) for constant depth of cut values. It was noted that milling force components increase with feed values. With the increase in feed value, amount of material in contact with the tool increases thereby increasing the chip load per tooth. It has to be noted that, the force component F x (cutting force) is higher that feed force component (F y ). Also, the axial force component (F z ) is relatively smaller in comparison to the other two force components. It can also be noted that the depth of cut in axial direction influences the three components of milling force. For a constant feed per tooth value, milling force components increase significantly with axial depth of cut. This is due to the increase in the interaction of tool and work material. It has to be noted that the force components increase almost linearly with axial depth of cut. The built-up-edge (BUE) formation was observed for cutting conditions viz. axial cut depth of 24 mm and feed per tooth of 0.06 mm/z (Fig. 31.2b). The formation of BUE is attributed to the increase in uncut chip thickness thereby increasing the cutting force value. Overall, it can be concluded that higher values of feed and axial depth enhance the MRR with the increase in milling force values.

31.3.2 Surface Roughness Figure 31.3 shows the surface profiles of the machined thin-wall parts obtained using optical profilometer. Cutting feed value was found to influence the surface roughness. As the feed increases, the roughness of the machined surface increases. This can be attributed to the increase in frictional contact between the work and tool. During the experiments, it was noted that the generated chips deposit between

31 An Experimental Investigation on Productivity and Product …

389

Fig. 31.2 a Influence of feed per tooth and axial depth of cut on milling force components, b Builtup-edge formation

the workpiece and the tool. This resulted in increase in the surface roughness. Figure 31.4a–f depicts the 3-D surface profile measured for the six experiments. It can be noted that the surface roughness increases with the feed values. Similarly, the surface roughness varies with the axial depth of cut. The roughness of the surface was found to be increased with the axial cut depth. For fixed feed values of 0.02, 0.04, and 0.06 mm/z, the surface roughness increased with the increase in axial cut depth. In addition, deterioration of machine surface due to chatter was also observed. Formation of mild chatter was observed at low feed (0.02 mm/z) and cut depth of 12 mm. This can be seen in Fig. 31.4a. But the severity of chatter increased with depth of cut and feed values (see Fig. 31.4b–d). At still higher levels of feed and axial depth, the formation of severe chatter marks was observed (see Fig. 31.4e, f).

390

G. Bolar and S. N. Joshi

Fig. 31.3 Influence of feed per tooth and axial cut depth on surface roughness

31.3.3 Dimensional Accuracy Thin-wall parts have very low rigidity during the final stages of machining. Due to this, the wall deflects under the action of milling forces thereby affecting the dimensional accuracy of the final finished part. The magnitude of deflections of the wall during the machining process at five different points along the free end of the workpiece is shown in Fig. 31.5. The deflection was lower when feed of 0.02 mm/z and axial depth of 12 mm were considered. The magnitude of deflection gradually increased with feed and cut depth. Maximum deflection was noted when a feed of 0.06 mm/z and axial depth of 24 mm were used. Also, the influence of axial cut on wall deflection was higher than feed per tooth. This increase in magnitude of wall deflection can be attributed to increase in chip load and tool-work contact length. This resulted in higher milling forces, which in turn leads to higher wall deflection. Another notable observation was the deflection variation along the workpiece length and height directions. Deflections were higher at unsupported free ends as compared to the wall center. This can be attributed to the lower rigidity of the two free ends as compared to the wall center which surrounded by bulk material. Also, the magnitude of wall deflection at top end was noted to be higher than at the base. This resulted in thin walls with thicker tops and thinner base.

31.4 Conclusions The paper presents an experimental investigation into the influence of material removal rate (MRR) on performance measures viz. milling forces, surface finish, and dimensional accuracy in thin-wall machining process. The increase in MRR was

31 An Experimental Investigation on Productivity and Product …

391

Fig. 31.4 3-D surface profile obtained for various process parameter combinations

incorporated by increasing the feed per tooth and axial cut depth while maintaining a constant radial depth of cut and spindle speed. The selected process parameters greatly influenced the studied performance measures viz. milling force, surface roughness, and wall deflection. Increase in the magnitude of milling force components was noted for increase in both depths of cut and feed per tooth. But, axial cut depth was found to prominently influence the milling forces. Increase in axial cut depth caused the force values to rise significantly due to the increase in the chip load and length of contact between the workpiece–tool. Also the formation of cut built-up-edges was noted which contributed to the rise in force values at higher level of feed and axial cut depth. Increase in the milling force values also influenced the wall dimensional accuracy. At higher axial cut depth and feed value, magnitude of wall deflection was found to be higher. Variations in wall deflection along the workpiece length and height

392

G. Bolar and S. N. Joshi

Fig. 31.5 Variation of wall deflection along the length of workpiece for various process parameter combinations

directions were noted. This resulted in the final thin-wall part with thicker top and thinner base. From the present study, it can be inferred that axial depth of cut and feed per tooth influence the performance measures such as milling force, surface finish, and wall deflection during thin-wall machining process. Employing the higher levels of process parameters in order to improve the material removal rate significantly affects part quality in terms of dimensional accuracy and surface roughness. Therefore, proper selection process parameters are essential to achieve the desired process performance during the thin-wall machining process.

References 1. Wu, Q., Li, D.P., Ren, L., Mo, S.: Detecting milling deformation in 7075 aluminum alloy thinwalled plates using finite difference method. Int. J. Adv. Manuf. Technol. 85(5–8), 1291–1302 (2016). https://doi.org/10.1007/s00170-015-8012-3 2. Gao, Y.Y., Ma, J.W., Jia, Z.Y., Wang, F.J., Si, L.K., Song, D.N.: Tool path planning and machining deformation compensation in high-speed milling for difficult-to-machine material thinwalled parts with curved surface. Int. J. Adv. Manuf. Technol. 84(9–12), 1757–1767 (2016). https://doi.org/10.1007/s00170-015-7825-4 3. Wang, S., Jia, Z., Lu, X., Zhang, H., Zhang, C., Liang, S.Y.: Simultaneous optimization of fixture and cutting parameters of thin-walled workpieces based on particle swarm optimization algorithm. Simulation 94(1), 67–76 (2017). https://doi.org/10.1177/0037549717713850 4. Bolar, G., Joshi, S.N.: Three-dimensional numerical modeling, simulation and experimental validation of milling of a thin-wall component. Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 231(5), 792–804 (2017). https://doi.org/10.1177/0954405416685387 5. Si-meng, L., Xiao-dong, S., Xiao-bo, G., Dou, W.: Simulation of the deformation caused by the machining cutting force on thin-walled deep cavity parts. Int. J. Adv. Manuf. Technol. 92(9–12), 3503–3517 (2017). https://doi.org/10.1007/s0017-0383-1

31 An Experimental Investigation on Productivity and Product …

393

6. Wang, J., Ibaraki, S., Matsubara, A.: A cutting sequence optimization algorithm to reduce the workpiece deformation in thin-wall machining. Precis. Eng. 50, 506–514 (2017). https://doi. org/10.1016/j.precisioneng.2017.07.006 7. Han, B., Ren, C.Z., Yang, X.Y., Chen, G.: Experiment study on deflection of aluminum alloy thin-wall workpiece in milling process. Mater. Sci. Forum 697–698, 129–132 (2012). https:// doi.org/10.4028/www.scientific.net/MSF.697-698.129 8. Bolar, G., Mekonen, M., Das, A., Joshi, S.N.: Experimental investigation on surface quality and dimensional accuracy during curvilinear thin-wall machining. Mater Today Proc. 5(2), 6461–6469 (2018). https://doi.org/10.1016/j.matpr.2017.12.259 9. Qu, S., Zhao, J., Wang, T.: Three-dimensional stability prediction and chatter analysis in milling of thin-walled plate. Int. J. Adv. Manuf. Technol. 86(5–8), 2291–2300 (2016). https://doi.org/ 10.1007/s00170-016-8357-2 10. Budak, E., Tekeli, A.: Maximizing chatter free material removal rate in milling through optimal selection of axial and radial depth of cut pairs. CIRP Ann. Manuf. Technol. 54(1), 353–356 (2005). https://doi.org/10.1016/S0007-8506(07),60121-8

Chapter 32

An Approach of Minimizing Energy Consumption in the Machining System Using Job Sequences Varying Technique Md. Shahzar Jawaid , S. C. Srivastava

and S. Datta

Abstract Globally, industry plays a real part in the energy consumption and energy is one of the most parameters which acts continuously over the process and increases the product price. Therefore, the main aim of the manufacturer is to reduce the manufacturing cost as the manufacturing cost also gets incurred in the product final price. A production system of five different parts is considered. Each part has different operations. Each operation consumes a certain amount of energy. The total energy consumption is calculated using the mathematical model and by considering the only direct source of energy. Further, three different factors are analysed using the rule-based system. Keywords Total energy consumption · Job sequence · The machining system

32.1 Introduction Manufacturers nowadays are facing an intensive challenge by fulfilling the customer needs and demand and also take care of the environmental impact. In the manufacturing sectors, time cost and qualities are the main objectives. In manufacturing industries designer and manufacturer both play a vital role in the energy efficiency designer need to consider the stages of manufacturing stage in the eco-design effort to make it energy efficient this was observed by Bonvoisin et al. [1]. The minimization of energy consumption is a very important part for the production facilities as it directly relates to the production cost, quality as well as environmental impact. Thus, it has attracted many researchers globally in recent years. In the recent years, many authors have given the different approaches of minimizing the total energy consumption just like Zhang et al. [2] propose the method for minimizing the energy consumption by integrating process planning and scheduling. Cao et al. [3] reviewed the existing technologies that is directly involved in improving the energy efficiency in manufacturing plant Deming Lei et al. [4] gave us research of minimization of Md. S. Jawaid (B) · S. C. Srivastava · S. Datta Department of Production Engineering, Birla Institute of Technology, Mesra, Ranchi, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_32

395

396

Md. S. Jawaid et al.

energy consumption and workload balance of flexible job shop scheduling; later on, shuffled frog leaping algorithm was proposed. Giles Mouzon et al. [5] develop an operational method for the minimization of energy consumption of manufacturing equipment. It was observed that they can be a significant amount of energy savings when underutilized machine lines/equipment was turned off when they get idle for a certain amount of time. Energy consumption can be direct as well as indirect which was noted by Seow et al. [6], and the total embodied product energy is a sum of direct and indirect energy for manufacturing industries. Gomes et al. [7] proposed a standard-based infrastructure to collect and monitor energy data for the manufacturing industries in real time with the concept of manufacturing energy management system. Anton et al. [8] gave the generic method that decides a model which was developed based on a statistical discrete event, simulation formation; this model was used by introducing and applications in real-time tactical and strategic decision-making process. Wang et al. [3] have considered energy consumption, cost and quality as their objective function for optimization. Pusavec et al. [9] have observed minimizing energy consumption was viewed as an important strategy for improving the sustainability of any manufacturing shop floor.

32.2 Problem Statement For establishing the corresponding mathematical model, the following assumption is made • • • • • • • •

Each machine can process at most one operation at a time, No jobs may be processed on more than one machine at a time, Operations cannot be interrupted, Set-up times and remove times are treated as auxiliary time and is included in processing times, Jobs are independent and have equal priority, After a job is processed on a machine, it is transported to the next machine immediately and the mean transportation time is fixed, All jobs and machines are available at the time zero, turning off the idle machines and machine failure is not considered, All jobs are loaded and processed continuously according to the predetermined sequence in the process plan (Table 32.1).

32.3 Procedure and Methodology The case study paper is divided into different sections: 1. Computation of total energy consumption without using rule-based system.

32 An Approach of Minimizing Energy Consumption …

397

Table 32.1 Nomenclature defined S. No.

Notation

Description

1

Pt

The processing time of yth operation of job on machine mz

2

Oxyz

yth operation on the job that can be processed on machine mz

3

Pst

Starting power of the respective servers/workstations

4

pt

Processing time

5

at

Auxiliary time

6

mt

Machining time

7

am

Actual machining

8

ft

Feeding time

9

d

The diameter of the job

10

D

The diameter of the cutter

11

l

Length

12

fl

Longitudinal feed (mm/rev)

13

fm

Motion axis feed(mm/min)

14

dc

Depth of cut

15

Ns

Spindle speed

16

A

Length of approach

17

O

Length of overrun

18

ρ

Half of cutting point angle of the tool

19

aw

Width of cut

2. Computation of total energy consumption by (a) Changing job sequence (b) Making machine order fixed. 3. Selection of best parent chromosome by using roulette wheel selection operator. 4. Computation of total energy consumption of the parent chromosomes by (a) Making job sequence fixed (b) Changing machine order. 5. Comparison of the result and finding the best-optimized result. Determine the value of total energy consumption E tot can be represented as → E tot =



Ez =

 (EMz + EIz )

(32.1)

EMz It is the energy consumption by machine MZ during processing/machining. EIz It is the energy consumption by the machine during idle time.

398

Md. S. Jawaid et al.



EMz =

x,y,z,Ox yz ∈m z



1 E cal × (b1 + b2 − 1) + |b1 + b2 − 1| 2

 (32.2)

E cal It is calculated energy which can be defined as the sum of primary energy, secondary energy and actual machining energy

E cal = Primary energy + Secondary energy + Actual machining energy

(32.3)

Primary energy (PE) = [Product of starting power and the processing time of each workstation] Secondary energy = [Sum of energy calculated by multiplying the power of each workstation by its corresponding execution time] Actual machining energy (AME) = [Product of material removal time and the power of each respective servers]

E cal = PE + SE + AME EIz =

 x,y,z,Ox yz ∈m z

(32.4)

1 (avg Pst ) × (total idleness) × (b1 + b2 − 1) + |b1 + b2 − 1|) 2 (32.5)

Thus, total energy consumption (E tot ) is the sum of EMz and EIz The total embodied energy is E tot , and the objective is to reduce that. To establish the corresponding mathematical model, the consideration of n jobs and m machines is taken. • There are five different parts taken to be manufactured. • Each part has a different number of operations. Each operation takes different processing (pt). Processing time differs according to the machine features and machining parameters. Figure 32.1 shows the layout of the machining system, the line diagram showing the different machine and part flow.

32.3.1 Determine the Value of Total Energy Consumption EMz It is the energy consumption by machine MZ during processing/machining. EIz It is the energy consumption by the machine during idle time.

32 An Approach of Minimizing Energy Consumption …

399

Fig. 32.1 Schematic layout of a virtual machining system as studied in the problem

Calculating the objective function using the following mathematical expression: For calculation of powers during turning operations Psr (spindle rotation power) = 1.029 Ns + 19.37 Vc (cutting speed) =

π d Ns 1000

(32.6)

P xf (x − axis feeding power) = 4 × 10−6 f m2 + 0.0211 f m

(32.7)

P z f (z − axis feeding power) = 4 × 10−8 f m2 + 0.0314 f m

(32.8)

Pc = 40.64 × Vc0.931 × f l0.662 × dc0.941

(32.9)

For calculation of powers during milling operations Psr (spindle rotation power) = 0.139 Ns + 138.22 Vc (cutting speed) =

π d Ns 1000

(32.10) (32.11)

P xf (x − axis feeding power) = −7 × 10−6 f m2 + 0.0602 f m

(32.12)

P yf (y − axis feeding power) = −2 × 10−6 f m2 + 0.0315 f m

(32.13)

P z f (z − axis feeding power) = 3 × 10−6 f m2 + 0.0371 f m

(32.14)

Pc (cutter power) = 3.353 × vc0.927 × f t0.764 × dc0.927 × aw0.942

(32.15)

For calculation of powers during milling operations Psr (spindle rotation power) = 0.086Ns + 14.76

(32.16)

400

Md. S. Jawaid et al.

Vc (cutting speed) =

π d Ns 1000

(32.17)

P z f (z − axis feeding power) = −10−7 f m2 + 0.0461 f m

(32.18)

Pc (cutter power) = 4.044 × vc0.958 × f m0.798 × dc0.923

(32.19)

For calculation of machining time during turning operations (In lathe) machining time(s) =

l+ A+O × 60 fm

(32.20)

For calculation of machining time during drilling operations (In lathe) machining time(s) =

l+ A+O × 60 fm

(32.21)

For calculation of machining time during milling operations (In milling machine) machining time =

l+ A+O+ fm

D 2

(32.22)

For calculation of machining time during drilling operations (In drilling machine) machining time =

l + A+ O +C × 60 fm

C = D cot ρ

(32.23) (32.24)

Feeding time(ft) = 0.4 × Processing time of each operation Auxiliary time(at) = 0.259 × Pt

(32.25) (32.26)

Total energy consumption can be computed by (Fig. 32.2). Fig. 32.2 Calculated energy is the sum of PE, SE and AME

Ecal

Primary energy(PE)

Secondary Energy(SE)

Actual machining energy(AME)

32 An Approach of Minimizing Energy Consumption …

401

Primary Energy(PE) = Pst × total processing time Secondaryenergy(SE) =

(32.27)



(Psr × processingtimesofmachine)  + (32.28) (Pf × feedingtime)

Actualmachiningenergy(AME) = [Productofmaterialremovaltimeand  thepowerofeachrespectiveservers (32.29) Actual machining energy(AME) =

 (Machining time × Pc )

(32.30)

E cal = Primary energy + Secondary energy + Actual machining energy (32.31)

32.3.2 Computation of TEC of Corresponding Parts Computation of TEC of Corresponding Parts

Job 1

Job 2

Operation 1

Operation 2

N s = 325 rpm

N s = 325 rpm

d = 25 mm

d = 20 mm

d c = 1 mm

d c = 1 mm

f m = 22.75 mm/min

f m = 22.75 mm/min

f l = 0.07 mm/rev

f l = 0.07 mm/rev

Operation 1

Operation 2

N s = 112 rpm

N s = 112 rpm

D = 75 mm

D = 75 mm

d c = 0.5 mm

d c = 5 mm

f m = 67.2 mm/min

f m = 67.2 mm/min

f t = 0.1 mm/tooth

f t = 0.1 mm/tooth

aw = 50 mm

aw = 50 mm

402

Job 3

Md. S. Jawaid et al.

Operation 1

Operation 2

Operation 3

N s = 325 rpm

N s = 325 rpm

N s = 325 rpm

d = 20.5 mm

d = 18.5 mm

d = 16.5 mm

d c = 0.01 mm

d c = 1 mm

d c = 1 mm

f m = 22.75 mm/min

f m = 22.75 mm/min

f m = 22.75 mm/min

f l = 0.07 mm/rev

f l = 0.07 mm/rev

f l = 0.07 mm/rev

Job 4

Job 5

Operation 1

Operation 2

Operation 3

N s = 325 rpm

N s = 325 rpm

N s = 325 rpm

d = 20.5 mm

d = 18.5 mm

d = 16.5 mm

d c = 0.1 mm

d c = 1 mm

d c = 1 mm

f m = 22.75 mm/min

f m = 22.75 mm/min

f m = 22.75 mm/min

f l = 0.07 mm/rev

f l = 0.07 mm/rev

f l = 0.07 mm/rev

Operation 1

Operation 2

Operation 3

N s = 93 rpm

N s = 72.15 rpm

N s = 509.55 rpm

D = 75 mm

D = 75 mm

D = 10 mm

d c = 0.5 mm

d c = 0.5 mm

d c = 0.5 mm

f m = 60 mm/min

f m = 50 mm/min

f m = 91.72 mm/min

f t = 0.1 mm/tooth

f t = 0.1 mm/tooth

f l = 0.18 mm/tooth

aw = 50 mm

aw = 20 mm

See Table 32.2. Table 32.2 Time information for corresponding parts Job

Processing time (s)

Actual machining time (s)

Feeding time (s)

O/p1

O/p2

O/p3

O/p1

O/p2

O/p3 0

1

213.6

47.5

47.5

0

31.6

31.6

2

307.3

79.0

57.6

0

52.7

38.4

0

3

516.2

94.9

82.3

52.2

63.3

54.8

34.8

4

532.9

94.9

82.28

59.7

63.3

54.8

39.8

5

499.7

88.5

27.4

59.0

70.8

18.2

106.2

32 An Approach of Minimizing Energy Consumption … Table 32.3 Summary of the processing time of corresponding jobs in respective machines

403

Job

M/C 1 (Lathe) (s)

M/C 2 (Milling) (s)

M/C 3(Drilling) (s)

1

213.6

0

0

2

0

307.3

0

3

516.2

0

0

4

532.9

0

0

5

0

324.5

45.6

32.3.3 Computation of TEC Using a Rule-Based System  Total energy consumption for corresponding five parts is (E tot ) = 1.8545 × 107 J, and this can be optimized using the rule-based system. For application of rule base system, the problem can be reduced to – (Table 32.3). Total energy consumption is computed by using the following mathematical expression EMz = (TET) × (avg Pst ) EIz =

 (EIz )

(32.32) (32.33)

EI1 = (IT1) × (Pst 1)

(32.34)

EI2 = (IT2) × (Pst 2)

(32.35)

EI3 = (IT3) × (Pst 3)

(32.36)

EIz = EI1 + EI2 + EI3

(32.37)

E total = EMz + EIz

(32.38)

Now, taking these shortlisted job sequences, the machine orders can be changed to further optimize the total energy consumption (E total ). Machine 3 should always be put after machine 2. Thus, the possible machine order could be 1-2-3, 2-3-1, 21-3. Now, by keeping the job sequences fixed for the shortlisted chromosomes, the machine order could be changed for getting the best result.

404

Md. S. Jawaid et al.

32.4 Results and Discussion The detailed analysis of the results is shown over here. The observations made from the results are represented on a graph, and the findings are highlighted. After changing the machine order by keeping the job sequence fixed, the total energy consumption can be further optimized. The main objective of this research is to minimize total energy consumption. Thus, from the above table, we got the best-optimized result. The best-optimized result is given in Table 32.4 and Fig. 32.3.

32.5 Conclusion This research addresses only the direct source of energy. The energy-saving method was incorporated. Firstly, energy consumption was computed for each job and the energy consumption for the whole system is the sum of energy consumption of the corresponding jobs. Firstly, energy consumption was computed for each job and the energy consumption for the whole system is the sum of energy consumption of the corresponding jobs to optimize this result. The total energy consumption was completed by varying the job sequences and keeping the machine order fixed. For further optimization, the objective function was calculated by varying machine order and by keeping the job sequence is fixed. Finally, at the end of the thesis comparison result was shown. To optimize this, result the total energy consumption Table 32.4 Optimized total energy consumption Case no

Job sequence

Machine order

Total elapsed time (×103 s)

E total (×107 J)

1

52143

1-2-3

1.2631

1.5798

2

14523

2-3-1

1.2628

1.5792

3

42153

2-1-3

1.2795

1.6047

4

13524

2-1-3

1.3565

1.8317

5

32154

2-1-3

1.2783

1.6212

6

54123

1-2-3

1.2631

1.5798

7

25431

1-2-3

1.2631

1.5798

8

12534

1-2-3

1.2631

1.5798

9

15243

1-2-3

1.2631

1.5798

10

51324

1-2-3

1.2631

1.5798

Fig. 32.3 Graphical representation of the optimized result

The optimized result 2 1.5 1

1

2

3

4

5

6

7

8

9

10

32 An Approach of Minimizing Energy Consumption …

405

was completed by varying the job sequences and keeping the machine order fixed and vice versa. The future scope of this might include the use of different hybrid AI techniques for minimizing energy consumption in the machining system. Address the uncertainty events, e.g. machine breakdown, the arrival of new jobs and cancellation of some jobs.

References 1. Cronrath, C., Lennartson, B., Lemessi, M.: Energy reduction in paint shops through energysensitive on-off control. In: IEEE International Conference on Automation Science and Engineering (CASE), pp. 1282–1288 (2016) 2. Zhang, Z., Tang, R., Peng, T., Tao, L., Lia, S.: A method for minimizing the energy consumption of machining system: integration of process planning and scheduling. J. Cleaner Prod. 1–33 (2016) 3. Le, C.V., Pang, C.K., Gan, O.P.: Energy Saving and Monitoring Technologies in Manufacturing Systems with Industrial Case Studies, IEEE, pp. 1916–1921 (2011) 4. Lei, D., Zheng, Y., Guo, X.: A shuffled frog-leaping algorithm for flexible job shop scheduling with the consideration of energy consumption. Int. J. Prod. Res. 3126–3140 (2017) 5. Mouzon, G., Yildirim, M.B., Twomey, J.: Operational methods for minimization of energy consumption of manufacturing equipment. Int. J. Prod. Res. 4247–4271 (2007) 6. Seow, Y., Rahimifard, S., Woolley, E.: Simulation of energy consumption in the manufacture of a product. Int. J. Comput. Integr. Manuf. 663–680 (2013) 7. Delgado-Gomes, V., Oliveira-Lima, J.A., Martins, J.F.: Energy consumption awareness in manufacturing and production systems. Int. J. Comput. Integr. Manuf. 84–95 (2015) 8. Dietmair, A., Verl, A.: A generic energy consumption model for decision making and energy efficiency optimization in manufacturing. Int. J. Sustain. Manuf. 123–133 (2009) 9. Pusavec, F., Kramar, D., Krajnik, P., Kopac, J.: Transitioning to sustainable production–part II: evaluation of sustainable machining technologies. J. Cleaner Prod. 1211–1221 (2010)

Chapter 33

Comparative Study on the Performance of Different Drill Bits for Drilling CFRP N. S. Sowjanya , V. G. Ladeesh , R. Manu

and Jose Mathew

Abstract In this work, experimental study was carried out to compare the machining performances of the three types of drills: (1) solid carbide drill, (2) diamond core drill, and (3) brad point bit for drilling CFRP. Comparative study of the performance of the tools was carried out in terms of delamination at hole entry and exit side. Experiment was conducted based on Taguchi’s orthogonal array technique. From the analysis of results, it is found that the performance of diamond core drill was better and resulted in lesser damages of holes. The lowest value for the delamination factor at entrance (1.093) and exit (1.101) was obtained at a spindle speed of 500 rpm, feed rate of 3 mm/min with the thrust force of 78 N. The use of diamond core drill is recommended for drilling CFRP in terms of economy and drilled hole quality. Keywords CFRP · Drilling · Delamination factor · Fiber pull-out · Cutting force

33.1 Introduction As the demand for CFRPs in aerospace industry is increased extensively, the research in the field of machining of CFRPs has been carried to a great extent. Carbon fiberreinforced polymers have the property of heterogeneity and anisotropy due to which the material is difficult to cut. It has got tremendous importance in aerospace industry because of its high strength to weight ratio. CFRP possesses some major important properties like its lightweight, high specific stiffness, and corrosion resistance which has made it to become the widely used material in the aerospace industry. Even though CFRPs can be produced to near net shape, it requires additional processes like drilling, milling, and grinding to get the necessary dimensional accuracy and required fit in assembly of structural components. Abrasive nature of the reinforcement makes the machining process complex and leads to many machining defects. Among the machining processes on CFRP, drilling is the primary machining process N. S. Sowjanya (B) · V. G. Ladeesh · R. Manu · J. Mathew Department of Mechanical Engineering, National Institute of Technology Calicut, Calicut, Kerala 673601, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_33

407

408

N. S. Sowjanya et al.

to produce the required hole feature for assembling components. The performance of the drilling is assessed by the hole quality obtained and the machining defects produced during the process. Delamination and fiber pull-out negatively affect the hole quality. Delamination is the interlaminar crack produced in the CFRP plies and has a most negative impact on the quality of the hole produced Ruslan et al. [1]. Cutting conditions and proper tool selection is the important factors to be considered for reducing machining defects and improving surface quality. Cutting force influences the selection of proper tool and parameters to be used during the machining process Gurumukh and Padam [2]. Tsao [3] carried out an experimental study on delamination and thrust force analysis during drilling of CFRP using Taguchi method. In his study, it shows that thrust force is influenced by feed rate. With the smaller feed rate, it is possible to produce lower thrust force which can reduce the delamination defects in CFRP. In the comparative study of the diamond-coated tools Pinho [4], the results show that the feed rate is the important factor to be considered in analyzing delamination. Delamination increases with thrust force. Tsao [5] conducted an experimental study of drilling of CFRP with different types of step core drills. In this study, it is found that the diameter ratio and feed rate are the two important factors influencing the thrust force. Paolo [6] presented the study of drilling of CFRP using diamond core drill with four kinds of grit sizes and reported that the feed rate has a significant impact on thrust force and delamination. Coarsest drill resulted in better performance. In the analysis of surface roughness in drilling CFRP using diamond-coated ball nose drill presented by Konneh [7], it is concluded that the spindle speed is the significant factor in drilling CFRP. Comparison of convention and orbital drilling using diamond-coated tools in drilling of CFRP was carried out by Robert [8] which focused on damages in workpiece during machining, variation in bore diameter, and tool wear. He concluded that less bore exit damages are obtained with orbital drilling. CFRP is a polymer matrix composite. The interfacial bond between the matrix and fiber is very important in the case of fiber-reinforced composites. Poor interfacial bond results in fiber pull-out, debonding of fibers from matrix during crack propagation and it results in low strength and stiffness. Strong bond prevents crack propagation and retains its material property even if the fiber breaks. Mechanical properties of the composite can be improved by the proper selection and orientation of fiber reinforcement in the matrix phase Robert Voss [9] and Eshetu [10]. Lower delamination damages can be obtained with lower thrust force. Koenig [11] showed the delamination is influenced by thrust force. Many researches have shown the comparison of either twist drills or other types of same kind of drills with different geometry. In the present study, comparison of three different types of drills is carried out which helps in proper selection of tools for drilling CFRPs. In the present study, the effect of different process parameters on the performance of three types of tools is conducted using systematic design of experiments. Solid carbide, diamond core drill, and conventional brad point drill were used in the experiment. Taguchi’s orthogonal array technique was used for designing the experiment for the comparison of three tools, and data were analyzed by using MINITAB 17 software.

33 Comparative Study on the Performance of Different Drill Bits …

409

33.2 Fabrication of CFRP CFRP lamina was prepared by simple hand lay-up process. The carbon fiberreinforced lamina was prepared using 13 layers of unidirectional carbon fiber fabric sheets, epoxy resin, and hardener. Initially, primer (HY 917) and epoxy resin (Araldite® LY 556) were mixed together in container with a specific ratio. Releasing agent was applied over the mold and carbon fiber sheets (200 gsm) were laid at 0°/90° in the mold. Epoxy and primer mixture was applied on each layer of carbon fiber sheets using roller. Using the compression molding machine, composite material was pressed to 50 bar at 120 °C to obtain CFRP of 5 mm thick. Composite was post-cured at 130 °C for 6 h in the oven. Brief details of the workpiece prepared by hand lay-up method are given in Table 33.1.

33.3 Experimental Setup and Methodology Experimental setup is shown in Fig. 33.1. Drilling operations were performed using vertical machining center (Make BFW, Model: BMV45 TC24 4-axis vertical machining center). Cutting force was measured with Kistler multicomponent 9257B dynamometer. Acquisition and evaluation of data were done by the easy to use universal Kistler Dynoware software. Three drilling tools were used for the comparative study. Solid carbide twist drill (Walter titex alpha® 2 of point angle 140°), brad point drill (Dewalt), and diamond core drill (Excel impex, India) were used as shown in Fig. 33.2. The tools specifications are given in Table 33.2. Taguchi’s L9 orthogonal array design was used for conducting experiments. The factors selected are spindle speed and feed rate which developed a design with nine experimental runs. The selected factors and their levels are shown in Table 33.3. Table 33.1 Brief details of the workpiece

Workpiece material

Carbon fiber-reinforced polymer

Mixing ratio

10:1

Fiber volume

45%

Matrix volume

55%

Thickness

5 mm

Width

100 mm × 170 mm

410

N. S. Sowjanya et al.

Fig. 33.1 Experimental setup

33.4 Results and Discussion The damages at the entrance and exit are common and critical while machining fiberreinforced polymers. A measuring index called delamination factor is used for measuring the delamination of the drilled hole. Delamination factor is the ratio of the maximum diameter of the delaminated region to the nominal diameter of the hole [12, 13]. Delamination is measured with the help of a vision measuring system (VMS) (Make: Insize, Model: ISD A100). For analyzing the effects of process parameters on the delamination, drilled surfaces were analyzed using micrographs obtained from VMS. Machining defects like delamination and fiber pull-out were analyzed using the images of machined holes taken. The image of the drilled hole is shown in Fig. 33.3. Delamination factor is the most common index used for assessing the delamination of drilled hole and is defined as the ratio of maximum diameter of delaminated zone (Dmax ) to the nominal diameter (Dn ) of the hole produced. Delamination factor(DF) = Dmax /Dn

(33.1)

Results of the delamination factor examined are given in Table 33.4. Damages are less at hole entry compared to hole exit in all the three types of tool used. Delamination is severe in the case of holes drilled with brad point bit. From the result, it is found that the feed rate has a maximum influence on delamination. As feed rate is increased, damages at the hole periphery are also increased. For low feed rates, low-speed delamination is found to be less. For larger feed rate, delamination is severe in case of brad point bit. In comparing the three drills, delamination with diamond core drill was found to be less even at the high feed rate.

33 Comparative Study on the Performance of Different Drill Bits …

411

Fig. 33.2 Drill bits used for comparative study

Table 33.2 Characteristic of tools used for the drilling experiment

Tool specification

Solid carbide drill

Diamond core drill

Brad point bit

Diameter

6 mm

6 mm

6 mm

Overall length

66 mm

65 mm

90 mm

Flute length

28 mm

Number of flutes

2

Grit: 80/100 US mesh

2

Shank diameter

6 mm

4 mm

6 mm

65 mm

412

N. S. Sowjanya et al.

Table 33.3 Factors and levels of process parameter

Levels

(A)

(B)

Speed (rpm)

Feed rate (mm/min)

1

500

3

2

1500

6

3

2500

9

Fig. 33.3 Assessment of delamination

Table 33.4 Results of the comparative experiment for delamination factor Parameters

Diamond core

Brad point

Speed (rpm)

Feed (mm/min)

Solid carbide DFentrance

DF-exit

DFentrance

DF-exit

DFentrance

DF-exit

500

3

1.121

1.102

1.093

1.101

1.123

1.152

500

6

1.162

1.123

1.117

1.131

1.154

1.203

500

9

1.101

1.144

1.193

1.182

1.205

1.234

1500

3

1.193

1.113

1.104

1.151

1.174

1.183

1500

6

1.182

1.133

1.104

1.172

1.195

1.193

1500

9

1.202

1.164

1.160

1.192

1.216

1.214

2500

3

1.151

1.154

1.117

1.151

1.174

1.173

2500

6

1.184

1.133

1.111

1.162

1.185

1.203

2500

9

1.172

1.174

1.103

1.222

1.247

1.254

33 Comparative Study on the Performance of Different Drill Bits …

413

33.4.1 Drilling Tests with Solid Carbide Drill From the main effect plot of Fig. 33.4a, it can be noticed that there exist some significant main effects with respect to speed. Effect of feed is not much significant compared to speed. Values for DF-entrance are increasing up to moderate speed and feed but with higher speed and feed rate, value of DF-entrance started decreasing. Delamination at entrance occurs probably because when the lip of the tool approaches the workpiece with higher speed, it tends to peel up the top layer easily. In the interaction plot of Fig. 33.4b, it can be seen that the line segments connecting the levels of feed and speed are diverging and there exist a significant interaction between the speed and feed. The results of DF at hole exit are quite different from the hole entry because the feed rate has more influence on damages at hole exit compared to hole entry. Delamination at hole exit is influenced by feed rate. Severe hole damages were observed in the hole number 9 drilled with speed of 2500 rpm and feed rate of 9 mm/min. Main effect plots given in Fig. 33.5a show that there exists a significant main effect with respect to both feed rate and speed; it can be noticed that the feed rate influences the (a)

(b) Speed (rpm)

Mean of DF-entrance

1.20

3

Feed (mm/min)

6

9 1.200 1.175

1.19 1.18

1.150

Speed (rpm)

Speed (rpm) 500 1500 2500

1.125

1.17 1.16

1.200

1.15

1.175

1.14

1.150

1.13

1.125

1.100 Feed (mm/min) 3 6 9

Feed (mm/min)

1.100

1.12 500

1500

2500

3

6

500

9

1500

2500

Fig. 33.4 a Main effect and b interaction plots of DF-entrance obtained for drilling operation with solid carbide drill

(a)

(b) Speed (rpm)

3

Feed (mm/min)

6

Mean of DF-exit

1.16

9

1.18 1.16 1.14

Speed (rpm)

1.15

Speed (rpm) 500 1500 2500

1.12 1.10

1.18

1.14

1.16 1.14

1.13

Feed (mm/min)

Feed (mm/min) 3 6 9

1.12 1.10

1.12 500

1500

2500

3

6

9

500

1500

2500

Fig. 33.5 a Main effect and b interaction plots of DF-exit obtained for drilling operation with solid carbide drill

414

N. S. Sowjanya et al.

variation in the response more compared to the speed. Initially, delamination factor was less when the feed rate was at low but as feed rate is increased from low to high, there is a drastic increase in the delamination factor. In the interaction plot of Fig. 33.5b, there exist a divergence in the lines of the plot which shows there is a significant interaction between speed and feed rate.

33.4.2 Drilling Tests with Diamond Core Drill Feed rate influences the delamination at both exit and entry side of the hole as observed from main effect plots of Figs. 33.6a and 33.7a. Delamination factor at hole entrance is observed to be less compared to delamination factor at hole exit. When compared with the other, two tools damages at hole entry were less and surface was clearer. When delamination at hole entry is considered, interaction between speed and feed is more significant. When the speed is increased from low to high, the delamination factor was decreased to lower values. With high speed of 2500 rpm and low feed rate of 3 mm/min, DF-entrance value was minimum as given from main (a)

(b)

Feed (mm/min)

Speed (rpm)

3

6

9 1.18

Mean of DF-entrance

1.15

1.16

1.14

1.14

Speed (rpm)

Speed (rpm) 500 1500 2500

1.12 1.10

1.13

Feed (mm/min) 3 6 9

1.18

1.12

1.16 1.14

1.11

Feed (mm/min)

1.12 1.10

1.10 500

1500

2500

3

6

500

9

1500

2500

Fig. 33.6 a Main effect and b interaction plots of DF-entrance obtained for drilling operation with diamond core drill

(a)

Speed (rpm)

1.20

(b)

Feed (mm/min)

3

6

1.20

1.19

Mean of DF-exit

9

Speed (rpm)

1.18

Speed (rpm) 500 1500 2500

1.15

1.17

1.10

1.16

1.20

1.15

Feed (mm/min)

1.15

Feed (mm/min) 3 6 9

1.14 1.10

1.13 500

1500

2500

3

6

9

500

1500

2500

Fig. 33.7 a Main effect and b interaction plots of DF-exit obtained for drilling operation with diamond core drill

33 Comparative Study on the Performance of Different Drill Bits …

415

effect plot of DF-entrance. With high feed and low speed delamination was critically severe compared with other levels of the cutting condition. Damages at the hole exit are severe when compared to hole entry. This is because when the tool is about to finish cutting, the thickness of the ply will be very less and the stress due to the thrust force exceeds the matrix bonding strength resulting in higher damages at hole exit. DF-exit was low at spindle speed of 500 rpm and feed rate of 3 mm/min. With high feed and high speed delamination was critically severe compared with other levels of the cutting condition. Since the divergence of the line segments is more as given in the interaction plot of Figs. 33.6b and 33.7b, a significant interaction between speed and feed rate can be identified.

33.4.3 Drilling Tests with Brad Point Bit When compared with the results of the other two tools, the delamination damages at entry were severe with brad point bit and fiber pull-out was observed with high feed rate of 9 mm/min and speed of 500 rpm. Even though the splinter problem was reduced by the spur of the bit, brad point peeled off the top layers producing higher DF-entry damage during the initial cutting. Since the thrust force of brad point reaches the higher values during cutting, it results in producing larger values of DF-exit compared to the other two drills. Effect of feed rate on delamination on both entry and exit side is more significant than speed as from main effect plots of Figs. 33.8a and 33.9a. Interaction plot for DF-entrance 8(b) shows that there is very less interaction between feed rate and speed and is less significant. Delamination is much more severe at hole exit for the holes machined with brad point bit. Severe delamination was observed at a speed of 2500 rpm and feed rate of 9 mm/min. With low feed and low speed delamination can be reduced. The interaction between feed rate and speed is much significant when interaction plot of DF-exit Fig. 33.9b is considered.

(a)

(b) Speed (rpm)

Mean of DF-entrance

1.23

3

Feed (mm/min)

6

1.22

1.25 1.20

Speed (rpm)

1.21

Speed (rpm) 500 1500 2500

1.15

1.20 1.25

1.19 1.18

1.20 Feed (mm/min)

1.17

Feed (mm/min) 3 6 9

1.15

1.16 1.15

9

500

1500

2500

3

6

9

500

1500

2500

Fig. 33.8 a Main effect and b interaction plots of DF-entrance obtained for drilling operation with brad point bit

416

N. S. Sowjanya et al.

(a)

Speed (rpm)

1.24

(b)

Feed (mm/min)

3

6

1.250

1.23

Mean of DF-exit

9 1.225 1.200

Speed (rpm)

1.22

Speed (rpm) 500 1500 2500

1.175

1.21

1.150

1.250

1.20

1.225

1.19

1.200

1.18

Feed (mm/min)

Feed (mm/min) 3 6 9

1.175

1.17

1.150

500

1500

2500

3

6

9

500

1500

2500

Fig. 33.9 a Main effect and b interaction plots of DF-exit obtained for drilling operation with brad point bit

Fig. 33.10 Microscope images of drilled hole a DF-entrance b DF-exit obtained at a spindle speed of 500 rpm and feed rate of 3 mm/min

The sample of microscopic images obtained for analyzing DF-entrance and DFexit for the three types of drill bits chosen for the experiment is given in Fig. 33.10. When analyzing delamination for entry and exit for spindle of 500 rpm and feed rate of 3 mm/min, clearer holes were obtained in the case of diamond core drill compared to solid carbide and brad point bit.

33.4.4 Study of Thrust Force During Drilling In the case of results obtained for thrust force Fig. 33.11, it is quite controversial with the results obtained for other response variables. Thrust force for brad point bit is high compared with solid carbide and diamond core drill. Diamond core tool reaches higher thrust force values during machining than solid carbide drill. Thrust force for diamond core drill was low at small spindle speed and low feed rates but as feed rate is increased from low to high, thrust force was increased to higher values. For all the three tools, thrust force increases as the feed is increased and decreases with increasing speed. Delamination reduces with lower thrust forces.

33 Comparative Study on the Performance of Different Drill Bits …

417

Fig. 33.11 Influence of parameters on thrust force: a effect of spindle speed on thrust force (F z ) b effect of feed rate on thrust force (F z )

In brief, the diamond core drill produces a better-quality hole compared to brad point and solid carbide tools. Diamond core drill gave a better result at higher speed and lower feed rate except with push-down delamination which is minimum at lower speed and lower feed rate. Results obtained with diamond cored drill are comparable with solid carbide bit. But the use of brad point bit resulted in poor surface quality. Machining defects like fiber pull-out and higher delamination defects were observed with holes machined using brad point bit. The use of diamond core drill can be considered economical compared to solid carbide bit which is costlier than diamond core drill.

33.5 Conclusions Based on the results of the comparative experimental conducted during drilling of CFRP using three different types of tool, the following results and key findings are drawn. These key findings help in the proper selection of tool and process parameters for machining CFRP.

418

N. S. Sowjanya et al.

Outcomes of the drilling tests with diamond core drill depict that the lowest DFentrance of 1.093 and DF-exit of 1.101 was obtained at a spindle speed of 500 rpm and feed rate of 3 mm/min with thrust force of 78 N. In the case of drilling performed with solid carbide bit, least DF-entrance of 1.101 at speed of 500 rpm and feed rate of 9 mm/min with thrust force of 99 N was obtained. Lowest thrust force of 69 N was observed at speed of 2500 rpm and feed rate of 6 mm/min. At spindle speed of 500 rpm, 3 mm/min lowest DF-exit of 1.102 was produced with 89 N of thrust force. For the drilling experiment with brad point bit, minimum thrust force of 100 N was observed at a speed of 1500 rpm and feed rate of 6 mm/min. DF-entrance of 1.123 and DF-exit of 1.152 were obtained at a spindle speed of 500 rpm and feed rate of 3 mm/min with the thrust force of 133 N. The use of diamond core drill is recommended for drilling CFRP in terms of economy and drilled hole quality.

References 1. Melentiev, R., Priarone, P.C., Robiglio, M., Settineri, L.: Effects of tool geometry and process parameters on delamination in CFRP drilling: an overview. In: 3rd CIRP Conference on Surface Integrity (CIRP CSI), pp. 31–34. North Carolina, USA (2016) 2. Das, G., Das, P.: Cutting forces in drilling operation: measurement and modeling for mediumscale manufacturing firms. Int. J. Comput. Appl. 121(8), 12–17 (2015). https://doi.org/10.5120/ 21559-4592 3. Tsao, C.C.: Thrust force and delamination of core-saw drill during drilling of carbon fiber reinforced plastics (CFRP). Int. J. Adv. Manuf. Technol. 37, 23–28. https://doi.org/10.1007/ s00170-007-0963-6 (2008) 4. Pinho, L.V., Carou, D., Davim, J.P.: Comparative study of the performance of diamond-coated drills on the delamination in drilling of carbon fiber feinforced plastics: assessing the influence of the temperature of the drill. J. Compos. Mater. 50(2), 179–189. https://doi.org/10.1177/ 0021998315571973 (2016) 5. Tsao, C.C.: Experimental study of drilling composite materials with step-core drill. J. Mater. Des. 29, 1740–1744. https://doi.org/10.1016/j.matdes.2008.03.022 (2008) 6. Priarone, P.C., Robiglio, M., Melentiev, R., Settineri, L.: Diamond drilling of carbon fiber reinforced polymers: influence of tool grit size and process parameters on workpiece delamination. In: 1st CIRP Conference on Composite Materials Parts Manufacturing, pp. 181–186. Karlsuhe, Germany (2017) 7. Konneh, M., Sudin, I., Abdullah Sidek, A.: Surface roughness in drilled carbon fiber reinforced polymer (CFRP) composite using diamond coated ball- nose drills. J. Adv. Mater. Res. 1115, 90–95. www.scientific.net/AMR.1115.90 (2015) 8. Robert, V., Henerichs, M., Kuster, F.: Comparison of conventional drilling and orbital drilling in machining carbon fiber reinforced plastics (CFRP). CIRP Ann. Manuf. Technol. 65, 137–140 (2016). https://doi.org/10.1016/j.cirp.2016.04.001 9. Robert, V.: Fundamentals of carbon fiber reinforced polymer (CFRP) machining. Ph.D. thesis, Eth Zurich (2017) 10. Eneyew, E.D., Ramulu, M.: Experimental study of surface quality and damage when drilling unidirectional CFRP composites. J. Mater. Res. Technol. 3(4), 354–362. https://doi.org/10. 1016/j.jmrt.2014.10.003 (2014)

33 Comparative Study on the Performance of Different Drill Bits …

419

11. Koenig, W., Wulf, C., Grass, P., Willerscheid, H.: Machining of fiber reinforced plastics. Ann. CIRP Manuf. Technol. 34(2), 537–548 (1985). https://doi.org/10.1016/S00078506(07)60186-3 12. Krishnamoorthy, A., Rajendra Boopathy, S.: Delamination analysis in drilling of CFRP composites using response surface methodology. J. Compos. Mater. 43, 2885–2902. https://doi. org/10.1177/0021998309345309 (2009) 13. Babu, J., Sunny, T., Alex Paul, N., Mohan, K.P., Philip, J., Davim, J.P.: Assessment of delamination in composite materials: a review. J. Eng. Manuf. 230(11), 1990–2003. https://doi.org/ 10.1177/0954405415619343 (2016)

Chapter 34

A Cyber-Physical System Improves the Quality of Machining in CNC Milling Machine—A Case Study Ganesh Kumar Nithyanandam , Saravana Kumar Sellappan and Selvaraj Ponnumuthu Abstract The authors have developed a portable cyber-physical system (CPS), U.S. patented (9971343), which helped to machine the components at a higher precision in a traditional CNC machine. The workpiece which needs to be machined is fixed to this device, which has its own controller, mounted on the table of the traditional CNC machine. This device learns, adopts, and adapts the in-process machining conditions and controls them autonomously using the artificial neural network. This paper discusses the development of this CPS device and how it helped to improve the machining quality. Keywords Cyber-physical system · Artificial neural network · Precision machining

34.1 Introduction Design engineers specify tolerances on a dimension based on either general tolerance or specific tolerance to meet the requirements of fit and assembly. Typically, tolerance specification is assigned to a dimension with upper specification limit (USL) and lower specification limit (LSL). For example, a component with rib thickness of 2 mm may be specified with a tolerance specification of ±0.022 mm. For precision machining, the dimension is to be realized with a tolerance specification narrower than the original tolerance specification. For instance, the same rib thickness of 2 mm may have to be machined with a tolerance specification of ±0.0055 mm. In order to achieve this precision machining, manufacturers have to use machines with higher precision. The reason for this is the limitation in the process capability of each CNC machine tool. G. K. Nithyanandam (B) · S. K. Sellappan Department of Mechanical Engineering, PSG College of Technology, Coimbatore 641004, India e-mail: [email protected] S. Ponnumuthu Aeronautical Research and Development Agency, Bangalore 560017, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_34

421

422

G. K. Nithyanandam et al.

Design and manufacturing engineers always strive for innovative means to machine the components at narrow tolerance specification. However, several factors such as cutting parameters, tool wear, vibration, thermal, deformation, etc., may influence the machining outcome. Today, industries try to implement Industry 4.0 (Fourth Industrial Revolution) solutions to solve several problems. The authors made an attempt to solve the machining problem (achieving the precision machining in a conventional CNC milling machine) using the Industry 4.0 approach. Industry 4.0 is a German term for achieving manufacturing excellence. Americans refer the Industry 4.0 as smart manufacturing. In Industry 4.0, machines are fully connected through wired or wireless networks, monitored using sensors, and controlled by advanced computational intelligence to improve product quality, system productivity, and sustainability while reducing costs [1]. The main feature of Industry 4.0 is a flexible mass production with full customization applied to several variant product lines using the automation of manufacturing processes. For this, machines need to operate independently. That implies that machines are capable to collect data, analyze data, and provide control for its actions. Internet of Things or Intranet of Things (IoT), cloud computing, and cyber-physical system (CPS) are key supporting technologies of smart manufacturing [2]. The key elements of Industry 4.0 are autonomous systems, Internet of Things (IoT), cloud computing, simulation, additive manufacturing, augmented reality, big data, cybersecurity, and simulation [3]. The autonomous systems are the machines and robots that act autonomously using computer algorithms/programs, tightly integrated with hardware components via Internet. It is also called as cyber-physical system (CPS). The cloud computing is used to store, manage, and process data remotely. The additive manufacturing helps to convert digital 3-D design data into a product layer-by-layer by depositing materials. The big data contains structured and unstructured data, which helps to improve processes. The cyber security protects valuable data from competitors. The IoT connect machines via the Internet (public access) or Intranet (private access) to send, receive, and process data. For Internet of Things CPS solutions, cyber security is a challenge from hackers; whereas, for the Intranet of Things CPS solutions, firewall is in-built, so cyber security is at low risk. The simulation is used to train employees and create situation planning. In these elements, CPS is the central, and all other elements are tightly integrated. Lee and Seshia [4] defined CPS as the integration of computing and physical processes such as embedded computers, network monitors, and controllers. Usually, they provide feedback where physical processes affect computations and vice versa. Lee [5] described a CPS as a mechanism where sensors and computer-based algorithms are tightly integrated with the Internet and its users. From the automation perspective, activities of CPS are controlled by computing and communication cores embedded in objects and structures of the physical environment [6]. Wang et al. [7] describe a CPS system, which has embedded and mobile sensing, cross-domain sensor sources and data flows, interaction of cyber and physical components, ability to train and adapt, interoperability through the Internet, robust system ensuring automated intellectual control, and human in and outside the loop.

34 A Cyber-Physical System Improves the Quality of Machining …

423

One of the key elements of CPS is machine learning. The machine learning algorithms such as artificial neural network, support vector machine, and random forest are used to predict tool wear in machining [8]. Helu et al. [9] used techniques such as fuzzy logic, neural networks, genetic algorithms, and hybrid systems to monitor machining operation and decision-making systems. Artificial neural network (ANN) is widely used in manufacturing applications; however, the performance of ANN is a challenge when a massive amount of data is to be handled [10]. The recent machine learning algorithm called deep learning overcomes this problem, and it performs well with big data [11]. Today, deep learning is widely used in applications related to speech recognition, image recondition, and natural language processing (translation, test questions and answers, etc.), multimodal image text, and games such as Alphago when compared to other artificial intelligence tools. This is because it is easy to solve problems related to nonlinear and complex feature abstraction, and its performance does not deteriorate even when the volume of data is high [10]. Several researchers worked on improving the quality of machining in CNC machines. Shi and Gindy [12] developed an on-line process monitoring system for hard turning operation using LabVIEW to detect tool wear and tool chipping. Kirby et al. [13] developed an in-process surface roughness adaptive control system for CNC turning operation using fuzzy logic. Saikumar and Shunmugham [14] demonstrated machining components using adaptive control system using an external open source controller for CNC turning center. From these, it is clear that these researchers used different techniques to improve the quality of machining but not at the domain of precision machining. This is because all of them were trying to communicate with an existing CNC controller by modifying the available CNC parameters, which restricts the precision in machining. There was no evidence found in the literature that the quality of machining was improved by using the external source (not open source controllers) or not modifying the existing parameters of the CNC controller. Having this in mind, the authors have developed a new cyber-physical system (CPS) device, which has the key elements of Industry 4.0. This device acts independent of CNC machine tool but augments the capability of a conventional CNC machine tool. In other words, it does not alter or re-program the available CNC machine tool but enhances the existing CNC machine tool to machine the components at higher precision. The development of this new CPS device is explained in the next section.

34.2 Development of Cyber-Physical System Typically, a CPS consists of five layers: physical, network, data storage, analytic, and application layers [1]. Figure 34.1 shows the cyber-physical system model of the developed technology.

424

G. K. Nithyanandam et al.

Fig. 34.1 Cyber-physical system of the developed model

34.2.1 Physical Layer Typically, the physical layer consists of sensors, actuators, tracking devices, and computing elements. The real-time machining data collected from the sensors by a controller is processed locally, and it is stored in the database, which resides in the cloud for further processing. Based on the precision machining embedded process algorithms, an instruction is commanded to the actuator for corrective action remotely. Here, the sensors are laser displacement systems (LDS), and the actuator is a portable modular fixture (PMF). In this layer, the actuator plays a very important role. Based on the embedded and ANN algorithms, the Netduino microcontroller commands actuator to perform the corrective action. Laser displacement system (LDS): It is used to collect real-time machining data. It consists of two non-contact sensors (Keyence LK-H0207), connected to a controller (Keyence 5001), as shown in Fig. 34.2. The Keyence 5001 controller transmits the machining data using API (processing layer) to the database (storage Fig. 34.2 Laser displacement system—two sensors connected to a local controller

34 A Cyber-Physical System Improves the Quality of Machining …

425

Fig. 34.3 Portable modular fixture

layer) via USB (network layer) using Inter-Integrated (I2C) communication protocol. A special fixture was designed to hold these two non-contact sensors to the spindle head of the CNC machine (application layer). Portable modular fixture (PMF): It is a uniquely designed and developed actuator called portable modular fixture (PMF). It consists of a modular fixture and portable base with a servomotor, as shown in Fig. 34.3. It can be mounted on any conventional CNC milling machine table (application layer). The workpiece which needs to be machined is fixed on this PMF. It helps to correct the error on thickness between actual machining thickness and desired thickness by feedback system using embedded and ANN algorithms (processing and analytic layer). In short, the PMF moves toward the cutting tool when the dimensional error is positive. Similarly, the PMF moves away from the cutting tool when the dimensional error is negative. Microcontroller: It is Netduino microcontroller, which is connected between the actuator and the database in the distributed cloud network. It executes the .NET embedded programs.

34.2.2 Network Layer The CPS accesses the CNC machine (application layer) using a wired serial communication port and using I2 C communication protocol. This protocol requires low power, and it is capable to transmit data between 100 Kbps to 3.4 Mbps. However, the connections between devices should be within a few meters only. On the other hand, the corrective action is transmitted to the actuator via Ethernet.

426

G. K. Nithyanandam et al.

34.2.3 Storage Layer The CPS system collects real-time data using sensors (physical layer), and it is stored in the cloud database. This database also contains historical machining data.

34.2.4 Processing and Analytic Layer The processing and analytic layer are used to process the data using vendor API, custom-embedded algorithm, simulation (PM software), and ANN algorithm. Using SQL queries, PM software is able to monitor the machining condition in real time. Data mining techniques are used to predict the future machining data point based on historical and real-time machining data. This layer helps to monitor and control actions that can be transmitted back to the physical layer to actuate the CNC milling machine. Vendor API program: It is Keyence API programs, which transmit the real-time machining data from local Keyence 5001 controller to database in the cloud. Custom-embedded algorithm: It is an embedded algorithm developed for 2-D straight machining condition, written in-house using C#. It carries out the data analytics functionality. It predicts the future machining data point based on customized artificial neural network (ANN). The ANN customization is based on the parameters such as number of hidden layers, number of neurons in the hidden layer, weight ratio of neurons in the hidden layer, and different types of transfer function. Simulation software: This software is developed in-house using .NET C#. It determines the optimum cutting parameters and the optimum ANN parameters based on the data available in the database and real-time machining data.

34.2.5 Application Layer The application layer is the user interface for manufacturing, and it is Makino S33 CNC milling machine.

34.3 Experimental Setup Al 6061 material is widely used in aircraft components like wing structures. For experimental feasibility study, a section of aircraft wing rib is selected as 100 mm length × 80 mm width × 25 mm height. For this work piece, a rib thickness of 2 mm with a narrow tolerance of ±0.011 mm (50% reduction in tolerance specification) is defined, to be machined in a conventional high-speed Makino S33 CNC

34 A Cyber-Physical System Improves the Quality of Machining …

427

milling machine. Ten-mm-diameter carbide tipped with two flute end mill cutter is used to machine the components. For the selected material and CNC machine, the optimum cutting parameters were determined from the preliminary study. With these parameters, the NC code was generated. Two sets of experiments were carried out. The first experiment is conventional machining, where the workpiece is fixed on the CNC table, and the machining is carried out using the NC code. The component is machined, and the rib thickness is measured using digital Vernier caliper. The second experiment is machining the workpiece using CPS, where the workpiece is fixed on the CPS device, which is then mounted on the conventional CNC milling machine table. Figure 34.4 shows the experimental setup using CPS technology. In this machine, there are two controllers: The first controller is the CNC controller, provided by the vendor; the second controller is developed by the authors. The second controller does not interact with the primary CNC controller. Both the controllers are closed-loop systems. The CPS technology acts as an auxiliary device and does not interface or interact with the existing CNC machine, but it complements the capability of the CNC milling machine. In this setup, the workpiece is machined using the same NC code. The machined rib thickness is measured automatically by CPS. A sample of ten work pieces was machined in Makino S33 CNC vertical milling machine between conventional machining and machining using CPS technology.

Fig. 34.4 Precision machining setup in a conventional CNC milling machine

428

G. K. Nithyanandam et al.

Fig. 34.5 Machining comparison between conventional machining and conventional machining using CPS technology

34.4 Results and Discussion Figure 34.5 shows the first 20 data points of the machined components of the two experiments. The X-axis represents the various data points of the machined component; whereas, the Y-axis represents the component thickness. The design engineer defined the original tolerance specification as ±0.022 mm. That means, for target mean of 2 mm thickness, the USL and LSL are defined as 2.022 mm and 1.978 mm, respectively, which is defined in blue dotted lines (in Fig. 34.5). When the narrow tolerance of 0.011 mm (±0.0055 mm) was defined for the same target, then the USL and LSL are defined as 2.0055 mm and 1.9945 mm, respectively, which are defined in yellow dotted line. The blue checkbox line represents traditional machining in a conventional CNC milling machined; whereas, the black dotted line represents the traditional machining using CPS technology in a conventional CNC milling machine. The interpretation of this study is that if any data points fall outside the narrow tolerance zone (yellow dotted line) is considered as not supported. From Fig. 34.5, even though all the data points of the first experiment (traditional machining in a conventional CNC milling machine) fall within the original tolerance specification (USL and LSL), they were out of control limits of narrow tolerance specifications (USL and LSL ). From this, it is clear that the first experiment did not support the study; whereas, the second experiment (traditional machining using CPS technology in a conventional CNC milling machine) supported the study.

34.5 Conclusions Following conclusions are drawn from the present study: • The conventional CNC milling machine cannot perform the precision machining when narrow tolerances beyond its machine capability are defined. But, the

34 A Cyber-Physical System Improves the Quality of Machining …

429

same conventional CNC milling machine with CPS technology can perform the precision machining with narrow tolerances are defined. • The conventional CNC milling machine is capable of machining the components at 3.2 sigma quality without CPS technology. But, the same conventional CNC milling machine is capable of machining the components at 4.8 sigma quality with CPS technology. • The limitation of this technology is that the algorithms are created for 2-D straight machining condition. It needs to be expanded to 3-D straight and 2-D/3-D curved machining conditions (contour surfaces).

References 1. Al-Ali, A.R., Gupta, R., Al Nabulsi, A.: Cyber-physical systems role in manufacturing technologies. In: 6th International Conference on Nano and Materials Science, AIP Conference Proceedings 1957, pp 050007(1–7) (2018). https://doi.org/10.1063/1.5034337 2. Wang, L., Torngren, M., Onori, M.: Current status and advancement of cyber-physical systems in Manufacturing. J. Manuf. Syst. 37, 517–527 (2015) 3. Essentra: What is Industry 4.0? Essentra PLC. http://www.essentracomponents.com/en-gb/ news/news-articles/industry-40-hub (2018). Retrieved 25 June 2018 4. Lee, E.A., Seshia, S.A.: Introduction to Embedded Systems: A Cyber-Physical Systems Approach. The MIT Press, Cambridge, England (2017) 5. Lee, S.: Internet of Things, Unist. http://isystems.unist.ac.kr/research/internet-of-things/ (2018). Retrieved 25 June 2018 6. Johansson, K.H.: Control of cyber-physical systems: fundamental challenges and applications to transportation networks. In: 27th International Conference on Architecture of Computing Systems, Lübeck Germany (2014) 7. Wang, J., Ma, Y., Zhang, L., Gao, R.X., Wu, D.: Deep learning for smart manufacturing: methods and applications. J. Manuf. Syst. (2018). https://doi.org/10.1016/j.jmsy.2018.01.003 8. Nitze, I., Schulthess, U., Asche, H.: Comparison of machine learning algorithms random forest artificial neural network and support vector machine to maximum likelihood for supervised crop type classification. In: Proceedings of the 4th GEOBIA, May 7–9, Rio de Janeiro, Brazil. pp. 35–40 (2012) 9. Helu, M., Libes, D., Lubell, L., Lyons, K., Morris, K.: Enabling smart manufacturing technologies for decision-making support. In: Proceedings of the ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, pp. 1–10 (2016) 10. Wuest, T., Weimer, D., Irgens, C., Klaus, D.: Machine learning in manufacturing: advantages, challenges, and applications. Prod. Manuf. Res. 4(1), 23–45 (2016) 11. Chen, X.-W., Lin, X.: Big data deep learning: challenges and perspectives. IEEE Translations 2, 514–525 (2014) 12. Shi, D., Gindy, N.N.: Development of an online machining process monitoring system: Application in hard turning. Sens. Actuators 135, 405–414 (2007) 13. Kirby, E.D., Chen, J.C., Zhang, J.Z.: Development of a fuzzy-nets-based in-process surface roughness adaptive control system in turning operations. Expert Syst. Appl. 30, 592–604 (2006) 14. Saikumar, S., Shunmugam, M.S.: Investigations into high-speed rough and finish end-milling of hardened EN24 steel for implementation of control strategies. Int. J. Adv. Manuf. Technol. 63(1), 391–406 (2012)

Chapter 35

Challenges in Machining of Silica–Silica Cone for Aerospace Application Ashish Tewari , T. Srinivasulu , B. Hari Prasad

and A. P. Dash

Abstract Silica–silica composites are generally used in high-end aerospace applications. They are highly temperature resistant, have low thermal conductivity and low density, and are abrasive in nature. Additionally, due to the soft and brittle nature of the material, it poses challenges during the machining and clamping of the workpiece for achieving the required shape and size. In realizing quality hardware, it becomes very critical to suitably plan the machining strategy. A manufacturing methodology is established to carry out machining of silica-based composite with primary focus on the development of special turning fixtures as a shop setup. This paper discusses the criticalities involved in the machining of the silica material using suitably developed fixtures. Keywords Machining · Silica · Fixture

35.1 Introduction In the present aerospace industry scenario, it is very important to develop suitable machining methodologies for achieving the maximum quality in machined parts, especially considering the nature of the workpiece raw material. The selection of costeffective and reliable manufacturing process is one of the key factors in aerospace industry to sustain in the present market. Silica-based composites are extensively used in the aerospace industry primarily due to their thermal characteristics. They are highly temperature resistant, have low thermal conductivity and low density, and are abrasive in nature. The raw material in the current work is silica-based composite. It is made up of silica fibers as reinforcement and silica powder as matrix. Additionally, the raw material is soft, fragile, porous, and brittle. This poses challenges during the machining and clamping of the workpiece for achieving the required shape and size. Sometimes customized and economical fixtures have to be made on need basis for prototype works. Most of the A. Tewari (B) · T. Srinivasulu · B. H. Prasad · A. P. Dash Defence Research & Development Laboratory, Hyderabad 500058, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_35

431

432

A. Tewari et al.

available literature describes the development/preparation of ceramic-based hard, brittle composites and study of their various thermal, mechanical and functional properties. However, few efforts have been made to study the machining aspect of silica-based high thermal performance composite addressing the tool wear and thermal aspects [1]. Advanced machining methodologies such as ultrasonic machining [2, 3], laser-based machining [4], and hybrid techniques [5] are utilized for achieving better machinability of these hard ceramic composites. Some studies are done with conventional machining method using different cutting tools [6]. Moreover, these studies are done on specimens. On the contrary, the composite material in the present study is very fragile/soft. This poses difficulty in clamping or fixturing while machining. It becomes even more challenging to hold the material when it has a thin wall section with a tapered geometry. This paper addresses machining methodology and development of special fixtures in order to machine a product made of highly abrasive, very soft, and brittle silica–silica composite for high-end aerospace application. It is required to machine a truncated cone profile made of silica–silica composite as shown in Fig. 35.1. The properties of the workpiece material are given in Table 35.1. Fixture elements play a vital and basic role in any machining process. There are wide varieties of standard fixtures available in the market, but, sometimes customized and economical fixtures have to be made on need basis for prototype works and these shop setups require ingenuity. This paper mainly emphasizes the ingenuity involved during the development of suitable fixtures in the realization of this hardware.

Fig. 35.1 Truncated cone profile

Table 35.1 Properties of the work material

1

Density

0.70 g/cc

2

Melting point

1700 °C

3

Hardness (shore A)

50

35 Challenges in Machining of Silica–Silica Cone …

433

35.2 Development of Special Fixtures In general, the fixtures and mandrels are designed out of metallic materials such as steels or aluminum, depending upon the need of the machining process, shape of the job, workpiece material, and available holding features. However, in the present case study, due to the inherent soft nature of the raw material, as it is prone to indentation and damages during machining and clamping, it is critical to develop work holding devices using suitable raw material. Wood and foam materials are selected for fixture development to ensure compatibility and avoid unwanted dents or damages into the workpiece. Multiple setups are required in order to complete the machining process. Clamps and fixtures are developed for each setup keeping into the mind, the machining processes involved and the feasibility of holding. An annular ring and a star-shaped wooden clamp are made as shown in Fig. 35.2 to carry out facing of top face during turning and milling modes, respectively. Facing in milling mode is done during the last stage, when total length is to be maintained. The star-shaped clamp serves a dual purpose. Firstly, it secures the workpiece. Secondly, the peripheral spaces available around the circumference allow easy access to the milling tool to carry out facing operation. A circular plate as shown in Fig. 35.3 made of wood is developed to carry out facing of bottom face while turning. Another wooden circular plate is made as shown in Fig. 35.4 to clamp the job for external profile turning. Finally, a foam-based fixture as shown in Fig. 35.5 is made with in built clamping provision to carry out internal profile turning. The foam fixture is machined such that its internal profile is suitable to locate the finished external profile of the workpiece. Additionally, holes are made into the fixture so as to provide clamping provision for holding the workpiece.

Fig. 35.2 Annular ring-type and star-shaped wooden clamps for top facing operation

434 Fig. 35.3 Circular plate-type clamp for bottom facing operation

Fig. 35.4 Circular plate-type clamp for external profile machining

Fig. 35.5 Foam-based internal machining fixture with holes for clamping provision

A. Tewari et al.

35 Challenges in Machining of Silica–Silica Cone …

435

Fig. 35.6 Sequence of operations

35.2.1 Sequence of Operations In order to realize a product from the raw material stage, it is required to conceptualize the type of machining operations involved and the same has to be in proper sequence along with inspection. A detailed planning of operations as shown in Fig. 35.6 is carried out, and a logical sequence is established to realize the hardware. To effectively meet the machining requirements, selecting an appropriate cutting tool plays a critical role during machining. Additionally, rigidity of the cutting tool is equally important as a vibrating tool may lead to the workpiece fracture and dimensional inaccuracies. Coated carbide cutting tools are used in machining as the material is very abrasive. All the machining operations are carried on 3-axes CNC turn mill center. The machine has a swing of Ø1000 mm and can cater job height of 1000 mm.

35.2.2 Prior Inspection of Raw Material As the raw material is supplied in near-net-shaped condition, it is required to estimate the machining allowances before starting the actual cutting operations. Therefore,

436

A. Tewari et al.

Fig. 35.7 Estimation of machining allowances by comparing inspection data and final required dimensions

inspection of the as supplied raw material is carried out using CMM and the inspection data is used to know the actual raw material profile. The CMM inspection for internal and external profile is carried out at 4 angular orientations (0°, 90°, 180°, and 270°). The data obtained from the inspection is exported to SolidWorks. This scan data helps to generate a 2-D profile for any given pair (0° and 180° or 90° and 270°). This planar profile is subsequently superimposed upon the final required profile. To estimate the machining allowances available, the inspection data is superimposed upon the final required sizes as shown in Fig. 35.7.

35.2.3 Facing of Top and Bottom Faces Facing both top and bottom faces is the primary machining operation carried out in order to generate proper references. The main challenge in the facing operation is clamping. Face clamping cannot be used because the clamps should not cover the faces and obstruct the cutting tool. Thus, a wooden annular ring is made with holes which serve as a purpose for clamping through bolts as shown in Fig. 35.8. As the initial thickness of the workpiece is sufficient, it is able to sustain the clamping load Fig. 35.8 Job setup during the facing of top face

35 Challenges in Machining of Silica–Silica Cone …

437

Fig. 35.9 Job setup during the facing of bottom face

and no indents or damages are observed on the workpiece. After the machining of top face, the job is inverted/reversed and a circular plate-type clamp is used to hold the job from inside to carry out facing of bottom face, as shown in Fig. 35.9. With this methodology, both the faces are machined without any hindrances and damages to the workpiece.

35.2.4 Turning External Profile A circular ring-type wooden pate is made for clamping the workpiece for the external profile machining as shown in Fig. 35.10. After the top face is machined, the same face is used for clamping as the wooden plate can now be properly seated. A T-slot through-bolt is used to hold the workpiece and clamping plate in place. External diameter is dialed to center the workpiece on the machine table. Bottom face is taken as a Z-reference (for maintaining height). CNC program is generated, and the external profile is machined to the required dimensions. In this process, dimensional inspection is carried out to ensure the accuracy. Fig. 35.10 Job setup during the external profile machining

438

A. Tewari et al.

35.2.5 Turning Internal Profile and Finish Facing To machine the internal profile, the cone has to be inverted/reversed for better accessibility of the cutting region. After the external profile is completed, the workpiece thickness is reduced further. Keeping in view the soft and brittle nature of the workpiece material, it now becomes difficult to hold the job in an inverted position using internal clamping. This is so because the clamping force will tend to dent and damage the workpiece. Since the workpiece cannot be clamped from inside, the only option is to clamp from outside with proper and complete support. Due to the above reasons, a new fixture is made of foam material. Based on the requirements, a single block of foam material measuring 1000 mm × 1000 mm × 300 mm was made by gluing multiple small foam blocks. Subsequently, a taper internal profile is machined into the block which is matched to the external profile of the workpiece as shown in Fig. 35.5. This taper will help in easy location and alignment of the workpiece. Additionally, four holes are made into the foam block which serves the purpose of face clamping. This fixture is used to securely and safely hold the workpiece during internal profile machining as shown in Fig. 35.11. In this process, inspection is done to ensure the dimensions. After machining the final workpiece, thickness of 9 mm is achieved. By post-completion of internal profile, the final height is maintained by reversing the job and finish facing the top face. Due to the soft and brittle nature of the workpiece material, it now becomes difficult to hold the job in the same setup as shown in Fig. 35.8 to machine the top face. Hence, a new star-shaped wooden clamp (Fig. 35.2) is developed which helps to clamp the job on the face itself, as well as, it allows the tool to face the material in milling mode as shown in Fig. 35.12. Top facing is done in two setups. In first setup, a T-slot through-bolt is used to hold the workpiece and star clamp plate in place. Fig. 35.11 Job setup during the internal profile machining

Fig. 35.12 Finish facing

35 Challenges in Machining of Silica–Silica Cone …

439

Fig. 35.13 Final CMM inspection

Material is removed in milling mode along the peripheral spaces available around the circumference. After completion of first setting, the star clamp is rotated and re-clamped to remove the remaining material.

35.2.6 Final Inspection To ensure the required quality, the dimensional and profile inspection is carried out using CMM. The CMM inspection for internal and external profile is carried out at four angular orientations (0°, 90°, 180°, and 270°). This scan data provides the internal diameter, external diameter, total height, and thickness. In this process, inspection results have helped to achieve the required results. The final inspections were checked on CMM as shown in Fig. 35.13 and found to meet the desired dimensional and geometrical requirements.

35.3 Results Suitable fixtures and mandrels are developed for clamping a soft, brittle material and carry out machining. The fixtures used to machine the workpiece played a critical role in achieving the required quality. No dents or cracks are observed on the finished job. The machined component achieved the dimensional accuracies within 150 microns. The maximum circularity error observed is within 200 microns. The surface roughness is not a critical parameter; however, finish of the order of N7 is achieved. The internal diameter dimensions are important for the assembly purpose, and they are achieved as per the tolerances.

440

A. Tewari et al.

35.4 Conclusions This paper mainly highlights the vitality of fixtures and clamps in the machining activity and discusses the ingenuity involved during the development of suitable fixtures required for machining a soft and brittle workpiece material. It is learnt and also emphasized, that work holding is a major contributing factor in the realization of these components. In the present study, not much emphasis is given to tool wear as satisfactory results were achieved using coated carbide cutting tools. A future study can be taken up to compare the performances of various types of cutting tools. A manufacturing methodology is established in order to hold and machine a soft/brittle material and achieve the required quality on the component. Acknowledgements The authors would like to acknowledge Director DRDL for his constant encouragement to take up challenging tasks.

References 1. Sreejith, P.S., Krishnamurthy, R., Narayanasamy, K., Malhotra, S.K.: Thermal aspects during machining of high silica/phenolic ablative composites. Mater. Manuf. Proces. 14(5), 647–659 (1999) 2. Singh, R.P., Singhal, S.: Investigation of machining characteristics in rotary ultrasonic machining of alumina ceramic. Mater. Manuf. Process. 32(3), 309–326 (2017) 3. Ning, F.D., Cong, W.L., Pei, Z.J., Treadwell, C.: Rotary ultrasonic machining of CFRP: a comparison with grinding. Ultrasonics 66, 125–132 (2016) 4. Dong, X., Shin, Y.C.: Improved machinability of SiC/SiC ceramic matrix composite via laserassisted micromachining. Int. J. Adv. Manuf. Technol. 90, 731–739 (2017) 5. Yan, B.H., Wang, A.C., Huang, C.Y., Huang, F.Y.: Study of precision micro-holes in borosilicate glass using micro EDM combined with micro ultrasonic vibration machining. Int. J. Mach. Tool Manuf. 42, 1105–1112 (2002) 6. Yan, B.H., Huang, F.Y., Chow, H.M.: Study on the turning characteristics of alumina-based ceramics. J. Mater. Process. Technol. 54, 341–347 (1995)

Chapter 36

Optimization of Cutting Parameters for Hard Turning of WC–Co–Ni–Cr (15% Binder) Mill Rolls on CNC Lathe with Polycrystalline Diamond Mahesh J. Hunakunti , Vaishali Jagannath , Ramesh S. Rao , S. Srinivas , S. Shyamsundar and D. Ashokkumar Abstract Hard turning has found to be a viable method for machining hard parts like cemented tungsten carbides. This method is effective and efficient with polycrystalline diamond (PCD)—a superhard cutting tool with a longer tool life. The objective of this paper is to optimize the cutting parameters for multimodal grade of PCD. Round full face PCD insert with hardness of greater than 9000 HV is used for turning WC–Co–Ni–Cr (15% Binder) mill rolls on CNC lathe. The responses like taper, cycle time, flank wear, and maximum spindle load are recorded. Optimal cutting parameters have been obtained from consideration of different responses. Abrasion on the flank is found to be the major wear mechanism. Keywords PCD · PCBN · Taper · Flank wear · Maximum spindle load

36.1 Introduction Hard turning refers to turning of workpiece material having greater than 45 HRC hardness on a lathe or turning center. It is often termed as a pre-grinding process or a replacement of grinding operations if desired surface finish is obtained. It is most often performed on post-heat-treated parts with surface hardness ranging from 45 to 68 HRC or even higher [1]. Cemented tungsten carbide is one such material which has hardness of over 61–63 HRC. Hence, these materials are considered under hard turning. These materials are hard, tougher, and wear resistant, and hence, they are used in many applications such as cutting tools, punches, dies and wear parts. Currently, WC-based materials are machined by grinding, electric discharge machining (EDM). But these methods have longer cycle time and lower material removal rate (MRR) and M. J. Hunakunti (B) · S. Srinivas Manufacturing Science and Engineering, Department of Mechanical Engineering, BMS College of Engineering, Bengaluru 560019, India e-mail: [email protected] V. Jagannath · R. S. Rao · S. Shyamsundar · D. Ashokkumar Kennametal India Limited, Bengaluru 560073, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_36

441

442

M. J. Hunakunti et al.

involve higher machining cost. In order to overcome these problems, hard turning is found to be a viable method for an acceptable quality of finish for tungsten carbides. The surface finish can be obtained after grinding process. Tungsten carbide rolls are widely used in metal forming industries especially in hot rolling industries where steel wires are drawn between cemented tungsten carbide rolls in the intermediate and finishing stands [2]. They replace the steel rolls because of their excellent wear resistance and longer life of almost 50 times than steel rolls. Carbide rolls are in sintered condition after processing through powder metallurgy (PM) process. Hence, they are to be finished to final dimensions according to the customer requirements. Grinding and hard turning are the only machining methods to finish the tungsten carbide rolls to the required specifications. Hard turning has always been preferred over grinding method for its higher MRR when precise tolerances are not required. Polycrystalline diamond (PCD) and polycrystalline cubic boron nitride (PCBN) are two known superhard cutting tools used in hard turning of hard metals like cemented tungsten carbide and cermets. While selecting PCD/PCBN, two most important properties must be considered: • Grain size and diamond/CBN content play an important role that determines the hardness of the tool and its application in machining operations like roughing, finishing, and interrupted cut. • Wear resistance and toughness effect the tool life. Hard turning of tungsten carbide rolls are succeeded by grinding; hence, a coarser grain size with higher diamond/CBN volume (%) is an optimal selection for hard turning. Nowadays, multimodal grades containing different grain sizes have shown excellent cutting action across various industries. PCD/PCBN comes in brazed tip form or full face form. Each has their own advantage. A brazed tip will save the cost and its regrinding can be eliminated. But a full face PCD/PCBN has better heat absorption capacity and can go higher DOC than tipped tools [3]. Even though their cost is high, they efficiently cut the material and can be used multiple times after regrinding. Hard turning of sintered cylindrical workpiece was tried with chemical vapor deposition (CVD) diamond coated tool. Cemented tungsten carbide with 12% Co as a major binder was used as workpiece material. A full factorial design has been considered with roughness and roundness as the responses. The result showed that the surface roughness increases when feed rate increases for any cutting speed value and the roundness is strongly influenced by depth of cut. In an opposite way, the increase of feed rate results in no influence in roundness [4]. Environmentally conscious hard turning and micro-cutting process in SEM was carried out by using four kinds of tungsten carbides with the PCD cutting tools. Result showed that the wear is more for higher binder content carbides. The thrust forces were larger in coarser WC grain size carbides. WC particle diameter had an influence in tool life besides increasing content of Co [5]. Optimization of cutting parameters for hard turning with triangular style brazed tip PCD tools of different nose radii has been tried by recording cutting forces. The

36 Optimization of Cutting Parameters for Hard Turning …

443

investigation showed the result analysis of the recorded total cutting force components—feed force (F f ), passive force (F p ), and main cutting force (F c ) during WC–Co (25% Co) turning. The biggest increase in passive cutting force component F p was indicated. Obtained values of the cutting forces were used to determine the mathematical equations, describing the growth of the particular components of the total cutting forces. This gave the hint of generated cutting forces [6]. Machinability of the PCD tools was studied while cutting sintered WC (15% Co) at different speed ratios by applying the UEVC method [7]. The influence of binder content of cemented tungsten carbide on PCD and CBN wear characteristic during hard turning was studied with different binder (Co) contents of 10, 15, 20, and 25% carbides without coolant. The results showed that the tool wear was mainly due to abrasion on flank and less wear on rake face. Flank wear was major in the case of higher cobalt content carbide among other work materials. And adhesion on flank face of 10% cobalt binder carbide was more. Turning of WC–Co (25% binder) resulted in continuous chips while WC–Co (10% binder) produced segmented chips. Turning with CBN tool showed that flank wear tends to be less when cobalt content increases and CBN showed less wear for WC–Co (25% binder) than PCD tool [8]. In the current field of hard turning, surface finish of the component, cutting force evaluation, and tool life estimation has been major responses. There has not been work carried out for measuring the taper of the workpiece material after hard turning. And, moreover, in most of the cases of studies either CVD-coated diamond or brazed PCD indexable inserts were used. Only in one of the case of the literature survey, there has been study on outer diameter (OD) turning with round full face diamond tool. The usage of round full face with multiple cutting edges has an added advantage over tipped inserts and their limited usage due to high initial cost is only a disadvantage. But considering its regrinding and usage capability, it has found to be an effective tool. Usually, a round tool is used for profile turning purposes. Their usage as an OD turning operation is limited and used only when surface finishing is required hence an attempt is made to use its full potential in hard turning of sintered tungsten carbide for OD rough turning application. Hard turning can be done on conventional or CNC lathe. A CNC lathe has a greater advantage over conventional in terms of rigidity, accuracy, and repeatability. It can drastically reduce cycle time once the cutting parameters are set. Hence, the objective is to hard turn the WC–Co–Ni–Cr (15% binder) on CNC lathe with full face round PCD inserts.

36.2 Experiments Full face PCD round inserts are used in OD turning of WC–Co–Ni–Cr (15% Binder) mill rolls. Physical and chemical properties of workpiece material and multimodal PCD tool (Grade 1) are given in Tables 36.1 and 36.2, respectively. The experiments are divided into two phases:

444

M. J. Hunakunti et al.

Table 36.1 Workpiece material composition and physical properties Grade

Grain size

Binder composition

Binder %

Hardness HV 30 kg/mm2

WC–Co–Ni–Cr

Coarse

Co–Ni–Cr

15

1000

Table 36.2 Cutting tool composition

PCD tool grade

Grain size (µm)

PCD (vol.%)

Grade 1

2–30

NA

• Hard turning with reground inserts of multimodal grade and • Fresh inserts of multimodal grade. Hard turning was done on CNC lathe. Soft inner diameter (ID) holding jaws were used to hold the workpiece up to 80% of its length to ensure firm clamping. Contact length of the tool is 65 mm/single pass. Coolant used is a water miscible, mineral oil-based, high-performance cutting fluid. All of the experiments were conducted in ambient condition. Depth of cut (DOC) is on overall diameter in mm. CRGCL25.4 X 25.4 M12 tool holder is used for the experiments. The first phase of the experiments with reground inserts of Grade 1 is done in order to understand the cutting action of round inserts. The results of the experiments have not been included in this paper as they did not have the proper tool micro-geometry. The recommended cutting parameters were taken from different companies PCD tools catalogue. The observations from these experiments were taken for the trials with fresh inserts. The tool and workpiece setup has been done on CNC lathe (Fig. 36.1a, b). Second Phase of experiments was done with fresh inserts of multimodal grade with RNGN120300F tool geometry designated by “Grade 1.” One of the parameters is selected considering the result from the previous experiments and the other parameters were randomized to understand the cutting action. In one of the parameter, cutting speed was increased to 40 m/min. In all the cases, DOC was constant, i.e., 0.5 mm. The responses are—taper, cycle time, flank wear, and maximum spindle load (%). For taper, 0.1 mm was the limit of criteria, tool life criteria for flank wear Fig. 36.1 a Full face round indexable PCD tool in a toolholder. b Work holding setup on CNC lathe

36 Optimization of Cutting Parameters for Hard Turning … Table 36.3 Experimental plan for fresh inserts of multimodal grade

S. No.

445

Cutting parameters Cutting speed (m/min)

Feed (mm/rev)

DOC (overall) (mm)

1

25

0.35

0.5

2

25

0.4

0.5

3

30

0.2

0.5

4

30

0.25

0.5

5

40

0.25

0.5

are 0.3 mm, and maximum permissible spindle load is 50%. The experimental plan is given in Table 36.3.

36.3 Results and Discussion 36.3.1 Results of Multimodal Grade The results of multimodal grade are discussed in this section. Taper is measured by differentiating the OD at start of the cut and the end of cut in mm. Cycle time in minute is calculated for a fixed stock. Flank wear measurements (V b ) were done using Nikon measuring microscope by measuring from referencing the top edge up to the maximum wear on flank face and maximum spindle load is displayed on the CNC panel. The results of the experiments are given in Table 36.4. Multiple experiments are carried out for each combination of parameters and the average of the responses is tabulated. Graphs were plotted based on following conditions: • For plotting flank wear graphs, the average values of replicates of particular cutting parameters are taken into account after every two passes, i.e., 130 mm of contact length. Table 36.4 Results for fresh inserts of multimodal grade S. No.

Cutting speed (m/min)

Feed (mm/rev)

Taper (mm)

Cycle time (min)

Flank wear (mm)

Max spindle load (%)

1

30

0.2

0.15

22.65

0.3610

26

2

30

0.25

0.046

18.06

0.1912

23

3

25

0.35

0.033

15.84

0.1629

27

4

25

0.4

0.02

13.95

0.1724

28

5

40

0.25

0.05

13.8

0.2624

31

446

M. J. Hunakunti et al.

• Spindle load is monitored and recorded for every pass and graphs are plotted based on maximum spindle load. • Plotted taper value is an average value in mm for replicates after every two passes of the tool. Figure 36.2 shows variation of different responses for different cutting parameters. Cycle time is lowest for V c = 40 m/min and f = 0.25 mm/rev. Theoretically, cycle time depends on cutting speed, feed, and DOC. Spindle load is minimum for V c = 30 m/min and f = 0.25 mm/rev. It is highest for V c = 40 m/min and f = 0.25 mm/rev as higher cutting speed generates larger cutting forces. Taper is minimum for higher feed rate with parameters: V c = 25 m/min and f = 0.4 mm/rev. Except at V c = 30 m/min and f = 0.2 mm/rev, all the parameters are within specified range of taper for the finished workpiece. In the graph for V c = 30 m/min, taper decreased when feed rate increased from f = 0.2 mm/rev to f = 0.25 mm/rev. Similarly, at V c = 25 m/min, taper decreased when feed rate of 0.35 mm/rev increased to 0.4 mm/rev. But while changing cutting speed from 30 to 40 m/min with constant feed rate of 0.25 mm/rev, the taper decreased by a small amount. Increase in feed rate had an influence over taper but increase in cutting speed for same feed rate decreased the taper by only few microns. This shows that increase in feed rate decreases the taper due to faster cutting action. Variations of flank wear for different cutting parameters indicated minimum value for at V c = 25 m/min, f = 0.35 mm/rev. For the same V c = 25 m/min, increased feed rate of 0.4 mm/rev resulted in slight increase of wear. But for V c = 30 m/min,

Fig. 36.2 Variation of cycle time, maximum spindle load, taper, and flank wear for different cutting parameters

36 Optimization of Cutting Parameters for Hard Turning …

447

increase in feed rate from 0.2 to 0.25 mm/rev decreased the flank wear. Only V c = 30 m/min and f = 0.2 mm/rev parameter crossed the tool life criterion of 0.3 mm. It can be concluded that flank wear increases with decrease in feed rate. Noise level of machining was considered for final selection of parameters and it was based on maximum load. From the analysis of different graphs, it is clear that optimum cutting parameters for multimodal full face PCD for hard turning WC–Co– Ni–Cr (15% Binder) are V c = 30 m/min, f = 0.25 mm/rev, and DOC = 0.5 mm at which the machining is smooth and even though surface finish was not considered as the response, and the optimized parameter gave a better finish than any of the above parameters. The next best option is V c = 25 m/min, f = 0.35 mm/rev, DOC = 0.5 mm. For any cutting parameter in the above experiments with overall DOC of >1 mm, the tool fails and it needs to be indexed. Otherwise, it affects the component severely.

36.3.2 Tool Wear Flank wear tends to be the major wear mechanism in the experiments. Figure 36.3 shows flank wear at different parameters. In Fig. 36.3a, thermal discoloration at the brazed layer was observed. In all the experiments, wear pattern was uniform except at V c = 40 m/min, f = 0.25 mm/rev as larger cutting speed lead to a greater flank wear and the wear reached the carbide substrate.

(a)

(b)

(c)

Vb

(d)

(e)

Fig. 36.3 Flank wear at different cutting parameters—a Vc = 30 m/min. f = 0.2 mm/rev, b Vc = 30 m/min, f = 0.25 mm/rev, c Vc = 25 m/min, f = 0.35 mm/rev, d Vc = 25 m/min, f = 0.4 mm/rev, e Vc = 40 m/min, f = 0.25 mm/rev

448

M. J. Hunakunti et al.

36.4 Conclusions From the experiments and analysis with reground, fresh inserts of multimodal grade, it is clear that• For any cutting parameter in the above experiments with multimodal grade, hard turning of 15% binder carbide rolls with overall DOC of >1 mm and a contact length of >130 mm for two passes (each pass of 65 mm), the tool fails and it needs to be indexed. Otherwise, it affects the component and tool severely. • Wear is less and consistent with multimodal grade. • Higher cutting speeds such as V c = 40 m/min should not be used as it abrades the tool very fast and the load on the spindle increases rapidly which in turn draws more power and affects the spindle on long service. • Abrasion is the major wear mechanism in turning 15% binder carbide rolls. • The best combination of parameters for hard turning of WC–Co–Ni–Cr (15% Binder) mill rolls with multimodal PCD grade is: V c = 30 m/min, Feed = 0.25 mm/rev, Overall DOC = 0.5 mm.

Essential Questions 1. What is your main contribution to the field? Once the objective of the project was set, I started doing literature survey on machinability of tungsten carbide in turning and arrived at a conclusion that I had to use a PCD tool which is efficient and cost-effective. Hence, we fixed the PCD tool based on its mechanical properties and its usage capability. I had to do preliminary experiments to understand the cutting action of selected tool. Based on the experiments, I was able to decide the parts per edge and wear mechanisms. The next phase was again with a motto of conducting experiments in order to understand the cutting action. After this, I was able to conclude on fixing the cutting parameters. Gathered the information in order to find cost per component and tried my best to suggest an optimized parameter for the hard outer diameter turning of cemented tungsten carbide rolls. 2. What is novel? In theory, in experimental techniques, or a combination of both? We used full face PCD tool than normal tipped PCD tools. The tool used is a different grade of PCD than optimal recommendation from diamond tool manufacturers for turning cemented tungsten carbides. In experimental technique, we used a different tool geometry for outer diameter turning which is usually used for profile turning operation. We were able to effectively use the cutting edges to reduce the CPC. 3. Does your paper have industrial applications? If yes, who are the likely user? Yes it has a wide usage in hard turning of cemented tungsten carbide rolls or any other component made of hardmetals. Hardmetal manufacturers are the likely users.

36 Optimization of Cutting Parameters for Hard Turning …

449

4. How this work is different from past research work (yours and others)?* (please include this statement in the paper). Our work stands apart from others in selecting a different grade of PCD and utilization of the PCD tool effectively to reduce the CPC. Past research works focused mainly on using tipped PCD tools and surface finish and flank wear were the major responses. We focused on industrial requirements and considered few other responses.

References 1. Koepfer, C.: Hard turning as an alternative to grinding, production machining. Gardner Business Media (2010) 2. Frayman, L.I.: Cemented carbide material for hot rolling application. General Carbide Corporation, Greensburg, PA 15601, USA (1) (2010) 3. Innovations Master catalogue @2012 by Kennametal Inc., Latrobe, PA 15650 USA, pp. B140 4. Coppini, N.L., Diniz, A.E., Bonandi, M., De Souza, E.M., Baptista, E.A.: Hard turning of sintered cemented carbide parts: a shop floor experience. 4th CIRP Conference on Modeling of Machining Operations (CIRP CMMO). Procedia CIRP 8, 368–373 (2013) 5. Heo, S.J.: Environmentally conscious hard turning of cemented carbide materials on the basis of micro-cutting in SEM: stressing four kinds of cemented carbides with PCD tools. J. Mech. Sci. Technol. 22, 1383–1390 (2008) 6. Z˛ebala, W., Kowalczyk, R., Matras, A.: Analysis and optimization of sintered carbides turning with PCD tools. Procedia Eng. 100, 283–290 (2015) 7. Nath, C., Rahman, M., Neo, K.S.: Machinability study of tungsten carbide using PCD tools under ultrasonic elliptical vibration cutting. Int. J. Mach. Tools Manuf. 49, 1089–1095 (2009) 8. Miyamoto, T., Fujiwara, J., Wakao, K.: Influence of WC and Co in cutting cemented carbides with PCD and CBN tools. Key Eng. Mater. 407–408, 428–431. Online: 2009-02-20. ISSN: 1662-9795

Chapter 37

Multi-response Optimization of End Milling on Al6061–Sicp Metal Matrix Composite–Hybrid GRA-PCA Approach B. Ravi Sankar

and P. Umamaheswarrao

Abstract In this study, optimization scheme for minimum machining force and amplitude during end milling of Al6061 reinforced with SiCp has been presented. The input parameters considered for the analysis are cutting speed, feed, depth of cut and weight fraction. Experiments are designed based on Taguchi L25 orthogonal array. The parametric influences on responses are discussed through main effects plot. Further, hybrid Grey Relational Analysis (GRA) and Principle Component Analysis (PCA) was performed. Results revealed that weight fraction is the considerably influencing factor affecting the responses followed by depth of cut, speed and feed. The optimum cutting parameters obtained are cutting speed 2500 rpm, feed rate 20 mm/rev, depth of cut 0.2 mm and weight fraction 4%. Keywords Al6061–SiCp composite · Machining force · Amplitude · GRA · PCA

37.1 Introduction Metal matrix composites (MMCs) find applications in various engineering fields over decades due to their low cost and superior properties [1]. Anyway the ceramic particulate reinforcements like metal oxides, carbides and so on in MMCs make the machining of composites harder [2, 3]. The end milling process was frequently used in the manufacturing industry due to the faster material removal rate and good surface quality [4]. Jeyakumar et al. [5] have optimized machining parameters in end milling of the Al6061/SiCp composite using a genetic algorithm. Arun Premnath et al. [6] concluded that cutting force and surface roughness are mainly influenced by speed and feed while milling Al356/SiC metal matrix composites (MMCs). Rajeswari and Sivasakthivel [7] reported that cutting speed is a significant parameter followed by rake angle, helix angle and nose radius. Jeyakumar et al. [8] reported that the reduction in feed rate and depth of cut leads to an increase of vibration. They also concluded that the feed rate has a significant influence on vibration. Low surface B. Ravi Sankar (B) · P. Umamaheswarrao Department of Mechanical Engineering, Bapatla Engineering College, Bapatla, AP 522102, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_37

451

452

B. Ravi Sankar and P. Umamaheswarrao

roughness is observed at higher speeds and high at lower feed rates. Jeyakumar et al. [9] demonstrated that at higher speeds, low surface roughness is observed and high at lower feed rates. Higher depths of cut result in higher cutting forces and surface roughness. Sujay et al. [10] showed that the feed was the most significant milling parameter affecting acceleration amplitude followed by weight fraction, depth of cut and speed for Al6061–SiC metal matrix composite milling. Hybrid GRA-PCA proved to be a prominent technique for obtaining optimum parameters for multiple responses and was adopted by several authors Ravi Sankar et al. [11, 12], Avinash et al. [13], Kishore Kumar et al. [14], Ravichandra et al. [15] and Umamaheswarrao et al. [16]. The past studies reported the effect of cutting parameters on the machining force and surface roughness during milling of MMCs. However, the induced vibrations have much effect on machining quality for MMCs and got less focus in the available literature. Hence, the present study is aimed to conduct parametric optimization for minimizing force and vibration amplitude during end milling of Al6061–SiC composite.

37.2 Experimental Work Aluminium alloy (Al6061) is heated in a furnace at 620 °C until complete melting in a furnace and SiCp particles at various weight fractions 4, 8, 12, 16 and 20% added to molten metal through the side of vertex created by mechanical stirring by the stir impeller. After mixing, the melt was poured into a prepared mould for the preparation of specimen in the dimensions 110 × 90 × 45 mm. The machining is done on CNC vertical machining centre model: BMV 40 T20. The tool used was carbide end mill cutter with four flutes, and size is Ø12 × 75 mm. Dytran 7543A equipment is used to measure vibration amplitude (shown in Fig. 37.1). It consists of a device which is made up of a Ti/4Al alloy. It is mounted with a USB cable which is connected to a computer that contains VibraScout software. Fig. 37.1 Vibration measuring instrument [10]

Acceleration Amplitude measuring instrument

Work piece

37 Multi-response Optimization of End Milling on Al6061–Sicp …

453

The experimental runs are designed in line with Taguchi’s L25 orthogonal array. Four parameters cutting speed, feed, depth of cut and weight fractions are varied at five levels, and their influences on machining force and acceleration amplitude are examined (Table 37.1). Table 37.1 Experimental matrix along with results Experiment Speed no. (rpm)

Feed (mm/rev)

Depth of cut (mm)

1

500

10

0.2

2

500

20

0.4

3

500

30

4

500

40

5

500

6 7 8

Weight fraction (%)

Machining force (N)

Amplitude (m/s2 )

4

24.84

8.243

8

69.85

8.042

0.6

12

73.98

9.531

0.8

16

67.56

8.369

50

1

20

89.93

6.964

1000

10

0.4

12

18.98

9.601

1000

20

0.6

16

33.7

6.747

1000

30

0.8

20

81.55

8.646

9

1000

40

1

4

70.82

7.489

10

1000

50

0.2

8

57.79

8.503

11

1500

10

0.6

20

34.21

9.426

12

1500

20

0.8

4

30.64

8.938

13

1500

30

1

8

33.65

7.3

14

1500

40

0.2

12

53.2

6.83

15

1500

50

0.4

16

66.22

8.327

16

2000

10

0.8

8

41.87

8

17

2000

20

1

12

87.42

8.113

18

2000

30

0.2

16

45.7

8.875

19

2000

40

0.4

20

33.5

8.549

20

2000

50

0.6

4

35.87

6.866

21

2500

10

1

16

87.71

9.054

22

2500

20

0.2

20

17.17

10.332

23

2500

30

0.4

4

40.85

6.782

24

2500

40

0.6

8

75.22

7.77

25

2500

50

0.8

12

88.05

5.977

454

B. Ravi Sankar and P. Umamaheswarrao

37.3 Methodology Adopted 37.3.1 Hybrid GRA-PCA Scanty information is enough to evaluate even the complex project performance with the help of grey relational analysis. The weight contributions of the parameters are estimated through PCA by explaining the variance and covariance structure of a set of defined variable [12, 16]. The steps executed are as follows: (1) experimental results/responses, (2) data pre-processing via normalization, (3) calculation of deviational sequence, (4) calculation of grey relational coefficient, (5) formulation of a covariance matrix, (6) compute the principal, and (7) calculate the grey relational grade. Normalized values and deviational sequences are shown in Table 37.2. Eigenvalues and eigenvectors are shown in Tables 37.3 and 37.4, respectively. The variance contribution for the first principal component characterizing the whole original variables, i.e. the two performance characteristics, is as high as 64.08%.

37.4 Results and Discussion The experiments were conducted as per the Taguchi L25 orthogonal array. The multiresponse optimization was performed by estimating GRG (shown in Table 37.5). The larger GRG indicates the better multiple performance characteristics, and therefore, the levels at which the largest average response was obtained were selected and marked with (*) for each factor and depicted in Table 37.6. A larger delta value indicates the higher significance of the parameter in controlling the response. In the response table (Table 37.6), it can be seen that weight fraction has been assigned a rank 1 which means it is the most significant parameter in controlling the response followed by the depth of cut, speed and feed. From the ANOVA, it is clear that weight fraction contribution is highest (43.48%) followed by depth of cut (24.86%), speed (20.55%), feed (11.08%) as shown in Table 37.7. Figure 37.2 illustrates the main effect plot for means of GRG. The peak value at each level of Fig. 37.2 represents the optimal result for GRG, i.e. A5 (cutting speed at 2500 rpm), B2 (feed rate at 20 mm/rev), C1 (depth of cut at 0.2 mm) and D3 (weight fraction at 4%), and the same was observed from the response table for the grey relational grades shown in Table 37.6. The higher GRG was noticed at 4% weight fraction for pure metal (Al6061) due to increased particle concentration which enhances machining force and amplitude. This is due to the increase in hardness of the material with particulate reinforcement. GRG for the obtained optimum combination of parameters was 0.715203 estimated from Eq. (37.1). The GRG for the optimum combination of parameters was

37 Multi-response Optimization of End Milling on Al6061–Sicp …

455

Table 37.2 Normalized values and deviation sequences for machining force and amplitude Experiment no.

Normalized values

Deviation sequences

Machining force

Amplitude

Machining force

Amplitude

1

0.894584

0.479678

0.105415

0.520321

2

0.275975

0.525832

0.724024

0.474167

3

0.219213

0.183926

0.780786

0.816073

4

0.307449

0.450746

0.692551

0.549253

5

0

0.773363

1

0.226636

6

0.975123

0.167853

0.024876

0.832146

7

0.772814

0.823191

0.227185

0.176808

8

0.115173

0.387141

0.884827

0.612858

9

0.262644

0.652812

0.737356

0.347187

10

0.441726

0.419977

0.558274

0.580022

11

0.765805

0.208036

0.234195

0.791963

12

0.814870

0.320091

0.185129

0.679908

13

0.773501

0.696211

0.226498

0.303788

14

0.504810

0.804133

0.49519

0.195866

15

0.325865

0.460390

0.674134

0.539609

16

0.660527

0.535476

0.339472

0.464523

17

0.034496

0.509529

0.965503

0.490470

18

0.60788

0.334557

0.392111

0.665442

19

0.775563

0.409414

0.224437

0.590585

20

0.742990

0.795866

0.257009

0.204133

21

0.030511

0.293455

0.969489

0.706544

22

1

0

0

1

23

0.674546

0.815154

0.325454

0.184845

24

0.202171

0.588289

0.797828

0.411710

25

0.025838

1

0.974162

0

Table 37.3 Eigenvalues and explained variation for principal components Principal component

Eigenvalue

Explained variations (%)

First

1.2304

0.6408

Second

0.6896

0.3591

Table 37.4 Eigenvectors for principal components and contribution Responses

Eigenvector First principal component

Machining force Amplitude

Contribution Second principal component

0.7071

0.7071

0.49999

−0.7071

0.7071

0.49999

456

B. Ravi Sankar and P. Umamaheswarrao

Table 37.5 Grey relational coefficient, grey relational grade and rank of the machining force, amplitude Experiment no.

Grey relational coefficient

Grey relational Grade

Rank

Machining force

Amplitude

1

0.82588

0.490041

0.657947

7

2

0.408489

0.513258

0.460864

19

3

0.390385

0.379917

0.385144

24

4

0.419269

0.476529

0.44789

21

5

0.333333

0.688102

0.510708

14

6

0.952605

0.375333

0.663956

6

7

0.687583

0.738761

0.713158

1

8

0.361056

0.449293

0.405167

23

9

0.404088

0.590188

0.497128

15

10

0.472468

0.46295

0.467701

17

11

0.681018

0.387007

0.534002

13

12

0.729789

0.423761

0.576764

10

13

0.688233

0.622053

0.65513

8

14

0.502417

0.718528

0.61046

9

15

0.425846

0.480949

0.453389

20

16

0.595612

0.518390

0.55699

12

17

0.34118

0.504810

0.422987

22

18

0.560468

0.429021

0.494735

16

19

0.690192

0.458469

0.574319

11

20

0.660494

0.71009

0.68528

2

21

0.340254

0.414406

0.377323

25

22

1

0.333333

0.666653

5

23

0.605728

0.730092

0.667897

4

24

0.385259

0.548419

0.46683

18

25

0.339176

1

0.669575

3

Table 37.6 Response table for means of grey relational grade Factor

Level 1

Level 2

Level 3

Level 4

Level 5

Delta

Rank

Speed

0.492511

0.549422

0.565949

0.546862

0.569656*

0.07714

3

Feed

0.558044

0.568085*

0.521615

0.519325

0.557331

0.04876

4

Depth of cut

0.579499*

0.564085

0.556883

0.531277

0.492655

0.08684

2

Weight fraction

0.617003*

0.521503

0.550424

0.497299

0.53817

0.11970

1

*The highest value for that factor

37 Multi-response Optimization of End Milling on Al6061–Sicp …

457

Table 37.7 ANOVA for grey relational grade Source

DF

SS

MS

F

P

% Contribution

Speed

4

0.0191

0.0048

0.38

0.819

20.55

Feed

4

0.0103

0.0026

0.20

0.936

11.08

Depth of cut

4

0.0231

0.0058

0.47

0.758

24.86

Weight fraction

4

0.0404

0.0101

0.88

0.493

43.48

Fig. 37.2 Main effects plot for GRG

0.286% more than highest GRG in Table 37.5. Hence, the combination of the milling parameters has provided a solution to minimization of machining force and amplitude. γ = γm +

q 

(γ j − γm )

(37.1)

i=1

37.5 Conclusions The present study aimed to optimize machining force and amplitude during end milling of Al6061 reinforced with SiCp composite. The statistical experimental plan L25 orthogonal array is devised and implemented. Multi-response optimization is achieved through GRA coupled with PCA technique. The following conclusions were drawn from this study:

458

B. Ravi Sankar and P. Umamaheswarrao

• Weight fraction is observed to be the most significant parameter affecting the multi-objective optimization followed by the depth of cut, speed and feed. • The recommended optimum combination of parameters for minimizing machining force and amplitude was speed—2500 rpm, feed rate—20 mm/rev, depth of cut— 0.2 mm and weight fraction—4%. • An increase in the value of predicted weighted GRG from 0.713158 to 0.715203 confirms the improvement in the performance of end milling.

References 1. Karabulut, S. U˘gur, G. Henifi, Ç.: Study on the mechanical and drilling properties of AA7039 composites reinforced with Al2O3/ B4C/SiC particles. Compos. Part B 93, 43–55 (2016) 2. Karabulut, S., Henifi, Ç., Halil, K.: Experimental investigation and optimization of cutting force and tool wear in milling Al7075 and open-cell SiC foam composite. Arab. J. Sci. Eng. 41, 1797–1812 (2016) 3. Davim, J.P.: Machining of metal matrix composites. Springer, London (2012) 4. Manna, A., Bhattacharayya, B.: A study on machinability of Al/SiC-MMC. J. Mater. Process. Technol. 140, 711–716 (2003) 5. Jeyakumar, S., Marimuthu, K., Ramachandran, T.: Optimization of machining parameters of Al6061 composite to minimize the surface roughness-modelling using RSM and ANN. Indian J. Eng. Mater. Sci. 22, 29–37 (2015) 6. Arun Premnath, A. Alwarsamy, T. Rajmohan, T.: Experimental investigation and optimization of process parameters in milling of hybrid metal matrix composites. Mater. Manuf. Process. 27, 1035–1044 (2012) 7. Rajeswari, S. Sivasakthivel, P.S.: Optimisation of milling parameters with multi-performance characteristic on Al/SiC metal matrix composite using grey-fuzzy logic algorithm. Multidiscipline Model. Mater. Struct. (2018). https://doi.org/10.1108/MMMS-04-2017-0027 8. Jeyakumar, S., Marimuthu, K., Ramachandran, T.: Prediction of vibration amplitude and surface roughness in machining of Al6061 metal matrix composites by response surface methodology. Int. J. Mech. Mater. Eng. 7(3), 222–231 (2012) 9. Jeyakumar, S., Marimuthu, K., Ramachandran, T.: Prediction of cutting force, tool wear and surface roughness of Al6061/SiC. J. Mech. Sci. Technol. 27(9), 2813–2822 (2013) 10. Sujay, P., Ravi Sankar, B., Umamaheswarrao, P.: Experimental investigations on acceleration amplitude in end milling of Al6061–SiC metal matrix composite. Procedia Comput. Sci. 133, 740–745 (2018) 11. Ravi Sankar, B., Umamaheswarrao, P., Srinivasulu, V., Kishore Chowdari, G.: Optimization of milling process on jute polyester composite using Taguchi based grey relational analysis coupled with principle component analysis. Mater. Today: Proc. 2, 2522–2531 (2015) 12. Ravi Sankar, B., Umamaheswarrao, P.: Optimisation of hardness and tensile strength of friction stir welded AA6061 alloy using response surface methodology coupled with grey relational analysis and principle component analysis. Int. J. Eng. Sci. Technol. 7(4), 21–29 (2015) 13. Avinash, K.L. Umamaheswarrao, P., Ravi Sankar, B.: Optimization of cutting force and surface roughness during hard turning of AISI 52100 steel using hybrid GRA-PCA. In: Proceedings of 3rd International Conference on Advanced Manufacturing & Automation, Krishnankoil (2018) 14. Kishore Kumar, M., Ravi Sankar, B., Umamaheswarrao, P., Suresh, K., Srikanth, K., Pavan Kumar, K.V.P., Manoj Kumar, M.: Optimization of tensile strength and hardness on friction stir welding of AA6061–AA7075 using hybrid GRA–PCA. In: Proceedings of 3rd International Conference on Advanced Manufacturing & Automation, Krishnankoil (2018)

37 Multi-response Optimization of End Milling on Al6061–Sicp …

459

15. Ravichandra, K., Ravi Sankar, B., Umamaheswarrao, P., Suresh kumar, K., Deva Sundar, G.J., Narendra Babu, J., Surendra, K.: Optimization of process parameters during friction stir welding of AA6061 using hybrid GRA–PCA. In: Proceedings of 3rd International Conference on Advanced Manufacturing & Automation, Krishnankoil (2018) 16. Umamaheswarrao, P., Ranga Raju, D., Suman, K.N.S., Ravi Sankar, B.: Multi objective optimization of process parameters for hard turning of AISI 52100 steel using hybrid GRA–PCA. Procedia Comput. Sci. 133, 703–710 (2018)

Chapter 38

Experimental Investigation on Dewaxed Tungsten Carbide-Based Self-lubricant Cutting Tool Material A. Muthuraja

Abstract In this paper, an attempt has been made to characterize the dewaxed tungsten carbide-based self-lubricant cutting tool material. In the course of characterization, crystalline size, particle size and morphological analysis were carried out using X-ray diffraction (XRD), transmission electron microscope (TEM) and scanning electron microscope (SEM), respectively. The XRD result confirms that the intensity of the diffraction peaks broadened for 40 h of milling as compared to raw powder sample. The result reveals the fracture resistance of beyond 40-h-milled particles increased, as strong cohesion between small particles occurs. The SEM analysis reveals the homogenously rounded grains of tungsten Carbide-Cobalt-5 wt% Calcium Fluoride system with 4 wt% stearic acid after 40 h of milling. The TEM analysis showed the homogeneous distribution of tungsten carbide particles in the considered milled samples in 40 h with the minimum possibility of agglomeration as compared to 100-h-milled powders. Keywords Tungsten carbide · Ball milling · Dewaxing · TEM and SEM

38.1 Introduction Tungsten carbide is a kind of traditional cutting material with the superior hardness and high wear and high heat resistance that can be found in many diverse applications especially cutting tool, radiation shielding materials and the rocket nozzle applications. In order to improve the performance of cutting tool with high strength and toughness, a certain amount of binders (Co, Ni, etc.) is added with tungsten carbide. The ductile phase of Co binder in WC enhances a relatively high toughness and transverse rupture strength, and also it enables sintering to relatively low temperatures. On the other hand, the ductile phase of binder reduces hardness and temperature stability. However, in the last few decades, performance requirements A. Muthuraja (B) Department of Mechanical Engineering, Sandip University, Mahiravani, Nashik 422213, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_38

461

462

A. Muthuraja

of tungsten carbide cutting tools have been more demanding due to excellent hightemperature strength and wear resistance without reacting with the workpiece in high-speed machining. Many articles have been reported on tungsten carbide-based materials in the last few decades [1, 2]. Tungsten carbide is mostly blended via powder metallurgy route [3, 4]. Cutting tool material is to be chosen to withstand high temperature, fatigue, abrasion, attrition and chemical-induced wear. But, very few articles are exclusively available on the overall characterization of WC-based selflubricant cutting tool material [1, 4]. A significant amount of the works has been carried out in the earlier to understand the effect of cobalt and lubricant content over friction and wear behaviour [5]. In cemented carbide industry, ball milling is one of the well-accepted methods for mixing and grinding of carbide materials [6–8]. In the light of achieving purity and minimum particle size without or with minimum agglomeration, a process control agent (PCA) is generally added in the range of 1–5 wt% to the powder mixture during ball milling [5]. The PCA minimizes cold welding between powder particles and thereby inhibits agglomeration. The weight proportion of the PCA to the powder is considered as the critical parameter in milling [3, 5]. The effects of ball milling, compaction pressures and solid lubricants for tungsten carbide-based cutting tool material were reported elsewhere [4]. In this article, tungsten carbide-based self-lubricant cutting tool material was characterized using X-ray diffraction (XRD), scanning electron microscope (SEM) and transmission electron microscope (TEM). The objective of the present study is to cultivate an idea to choose the appropriate parameters for tungsten carbide-based cutting tool material.

38.2 Experimental Procedure 38.2.1 Materials and Methods Tungsten carbide (WC) is chosen as a base material due to its high hardness, good strength and excellent wear resistance. Cobalt (Co) is chosen as a binder and 10 wt% of Co is commonly used for the WC-based composites [4, 8]. The calcium fluoride (CaF2) is considered as a solid lubricant, for cutting tool application due to its lubrication stability even at a higher temperature. The stearic acid of 1–5 wt% is chosen and used as a solvent (processing controlling agent) to avoid cold welding during ball milling. The proposed materials are milled in the high-speed planetary ball mill (Insmart Systems, PBM07) under nitrogen (0.5 kg/cm2 ) atmosphere. Plate and bowl speed was set at 90 and 207 rpm, respectively, and powder-to-ball ratio was taken as 1:5. Plate speed and powder ball ratio was chosen to avoid agglomeration as well as contamination during milling [4, 8]. In the course of the high-temperature sintering process, the dewaxing process is considered an essential step in the development of materials in powder metallurgy route. The milled powders were taken for the

38 Experimental Investigation on Dewaxed Tungsten Carbide …

463

dewaxing process, as the high tendency of crack formation in the developed sample during sintering is due to the presence of process controlling agent [6]. The milled powders were measured before and after dewaxing using the precious weighing machine. The powders were kept in the smallest container, and sequentially, container is put inside the dewaxing set-up. The dewaxing set-up consists of a Variac, temperature controller and vacuum pump. The Variac (variable transformer of 10 A) is for equipment testing on low and high voltages, and temperature controller is to control the temperature and vacuum pump for removing vapour stearic acid (PCA) during dewaxing. The dewaxing process was carried out in a vacuum atmosphere at 400 °C for an hour. Then the dewaxed powders were collected and kept in airtight container (desiccator). To characterize the considered materials, the effect of PCA wt. % was taken into account and followed by effect of milling time was verified, as it is most significant on powder properties. WC was milled with 10 wt% of Co and 5 wt% of CaF2 for different duration of 20, 40, 60, 80, and 100 h for the milling characterization. During milling, PCA was added to minimize or eradicate the cold welding. The detailed methodology adopted for the cutting tool material development has been reported elsewhere [3, 4]. In order to determine the crystallinity of the developed material, XRD investigations were performed on the WC–10Co–5CaF2 samples before and after ball milling, and the milled powders were characterized with the aid of X-ray diffractometer (Bruker D8 Advance). XRD patterns were obtained with Cu Kα radiation (λ = 0.1504 Å) in 10° ≤ 2θ ≤ 75° in steps 0.05°. The full width at half maximum (FWHM) is denoted by β hkl , and it is obtained with the aid of Powder X, diffraction data analysis software. Microstrain (ε) generated on the milled powders was also obtained by computing ratio between d-spacing of milled and raw pattern. Peak broadening due to microstrain is determined using Wilson’s formula. Borah et al. [9] also characterized the material by utilizing Wilson’s formula. Crystallite sizes of milled powders were acquired from broadening due to the size effect (β) using Scherer’s equation [10]. The milled samples were examined in JEOL, JEM 2100, and transmission electron microscope (TEM). The milled powders were ultrasonically agitated in a propanol solution for 10 min to reduce the agglomeration of powders. Then, TEM samples are dispersed on the copper grid and allowed to dry and followed by TEM grid which is loaded on the sample arm. The grid is safely guarded with no mechanical binding of the specimen cradle at tilt limits. Then the vacuum starts generated for TEM analysis.

464

A. Muthuraja

38.3 Result and Discussion 38.3.1 SEM Analysis on Morphology of Milled Particles After Dewaxing Dewaxed samples were examined in a scanning electron microscope (SEM) to study morphology. Figure 38.1a–c shows the morphology of WC–10 wt% Co–5 wt% CaF2 powder with 2, 3 and 4 wt% of stearic acid after 40 h of ball milling. Figure 38.1a revealed that the most particles are found to be agglomerated due to rapid heat generation by striking of carbide balls on bowl and powders at the beginning of milling. Figure 38.1b showed that agglomeration between particles are reduced progressively, due to the presence of 3 wt% of stearic acid. Figure 38.1c showed that agglomeration between particles are almost eliminated progressively and achieved homogeneous powder mixture due to the presence of 4 wt% of stearic acid. From this SEM analysis, the influence of stearic acid on reduction/elimination of cold welding during ball milling is absorbed. Borah et al. [9] also investigated the effect of stearic acid in milling. It was observed that stearic acid can prevent the agglomeration to a certain limit. With considering the effect of stearic acid on agglomeration, the particles’ morphology with 2, 3 and 4 wt% was considered for present investigation.

38.3.2 XRD Analysis on Powder Crystallinity The XRD patterns of powder mixture are illustrated in Fig. 38.2, and mechanical milling results revealed progressive broadening effect on the diffraction peaks of sample with increasing milling time, as compared to unmilled sample [4]. The X-ray diffraction patterns in Fig. 38.2 revealed that the diffraction peaks broadened with an increase in milling hours. With respect to the starting pattern, the sample milled for 40 h shows the progressive weakening of the peaks by refinement of the powder particles in X-ray pattern lines, which are detectable after about 40 h of milling due to increased fracture resistance by agglomeration. It indicates that the crystallite of milled particles is reducing with milling time of up to 40 h [2]. Moshtaghioun et al. [11] also identified the effect of milling time on nanoparticles by utilizing the XRD pattern. The crystallite size was rapidly declined after 20-h milling and the small reduction in crystallite size after 40-h milling. From Fig. 38.3, it is witnessed that the average crystallite size of the milled powder reduces considerably up to 40 h of milling. Crystallite size of the milled particles is in the range of 16–26 nm. From this, it is observed that the crystallite size of the milled powders reduces considerably up to 60 h of milling. The brittle characteristics of the tungsten carbide contribute to the rapid fracturing of the particles, and hence, crystallite size also reduces. The result showed the excellent morphological features and a uniform size. Figure 38.3 [4] shows microstrain of considered powders after different milling

38 Experimental Investigation on Dewaxed Tungsten Carbide … Fig. 38.1 SEM images of 40-h-milled powders with a 2 wt%; b 3 wt%; c 4 wt% of stearic acid after dewaxing

465

466

A. Muthuraja

Fig. 38.2 XRD patterns of WC–10Co–5CaF2 powders before and after milling in different hrs

Fig. 38.3 Effect of milling time on crystallite size and fracture resistance of material

hours in the range of 0.5–0.6%. Microstrain is directionally proportional to the energy required to fracture the particles. From this result, it was observed that beyond 60 h of milling, the energy required to fracture the particle increases due to the increased particle and crystallite size [2].

38.3.3 TEM Analysis on Shape and Distribution Figure 38.4a, b shows the TEM micrograph of the composite WC–10 wt% Co–5 wt% CaF2 after 40 and 100 h of ball milling. Figure 38.4a shows a homogenous distribution of tungsten carbide particles in the considered milled powders in 40 h with the minimum possibility of agglomeration. TEM analysis utilized to obtain average particle size [5]. It was shown that after 30 h

38 Experimental Investigation on Dewaxed Tungsten Carbide … Fig. 38.4 a TEM micrograph of dewaxed 40-h-milled powder. b TEM micrograph of dewaxed 100-h-milled powder

(a)

(b)

467

468

A. Muthuraja

of milling, the particle size stabilized. Figure 38.4b shows the maximum possibility of agglomeration with 100 h of milling. The result reveals that particles embedded into each other, due to cohesion between small particles and it was obtained that the beyond 40 h of milling, the particles lean towards to agglomeration. It is noted that, however, as the area of analysis for TEM samples is very small, an absolute result could not be determined.

38.4 Conclusions The present study revealed that there is the significant role of ball milling time on Crystallite size, particle size and morphology of milled powders after dewaxing. Following major conclusions were made from the present work. 1. SEM analysis shows the importance of stearic acid and dewaxing process. It indicates that developed material during ball milling with 4 wt% of stearic acid has superior homogenous morphology. 2. X-ray diffraction patterns for 40 h of milled powders show most progressive broadening. The average crystallite size of milled powders is 15–17 nm at 60 h milling and beyond 60 h of milling, and the increase in average crystallite size was witnessed due to the particle agglomeration. From microstrain analysis, it was observed that beyond 60 h of milling, the energy required to fracture the particles increases due to particle agglomeration. 3. TEM analysis shows the maximum possibility of agglomeration in developed materials with 100 h of ball milling. It is noted that, however, as the area of analysis for TEM samples is very small, an absolute result could not be determined. Acknowledgements Authors wish to thank Aerospace Manufacturing and Value Engineering Panel of ARDB, India (DARO/08/1103/M/I), for funding this project. Authors also wish to thank his supervisor Prof. Dr. S. Senthilvelan and Professor Dr. PS. Robi, IIT Guwahati, and Centre for Instrument Facility, IIT Guwahati, for their kind assistance.

References 1. Deng, J.X., Wenlong, S., Hui, Z.: Design, fabrication and properties of a self-lubricated tool in dry cutting. Int. J. Mach. Tools Manuf. 49(1):66–72 (2009) 2. Sarin, V.K.: Cemented carbide cutting tools. In: Chin, D.Y. (ed.) Advances in Powder Technology. ASM Material Science, Louisville, KY (1981) 3. Muthuraja, A., Senthilvelan, S.: Milling characteristics of tungsten carbide based self-lubricant cutting tool material. Procedia Mater. Sci. 5, 1998–2003 (2014) 4. Muthuraja, A., Senthilvelan, S.: Development of tungsten carbide based self-lubricant cutting tool material: preliminary investigation. Int. J. Refract Metal Hard Mater. 48, 89–96 (2015) 5. Dabhade, V.V., Rama Mohan, T.R., Ramakrishnan, P.: Nanocrystalline titanium powders by high energy attrition milling. Powder Technol. 171, 77–83 (2007)

38 Experimental Investigation on Dewaxed Tungsten Carbide …

469

6. Suryanarayana, C.: Mechanical alloying and milling. Prog. Mater Sci. 46, 1–184 (2001) 7. Haglund, B.O.: A DSC study of the dewaxing of cemented carbide powder. Thermochim. Acta 65, 323–326 (1985) 8. Hewitt, S.A., Kibble, K.A.: Effects of ball milling time on the synthesis and consolidation of nanostructured WC–Co composites. Int. J. Refract Metal Hard Mater. 27, 937–948 (2009) 9. Borah, A., Robi, P.S., Srinivasan, A.: Synthesis of nano-crystalline RuAl by mechanical alloying. Met. Mater. Int. 13, 293–302 (2007) 10. Langford, J.I., Wilson, A.J.C.: Scherrer after sixty years, a survey and some new results in the determination of crystallite size. J. Appl. Crystallogr. 11, 102–113 (1978) 11. Moshtaghioun, B.M., Monshi, A., Abbasi, M.H., Karimzadeh, A.: A study on the effects of silica particle size and milling time on synthesis of silicon carbide nanoparticles by carbothermic reduction. Int. J. Refract Metal Hard Mater. 29, 645–650 (2011)

Chapter 39

An Experimental Investigation on Horizontal Surface Grinding of Mild Steel Using Different Lubricating Oils Soutrik Bose

and Nabankur Mandal

Abstract Abrasives bonded in a rotary wheel are the prime agents to confiscate material in grinding. Surface finish is a customer-specified quality for machined parts. Various parameters like speed, feed and infeed affect surface finish. In grinding operation, there are many parameters like lubricating oil, quantity of lubricating oil and infeed that are directly or indirectly proportional to the surface finish. In this paper, the surface roughness is measured in a horizontal surface grinding machine with input machining variables as lubricating oils, quantity of lubricating oil and different infeed values. Different experimental results are obtained by using different lubricating oil (engine oil, mustard oil and rapeseed oil), different quantity of lubricating oil (0.08, 0.13, 0.9 ml) and different infeed (10, 20, 30 µm) on the surface finish of the mild steel plate. Through ANOVA, it is observed 52.41% dependency of surface roughness of the mild steel on types of lubricating oil, 7.08% on quantity of lubricating oil, 10.40% on the variation of infeed and 30.11% on machine error. Experimental results are obtained at 10 µm infeed, 0.9 ml quantity of lubricating oil using rapeseed oil, least value of surface roughness (Ra) is obtained than the others. Keywords Optimization · Horizontal surface grinding · Surface roughness · ANOVA · Optimal conditions

39.1 Introduction Grinding is vastly used in industries to eliminate unwanted material to get good quality of surface finish. Lubricants must be applied straight to the cutting zone to ensure that the fluid is not carried away by the wheel. Surface condition, appearance of the finished products and its reliability are the key factors for improved product quality. In 2008, Taghi et al. [1] investigated the influence of ultrasonic vibrations on dry grinding of soft steel, which eradicated the thermal spoils and grinding force was decreased. Shen et al. [2] worked on the grinding of impenetrable CBN wheels under S. Bose (B) · N. Mandal MCKV Institute of Engineering, 243 G.T. Road (N), Liluah, Howrah, West Bengal 711204, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_39

471

472

S. Bose and N. Mandal

minimum quantity lubrication (MQL) method in 2009. Results concluded inadequate cooling problem of MQL grinding with impenetrable CBN wheels. Mane et al. [3] worked on surface finish enhancement of grinding process using compressed air in 2012. Setti et al. [4] worked on MQL method on cutting fluid (nano) to improve machining of Ti–6Al–4V alloy in 2012. Pal et al. [5] worked on Taguchi parametric optimization technique in 2012 on optimizing grinding parameters. Kumar et al. [6] found under high compression of gas, depth of cut increased with decrement in feed rate; hence, material removal rate also increased during grinding of AISI H11 in 2014. Aravind et al. [7] optimized the progressive parameters of grinding for AISI 1035 steel in 2014. Mandal et al. [8] improved grinding performance by introducing pneumatic barricade in 2015. Majumdar et al. [9] experimentally investigated formation of a layer of air around the high velocity rotating grinding wheel in 2016. Saravanakumar et al. [10] analyzed the optimum process parameters like table speed, grinding wheel speed and also depth of cut in surface grinding process in 2018 by using Taguchi method. This paper aims to obtain the minimum surface roughness using different lubricating oils, quantity of lubricating oil and different infeed on mild steel by horizontal grinding machine. The experimental results are obtained by varying the different lubricating oil (engine oil SAE 30, mustard oil and rapeseed oil), quantity of lubricating oil (0.08, 0.13, 0.9 ml) and infeed (10, 20, 30 µm) on the mild steel plate.

39.2 Experimental Setup and Conditions An experimental design methodology is proposed for two basic objectives, number of trials determination and specifications of the conditions for each trial. Input parameters are lubricating oil, quantity of lubricating oil (number of drops or equivalent ml) and infeed (µm) as shown in Table 39.1. Surface roughness is a surface texture measurement, measured in Ra and Rz units. Table 39.1 Factors with values of different level S. No.

Factors

Level 1

Level 2

Level 3

1 2

Lubricating oil

Engine oil (SAE-30)

Mustard oil

Rapeseed oil

Quantity of oil (drop)

3 (0.08 ml)

5 (0.13 ml)

3

7 (0.9 ml)

Depth of cut (µm)

10

20

30

39 An Experimental Investigation on Horizontal Surface Grinding … Table 39.2 Mechanical properties of mild steel

473

S. No.

Name

Metric

1.

Hardness, Brinell

126

2.

Tensile strength, ultimate

440 MPa

3.

Modulus of elasticity (steel)

205 GPa

4.

Poisson’s ratio (steel)

0.290

Fig. 39.1 Workpiece of the MS material

39.2.1 Workpiece Material In this research work, mild steel (MS) is used as workpiece material. MS is costeffective and possesses excellent material properties for a range of applications. Table 39.2 represents the material library of MS. MS has a comparatively low tensile strength, lesser price and simple to manufacture. Besides carbon, MS consists of chromium, manganese, tungsten and vanadium. The various chemical compositions of MS are C [0.14%], Fe [98.81%], Mn [0.60%] and P [= 0 Psi < 0

Phi = 90.00º

1.0862

d-spacing (A)

1.0860 1.0858 1.0856 1.0854 1.0852 1.0850 1.0848 0.0

0.1

0.2

sin2 (Psi)

0.3

0.4

0.5

Fig. 42.6 Measurement of d-spacing at different ‘φ” angles

42.4 Conclusion In this study, an experimental investigation into the end milling of nickel-based superalloy Inconel 625 was conducted in order to analyze the influence of the machining parameters on surface integrity characteristics, such as surface roughness, microhardness, and residual stresses. The surface roughness was analyzed using ANOVA and main effect plot. It was observed from results of analysis of variance that the surface roughness is significantly influenced by feed per tooth, followed by cutting speed. While radial depth of cut and radial rake angle have a smaller influence. The percent contribution of feed per tooth (64.14%), cutting speed (17.15%), and depth of cut (10.38%) in affecting the deviation of surface finish were significantly larger (95% confidence level) as compared to the contribution of the other parameters. It was observed from the main effects plot that the surface roughness increases drastically as feed per tooth increases from 0.05 to 0.17 mm/min. The experimental results revealed that the minimum surface roughness (0.084 μm), residual stress (123.2 Mpa) and maximum micro-hardness (344) are observed at higher cutting speed (90 m/min), positive radial rake angle (13°), lower feed per tooth (0.05 mm/tooth), and lower radial depth of cut (0.2 mm). Based on the experiment analyzed, it was observed that the higher cutting speed, the lowest feed per tooth, and lower radial depth of cut coupled with the use of positive radial rake angle could ensure induction of superior surface integrity in the machined surfaces. Acknowledgements This study would not have been completed without the immense cooperation and help given by National Facility of texture and OIM IIT Bombay, Advance machining centre

514

R. Rajguru et al.

Walchand College of Engineering Sangli, and Salbro engineers, Andheri, Mumbai, and the authors thank them for their support and gesture.

References 1. Hanasaki, S., Fajiwara, J., Touse, M.: Tool wear of coated tools when machining and high nickel alloy. Annu. CIRP 39(1), 77–80 (1990) 2. Davim, J.P.: Surface Integrity in Machining. Springer-Verlag, London (2010) 3. Ren, X.D., Zhan, Q.B., Yang, H.M., Dai, F.Z., Cui, C.Y., Sun, G.F., Ruan, L.: The effects of residual stress on fatigue behavior and crack propagation from laser shock processing-worked hole. Mater. Des. 44, 149–154 (2013) 4. Henriksen, E.K.: Residual stresses in machined surfaces. Trans. ASME 73, 69–74 (1951) 5. Ee, K.C., Dillon, O.W., Jawahir, I.S.: Int. J. Mech. Sci. 47(2005) 1611 6. Guo, Y.B., Li, W., Jawahir, I.S.: Surface integrity characterization and prediction in machining of hardened and difficult-to-machine alloys: a state-of-art research review and analysis. Mach. Sci. Technol. 13, 437–470 7. Ulutan, D., Ozel, T.: Int. J. Mech. Tools Manuf. 51(2011), 250 8. Kundrák, J., Mamalis, A.G., Gyani, K., Bana, V.: Surface layer microhardness changes with highspeed turning of hardened steels. Int. J. Adv. Manuf. 53, 105–112 (2011) 9. Thakur, D.G., Ramamoorthy, B., Vijayaraghavan, L.: Effect of cutting parameters on the degree of work hardening and tool life during high-speed machining of Inconel 718. Int. J. Adv. Manuf. 59, 483–489 (2012) 10. Krolczyk, G., Nieslony, P., Legutko, S.: Microhardness and Surface Integrity in turning process of duplex stainless steel (DSS) for different cutting conditions. J. Mater. Eng. Perform. 23, 859– 866 (2014) 11. Jiang, W., More, A.S, Brown, W.D., Malshe, A.P.: A cBN-TiN composite coating for carbide inserts: Coating characterization and its applications for finish hard turning. Surf. Coat. Tech. 201(2006), 2443–2449 12. Dogra, M., Sharma, V.S., Sachdeva, A., Suri, N.M., Dureja, J.S.: Tool wear, chip formation and workpiece surface issues in CBN hard turning: a review. Int. J. Precis. Eng. Man. 11, 341–358 (2010) 13. Groover, M.P.: Fundamentals of Modern Manufacturing: Materials, Processes, and Systems, 4th edn, pp. 585. Wiley, New York (2010) 14. Production Technology HMT, Ch. 3 Machinability Aspects in Machining, pp. 53–68. Tata McGraw Hill, Bangalore, India (1980) 15. Ezugwu, E.O.: Key improvements in the machining of difficult-to-cut aerospace superalloys. Int. J. Mach. Tools Manuf. 45, 1353–1367 (2005) 16. Dry Machining’s Double Benefit, Machinery and Production Engineering, pp. 14–20 (1994) 17. Ezugwu, E.O., Wang, Z.M., Machado, A.R.: The machinability of nickel-based alloys: a review. J. Mater. Process. Technol. 86, 1–16 (1999) 18. Kaya, Eren, Akyuz, Birol: Effects of cutting parameters on machinability characteristics of Ni-based super alloys: a review, published by De Gruyter Open. Open Eng. 7, 330–342 (2017) 19. Pawade, R.S., Joshi, S.S., Brahmankar, P.K.: Int. J. Mach. Tools Manuf. 48, 15–28 (2008) 20. Bhopale, Nandkumar N., Joshi, Suhas S., Pawade, Raju S.: Experimental investigation into the effect of ball end milling parameters on surface integrity of Inconel 718. J. Mater. Eng. Perform. 24, 986–998 (2015) 21. Moufki, A., Le Coz, G., Dudzinski, D.: End milling of Inconel 718 super alloy—an analytical modeling 22. Singaravel, B., Selvaraj, T.: A review of micro hardness measurement in turning operation, applied mechanics and materials. ISSN: 1662–7482, Vols. 813–814, pp. 274–278. https://doi. org/10.4028/www.scientific.net/AMM.813-814.274

42 Effect of Machining Parameters on Surface Integrity …

515

23. El-Wardany, T.I., Kishawy, H.A., Elbestawi, M.A.: Surface integrity of die material in high speed hard machining, part 2: microhardness variations and residual stresses. J. Manuf. Sci. Eng. 122/633 (2000) 24. Ezugwu, E.O., Pashby, I.R.: High speed milling of nickel-based superalloys. J. Mater. Process. Technol. 33, 429–437 (1992) 25. Phadke, M.S.: Quality Engineering Using Robust Design. Prentice Hall, Englewood Cliffs (1989)

Chapter 43

Tool Wear Behavior in Milling of Hardened Custom 465 Steel V. Prasath , V. Krishnaraj , J. Kanchana and B. Geetha Priyadharshini

Abstract With current advancements in hard-milling technology, tool wear and surface integrity have turn out to be a vital attention to improve machining efficiency and the quality of machined components. The aspiration of this current research paper is to examine tool wear and chip morphology when machining hardened Custom 465 (~55HRC) steel with uncoated, commercial TiAlN, and nanolayer silicon nitride coated inserts. The tests were carried out using constant cutting speed of 150 m/min, feed of 0.1 mm/tooth, and depth of cut of 0.3 mm in dry atmosphere. Wear measurements were made with an optical microscope aided by imaging software. Subsequently after each cutting pass, the milling process was clogged; the tool was detached from the tool holder and carried to the optical microscope where the wear of the VBmax flank wear was measured. Likewise, chips were collected and examined during each cutting pass to study its morphology. Flank wear of 0.5 mm was postulated as a denial standard. Wear investigation was executed using an optical microscope for flank and rake faces. Several wear patterns such as abrasion, flank wear, and unwarranted chipping on the tool edge have been studied. The nanolayer silicon nitride coating offers the finest wear resistance in machining and hence responsible for the extended life with carbide inserts. Keywords Tool wear · Coated inserts · Flank wear · Wear patterns

43.1 Introduction Primarily, hard milling implicates executing a rough or else finished milling operations, subsequently, next to heat treatment processes rather than milling when the steel is in its soft state. The key benefits of hard milling are the capability to produce difficult geometric profiles, better surface roughness, quicker material removal, decreased finishing time, and reduced costs [1]. Yet, hard milling produces substantial tool wear and deviations in product attribute and performance owing to elevated V. Prasath (B) · V. Krishnaraj · J. Kanchana · B. Geetha Priyadharshini PSG College of Technology, Coimbatore, Tamil Nadu 641004, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_43

517

518

V. Prasath et al.

mechanical stress and increased heat formation at the tool–chip boundary. Hence, suitable measures need to be implemented in order to maintain an extended tool life and sustain the superiority of the machined surface. Normally, the tool wear is used as a measure to evaluate the efficiency of the machining process [2]. In hard-milling processes, the tool–chip interface temperature and the stresses have a greater influence at the cutting edge which in turn determines the tool life [3]. Carbide inserts remained predominantly suitable in these machining processes due to their remarkable properties for instance, high hardness and hot hardness. However, premature wear of the carbide inserts in machining hardened work materials limits using those tools in hard-machining practices. The solitary key for the use of carbide inserts in hard milling is to shield them through an appropriate hard coating [4]. High hardness, low thermal conductivity, and high adhesion to the substrate stood important for wear-resistant coatings because these coatings want to sustain substantial stresses and high-temperature working environments in cutting [5, 6]. Recent tough physical vapor deposition (PVD) coatings improve the multi-functionality of the cutting tool insert [7]. Among, nanolayer (nl) coatings own improved mechanical properties than multilayer- and singlelayer-coatings owing to the consequence of large number of interfaces on surface crack propagation [8, 9]. Similarly, the arrangement of altered materials in layered constructions produces a coating near to good mechanical properties. For instance, Bouzakis et al. [5] located that adding silicon (Si) in titanium nitride (TiN) structure avoids the creation of a columnar structure and enables in formation of a dense nanocomposite structure (nc). Titanium silicon nitride (TiSiN) coatings exhibit improved mechanical strength and wear resistance properties on ambient and preeminent temperatures. However, the overall performance of TiSiN-coated inserts throughout machining was reduced owing to their higher stresses at machining interfaces and poor adhesion among coating and the substrate interfaces, on the other hand, TiAlSiN coatings accomplished better cutting performance, enhanced hardness, and oxidation resistance. Chen and their research team [10] altered the coating in multilayer TiAlN/TiAlSiN structure toward increasing the adhesion and hardness which consecutively enhanced the milling performance. An upsurge of tool life was attained through milling process using TiAlN/TiAlSiN multilayer-coated inserts. Recently, established nl-TiAlN/Si3 N4 tough coatings remained smooth and chemically stable at an extreme application temperature of 1100 °C. Despite, research investigations on the performance evaluation of nl-TiAlN/Si3 N4 -coated inserts in dry hard-milling conditions were found deficient. So, this research studies the influence of uncoated, commercial TiN/TiAlN, and nl -Si3 N4 -coated inserts with constant cutting parameters such as cutting speed (150 m/min), feed (0.2 mm/min), and depth of cut (0.3 mm) on tool wear behavior and chip morphology were examined throughout end milling of hardened Custom 465 steel (~56 HRC).

43 Tool Wear Behavior in Milling of Hardened Custom 465 Steel

519

Fig. 43.1 Layer structure of coatings on inserts. a Uncoated, b commercial TiAlN, c nl -Si3 N4

43.2 Experimental Setup 43.2.1 Substrate, Coatings, and Characterization The silicon nitride (Si3 N4 ) thin films of size 1 µm were superimposed on a commercial TiAlN-coated inserts and silicon substrates (100) using radio frequency (RF) magnetron sputtering technique using a 3 inch diameter silicon nitride target of purity >99.5 wt% with 0.125 in copper backing plate in argon (99.995%) and nitrogen (99.95%) atmosphere. The flow rate of argon and nitrogen was kept constant by 50 and 10 sccm, RF power in the range of 300 W, and deposition time of 2.5 h. Before introducing the gas inside, the chamber was pumped to a high vacuum pressure of 2 × 10−6 Mbar using a turbo molecular pump with a maximum pumping speed of >500 L/min and the rotational speed of 31,200 RPM. The argon and nitrogen gases were led inward the chamber using mass flow controllers, and working pressure was kept at 20 mTorr with the help of the throttle valve. The substrate temperature was kept maintained at 300 °C. During the deposition process, samples were rotated at 20 RPM for uniform deposition. Film thicknesses were experimentally confirmed by Bruker’s Dektak XT benchtop stylus profilometer. Figure 43.1 shows the layer structure of coatings on inserts (a) Uncoated, (b) Commercial TiAlN, and (c) nl -Si3 N4 . The surface topography of the coated samples was studied using atomic force microscopy (AFM, NTMDT, Ireland). Elemental composition of the samples was conducted within the SEM, CARL ZEISS, Germany.

43.2.2 Workpiece Material Hardened Custom 465 steel workpiece material was used in the machining test. Figure 43.2 shows the microstructure of hardened Custom 465 steel. From the observations, it has lath martensitic structure, which is a precise rigid form of steel crystallike structure. When noticed in crosssection, the crystal grains have an acicular structure (needle shaped) which leads to brittle behavior of the material which in turn makes the machinability poor. The chemical composition of hardened Custom 465 steel is given in Table 43.1. The hardness of hardened Custom 465 steel materials is 550–565 VHN at 1 kg load.

520

V. Prasath et al.

Fig. 43.2 Microstructure of hardened custom 465 steel

Table 43.1 Chemical composition of hardened custom 465 steel work material (wt%) C

Cr

Ni

Ti

Fe

Mo

V

0.015

11.23

13.95

1.775

71.56

0.929

0.018

43.2.3 Cutters and Machine Tool The cutting tool used in the machining test was standard 16 mm Kennametal 12A01R020A16ED10 end milling cutter with three different sets of cemented carbide inserts such as uncoated, commercial TiAlN (K313), and superimposed Si3 N4 nanolayer were used for the experimentation. MAKINO S33 three-axis vertical machining center focus with an extreme spindle speed of 20,000 rpm and cutting feed rate of 40,000 mm/min with a PC-based NC controller machine tool were used in this research. The work material was mounted in particularly designed fixture. The investigational setup is shown in Fig. 43.3.

43.2.4 Experimental Procedure The machining investigations were conducted with a constant cutting speed of 150 m/min, feed of 0.1 mm/rev, and depth of cut of 0.3 mm as an input response. Wear measurements were executed using a Dino-Lite transportable optical microscope. After each cutting pass, milling process was clogged; cutting tools existed were detached from the tool holder and engaged to the optical microscope and flank wear width on the cutting tool inserts were inspected. As per the standards, flank wear 0.5 mm was postulated as the norm for tool elimination. The machining will

43 Tool Wear Behavior in Milling of Hardened Custom 465 Steel

521

Fig. 43.3 Experimental setup

be stopped when VBmax reached 0.6 mm. Similarly, chip morphology analysis was performed using Dino-Lite portable optical microscope aided with imaging software.

43.3 Results and Discussions 43.3.1 Coating Characterization Figure 43.4, elucidates the topography structure of Si3 N4 nanolayer for a deposition time of 2.5 h. A dense uniform columnar film structure was observed owing to the unsystematic angle incident of the incoming atoms due to the unchanging rotation of the substrate. From Table 43.2, the percentages of silicon and nitrogen on inserts were observed. The silicon to nitrogen atomic percentage was in the ratio of 75:25.

43.3.2 Tool Wear and Wear Mechanism Figure 43.5 describes the variation of wear depth VB in conjunction with the cutting time T. Among the variations, tool wear during the milling process can be subdivided into early wear, regular wear, and serious wear. At the beginning of the wear time (0–0.5 min), the cutting tool is quickly damaged owing to enormous irregularities of the tool surfaces and the tool is damaged rapidly in short time at the beginning of cutting process.

522

V. Prasath et al.

Fig. 43.4 Atomic force microscopy topography of Si3 N4 coatings on insert

Table 43.2 Elemental composition of the Si3 N4 coating Fig. 43.5 Tool life and optical microscope images of the worn insert during machining of hardened Custom 465 steel

Coating material

Silicon at %

Nitrogen at %

Si3 N4

74.90

25.09

43 Tool Wear Behavior in Milling of Hardened Custom 465 Steel

523

Fig. 43.6 Flank wear a insert before machining, b uncoated, c TiAlN, d nl- Si3 N4

As soon as the depth of the wear slowly increases from 0.15 to 0.40 mm with a small percentage of wear, the wear will shift during the regular wear period (0.5– 2 min). Meanwhile, owing to the presence of irregular surfaces they were scraped off due to their lower wear resistance and working pressures on tool faces were reduced and become constant. As a result, the tool has been slowly damaged, which shortens the effective machining time of the tool. Later, in the phase of intense wear, the friction force and the cutting temperature increases, which causes wear or burn of the tool and significantly increases the depth of wear of the flank face. In addition to its limitation, the wear patterns of the tool are mainly exposed to wear on the rake surface and the flank surface. Due to the practice of complex cutting, several wear models have acted simultaneously. During this cutting process, considerable friction has occurred between the cutting surface of the tool and the chips, as well as the major flank surface and the machined surfaces, creating very high contact pressures and high temperatures. Small chips were fragmented from the cutting edge of the uncoated insert (Fig. 43.6b) due to early wear or edge fragmentation caused mainly by intermittent cutting and abrupt shocks. Breakage or fragment of the cutting edge is the result of excessive mechanical tensile stress. These stresses can be due to various causes, such as chip breaking, higher depth or feed, additional edge formation (build-up edge), vibration or extreme wear of the cutting tool inserts. Crater wear is another type of pragmatic wear limited to the rake surface of the TiAlN insert (Fig. 43.6c). This is due to a chemical reaction between working material and cutting tool and is increased further by the cutting speed. Extreme crater wear damages the cutting edge and can lead to breakage. The nl -Si3 N4 -coated cutting tool inserts extended tool life by about 3.2 min, but flank wear was noted. This wear is the most commonly desired wear and is preferred because it provides a consistent and predictable life. This wear of the flanks is triggered due to abrasion by hard components of the working material.

43.3.3 Chip Morphology The perceptive of chip formation is of great importance for the optimization of machining processes and the integrity of the surface. Therefore, the chips collected in the sequence of cutting tests were examined for their shapes and texture. Figure 43.7

524

V. Prasath et al.

Fig. 43.7 Chip morphology a uncoated, b TiAlN, c nl -Si3N4

shows the chip images obtained besides the first passes. The shape and geometry of the chip formed during machining using uncoated TiAlN, n1 -Si3 N4 cutting tool insert were completely different (Fig. 43.7a–c). Short ribbon type chips were observed (Fig. 43.7a) for uncoated inserts. The uncoated tool contributed blunted chips owing to the early tool wear. Saw-tooth chips were developed in commercial TiAlN inserts (Fig. 43.7b). Short tubular chips were observed for nl -Si3 N4 inserts (Fig. 43.7c). During machining, the chip color turns to golden yellow when using uncoated inserts because most of the heat engendered at the tool–chip boundary was shifted to the cutting insert as a result of the indigent thermal conductivity of the work material. When using TiAlN inserts, the chip color turns to pale yellow because most of the heat generated at the interface was carried out by chips and not transferred to the inserts. In the case of nl -Si3 N4 , the chip color turns to a mixture of green and blue in nature because the silicon nitride layer acts as an excellent thermal barrier protecting the inserts while transferring the most of the heat generated via chips. After the nl -Si3 N4 -coated tools had damaged out, the chips worsened and blunt saw-tooth chips were stayed.

43.4 Conclusions Through cutting experiments using an uncoated, commercial TiAlN and nl -Si3 N4 inserts to mill Custom 465 steel, the subsequent inferences are drawn. Once the Custom 465 steel is machined using an uncoated, commercial TiAlN, and nl -Si3 N4 inserts, edge chipping was observed for uncoated inserts due to the interrupted cutting and sudden shock on the inserts. Crater wear is another wear type observed confined to a small area to the rake side of the TiAlN insert which quickens tool wear. The nl -Si3 N4 -coated inserts have produced longest tool life but flank wear was observed, these are the utmost typical type of wear and desired wear type, confirming anticipated and unchanging tool life. The shape and geometry of the chip formed during machining with uncoated, TiAlN, nl -Si3 N4 inserts were significantly different in morphology and color changes were distinct. Thus, the nanolayer silicon nitride coating provides the finest wear resistance in machining and as a result delivers the extended life with carbide inserts.

43 Tool Wear Behavior in Milling of Hardened Custom 465 Steel

525

Acknowledgements We would like to express our thanks to DRDL, Hyderabad for funding this project (Grant No. 42141/2015). We also acknowledge the faculty members of Center for advanced tooling and precision dies and PSG Institute of Advanced Studies, Coimbatore.

References 1. Takacs, M., Farkas, B.Z.: Hard cutting of AISI D2 steel, No. 176, pp. 1–7 (2014) 2. Lajis, M.A., Amin, A.K.M.N., Karim,A.N.M., Hafiz, A.M.K.: Preheating in end milling of AISI D2 hardened steel with coated carbide inserts (2008) 3. Ning, L., Veldhuis, S.C., Yamamoto, K.: Investigation of wear behavior and chip formation for cutting tools with nano-multilayered TiAlCrN/NbN PVD coating. Int. J. Mach. Tools Manuf. 48(6), 656–665 (2008) 4. Gopalsamy, B.M., Mondal, B.: Investigations on hard machining of Impax Hi Hard tool steel, pp. 145–165 (2009) 5. Bouzakis, K.D., Michailidis, N., Skordaris, G., Bouzakis, E., Biermann, D., M’Saoubi, R.: Cutting with coated tools: coating technologies, characterization methods and performance optimization. CIRP Ann. Manuf. Technol. 61(2), 703–723 (2012) 6. Fox-rabinovich, G.S., Yamamoto, K., Beake, B.D., Kovalev, A.I., Aguirre, M.H., Veldhuis, S.C.: Surface & coatings technology emergent behavior of nano-multilayered coatings during dry high-speed machining of hardened tool steels. Surf. Coat. Technol. 204(21–22), 3425–3435 (2010) 7. Fox-Rabinovich, G.S., Yamamoto, K., Beake, B.D.: Hierarchical adaptive nanostructured PVD coatings for extreme tribological applications: the quest for nonequilibrium states and emergent behavior, Vol. 043001 (2012) 8. Ducros, C., Cayron, C., Sanchette, F.: Multilayered and nanolayered hard nitride thin films deposited by cathodic arc evaporation. Part 1 : deposition, morphology and microstructure, Vol. 201, pp. 136–142 (2006) 9. Yashar, P.C., Sproul, W.D.: Nanometer scale multilayered hard coatings, Vol. 55, pp. 179–190 (1999) 10. Liao, Y.S., Lin, H.M., Chen, Y.C.: Feasibility study of the minimum quantity lubrication in high-speed end milling of NAK80 hardened steel by coated carbide tool. Int. J. Mach. Tools Manuf 47(11), 1667–1676 (2007)

Chapter 44

Experimental Study on Machining of EN24 Using Minimum Quantity Lubrication Gaurav Tyagi , J. Bhaskar, S. K. Singhal and G. Bartarya

Abstract Machining under environment of cutting fluids has always been preferred due to various advantages. This is more relevant while machining hardened materials. But excessive use of cutting fluids is not good due to machining cost and its negative issues in the environment nearby the machine tool. All these factors inspire for minimal use of cutting fluids whenever possible. In view of this, an experimental effort has been made to study the machining performance while machining hardened steel workpiece under the condition of minimum quantity lubrication. It has been observed that minimum quantity lubrication reduces tool wear and increases surface finish achieved during hard turning. Keywords Hard turning · Minimum quantity lubrication · Surface roughness · Tool wear

44.1 Introduction Nowadays, hard turning technology is preferred over traditional grinding operations because of its significant improvement in the quality of machined surfaces at high material removal rate. Hard turning can possibly facilitate low process cost, low process time, better surface quality, and lower waste. The turning of materials having hardness value in the range between 46 and 69HRC is termed as hard turning. This G. Tyagi · J. Bhaskar (B) · S. K. Singhal Department of Mechanical Engineering, Harcourt Butler Technical University, Kanpur 208002, India e-mail: [email protected] G. Tyagi e-mail: [email protected] S. K. Singhal e-mail: [email protected] G. Bartarya School of Mechanical Sciences, Indian Institute of Technology, Bhubaneswar 752050, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_44

527

528

G. Tyagi et al.

machining process exhibits some unique behavior and includes hardened workpiece, cutting tools, mechanism of chip formation. Grinding produces good surface finish at high feed rates but hard turning produces better surface finish at higher material removal rates [1]. In hard turning, the selection of process parameters is based on hardness of material, roughness of material and tool wear, etc. The machining cost, machining time, tool wear, and surface roughness are decreased as the hardness of the material is increased. The process parameter factor such as feed rate, speed, cutting time, cutting fluid, workpiece hardness, depth of cut and tool geometry affects surface finish and tool wear. Alternative methods to apply cutting fluids, such as flood cooling, close to dry machining, cryogenic cooling, and lubricant, have been developed in recent years [2]. Issues related to environment, health, and manufacturing cost suggest minimal use of cutting fluids whenever possible. Efforts have been made to develop methods to reduce quantity of lubrication while machining [3]. Minimal quantity lubrication (MQL), which sprays small amount of cutting fluid (in the range of approximately 10–100 mL/h) to the cutting zone area with the aid of compressed air, is an alternative for this hazard. In many machining operations, minimum quantity cooling lubrication (MQL) is the key to successful dry machining. According to Rahim et al. [4], the sustainable manufacturing introduced various machining conditions, such as dry, near dry, also called minimum quantity lubrication (MQL) and cryogenic machining. Researcher have focused to optimize the effects of cutting parameters on surface finish and MRR of EN24/AISI4340 work material [1]. Results showed that nose radius, depth of cut, feed rate, cutting speed, and coolant condition affect the material removal rate and also affect the surface roughness produced. The effect of temperature was studied on insert, workpiece, and chip by Mohandas et al. [5]. Experimental work indicated that tool temperature produced due to shearing action of the tool on the workpiece surface has very less effect on surface roughness of the hard chrome plating. Nose radius and cutting speed play an important role in achieve lowest surface roughness, and feed, depth of cut play least role [1]. Flank wear was studied using multilayer-coated carbide and mixed ceramic inserts. Constant depth of cut and speed along with feed increases flank wear. Higher flank wear was produced on coated carbide tool in comparison with ceramic insert [6]. The effect of cryogenic cooling was studied experimentally and found that the temperature was significantly reduced in the cutting zone [7]. The preferable tool material used for machining of these materials is CBN tool. CBN has hardness and wear resistance to a high coefficient of thermal conductivity, good chemical stability, and high hot hardness. Sahoo et al. [8] investigated that coated insert shows better performance than an uncoated insert. Titanium based hard thin films are mostly used for coating due to better wear resistance. Dhar et al. [9] experimentally found that MQL decreases the cutting temperature, dimensional inaccuracy on the basis of the levels of the cutting velocity, and feed rate. A review of the works done in the field of hard turning [10] provides the basis of selection of cutting parameters for efficient machining of hard steels. Three parameters such as speed, feed, and depth of cut have been selected for experimental study.

44 Experimental Study on Machining of EN24 …

529

44.2 Experimental Details 44.2.1 Selection of Material EN24 is a medium carbon low alloy steel. It finds its typical applications in the manufacture of automobile and machine tool parts, low specific heat and tendency to strain harden, and diffuse between tool and work material. Problems like large cutting forces arise while machining on EN24.

44.2.2 Selection of Cutting Tool Cutting tools used for hard turning require high hardness, high compressive strength, and high resistance to abrasive wear, thermal resistance, and chemical stability at elevated temperatures. Various cutting inserts are available such as cubic boron nitride (CBN), ceramic, and cerements. Carbides have very low tool life, whereas CBN exhibits maximum tool life for hard turning operation. It was decided to use low-costcoated carbide inserts in the present work and to validate their utility in hard turning under MQL conditions. Chamfered edge-coated carbide insert (type: CNMG120408HM) was selected for machining of hardened EN24 steel rod as chamfered edge tools have a better tool life. A tool holder (Seco make) CCLNR2525M12 was employed (Fig. 44.1).

44.2.3 Design of Experiments A 15-run experiment was designed based on Box–Behnken method in which the input variables such as cutting velocity, feed, depth of cut were varied at three levels (low, medium, and high). Box–Behnken design of experiments was considered because it provides lesser number of experiments when compared with Taguchi’s method or a full factorial design (Table 44.1).

Fig. 44.1 a Coated carbide insert and b tool holder

530 Table 44.1 Range of velocity, feed, and depth of cut

Table 44.2 Cutting parameter of Box–Behnken design

G. Tyagi et al. Parameter

L.L.

M.L.

H.L.

Velocity (m/min.)

100

130

160

Feed (mm/rev.)

0.08

0.12

0.16

Depth of cut (mm)

0.2

0.3

0.4

Run

Velocity (m/min)

Feed (mm/rev)

Depth of cut (mm)

1

100

0.08

0.3

2

100

0.16

0.3

3

160

0.08

0.3

4

160

0.16

0.3

5

100

0.12

0.2

6

100

0.12

0.4

7

160

0.12

0.2

8

160

0.12

0.4

9

130

0.08

0.2

10

130

0.08

0.4

11

130

0.16

0.2

12

130

0.16

0.4

13

130

0.12

0.3

14

130

0.12

0.3

15

130

0.12

0.3

The ranges of velocity, feed, and depth of cut (shown in Table 44.1) were selected based on lathe machine capability, tool manufacturer criteria, and surface finish parameters according to the literature. Table 44.2 shows different machining experiment conditions as suggetsed by Box-Behnken design.

44.2.4 Working Conditions and Experimental Setup In MQL condition, workpiece was turned using very low flow rate of cutting fluid. Figure 44.2 demonstrates the experimental setup introduced on CNC lathe machine to use amid MQL. In Fig. 44.2, a compact compressor of limit 50 L was associated with inlet of flow control unit by PU pipe. In flow control unit, a regulating valve was utilized to control the amount of compressed air according to the requirement. Another PU pipe was associated with the outlet of flow control unit. A pressure gauge was associated just before nozzle keeping in mind to check pressure of compressed air at the exit. Cutting fluid was put in a bottle which was associated with a nozzle.

44 Experimental Study on Machining of EN24 …

531

Fig. 44.2 Setup of MQL in on a CNC lathe machine

Angle of nozzle was kept at 90° with the axis of workpiece. The quantity of cutting fluid was control by regulating knob. The proportion of oil (servo cut S water-based cutting oil) to water was kept at 1:20 ratio.

44.2.5 Hard Turning with MQL EN24 hardened steel was turned under two conditions which were dry cutting and MQL cutting. In dry cutting, material was turned without use of any cutting fluid, whereas in MQL cutting, material was turned with the use of cutting fluid. After the complete setup in CNC lathe machine, hardened steel was machined using MQL. The compressed air which was obtained using electric-powered air compressor was adjusted to 4 bar pressure using the pressure regulator and was kept constant throughout the experiment. A cutting fluid was also supplied simultaneously to the nozzle.

44.2.6 Surface Integrity Test After the turning of hardened EN24 steel with and without MQL, various experimental tests were made to check the integrity of machined surface. These tests are:

44.2.6.1

Surface Roughness

Taylor Hobson made Surtronic 3P instrument that was used to measure the roughness. Roughness was measured for each workpiece at three different points. Then average

532

G. Tyagi et al.

was calculated of three points and recorded as Ra value. Range and cutoff set in Surtronic 3P instrument during measurement are 99.99.

44.2.6.2

Tool Wear Measurement

A USB-connected Dino-lite optical microscope with Dino-capture 2.0 software, available at Manufacturing Science Lab of IIT Kanpur, was used to measure the tool flank wear. It has the freedom to magnify from 20× to 230×.

44.3 Results and Discussion The hard turning experiments were performed as per the experimental plan given in Table 44.2. The results have been analyzed though regression analysis and response surface methodology, and effect of the cutting parameters, i.e., speed, feed, and depth of cut, on the surface roughness and flank wear have been analyzed for hard turning with and without MQL.

44.3.1 Surface Roughness Figure 44.3 has been plotted for better understanding of variation of surface roughness with and without MQL for same cutting conditions prevailing for various test runs. As surface roughness achieved with MQL is found better than that achieved without using MQL, thus it may be said that use of MQL for hard turning of steel is certainly advantageous. Regression analysis was applied to study the effect of cutting Fig. 44.3 Comparison of surface roughness between MQL and without MQL

1.6

Roughness without Lubrication

Roughness (μm)

1.4

Roughness with Lubrication

1.2 1 0.8 0.6 0.4 0.2 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 15

Number of Runs

44 Experimental Study on Machining of EN24 …

533

conditions on the surface roughness work. The response surface equation in second order for three factors is given by: Ra = a11 v2 + a22 f 2 + a33 d 2 + a12 f v + a23 f d + a13 dv + a1 v + a2 f + a3 d + a0 where Ra is desired response and a0 , a1 , a2 , a3 , a11 , a22 , a33 , a12 , a23 , a13 are coefficients. Surface roughness equation when machining was performed without MQL: Ra = 0.000151852 ∗ v ∗ v − 5.20833 ∗ f ∗ f − 24.0833 ∗ d ∗ d + 0.0166667 ∗ v ∗ f + 16.875 ∗ f ∗ d − 0.0108333 ∗ v ∗ d − 0.0313565 ∗ v − 4.69792 ∗ f + 13.4583 ∗ d + 0.63046 Surface roughness equation when machining was performed with MQL: Ra = 0.000143056 ∗ v ∗ v − 17.9687 ∗ f ∗ f − 22.875 ∗ d ∗ d + 0.0208333 ∗ v ∗ f + 21.875 ∗ f ∗ d − 0.015 ∗ d ∗ v − 0.0283611 ∗ v − 4.11458 ∗ f + 12.7125 ∗ d + 0.479306 Response surface graphs generated with the help of MATLAB are shown in Figs. 44.4, 44.5, and 44.6. The graphs show the trend of change in surface roughness as the cutting velocity and the feed are varied keeping the depth of cut constant for machining with MQL and without MQL. From the graphs plotted at different depth of cut, for both MQL and without MQL conditions, it could be observed that: 1. Surface roughness increases with cutting velocity for both cutting conditions. 2. On increasing feed, surface roughness increases in both with MQL and without MQL cases. at DOC=.2 without MQL

(a)

0.95

at DOC=.2 with MQL

(b)

0.9

Roughness (µm)

1

0.85

0.9 0.8 0.8 0.75

0.7

0.7

0.6

0.65

0.5

0.6

0.4 0.16

0.55 0.14

160 140

0.12 120

0.1

Feed (mm/rev)

0.08 100

Velocity (m/min)

Roughness (µm)

0.9 1

0.8

0.8 0.7 0.6 0.6 0.4 0.5

0.2 0.16 0.14

0.5 0.45

160 140

0.12

Feed (mm/rev)

120

0.1 0.08 100

Velocity (m/min)

Fig. 44.4 Surface roughness at 0.2 mm depth of cut, a without MQL, b with MQL

0.4

534

G. Tyagi et al. at DOC=.3 without MQL

(a)

at DOC=.3 with MQL

(b)

1.2

1.1 1.15 1.05

1.05

1.2

Roughness (µm)

Roughness (µm)

1.1 1.3

1

1.1 1

0.95

0.9 0.9

0.8 0.7

0.85

160

0.14

0.9

1 0.85

0.9 0.8

0.8

0.7 0.16

0.75

0.14

140

0.12

Feed (mm/rev)

0.95

1.2 1.1

0.75

0.8

0.16

1

1.3

120

0.1 0.08 100

160 140

0.12

0.7

Feed (mm/rev)

Velocity (m/min)

0.1 0.08 100

120 Velocity (m/min)

0.7 0.65

Fig. 44.5 Surface roughness at 0.3 mm depth of cut, a without MQL, b with MQL at DOC=.4 without MQL

(a)

at DOC=.4 with MQL

(b) 0.9

0.7

0.6 0.4

0.6

0.2 0.16

0.5 0.14

160 140

0.12

Feed (mm/rev)

120

0.1 0.08 100

Velocity (m/min)

0.8

1

0.8 0.8

Roughness (µm)

Roughness (µm)

1

0.8

0.7

0.6

0.6

0.4 0.5 0.2 0.16 0.14

0.4

160 0.12

Feed (mm/rev) 0.1

0.4

140 120 0.08 100

0.3

Velocity (m/min)

Fig. 44.6 Surface roughness at 0.4 mm depth of cut, a without MQL, b with MQL

3. On increasing the cutting speed beyond a certain limit, surface roughness start increasing quickly when cutting without MQL. The reason could be given that as the cutting speed increases, surface temperature of workpiece increases which makes it soft. This is called thermal softening. Turning of a softer steel gives rougher surface than hardened steel. 4. There is best improved surface roughness 23.68% at 130 m/s velocity, 0.16 mm/rev feed, 0.4 mm depth of cut in Experiment No. 12. 5. From graph 44.4, 44.5, 44.6 on increasing doc, first surface roughness increases at 0.3 mm after that decrease at 0.4 mm in both with and without MQL.

44 Experimental Study on Machining of EN24 … Fig. 44.7 Comparison of surface roughness between MQL and without MQL

535

0.7

Flank wear (mm)

0.6 0.5 0.4 0.3 0.2 0.1 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 15

Number of Runs Flank wear without Lubrication

Flank wear with Lubrication

44.3.2 Tool Wear Figure 44.7 has been plotted in for better understanding of variation of flank wear with and without MQL.teh comparison again shows that application of MQL reduces the tool wear generated during each test run. The regression condition for flank wear while machining was performed without MQL is as below: FW = 0.00000537037 ∗ v ∗ v−4.94792 ∗ f ∗ f −2.69167 ∗ d ∗ d − 0.004375 ∗ v ∗ f + 7.25 ∗ f ∗ d − 0.0164167 ∗ d ∗ v + 0.00807037 ∗ v + 0.478125 ∗ f + 3.05292 ∗ d − 0.863824 Flank wear equation when machining was performed with MQL is: FW = 0.0000733796 ∗ v ∗ v − 6.84896 ∗ f ∗ f − 0.420833 ∗ d ∗ d − 0.0160417 ∗ v ∗ f − 1 ∗ f ∗ d − 0.0205 ∗ d ∗ v − 0.0077412 ∗ v + 5.45104 ∗ f + 3.08 ∗ d − 0.327968 The response surface graphs are shown in Figs. 44.8, 44.9, and 44.10. The graphs clearly show the effect of the cutting velocity and feed on the flank wear observed keeping the depth of cut constant for both MQL and without MQL conditions. From the graphs plotted at different depth of cut, for both MQL and without MQL conditions, it could be observed that: 1. Tool wear starts increasing with cutting velocity for both cutting conditions. 2. At maximum feed rate, tool wear initially shows decreasing trend and later shows increasing trend at 0.4 mm depth of cut with MQL. 3. On increasing feed, tool wear gets increasing.

536

G. Tyagi et al. At DOC=.2mm without MQL

(a)

AT DOC=.2mm with MQL

(b) 0.6

0.45

0.55

0.8

0.5

0.6

0.4

0.6

Tool wear (mm)

Tool wear (mm)

0.7

0.5

0.5 0.45 0.4 0.4

0.3 0.2 0.16

0.35

0.4

0.3

0.2

0.25 0.2

0

0.35

0.16 160

0.14 0.12

Feed (mm/rev) 0.1

140

0.12

Feed (mm/rev)

120 0.08 100

160

0.14

0.3

140

0.1

120

0.1 0.08 100

Velocity (m/min)

0.15

Velocity (m/min)

Fig. 44.8 Tool wear at 0.2 mm depth of cut, a without MQL, b with MQL At DOC=.3mm without MQL

(a)

AT DOC=.3mm with MQL

(b)

0.45

0.65

0.55

Tool wear (mm)

Tool wear (mm)

0.6

0.6 0.55 0.5

0.5 0.45

0.45

0.4

0.5

0.4

0.4

0.35

0.3

0.3

0.2 0.25

0.35 0.1 0.16

0.4

0.16 0.14

0.14

160 0.12

140

Feed (mm/rev) 0.1

120 0.08 100

160 0.12

0.35

Feed (mm/rev) 0.1

120 0.08 100

Velocity (m/min)

0.2

140 0.15

Velocity (m/min)

Fig. 44.9 Tool wear at 0.3 mm depth of cut, a without MQL, b with MQL At DOC=.4mm without MQL

(a)

(b)

0.6

AT DOC=.4mm with MQL 0.38 0.36

0.55 0.4

0.6 0.55

0.5

0.5 0.45 0.45

0.4 0.35

Tool wear (mm)

Tool wear (mm)

0.65

0.34 0.35

0.32

0.3

0.3

0.25

0.28

0.2

0.26

0.4

0.16

0.24 0.16

0.14

160 0.12

140

Feed (mm/rev) 0.1

120 0.08 100

Velocity (m/min)

0.14 0.35

160 0.12

Feed (mm/rev)

140 0.1

120 0.08 100

Fig. 44.10 Tool wear at 0.4 mm depth of cut, a without MQL, b with MQL

Velocity (m/min)

0.22 0.2

44 Experimental Study on Machining of EN24 …

537

4. The most improved tool wear value (reduction by 62.16%) is found at 100 m/s cutting velocity, 0.12 mm/rev feed, 0.2 mm depth of cut in Experiment No. 5. In experiment No. 15 achieved tool wear reduces by 55.47% for MQL case. But in case of similar experiment No. 13 and 14 tool wear is different due to hardening of work piece material. Also, the experiment No. 9 achieved good tool wear reduction of 53.34% for machining with MQL at 130 m/s velocity, 0.08 mm/rev feed, 0.2 mm depth of cut.

44.4 Conclusions Using hard turning of CNC lathe machine, the results and analysis of the machinability study of EN24 steel with coated carbide inserts with and without minimum quantity lubrication, the following conclusions could be made. 1. Turning of hardened steel with MQL provides better surface roughness when compared with turning without MQL. 2. On increasing the feed, surface roughness increases for both MQL and without MQL cases. 3. Tool wear reduces when turning of hardened steel with MQL than turning without MQL. 4. On increasing the feed, the tool wear increases for both MQL and non-MQL cases. So, an overall remark from the work can be made that the minimum quantity lubrication would work well for hard turning of steels as it improves surface finish and reduces flank wear. This also shows that the coated carbide inserts can be effectively used for hard turning with MQL and may become economical alternative to the costly CBN tools.

References 1. Khan, P.L., Bhivsane, S.V.: Experimental analysis and investigation of machining parameters in finish hard turning of AISI 4340 steel. In: 2nd International Conference on Materials Manufacturing and Design Engineering, pp 265–270 (2018) 2. Korat, M., Agarwal, N.: Optimization of different machining parameters of En24 alloy steel in CNC turning by use of Taguchi method. Int. J. Eng. Res. Appl. 2(5), 160–164 (2012) 3. Mittal, D., Garg, M.P., Khanna, R.: An investigation of the effect of process parameters on MRR in turning of pure titanium (Grade-2). Int. J. Eng. Sci. Technol. 3 (2011) 4. Rahim, E.A., Ibrahim, M.R., Rahim, A.A., Aziz, S., Mohid, Z.: Experimental investigation of minimum quantity lubrication (MQL) as a sustainable cooling technique. 12th Global Conf. Sustain. Manuf. Proced. CIRP 26, 351–354 (2015) 5. Mohandas, K.N., Ramesh, C.S., Eshwara Prasad, K., Balashanmugam, N.: Study of the surface integrity and heat measurement of hard turning of hard chrome coated EN24 substrate. Int. Conf. Adv. Manuf. Mater. Eng. Proced. Mater. Sci. 5, 1947–1956 (2014)

538

G. Tyagi et al.

6. Panda, A., Sahoo, A.K., Rout, A.K., Kumar, R., Das, R.K.: Investigation of flank wear in hard turning of AISI 52100 grade steel using multilayer coated carbide and mixed ceramic inserts. 2nd Int. Conf. Mater. Manuf. Des. Eng. 20, 365–371 (2018) 7. Aramcharoen, A., Chuan, S.K.: An experimental investigation on cryogenic milling of Inconel 718 and its sustainability assessment. Singapore Inst. Manuf. Technol. 14, 529–534 (2014) 8. Sahoo, A.K., Sahoo, B.: Experiment investigation on machinability aspect in hard turning of AISI 4340 steel using uncoated and multilayer coated carbide insert. Measurement 45, 2153– 2165 (2012) 9. Dhar, N.R., Islam, M.W., Islam, S., Mithu, M.A.H.: The influence of minimum quantity of lubrication (MQL) on cutting temperature, chip and dimensional accuracy in turning AISI1040 steel. J. Mater. Process. Technol. 171, 93–99 (2006) 10. Bartarya, G., Choudhury, S.K.: State of the art in hard turning. Int. J. Mach. Tools Manuf. 53, 1–14 (2012)

Chapter 45

Investigative Study of Temperature Produced During Turning Operation Using MQL and Solid Lubricants Anand S. Patel , Mayur A. Makhesana

and K. M. Patel

Abstract Temperature produced during machining plays a very important role as it affects tool wear and leading to variation of surface roughness. Conventional metalcutting fluids serve the purpose of reducing the cutting temperature, but, on the other side, it also affect worker’s health and create environmental issues. Hence, in the present work, effort has been made to use MQL and solid lubricants as promising options compared to metal-cutting fluid to analyze its effect on chip–tool interface temperature as output response. Tool–work thermocouple setup is developed and calibrated for the measurement of chip–tool interface temperature. Graphite and MoS2 are applied as solid lubricants, and process performance is measured in the form of power consumption and chip–tool interface temperature and compared with that of dry machining. Experimental results are compared with simulated results to check the adequacy of developed relations. Results showed effectiveness of solid-lubricantassisted MQL and can be considered as sustainable alternative for lubrication. Keywords Minimum quantity lubrication · Graphite · Molybdenum disulfide

45.1 Introduction Machining is a process in which cutting tools are used to remove material from the workpiece. Removal of material is in the form of chip and to form the chip from material, relative motion is required between tool and workpiece. Relative motion is achieved by two types of motion—primary motion in the form of cutting speed and secondary motion in the form of the feed. Primary and secondary motions combined with the shape of tool and its penetration in the workpiece result in the desired shape and size of the work surface [1]. The quality of the machined workpiece, i.e. surface finish depends on all the input parmeters. Heat generated during turning operation plays a very critical role. It has a large impact on the tool wear and surface integrity of the product formed; tool life is reduced by a large extent. Heat generated during A. S. Patel (B) · M. A. Makhesana · K. M. Patel Institute of Technology, Nirma University, Ahmedabad, Gujarat 381482, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_45

539

540

A. S. Patel et al.

turning operation is distributed in three different regions—shear zone (primary), deformation zone (secondary), and worn out of flank (tertiary). In primary zone, heat generated is due to plastic deformation, and shear energy is completely converted into heat energy; hence, major portion of heat generated is in primary zone. In secondary zone, heat generated is due to rubbing between chip and rake surface of tool; heat generated in this region is very low as compared to primary zone; therefore in many cases, it is neglected. In tertiary zone, heat generated is due to rubbing of tool and finished surface of the workpiece; heat generated in this region affects surface finish of the product produced.

45.2 Literature Review Xavior and Adithan [2] have studied the influence of cutting fluids on tool wear and surface roughness during turning of ANSI 304 with carbide tool. The significant parameters influencing the surface roughness and tool wear were found using ANOVA procedure. It was found that feed rate has greatest influence on surface finish and cutting speed has greatest influence on tool wear. Among three lubricants, coconut oil was found to be a best lubricant for reducing tool wear and surface roughness. Reddy and Rao [3] investigated the effect of solid lubricant on machining process with high material removal rate. They compared the experimental result of machining with solid lubricants and conventional cutting fluids. It was concluded that there is no such improvement in machining process with solid lubricant as compared to conventional lubricants. Singh and Rao [4] worked on enhancing surface quality by solid lubricant in hard turning. MoS2 as a solid lubricant was applied as an alternative to the use of cutting fluids. Result originates from these was that the execution enhanced in the machining zone when MoS2 solid lubricant was utilized furthermore enhanced machinability with noteworthy change in surface quality. Patel et al. [5] studied the effect of dry and wet turning. External turning on EN31 material was performed with tungsten carbide insert tool. Effect of feed, depth of cut, and cutting speed on turning operation in both operations were observed. It was concluded that using wet environments leads to 20–30% reduction in process temperature and there is lower power consumption in wet environment. Tool wear is reduced significantly, which leads to better surface finish as compared to dry cutting. Korkut [6] explored the impact of cutting temperature on workpiece surface. It has been demonstrated that cutting temperature affected the integrity over the machined surfaces. Schubert and Nestler [7] have done work to improve surface finish and stress condition in shear zone in machining of aluminum matrix composites (AMCs) by modifying tool geometry. For experiment, only feed was varied keeping all other parameters constant and each parameter chosen was tested thrice. Experimental readings concluded that tools with large wiper radius or trailing edge are particularly good for generating surface with small roughness values.

45 Investigative Study of Temperature Produced During Turning …

541

Mantle et al. [8] have checked the performance of various ranges of polycrystalline diamond tools (PCD) for turning of Ti–6Al–2Sn–4Zr–6Mo with pressurized cutting fluid of 150 bar. Five different grades of PCD insert tools were used. Tools were differentiated by their grain size. Flank wear was taken as end criteria for test keeping all other parameters constant. From experimental readings, it was found that tool with grain size of approximately 14 µm leads to increase in tool life. Silberschmidt et al. [9] investigated effect of vibration on cutting forces and temperature levels in cutting region for different cutting conditions. Low-energy highfrequency vibration was superimposed on tool movement using ultrasonic-assisted turning (UAT). Dry machining was done to obtain all experimental readings. Using MSC MARC/MENTAT 3D, thermo-mechanically coupled finite element method of cutting tool and ultrasonic-assisted turning was developed. At tool–workpiece interface, friction behavior was simulated using modified shear friction law. Tangential component of cutting forces was reduced by 69%. Sullivan and Cotterell [10] have measured reaction forces and temperature during turning operation. They also observe the effect of cutting forces on surface temperature. Temperature was measured using a welded tip PTEE-insulated K-type thermocouples and infrared thermal imagining camera. Measurement of forces was done by Kistler quartz four-component dynamometer. Cutting forces are very important variables in generation of surface temperature. It was also concluded that tool wear resulted in increased cutting forces and machined surface temperature. Byrne [11] evaluated average interfacial temperature generated in turning of various metals experimentally using by Herbert/Gottwein dynamic thermocouple method. High-speed steel (HSS) tools were used to machine mild steel, aluminum, brass, and stainless steel. Calibration by furnace method was done to evaluate the thermoelectric relation between HSS and each metal. Influence of cutting tool condition on the emf generated was investigated as well as AC and DC component of emf were explained. Extensive range of process parameter for all individual material was selected to obtain results for interfacial temperature. For wide range of interfacial temperature, results were presented and discussed. From the above review of research papers, it can be concluded that machining process can be enhanced in terms of the quality of the product produced as well as improved production rate. It is the time to improve methods of conventional machining in terms of cutting parameters and lubrication method. From the data, it can be concluded that proper use of solid lubricants can help in reducing friction between tool–workpiece interface and also in cooling which leads to better surface integrity of the workpiece.

45.3 Experimental Details From the wide range of material available in the market and their application in industry, AISI 4140 steel has been selected as the workpiece material for turning. It is also known as high tensile steel, used in manufacturing axles, conveyor parts, crowbars,

542

A. S. Patel et al.

gears, logging parts, spindles, shafts, sprockets, studs, pinions, pump shafts, rams, ring gear, etc. AISI 4140 has wide application in the industries, and many of its application do not allow dimensional tolerance, hence machining of such metal should be done precisely. Chemical composition of the material is given in Table 45.1. Response variables needed for experiment are power consumption and temperature at tool–chip interface. Carbide tools are selected because workpiece material has good tensile strength, yield strength, and hardness values. Considering the workpiece material and manufacturer’s recommendation, CNMG120404 grade TN2000 as the cutting insert is selected. Experiments are performed to obtain optimal process parameter for machining of AISI 4140 steel with CNMG120404 carbide tool of TN400 grade. Experiments are conducted to assess the performance of graphite and MoS2 as solid lubricants and the results are compared with dry machining. Experiments are conducted on heavy duty lathe machine KIRLOSAKR TURNMASTER 35. Workpiece specification for initial cutting will be 50 mm in diameter and 1000 mm in length. Based on the available spinde speeds on the machine, the workpiece diameter is changed to obtain different cutting speeds. From the available values of feed rates 0.2, 0.355, and 0.5 mm are selected for experiment work. Higher value of feed rate is selected to check its influence on response, when there is a need for higher production rate. Values of depth of cut selected for experiments are 0.5, 0.75, and 1 mm. Higher depth of cut value is selected to check the influence of parameter on reaction forces produced and power consumption. For selected workpiece material and available raw material dimensions, values of cutting speed levels as 85, 125, and 190 m/min are selected. So from available parameters, raw material has to be turned in desired diameter. Selected parameters and their levels are shown in Table 45.2. Heterogeneous mixture of SAE 40 and solid lubricant is used as cutting fluid for machining. Mixture can be formed by different weight ratios of solid lubricant and SAE 40 oil. Many researchers had worked to find proper weight ratio of solid lubricant in heterogeneous mixture for better surface finish and less temperature generation. Krishna and Rao [12] used 5, 10, 15, 20, 30, and 40% solid lubricant by weight with SAE 40 oil to test the lubricating and cooling properties of the cutting fluid. It was found that increasing the solid lubricant content in mixture improves Table 45.1 Chemical composition of the material C

P

S

Si

Mn

Cr

Mo

Ni

Fe

0.386

0.032

0.029

0.377

0.67

1.04

0.091

0.143

Balance

Table 45.2 Process parameters

Process parameter

Level 1

Level 2

Level 3

Cutting speed (m/min)

85

125

190

Feed (mm/rev)

0.2

0.335

0.5

Depth of cut (mm)

0.5

0.75

1

45 Investigative Study of Temperature Produced During Turning …

543

machining process, 15% by weight of solid lubricant used in mixture gives the most optimum results. As it is required to conduct experiment for turning operation by varying three cutting parameters at three different levels, number of trials will be very high; so to optimize number of trials covering all different cutting parameters, it is preferred to use response surface methodology (Optimal 3 × 3), where 3 indicates number of levels and 3 indicates number of factors. Design of experiment was done using DESIGN EXPERT 10 software. Using response surface methodology (optimal 3 × 3), ten optimized trials were obtained, which cover all possible combinations of cutting parameters to calculate influence of process parameters on responses. Using DESIGN EXPERT 10, analytical formula for calculating surface roughness and temperature is developed. Thermocouples have been a mainstream transducer utilized as a part of temperature estimation. Thermocouples are exceptionally tough and economical and can work over a wide temperature range. Temperature has been measured by tool–work thermocouple. Complete setup of tool–work thermocouple was assembled on lathe machine as shown in Fig. 45.1. Calibration by the means of furnace has some limitations, and we can obtain linearly increasing temperature with steady-state condition, but we cannot go for higher values of temperature. To avoid emf leakage, workpiece metal chip and copper wires were insulated by glass fiber tape. For higher readings of temperature, metal plate was heated by oxy-acytelene flames for obtaining higher temperatures. For each value of emf, average value of temperature was taken and given an input to the graphpad for obtaining the thermoelectric relationship. It was found that emf and temperature are linearly dependent as shown in Fig. 45.2. A multiple correlation

Fig. 45.1 Temperature measurement setup mounted on machine

544

A. S. Patel et al.

Fig. 45.2 Temperature measurement setup mounted on machine

coefficient of 0.9998 was obtained from graphpad. Equation obtained from graphpad was Temperature = 70.81 × EMF + 36.29.

45.4 Results Experimental readings were analyzed using ANOVA (response surface methodology). Readings obtained for all different cutting conditions were compared to obtain more precise effect of the solid lubricant on the machining process. For each set of trials, new tool was used. As a response, temperature at tool–chip interface was measured as well as power was measured by wattmeter (Table 45.3). Graphs for the obtained results were plotted for comparing the effect of lubrication on the machining. From the graph, it can be concluded that molybdenum disulfide is better lubricant than graphite. By use of solid lubricant, we can decrease cutting temperature by considerable amount (Fig. 45.3). It is clear from the graphs that there is a considerable change in the value of temperature at tool–chip interface with the change in value of depth of cut. Whereas value of temperature at tool–chip interface is almost constant for all values of feed and cutting speed. With increase in the value of depth of cut, the value of surface roughness increases almost linearly. Effect on temperature by individual parameter as well as by combined effects of parameter was calculated using ANOVA. Following equations are obtained for various cutting conditions. T (Dry machining) = 332.709 − 182.042 × F + 550.142 × D + 1 : 029 × V + 167.336 × F × D + 1.575 × F × V − 1.47975 × D × V − 224.801 × F − 138.372 × D2 (45.1)

0.5

0.2

0.355

0.355

0.2

0.355

0.5

0.2

0.5

0.5

2

3

4

5

6

7

8

9

10

Feed (mm/rev)

1

S. No.

1

1

0.5

0.5

1

1

0.5

0.75

0.75

0.75

D.O.C (mm)

Table 45.3 Experimental results

190

85

190

125

190

125

85

125

85

190

Cutting speed (m/min)

110

70

100

90

150

60

80

110

50

130

694.743

673.5

638.095

595.609

694.743

680.581

574.366

659.338

616.852

701.824

160

105

90

90

175

75

50

100

40

150

652.337

645.256

624.013

546.122

680.661

659.418

531.96

638.175

602.77

652.337

Temperature (°C)

Power consumption (W)

Power consumption (W)

Temperature (°C)

Machining with MoS2

Machining with graphite

155

100

95

90

160

90

55

100

50

150

Power consumption (W)

Dry machining

746.258

708.985

659.418

602.77

730.228

716.066

581.527

680.661

666.499

716.066

Temperature (°C)

45 Investigative Study of Temperature Produced During Turning … 545

546

A. S. Patel et al.

Fig. 45.3 Comparison of lubricating conditions, a depth of cut versus temperature, b cutting speed versus temperature, c feed versus temperature

T (Machining with Graphite) = 206.992 + 1050.799 × F + 91.093 × D + 2.073 × V − 78.356 × F × D + 5.968 × F × V − 4.844 × D × V − 2451.010 × F + 483.501 × D2

(45.2)

T (Machining with MoS2 ) = −6.72072 + 652.673 × F + 804.307 × D + 2.015 × V − 8.546 × F × D + 0.368 × F × V − 2.114 × D × V − 1042.033 × F − 221.759 × D2

(45.3)

45 Investigative Study of Temperature Produced During Turning …

547

45.4.1 Numerical Analysis Mathematical model generated by experimental readings is adequate to give value of temperature at tool–chip interface for machining of AISI 4140 with tungsten carbide tool of given specification. For the selected process parameter, simulations were performed to obtain maximum temperature at tool–chip interface. Simulations were performed on ANSYS workbench 16.2. The tool body was kept at room temperature, whereas at the tool tip, heat flow was given input. It was observed that temperature obtained by the mathematical models is within 5% range of the temperature readings obtained by the simulation. Figure 45.4 shows the simulation results for process parameter 0.355 mm/rev as feed, 0.5 mm as depth of cut, and 85 m/min as cutting speed. Experimental value of tool tip temperature was 582.034 °C. Value obtained from the simulation is 581.05 °C, which is 0.169% error. Figure 45.5 shows the simulation results for process parameter 0.5 mm/rev as feed, 0.5 mm as depth of cut, and 125 m/min as cutting speed. Experimental value of tool tip temperature was 602.37 °C. Value obtained from the simulation is 603.67 °C, which is 0.216% error. Figure 45.6 shows the simulation results for process parameter 0.355 mm/rev as feed, 0.75 mm as depth of cut, and 125 m/min as cutting speed. Experimental value of tool tip temperature was 678.866 °C. Value obtained from the simulation is 685.75 °C, which is 1.01% error. Figure 45.7 shows the simulation results for process parameter 0.355 mm/rev as feed, 0.5 mm as depth of cut, and 85 m/min as cutting speed. Value obtained by mathematical for tool tip temperature was 744.896 °C. Value obtained from the simulation is 745.85 °C, which is 0.128% error. Values obtained from the simulation and mathematical models are within 5% range of each other. Hence, mathematical model generated can be used to calculate surface finish and temperature at tool–chip interface at any value of process parameters.

Fig. 45.4 Temperature at tool tip

548

Fig. 45.5 Temperature at tool tip

Fig. 45.6 Temperature at tool tip

Fig. 45.7 Temperature at tool tip

A. S. Patel et al.

45 Investigative Study of Temperature Produced During Turning …

549

45.5 Conclusions In order to obtain temperature at tool–chip interface and surface roughness under different cutting conditions, experiments were conducted by varying process parameters of machining such as feed, depth of cut, cutting speed, and lubricants. By the use of MQL with solid lubricants, temperature at tool–chip interface can be reduced. Due to its excellent lubrication properties even at higher temperatures, Molybdenum disulfide in SAE 40 oil is best suited to reduce temperature as compared to graphite and other cutting fluids. Depth of cut is the most influencing process parameter for the temperature produced in machining. Mathematical model was generated using ANOVA (response surface methodology) for temperature generated at tool–chip interface and surface roughness of the workpiece for dry cutting condition and each lubricant separately. Validation of mathematical model was done by comparing the results of simulation done on ANSYS.

References 1. Sharma, V.S., Dogra, M., Suri, N.M.: Cooling techniques for improved productivity in turning. Int. J. Mach. Tools Manuf. 49(6), 435–453 (2009) 2. Xavior, A., Adithan, M.: Determining the influence of cutting fluids on tool wear and surface roughness during turning of AISI 304 austenitic stainless steel. J. Mater. Process. Technol. 209(2), 900–909 (2009) 3. Reddy, N.S.K., Rao, P.V.: Experimental investigation to study the effect of solid lubricants on cutting forces and surface quality in end milling. Int. J. Mach. Tools Manuf. 46(2), 189–198 (2006) 4. Singh, D., Rao, P.V.: Improvement in surface quality with solid lubrication in hard turning. In: Proceedings of the World Congress on Engineering, London, UK, vol. 3, pp. 2–4 (2009) 5. Patel, M.H., Makhesana, M.A., Patel, K.M.: Performance assessment and selection of solid lubricants for environmentally benign lubrication in machining. In: Proceedings of 3rd International Conference on Industrial Engineering (ICIE-2015), SVNIT, Surat, ISBN: 978-9384935-56-6, pp. 437–441 (2015) 6. Ihsan, Korkut, Boy, Mehmet, Karacan, Ismail, Seker, Ulvi: Investigation of chip-back temperature during machining depending on cutting parameters. Mater. Des. 28(8), 2329–2335 (2007) 7. Schubert, A., Nestler, A.: Enhancement of surface integrity in turning of particle reinforced al minimum matrix composites by tool design. Procedia Engineering 19, 300–305 (2011) 8. Mantle, A.L., Pretorius, C.J., Soo, S.L., Aspinwall, D.K., Harden, P.M., M’Saoubi, R.: Tool wear behaviour and workpiece surface integrity when turning Ti–6Al–2Sn–4Zr–6Mo with polycrystalline diamond tooling. CIRP Ann. Manuf. Technol. 64(1), 109–112 (2015) 9. Muhammad, R., Ahmed, N., Roy, A., Silberschmidt, V.V.: Numerical modelling of vibrationassisted turning of Ti-15333. Proced. CIRP 1(1), 347–352 (2012) 10. Sullivan, D., Cotterell, M.: Temperature measurement in single point turning. J. Mater. Process. Technol. 118(1), 301–308 (2001) 11. Byrne, G.: Thermoelectric signal characteristics and average interfacial temperatures in the machining of metals under geometrically defined conditions. Int. J. Mach. Tools Manuf 27(2), 215–224 (1987) 12. Nageswara Rao, D., Vamsi Krishna, P.: The influence of solid lubricant particle size on machining parameters in turning. Int. J. Mach. Tools Manuf. 48(1), 107–111 (2008)

Chapter 46

Effect of Dressing Infeed on Alumina Wheel During Grinding Ti–6Al–4V Under Varying Depth of Cut Manish Mukhopadhyay , Souvik Chatterjee , Pranab Kumar Kundu and Santanu Das Abstract Dressing removes the outermost layer and helps in exposing new sharp grits from subsequent layer of a grinding wheel. Grinding wheel becomes loaded with chips and blunted during its continuous engagement in the grinding process. In order to maintain sharpness in the cutting edges, which would eventually improve the grindability, dressing is important. Dressing effectively increases the shearing ability of the cutting wheel and reduces the unproductive rubbing and ploughing actions. Dressing parameter, like infeed, plays an important role in obtaining enhanced grindability, which is extremely significant for processing the super alloys like Ti–6Al–4V. In the present experimental study, identical alumina wheels are dressed with single point diamond dressers at infeed values of 5, 10, 15, 20 and 25 µm. The effect of dressing infeed on alumina wheel for grinding Ti–6Al–4V has been investigated based on the grinding performance at 5, 10 and 15 µm infeed values. Twenty passes of up-grinding have been performed for each set of grinding operations. The inferences have been drawn based on the grinding force requirement, ground surface characterization, grinding ratio, etc., obtained herein. Keywords Dry grinding · Titanium · Dressing infeed optimization · Single point diamond

46.1 Introduction The recent advancement in material science and technology has led to the use of various alloys, which exhibit inherent favourable properties. One such alloy is Ti– 6Al–4V of titanium, comprising of aluminium (6wt%) and vanadium (4wt%) that serves as the ‘workhorse’ of the titanium industry due to its excellent corrosion resistance, good creep resistance, high temperature strength, high strength to weight M. Mukhopadhyay · P. K. Kundu (B) National Institute of Technology, Sikkim, South Sikkim 737139, India e-mail: [email protected] S. Chatterjee · S. Das Kalyani Government Engineering College, Kalyani, Nadia, West Bengal 741235, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_46

551

552

M. Mukhopadhyay et al.

ratio and biocompatibility among other useful properties [1]. Ti–6Al–4V is widely used in different industries like aerospace, automobiles, engines, gas turbines and biomedical applications and other sectors. The extensive applications of Ti–6Al–4V require it to be processed into a product having high dimensional precision and good surface form, making grinding an indispensable processing operation for it [2]. In grinding, abrasive grits are held together with a bond material in the form of a wheel. This wheel acts as a cutting tool in which material removal occurs due to the relative interaction between the fast-rotating abrasive (grinding) wheel and workpiece, in the form of tiny chips [3, 4]. As, in general, grinding is performed at high cutting velocity, it produces large forces, that leads to requirements of high specific energy and power consumption. Furthermore, due to high rate of plastic deformation, high heat generation at the grinding zone is detrimental to the surface quality and integrity of the workpiece. Additionally, surface burn and chip re-deposition along with surface cracks are very common in grinding of difficult-to-machine (DTM) material like Ti–6Al–4V. Apart from that, wheel loading and rapid tool wear are very common phenomena, frequently encountered during grinding, especially for DTM [5]. The problems associated can be controlled by selection of proper process parameters, application of cutting fluids, etc. [6–9]. Apart from these, from the recent studies, it has been observed that dressing parameters can play a significant role in obtaining enhanced grindability. The fundamental objective of the dressing is to produce topography and conducive geometry in the grinding wheel through removal of abrasive grits which have lost their sharpness in a controlled way [10]. Pacitti and Rubenstein [11] stated that dressing of vitrified bonded wheel should not be performed at too high or too low dressing depths. An intermediate strength while dressing could lead to a superior grindability of the wheel. Buttery et al. [12] observed that wheels dressing with very low dressing infeed results in better workpiece surface finish. Pande and Lal [13] suggested high dressing depth while using aluminium oxide wheel to exhibit better grindability on mild steel. Baseri et al. [14] reported that superior ground surface can be generated by lowering the infeed during grinding of SPK 1.2080 using an alumina wheel, dressed by a disc type dresser. Klocke and Linke [15] reported that the grinding face structure of the wheel gets enfeebled by previous truing infeed strokes, along with grinding infeed. In a more recent and relevant study, Mukhopadhyay and Kundu [16] performed experimental analysis on finding the optimal dressing infeed value for grinding Ti– 6Al–4V while employing alumina wheel. They reported that 20 µm infeed value to be the most suitable dressing infeed for the work tool combination. However, in their experimental investigation, they did not elaborate on the possibility of change in optimized value of dressing infeed with varying grinding infeed values, making the present study more relevant and extensive. The present experimental investigation deals in finding the effects on the performance of varying grinding infeed of differently dressed alumina wheels. The alumina wheels are dressed at identical dressing infeed values as considered by Mukhopadhyay and Kundu [16]. These wheels are subsequently, employed for grinding of Ti–6Al–4V at 5, 10 and 15 µm infeed values. The results at 10 µm grinding infeed have been found to be identical to that of the

46 Effect of Dressing Infeed on Alumina Wheel …

553

results of Mukhopadhyay and Kundu [16], which signifies the reliability of the system used herein. Based on the results obtained for other grinding infeed values, the performance is critically analysed with the help of the parameters like grinding force requirement, grinding ratio, average surface roughness parameter, and micrographs of the ground substrate obtained under SEM.

46.2 Materials and Methods Grinding of Ti–6Al–4V has been performed using a surface grinder of HMT make in plunge grinding mode. Dressing operations have been accomplished using a single point diamond dresser of 0.75 carat. A new dresser has been used for each set of dressing infeed condition to minimise the experimental errors associated with it. Separate alumina wheels with identical specifications have been employed for each set of experiments to avoid possible uncertainties in the process. While truing a given wheel, identical depth is considered as per the intended dressing infeed, i.e. for 5 µm dressing infeed (depth), truing is achieved using 5 µm infeed only. Thereafter, ten dressing passes with the particular depth of cut are accomplished at constant crossfeed of 30 mm/sec. The dressing infeed values 5, 10, 15, 20 and 25 µm have been conceived for this experiment identical to Mukhopadhyay and Kundu [16, 17]. For each dressing infeed, up-grinding operations have been performed at 5, 10 and 15 µm infeed values. To ensure dimensional precision, spark out of the work material is performed before every set of grinding operation. Each set of up-grinding operation contains 20 passes. Table 46.1 represents the details of the experimental investigation. The variations of tangential (F t ) and normal (F n ) force components of the grinding force have been accurately measured for each grinding pass using a strain gauge dynamometer. The ground surfaces have been characterized by measuring the average surface roughness using mechanical type stylus at five different points on the ground surface and by observing the micrographs under scanning electron microscope. Radial wheel wear has been estimated using a high precision dial gauge and the volume of material removal has been calculated using a high precision height gauge that subsequently, be used to assess the grinding ratio. The dressing set-up and the pictorial view of the grinding for the present study are represented in Fig. 46.1.

46.3 Results and Discussion In the present experimental study, grinding performance is analysed based on the parameters like grinding force, average surface roughness, grinding ratio, surface morphology and chip form observed herein. Grinding force requirement provides a direct indication about the ease of the process. Generally, lower force requirement suggests an enhancement of grinding performance. Figure 46.2 represents the aver-

554

M. Mukhopadhyay et al.

Table 46.1 Experimental details Surface grinding machine

Make: HMT Limited, Praga Division, India Spindle speed: 2880 rpm Main motor power: 1.5 kW

Grinding wheel

Make: Carborundum Universal Limited, India Size: 200 mm × 20 mm Specification: AA60 K5 V8

Wheel dresser

Make: Norton Abrasives, India

Work material

Material: Ti–6Al–4V

Specification: 0.75 carat single point diamond tip Size: 60 mm × 60 mm × 6 mm Hardness: 33 HRC Composition (by weight): Ti—88.77%; Al—6.19%; V—4.25%; Fe—0.34% Dressing environment

Dry

Grinding environment Scanning electron microscope

Make: Zeiss, USA Serial: EVO-18

Surface roughness tester

Make: Mitutoyo, Japan; Model: Surftest 301

Grinding force dynamometer

Make: Sushma Industries, Bengaluru, India

Stereo microscope

Make: Gippon, Japan

Resolution: 0.05 µm Resolution: 0.98 N Magnification: 40×

(a)

(b)

Dresser Fig. 46.1 a Plunge surface grinder showing dressing operation. b Magnified view of the dressing operation

46 Effect of Dressing Infeed on Alumina Wheel …

555

Grinding infeed: Ft for Fn for

Average Grinding Force (N)

90

Ft for Fn for

Ft for Fn for

80 70 60 50 40 30 20 10 0 5 µm

10 µm

15 µm

20 µm

25 µm

Dressing infeed

Fig. 46.2 Variations of average tangential and normal force components using wheels dressed at different infeed values

age force requirement for different grinding environment. The average grinding force is calculated considering the forces experienced during 8th–20th passes because of the fact that initial couple passes (generally up to 7th pass) the force requirement shows an increasing trend due to the incapability of the system to overcome the stiffness. From Fig. 46.2, the trend of grinding force requirements suggests that the normal force is always greater than their corresponding tangential counter parts, which is due to the predominantly high negative rake of the randomly oriented grits, participating in material removal process. Also, the force requirement is found to be increased with the increasing grinding infeed; quite understandably because of the higher penetration and increase in frictional force at a higher grinding infeed. Comparing the interrelation of the dressing parameters, it can clearly be seen that generally, force requirement decreases with the increase in dressing infeed values. It is in accordance with arguments of retention of sharpness of the abrasive grits and surges in intergrit spacing as delineated by Mukhopadhyay and Kundu [16, 17]. With the increase and retention of grit sharpness, material is primarily removed through shearing mechanism, thus force requirement decreases. Overall characteristics of the graph suggest that dressing with 20 µm dressing infeed resulted in lowest grinding force generation in every case of grinding infeed values. However, when the dressing infeed is 25 µm, the wheel grits might have retained their sharp cutting edges for a prolonged period and subsequently, results in higher force requirement. The phenomenon is supported by the report presented by Wegener and his co-workers [18] where they reported that smaller dressing depths result in inappropriate ‘topology generation’ and hinder ‘opening up of the wheel’ as less cutting edges are formed. An anomaly in the behaviour of grinding force can be seen when 15 µm grinding infeed is provided to wheel dressed at 25 µm. This is due to the high reaction force acting

556

M. Mukhopadhyay et al.

0.6

1.412

1.5

1.354

1.112

1.02

0.832

0.666

0.774

0.714

0.59

0.49

1 0.8

(b) (d)

0.676

1.2

0.62

1.4

0.72

(a) (c) (e)

1.6

0.46

Average Surface Roughness (µm)

on the wheel during the operation and as wheel dressed at 25 µm is subjected to greater force while being dressed, it experiences higher grit dislodgement, resulting into lower penetration but comparatively higher material removal. Generally, force requirement for grinding using the wheels dressed at 5, 10 and 15 µm infeed values are found to be greater than that of 20 µm. This may have happened due to the insufficient intergrit spaces for the microchips to adhere thus, promoting wheel loading and high heat generation resulting in the loss of grit sharpness as reported by Mukhopadhyay and Kundu [16, 17]. Similar inferences were made by Baseri et al. [14] where they reported that for lower dressing infeed conditions, heat generation at the grinding zone is always high. Surprisingly from Fig. 46.2, grinding force requirement has been found to be least with 5 µm dressing infeed. This is due to the presence of blunt grits that might have participated in the grinding operation, thus removing material predominantly by rubbing action. This would have resulted into a polishing action on the ground surface thus, reducing the force requirement. This is further sustained by the fact that bond breakage and grit deformation are the two most important effects that are reported during grinding process having low dressing infeed as reported by Mukhopadhay and Kundu [16, 17]. Dressing and grinding infeed values are the most important factors determining the surface quality of any grinding process [14]. Figure 46.3 represents the variations of average surface roughness values obtained during different grinding conditions. Figure 46.3 suggests that increase in dressing infeed has resulted in rougher surface. This is in acceptance to the inference of Buttery et al. [12]. This trend is seen because with the increase in dressing infeed the grits have opened up and grit sharpness has increased considerably as also stated by Mukhopadhyay and Kundu [16, 17] and thus, shearing operation becomes predominant. Additionally, during grinding with lower dressing infeed, the smaller grits undergo glazing and plastic deformation thus, losing their sharpness and become blunt subsequently; hence, polishing action occurs, generating a smoother surface. The trend has violated in case of grinding

0.4 0.2 0

5 µm

10 µm

15 µm

Grinding Infeed (µm)

Fig. 46.3 Average surface roughness observed during different grinding infeed a 5, b 10, c 15, d 20 and e 25 µm dressing infeed

46 Effect of Dressing Infeed on Alumina Wheel …

557

with 25 µm infeed where it has generated smoother surface than that obtained during grinding with 20 µm dressing infeed. This may be due to the greater impact forces and high temperature rise provided by the dresser as reported by Klocke and Linke [15], making the wheel topology fragile and resulting into more bond breakage than that for 20 µm infeed. This bond breakage has resulted into reduced effective infeed during grinding that limits the shearing action to some extent. The ground surfaces, observed under SEM are shown in Fig. 46.4, which give clear visual indications to the effects of the dressing infeed. From Fig. 46.4, it can clearly be seen that the surface quality obtained during grinding with greater dressing infeed is better as compared to that using lower dressing infeed. As all the experiments have been performed under dry environment, without the application of any coolants, therefore, presence of surface defects is a common phenomenon. However, the surface defects are comparatively reduced with the increase in grinding using greater dressing infeed, indicating positive effect of better retention of grit sharpness and increased intergrit spacing, which reduce the chances heat generation. Economy in any machining operation is a paramount factor under consideration. In grinding, similar significance is given by the analysis of the grinding ratio or G-ratio, which is defined by the ratio of volume of work material removal by the volume of wheel material removal. Thus, higher values of grinding ratio imply the impression of higher material removal and comparatively, lower wheel wear rate. The life of a grinding wheel is directly related to its dressing operation, thus analyzing the G-ratio at different dressing infeed is important. Figure 46.5 represents the grinding ratio obtained during grinding of Ti–6Al–4V using the alumina wheels dressed at different infeed values. It can be seen from the figure that with the increase in dressing infeed, grinding ratio is improved. This is due to the lowering of force requirement with increasing dressing infeed, which culminates in lesser bond breakage and grit pullout. In case of 25 µm dressing infeed, grinding ratio is found to be inferior compared with 20 µm dressing infeed. This may be due to the large values of force and temperature generated while grinding with higher dressing infeed thus, resulting into a weakened surface topology with 25 µm dressing infeed as compared to 20 µm dressing infeed, well reported by Klocke and Linke [15].

46.4 Conclusions The present experimental investigation deals in finding the effects on the performance of varying grinding infeed of differently dressed alumina wheels. The alumina wheels are dressed at identical dressing infeed values as considered by Mukhopadhyay and Kundu [16]. These wheels are subsequently employed for grinding of Ti–6Al–4V at 5, 10 and 15 µm infeed values. The results at 10 µm grinding infeed have been found to be identical to that of the results of Mukhopadhyay and Kundu [16], which signify the reliability of the system used herein. Based on the results obtained for other grinding infeed values, the performance is analysed critically with the help

558

M. Mukhopadhyay et al.

(a)

(b)

Subdued ground surface with surface damage Subdued ground surface with surface damage

(c)

(d)

Surface Damage Improved grinding surface

(e)

Surface Damage

Fig. 46.4 Micrographs of the ground surfaces, generated using wheels, dressed at a 5 µm, b 10 µm, c 15 µm, d 20 µm and e 25 µm infeed values

0.2

0.52

0.45

0.38

0.32

0.31

0.28

0.23

0.2

0.29

(b) (d)

0.16

0.3

0.29

0.27

0.4

0.15

Grinding Ratio

0.5

0.3

(a) (c) (e)

0.6

559

0.48

46 Effect of Dressing Infeed on Alumina Wheel …

0.1 0

5 µm

10 µm

15 µm

Grinding Infeed (µm)

Fig. 46.5 Grinding ratio obtained during different grinding infeed a 5, b 10, c 15, d 20 and e 25 µm dressing infeed

of the parameters like grinding force requirement, grinding ratio, average surface roughness parameter, and micrographs of the ground substrate obtained under SEM. The following conclusions are drawn from the current study. • Grinding forces are found to decrease with increasing dressing infeed for every set of grinding infeed. Wheel dressed at 25 and 20 µm requires lesser grinding force for material removal than those dressed at 5, 10 or 15 µm. • Wheels dressed at 25 and 20 µm dept of cut are expected to have a greater life than the rest. • SEM micrographs of the ground substrate indicate that wheel dressed at 20 µm infeed provides the most satisfactory results for every set of grinding infeed values, considered herein. Analysing the different results, it can be stated that 20 µm is found to be the most optimal dressing dept of cut condition when grinding Ti–6Al–4V using alumina wheel at different grinding infeed values.

References 1. Mukhopadhyay, M., Kundu, P.K.: Performance evaluation of conventional abrasive wheels for grinding Ti–6Al–4V. IOP Conf. Series. Mater. Sci. Eng. 377(012043), 1–7 (2018) 2. Kundu, P.K., Das, S., Sinha, S., Chowdhury, P.P.: On grinding wheel performance in dry and wet conditions. In: 4th International Conference on Mechanical Engineering, BUET Dhaka, pp. 19–24 (2001) 3. Mukhopadhyay, M., Kundu, P.K.: Development of a simple and effective coolant delivery technique for grinding Ti–6Al–4V. Int. J. Mach. Mach. Mater. 20(4), 345–357 (2018) 4. Kundu, A., Mukhopadhyay, M., Banerjee, A., Mandal, B., Das, S.: Grinding of Inconel 718 using soap water jet and liquid carbon dioxide. Adv. Manuf. Mater. Sci. 293–300 (2018)

560

M. Mukhopadhyay et al.

5. Kundu, A., Mukhopadhyay, M., Mahata, S., Banerjee, A., Mandal, B., Das, S.: Grinding titanium grade 1 alloy with an alumina wheel using soap water. Procedia Manuf. 20, 338–343 (2018) 6. Mukhopadhyay, M., Kundu, P.K.: Laser assisted conditioning of aluminium oxide grinding wheel using Nd:YAG laser: a review. In: Proceedings of National Conference on Advanced Functional Materials Processing and Manufacturing, CMERI, Durgapur, pp. 63–66 (2017) 7. Mukhopadhyay, M., Kundu, P.K., Das, S.: Experimental investigation on enhancing grindability using alkaline-based fluid for grinding. Mater. Manuf. 33(16), 1775–1781 (2018) 8. Rai, B.R., Mukhopadhyay, M., Kundu, P.K.: Evaluating the grinding ratio and surface quality of Ti-6Al-4V under varying grinding pass count and depth of cut, IOP Conf. Series. J. Phys. 1240(012143), 1–7 (2019) 9. Mukhopadhyay, M., Kundu, P.K.: Evaluating application potentiality of unconventional fluids for grinding Ti-6Al-4V using alumina wheel. Mater. Manuf. Process. 34(10), 1151–1159 (2019) 10. Mukhopadhyay, M., Kundu, P.K.: Laser dressing of grinding wheels—a review. Int. J. Mecha. Manuf. Syst. 11(2/3), 167–181 (2018) 11. Pacitti, V., Rubenstein, C.: The influence of the dressing depth of cut on the performance of a single point diamond dressed alumina grinding wheel. Int. J. Mach. Tool Des. Res. 12, 267–279 (1972) 12. Buttery, T.C., Statham, A., Percival, J.B., Hamed, M.S.: Some effects of dressing on grinding performance. Wear 55, 195–219 (1979) 13. Pande, S.J., Lal, G.K.: Effect of dressing on grinding wheel performance. Int. J. Mach. Tool Des. Res. 19, 171–179 (1979) 14. Baseri, H., Rezaei, S.M., Rahimi, A., Saadat, M.: Analysis of the disc dressing effects on grinding performance—part 1: simulation of the disc dressed wheel surface. Mach. Sci. Technol. 12, 183–196 (2008) 15. Klocke, F., Linke, B.: Mechanisms in the generation of grinding wheel topography by dressing. Prod. Eng. 2, 157–163 (2008) 16. Mukhopadhyay, M., Kundu, P.K.: Optimization of dressing infeed of alumina wheel for grinding Ti–6Al–4V. Mater. Manuf. Process. 33(13), 1453–1458 (2018) 17. Mukhopadhyay, M., Kundu, P.K.: Impact of dressing infeed on SiC wheel for grinding Ti– 6Al–4V. Mater. Manuf. Process. 34(1), 54–60 (2019) 18. Wegener, K., Hoffmeister, H.W., Karpuschewski, B., Kuster, F., Hahmann, W.C., Rabiey, M.: Conditioning and monitoring of grinding wheels. CIRP Ann.—Manuf. Technol. 60, 757–777 (2011)

Chapter 47

Experimental Evaluation of Surface Roughness, Dimensional Accuracy, and MRR in Cylindrical Grinding of EN 24 Steel Pankaj V. Mohire

and Raju S. Pawade

Abstract The present paper discusses the experimental results and analysis of PRSR, dimensional accuracy, and material removal rate as the responses in cylindrical grinding of EN24 steel. The work material chosen was EN24, a popular grade of thorough-hardening alloy steel. The Taguchi design of experiments using L18 orthogonal array was used. It is observed from the experimental work that the input factors depth of cut has statistically significant effect on the PRSR. Material removal rate increases linearly with increase in wheel speed. Dimensional deviation is maximum for work speed of 180 rpm and it is minimum for work speed of 250 rpm. Keywords Cylindrical grinding · Surface roughness · PRSR · MRR · Taguchi

47.1 Introduction Cylindrical grinding is an essential process for final machining of components requiring smooth surface and precise tolerances. External cylindrical grinding is used to finish parts that have been machined to approximate size and heat treated to desired hardness. The various process parameters of a cylindrical grinding machine include depth of cut, number of passes, workpiece speed, grinding wheel, and grinding wheel speed. EN24, a popular grade hardened alloy is extensively used in motor vehicle and machine tool industries, axles, crankshafts, spindles, bearings, bushings, and rolls for rolling mills. Many researchers have focused their studies in cylindrical grinding of different materials which include the following: George et al. [1] performed the experimentation on cylindrical grinding machine and observed the decrease in surface roughness as the material hardness increased. The surface roughness decreased when wheel speed increases from 60 to 120 rpm; similarly when depth of cut increases from 10 to 20 µm, the surface roughness decreases. Kumar et al. [2] have carried out experimental study on cylindrical grinding of C40E steel and concluded that the P. V. Mohire (B) · R. S. Pawade Department of Mechanical Engineering, Dr. Babasaheb Ambedkar Technological University, Lonere, Raigad 402103, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_47

561

562

P. V. Mohire and R. S. Pawade

work speed is most influential factor for C40E work material. They found that the grinding wheel speed, number of passes, and depth of cut has influence on surface roughness. Hence, to achieve the minimum surface roughness of AISI 1040 steel, low depth of cut (300 µm), highest work speed (630 rpm) with moderate number of passes and high grinding wheel speed (1910 rpm) are the optimum parameters. Kumar et al. [3] performed cylindrical grinding on EN24 hardened grade alloy and observed that the effect of cutting speed is more significant than that of depth of cut. Panthangi and Naduvinamani [4] has performed a cylindrical grinding process on EN 24 to obtain better surface finish. They studied effect of material hardness, depth of cut, and work speed using the Taguchi method to optimize the surface finish. Karande et al. [5] perform experiments on surface grinding operation of EN19 steel and found optimum machining parameter which lead to minimum surface roughness and maximum material removal rate. It was concluded that optimum value of the surface roughness is noted at 20 mm depth of cut, 0.18 mm/rev is the feed rate and 415 rpm is the spindle speed with hardness of 50 HRc.

47.2 Experimental Details 47.2.1 Workpiece, Grinding Wheel, and Machine In this experimental work, EN24 alloy steel was chosen as work material. It is a readily machinable in the “T” heat treated condition. The workpiece in rod form with 32 mm diameter and 300 mm length was selected. These workpieces were first center drilled for accurate mounting in the cylindrical grinding machine. From the literature survey, it has been observed that aluminum oxide grade (A) wheel is recommended for machining of EN24 steel alloy, therefore it is used. The specification of grinding wheel selected was carborundum grinding wheel-AA46K5V40 having dimension 300 mm OD, 40 mm width, and 127 mm ID. Hydraulic cylindrical grinding machine, Make-Parishudh Model-GCU-200A was for conducting experimental trials (Fig. 47.1).

47.2.2 Selection of Input Parameters In present experimental work, Taguchi L18 array was used. Eighteen experiments were conducted for three factors. Three levels each was selected for cutting speed and feed rate both. MINITAB 15 software was employed to analyze the experimental data. Table 47.1 presents the experimental observations with standard designed runs of Taguchi array. Table 47.2 presents the L18 orthogonal array for selecting process parameters.

47 Experimental Evaluation of Surface Roughness …

563

Fig. 47.1 Workpiece and wheel Table 47.1 Input process parameter and their levels

Process parameters

Levels Low

Number of passes

3

Work speed (rpm)

90

Wheel speed (rpm)

1975

Depth of cut (µm)

20

Medium –

High 5

180

250

2200

2450

35

50

Table 47.2 L18 orthogonal array for process parameter Experiment no.

No. of passes

1

3

2 3

Work speed (rpm)

Wheel speed (rpm)

Depth of cut (µm)

90

1975

20

3

90

2200

35

3

90

2450

50

4

3

180

1975

20

5

3

180

2200

35

6

3

180

2450

50

7

3

250

1975

35

8

3

250

2200

50

9

3

250

2450

20

10

5

90

1975

50

11

5

90

2200

20

12

5

90

2450

35

13

5

180

1975

35

14

5

180

2200

50

15

5

180

2450

20

16

5

250

1975

50

17

5

250

2200

20

18

5

250

2450

35

564

P. V. Mohire and R. S. Pawade

Percentage Reduction in Surface Roughness (PRSR) Surface roughness of the workpiece before and after grinding was measured using Mitutoyo surface roughness tester SJ 310. Then, the percentage reduction in surface roughness was calculated as shown in Eq. (47.1). PRSR = (Ra value before grinding − Ra value after grinding)/(Ra before grinding) × 100

(47.1)

Material Removal Rate It is a measure of amount of material removed in specific period of time and determined using following formula [6] as shown in Eq. (47.2). MRR = wt. of wp. before grinding − wt. of wp. after grinding/Grinding time (47.2) Dimensional Accuracy: Accuracy in dimensions is important consideration in any machining operation. If the dimensions of the part are not within the tolerance limit, then part is rejected. Actual diameter of workpiece is measured by using digital Vernier caliper. Expected diameter of the work piece after grinding can be calculated by using Eq. (47.3) as shown in Eq. (47.3). Expected dia. after grinding = (dia. before grinding − dia. after grinding) − (2 × Depth of cut × No. of passes)

(47.3)

47.3 Experimental Procedure As cylindrical grinding is used as finishing operation, it should give better surface finish than finish obtained in preceding machining operation. So, initially, we performed a turning operation on round EN 24 steel workpiece with 0.1 mm/rev feed, 0.05 mm depth of cut, 100 m/min cutting speed, and 1400 rpm spindle speed. The experiments were performed on cylindrical grinding machine. Initially, dressing process was done to improve the efficiency of process. Therefore, the job was fixed between centers in work head and tail stock for dressing operation. Figure 47.2 shows the grinding of a specimen when it is fitted on chuck subjected to high-speed grinding wheel which is assisted with coolant system to meet better accuracy and finish. Surface roughness of the workpiece before and after grinding was measured with the help of Mitutoyo surface roughness tester (SJ 310).

47 Experimental Evaluation of Surface Roughness …

565

Fig. 47.2 Cylindrical grinding machine

47.4 Results and Discussion To analyze the effect of selected grinding parameters on response variables, the measurements were taken after performing the grinding trials. These readings were further used to find out the main effect plots and to obtain optimum conditions for cylindrical grinding.

47.4.1 Analysis of PRSR The experimental observation of percentage reduction in surface roughness is shown in Table 47.3. The optimum PRSR value obtained at work speed is 180 rpm, wheel speed is 2450 rpm, and depth of cut is 20 mm. It is observed from the main effect plots in Fig. 47.3 that as the number of passes increases from 3 to 5, the PRSR decreased. It is because of the larger wheel wear occurred with increase in number of passes. Increase in the work speed and depth of cut relatively increases the PRSR, and this might be due to rapid impingement of wheel on workpiece which affects the uneven travel of tool on workpiece. The lower work and depth of cut exhibits better surface finish, this similar results found by Saglam et al. [7]. When the wheel speed increases, the surface roughness increased because at higher wheel speed, grinding tool abrasives wears fast and new face of grits comes in contact with the exposed surface. Kremen et al. [2] found similar results wherein they observed at higher speed where the increased surface roughness increased.

566

P. V. Mohire and R. S. Pawade

Table 47.3 Experimental results for the output responses Run No.

No. of passes

Work speed (rpm)

Wheel speed (rpm)

Depth of cut (µm)

PRSR(%)

MRR (g/s)

Dimensional deviation

7

3

90

1975

20

37.84

0.242947682

0.01

7R

3

90

1975

20

16.39

0.293880087

0.07

18

3

90

2200

35

6.57

0.436609687

0.05

18R

3

90

2200

35

24.79

0.256794752

0.54

3

3

90

2450

50

76.1

0.473849935

0.1

3R

3

90

2450

50

88.84

0.315290004

0.09

6

3

180

1975

20

52.39

0.290296513

0.11

6R

3

180

1975

20

52.7

0.567433188

0.24

13

3

180

2200

35

18.88

0.344557557

0.12

13R

3

180

2200

35

22.52

0.424787606

0.21

4

3

180

2450

50

62.9

0.717644293

0.27

4R

3

180

2450

50

66.29

0.529294688

0.06

10

3

250

1975

20

47.88

0.476372534

0.26

10R

3

250

1975

20

24.11

0.425471159

0.11

17

3

250

2200

50

15.09

0.425001558

0.04

17R

3

250

2200

50

24.82

0.345400199

0.13

1

3

250

2450

20

60.93

0.476970172

0.04

1R

3

250

2450

20

71.2

0.501333333

0.11

14

5

90

1975

50

12.03

0.19551562

0.55

14R

5

90

1975

50

20.35

0.203798456

0.5

15

5

90

2200

20

23.89

0.512779616

0.52

15R

5

90

2200

20

19.52

0.521829282

0.63

11

5

90

2450

35

23.37

0.20052839

0.15

11R

5

90

2450

35

32.18

0.488249068

0.16

8

5

180

1975

35

43.68

0.179772482

0.39

8R

5

180

1975

35

40.14

0.328408648

0.98

2

5

180

2200

50

85.7

0.259650037

0.41

2R

5

180

2200

50

82.8

0.402267574

0.36

16

5

180

2450

20

0.168800708

0.06

16R

5

180

2450

20

14.8

0.331720802

0.36

12

5

250

1975

50

32.4

0.266622368

0.18

4.09

12R

5

250

1975

50

21.03

0.222795986

0.33

5

5

250

2200

20

53.74

0.19319738

0.83

5R

5

250

2200

20

64.32

0.211862843

0.05

9

5

250

2450

35

5.06

0.159833218

0.33

9R

5

250

2450

35

34.22

0.44756034

0.19

47 Experimental Evaluation of Surface Roughness …

567

Fig. 47.3 Effect of input parameter on PRSR

47.4.2 Analysis of Material Removal Rate The experimental observation of material removal rate is shown in Table 47.3. The higher cutting speed and depth of cut are more significant parameter during material removed. It was observed that with an increase in depth of cut, the surface finish decreases. It was observed that increase in number of passes increases the material removal rate because when the number of passes increases the wheel wears more and hence material removal rate decreased. It is observed that the work speed MRR are in direct proportion, whereas at work speed of 250 rpm the MRR is reduced. Because at higher work speed, the grits wear out rapidly that dulls the grinding wheel which turns into lower MRR. EN24 is a hardened material, at higher depth of cut, MRR decreased as the material is harder and not easily cut but at 50 µm depth of cut, the MRR slightly increases due to strain hardening of material. Jeevanantham et al. [8] found similar results wherein they noticed increase in surface roughness with increase in depth of cut and higher cutting speed led to more volume of material removal from workpiece within a short period of time. Kumar et al. [3] found similar results wherein they noticed material removal rate increase in MRR with increase in wheel speed (Fig. 47.4).

47.4.3 Analysis of Dimensional Deviation As the number of passes increases, the dimensional deviation is more. The volume of material removal is more with increase in number of passes also dimensional deviation reduced with the increase in wheel speed. At higher cutting speed grits wear out faster; therefore, all grits remained in contact each time with the workpiece; hence, dimensional deviation increased. When work speed increases, the dimensional

568

P. V. Mohire and R. S. Pawade NO. OF PASSES

26

WORK SPEED (rpm)

24 22

Mean of MRR

20 18 3

5

90

180

26

250

DEPTH OF CUT (mm)

WHEEL SPEED (rpm)

24 22 20 18 1975

2200

2450

20

35

50

Fig. 47.4 Effect of input parameter on MRR

NO. OF PASSES

Mean of Dimensional Deviation

0.4

WORK SPEED (rpm)

0.3

0.2

3

90

5

0.4

180

250

DEPTH OF CUT (mm)

WHEEL SPEED (rpm)

0.3

0.2

1975

2200

2450

20

35

50

Fig. 47.5 Effect of input parameter on dimensional deviation

deviation reduced because at higher work speed wheel moved fast and has less contact with machined surface (Fig. 47.5).

47.5 Conclusions The aim of the present study was to examine the effects of cylindrical grinding parameters on EN24 alloy steel. Based on the analysis of results, the following conclusions are drawn.

47 Experimental Evaluation of Surface Roughness …

569

1. The effect of number of passes on the MRR is found most significant. The MRR increase in number of passes from 3 to 5, because of increase in volume of material removal. However, the MRR decreased with increase in wheel speed as the number of grits in contact reduces due to large wear of abrasive grits at higher wheel speed. 2. The surface quality is assessed in terms of PRSR. Three number of passes produced the higher PRSR. Work speed and infeed relatively increase the PRSR. This is because of rapid impingement of wheel grits on the workpiece causing uneven travel of tool during grinding operation. 3. Dimensional deviation increased when the number of passes increases from 3 to 5 due to higher amount of material removal. However, the dimensional deviation decreased when the work speed increases because of reduced wheel contact with the workpiece.

References 1. George, L.P., Job, K.V., Chandran, I.M.: Study on surface roughness and its prediction in cylindrical grinding process based on Taguchi method of optimization. Int. J. Sci. Res. Publ. 3(5), 1–5 (2013) 2. Kumar, N., Tripathi, H., Gandotra, S.: Optimization of cylindrical grinding process parameters on C40E steel using Taguchi technique. Int. J. Eng. Res. Appl. 5(1), 100–104 (2015) 3. Kumar, P., Kumar, A., Singh, B.: Optimization of process parameters in surface grinding using response surface methodology. IJRMET 3(2), 245–252 (2013) 4. Panthangi, R.K., Naduvinamani, V.: Optimization of surface roughness in cylindrical grinding process. Int. J. Appl. Eng. Res 12, 7350–7354 (2017) 5. Karande, M.R.J., Patil, M.K.R., Jadhav, S.M., Nanwatkar, R.K.: Optimization of cylindrical grinding machine parameters for minimum surface roughness and maximum MRR. GRD J. Glob. Res. Dev. J. Eng. 2(5) (2017) 6. Pawade, R.S., Joshi, S.S.: Multi-objective optimization of surface roughness and cutting forces in high-speed turning of inconel 718 using Taguchi grey relational analysis (TGRA). Int. J. Adv. Manuf. Technol. 56(1–4), 47–62 (2011) 7. Saglam, H., Unsacar, F., Yaldiz, S.: An experimental investigation as to the effect of cutting parameters on roundness error and surface roughness in cylindrical grinding. Int. J. Prod. Res. 43(11), 2309–2322 (2005) 8. Jeevanantham, S., Sivaram, N.M., Smart, D.R., Nallusamy, S., Prabu, N.M.: Effect of machining parameters on MRR and surface roughness in internal grinding using EN8, EN31 steel. Int. J. Appl. Eng. Res. 12(11), 2963–2968 (2017)

Part III

Automation

Chapter 48

Bio-inspired Knowledge Representation Framework for Decision Making in Product Design Varun Tiwari , Prashant Kumar Jain

and Puneet Tandon

Abstract A lot of information, mostly disorganized, is available to the designer during early stages of design in the form of raw data. To extract useful information from raw data, its storage and analysis is important, for which knowledge representation plays prominent role. While new product variants are generated, a few previous designs are generally accessible to the designers to take inspiration. If efficient knowledge representation of relationships among previous designs exists, it would help the designers in new product development. In this work, analogy is sought from biological phenomenon in nature, i.e., phylogenetic, to store, depict, and retrieve knowledge from previous design concepts in the form of phylogenetic tree. The tree is developed using two Phenetic approaches, i.e., unweighted pair group method with arithmetic mean (UPGMA) and neighbor joining algorithm. The tree of UPGMA represents similarity of product features for different design concepts, and the tree of neighbor joining depicts number of modifications a designer performs to develop a new variant from the previous variants. An example of power drill is taken to illustrate the application of both the algorithms in product design. Keywords Knowledge representation · New product development · Phylogenetic · Phylogenetic tree

48.1 Introduction Product design is an evolutionary process, in which the design solutions evolve, adapt, and improve over a period in response to continuously changing customers’ demands and technological developments. Designer develops new design variants (also called design concepts) most of the time, by modifying design specifications V. Tiwari School of Engineering, Avantika University, Ujjain 456006, India P. K. Jain · P. Tandon (B) Department of Mechanical Engineering, Indian Institute of Information Technology Design and Manufacturing, Jabalpur 482005, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_48

573

574

V. Tiwari et al.

of previous variants. If the designers can depict relations among previous design solutions and develop a roadmap for modifications among design specifications of the previous products, so as to achieve new designs from previous designs, it would help the industries to keep track of the related changes in the existing products to anticipate future product developments. This work develops a mechanism to represent knowledge of similarity in product feature specifications among previous design variants by understanding and mimicking a biological phenomenon in nature called Phylogenetic. This work uses Phenetics-based Phylogenetic tree to depict and retrieve information regarding similarities among previous products’ specifications. In biology, Phylogenetics stands for the study of evolutionary relations among groups of entities, which are established by comparing molecular sequencing data or morphological features. Phenetics and cladistics methods are the two methods that can be used to build phylogenetic tree. Phenetics is a distance-based method, while cladistics is a character state method. Phenetics categorizes organisms, genes, or species on the basis of comparison of overall similarity (or difference) coefficients (i.e., distance) in morphological features, without considering evolutionary relationships. Cladistics considers the information only where the evolutionary change occurs in morphological features among organisms, genes, and species. There exists very few fundamental theories, methodologies, and knowledge representation frameworks that can depict, record, and retrieve changes in product features specifications from previously developed concepts to effectively anticipate new product developments. The present work develops knowledge representation framework in the form of a phylogenetic tree, which helps the designer to understand how product variants or design concepts are evolved from a set of product specifications of previous design variants. The developed phylogenetic tree would help to understand the intent of the designer, in the form of ‘what’ and ‘how,’ by representing the grouping of product variants that share common product features. ‘What’ here means what are the previously generated product variants available to the designers during new product development (NPD) process, which shares similar product characteristics with the new design, and ‘how’ represents how many design modifications the designers perform in previous designs to evolve the new design concepts. This paper uses two Phenetic-based approaches, unweighted pair group method with arithmetic mean (UPGMA) and neighbor joining algorithm [1] to develop the phylogenetic tree. The tree developed using UPGMA depicts the efficient grouping among product variants based on similarity of their product features, and tree developed using neighbor joining depicts the numbers of mutations the designers perform to achieve the new designs from the previous designs. This helps the designer to effectively manage the knowledge of similarities among previously existed product variants and use it for future product developments.

48 Bio-inspired Knowledge Representation Framework …

575

48.2 Related Work The approach proposed in this work helps to represent the product evolution by establishing the analogy with phylogenetic analysis. Very few researchers have developed knowledge representation framework to predict new product developments from previous trends of product evolution. For developing efficient knowledge representation, it is very important to understand classification of knowledge [2]. Chandrasegaran et al. [3] proposed classifications of knowledge along several dimensions as formal/tacit knowledge, product/process knowledge, and compiled/dynamic knowledge. Knowledge representation has found its applications in many fields including design process. Knowledge representation can be classified into five categories: pictorial, linguistic, symbolic, virtual, and algorithmic [4]. Bryan et al. [5] proposed assembly system reconfiguration planning approach for tackling continuous design changes in product families. A nonlinear optimization method was proposed to select cost-effective assembly sequence configuration from previously generated assembly sequences. In recent years, limited work is reported in the literature regarding design knowledge modeling based on Phenetic-based methods. Most of the work of representing previous products’ evolution was based on cladistic-based methods. Hölttä-Otto et al. [6] developed a quantitative algorithm to depict similarity based on Euclidean distance. This distance compares input and output of different modules. Based on their commonality, dendrograms depicted the similar modules in a product family. To predict future of novel product developments and manufacturing systems, AlGeddawy and ElMaraghy [7] used cladistics to track co-evolution path of previous products. Kashkoush and ElMaraghy [8] had proposed an algorithm for design knowledge modeling and retrieving that automatically identifies the previous design that is most similar to new product being designed. This method was based on ‘tree reconciliation’ (matching phylogenetic trees) algorithm. A tree in that work represents product’s bill of materials (BOM), which captures all the components of the product and hierarchical relationships among them. The main difference between the tree reconciliation algorithm proposed by [8], and the work reported in this manuscript is that the approach of Kashkoush and ElMaraghy [8] is based on cladistics and generates cladogram, whereas proposed method in this paper is based on Phenetics and generates Phenogram. In the approach proposed by [8], for each product variant, a cladogram needs to be generated so as to match with new products’ cladogram, whereas, in the proposed approach, the similarities among all the previous products and new product can be depicted by a single (common) tree. The proposed framework in this manuscript depicts efficient grouping among products based on their similarities of product features through a single tree. The true number of design modifications the designer performs in one product variant to achieve other variant can also be depicted by a single tree in this manuscript. ElMaraghy et al. [9] used cladistics-based approach to generate cladogram, which models evolution of the products along with manufacturing systems. AlGeddawy and ElMaraghy [10] hypothesized that product and

576

V. Tiwari et al.

manufacturing systems share common evolutionary course as different species. Azab and ElMaraghy [11] proposed mathematical model for reconfigurable process planning and compared this model with classical process planning models. This model is applied to evolvable product and part families by introducing process planning reconfigurable index (RI) to capture extent of changes. Cladogram represents evolution history of a group of genes and also shows relationship and differences among groups of genes. The proposed tree (Phenogram) in the manuscript is developed using information regarding total number of differences in product features among previous products, whereas cladogram proposed by [9] for previous product variants is developed by using information only from the features (characters) of previous product, where difference occurs. In other words, Phenogram uses macro-information about total number of differences in product features, whereas cladogram uses microinformation about difference in each product feature. The major advantage gained by using Phenetics in this work is that it helps to represent evolutionary relationships among product variants through one tree only.

48.3 Methodology The methodology starts with generation of product variants, which themselves are the source of knowledge. The basic steps for generating product variants are as follows: 1. Customer requirements are accessed by conducting a user survey. The customer requirements are denoted by a layered vector set CR = {CR I } where I → 1 to N 2. Product characteristic features are developed based on identified customer requirements. These characteristic features are denoted by layered vector set, product feature (PF), where PF = {PFi } where i → 1 to n 3. Each product characteristic feature can have values in the form of string, range, numbers, etc., which are called here as product characteristic feature values, denoted as FV.   FV = FVi j where j → 1 to m Here ‘i’ represents index of product characteristic feature, and ‘j’ represents feature value index for that product feature. 4. Product variants are generated by suitable combination of FVs from each PF. Each product variant has feature values which are raw data in our work. This raw data is compared for a number of changes in feature values to generate distance data.

48 Bio-inspired Knowledge Representation Framework …

577

Customer requirements Generation of Product variants

Contains hidden Knowledge

UPGMA Algorithm & Neighbour joining

Comparing feature values among product variants

Extracted knowledge

Raw data Computation of Distance matrix

Phylogenetic tree Useful Information

Knowledge Representation

Fig. 48.1 Proposed framework for knowledge representation in product design

The distance between two product variants is defined here as number of differences in feature values of product variants. This data is stored in the form of matrix, known here as distance matrix which is a useful information. UPGMA algorithm and neighbor joining algorithm are applied on this distance matrix to generate a phylogenetic tree to represent knowledge. This tree can yield useful information while generating new product variants or selecting existing product variants from design repositories based on the set of customer requirements. The proposed methodology is shown in Fig. 48.1.

48.3.1 UPGMA Algorithm for Knowledge Representation in Product Design UPGMA algorithm, in this work, is based on the idea of combining two product variants that are closest in terms of similarity of characteristic feature values. The algorithm includes recalculating the distances and then combining them again in an iterative manner. The algorithm terminates when all the product variants have been clustered in the form of a tree. The distance between the two product variants P i and P j in terms of number of differences in the product characteristic feature values is denoted as d i j . The distance represents the number of differences in product feature values, which are calculated by the designer. Input: A set of ‘n’ product variants in the form of distance matrix M. Output: A Phylogenetic tree that represents relationship among generated product variants.

578

V. Tiwari et al.

Algorithm 1. Find two product variants P i and P j from M having the smallest distance d i j . 2. Place the two P i and P j on the leaf node of the tree and create intermediate node, i.e., a new hypothetical product variant, as P ij , with branch length as d i j /2. 3. Compute the distance from the new P ij to all other product variants (except for P i and P j ) as arithmetic mean of their distances. For example, distance of P ij   from P k is d i j k = d i k + d j k 2. 4. Remove the columns and rows in M that corresponds to P i and P j . 5. Add new column and row for P ij and compute distances of P ij from all other product variants as explained in Step 3. 6. Return to Step 1, and repeat all the steps until each product variant that has been inputted are placed on the tree.

48.3.2 Neighbor Joining Algorithm The algorithm, in this paper, is based on the idea of combining two product variants that are neighbors in terms of similarity of characteristic feature values. Among the available product variants, the two variants having the smallest value of Dij (sum of branch length) are classified as closest neighbors. The process includes recalculating the new distances and then combining them again in an iterative manner, which terminates when all the product variants have been clustered in the form of a tree. Edge length can depict approximately correct evolution history. If distance matrix is additive, neighbor joining algorithm can capture true evolution history among different species. If distances are non-additive, in that case, neighbor joining algorithm can depict approximately actual history of evolution. A distance matrix denoted by ‘M’ is additive if it satisfies Buneman’s four-point condition [12]. M is additive if and only if for any four species, i, j, k, l, it satisfies condition; d(i, j) + d(k, l) ≤ max{d(i, k) + d( j, l), d(i, l) + d( j, k)}. In the proposed work, both the cases, where distances are additive (ideal case) and non-additive are considered to develop phylogenetic tree for depicting the number of changes, the designer preforms in product features of previous products to generate new variants. Input: A set of ‘n’ product variants in the form of distance matrix M. Output: A Phylogenetic tree that represents relationship among generated product variants. Algorithm 1. Find two product variants P i and P j from M having the smallest Dij (closest 1  neighbor) defined by D i j = d i j − r i + r j , where r i = q−2 k∈n d i k , q is number of product variants in each step of iteration.

48 Bio-inspired Knowledge Representation Framework …

579

2. Place the two P i and P j on the leaf node of the tree and create intermediate node, product variant, as P k , with branch length as d i k =  i.e., a new hypothetical  1 d and d + r − r = d i j − d i k. ij i j jk 2 3. Compute the distance of P with k  all other product variants (m) (except P i and P j ) as d km = 21 d i m + d j m − d i j . 4. Remove the columns and rows in M that corresponds to P i and P j . 5. Add new column and row for P ij and compute distance of Pij from all other product variants as explained in Step 3. 6. Return to Step 1 and repeat all the steps until all the product variants are placed on the tree.

48.4 Design Case Study A design case study of power drill is presented to illustrate the proposed methodology as well as the phylogenetic tree representation scheme for UPGMA, neighbor joining algorithm. The PFs and FVs of power drill are developed according to the identified customer requirements. For example, a customer requirement, ‘no wired drill’ (i.e., wireless drill), would lead to PF for this requirement as ‘power source,’ and FVs would be ‘wireless’ (and accordingly, ‘wired’). For power drill, based on the user study performed and available literature, the identified PFs are type (of drill), chuck size, no load speed, power source, variable speed trigger, reverse rotation, usage type, and estimated price. These features represent the customer requirements and would have different values (product characteristic feature values) for different designs. The initial concepts can be combined in various ways to generate a new set of product concepts. In this case study, eight product variants are initially developed as sequences to compare with UPGMA and neighbor joining algorithms. The product features along with their values are shown in Table 48.1 for product variants A, B, C, D, E, F, G, and H. The number of differences among product variants is computed to construct the distance matrix. This distance matrix becomes input for UPGMA algorithm and tree is constructed using steps described in Sect. 48.3. For neighbor joining algorithm, case where distance matrix is not additive, the product features along with their values are shown in Table 48.1 for product variants A, B, C, D, E, F, G, and H, and for the case where distance matrix is additive (Buneman’s four-point condition [12]), the product features along with their values are shown in Table 48.2 for product variants A1 , B1 , C1 , D1 , and E1 .

Pistol grip drill

10

1200–3000

Corded

Yes

Yes

Home

2000–3000

Chuck size (mm)

No load speed (rpm)

Power source

Variable speed trigger

Reverse rotation

Usage type

Estimated cost (INR)

A

Product variants

Type

Product features

2000–3000

Home

Yes

Yes

Corded

Above 3000

6.5

Pistol grip drill

B

Above 5000

Professional

Yes

Yes

Cordless

0–1200

10

Angle drill

C

Above 10000

Professional

Yes

Yes

Cordless

1200–3000

10

Angle drill

D

Table 48.1 Product characteristic features of variants for UPGMA and neighbor joining

Above 5000

Home & Prof.

Yes

Yes

Cordless

Above 1200

13

Pistol grip drill

E

2000–5000

Home

No

No

Corded

0–1200

13

Pistol grip drill

F

2000–5000

Home

Yes

Yes

Cordless

Above 1200

13

Pistol grip drill

G

Above 5000

Professional

Yes

No

Cordless

Above 3000

6.5

Angle drill

H

580 V. Tiwari et al.

48 Bio-inspired Knowledge Representation Framework …

581

Table 48.2 Product characteristic features’ variants for neighbor joining algorithm (additive distance matrix) Product features

Product variants A1

B1

C1

D1

E1

Type

Pistol grip drill

Pistol grip drill

Pistol grip drill

Angle drill

Angle drill

Chuck size(mm)

6.5

10

10

13

6.5

No load speed(rpm)

0–1200

Above 3000

Above 3000

1200–3000

1200–3000

Power source

Cordless

Corded

Cordless

Cordless

Cordless

Variable speed trigger

No

No

Yes

No

No

Reverse rotation

No

No

No

No

Yes

Usage type

Home and professional

Home

Home

Home and professional

Professional

Estimated cost (INR)

2000–3000

2000–3000

2000–3000

2000–3000

2000–3000

48.4.1 Phylogenetic Tree Creation for Power Drill by UPGMA Algorithm The distance matrix M calculated by computing number of differences among feature values of product variants A to H from Table 48.1 is reproduced as follows: ⎡ A 0 B⎢ ⎢2 C⎢ ⎢5 ⎢ D⎢4 M= ⎢ E ⎢4 ⎢ F⎢ ⎢5 G⎣3 H 7

2 0 6 6 5 5 4 5

5 6 0 2 4 7 5 3

4 6 2 0 4 8 5 4

4 5 4 4 0 6 2 5

⎤ 537 5 4 5⎥ ⎥ 7 5 3⎥ ⎥ ⎥ 8 5 4⎥ ⎥ 6 2 5⎥ ⎥ 0 4 7⎥ ⎥ 4 0 6⎦ 760

Algorithm described in Sect. 48.3.1 is applied on the distance matrix to generate phylogenetic tree shown as shown in Fig. 48.2. Tree depicts relationship among generated product variants. The figure shows product variants A, B, C, D, E, F, G, and H as taxa or product variants under comparison. Internal nodes represent hypothetical product variants, which possess all the product feature values of the product variants that branch out from that node. This node would help the designer to generate a new design possibility due to the fact that novel product variants would emerge

582

V. Tiwari et al.

Fig. 48.2 Phylogenetic tree for the power drill case study based on UPGMA

by including or removing FVs from the identified node. Branch length depicts distances calculated through UPGMA. Phylogenetic tree helps to form efficient grouping among product variants. It can be shown from Fig. 48.2 that phylogenetic tree can be divided into two groups. Product variants A, B, E, G, and F share common FVs, as they belong to the category of pistol grip drills, whereas product variants C, D, and H form angle grip drill. From the tree, designers’ intent can also be captured, as it is easy to find evolution trend of product variants. The distance matrix calculated by computing number of differences among feature values of product variants A1 to E1 from Table 48.2 is reproduced as follows

A1 B M= 1 C1 D1 E1

⎡A1 0 ⎢4 ⎢ ⎢ ⎢4 ⎢ ⎣3 4

B1 4 0 2 5 6

C1 4 2 0 5 6

D1 3 5 5 0 3

E⎤ 1 4 6⎥ ⎥ ⎥ 6⎥ ⎥ 3⎦ 0

48 Bio-inspired Knowledge Representation Framework …

583

Algorithm described in Sect. 48.3.2 is applied on both the distance matrices to generate phylogenetic tree for the non-additive matrix as shown in Fig. 48.3 and for additive matrix as shown in Fig. 48.4. Figure 48.3 shows product variants A, B, C, D, E, F, G, and H as taxa or product variants under comparison. Computed branch lengths between any pair of product variants depict distances, which are very close to actual numbers of modifications designers perform to generate new variants from previous variants. For example, sum of branch length from H to C is 3.4, which is close to 3 (actual number of modifications designers perform in H to generate C or vice versa). For the pairs of variants AB, EG, and CD, the neighbor joining algorithm provides true number of modifications, i.e., 2 for all the three pairs. Figure 48.4 shows product variants A1 , B1 , C1 , D1 , and E1 as taxa or product variants under comparison for additive distance matrix. Branch length depicts distance, which are the actual numbers of modifications designers perform to generate new variant from previous variant. For example, sum of branch length from A1 to B1 is four, i.e., designers perform four modifications in A1 to generate B1 . Internal nodes in the tree

2.03125

.46875 3.46875 F 1.125

2 0.5

1.125 H

.875

A

1.125

B

1.083333

.916667

E

0.9

G

1.1 D

C

Fig. 48.3 Phylogenetic tree for the power drill by neighbor joining algorithm (non-additive distance matrix)

Fig. 48.4 Phylogenetic tree for the power drill by neighbor joining algorithm (additive distance matrix)

1 1

2

B 1

A1

1 1 2

C1 D1

E1

584

V. Tiwari et al.

generated by neighbor joining algorithm can represent potential new design variants, which would help the designer to generate a new design possibility due to the fact that novel product variants would emerge by including or removing FVs from the identified node. For example, common node from which B1 and C1 evolve would generate new design solution B1 C1 . Design variant B1 C1 is generated by modifying one FV in each variant B1 and C1 and three FVs in variant A1 .

48.5 Advantages of Proposed Framework Some of the salient features of this representation framework are as follows: (a) Representing evolution of design variants Evolution of product variants is characterized by changes and modifications of product characteristic feature values for any product. This can be tracked and represented with the help of generated phylogenetic tree. For example, as shown in Fig. 48.2, variants ‘A,’ ‘B,’ ‘E,’ and ‘G’ are evolved by the designer by adding variable speed trigger and reverse rotation feature in the variant F. It is easy to identify the number of modifications; the designers perform in one product variant to generate another product variant through neighbor joining algorithm. (b) Improving product design decisions Phylogenetic tree helps the designers to identify the trend and actual number of design modifications among previous design variants. This would guide the designer to save their time and energy, by adopting preferably less complex, but correct modifications on the existing product variants. (c) Anticipating new design possibilities Evolution trend established by phylogenetic tree can help the designer to look for new design possibilities and novel alternatives for future product developments. For example, in Fig. 48.2, new product variants can be developed from product variant ‘F,’ due to the possibility of modifying some feature values. Similarly, as visible in Fig. 48.4, each internal node may represent new design solution, which can be generated by modifying feature values of previous design variants.

48.6 Conclusions A novel bio-inspired knowledge representation framework is developed to represent designer’s intent and to provide decision support tool to the designer for new product development. The technique of Phenetic-based representation scheme was developed to study similarities of product features specifications, among product variants or concepts. Its applicability to product design domain was demonstrated

48 Bio-inspired Knowledge Representation Framework …

585

using Phylogenetic tree developed by two algorithms, UPGMA and neighbor joining algorithms, and validated using a case study of power drill. It has been shown in the present work that the phylogenetic tree can yield additional promising information. These include rearranging previous product variants under study to form more meaningful grouping, generating more novel product concepts by combining product features from existing tree, enhancing product design decisions, and anticipating future directions for the product development process.

References 1. Saitou, N., Nei, M.: The neighbour-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425 (1987). https://doi.org/10.1093/oxfordjournals.molbev. a040454 2. Ammar-Khodja, S., Bernard, A.: An overview on knowledge management. In: Methods and tools for effective knowledge life-cycle-management, pp. 3–21. Springer (2008) 3. Chandrasegaran, S., Ramani, K., Sriram, R., Horváth, I., Bernard, A., Harik, R., Gao, W.: The evolution, challenges, and future of knowledge representation in product design systems. Comput. Aided Des. 45(2), 204–228 (2013). https://doi.org/10.1016/j.cad.2012.08.006 4. Owen, R., Horváth, I.: Towards product-related knowledge asset warehousing in enterprises. In: Proceedings of the 4th international symposium on tools and methods of competitive engineering, TMCE, pp. 155–170. China (2002) 5. Bryan, A., Wang, H., Abell, J.: Concurrent design of product families and reconfigurable assembly systems. J. Mech. Des. 135(5), 051001 (2013). https://doi.org/10.1115/1.4023920 6. Hölttä-Otto, K., Tang, V., Otto, K.: Analyzing module commonality for platform design using dendrograms. Res. Eng. Des. 19(2–3), 127–141 (2008). https://doi.org/10.1007/s00163-0080044-3 7. AlGeddawy, T., ElMaraghy, H.: A co-evolution model for prediction and synthesis of new products and manufacturing systems. J. Mech. Des. 134(5), 051008 (2012). https://doi.org/10. 1115/1.4006439 8. Kashkoush, M., ElMaraghy, H.: Product design retrieval by matching bills of materials. J. Mech. Des. 136(1), 011002 (2013). https://doi.org/10.1115/1.4025489 9. ElMaraghy, H., AlGeddawy, T., Azab, A.: Modelling evolution in manufacturing: a biological analogy. CIRP Ann. 57(1), 467–472 (2008). https://doi.org/10.1016/j.cirp.2008.03.136 10. AlGeddawy, T., ElMaraghy, H.: Co-evolution hypotheses and model for manufacturing planning. CIRP Ann. 59(1), 445–448 (2010). https://doi.org/10.1016/j.cirp.2010.03.032 11. Azab, A., ElMaraghy, H.: Mathematical modeling for reconfigurable process planning. CIRP Ann. 56(1), 467–472 (2007). https://doi.org/10.1016/j.cirp.2007.05.112 12. Buneman, P.: A note on the metric properties of trees. J. Comb. Theory (B) 17, 48–50 (1974). https://doi.org/10.1016/0095-8956(74)90047-1

Chapter 49

Heuristic Algorithmic Approach for Automatic Generation of Pin Layout for Robotic Unloading of Sheet Metal Parts A. Ramesh Babu Abstract In typical robotic sheet metal bending process, unloading of sheet metal parts is a challenging task due to complex nature of three-dimensional part geometry, and it is not possible to place the parts on a flat table. Hence, there is a need to design proper unloading station that allows the robotic gripper to unload the part with stability and without any collision. For this purpose, this paper proposes configuration of unloading station with a perforated table with a specific pattern of holes in which cylindrical pins are inserted. Then, a CAM system with a heuristic algorithm, along with a graphical interface, is developed to identify the number and distribution of pins on a perforated table that supports the part in a stable manner. The results are validated, for a variety of complex-shaped parts, using a graphical simulation tool. Keywords Robotic bending · Sheet metal bending · CAM system · Automation · Unloading of sheet metal parts

49.1 Introduction In the present manufacturing scenario, the industry is looking toward the implementation of virtual manufacturing, digital manufacturing, and smart manufacturing technologies to meet dynamic changes in the product development. This paper focuses on one of the modules in automating the manufacturing of the sheet metal parts. Sheet metal products such as enclosures and cabinets in various applications such as machine tools, automobiles, compressors, generators, electronic devices, furniture, and kitchen utensils are in huge demand with increasing complexity of part geometry. Sheet metal products are fabricated by assembling a variety of sheet metal parts that are produced by cutting and bending processes. At first, CAD model of the sheet metal part is taken as input, and the flat or unfold of blank is generated. Then, these 2D blanks are nested on a larger sheet, using an appropriate CAM system, followed A. Ramesh Babu (B) Department of Mechanical Engineering, PSG College of Technology, Coimbatore 641004, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_49

587

588

A. Ramesh Babu

by blanking or laser cutting or combination of these processes [1]. Subsequently, these blanks are bent using hydraulic press brakes to produce 3D components [2]. It is clear that many researchers and sheet metal machinery manufacturers are working to develop algorithms and CAM systems to virtually simulate many stages right from creating the part drawing, unfold generation, 2D nesting, NC code generation cutting, tooling selection for bending, bend sequence selection, and bending process simulation, etc. [3, 4]. In general, a bending operation is performed manually by an operator by feeding the blank between the punch and die. To make complete automation in sheet metal component production, manufacturers develop the robotic bending technologies [5, 6]. However, due to huge investment, robotic bending is being used by high-end customers. In this context of robotic bending, loading and unloading activities are to be automated to attain the complete automation. This paper focuses on the automation of unloading of sheet metal parts considering the robot assisted bending process. Typical robotic work cell consists of a press brake and a robot for bending, and a robot for loading/unloading, as shown in Fig. 49.1. The operational flow of the work cell starts from the loading of flat sheet material from the loading station. Flat sheets are stacked either horizontally or vertically. First, loading/unloading robot picks up a sheet and brings it close to bending robot so that bending robot grasps it by its gripper. Then, the bending operation is carried out by bending robot [3, 4]. At the end of the bending operation, bending/unloading robot unloads the product on unloading station. Typical loading and unloading activities are shown in Fig. 49.2. It may not be possible to unload 3D sheet metal part on a flat table due to the complexity of the part geometry. It needs some sort of support configured with a set of pins as shown in Fig. 49.2b. It can be seen that certain cylindrical pins are inserted on a flat table, and 3D part is positioned on pins. However, there is no literature available on a suitable method to generate the pin layout. The proposed algorithmic method is first-of-its-kind. At present, sheet metal parts are located manually by the operator, after bending. There is no method available to address this issue. This paper presents a method to arrive at pin layout suitable for unloading complex geometrical sheet metal parts and Fig. 49.1 Typical robotic work cell (Reference: Google images)

49 Heuristic Algorithmic Approach for Automatic Generation …

589

Fig. 49.2 Loading and unloading of a sheet metal part (Reference: Google images). a Loading the flat part and b unloading the finished part

ensures with a simulation system. The pin layout can be described as the location of pins on the unloading table that avoids the collision of the part during unloading and supports the part after unloading. In this context, the automatic generation of pin layout is the challenging task, since it strongly depends on the part geometry of the sheet metal part. Detailed approach for automatic generation of pin layout is as follows.

49.2 Approach As described above, in robot assisted sheet metal bending work cell, it is necessary to find the suitable pin layout for unloading a part. In this work, it is assumed that all the pins (like cylindrical locators with head) are of the same length. Figure 49.3 shows the photograph of a typical unloading table comprises a perforated base and pins inserted into holes. It can be seen that several holes are drilled on its base plate, and each hole is indexed to represent its location with respect to the bottom left corner of the table. Each row of the holes is represented as A, B, C, etc. Each column of a hole is represented as 1, 2, 3, etc. Indexing of the holes is shown in Fig. 49.4. It can be observed that certain holes at the bottom left side are closer to each other compared to other holes. This structure is suitable to handle both smaller and larger parts. Smaller parts need closer pins configuration, whereas larger parts can be located on pins that spread wider. With the given positions of holes, it is required to identify locations where pins are to be inserted. This depends on the geometry of parts. For example, consider a part as shown in Fig. 49.5a, which is constructed with four flanges attached to the base face. In this case, the base face, which is horizontal, is suitable for the unloading the part on the pins of the unloading table. Figure 49.5 shows different stages in unloading the sheet metal.

590

A. Ramesh Babu

Fig. 49.3 Typical unloading table

D C B A 1

2 3 4

5

Fig. 49.4 Indexing of perforated table

From Fig. 49.5c, it can be noted that 10 pins are used for unloading the part. However, if the complexity of the part increases, it is quite difficult to identify the indexing position of pins. The following methodology explains the heuristic method to identify the pin layout. Figure 49.5b, c shows part before and after unloading, respectively.

49 Heuristic Algorithmic Approach for Automatic Generation …

(a)

591

(b)

Base face Gripper (c)

(d)

Fig. 49.5 Unloading a sheet metal part on pin layout. a Base face, b grasping the part with gripper, c placing the part on pin layout, and d pin layout

592

A. Ramesh Babu

49.3 Heuristic Algorithm As explained above, the automatic pin layout generation strongly depends on the geometry of the sheet metal part. It is common practice that the sheet metal part is constructed with one base face and several flanges attached to it subsequently. Further, the sheet metal part may consist of hole and form features. Since it is considered that all the pins are of the same size, it is possible to unload the part on pins only with horizontal faces supported by pins. It is unnecessary to have pins under hole features of the part. The part cannot be supported at inclined faces or curved faces (Rbend). Thus, the scope of this work is limited to unload the parts that have sufficient horizontal faces. A simple representation of the part, pins, and pin layout is shown in Fig. 49.6. To develop the automatic pin layout algorithm, the following assumptions and conditions are considered. Part and robot gripper should not collide with pins. Pins should not collide themselves. All pins are the same size. There must be sufficient horizontal faces in a part. With the above, the algorithm is divided into two stages. Stage 1: Identifies the faces to be considered for supporting by pins and all the possible pins that fall on these faces. Stage 2: Select the best face to be supported and pins needed for it. To understand the first stage of the algorithm, consider the part shown in Fig. 49.7a with horizontal, vertical, inclined, and curved features. Detail steps in the first stage are as follows. Fig. 49.6 Schematic view of a part on pin layout

49 Heuristic Algorithmic Approach for Automatic Generation …

593

(a)

Hole

(b) Layer Layer Layer Layer

(c) Not useful

Table Fig. 49.7 Schematic representation of major stages in generation of pin layout: a part geometry, b layer representation, and c possible pins on each layer

Stage-1: Step 1: Consider horizontal faces (flanges) and ignore inclined and curved flanges. Step 2: Group horizontal faces into layers based on the height of the face. The condition is that the height of all faces in a layer is the same, as shown in Fig. 49.7b. Step 3: Sort the layers in ascending order of height. Step 4: Translate the sheet metal part in such way that bottom left corner of part align with table corner and touching the first layer with pins. Then, from the top it looks as superimposing of faces on the holes of the table as shown in Fig. 49.8. Step 5: Identify the pins that are within the boundary of faces. Offset the part in x, y directions if needed to maximize the number of pins that are within the part faces.

594

A. Ramesh Babu

Table

Centroid

Pin

Convex hull Pin

Fig. 49.8 Superimposed part layer with pin and hole pin positions

Step 6: Step 7:

Superimpose the next layer with pins and identify maximum number of pins that fall within the boundary. Repeat it for all layers. Figure 49.7c shows the pins’ position for all layers.

As explained earlier, pins position on the table with respect to position of the part is important in estimating the pins suitable for unloading. Let us consider that positions of holes under the part are as shown in Fig. 49.7c. Dotted vertical lines are projected lines from the center of holes on the table. With this data and given position of the part, pins (shown as a small arrow) on each layer of the product can be seen. Now the question is which layer of pins to be considered and how many pins to be considered, and which pins to be considered. All these points will be dealt with the second stage of the algorithm as follows with continued step number. Stage-2: Select the best face to be supported and pins needed for it. Step 8: Remove pins that are “not useful” from each layer. Pins that either collide with other flanges or come under the hole features can be deleted. Step 9: Draw the convex hull considering all pins in a layer. Step 10: Calculate the centroid of the part. Step 11: Identify the layers for which the centroid of the part falls within the convex hull correspond to each layer. Step 12: Select the layer that satisfies step 11 and has a maximum number of pins. To illustrate the logic, consider the example shown in Fig. 49.8. Here, the part has a base face with certain larger and very small hole features, and small flanges attached to it. It can be seen the position holes on the table and selected positions of pins with logic explained before. It can be observed that the convex hull formed with these pins, and the centroid of the product falls well into this convex hull and satisfying the condition of stability. In this example, an algorithm chooses only a layer (faces) that has a maximum number of pins. Maximum number can be

49 Heuristic Algorithmic Approach for Automatic Generation …

595

set by the user. In this work, the maximum number of pins is limited to 20. This automatic generation of pin layout took just couple seconds to generate the result, and hence, it is suitable for online unloading scenario.

49.4 Results and Discussion This work is implemented using VC++ development environment with OpenGL graphical display and tested with a variety of parts. This section shows some of the typical results obtained by executing the algorithm proposed in this paper to generate pin layout on the table automatically for a given sheet metal part. Figure 49.9 shows a pin layout for a part with certain hole features. In this picture, the grasping position of the robotic gripper is also shown. It can be observed that the positions of pins in layout form a convex hull in such a way that centroid of the part falls within this convex hull, and thus making the product more stable on the pins. In contrast to the above, Fig. 49.10 shows a larger part that occupies more area of the table. Further, it has a very larger hole feature at the center of the part and leaving narrow material around the boundary. From this result, it can be noted that the pins are arranged along the boundary and the proposed algorithm is successful in generating result where narrow space is available for locating pins. Similarly, Fig. 49.11 shows the pin layout for a sheet metal part with many small holes. These hole features can be ignored while generating the pin layout because their size is very small compared to the diameter of the pin. Pin layout shown in Fig. 49.11 is one such example. Further, it can be noted that the algorithm could able to find 9 pins suitable to support the part. The reason to get such a layout includes the positions of holes on the table, pin diameter, and grasping position of the gripper. In this example, grasping area of the gripper is noticeably larger to influence the layout. Hence, the grasping area of the gripper and its position also influence the pin layout. The present method is also effective in generating the pin layout for very small and very complex parts. From Fig. 49.12, it can be seen that the sheet metal parts are so small that it can be supported with only one pin. The offset of the part on the table is adjusted in such a way that pin is located at the centroid of the product to ensure the stability of the product. In reality, top of the pin is attached with a vacuum pad in order to provide better stability to the product on pins.

596

A. Ramesh Babu

Fig. 49.9 Generated pin layout for a part with hole features: a view-1 and b view-2

49 Heuristic Algorithmic Approach for Automatic Generation …

Fig. 49.10 Pin layout for a part with narrow material boundary: a view-1 and b view-2

597

598

A. Ramesh Babu

Fig. 49.11 Pin layout for a part with many small hole features: a view-1 and b view-2

49 Heuristic Algorithmic Approach for Automatic Generation …

599

Fig. 49.12 Pin support for very small parts: a smaller complex part-1 and b smaller complex part-2

600

A. Ramesh Babu

49.5 Conclusions The proposed method, to generate the pin layout to support the sheet metal parts at unloading station, found to be effective and suitable for simple to complex geometrical sheet metal parts. Further, it can be noted that this kind of pin layout is much suitable for complete automation of sheet metal bending process where a robot loads the flat 2D sheet, bends the part as per bending order, and unloads 3D part at unloading station. Since the product after bending may be having the complex bends, it is not possible to unload on the flat table. In this context, the proposed pin layout on table is found to be effective to unload the part. The proposed pin layout generation process is suitable for unloading the sheet metal parts even in case of the manual bending process, even though it is aimed for the robotic bending process.

References 1. Vijayanand, K., Ramesh Babu, A.: Heuristic and genetic approach for nesting of two-dimensional rectangular shaped parts with common cutting edge concept for laser cutting and profile blanking processes. Comput. Ind. Eng. 80, 111–124 (2015) 2. Amada Sheet Metal Working Research Association: Bending Technique: New Knowhow on Sheet-Metal Fabrication. Machinist Publications, New York (1980) 3. Aomura, S., Koguchi, A.: Optimized bending sequences of sheet metal bending by robot. Robot. Comput. Integr. Manuf. 18, 29–39 (2002) 4. Duflou, J.R., et al.: Computer aided process planning for sheet metal bending: a state of the art. Comput. Indus. 56, 747–771 (2005) 5. Bending Automation: http://www.amada.com. Last accessed 15 Aug 2018 6. TruBend Cell 7000: https://www.trumpf.com. Last accessed 15 Aug 2018

Chapter 50

Voxel-Based Strategy for Efficient CNC Machining A. Kukreja , H. D. Mane , M. Dhanda

and S. S. Pande

Abstract Complex industrial parts such as dies and molds are produced on multiaxis CNC machines. Efficiency and utilization of such machines are dictated to a great extent by the availability of efficient part programs to carry out various roughing and finishing operations. This paper presents a novel approach for the tool path planning of roughing operation of complex machining features using voxelization. The developed system takes the CAD part model in STL format as the input. Machinable and non-machinable regions are identified using voxelization followed by the generation of isoplanar zig-zag roughing tool path for 3-axis milling machine. The system was tested for different parts with varying complexities. The developed voxel-based tool path strategy was found to be generic, computationally efficient, and productive compared to those in commercial software. Keywords CNC rough machining · Tool path planning · STL · Voxelization

50.1 Introduction Complex freeform surfaces used in industrial products like dies and molds are machined on multi-axis (3–5) CNC milling machines. Efficiency and utilization of such machines are dictated to a great extent by the availability of efficient part programs to carry out various roughing and finishing operations. Rough tool planning is thus a major step in the process planning for CNC machining. Current tool path planning strategies involve recognizing machining features either by automatic feature recognition technique or by selecting geometric entities manually during process planning in CAM systems [1]. This makes the rough tool path planning process computationally complex, time-consuming, and less efficient. A need, thus, exists to address this issue.

A. Kukreja · H. D. Mane · M. Dhanda · S. S. Pande (B) CAM Lab, Department of Mechanical Engineering, Indian Institute of Technology Bombay, Mumbai 400076, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_50

601

602

A. Kukreja et al.

Literature documents various tool path planning methods for roughing operation. Popescu et al. [2] developed a rough milling tool path on point data based on graph theory. The cutting areas are identified using bicolored binary map, and path planning is done using the Dijkstra algorithm. The procedure generated verbose code and was computationally complex. Han et al. [3] propose an algorithm for generating roughing and finishing tool paths on measured data directly by using moving least squares (MLS) method. Roughing is performed in a slice-by-slice manner to remove material rapidly. Makki et al. [4] proposed an area-wise segmentation and overlapping method for 5-axis machining of rough clouds of points. The method is applied with success to the direct duplication of a human femur in a reasonable computation time. However, its efficiency depends on the technique of cloud segmentation and overlapping. Olam et al. [5] reported on converting a 2D-shaded pixel graphic into tool path for machining on a CNC milling machine. The technique was quite primitive in nature. Tchantchane and Bey [6] proposed a methodology permitting the automation of the process by generating the Z-map model and parallel plane machining strategy. This methodology reduces the product development cycle by eliminating the reconstruction of the CAD model. However, selection of an optimum tool radius for each cutting plane and tool path optimization is still a concern. The voxel-based CAD model representation, display, and thickness analysis of Intricate shapes were reported by Patil and Ravi [7]. The binary voxel model is stored as a stack of layers represented as bit arrays and coupled to traditional CAD/CAM systems providing additional capabilities to product engineers such as modeling heterogeneous objects, compact memory usage, and computation time. Srinivasan and Pande [8] reported the integrated system for additive manufacturing (AM) process. The system takes the CAD model (STL) as input which is voxelized and hollowed. Then the optimal part orientation using GA-based strategy was found out, and support structure was generated. The methodology was found to be generic and can be used to produce AM parts with significant improvement in part quality, productivity, and material utilization. For tool path planning, Tarbutton et al. [9] explored the use of the voxel-based model for tool path planning. Cutter contact points were generated using the ray tracing concept. Literature tends to suggest that a hybrid scheme that combines voxel and surface-based models can be explored to improve the efficiency of algorithms. Majority of tool path planning algorithm for roughing uses slicing technique using Z-map model or pixel data. Hardly any work has been reported for rough tool path planning by directly using voxel-based CAD model. This paper reports the design and implementation of a voxel-based rough tool path planning strategy for 3-axis CNC machining. CAD model in STL format is initially converted into the voxelbased format and used to generate isoplanar tool paths. In what follows, various functional modules of the system are discussed one by one.

50 Voxel-Based Strategy for Efficient CNC Machining

603

50.2 Methodology The task of recognizing features and generating tool path is computationally expensive and inefficient. The proposed methodology (Fig. 50.1) tries to eliminate the same by converting the CAD part model into binary voxel data without actually recognizing any features. Voxelized CAD model is then pre-processed to provide stock-to-be-left for finish machining. Blind (inaccessible) features, if any, that are impossible to be machine using 3-axis CNC milling machines are also removed during pre-processing. The modified voxelized model is then processed to generate CL (cutter location) data in order to perform rough machining operation layer by layer, followed by post-processing to generate the NC file.

50.3 Voxelization Voxelization is concerned with converting geometric objects from their continuous geometric representation into a set of voxels that best approximates the shape. In this work, we have used binary voxels (presence {gray} or absence {white} of material) (Fig. 50.2). The voxel is cuboidal in shape and aligned with the Cartesian coordinate system. To ensure quick access to the voxel information coupled with storage efficiency, the solid model is represented by a stack of plane layers. Each layer is represented by an array of voxels. The index of a voxel in the array is calculated from Fig. 50.1 Modular diagram

604

A. Kukreja et al.

Fig. 50.2 Voxels’ representation

its X, Y, and Z coordinate values. A MATLAB function file ‘VOXELISE’ was used [10]: [OUTPUTgrid, gridCOx, gridCOy, gridCOz] = VOXELISE (xd, yd, zd, STL, ‘xyz’); The function takes STL and xd, yd, zd as inputs, where xd, yd, and zd are the number of voxels in the x, y, and z directions, respectively; this number depends on the size of voxels that can be different in different directions. The extremities of the STL file form the cuboidal raw stock. The cuboid is voxelized, and the voxels coming inside the part volume (non-machinable region) are marked as ‘1’ and all other voxels as ‘0’. That is saved in OUTPUTgrid. The location of voxels is specified by gridCOx, gridCOy, and gridCOz in x, y, and z coordinates. For convenience, voxels marked as ‘1’ and ‘0’ would be referred as vx‘0’ and vx‘1’ in the subsequent sections. Voxelized data is then passed to the pre-processing module for further analysis.

50.4 Pre-processing Pre-processing module consists of two steps, namely offsetting and voxel projection which are described as follows.

50.4.1 Offsetting Offsetting on each voxel layer of the part model is carried out to leave the stock allowance for the finishing cut. The cutter path is computed from the offset voxels using the CL (cutter centerline) method of programming. Offsetting in the layer is done by expanding the non-machinable region. The voxels in the vicinity of all vx‘1’ are remarked as ‘1’ as shown in Fig. 50.3. The number of voxels to be offset is

50 Voxel-Based Strategy for Efficient CNC Machining

(a) Initial Voxels

605

(b) Offset Voxels

Fig. 50.3 Offsetting of voxels

calculated from Eq. (50.1). No = St /w

(50.1)

where N o represents a number of voxels to be offset, S t is the stock allowance, and w is the width of a voxel. The size of the voxel is chosen 4 mm (two-thirds of the tool diameter) in this work. This, in essence, takes care of the CL point computation for the tool. As shown in Fig. 50.3b, the tool is placed at the center of the voxel (specified as cutter location). This would imply the engagement of the tool with the offset voxel to be one-fourth of the voxel size and hence leaving 3 mm (3/4 of the voxel size) as stock allowance. The voxel size is automatically chosen by our system once the tool diameter is input by the user. These proportions can, of course, be tuned for any specific requirements.

50.4.2 Voxel Projection Voxel projection along the z-axis is done to avoid machining of blind features. The upper voxel layer data is superimposed on the subsequent lower voxel layer to fill the blind areas. In our algorithm, it is done as per Eqs. (50.2) and (50.3). Temp_vxi+1 = vxi + vxi+1 vxi+1 = ∀ x, y, z, Temp_vx(x, y, z)i+1 ‘2’ → Temp_vx(x, y, z)i+1 ‘1’

(50.2) (50.3)

where vxi refers to the voxel layer i, Temp_vx(x, y, z)‘2’ as temporary voxel which gets marked as ‘2’ because of the operation performed using Eq. (50.2). This observation is done for each layer for subsequent tool path planning. The voxel projection is graphically shown in Fig. 50.4.

606

A. Kukreja et al.

Fig. 50.4 Projection of voxels

50.5 Tool Path Generation Two kinds of tool path strategies are developed in this work to machine the part using the voxelized data as input, and NC file is generated as an output. The following sections explain them in detail.

50.5.1 Tool path with Lift To machine the part, the tool is moved only over vx‘0’. This strategy covers these voxels, layer by layer in the z-direction. Voxels of one layer are traversed in zig-zag parallel tool paths from one extremity to another. Forward and side steps are taken along X-direction and Y-direction, respectively. The main steps of the algorithm used for tool path with lift strategy are mentioned below: • Start moving in the positive x-direction from first voxel (0, 0, 1) of the top layer while checking subsequent voxels, and if vx‘1’ encountered, lift the tool to the feed plane and keep moving in the feed plane until vx‘0’ is found. • If the last voxel in the x-direction is reached, do ramp off. • Move to the next y layer (i.e., side step), and start moving in the negative x-direction and follow the same rule for lift. • Keep following zig-zag pattern till y extremity is reached. • Move to the starting voxel in the next z layer, and repeat steps 1–4. • Repeat until last z layer is reached. Figure 50.5 shows the tool path with lift for one z layer of a typical part model having two rectangular protrusions.

50 Voxel-Based Strategy for Efficient CNC Machining

607

Fig. 50.5 Tool paths with lift

50.5.2 Tool Path with Minimal Lift In order to minimize the air-cutting, a new algorithm is proposed. The algorithm creates the zig-zag tool path without lifting the tool considering one layer at a time. If the tool gets stuck somewhere in between, it lifts the tool, identifies the remaining area to be machined, and goes to the starting point of the remaining non-machined area. The main steps of the algorithm used for tool path with minimal lift strategy are as follows. • Start moving in the positive x-direction from voxel (0, 0, 1) while checking subsequent voxels, and if vx‘1’ encountered, find vx‘0’ in the vicinity of the current voxel in the next y layer and move to that voxel. • If there is no vx‘0’ in the vicinity, then lift the tool to the feed plane and move to the starting point of next y layer where vx‘0’ is found. • Keep following the zig-zag path without lifting the tool until the y extremity is reached. • Find the remaining non-machined area (trapped area). • Repeat steps 1–4 until all the machining area is removed. • Move to the starting voxel in the next z layer, and repeat steps 1–5. • Repeat until the last z layer is reached. To better explain the concept, a simple part is taken as an example. Figure 50.6 shows tool paths for different regions (1, 2, and 3), and while creating tool path on region 1, it goes on finding the machining areas without lifting the tool until it reaches Y extremities. So this creates regions 2 and 3 as trapped areas which are machined by the minimal lifting of the tool. The process of tool path planning goes on until all voxels are visited once.

608

A. Kukreja et al.

Fig. 50.6 Tool paths with minimal lift

50.6 Post-processing Tool path generation module gives CL data that is processed to generate an NC file. The CL path so generated is post-processed to ISO neutral as well as for FANUC (G/M) format to create the final CNC part program.

50.7 Results and Discussion The algorithm is implemented using Matlab2015b on windows platform. This section describes the different case studies. Finally, the results are compared with those obtained from the commercial software, Mastercam X5. The tool path is created using three different strategies, viz. by lifting the tool, with minimal lift and using Mastercam roughing. For all the case studies, feed rate (159.1 mm/min), depth of cut (0.5 mm), tool diameter (6 mm, flat end mill), stepover (4 mm), and stock to leave (3 mm) are kept same. Results are presented below.

50.7.1 Case Study 1 The developed system takes input as closed volume part model STL (Fig. 50.7a). The closed volume part model containing various features such as cuboid, cylinder, and rotated cuboid placed arbitrarily on the top of the flat base to create a variety of trapped regions intentionally with varying heights. Part model is then voxelized, and

50 Voxel-Based Strategy for Efficient CNC Machining

609

(a) STL part model

(b) Cut stock with lift

(c) Cut stock with minimal lift

(d) Cut stock by Master CAM

Fig. 50.7 Case study 1

CL data is post-processed using FANUC OM control and to generate CNC tool path. Figure 50.7b–d shows the typical tool path for one layer of rough-cut stock with lift, with minimal lift and using Mastercam X5 roughing, respectively, generated using simulation software VERICUT 8.0. The green and blue color tool paths show rapid motion and linear feed motion of the tool.

50.7.2 Case Study 2 To check the robustness of the developed system, another complex part model with freeform surface (ffs) features is considered (Fig. 50.8a). All the parameters were kept the same as earlier. This case study aims to check the ability of the developed system to maintain the form of the surface. Figure 50.8b–d shows the tool path for one layer for rough-cut stock of freeform surface with lift, with minimal lift, and using Mastercam X5 roughing, respectively, generated using simulation software VERICUT 8.0. In the developed strategy, the staircase effect was observed (Fig. 50.8b, c) because the machining is performed layer by layer. As can be shown in all Fig. 50.8b–d, some part of the material is left uncut which is due to stock to leave option enabled in each case.

610

A. Kukreja et al.

(a) STL ffs model

(c) Cut stock of ffs with minimal lift

(b) Cut stock of ffs with lift

(d) Cut stock of ffs with Mastercam

Fig. 50.8 Case study 2

Results of the developed system/strategies are compared with those from commercial software Mastercam X5 (Table 50.1). The comparison between strategies is done based on machining time and code length in no. of blocks. For case 1, the developed strategy with minimal lift was found to be more efficient (16.23%) than with lift and (16.72%) than the Mastercam roughing strategy in terms of machining time. In terms of code length, developed strategies are giving large values as compared to Mastercam because new NC block is created for each visited voxel. This can, of course, be minimized by further postprocessing to join the voxels together along a linear path to form a single block. Table 50.1 Comparison of developed strategies with Mastercam Machining strategy With lift With minimal lift Mastercam X5

Machining time (h: min: s)

Code length (no. of blocks)

Case 1

Case 2

Case 1

5:05:08

85:00:40 71:20:47 85:50:59

Case 2

223 ×

103

19 × 103

3:53:50

171 ×

103

10 × 103

7:37:32

59 ×

103

13 × 103

50 Voxel-Based Strategy for Efficient CNC Machining

611

Fig. 50.9 Machined foam component

Similarly, for case 2, the developed strategy with minimal lift was found to be 23.6% more efficient than with lift and 49.05% better than with the Mastercam strategy. For this case, the code length of the strategy with minimal lift was the least. The simulated components from the developed system show staircase effect in the rough stock. The amount of material left as stairs depends on the depth of cut and slope of the machined surface. This is subsumed in the stock allowance and will be removed during finishing cut. Voxel size can be properly chosen to be small to control the stair step errors under tolerance in the finishing cut. A machining trial was carried out on a 3-axis Denford CNC milling machine with a FANUC controller to validate the NC program generated by our system. The raw material was PU foam block of 100 × 100 × 50 mm. The cutting tool used was 6.35 mm (0.25 ) flat end mill. Figure 50.9 shows the machined component with a shape conforming to the CAD model for case study 2 with minimal lift. The developed system has thus been validated successfully.

50.8 Conclusions In this work, a novel voxel-based rough machining algorithm has been developed and implemented. The present methodology is generic can be used for any close-bound STL part model with different machining features. As compared to the commercial software, the developed system does not need to recognize any feature to generate a roughing tool path. The algorithm uses the voxelization technique to differentiate between machinable and non-machinable areas without using surface properties, which reduces the computational complexity. Case studies with various machining features show the robustness of the developed system with a reduction in machining time and code length. Future work would include a selection of the optimum voxel size for each cutting plane and tool path optimization.

612

A. Kukreja et al.

References 1. Hou, M., Faddis, T.N.: Automatic tool path generation of a feature based CAD/CAPP/CAM integrated system. Int. J. Comput. Integr. Manuf. 19, 350–358 (2006) 2. Popescua, D., Popistera, F., Popescua, S., Neamtua, C., Gurzaua, M.: Direct tool path generation based on graph theory for milling roughing. Procedia CIRP 25, 75–80 (2014) 3. Han, Y.A., Zhang, Y.J.: Tool path generation on measured data using moving least squares method. Comput. Integr. Manuf. Syst. CIMS 17, 638–642 (2011) 4. Makki, M., Tournier, C., Thiebaut, F., Lartigue, C., Souzani, C.: 5-axis direct machining of rough clouds of points. Comput. Aided Des. Appl. 7, 591–600 (2013) 5. Olam, M., Sanliturk, I. H., Tosun, N.: Converting a pixel graphic into tool path for machining on a CNC milling machine. In: Proceedings of the International Conference on Advances in Information Processing and Communication Technology, pp. 59–62, Turkey (2016) 6. Tchantchane, Z., Bey, M.: Toward the roughing of sculptured surfaces from a regular cloud of points. In: Proceedings of 13th International Research/Expert Conference on Trends in the Development of Machinery and Associated Technology, pp. 16–21, Hammamet, Tunisia (2009) 7. Patil, S., Ravi, B.: Voxel-based representation, display and thickness analysis of intricate shapes. In: Proceedings of the Ninth International Conference on Computer Aided Design and Computer Graphics, pp. 415–422. IEEE Computer Society, Washington, DC, USA (2005) 8. Srinivasan, U., Pande, S.S.: Efficient process planning strategies for additive manufacturing. In: Proceedings of the ASME 2017 12th International Manufacturing Science and Engineering Conference, Los Angeles, CA, USA (2017) 9. Tarbutton, J.A., Kurfess, T.R., Tucker, T.M.: Graphics based path planning for multi-axis machine tools. Comput. Aided Des. Appl. 7(6), 835–845 (2016) 10. Mesh Voxelization: http://www.mathworks.com/matlabcentral/fileexchange/27390-meshvoxelisation. Last accessed 20 May 2018

Chapter 51

Effect of Geometrical and Process Parameters on Utilization of Sheet Material in Plasma and Laser Cutting Processes N. Venkatesh , S. Sabari Sriram , V. Satish Chandran and A. Ramesh Babu Abstract Nesting, an optimal arrangement of 2D sheet metal blanks on a large sheet, is a very important CAM activity in sheet metal industry since it influences the sheet material wastage significantly. In this work, the effect of part geometry, part size, part quantity and cutting process parameters on material wastage while nesting of two-dimensional shapes in rectangular sheets is analyzed. The nested patterns are generated by a professional commercial package, and material wastage is captured by varying the part geometry, size and quantity. The results are analyzed for laser and plasma cutting processes by considering corresponding parameters such as kerf width, bridge width, part-to-sheet distance, lead-in and lead-out distances. The results are compared with data available in the literature. Keywords Sheet metal cutting · 2D nesting · Material utilization · Laser cutting · Plasma cutting

51.1 Introduction With increasing demand for sheet metal products such as enclosures and cabinets in various applications such as machine tools, automobiles, compressors, defense, aerospace, railways and domestic products, sheet metal machinery manufacturers face several challenges in delivering the products with shorter lead time, less cost and high quality. These types of sheet metal products are produced with cutting followed by bending processes. Initially, 3D sheet metal product is created in CAD system by assembling several sheet metal parts. Each 3D part is unfolded to obtain a 2D flat profile. These 2D profiles are arranged on large sheet, known as nesting, and then cut with blanking or laser cutting or combination of these processes. Subsequently, 2D blanks are subjected to bending process to make 3D parts.

N. Venkatesh · S. Sabari Sriram · V. Satish Chandran · A. Ramesh Babu (B) Department of Mechanical Engineering, PSG College of Technology, Coimbatore 641004, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_51

613

614

N. Venkatesh et al.

Nesting is an important task in sheet metal cutting industry since the less efficient nesting process results in more material wastage. Generally, sheet metal wastage is represented in terms of percentage of scrap. Efficiency of nesting process depends on several parameters including part geometry, sheet geometry, part size, sheet size, process parameters of cutting process, arrangement strategy and batch size, etc. Since the geometry of parts may vary from simple rectangular shapes to highly complex shapes, several researchers proposed various algorithms to generate the nested pattern. Sherif et al. proposed a method using simulated annealing algorithm to optimize the tool path in addition to material utilization, for laser cutting process [1]. In this work, a professional nesting suite is used to obtain the material utilization and observed that it is around 55–60% considering the part’s size around 100 mm × 100 mm. This is due to rectangular approximation of parts, and justification for above utilization is not provided. Considering part sizes in range of 100 mm × 100 mm to 500 mm × 500 mm and kerf and bridge allowances as 0.5 mm and 4 mm, respectively, Vijayanand and Ramesh Babu proposed genetic algorithmic (GA)-based rectangular nesting approach with common cutting edge for laser cutting process and found that pattern efficiency is around 90% [2]. In contrast to rectangular nesting problems, Elkeran proposed cuckoo search combined with guided local search algorithms to arrange irregular polygons in rectangular sheet [3]. However, this nesting approach is based merely on geometry, and process is not considered. It is reported that 70–90% utilization is observed with variation in geometry. Similar to above geometry-based methods, Siasos and Vosniakos applied genetic algorithm as an optimization tool combined with bottom-left-fill-left (BLFL) approach to address irregular shape nesting problem [4]. Here, irregular parts are converted into raster form and then precision movement is carried out with shape points. Further researchers explored algorithms for nesting of irregular parts in rectangular or irregular boundaries suitable for garment and leather industries [5–7]. Though geometric parameters were considered in algorithms proposed by several researchers for 2D nesting of sheet metal parts for optimal utilization of sheet material, process parameters such as kerf width, bridge width, part-to-sheet distance, lead-in and lead-out distances are not included. These parameters are very significant to arrive at the actual utilization of sheet material while cutting. The analysis proposed in this work is unique, and it helps sheet metal part manufacturers to find suitable strategy to nest the parts in order to minimize the wastage of sheet material.

51.2 Process Parameters In this work, the cutting parameters, viz. kerf, part-to-part distance (bridge width), part-to-sheet distance, lead-in and lead-out distance are considered, for a fiber laser and hypertherm-45 plasma processes, while estimating the material wastage. For example, consider a simple arrangement of squares of 100 mm × 100 mm size on a rectangular sheet of 6 mm thickness as shown in Fig. 51.1. Considering the process

51 Effect of Geometrical and Process Parameters on Utilization …

615

Fig. 51.1 Simple nested pattern with process parameters

parameters for plasma cutting process: bridge width = 4.8 mm, part-to-sheet distance = 4 mm, lead-in = 8 mm and lead-out = 4.8 mm, minimum size of required rectangular sheet = 333.6 mm × 112.8 mm, as per the layout given in Fig. 51.1. Hence, there will be a scrap of 20.2%. This result is generated by commercial nesting software considering the parameters recommended by the industry. If the nesting is carried out considering only part geometry without considering the process parameters, expected sheet size = 300 mm × 100 mm and scrap = 0%, which is not feasible in practice. The above initial study motivated to analyze the effect of process parameters on wastage of sheet material, in practical situation, for different geometrical shaped parts with varying sizes, complexity and order quantity. At first, data for various process parameters is collected. Figures 51.2, 51.3, 51.4, 51.5 and 51.6 show values of these parameters for different thicknesses of mild steel material for typical plasma and laser cutting processes. These values are proven values and being considered in industry (Courtesy: Messer Cutting Systems India Private Limited). It is well known that laser or plasma beam burns away a portion of the material when it cuts through the material, and width of this material is known as kerf width.

Fig. 51.2 Kerf values

616

Fig. 51.3 Part-to-part distance (bridge width)

Fig. 51.4 Part-to-sheet distance

Fig. 51.5 Lead-in distance

N. Venkatesh et al.

51 Effect of Geometrical and Process Parameters on Utilization …

617

Fig. 51.6 Lead-out distance

The kerf width depends on the thickness of the material as well as cutting process as shown in Fig. 51.2. It can be observed that laser cutting process needs less kerf width in the range of 0.2–0.4 mm compared to that of plasma cutting process which needs higher kerf values in the range of 0.6–1.6 mm, for the mild steel material in the thickness range of 0.5–8 mm. During the nesting process, it is necessary to place the parts with certain gap from the sheet boundary as well as among the parts to avoid distortion of the material and collision with cutting head. The recommended values for part-to-part distance and part-to-sheet distance are shown in Figs. 51.3 and 51.4, respectively. It can be observed that these values are in the range of 1.5–5 mm. Even though kerf width is very less for laser cutting process, part-to-part distance and part-to-sheet distance values significantly influence the wastage of sheet material. Further, while cutting with laser or plasma, the jet has to pierce at some distance (lead-in) away from the part profile, follow the profile and exit the path at certain distance (lead-out) away as shown in Fig. 51.1, in order to ensure the edge quality of the part. Recommended lead-in and lead-out values are in the range of 2–8 mm as shown in Figs. 51.5 and 51.6, respectively. From the above data, it can be observed that, except kerf values, there is less difference in the values of part-to-part distance, part-to-sheet distance, lead-in and lead-out distance for both laser and plasma cutting processes. These process parameters affect wastage of sheet material significantly, as a simple case shown in Fig. 51.1. In this paper, it is focused to analyze this effect as explained in the following section.

51.3 Evaluation of Nested Patterns At first, efficiency of the nested pattern is analyzed, by considering the different geometrical shapes of 100 mm × 100 mm boundary. These geometrical shapes are square (Sq1-100), circle (Crc1-100), hollow square (Hosq1-100), circle with holes (Hoci1-100) and typical irregular part (0018–100) as given in Fig. 51.7. Considering the quantity each part equal to 100, nested patterns are generated for varying

618

N. Venkatesh et al.

Fig. 51.7 Part data

sheet thickness, considering the plasma cutting process, and results are presented in Fig. 51.8. From Fig. 51.8, it can be observed that percentage of scrap increases with complexity of part geometry, and it is in the range of 18–51% (approx). It is clear that percentage of scrap is around 50% for hollow square part (Hosq1-100) compared to that of square shape (Sq1-100) that produces around 20% scrap. It is evident that a hollow portion of the part-Hosq1-100 cannot be utilized and results in higher material wastage. Further, it can be noted that there is around 5% variation in sheet material wastage for sheets of 1–6 mm thickness. This is due to the fact that thicker sheet material needs higher values of part-to-part distance, part-to-sheet distance, lead-in and lead-out distance, compared to that of thinner sheets.

51 Effect of Geometrical and Process Parameters on Utilization …

619

Fig. 51.8 Wastage of the sheet material for different geometrical shapes of single variety and fixed quantity

Fig. 51.9 Wastage of the sheet material for different geometrical shapes of single variety with varying sizes and fixed quantity

To analyze the effect of variation in part size on material wastage for different geometrical objects, two varieties of parts including hollow square and perforated circle of 6 mm thickness with varying sizes are considered. These parts include Hosq1-100, Hosq2-120, Hosq3-140, Hosq4-175, Hosq5-180 and Hoci1-100, Hoci2120, Hoci3-140, Hoci4-175, Hoci5-180, as given in Fig. 51.7. Sizes of these parts (side of hollow square or diameter of perforated circle) vary from 100 to 180 mm. The batch size is taken as 100. The percentage of scrap while nesting these parts is shown in Fig. 51.9. From this, it can be observed that there is around 7% reduction in material wastage while increasing the part dimension from 100 to 180 mm.

51.4 Results and Discussion From the above evaluation, it is clear that wastage of sheet material while nesting 2D parts on 2D sheet depends on several parameters including part geometry, part/sheet size, process parameters. In this work, nested patterns are generated by commercial software (OmniWin) considering the process parameters part-to-part distance, partto-sheet distance, lead-in and lead-out distance from industrial use (courtesy: Messer

620

N. Venkatesh et al.

Table 51.1 Sheet material wastage during nesting for variation in part geometry, size, thickness Cutting process: Laser cutting and plasma cutting part quantity =100 Process parameters: Kerf width: 0.2–0.4 mm (laser) 0.6–1.6 mm (plasma) Bridge width: 2–5 mm Part-to-sheet distance: 1.5–4 mm Lead-in: 3–8 mm Lead-out: 2–6 mm

Sheet thickness (mm)

Part size

Part geometry

% of scrap (in the range)

1–6

100 mm × 100 mm

Simple square

Around 20

Complex part (level 2)

Around 50

100 mm × 100 mm to 180 mm × 180 mm

Complex part (level 1)

42–35

Complex part (level 2)

50–45

6

Cutting Systems India Private Limited). The consolidated results from the above data and evaluation process are presented in Table 51.1. From Table 51.1, sheet material wastage during nesting for variation in part geometry, size and thickness can be observed. It can be seen that complexity of part geometry has more effect on wastage of sheet material for a given size of parts. Complex parts may result in around 50% of wastage compared to that of simple shapes like square that generate around 20% wastage, for given size of parts of 100 mm × 100 mm boundary. This 20% loss is due to the bridge width, lead-in and lead-out. In addition to this, nesting logic also matters a lot. Similarly, part size also has certain influence on material wastage. While increasing the part size from 100 mm × 100 mm to 180 mm × 180 mm, it is observed that material wastage is reduced by around 5–7%, when single variety of parts with same quantity is nested. To understand the influence of different geometrical shapes with different quantities, on material wastage while nesting, around 10 sets of parts as shown in Table 51.2 are considered. Part IDs in this table are corresponding to the parts given in Fig. 51.7. The percentage of scrap for each set of parts is shown in Fig. 51.10. It can be observed that percentage of scrap is in the range of 37–52%. The trend is similar to the results presented in Table 51.1 where the single variety of parts is considered. Further, it can be noted from the nested patterns shown in Figs. 51.11, 51.12 corresponding to set 6 and set 1, when parts-in-parts nesting is possible, material wastage can be minimized. In the case study observed (Figs. 51.11, 51.12) around 15% of reduction in the scrap is observed with parts-in-parts nesting compared to case where parts cannot fit into other parts.

1

5

6

4

2

3

12

1

8

10

6

6

3

0006

0007

0008

0009

0010

0011

0012

0013

0014

0015

0016

0017

0019

37.53

9

0005

% Scrap

2

0004

6

7

0003

100

3

0002

Total no. of parts

6

0001

0020

Set1

Part ID

38.35

100

2

2

5

8

8

2

5

12

5

7

8

4

5

6

7

6

1

2

5

Set2

35.06

100

4

4

8

3

6

7

2

13

6

4

1

5

7

9

2

4

8

6

1

Set3

42.15

100

3

6

5

4

3

1

3

12

10

6

5

7

4

8

6

2

1

5

9

Set4

Table 51.2 Different sets of parts with different quantity and complexity

40.92

100

7

3

3

9

13

6

6

8

4

7

4

1

3

2

8

5

6

2

3

Set5

52.7

100

*

*

*

*

*

*

*

*

*

*

*

*

*

*

20

20

20

20

20

Set6

48.74

100

*

*

*

*

*

*

*

*

*

20

*

20

*

20

*

20

*

20

*

Set7

48.51

100

*

*

*

*

*

*

*

*

*

*

20

*

*

20

20

*

*

20

20

Set8

48.0

100

*

*

*

*

*

*

*

*

*

20

*

*

20

20

*

*

20

20

*

Set9

47.04

100

*

*

*

*

*

*

*

*

*

*

20

*

20

*

20

*

20

*

20

Set10

51 Effect of Geometrical and Process Parameters on Utilization … 621

N. Venkatesh et al.

Scrap (%)

622 60 50 40 30 20 10 0

set1

set2

set3

set4

set5

set6

set7

set8

set9

set10

Fig. 51.10 Wastage of sheet for different sets of parts with different quantity and complexity as given in Table 51.2

Fig. 51.11 Nested pattern corresponding to the data of set 6 given in Table 51.2 (scrap = 52.7%)

Fig. 51.12 Nested pattern corresponding to the data of set 1 given in Table 51.2 (scrap = 37.5%)

51 Effect of Geometrical and Process Parameters on Utilization …

623

Table 51.3 Comparison of sheet material utilization with literature data Instance

Nesting method or algorithm

Process parameters

Part size

Part geometry

% Utilization

Sherif et al. [1]

Not available (commercial software)

Not available

Around 100 mm × 100 mm

Irregular shapes with rectangular approximation

55–60

Vijayanand and Ramesh Babu [2]

Genetic algorithm combined with bottom-left heuristic and common cutting edge concept

Kerf width: 0.5 mm Bridge width: 4 mm

100 mm × 100 mm to 500 mm × 500 mm

Rectangular shapes

Around 90

Present work

Not available (commercial software)

Refer Table 51.1

Around 100 mm × 100 mm

Simple to irregular shapes

80–50

To compare the results with data available in the literature, two instances are considered as shown in Table 51.3. From this, it can be observed that percentage of utilization is around 50% while nesting irregular shapes of around 100 mm × 100 mm boundary size as given in the present work as well as the results given in Sherif et al. work. In both cases, nested layouts are generated using the commercial software, and hence, the algorithms used for nesting are not known. For same size parts, if the geometry is simple rectangular shape, around 80% of material utilization is observed, whereas 90% sheet material utilization is observed with Vijayanand and Ramesh Babu approach, where the parts’ sizes are taken in the range of 100 mm × 100 mm to 150 mm × 500 mm with common cutting edge concept. However, here lead-in and lead-out are not considered.

51.5 Conclusions In this paper, cutting of two-dimensional sheet metal blanks from rectangular sheets using laser and plasma cutting processes is considered and wastage, i.e., scrap, of sheet material is analyzed while generating the nested layout. It is observed that the sheet metal wastage, while nesting, depends on several parameters including type of nesting algorithms, part’s geometry, part’s size, part’s quantity, sheet size, cutting process, etc. In general, many researches proposed several nesting strategies using optimization heuristics and algorithms by considering only the geometry of parts. In fact, cutting process has significant impact on efficiency of the nested pattern. This

624

N. Venkatesh et al.

is due to process parameters such as kerf width, sheet-to-part distance, part-to-part gap, lead-in and lead-out positions, which affect the nested pattern. It is observed that about 2–5 mm distance is required between two parts while cutting them, depending on thickness of the sheet, in order to avoid the distortion and collision with nozzle. Similarly, part-to-sheet distance has to be about 1.5–4 mm. In addition to this, laser or plasma jet has to pierce into the material before cutting along the part profile. Hence, it has to pierce certain distance away from the cut path to get smooth cut edge. This distance is known as lead-in distance. Similarly, cutting beam exits the tool path at a certain distance away from the profile, referred as lead-out. As per industry practice, the lead-in and lead-out values are about 2–8 mm. All these process parameters contribute to efficiency of the nested pattern. With the above, it is clear that nesting is the critical task, and if proper approach is not adopted, it results in more material wastage and subsequently more investment on raw material to sheet metal product manufacturers. Acknowledgements The authors express their sincere thanks to M/s. Messer Cutting Systems India Private Limited for providing necessary data and inputs.

References 1. Sherif, S.U., et al.: Sequential optimization approach for nesting and cutting sequence in laser cutting. J. Manufact. Syst 33, 624–638 (2014) 2. Vijayanand, K., Ramesh Babu, A.: Heuristic and genetic approach for nesting of two-dimensional rectangular shaped parts with common cutting edge concept for laser cutting and profile blanking processes. Comput. Ind. Eng. 80, 111–124 (2015) 3. Elkeran, A.: A new approach for sheet nesting problem using guided cuckoo search and pairwise clustering. Eur. J. Oper. Res. 231, 757–769 (2013) 4. Siasos, A., Vosniakos, G.C.: Optimal directional nesting of planar profiles on fabric bands for composites manufacturing. CIRP J. Manufact. Sci. Technol. 7, 283–297 (2014) 5. Martinez-Sykora, A., et al.: Matheuristics for the irregular bin packing problem with free rotations. Eur. J. Oper. Res. 258, 440–455 (2017) 6. Baldacci, R. et al.: Algorithms for nesting with defects. Discret. Appl. Math. 163, 17–33 (2014) 7. Alves, C., Brás, P., de Carvalho, J.V., Pinto, T.: New constructive algorithms for leather nesting in the automotive industry. Comput. Op. Res. 39, 1487–1505 (2012)

Chapter 52

Development of Manufacturability Indices for Prismatic Parts Manish Kumar Gupta , Pramod Kumar Jain and Abinash Kumar Swain

Abstract This paper describes a computer-aided tool for the quantitative evaluation of manufacturability of prismatic machining parts. The manufacturability of a part is expressed in terms of relative manufacturability indices of its constituting features. The present work considers the geometrical aspects of the designed product along with manufacturing issues at a very early stage of design. Geometrical and technological complexity of the design is established using several parameters such as feature intricacy, tool access direction, feature face orientation, feature accessibility, approach direction depth, feature neighbourhood, feature hierarchy, parent and child feature complexities, tolerances, surface finish, tooling complexities which affect the manufacturability of a feature on the part directly or indirectly. Best worst method (BWM) is used to assign weights to manufacturability parameters to reflect their relative importance. A case study is presented to show the capability of the system to generate sound indices which could make designs easier to manufacture without compromising on the functional requirements. Keywords Manufacturability · Geometrical complexity · Technological complexity · Prismatic parts

52.1 Introduction Design for manufacturability is the process of concurrently considering the design goals and manufacturing constraints imposed by the manufacturing engineer. This helps to identify and eliminate the manufacturing problems, while the product is being designed. Broadly, two approaches, i.e. rule-based and rating-based, are reported to access manufacturability. The rule-based approaches evaluate a design for its feasibility based on the rules embedded in the system and provide redesign suggestions as feedback to the designer. These approaches do not compare designs M. K. Gupta (B) · P. K. Jain · A. K. Swain Department of Mechanical and Industrial Engineering, Indian Institute of Technology Roorkee, Roorkee 247667, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_52

625

626

M. K. Gupta et al.

for their manufacturability and useful for near-net shape manufacturing. Ratingbased approaches are quantitative which provide manufacturability indices based on product cost, quality and production times. These approaches are used to compare alternative designs/process plans based on a specific objective. Some of the reported rating-based DFM systems generate time and cost measures of manufacturability [1]. The output of these systems is a rating based on cost estimation considering different manufacturing costs. Estimating the cost accurately requires detailed information on manufacturing processes and resources which is time-consuming. Also, these measures may not be directly helpful for determining if the designer has achieved a satisfactory level of manufacturability in the design. Manufacturability of the designed part can be estimated through evaluation of manufacturing difficulty and design intent or function of the individual feature. In other words, features can be rated for their importance with respect to manufacturing and design implications.

52.2 Related Work Very few studies have been published for the manufacturability evaluation of prismatic parts using rating-based methodologies. Gu et al. [2] proposed a method to find the important features through manufacturability evaluation for feature sequencing of prismatic parts. Gebresenbet et al. [3] proposed manufacturability indices for turned components using feature–function–resource considerations. Chan [4] has developed an evaluation tool to access manufacturability of product designs. The tool developed is independent of manufacturing domain and covers a comprehensive range of manufacturing processes and commonly used engineering materials. The system is using knowledge base and production rules embedded in it. Arivazhagan et al. [5] and Hebbal and Mehta [6] have proposed machining feature recognition approaches for prismatic parts which is an integral part of CAD/CAM integration. Syaimak and Axinte [7] analysed manufacturability of micro-parts for milling and drilling operations using primitive feature analysis technique (PFA). Hoque et al. [8] and Amaitik and Kiliç [9] developed a platform for CAD-CAM integration using design feature library and knowledge base embedded in the system. It can be concluded from the literature review that most of the developed DFM systems are rule-based that makes them domain-specific and knowledge-intensive. The present systems lack simultaneous consideration of design and manufacturing complexities of the designed product.

52.3 Manufacturability Evaluation The manufacturability of a part could be assessed by a systematic analysis of different design and manufacturing factors posing difficulties towards producing it. It is a

52 Development of Manufacturability Indices for Prismatic Parts

627

measure of the feasibility and ease of achieving designer-assigned specifications and attributes to be fulfilled by the selected manufacturing environment. The parameters influencing manufacturability of the part have been identified and correlated to give an overall manufacturability index. A higher membership value indicates that the considered feature has higher manufacturability, i.e. easier to machine with respect to corresponding parameter.

52.3.1 Geometrical Complexity Parameters Geometrical and topological information has significance in analysing complexity related to feature shape with respect to manufacturing the considered feature. The investigation of geometrical attributes of the feature revealed that geometrical complexity of the feature can be estimated from the following eight parameters. Feature Intricacy (FI) Feature intricacy parameter takes care of complexity with respect to amount of details in the feature. It also accounts for complexity of combined features within the specific type. A feature with more number of machining surfaces is considered to be more complex. Membership value of a feature to the parameter FI can be explicitly determined using Eq. 52.1.

G FI =

(rFI )min × 0.9 (rFI )f

(52.1)

rFI is calculated as N f /N min , where N f is the number of machined surfaces of the considered feature and N min belongs to the machined surfaces count of the simplest feature of respective type (pocket, hole, slot, step and compound feature). Feature Face Orientation (FFOR) The orientation of a feature face can be defined based on the three mutually perpendicular principal directions of the part. Feature faces which are oriented parallel or perpendicular to the principal directions of the part are easier to machine compared to those which are inclined. So when comparing features having same number of machining faces, the feature consisting of more number of inclined faces will be more difficult to machine (Table 52.1). Table 52.1 Ratings to different surfaces w.r.t. its orientation

Face inclination type

Rating (r s )

Face normal is parallel or perpendicular to all principal axes

0.9

Face normal is perpendicular to any one principal axis

0.5

Face normal is inclined to all principal axes

0.1

628 Table 52.2 Ratings to features w.r.t. number of feasible TADs

M. K. Gupta et al. Number of feasible TADs

r NTAD

4

0.1

3

0.3

2

0.6

1

0.9

Membership value of FFOR will be average rating of all the machined faces constituting the feature under consideration. Feasible Tool Access Directions (TADf) A feature with its selected TAD common with more number of other features on the part will contribute towards reducing number of setups and thus will have more manufacturability. On the other hand, a feature with more number of feasible tool access directions will increase the set-up time and thus will have lower manufacturability. Membership value of this parameter can be assigned using Eq. 52.2. G TAD

  Nc × rNTAD = 1− N

(52.2)

where N c is number of features with tool access direction selected same as that of considered feature, N is total number of features on the part and r NTAD is comparative rating given to features as per number of feasible TADs (Table 52.2). Feature Neighbourhood Categories (FNC) Feature neighbourhood could be useful in determining feature accessibility for machining with respect to the existence of safe and sufficient space for cutting tool to access the feature and relationship between neighbouring features. To account this, several categories are identified for different instances of feature under consideration with respect to other features such as independent, nesting, nested, attached and volumetric. Further, these categories have been assigned membership values relatively (0–0.9 scale) with respect to difficulty in machining. Feature Hierarchy (FH) Feature hierarchy will account for parent or child nature of features. Parent features are considered to be more manufacturable compared to their child as they do not require further machining. Membership value of FH can be assigned as 0.9 to parent feature and 0.5 to child feature. Approach Direction Depth (ADD) The approach direction depth (ADD) also affects the manufacturability of the feature. Approach direction depth of a feature is the depth from the stock face to the lowest machined point on the feature measured along the selected tool access direction. The feature with lesser depth will be easier to access thus its manufacturability will be higher compared to the feature with higher depth. Membership value of the parameter ADD can be estimated using Eq. 52.3 where ADDf is approach direction depth of the considered feature and ADDmin corresponds to the lowest ADD of a feature on the part.

52 Development of Manufacturability Indices for Prismatic Parts

 G ADD =

ADDmin ADDf

629

 × 0.9

(52.3)

Feature Accessibility (FACC) Accessibility of a feature is defined as the ability of its constituting surfaces to get easily reached for machining by the standard tools available in the selected approach direction. The accessibility of a feature face depends on the geometry and shape of the feature. For this purpose, the method developed by Shankar and Jansson [10] has been modified and used in this work. Accessibility rating of a feature surface can be calculated using primary (θ ) and secondary (γ ) access angles using Eq. 52.4.  racc =

2θ + γ 270

 + 0.1

(52.4)

Thus, accessibility of a feature will be average rating of all the machined faces constituting the feature under consideration (Eq. 52.5). Membership value of FACC can be determined using Eq. 52.6. Nf 1  (racc )k Nf k=1   ACCmin × 0.9 = ACCf

ACCf = G FACC

(52.5) (52.6)

Parent or Child Feature Complexity (PFC or CFC) Parent feature complexity is determined to capture the relative difficulty of realizing the shape or form of the parent features with respect to the type of machining processes and other resources required. Drilling features are considered less manufacturable as they require more cutting forces compared to milling features. The features with higher machining volumes are rated lower as they require more machining time. Comparative ratings are given to the type of machining features as per Table 52.3. Similarly, child feature complexity is determined to capture the relative difficulty of realizing the shape or form of the child features on their corresponding parents. Three types of categories of child features with respect to their parents are preferentially rated as given in Table 52.4. Table 52.3 Relative ratings to type of machining features

Type of feature (t)

Rating (r t )

Milling feature

0.9

Drilling feature

0.5

630 Table 52.4 Relative ratings to different categories of CFC

M. K. Gupta et al. Category

Child FT

Parent FT

Rating (r CFC )

1

Milling

Milling

0.9

2

Drilling

Milling

0.5

3

Drilling

Drilling

0.1



 Vmin × rt Vf p   Vmin = × rCFC Vf c

G PFC = G CFC

(52.7) (52.8)

Membership values of CFC and PFC are determined using Eqs. 52.7 and 52.8.

52.3.2 Technological Complexity Parameters Among various technological attributes of a feature, dimensional tolerances and surface finish are the most important attributes which affect directly or indirectly the cost and quality of the final part. In general, a feature on a part can be produced by a number of alternative processes. Each of these processes will have different set of tooling requirement such as cutting tools, machine tools which have different costs associated. Therefore, the technological attributes (dimensional tolerances and surface finish) and tooling complexities (machine tool and cutting tool) are selected as the parameters to evaluate the technological complexity of the feature. Tolerances (TOL) General cost-tolerance reciprocal power model [11] given by Eq. 52.9 is used to estimate relative cost factor of tolerance assigned to a feature. Based on these cost factors, Eq. 52.10 is used to determine membership value of a feature towards TOL parameter (TTOL) for its manufacturability. TOLCf = B/δfk  TTOL =

TOLCmin TOLCf

(52.9)

 × 0.9

(52.10)

Surface Finish (SF) Cost-surface finish relationship [12] given by Eq. 52.11 has been used here to estimate relative cost factor of surface finish required by a feature which will be later converted to membership value towards feature’s SF (TSF) using Eq. 52.12. 

SFCf = a1 e

−

Raf λ1



+ a2 e

  Ra −λf 2

(52.11)

52 Development of Manufacturability Indices for Prismatic Parts

 TSF =

SFCmin SFCf

631

 × 0.9

(52.12)

Machine Tool Complexity (MTC) Machine tool complexity reflects the difficulty in producing the part due to the cost of machine tools selected. Different machines would impose different costs. The machine tool complexity is determined by using the cost factors assigned to the machine tools available at the shop floor (Eqs. 52.13– 52.14). The membership value towards MTC can be estimated using Eq. 52.15, where m is the number of machines used for a single feature, MCFi is the cost factor for using ith machine, MCCF is the cost factor for machine change and M i is machine used for an operation. MCf =

 m 

 + MCCF ×

MCFi

i=1

f

m−1 

(Mi+1 , Mi )



1, if x = y 0, if x = y   MCmin × 0.9 = MCf

Ω(x, y) = TMTC

(52.13)

i=1

(52.14) (52.15)

Cutting Tool Complexity (CTC) Cutting tool complexity (CTC) is determined on the similar fact as MTC is established. Equations 52.16 and 52.17 are used to calculate membership value of the feature towards CTC, where t is the number of cutting tools used for a single feature; TCFi is the cost factor for using ith cutting tool, TCCF is the cost factor for machine change and C i is cutting tool used for an operation TCf =

 t  i=1

 + TCCF ×

TCFi

t−1 

[{1 − (Mi+1 , Mi )} × (Ci+1 , Ci )]

i=1

f

 TCTC =

TCmin TCf

(52.16)

 × 0.9

(52.17)

52.3.3 Determination of Manufacturability Indices (TCI and GCI) The geometrical complexity index (GCI) and technological complexity index (TCI) of a feature are obtained as a weighted sum of the various geometrical and technological parameters considered using Eqs. 52.18 and 52.19.

632

M. K. Gupta et al.

GCI =

n 

wi G i

(52.18)

w j Tj

(52.19)

i=1

TCI =

p  j=1

52.3.4 Determination of Weights for Geometrical and Technological Parameters Best worst method (BWM) is the latest technique developed by Rezaei [13] which is utilized for computing weights. Among all the geometrical and technological parameters, respectively, best and worst in terms of their influence on manufacturability of the feature are decided by experts. Other parameters are rated with respect to best and worst parameters on the scale of 1–9. The optimized weights for geometrical and technological parameters such obtained as (0.056, 0.062, 0.216, 0.076, 0.046, 0.121, 0.249, 0.174) and (0.34, 0.30, 0.23, 0.13), respectively.

52.4 Illustrative Example A prismatic part shown in Fig. 52.1 is taken as an example to explain the manufacturability evaluation methodology. The prismatic part contains 15 manufacturing features. The results of geometrical and technological complexity index evaluation of individual features of the part are shown in Tables 52.5 and 52.6. The overall manufacturability indices (OMIs) of all the features are graphically presented in Fig. 52.2. Results show that overall manufacturability index of the feature through hole (F5) is the lowest (0.4953); hence, it is found to be most difficult to manufacture followed by feature F7, F3, F9, F15, F11, F13, F6 and so on. F5 is assessed to be most difficult as its both geometrical (GCI) and technological complexity indices (TCI) are lower as compared to other features on the part. Specifically, the difference between TCI of F5 and other features is much higher which indicates need for revisiting its technological requirements. Geometrical complexity index can also be improved by making adjustments in different geometrical parameters of the feature as far as it satisfies the functional requirements. Similarly, other features on the part can be compared among each other with respect to their ease of manufacturing with geometrical and technological point of view using these quantitative indices. These indices will help designers to adjust their designs to achieve greater manufacturability up to their highest level of satisfaction.

52 Development of Manufacturability Indices for Prismatic Parts

633

Fig. 52.1 Sample prismatic part Table 52.5 Membership values to different geometrical complexity parameters with GCIs Feature ID

GCFC

GPFC

GACC

GFNC

GADD

GFH

GTAD

GFFOR

GFI

GCI

F1

0

0.329

0.259

F2

0

0.492

0.556

0.9

0.45

0.9

0.14

0.9

0.3

0.389

0.84

0.9

0.9

0.42

0.9

0.36

F3

0.5

0

0.605

0.402

0.54

0.45

0.5

0.42

0.9

0.45

0.477

F4

0.067

F5

0.045

0

0.169

0.72

0.113

0.5

0.42

0.9

0.45

0.32

0

0.149

0.72

0.075

0.5

0.093

0.9

0.9

F6

0.258

0

0.117

0.9

0.84

0.75

0.9

0.046

0.9

0.45

0.531

F7

0.052

0

0.433

0.3

0.281

0.5

0.093

0.9

0.225

0.285

F8

0

0.420

0.186

0.9

0.45

0.9

0.84

0.63

0.3

0.521

F9

0.433

0

0.335

0.48

0.346

0.5

0.093

0.9

0.225

0.349

F10

0

0.349

0.128

0.9

0.069

0.9

0.133

0.9

0.9

0.346

F11

0

0.459

0.347

0.66

0.45

0.9

0.6

0.9

0.36

0.518

F12

0

0.9

0.374

0.66

0.643

0.9

0.6

0.9

0.36

0.625

F13

0

0.155

0.134

0.9

0.069

0.9

0.133

0.9

0.9

0.313

F14

0

0.367

0.9

0.78

0.45

0.9

0.2

0.9

0.3

0.559

F15

0

0.354

0.259

0.78

0.409

0.9

0.28

0.9

0.3

0.409

634

M. K. Gupta et al.

Table 52.6 Membership values to different technological complexity parameters with feature technological complexity indices Feature ID

T TOL

T SF

T MTC

T CTC

F1

0.625

0.9

0.09

0.27

0.5383

F2

0.4911

0.3472

0.3

0.27

0.3752

F3

0.4911

0.1432

0.0346

0.0342

0.2224

F4

0.5464

0.3472

0.9

0.9

0.6139

F5

0.5464

0.1432

0.0173

0.0338

0.2371

F6

0.4093

0.3472

0.3

0.27

0.3474

F7

0.4093

0.3472

0.225

0.225

0.3243

F8

0.9

0.1914

0.09

0.108

0.3982

F9

0.5894

0.3472

0.225

0.27

0.3914

F10

0.5153

0.1698

0.9

0.9

0.5501

F11

0.5894

0.1914

0.225

0.108

0.3236

F12

0.5894

0.1914

0.3

0.225

0.3561

F13

0.5153

0.1698

0.9

0.9

0.5501

F14

0.5894

0.1914

0.225

0.18

0.333

F15

0.5894

0.1914

0.3

0.27

0.3619

Fig. 52.2 Overall manufacturing indices of features on the example part

TCI

52 Development of Manufacturability Indices for Prismatic Parts

635

52.5 Conclusions This paper presents a rating-based methodology for early manufacturability assessments of manufacturing features on the proposed design. The developed manufacturability indices help designers and machinists to detect and rectify unintentional errors while assigning technological attributes and to know the key areas of focus for redesign or modification to reduce cost and time to market. Most of the interrelationships are built up using inputs of experienced personnel who are well acquainted with workshop conditions. Compared to other rating-based approaches, the present approach incorporates more number of parameters. Future work involves evaluation of manufacturability precisely with respect to design intended function.

References 1. Jiang, B.C., Hsu, C.-H.: Development of a fuzzy decision model for manufacturability evaluation. J. Intell. Manuf. 14(2), 169–181 (2003) 2. Gu, Z., Zhang, Y.F., Nee, A.Y.C.: Identification of important features for machining operations sequence generation. Int. J. Prod. Res. 35(8), 2285–2308 (1997) 3. Gebresenbet, T., Jain, P.K., Jain, S.C.: Preliminary manufacturability analysis using featurefunction—resource considerations for cylindrical machined parts. Int. J. Comput. Integr. Manuf. 15(4), 361–378 (2002) 4. Chan, D.S.K.: Expert system for product manufacturability and cost evaluation. Mater. Manuf. Process. 18(2), 313–322(2003) 5. Arivazhagan, A., Mehta, N.K., Jain, P.K.: Development of a feature recognition module for tapered and curved base features. Int. J. Adv. Manuf. Technol. 39(3–4), 319–332 (2008) 6. Hebbal, S.S., Mehta, N.K.: Setup planning for machining the features of prismatic parts. Int. J. Prod. Res. 46(12), 3241–3257 (2008) 7. Syaimak, A.S., Axinte, D.A.: An approach of using primitive feature analysis in manufacturability analysis systems for micro-milling/drilling. Int. J. Comput. Integr. Manuf. 22(8), 727–744 (2009) 8. Hoque, A.S.M., Halder, P.K., Parvez, M.S., Szecsi, T.: Integrated manufacturing features and design-for-manufacture guidelines for reducing product cost under CAD/CAM environment. Comput. Ind. Eng. 66(4), 988–1003 (2013) 9. Amaitik, S.M., Kiliç, S.E.: An intelligent process planning system for prismatic parts using STEP features. Int. J. Adv. Manuf. Technol. 31(9–10), 978–993 (2007) 10. Shankar, S.R., Jansson, D.G.: A generalized methodology for evaluating manufacturability. Concurr. Eng. 248–263 (1993) 11. Chase, K.W., Greenwood, W.H., Loosli, B.G., Hauglund, L.F.: Least cost tolerance allocation for mechanical assemblies with automated process selection. Manuf. Rev. 3(1), 49–59 (1990) 12. Ong, S.K., Sun, M.J., Nee, A.Y.C.: A fuzzy set AHP-based DFM tool for rotational parts. J. Mater. Process. Technol. 138(1–3), 223–230 (2003) 13. Rezaei, J.: Best-worst multi-criteria decision-making method. Omega 53, 49–57 (2015)

Chapter 53

A Cyber-Physical System Architecture for Smart Manufacturing Jitin Malhotra , Faiz Iqbal , Ashish Kumar Sahu

and Sunil Jha

Abstract A cyber-physical system (CPS) architecture is proposed for achieving the aim of smart manufacturing. The proposed architecture lays a framework to implement a CPS in shop floor from data acquisition, processing it, running algorithms to acquire some useful information from the raw data and studying the patterns generated by the machine and take a step toward implementing Industry 4.0 concepts in the shop floor. The two major areas, i.e., CPS configuration and operation are focused, and further, a step-by-step description of various modules of CPS architecture is explained. In the end, a small-scale implementation of CPS is done on i5-B CNC BEMRF machine. A CPS-based application is also developed to verify the feasibility of proposed architecture, which calculates finishing time, no. of cycles required to achieve target roughness value and what is the best possible roughness value achievable based on the input process parameters and roughness value given by the operator. Keywords Cyber-physical system · Industry 4.0 · Ball end · MR finishing

Abbreviations Used AGV API AWS BEMRF CNC CPPS CPS HMI

Automated guided vehicle Application programming interface Amazon Web Services Ball end magnetorheological finishing Computer numerically controlled Cyber-physical production systems Cyber-physical systems Human–machine interface

J. Malhotra · F. Iqbal · A. K. Sahu · S. Jha (B) Department of Mechanical Engineering, Indian Institute of Technology Delhi, New Delhi, Delhi 110016, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_53

637

638

I/O IOT IPC MQTT MRP PLC Ra USB

J. Malhotra et al.

Inputs/Outputs Internet of things Industrial PC Message queuing telemetry transport Magnetorheological polishing fluid Programmable logic controller Roughness average Universal serial bus

53.1 Introduction Digital electronics advancement has led to a considerable growth in the systems that link the digital part to the physical world, popularly known as cyber-physical system (CPS). The term, cyber-physical systems, was introduced at National Science Foundation (NSF) in the USA in the year 2006, and it defines CPS as a multi-disciplinary, intricate, next-generation systems with physical awareness that add embedded computing capabilities (cyber) with the physical machines [1]. In the manufacturing context, CPS exploits the advancements achieved in computing systems on modeling in combination with the enormous amount of data produced from the surrounding environment through low-power sensors and actuators. These technologies led to the creation of smart factories in which machines can reconfigure themselves in line with the external conditions and able to plan the factory’s resources and the information throughout the entire value chain by utilizing IoT, cloud computing, and smart products paradigm. The effects of CPS in manufacturing sector focusing on process automation and control are in the form of cyber-physical production systems (CPPS). CPPS consist of elements and sub-systems usually autonomous and cooperative in nature that gets connected with each other in different ways based on varied situations, on all available levels of production, starting from various individual processes through different machines up to the logistics and production networks available. The challenges come in operating sensor networks, handling big data, and then retrieving useful information, its representation, and interpretation, with a security check. The CPPS is different from the traditional automation pyramid as it is more of a decentralized system. The standard control and field level systems still exist including the standard PLCs which are near to the technical processes and giving good performance in critical control loops, but at the upper levels CPPS exist which act in a more decentralized way. The elements of a CPPS can acquire and process data, can control some tasks on their own, and further interact with humans via different interfaces [2].

53 A Cyber-Physical System Architecture for Smart Manufacturing

639

Various researchers have proposed the architecture for implementing CPS in manufacturing scenarios like in Lee et al. [3] have proposed a 5C architecture for CPS implementation in manufacturing. The first level is Connection level which acquires accurate and reliable data from different machines and components at shop floor and is passed onto the second level called Conversion level which converts the raw data into useful information by running algorithms. The third level is a Cyber level which acts as a central hub, and all data is pushed to it from different networked machines. Fourth is Cognition in which this information is depicted in graphical form and is presented to users and experts for analysis and decision making. The last level is Configuration which takes the feedbacks from Cyber level back to Connection level and works as a supervisory control to make machines configure and adapt themselves based on feedbacks. After describing this architecture, authors have discussed CPS-based physical health monitoring system. Zhang et al. [4] presented an architecture for CPS implementation, in which the architecture is divided into three layers. Connection layer is the first layer which collects data in four different groups; then this data is passed on to the second layer that is data network layer, which analyzes the data, and preprocessing is done. Last is the Configuration layer which presents the data to user and experts and further provides the supervisory controls to the machine making them self-configure and self-adaptive. This add-on CPS system is implemented further on EMCO MILL55 by the author in his lab. Liu and Jiang [5] proposed architecture for smart manufacturing in the shop floor. This architecture comprises of three layers, i.e., Connection layer in which the installation of various sensors as required for CPS is done. It also solves the issues related to protocols, processing of sensor data, location of sensors, distance, and storage capacity needs for considering the embedded component for systems. Then comes the Middleware layer which acquires the data from different machine tools, AGVs, industrial robots, and various sensors installed in the factory in the form of nodes, units, and system. Moreover, this collected data is transferred to the third layer which processes this data into useful information and stores it in the cloud. Also, from the second layer, the commands are passed on to Connection layer from external applications, i.e., job scheduling, condition monitoring, and quality control groups. However, all the architectures implemented or proposed by the researchers for implementing CPS in manufacturing are not suitable for our BEMRF machine tool or machine tools having a soft controller-based architecture with a high computing power available on machine tool as all the architectures are focused on adding more hardware for CPS implementation like additional embedded processors for capturing the data from sensors, running algorithms for pre-processing on embedded processors, and incorporating edge processing functionality by adding a high end embedded processor or a desktop at machine level. This additional embedded hardware is not required in case of BEMRF tool as it has the built-in Intel i3-based IPC which can acquire data from sensors, does the pre-processing at the machine level, and further sends the useful information to the cloud. Also, the Internet connectivity is directly available in the IPC, so additional processors are not required. The layout for such kind of systems having the inbuilt processing power at machine level is not available

640

J. Malhotra et al.

in literature as most of them focus on PLC which do not have Internet connectivity and cannot be programmed to run some customized scripts for pre-processing and information models cannot be deployed on them. This work focusses on laying down the framework for implementing CPS on such machine tools.

53.2 Proposed CPS Architecture for Smart Manufacturing The proposed architecture is divided into two major components which are as follows: • Data Acquisition and Processing. • Cloud Network. Now, these components are the backbone of the cyber-physical systems as proposed by the researchers [3–5]. These components and some additional modules form the CPS architecture for making machines smarter (Fig. 53.1).

53.2.1 Data Acquisition and Processing The first component of CPS includes all the physical components present in the shop floor level, starting from the sensors to actuators, to spindles and the controllers. Establishing heterogeneous types of communication between the sensors and the

Fig. 53.1 Proposed CPS architecture

53 A Cyber-Physical System Architecture for Smart Manufacturing

641

controller can be done with ease as the high computing power is already available at the machine tool. This architecture focusses on the cases where a desktop PC or Industrial Panel PC is already installed on the machine tool, so there is no requirement of any additional embedded processors for connecting sensors or for establishing network connectivity. The sensors which are present in the machine can be directly integrated into analog or digital I/O modules already installed in the machine. In addition to interfaces other than analog or digital I/O like an Ethernet or USB interface on a dynamometer or RS232/485 from some sensors, these can be directly integrated into machine computer and data can be acquired either through APIs or through socket connections and further utilized for pre-processing. One additional benefit for using this integrated computer is the computing power available on them, which further pre-processes the data acquired earlier. So, the acquired data either comes through the control software of PLC or through some socket listener running on the machine computer which can further run the pre-processing modules on the acquired data, like data filtering module can be used to filter the data from the noise which occurs in the data through various environmental sources or hardware limitations; data conversion module does the conversion of available raw data into some useful information and storing it into a format acceptable by the other modules in pre-processor and cloud-based data servers. The data pre-processor is not limited to these mentioned modules and varies from case to case. Some data pattern recognition algorithms can also be used for getting some insight into the data, and some information model created on cloud can be deployed here on the machine level. Talking about the interoperability, the sensor data, machine data, or process data, all of them are already available at the IPC through standard interface, and there is no requirement of additional embedded hardware with interface shields to acquire them and send it to cloud, instead a high-level language like C#, C, C++, or Python-based script can be developed and deployed on the machine computer itself. Also, to some extent EDGE processing can be done at the IPC itself. This EDGE processing can further help in converting the raw data into useful information and sending it further to cloud for post-processing, analysis, and decision making. Also, as most of the task from acquiring to processing can be done at the machine level, the communication loop is smaller so a better response time is achievable in comparison with embedded hardware-based system architectures. This unit also takes the supervisory commands given by the decision-making module in cloud network.

53.2.2 Cloud Network This unit deals with the pre-processed data and further storing it in a cloud. These clouds can be Amazon Web Services cloud, Microsoft Azure, IBM Watson, or any third-party cloud which has the data servers for storing the useful information and has the capability to run the analytical algorithms on it. At this level, the information from data servers is also available to the web or mobile applications, through which users can get into the insight of data, monitor the machine, and can give

642

J. Malhotra et al.

supervisory commands in case of emergency. Also, the data acquired at this level is a post-processed, and detailed analysis of them are taken out which is useful at various management levels in any industry. The data is also utilized in forming the information model which helps in predicting and monitoring purposes. The supervisory commands can also be generated by decision-making module by running the algorithms for post-processing on the data and taking inputs from the human expert, which can further help in health monitoring scenarios and emergencies.

53.3 Implementation of CPS Architecture 53.3.1 Ball End Magnetorheological Finishing Process Ball end magnetorheological finishing (BEMRF) [6] is a finishing process, based on smart fluids which have the capability of finishing 3D surfaces and getting roughness value in range of nanometers. Different types of ferromagnetic and nonmagnetic materials including aluminum, silicon [7], copper [8], etc., are finished using this process. In this finishing method, magnetorheological polishing (MRP) fluid (containing abrasives base medium and carbonyl iron particles) is utilized as a finishing medium, and it changes its rheological characteristics under the magnetic field. In this process, the spindle is centrally hollow cylindrical in nature, usually made of ferromagnetic material and is kept axially inside an electromagnetic core. The pressurized MRP fluid is passed on to the tip of this tool spindle via the hollow space, and the electromagnet is energized, which results in the formation of a hemispherical ball of this MRP fluid at the tip of the tool spindle. The stiffness of this ball is also variable as it changes with a change in the value of the magnetizing current supplied to the electromagnetic core [9]. The BEMRF machine tool is in-house manufactured at IIT Delhi, which is a five axes CNC machine with process and motion controlled by the customized controller and has a capability of in situ roughness measurement [10].

53.3.2 Description of i5-B CNC BEMRF Machine Tool This section depicts the detailed description of i5-B CNC BEMRF system installed in Automation Lab, Mechanical Engineering Department, IIT Delhi, as an example to verify the proposed CPS architecture. The BEMRF machine tool is an in-house developed five axes CNC controlled fully automated machine with in situ roughness measurement functionality. Detailed specifications of BEMRF machine are given in Table 53.1.

53 A Cyber-Physical System Architecture for Smart Manufacturing

643

Table 53.1 Specifications of BEMRF machine tool • Machine dimensions: 1000 mm × 1000 mm × 1500 mm • Five axes system with individual travels – X: 150 mm – Y: 150 mm – Z: 300 mm – B: +45° to −45° (limited by software) – C: 0° to 360° • CNC functionality achieved by the customized controller • In situ roughness measurement using the confocal sensor • Intel i3 processor-based IPC system with Windows operating system and 12-in. multitouch screen • Temperature measurement sensors at three level in BEMRF tool • Spindle RPM range: 300–3000 • Electromagnet current range: 0–10 A • On machine MR fluid preparation and fluid delivery mechanism • Workpiece cleaning mechanism

53.3.3 Configuring CPS on i5-B CNC BEMRF Machine Tool According to the proposed CPS architecture in Sect. 53.2, the details of implementation of CPS done on the i5-B CNC BEMRF system are given in this section. The two major levels of the CPS architecture are data acquisition and processing unit, and cloud network. The required elements used in each unit are described below. Data Acquisition and Processing Unit In data acquisition, we have all the sensors which are present in the BEMRF machine tool, i.e., three sensors for temperature measurement installed inside the coil of the electromagnet, one confocal sensor for roughness measurement, limit and home sensors for individual axis, and linear encoder data from each axis. This unit includes the process parameter’s controls of BEMRF through the analog and digital I/O modules, which are current for the electromagnet, stirrer on/off for fluid preparation, and fluid delivery pump on/off for delivery of prepared fluid to the tooltip. The captured data from each sensor is filtered and converted into useful information, like data from the confocal sensor is captured and converted into roughness values, temperature data is captured and filtered, and patterns are formed based on the time of finishing using the EDGE processing. Based on the models which are developed in the cloud network, the processing unit preprocesses the data and compresses it further for sending it to the cloud. One additional module is added to C#-based HMI for this machine to send the pre-processed data to AWS cloud and for pre-processing the algorithms are programmed in C# in Visual Studio. This unit also takes and processes the supervisory commands generated from the decision-making module running in cloud network.

644

J. Malhotra et al.

Cloud Network In this unit, the pre-processed data is further analyzed by running some machine learning algorithms; some information models are created which can help us in finding the estimated time for machining, and also with some mobile or web applications the data from the machine like the current part which is machined, last roughness value measured, required roughness value, current machine parameters, and a digital twin of the machine can be seen. In case of machine malfunction or change in the machining pattern of the machine, the human expert is notified who can analyze the health of the machine and take decisions based on the condition, like in case of increase in temperature above the threshold value can generate alarms or personalized emails/SMS and the human operator can take the supervisory control by either switching the machine off or stopping it and checking the reason for increase in the temperature.

53.3.4 Example of EDGE Processing on i5-B CNC BEMRF Machine Tool This section shows an application implemented on the BEMRF machine tool for predicting the finishing time for finishing with current set of process parameters of BEMRF, i.e., electromagnet current, spindle speed, working gap, and feed to the system; number of cycles required to achieve target Ra value and finishing time for current cycle and when the parameters need to be changed by taking input from user, a set of process parameters, current Ra , and target Ra are required. A database [11] is already created in the cloud with values of process parameters, achieved roughness value, and time taken in achieving that roughness value; based on this data, an information model is created in the cloud and deployed in the BEMRF machine. This database is continuously updated in real time with each finishing cycle completed on the machine, and a self-improving type of scenario is developed. The application deployed on BEMRF machine tool predicts about the estimated finishing time, the best possible roughness value achieved with current set of process parameters, and how many further cycles are required to achieve the target Ra . The algorithm running behind this application has been explained below. Figure 53.2 shows the interface of the application deployed on the BEMRF machine. In the backend of this C#-based application, the model is regularly updated with each machining by analyzing the data stored on AWS cloud by machine HMI through MQTT protocol. Finishing Time Calculator Algorithm This section explains the algorithm running behind the finishing time calculator application shown in the above section.

53 A Cyber-Physical System Architecture for Smart Manufacturing

645

Fig. 53.2 Finishing time calculator application

In Fig. 53.3, a flowchart for finishing time calculator application is shown which is running in the background of finishing time calculator application shown in Sect. 53.3.4. Initially, the user inputs the value of Ra required, current Ra value, and a set of process parameters, then a comparison is done whether the current value is equal to the required value; if it is true, then a message is pop up with a text that current value of Ra is equal to the required Ra value. If the current value is not equal to required value, then a comparison algorithm comes into picture which compares the set of input process parameters and calculates the finishing time, no. of cycles required to achieve target Ra value, and after what time the process parameters be changed so as to achieve the required value. For doing all this, a parameter calculation algorithm is proposed in [11]. This algorithm has been scripted in C# and target values are printed on the screen.

646

J. Malhotra et al.

Fig. 53.3 Finishing time calculator algorithm

53.4 Conclusions In this paper, a CPS architecture is proposed for intelligent manufacturing. The proposed architecture aims to provide a solution by giving a framework for configurating CPS and then reveals how the CPS-based machine tool or a factory operates. The CPS structure followed is capturing the mixed data from various sources, integrating it, converting it to standard format, pre-processing it and doing some EDGE processing through information models created based on the previous data for initial

53 A Cyber-Physical System Architecture for Smart Manufacturing

647

level information from raw data. Then, this pre-processed data is stored on the cloud network for post-processing and analysis, where a large number of machine learning algorithms are running, and an information model is created, based on which intelligent decision making can be done. The data after analysis is further stored in the data servers, and web or mobile applications have access to this data. In future, CPS configurations and operational aspects for smart manufacturing in shop floor will be further studied based on the proposed CPS architecture, and some more advanced algorithms for pre-processing and post-processing will be developed based on this architecture.

References 1. Gunes, V., Peter, S., Givargis, T., Vahid, F.: A survey on concepts, applications, and challenges in cyber-physical systems. KSII Trans. Internet Inf. Syst. 8, 134–159 (2014). https://doi.org/ 10.3837/tiis.0000.00.000 2. Monostori, L.: Cyber-physical production systems: roots, expectations and R&D challenges. Procedia CIRP 17, 9–13 (2014). https://doi.org/10.1016/j.procir.2014.03.115 3. Lee, J., Bagheri, B., Kao, H.: A cyber-physical systems architecture for Industry 4.0-based manufacturing systems. Manuf. Lett. 3, 18–23 (2015). https://doi.org/10.1016/j.mfglet.2014. 12.001 4. Zhang, C., Jiang, P., Cheng, K., Xu, X.W., Ma, Y.: Configuration design of the add-on cyberphysical system with CNC machine tools and its application perspectives. Procedia CIRP 56, 360–365 (2016). https://doi.org/10.1016/j.procir.2016.10.040 5. Liu, C., Jiang, P.: A cyber-physical system architecture in shop floor for intelligent manufacturing. Procedia CIRP 56, 372–377 (2016). https://doi.org/10.1016/j.procir.2016.10.059 6. Singh, A.K., Jha, S., Pandey, P.M.: Magnetorheological ball end finishing process. Mater Manuf. Process 27, 389–394 (2012). https://doi.org/10.1080/10426914.2011.551911 7. Saraswathamma, K., Jha, S., Rao, P.V.: Experimental investigation into ball end magneto rheological finishing of silicon. Precis. Eng. 42, 218–223 (2015). https://doi.org/10.1016/j. precisioneng.2015.05.003 8. Khan, D.A., Alam, Z., Jha, S.: Nanofinishing of copper using ball end magnetorheological finishing (BEMRF) process. In: Proceedings of the ASME 2016 International Mechanical Engineering Congress and Exposition, Phoenix, Arizona, USA, 11–17 November 2016. https:// doi.org/10.1115/imece2016-65974 9. Iqbal, F., Jha, S.: Nanofinishing of Freeform Surfaces Using BEMRF. Nanofinishing Science and Technology: Basic and Advanced Finishing and Polishing Processes, pp. 235–264 (2017) 10. Alam, Z., Iqbal, F., Ganesan, S., Jha, S.: Nanofinishing of 3D surfaces by automated five-axis CNC ball end magnetorheological finishing machine using customized controller. Int. J. Adv. Manuf. Technol. 1–12 (2018). https://doi.org/10.1007/s00170-017-1518-0 11. Iqbal, F., Jha, S.: Experimental investigations into transient roughness reduction in ball-end magnetorheological finishing process. Mater. Manuf. Process (2018). https://doi.org/10.1080/ 10426914.2018.1512131

Chapter 54

Measurement of Bores Using Scanning Mode of Articulated Arm Coordinate Measuring Machines Ashik Suresh

and P. B. Dhanish

Abstract Articulated arm coordinate measuring machines (AACMM) have become popular because of their portability, flexibility, and reduced cost though it is generally reputed that they are less accurate compared to CNC CMMs. We have compared the performance of a CNC CMM with a touch-trigger probe to that of an AACMM for the measurement of inner diameter of bores. AACMM was used in three different methods, viz., manual triggering, probe scanning for two revolutions, and finally scanning for the same time as that for manual triggering. Using each method, 20 bores were measured twice in random order and the uncertainties were computed using the measurement systems analysis approach. It was found that the measurement with AACMM in scanning mode gave overall lower uncertainties in the measurement of diameter. This may be because even though the single point coordinate uncertainty is lower in a CNC CMM, a larger number of points can be captured at the same time with an AACMM in scanning mode, thereby resulting in a lower task-specific measurement uncertainty. Keywords AACMM · MSA · Probe scan · Uncertainty · Coordinate metrology

54.1 Introduction The articulated arm coordinate measuring machine (AACMM) consists of several articulated arms equipped with angular encoders. It measures the coordinates of an object in space by reading the rotated angles of the articulated arms from angular encoders [1]. The articulated arm has two linkage arms with a spherical measurement volume. AACMMs can be moved and set up anywhere in the manufacturing plant for immediate use. Its six degrees of freedom can be used to point in any orientation. A. Suresh · P. B. Dhanish (B) Department of Mechanical Engineering, National Institute of Technology Calicut, Kozhikode 673 601, India e-mail: [email protected] A. Suresh e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_54

649

650

A. Suresh and P. B. Dhanish

Further, its cost is lower compared to conventional coordinate measuring machines (CMM). However, its accuracy is less compared to CMMs, when determining the coordinates of a single point. In general, the measurement of a feature on a workpiece requires a number of points. AACMMs in scanning mode can capture a large number of points in a very short time which may compensate for the higher point coordinate uncertainty. A series of experiments was planned to test this idea, and the results are presented in this paper. A number of papers on AACMMs have recently been published and a brief review follows. Santolaria et al. [2] proposed a correction model to rectify the errors caused due to temperature changes. Filho et al. [3] developed a virtual sphere gauge for coordinate measuring arm performance test. Li et al. [4] proposed calibration and error compensation techniques for AACMMs with two rotational axes. Mutilba et al. [5] presented a process to carry out performance calibration as per ASME standards. Madruga et al. [6] proposed a feature-based gauge for measuring operator and AACMM performance. Ostrowska et al. [7] determined the uncertainty of AACMM measurement with the help of Denavit–Hartenberg modeling and Monte Carlo simulation. Brau et al. [8] have done a comparison between four different probing systems: passive spherical probe, active spherical probe, self-centering passive probe, and self-centering active probe based on parallel kinematics for an articulated arm CMM. Precision reference ball, Trihedral seat, and Ball bar was used as a reference for measurement. Ostrowska et al. [9] have reported a verification test done on AACMM with the help of laser tracker. Acero et al. [10] proposed an indexed metrology platform as an alternative method to assess the volumetric accuracy of AACMMs. Based on the available literatures, it was seen that there was no work that quantitatively compared the performance of AACMMs for measurement of bores, especially using the scanning mode approach (Fig. 54.1). Fig. 54.1 Articulated arm CMM

54 Measurement of Bores Using Scanning Mode of Articulated …

651

54.2 Experimental Methodology 54.2.1 Workpiece While a variety of shapes can be measured using AACMM, in this work, the measurement of a bore alone was considered. For the purpose of experimentation, 20 specimens of mild steel material each having an inner diameter of 20 mm were prepared, by drilling and boring round workpieces with an outer diameter of 45 mm and thickness of about 25 mm (Fig. 54.2).

54.2.2 Procedure The workpiece was placed on a magnetic V block and clamped. The inner diameter of each ring was measured using (a) CNC CMM (b) AACMM with (i) manual triggering, (ii) scanning for two rotations, and (iii) scanning for a fixed time and are described in detail below. The details of the instrument used for CNC CMM are given in Table 54.1. A total of 16 points each were probed with uniform spacing at the top, middle, and bottom of Fig. 54.2 Workpiece

652 Table 54.1 CNC CMM specifications

Table 54.2 Articulated arm CMM specifications

A. Suresh and P. B. Dhanish Make

Mitutoyo Japan

Model

Bright A 504

Measuring range

X 505, Y 405, Z 405

Repeatability

3 µm

Working area

638 × 860 mm

Maximum workpiece weight

100 kg

Volumetric accuracy U3 (MPEe )

4 + 5 L/1000 µm

Make

Hexagon metrology

Model

Absolute Arm 7312

Measuring envelope

1.2 m

Arm weight

10.2 kg

MPEp

8 µm

MPEe

5 + L/40 ≤ 18 µm

the inner diameter, thus resulting in a total of 48 points. The height difference between the top and bottom was 10 mm. The measuring speed was 3 mm/min; movement speed 10 mm/min; and safety distance 1 mm. The specifications of the instrument for AACMM are given in Table 54.2. Three different methods were used to measure the diameter which are explained below. 1. Manual trigger method for 48 points (AACMMm). Here, the operator touches each point on the workpiece surface with the stylus and presses the probe trigger. The 48 points are chosen randomly at two heights along the circumference, trying to keep uniform spacing as far as possible. 2. Scanning for two rotations (AACMMs2r). In this method, the operator touches the workpiece surface continuously using the stylus. He will also continuously press and hold the probe trigger button and move around the surface of the bore for two rotations in a spiral pattern. 3. Scanning for a fixed time (AACMMsft). Here also, the operator touches the workpiece surface continuously using the stylus, continuously pressing and holding the probe trigger button and moving the stylus around the surface of the bore in a spiral pattern. But in this case, the time of contact was kept equal to that required for manual triggering. It was found that approximately two and a half revolutions were completed during each scan. Measurements were taken for all the methods by a single operator in a controlled room temperature of 25 ± 1 °C though 20 °C is the standard temperature; because, the operator felt uncomfortable at the lower temperature. Further, the AACMM is meant for shop-floor measurement where the temperatures are higher, and for comparing the uncertainties, only the variation in temperature during a measurement is relevant.

54 Measurement of Bores Using Scanning Mode of Articulated …

653

54.3 Results and Discussion 54.3.1 Measurement Times The time taken for different measurement methods was also recorded. The average time taken for different methods is given in Table 54.3. It can be seen that CNC CMM generally takes more time compared to AACMM measurements. That is because the CMM moves in fast mode from point to point, and then moves slowly while carrying out the actual measurement, thus resulting in overall more time.

54.3.2 Determination of Precision Precision is defined as closeness of agreement between indications or measured quantity values obtained by replicate measurements on the same or similar objects under specified conditions [11]. To determine the precision of each method, we used the procedure of measurement system analysis as outlined by Wheeler [12]. Each workpiece was measured twice in random order. The absolute value of the difference between the two readings on each workpiece is the range R. The average of all the range values was computed as R. Control charts were plotted for R with upper control limit D4R and lower control limit D3R. It was verified that all the points were in control, or else, the point removed and the measurement repeated. The control limits for x¯ can be calculated as upper control limit x + A2R and lower control limit x − A2R, where D4, D3, and A2 are constants obtained from statistical tables. The standard uncertainty of the measurement is determined as R/d2. The calculations are illustrated partially in Table 54.4 for a single case. The R charts for the four different methods are shown in Fig. 54.3. For subgroup 2, D3 = 0, D4 = 3.267, A2 = 0.187 UCLR = D4R = 0.027536562; LCLR = D3R = 0 UCL X = X + A2R = 20.00418672; LCL X = X − A2R = 19.97254338 The range chart appears to be in statistical control. This shows that the operator has been capable of consistently repeating the measurements. x¯ chart helps in determining whether the measurement system is capable of distinguishing between Table 54.3 Time taken for different methods Method

Time

CNC CMM

1 min 45 s

AACMMm

33 s

AACMMs2r

8s

AACMMsft

33 s

654

A. Suresh and P. B. Dhanish

Table 54.4 Partial calculation explaining one case Part No.

1

2

3….

20

Reading 1

19.860973

20.056394

…

20.162058

Reading 2

19.850597

20.053955

…

20.157857

R

0.010376

0.002439

…

0.0042010

X

19.855785

20.0551745

…

20.1599575

R = 0.0084287 X = 19.98832505

Fig. 54.3 X -R charts

samples [13]. The greater the number of points beyond the control limits, the better is the measurement method in discriminating between the different units of product. In the xbar charts of Fig. 54.3, all methods show good discriminating power. Further, the distance between the control limits appears to be the least for AACMM scan for fixed time.

54 Measurement of Bores Using Scanning Mode of Articulated …

655

54.3.3 Comparison of Methods Here, the standard deviations measured from four different methods are compared. The 95% confidence intervals for the measured standard uncertainties were computed based on the chi-square distribution as explained in [14] as      (N − 1)s 2  (N − 1)s 2  2 ≤σ ≤ 2 χ α ,N −1 χ 1− α ,N −1 (2 ) ( 2 )

(54.1)

Here, s is the standard deviation computed from a sample of size, N and, χ 2 is the chi-square variable for significance level of α. The critical value of the chi-square 2 2 = 8.91 and χ0.02519 = 32.85 for a 95% confidence interval. distribution is χ0.97519 The confidence intervals for the standard deviations of the different methods are compared in Fig. 54.4. We can see that there is no significant difference between the scanning methods (AACMMsft and AACMMs2r) and between point trigger methods (AACMMm and CNC CMM). However, it is observed that the scanning methods give lower uncertainty when compared to that of point trigger methods.

54.3.4 Cylindricity While carrying out the diameter measurement, the peak-to-valley cylindricity deviation CYLt based on the least squares reference cylinder (LSCY) [15] were also obtained, and the results of the four methods compared. The range chart for different methods is shown in Fig. 54.5, and the process appeared to be in control. However,

Fig. 54.4 Comparison of 95% chi-square confidence intervals between various methods

656

A. Suresh and P. B. Dhanish

Fig. 54.5 X -R charts for form error values

only a few points were out of control in the average chart, showing that the measurement method is not capable of distinguishing between the cylindricity values in the different units of product. It was expected that the rings with high cylindricity error will have more uncertainty in the measured values and vice versa. To test this, we divided the rings into two categories low and high, based on the mean cylindricity values which were calculated as the average of the values using all different methods. The measurement results for the two different groups were compared in Fig. 54.6. However, the results obtained were not in the expected pattern. This may be due to the insufficient capability of the instrument to resolve the cylindricity error, as indicated by the average charts.

54 Measurement of Bores Using Scanning Mode of Articulated …

657

Fig. 54.6 Range plot categorized into high and low form error

54.4 Conclusions Measurement of bores has been carried out using CNC CMM and AACMM in three different ways, viz., manual trigger, scanning for two rotations, and scanning for fixed time. Measurement using AACMM scanning appears to have much less uncertainty compared to measurement using a fixed number of points using AACMM or CNC CMM. The time taken for measurement in AACMM scanning is significantly lower compared to CMM. The CNC CMM used for this study used a touch trigger probe, and it is admitted that the time as well as uncertainty may be reduced by using a scanning probe in the CNC CMM. But the cost of such a CMM will be much higher so that in that case the comparison with an AACMM will not be meaningful. Industries which are planning to buy AACMMs or CMMs can use the results to evaluate whether the equipment will meet their requirements. Further, industries which use AACMMs can adopt the scanning mode with benefits as reported.

References 1. Hocken, R.J., Pereira, P.H.: Coordinate Measuring Machines and Systems, 2nd edn. CRC Press, Boca Raton (2017) 2. Santolaria, J., Yague, J.A., Jimenez, R., Aguilar, J.J.: Calibration based thermal error model for articulated arm coordinate measuring machines. Precis. Eng. 33, 476–485 (2009). https:// doi.org/10.1016/j.precisioneng.2009.01.002 3. Filho, A.P., Fernandes, F.H.T., Arencibia, R.V.: Application of virtual spheres plate for AACMMs evaluation. Precis. Eng. 36, 349–355 (2012). https://doi.org/10.1016/j.precisioneng. 2011.10.004 4. Li, K.H., Chen, B., Qiu, Z.R.: The calibration and error compensation techniques for an Articulated Arm CMM with two parallel rotational axes. Measurement 46, 603–609 (2013). https://

658

A. Suresh and P. B. Dhanish

doi.org/10.1016/j.measurement.2012.08.020 5. Mutilba, U., Kortaberria, G., Olarraa, A., Gutierreza, A., Acedoa, E.G., Zubietaa, M.: Performance calibration of articulated arm coordinate measuring machine. In: The Manufacturing Engineering Society International Conference, MESIC, pp. 720–727 (2013). https://doi.org/ 10.1016/j.proeng.2013.08.264 6. Madruga, D.G., Barreiro, J., Cuestab, E., Gonzaleza, B., Pelliteroa, S.M.: AACMM performance test: influence of human factor and geometric features. In: 24th DAAAM International Symposium on Intelligent Manufacturing and Automation, pp. 442–448 (2014). https://doi. org/10.1016/j.proeng.2014.03.010 7. Ostrowska, K., Gaska, A., Sladek, J.: Determining the uncertainty of measurement with the use of a virtual coordinate measuring arm. Int. J. Adv. Manuf. Technol. 71, 529–537 (2014). https://doi.org/10.1007/s00170-013-5486-8 8. Brau, A., Valenzuela, M, Santolaria, J., Aguilar, J.J.: Evaluation of different probing systems used in articulated arm coordinate measuring machines. Metrol. Meas. Syst. 21, 233–24 (2014). https://doi.org/10.2478/mms-2014-0020 9. Ostrowska, K., G˛aska, A., Kupiec, R., Gromczak, K.: Verification of articulated arm coordinate measuring machines, accuracy using laser tracer system as standard of length. MAPAN 31, 241–256 (2016). https://doi.org/10.1007/s12647-016-0176-2 10. Acero, R., Brau, A., Santolaria, J., Pueo, M.: Evaluation of a metrology platform for an articulated arm coordinate measuring machine verification under the ASME B89.4.22-2004 and VDI 2617 9-2009 standards. J. Manuf. Syst. 42, 57–68 (2017). https://doi.org/10.1016/j.jmsy. 2016.11.002 11. Vocabulary of Metrology (VIM) site: https://jcgm.bipm.org/vim/en/2.15.html. Last accessed 13 Aug 2018 12. Wheeler, D.J.: EMPIII Using Imperfect Data. Statistical Quality Control Inc., Tennessee (2006) 13. SPC for Excel site: https://www.spcforexcel.com/knowledge/measurement-systems-analysis/ measurement-systems-yours-any-good 14. NIST/SEMATECH e-Handbook of Statistical Methods, site: http://www.itl.nist.gov/div898/ handbook/. Last accessed 27 June 2018 15. ISO 12180-1:2011, site: https://www.iso.org/obp/ui/#iso:std:iso:12180:-1:ed-1:v1:en. Last accessed 27 June 2018

Chapter 55

A Method for Evaluation of Simple Torus Surfaces T. S. R. Murthy

Abstract Standards are available only for the evaluation of simple surfaces like planes and cylinders to evaluate flatness and circularity. For many other surfaces like spheres and elliptical surfaces, methods have been developed by the author. In this paper, it is proposed to develop a method for the evaluation of simple torus surfaces like ring doughnut for which there are no standards or methods. The torus surface (torus of revolution) has application in chemical and enzymatic modifications of plant proteins. The CMM or flexible CMM like ROMER is suitable for measuring the surface data. Data measurement at equally spaced locations on the surface is preferred for better results. MATLAB has been used to simulate the measured data and test the developed mathematical model for its evaluation and a new term toricity is used to define the error. Keywords Circularity error · Toricity error · Coordinate measurement data

55.1 Introduction Any component manufactured has to be inspected to ensure that the manufactured component satisfies the design specifications. For that purpose in the digital era, a standard is available [1]. Methods and standards are available only for simple surface characteristics like flatness, circularity and cylindricity. For other types of surfaces for which no standards or methods are available, the author has established methods for the evaluation of interrelated tolerances [2], evaluation of different types of reference axes [3–11], spherical surfaces [12], elliptical surfaces [13], paraboloid surfaces [14], minimum zone evaluation of surfaces [15], etc. Regular coordinate measuring machines (CMM) and flexible coordinate measuring machines like robot called ROMER [16] are used for the measurement of coordinate data on the surface. Then, evaluation like circularity is done online by CMM for which algorithms are T. S. R. Murthy (B) Department of Mechanical Engineering, Sreenidhi Institute of Science and Technology, Hyderabad 501301, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_55

659

660

T. S. R. Murthy

available. Where the component surface is different which requires special algorithms, their evaluation has to be done offline with the measured data. In the present paper as there are no methods for the evaluation of torus surfaces, an attempt is made to find methods for its evaluation.

55.2 Problem Definition Torus is a surface of revolution generated by revolving a circle in a three-dimensional space about an axis coplanar with the circle. If the axis of revolution (Fig. 55.1b) does not touch the circle, the surface has a ring shape and is called a torus of revolution as shown in Fig. 55.1a. Figure 55.1a shows a torus with major radius R of 15 units and minor radius r of 5 units.

(a)

z - axis

Fig. 55.1 a Torus of revolution with major radius R = 15 and minor radius r = 5 units. b Tori based on rotation axis and circle location

Torus with Major Radius R=15 and Minor radius r=5 (Aspect ratio R/r=3)

5 0 -5 20 10

y-

20 10

0

ax

is

0

-10

-10 -20

(b)

-20

x-

s

axi

55 A Method for Evaluation of Simple Torus Surfaces

661

The ratio R/r is called aspect ratio which is 3 in this case. If the aspect ratio is 1 (R = r), then the torus is called horn torus. If R < r (aspect ratio TIBKAT:510998585 4. Murthy, T.S.R.: Measurement and evaluation of three intersecting axes of precision equipment. In: 9th International Symposium on Measurement and Quality Control (9th ISMQC), 21–24 Nov 2007, IIT Madras, pp. 292–295 (2007). ISBN:8190423525

55 A Method for Evaluation of Simple Torus Surfaces

669

5. Murthy, T.S.R.: Measurement and evaluation of three closely approaching axes of precision equipment. In: 5th National Conference on Precision Engineering COPEN 2007, Dec 2007, CET, Trivendrum, pp. 337–342 (2007) 6. Murthy, T.S.R.: Methods for evaluation of 3D reference axes. In: International Conference on Advanced Manufacturing Technologies (ICAMT2007), Nov 2007, CMERI, Durgapur, pp. 644– 650 (2007). www.books.google.co.ib/books, ISBN:81-8424-259x 7. Murthy, T.S.R., Shravan Kumar, T.: A method for specifying the deviation of 3D reference axes. In: 7th National Conference on Precision Engineering, COPEN7, 10–11 Dec 2011, College of Engineering, Pune, pp. 399–402 (2011) 8. Murthy, T.S.R., Shravan Kumar, T.: Methods for specifying the deviation of 3D reference axes. In: All India Machine Tool Design and Research Conference (AIMTDR 2012), Jadavpur University, Kolkata, pp. 844–847, Dec 2012 9. Murthy, T.S.R., Shravan Kumar, T.: Special cases in specifying the deviation of 3D reference axes. In: 8th National Conference on Precision Engineering, COPEN8, NIT Calicut, pp. 883– 887, Dec 2013 10. Murthy, T.S.R., Shravan Kumar, T.: More special cases in specifying the deviation of 3D reference axes. In: All India Machine Tool Design and Research Conference (AIMTDR 2014), IIT Guwahati, paper O687, 12–14 Dec 2014. www.iitg.ac.in>aimtdr2014>proceedings 11. Murthy, T.S.R., Shravan Kumar, T.: A simplified way of specifying the deviation of 3D reference axes. In: International Conference on Meso, Micro and Nano Technology, COPEN9, IIT Mumbai, paper 106, 10–12 Dec 2015 12. Murthy, T.S.R., Rao, B.R., Abdin, S.Z.: Evaluation of spherical surfaces. Wear 57, 167–184 (1979). ISSN:0043-1648 13. Murthy, T.S.R.: Methods for evaluation of elliptical profiles. Int. J. MTDR 25, 209–312 (1985). ISSN:0890-6955 14. Murthy, T.S.R.: A method for evaluation of paraboloid surfaces. In: CIRP International Seminar on Intelligent Computations in Manufacturing Engineering, Capri, Italy, July 98, pp. 439–445 (1998) 15. Murthy, T.S.R., Abdin, S.Z.: Minimum zone evaluation of surfaces. Int. J. Mach. Tool Des. Res. 20(2), 123–136 (1980). https://doi.org/10.1016/0020-7357(80)90024-4, ISSN:0890-6955 16. ROMER, Hexagonal manufacturing company. www.hexagonmi.com 17. https://en.wikipedia.org/wiki/Torus 18. Legrand, J., Popineau, Y., Berot, S., Gueguen, J., Nouri, L.: Application of torus reactor to chemical and enzymatic modifications of plant proteins. In: Plant Proteins from European Crops, pp. 297–302 (1998). https://doi.org/10.1007/978-3-662-03720-1_50

Chapter 56

An Artefact-Based Continues Performance Verification of Coordinate Measuring Machine Goitom Tesfay

and Rega Rajendra

Abstract Coordinate Measuring Machine (CMM) is a vital measuring machine to maintain the quality of the manufactured products. The movement along the threedirection XYZ of the machine during measurement affected by an error that arises from different sources. The measurement uncertainty range of the CMM should maintain through regular calibration. However, the interim calibration is costly and time-consuming as well as there is no specific time schedule for the next interim calibration. In this work, an artefact route for evaluation of CMM is suggested. The artefact is manufactured with certain geometric features and calibrated at NMI (India). The same artefact is again used for calibration the own CMM at shop floor. By this, it is possible to calibrate own CMM any time and as frequently as required. Comparing calibration measurement of the artefact at NMI and shop floor CMM, reliable results are obtained and suggested these routes to industries. Keywords CMM · Uncertainty analysis · Artefacts · Performance evaluation

56.1 Introduction Quality is the main factor for companies to compete with their sectors. Not only the quality of the product, but the processes of the production should also maintain quality because the good production processes only lead to good quality of products [1]. The touch trigger type CMM is the most available measuring machine among industries. The vibration of operation, geometrical contact of the different components of the CMM part, the temperature of the measuring room, part accessibility, deflection of measuring probe, sampling strategy, CMM hardware, fitting algorithm, etc. are the main cause of the deviation on the uncertainty of the machine from its specification [2]. The uncertainty deviation of the CMM can be reduced by frequently checking of the CMM. The need for verification and test using regular calibration is explained in G. Tesfay (B) · R. Rajendra Department of Mechanical Engineering, College of Engineering, Osmania University, Hyderabad 500007, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_56

671

672

G. Tesfay and R. Rajendra

detail on ISO 10360-2:2009 [3]. The ISO 17025:2005 explained how the calibration laboratories, the competency test are important and the interim check should do to maintain the calibration equipment status within the define confidence [4]. The interim evaluation of CMM is costly and time-consuming processes starting from the specialised owner of the CMM Company slot booking, calibration instrument setup and so on. Due to the above reasons, frequently interim evaluation of the CMM is not feasible. Above all, if the cost calibration of CMM is above the objective of the measurement, the company is not profitable. It is better to check the CMM whether it needs calibration or not with the help of the fast, economical and easy way of evaluating performance using the developed artefact. If the status of the CMM is good for the specified objective, it is better not to calibrate. The researches done related to this objective of the paper are explained below. Fernando et al. [5] they developed their own calibrated artefacts with a high number of geometrical components and the calibrated artefact maintains both good metrological quality and geometric features without considering the effect of thermal stability. Concluded that there is still a need to develop more types of simple and special artefacts. Peggs [6] states that procedure to estimate the magnitude of the geometrical errors may be to measure the departures from the correct geometry of all degrees of freedom of the machine using a laser interferometer, straight-edge, precision level, autocollimator, and other instruments. These methods are, however, time-consuming because many different experimental arrangements are needed to cover the principal error sources. Suggested that because these techniques are expensive and require skilled operators, considers the use of artefacts as an alternative calibration technique. Miguel [7] CMM can be calibrate using a wide range of artefacts such as length bar, ball-ended plate, hole plates, gauge blocks or user define artefacts, etc. The use of some of the artefacts is reviewed in his work. Finally, concluded that the developed artefact should consist more number of geometry based on the requirement of the industry. In this work, a new type of artefact as shown in Fig. 56.1 is developed. Measurement is done in two CMM’s, the CMM (laboratory) that is under investigation and the CMM with good uncertainty as compared to the CMM (laboratory) from the calibration centre CMM (NMI). The performance of the CMM (laboratory) checked based on the CMM at NMI (India).

56 An Artefact-Based Continues Performance Verification …

673

Fig. 56.1 New designed and machined artefact model (dimensions are given in mm)

56.2 Theory of CMM Uncertainty 56.2.1 Uncertainty Factor of the CMM The major factors that affect the accuracy of the coordinate measuring machine are quantified using both Types A uncertainty and Type B uncertainty. As shown in Table 56.1 and according to the international vocabulary of basic and general terms in metrology [ISO VIM (GUIDE 99999)] [8], the Type A uncertainty is uncertainty that quantifies using statistical methods and the Type B uncertainty quantify using experimental methods. To compare the measurement result of the reference CMM to that of the laboratory CMM, both Type A and B measurement uncertainties used to calculate the expanded uncertainty. The combined uncertainty is calculated based on the standard uncertainty of each factor and the proportional constant, i.e. sensitivity coefficient. Table 56.1 The major measurement uncertainty sources of CMM Uncertainty factors

Uncertainty

Uncertainty distribution

Uncertainty associated repeatability (u rep )

A

Normal

Uncertainty associated temperature (u t )

B

Rectangular

Uncertainty associated reproducibility (u rp )

A

Normal

The uncertainty of measuring tip radius correction (u p )

B

Normal

Uncertainty related to the sampling strategy (u s )

A

Normal

674

G. Tesfay and R. Rajendra

  n  Uc (y) =  [C1 U (xi )]2

(56.1)

i=1

where Uc (y) is combined standard uncertainty, C1 is sensitivity coefficient, U (xi ) is the standard uncertainty of ith number of uncertainty factors. Calculation of Measurement Uncertainty Budget (Based on [2] used new methodology). From Table 56.1, the five major uncertainty contributors to the CMM measurement are listed. The method of statistical uncertainty analysis of this fiveuncertainty is explained in [9]. The detail uncertainty analysis of each factor is explained below. Uncertainty associated with repeatability (u rep ). This uncertainty is due to the error arises during repeated measurement of the outer diameter circle. 1 1  SD u rep = √ ∗ √ N1 N2  1 SD = ( j X i − j X )2 N1 − 1

(56.2)

(56.3)

where N1 is the number of measurement repetitions in one orientation, N2 is a number of orientations, SD is the standard deviation of the measurement in one orientation, j X i is the ith measurement in the jth orientation and j X is the jth orientation mean value measurement. Uncertainty associated with temperature (u t ). Measurement is correct and reliable if the measurement is taken at a controlled room temperature. The ISO 10360-2 states that the ideal measurement temperature of the CMM is 20°C. However, in practical achieving, this temperature is very difficult, and uncertainty due to temperature is always not zero. From different literature, temperature uncertainty is the most dominant factor.  ( j X i,max − j X i,min )2 ∗ α ∗ 0.1 ∗ t (56.4) ut = 12 where ( j X i,max ) is the jth orientation ith maximum measurement, ( j X i,min ) is the jth orientation ith minimum measurement value, α coefficient of thermal expansion and t change in temperature between the maximum and minimum. Uncertainty associated with reproducibility (u rp ). The condition of measurement, i.e. out of a set of conditions that includes different locations, operators, measuring systems and replicate measurements on the same or similar objects (ISO 57251:1994).  1 u rp = √ N2

1  j ( X − X) N2 − 1

(56.5)

56 An Artefact-Based Continues Performance Verification …

675

where ( j X ) is the jth orientation measurement of X. Sampling Strategy Uncertainty (u s ). Uncertainty due to sampling is the uncertainty induced in considering the small size of sample for evaluation. It is Type A uncertainty. It calculated as: us = √

σpc n−x

(56.6)

where σpc is the point coordinator error, n is the number of sample points probed from the feature and x is the minimum number of point required to define a given feature.  2 2 + σstylus (56.7) σpc = σMPE where σMPE is probing error and σstylus the probing error due to stylus length. Combined Uncertainty (u c ). Uncertainty is the combined effect of different standard uncertainties. For the above given standard uncertainties, the magnitude of combined uncertainty is calculated the square root of the sum of individual standard uncertainties.  (56.8) u c = u 2rep + u 2t + u 2rp + u 2s Expanded Uncertainty (u e ). The confidence band interval expresses based on the uncertainty interval and the actual measurement found within this band of the interval. To calculate the bandwidth, the coverage factor of the measurement used. In this paper for 95.45% confidence, the coverage factor is 2. Based on the coverage factor and the combined uncertainty, the expanded uncertainty of the machine is ue = k × uc

(56.9)

where k is coverage factor according to the t-distribution for the number of degree of freedom and confidence level.

56.3 Experimental Work Done The artefact shown in Fig. 56.1 of hole diameter 12 mm circle is measured at NMI (India), and the same is measured at the laboratory CMM. The CMM (laboratory) specification is MITUTOYO: a model of Crysta-plus M544 CMM with, TP2 probe and stylus diameter 2 mm. Maximum permissible error of the machine (MPE) = ±(3.5 + 4.5 L/1000) µm. Measurement took at a protected environmental temperature of 20 ± 2 °C based on the CMM standard. For the same artefact and location of the sample measurement, the artefact measured using the reference CMM

676

G. Tesfay and R. Rajendra

(Legex9106) from the calibration centre at NMI (India) of uncertainty MPE = ±(0.55 + L/1000) µm.

56.3.1 New Design Aluminium Alloy (Al6061) Artefact Aluminium alloy (Al6061) is used for artefact due to workability, availability and good surface finish. The developed artefact is consists of the basic geometry and a free-form surface for dimensional and geometrical measurement. After the artefact is designed in solid works 2016, the same machined in the vertical milling. Measurement is done using the CMM (NMI) and the CMM (laboratory). After measurement, to validate the status of the laboratory, CMM comparison done between the CMMs. Since in laboratory and industries different type of CMM exists, the methodology should work for all types of CMM. To obtain the measurement capability index, the expanded uncertainty of the CMM (NMI) divided by its maximum permissible error (MPE) as given by the manufacturing company. In this work, the proposed hypothesis is, the ratio of expanded uncertainty of the CMM (laboratory) to its maximum permissible error is less than to the CMM (NMI) ratio as shown in Eq. (56.10). The ratio of tolerance to expanded uncertainty approximately constant irrespective to change the type, brand or model of CMM as explained in detail in [9], because of the maximum permissible error of the laboratory CMM is always greater than the CMM (NMI). If the temperature factor of the laboratory maintains properly. From ISO/TS 15530-3 [10], the expanded uncertainty is expressed ue = |b| + 2σ , which is the expanded uncertainty is equal to the absolute of bias plus the twice of the measurement standard deviation (σ ). u e laboratory CMM u e NMI CMM < MPE laboratory CMM MPE NMI CMM

(56.10)

The measurement capability index F u e = |Measured value − Reference value| + 2σ F=

Expanded uncertainty of NMI CMM MPE of the NMI CMM

[|Measured value − Reference value| + 2σ  < F ∗ MPE laboratory CMM

(56.11) (56.12)

(56.13)

For the comparison, the maximum mean diameter of the circle measured using the CMM (laboratory) is considered. The expanded uncertainty from the CMM (NMI) is used as reference uncertainty, i.e. 0.8 µm (Table 56.2). After the measurement, the capability index is determined by Eq. (56.12). The performance of the CMM

56 An Artefact-Based Continues Performance Verification …

677

Table 56.2 Comparative measured results from CMM (NMI) and (laboratory)

Avg. (mm)

Measurements from the laboratory CMM

Measurements from the NMI CMM

XY plane

XZ plane

YZ plane

XY plane

XZ plane

YZ plane

11.9990

12.0024

11.9987

11.9998

12.0024

12.0015

12.0026

12.0530

12.0870

12.0016

12.0031

12.0018

12.0390

11.9943

11.9998

12.0090

11.9993

11.9998

11.9980

11.9958

12.0370

11.9980

11.9998

12.0015

11.999

12.0047

12.0078

11.9990

12.0017

12.0028

12.0037

12.0015

12.0092

12.0007

12.0019

12.0012

12.0081

11.9992

12.0011

12.0021

11.9995

12.0018

11.9990

12.0009

12.0081

11.9990

12.0039

12.0120

11.9980

12.0081

12.0011

11.9988

12.0019

12.0019

12.0077

12.0052

12.0009

12.0027

12.0020

11.9998

11.9980

11.9996

12.0082

11.9980

11.9997

12.0080

12.0510

11.9989

12.0052

12.0013

11.9989

12.0251

12.00859

12.0053

12.0137

12.0008

12.0012

12.0048

Expanded uncertainty (mm) (NMI) Certificate 0.0008

(laboratory) is determined through comparison to the measurement uncertainty as explained in Eq. (56.13). F=

0.8 = 1.455 0.55

The above result means that the laboratory CMM to be close accuracy to that of the NMI CMM the maximum permissible error should increase by 45.5%. [12.0137 − 12.0048] + 2 ∗ 0.0176 < 1.455 ∗ 0.00350.0441 < 0.00509(false) Based on the above result, the laboratory CMM needs calibration. Since the interim calibration did not perform for a long time. This performance evaluation is effective as compared to the other similar works [5–7] so that biases are ignored in most of these literatures. In addition, most of the above literature uncertainty analysis of the CMM is based on Guide to the Expression of Uncertainty in Measurement (GUM) methodology but in this work, new type of uncertainty analysis is introduced.

678

G. Tesfay and R. Rajendra

56.4 Conclusions Performance of the CMM (laboratory) analysed to that of the CMM (NMI) which is located at the NPL India, New Delhi. The CMM (laboratory) performance is good if the proposed hypothesis of Eq. (56.13) is satisfied. The right-hand value of the above equation is 0.0441, which is less than the left-hand value 0.00509 that is false based on the proposed hypothesis. The CMM (laboratory) needs calibration. This is due to the CMM (laboratory) did not calibrate for a long time. From the above numerical results, satisfactory results obtained for performance evaluation of the shop-floor CMMs.

References 1. Giovanni, M., Stefano, P.: CMM measurement uncertainty reduction via sampling strategy optimization. In: 9th Biennial ASME Conference on Engineering Systems Design and Analysis, pp. 1–8, Haifa, Israel (2008) 2. Jerzy, A.S.: Coordinate Metrology: Accuracy of Systems and Measurements. Springer, Berlin, Germany (2016) 3. International standard ISO, 10360-2:2009: Geometrical Product Specifications (GPS)—Acceptance and Reverification Tests for Coordinate, Switzerland (2005) 4. International standard ISO, 17025:2005: General Requirements for the Competence of Testing and Calibration Laboratories, Switzerland (2005) 5. Fernando, A.M.F., de Jesus, V.O., Angel, M.S.P.: Evaluation of the performance of coordinate measuring machines in the industry, using calibrated artefacts. In: Manufacturing Engineering Society International Conference, MESIC, pp. 659–668, Porto, Portugal (2013) 6. Peggs, G.N.: Creating a standards infrastructure for co-ordinate measurement technology in the UK. Ann. CIRP 38, 521–523 (1989) 7. Paulo, C.M.: CMM Performance Verification Considerations for a Co-ordinate Measurement Technology in the UK. Ph.D. Thesis, University of Birmingham (1998) 8. International standard ISO, VIM (DGUIDE 99999): International Vocabulary of Basic and General Terms in Metrology (VIM) 9. Paulo, H.P., Robert, J.H.: Measurement uncertainty for coordinate measuring the machine. In: Coordinate Measuring Machine and Systems. CRC Press, Florida, USA (2015) 10. International standard ISO, 15530-3:2011(E): Geometrical Product Specifications (GPS)Coordinate Measuring Machines (CMM): Technique for Determining the Uncertainty of Measurement. Part 3: Use of Calibrated Workpieces or Standards, Switzerland (2011)

Chapter 57

Development of Welding Fixture for Rocket Motor Casing Assembly Venkateswarlu Chepuru, P. Kiran and B. Hari Prasad

Abstract Rocket motor casing assembly is the critical and most significant assembly for an aerospace vehicle. Rocket Casing assembly is welded construction, consists of flow formed shell, front flange, aft flange, and external components like launch shoe, wing brackets, and wire tunnel lugs. The assembly of these components is carried out in two stages by tungsten insert gas welding. First stage is assembly of flanges to the shell and second stage is the assembly of external components to the flange subassembly. Four welding fixtures are developed for realization of rocket motor casing assembly. First fixture is designed for stage one (flange subassembly) to maintain correct relationship and alignment between flanges and shell. Remaining fixtures to align, locate the launch shoe, wing brackets, and wire tunnel lugs in a given orientation and position with respect to motor casing. Designed, developed, and utilized welding fixtures for realization of rocket motor casing. Keywords Aerospace · Welding fixture · Rocket · Motor · Orthogonality

57.1 Introduction The motor casing is a subassembly of a rocket motor which is made of maraging steel 250 grade a well-known material in the aerospace domain. Maraging steel is an age-hardened (maraging) iron–nickel steel, combining ultrahigh strength, toughness, and resistance to crack propagation. The main application of the motor case is to store the high energy propellant. Hence, the total subassembly is located at the aft end of the missile mechanical system. The subassembly consists of six parts, flow formed shell, front flange, aft flange, launch lug, wing brackets, and wire tunnel lugs which have to be welded by using TIG welding in the respective order. The flanges have to be welded to motor case shell by using C-seam (circular-seam) semiauto TIG welding process. Launch lug, wing brackets, and wire tunnel lugs to be fillet welded V. Chepuru (B) · P. Kiran · B. Hari Prasad Defence Research and Development Laboratory DRDO, Ministry of Defence, Kanchan bagh, Hyderabad 500035, Telangana, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_57

679

680

V. Chepuru et al.

Fig. 57.1 Flanges assembled on shell

on motor case using manual TIG welding. Semiautomatic TIG welding machine with dia. 1.2 mm filler wire is used for welding process. Design of fixture for welding of flanges needs to be carried out to control the concentricity 0.3 mm between the end machined flanges which are 2000 mm apart and to achieve the minimum [1] mismatch 0.2 mm of the components. The shaft/support pipe is designed as main part of the welding fixture of flange and different parts of the fixture are mounted on this shaft/support pipe. Fabricated all the items of the firststage welding fixture and assembled the three components during c-seam welding operation (Fig. 57.1). Design of fixture for welding of wing bracket needs to be carried out to control the orthogonality of 45° (±0.1°), canting angle 90° (±0.1°), and axial distance of the inboard panel from front flange 300 mm (±0.05) of the rocket motor. The motor is consists of four-wing brackets which are oriented at 90° with reference to each other and the Cant angle of the wing brackets is defined as an angle between the symmetry plane of the wing brackets and missile axis. Design of launch lug welding fixture is to control symmetry [2] of 0.2 w.r.t central plane of the rocket motor and distance between two lugs. The wire tunnels lugs which are of 39 symmetry and various critical distances from front flange of the rocket motor casing and fixture to be developed for the above precise requirements of the rocket motor.

57.2 Design of Fixtures 57.2.1 Welding Fixture for Assembly of Flanges to Motor Case The C-seam welding fixture is designed with minimum miss-match tolerance of 0.2 mm and maintaining concentricity of 0.3 mm at both end flanges (Fig. 57.2). The cross-sectional view of the c-seam welding fixture is shown in Fig. 57.3iii. Expanding mandrel with copper backup is shown in Fig. 57.4. ‘Expanding Mandrels’ are used inside the welding fixture to correct the ovality of rocket motor shell during c-seam welding. Expanding mandrels used in the welding fixtures have the following two functions in the fixture assembly: it supports the components from inside during welding

57 Development of Welding Fixture for Rocket Motor Casing Assembly

(b) °2

5'

(c)

25 0°

0.2 A

'

°

45° 0

4

x

90

(a)

681

8 .1

1

+0

0. A

= 39 =

0°25'

(Typ)

90°

Fig. 57.2 Rocket motor with external components. a wing brackets, b launch lug, c wire tunnel lugs

(i)

(ii)

Fig. 57.3 C seam welding fixture

Fig. 57.4 Expanding mandrel assembly

(iii)

682

V. Chepuru et al.

process. And also provides copper backup to extract heat generated during welding so that less heat will be taken away by the maraging steel components. The expanding mandrel assembly has a central conical shaft which is held between the six equally sized segments of the expanding mandrel. The six segments of expanding mandrel expand or contract, when the threaded nut is rotated clockwise or anticlockwise directions, respectively as shown in Fig. 57.5. The copper segments are provided to extract heat generated during welding. • Welding fixture assembly for launch lug to motor casing: The rocket motor is mounted between two locators, i.e., Front End Locator and Aft End Locator. The four flat faces in each locator, along with ‘T Section rods,’ at the ends ensure the orientations of both the end flanges. The wire tunnel is held and supported by other elements, i.e., supporting blocks and clamping beam of the fixture (Fig. 57.6). • Welding fixture assembly for wing brackets to motor casing (Fig. 57.7)

Fig. 57.5 Schematic of functioning of expanding mandrel assembly

Fig. 57.6 Launch lug welding fixture

57 Development of Welding Fixture for Rocket Motor Casing Assembly

683

Fig. 57.7 Wing bracket welding fixture

Fig. 57.8 Wire tunnel bracket welding fixture

• Welding fixture assembly for wire tunnel brackets to motor case (Fig. 57.8)

57.3 Major Components and Manufacturing Criticalities of Fixture Front Locator (Fig. 57.9) is used to mount the rocket motor on both ends. The accuracy of flat surface on the outer diameter of the front locator to be maintained to control the orientations of all the external components with reference to the central axis of motor. During the realization of end locator, orthogonality [3] of flat faces needs to be ensured within tolerance. One of the faces is oriented by 45° from an orientation controlling hole, i.e., R2T, and then other faces are machined perpendicular to adjacent faces. These components are finish machining is done with grinding and wire EDM process to achieve the critical tooling requirements. Aft Locator (Fig. 57.10) is used to support the rocket motor from rear side. Similar to Front End Locator it has four flat faces oriented at 90° to each other. The

684

V. Chepuru et al.

Fig. 57.9 Front locator

Fig. 57.10 Aft locator

Fig. 57.11 a Wing lug stopper, b wing lug stopper assembly

alignment of flat faces of Rear End Locator with corresponding faces of Front End Locator is critical. Wing Lug Locator—wing lug locator is used to locate the launch lug to rocket motor case through Aft locator. It controls the position [2] and orientation of launch lug (Fig. 57.11). Copper Backup Bar: Copper backup bar is used to extract heat during welding. Backup supports are used during welding of significant thickness for easy penetration control where welding is

57 Development of Welding Fixture for Rocket Motor Casing Assembly

685

carried out from one side only. The role of backup support is also essential to protect the underbead from atmospheric oxidation and to keep it smooth and of high quality. This is achieved by Argon gas purging through a number of holes.

57.4 Calculation and Analysis

Mass of Weld Backup Support The heat to be dissipated depends largely on the mass of the backup support. Hence, calculation of the mass has been carried out to ensure that the mass of copper backup in fixture is above the required mass as dictated by thermal analysis. Welding parameters: Voltage, E = 10 V Current, I = 55 A Travel speed, v = 63.5 mm/min = 1.06 mm/s Total heat input [H]: Rate of heat input (q) = E I H/t = E I H/L = E I /V H = E I /V in J/mm H = (10 × 55)/1.06 H/L = 518.86 J/mm Total heat input (H ) = 518.86 ∗ 989.1 = 513.212 kJ Let, H n be the actual heat transferred to the workpiece considering small electrical losses in arc. Then Hn = ( f 1 × E × I )/v = f 1 × H where f 1 = Heat transfer efficiency = 0.7 for TIG welding (considered as 70% efficient).

Hn = 0.7 × 513.212 = 363.20 kJ

686

V. Chepuru et al.

Specific melt energy of base metal MDN 250 (u) = Hm /Volume = [mC(Tm − To ) + m Hf ]/volume = r [C(Tm − To ) + Hf ] where Hf C Tm To r

Heat of fusion = 1126 × 103 J/kg (for M-250) Specific heat of weld metal = 0.813 × 103 J/kg (for M-250) Melting temperature of base metal = 1413 °C Temperature of base metal prior to weld = 20 °C Density of weld metal (M-250) = 8.1 × 103 kg/m3

Therefore, u = Hm /vol = r [C(Tm − To ) + Hf ) = 8.1 × 103[813(1393) + 1126 × 103] = 1.8 × 1010 J/m3 Heat dissipated = Q = Hn − Hm = Hn − [u × A × L] where, A Cross-sectional area of weld bead = 16.72 mm2 L Length of weld metal = 989.1 mm H m 18 × 16.72 × 989.1 = 233,674.875 J = 233.674 kJ So, Heat dissipated to component and tool is given by the expression [4]: Q = Hn − Hm Heat dissipated = 363.207 − 233.674 = 129.53 kJ Mass of backup bar Let, Q = heat dissipated in joules Q Total = Q M250 + Qcu 70 × 103 = (mCdT )M−250 + (mCdT )cu Now (mCdT )M−250 where m

Mass of maraging steel

57 Development of Welding Fixture for Rocket Motor Casing Assembly

687

=r×A×L = (8.1 × 103) × 13.12 × 10−6 × 0.989 = 0.105 kg C Specific heat of maraging steel = 813 J/kg K dT Change in temperature = 1393 K Heat absorbed by copper backup bar H cu = Q − (H ) M−250 (mcdT )cu = 129.5 × 103−(0.105 × 813 × 1393) = 10, 586.56 J mcu = 10, 586.56/(380 × 150) = 2.086 kg Since along the total length of rocket motor, copper backups are required. So, mass of each copper backup mcu = 2.08 kg Now, Volume = mass/density = 2.081/8950 = 2.324 × 10−4 m3 So, calculated volume = 232.4 cm3 Based on the above calculations, volume of backup bar should be above 232.4 cm3 .

57.5 Result and Conclusions In this paper, the welding fixture for motor casing assembly is designed successfully and in this process, the following conclusions are made. Fixtures are designed to meet the requirements of concentricity of the flanges, orthogonality, canting and distance of wing brackets from front flange of rocket motor, launch lug position, and wire tunnel lugs position during welding of the rocket motor assembly. Welding fixtures are designed to hold and position the components of the rocket motor casing assembly for TIG welding. The fixtures are aided to welding process as these have provision Argon gas purging below the weld pool and copper backup for extraction the heat during the welding process. The calculations using the thermal aspects of the arsenic copper and the maraging steel enable to arrive at the optimum mass of the backup bar, which was found to be 2.08 kg and this ensures that the heat evolved during the welding of the joint is absorbed toward the backup support ring, and the mass of this backup support is calculated in order to sustain this hot gas without much thermal expansion.

688

V. Chepuru et al.

Manufacturing process plan is made corresponding to fixture design. Critical geometrical tolerances are controlled in fabrication of fixture to achieve final required tolerance specifications in rocket motor assembly. The rocket motor is realized within requires accuracy using welding fixtures by maintaining concentricity 0.3 mm and minimum mismatch 0.2 mm of the components and orthogonality of 45° (±0.1°), canting angle 90° is achieved.

References 1. Murthy, T.S.R., Abdin, S.Z.: Minimum zone evaluation of surfaces. Int. J. Mach. Tool Des. Res. 20, 123–136 (1980) 2. Kanada, T.: Modern Metrology Concerns, pp. 281–289. InTech (1995) 3. Samuel, G.L., Shunmugam, M.S.: Use of limacon and limacoid in evaluation of circularity and sphericity errors. In: All India Manufacturing Technology Design and Research Conference, Chennai, India, pp. 145–147 (2000) 4. Yoshizawa, T.: Handbook of Optical Metrology: Principles and Applications. CRC Press, Florida, U.S.A. (2009)

Chapter 58

Generation of Sequence of Machining Operations Through Visualization of End Product G. V. S. S. Sharma , P. Srinivasa Rao

and B. Surendra Babu

Abstract Geometric modeling in a computer facilitates spatial visualization of the component. This visualization is applied for generating the manufacturing process sequence of the component. In the present work, a generic procedure for generating machining process sequence through spatial visualization is proposed and specifically applied onto a case study of connecting rod manufacturing. A graphical tree approach is adopted for identification of locating surface for succeeding machining operations. Cluster diagram groups the similar machining operations in order to simplify the machining information flow and lays the foundation for the machining process route. The spatially visualized process sequencing (SVPS) procedure charted in this paper helps the process engineers in charting the manufacturing process sequence in the design stage and before production of the component. The present study employs the spatially visualized computer-aided geometric model, rooted tree graph and cluster diagram for the industrial case study on connecting rod process sequencing. Keywords Spatial visualization · Machining process sequence · Cluster diagram · Spatially visualized computer aided process sequencing

58.1 Introduction Computer-aided geometric modeling offers the unique advantage of visualizing the component even before it is manufactured [1]. This advantage of component visualization enables the process engineers to determine the machining process sequence of the product. G. V. S. S. Sharma (B) GMRIT, GMR Nagar, Rajam, AP 532127, India e-mail: [email protected] P. Srinivasa Rao Centurion University, Parlakhemundi, Odisha, India B. Surendra Babu GITAM, Visakhapatnam, AP, India © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_58

689

690

G. V. S. S. Sharma et al.

Process planning in the present-day approach encompasses a large domain of activities ranging from process sequencing, measurement system analysis, tool management, process capability and process control. A prototype-based incremental process planning structures the process plan by identifying the crucial part surfaces early in the product design phase [2]. Process sheets are formulated by emphasizing upon feature recognition and feature extraction [3]. Before charting the process sequence, the geometry of the product must be thoroughly studied. There exists a research gap in the area of process sequence generation taking the spatial visualization of the product into consideration. This research gap is fulfilled by formulating a novel Spatially Visualized Process Sequencing (SVPS) approach which proposes the machining operations based on the surface profile, topography, spatial views and locating surfaces defined through a rooted tree structure. Spatial visualization stands the test of time in clarifying the engineering concepts by generating 3D spatial views from the orthographic projections [4]. Thus, spatial visualization paves way for suggesting an appropriate manufacturing operation suiting the geometry, surface profile and features of the product. This research paper focuses on generation of initial process sequence by spatially visualizing the product. The initial study of the process is done on the basis of identification of locating surfaces for the interdependent machining operations. This interlinking of machining locating surfaces is obtained through rooted tree graph. This research work also proposes a cluster diagram where the machining operations of similar nature are grouped into a cluster for facilitating the tool management. Generic procedure for generating the manufacturing process sequence is established by spatially visualizing the geometric model of the component in a computer. This generic procedure of SVPS is applied onto the industrial case study of internal combustion engine connecting rod.

58.2 Methodology The present-day process planning software extracts the features of the product [5, 6] but generates the process sequencing intermittently, which at a later stage must be integrated to suit the process. In feature-based design, focus is set on interconnecting of the CAD software systems in a network-based environment in order to achieve the distributed concurrent engineering [7]. The feature recognition science is based on the Boolean algebraic basics applied on formation of surfaces, detection of boundaries and form extraction. But there exists a deficiency of correlation of manufacturing information with the recognized part features [8] and hence leaves a scope for further research. In the present study, focus is set on assigning the manufacturing operation for corresponding part surface profile and topology. To empower the present-day feature extraction aspect, spatial visualization study of the product forms the need of the hour. Hence, integrating process sequencing with spatial visualization imparts a new focus to this area of study. The generic procedure involving the different stages

58 Generation of Sequence of Machining Operations …

691

Fig. 58.1 Flowchart depicting the stages in SVPS approach

in this SVPS methodology is summarized and explained in the form of a flowchart as in Fig. 58.1. Case study of connecting rod machining process sequencing through spatial visualization of the product is explored and described through the stages in Fig. 58.1. The ideology behind selecting the connecting rod component is that it is subjected to severe dynamic loading conditions and involves many critical-to-quality characteristics which are to be taken into consideration during manufacturing.

58.3 Case Study: Connecting Rod Connecting rod machining process sequencing through spatial visualization of the component is charted in this section by tracing the stages in spatially visualized process sequencing of Fig. 58.1.

58.3.1 Stage-1: Geometric CAD Model The geometric CAD model of the connecting rod is generated and displayed in Fig. 58.2.

692

G. V. S. S. Sharma et al.

Fig. 58.2 Geometric CAD model of the anticipated end-product connecting rod

58.3.2 Stage 2: Spatial Orientation Spatial visualization is applied onto the horizons of manufacturing process sequencing. The product proposed to be manufactured is spatially visualized in directions as shown in the Fig. 58.3. It is seen that the hidden or intricate features and all the major features are directly visible by spatially rotating the geometric model in the North (N), West (W), South (S) and East (E) directions. Depending on the surface profile and topography of the product, the machining operation is proposed. Figure 58.4 shows the surface profile and topography of the connecting rod forging. The suggested machining operation based on the spatially visualized surface profiles is delineated in Table 58.1.

Fig. 58.3 Spatial orientation of the connecting rod

58 Generation of Sequence of Machining Operations …

693

Fig. 58.4 Surface profile and topography of the connecting rod

Table 58.1 Suggested machining operation based on the spatially visualized surface profiles Profile code

Surface/profile topography

Suggested machining operation

A

Thrust face flat surface

Surface milling Surface grinding Surface lapping

B

Gudgeon pin end cylindrical surface

Drilling Boring Reaming Honing

C

Crank pin end cylindrical surface

Drilling Boring Reaming Honing

D

Side planar surfaces

Milling Side broaching Side surface grinding

E

Parallel planes as reference for separation of the rod and cap portions

Milling Side broaching Side surface grinding

58.3.3 Stage 3: Rooted Tree Graph In the rooted tree graph, the machining operation is represented by corresponding machining operation number within a circular node [9]. This node number is synonymous to machining operation number. Figure 58.5 depicts the rooted tree graph for connecting rod machining based on the information in Table 58.2. This helps in generating the process sequence.

694

G. V. S. S. Sharma et al.

Fig. 58.5 Rooted tree graph for connecting rod machining

58.3.4 Stage 4: Cluster Diagram In the cluster diagram, similar operations are categorized into clusters. Operation numbers 10, 60 and 100 form cluster A comprising of thrust face rough, intermediate and finish grinding operations. Operation numbers 20, 30, 120 and 130 form cluster B comprising of GP and CP rough and finish boring operations. Operation number 40 forms cluster C comprising of broaching operation. Operation number 70 forms cluster D comprising of bolt hole drilling and reaming operations. Notch milling for journal sleeves positioning form cluster E, and finally, the honing of CP bore forms the cluster F. Arrangement of identical machining operations into clusters facilitates the generation of machining process route and supports the tool management to facilitate tool planning, tool procurement and tool presetting by-and-large. The machining process route lays the stepping stone for calculating the stock removal at each operation step and subsequently dimensioning and tolerancing the component at each machining intervals [10]. Figure 58.6 depicts the cluster diagram of connecting rod machining. Thus, the sequence of machining operations is tabulated in the Table 58.3.

58 Generation of Sequence of Machining Operations …

695

Table 58.2 Generation of rooted tree graph Node number

Discussion

Profile code

140

Crank pin (C.P) bore finishing through honing operation

C

130

(C.P) bore fine boring operation. Location through gudgeon pin (G.P) fine bore at node 120

C

120

G.P. bore finishing through fine boring operation. Location through C.P rough bore at node 30 and finish grinded thrust face at node 100

B

110

Oil hole drilling

100

Thrust face finish grinding. Location through thrust face intermediate grinded surface at node 60 after rod and cap assembly at node 90

A

90

Rod and cap assembly through nuts and bolts fitment. Location through reamed bolt holes at node 70

Assembly operation

80

Keyway milling

Reference surface C

70

Bolt hole drilling and reaming. Location through GP bore rough diameter from node 20, side face thickness and collar thickness at node 40 and intermediate thrust face flatness at node 60

Reference surfaces E and F

60

Thrust face intermediate grinding and joint face grinding operation. Location through deburred component at node 50 and after cap and rod separation at node 40

A

50

Manual deburring

Sharp edges

40

Separation of rod and cap portions through broaching. Location through CP rough bore at node 30 and rough grinded thrust face at node 10

D, E, F

30

Crank pin (CP) rough bore diameter. Location through rough grinded thrust face at node 10 and rough bore G.P. bore at node 20

C

20

Gudgeon pin (GP) rough boring operation. The perpendicularity of GP bore with respect to the thrust face depends on the node 10

B

10

Thrust face rough grinding. Location through the thrust face of the raw material forged connecting rod

A

696

G. V. S. S. Sharma et al.

Fig. 58.6 Cluster diagram of connecting rod machining

Table 58.3 Sequence of machining operations Op. No.

Machining operation description

Critical-to-quality characteristic

10

Thrust face rough grinding is performed on both the thrust face flat surfaces on vertical axis horizontal grinding machine

Thrust face thickness

20

Gudgeon pin bore rough boring is the primary drilling and boring operation for the formation of gudgeon pin bore

Gudgeon pin bore diameter

30

Crank pin bore rough boring is the primary boring operation leading to formation of crank pin bore by leaving stock allowance for the succeeding broaching operation

Crank pin bore diameter

40

Broaching is the operation for separating the rod and cap portions, suitably performed on vertical broaching machine

Side face width Cap collar width Rod collar width

50

Deburring

Edges free from chips

60

Thrust face intermediate grinding is performed on both the thrust face flat surfaces on vertical axis horizontal grinding machine

Thrust face intermediate thickness

(continued)

58 Generation of Sequence of Machining Operations …

697

Table 58.3 (continued) Op. No.

Machining operation description

Critical-to-quality characteristic

70

Bolt hole drilling and reaming is performed on both rod and cap joint faces for the formation of bolt holes

Bolt hole diameter Bolt hole center distance

80

Notch milling is performed for notch formation on both rod and cap portions, for location of the journal bearing sleeves covering the crank pin bore

Notch thickness Notch position

90

Rod and cap assembly is done by fastening the rod and cap portions through applying predetermined torque to the joining bolts and nuts

Nut and bolt assembly torque

100

Thrust face finish grinding is performed on both the thrust face flat surfaces on vertical axis horizontal grinding machine

Thrust face finish thickness

110

Oil hole drilling is for the formation of oil hole for lubricating the connecting rod while in engine assembly

Oil hole diameter

120

Gudgeon pin bore finish boring is the finishing operation of gudgeon pin bore suitable for assembling the sleeve bush

Gudgeon pin bore finish diameter

130

Crank pin bore finish boring is the operation for obtaining the final diameter of crank pin bore and for leaving the necessary machining stock allowance for the succeeding honing operation

Crank pin bore finish diameter after boring

140

Crank pin bore honing is the final super finishing operation done over the crank pin bore in order to improve over its surface finish and ovality

Crank pin bore finish diameter after honing

58.4 Conclusions With the advent of computers, the lead time in process sequencing has drastically reduced. A new SVPS procedure is charted where spatial visualization through computer-aided geometric modeling is applied into the horizons of generating the machining process sequence, in context of case study pertaining to connecting rod. Graphical tree model depicts the flow of machining sequence based on the locating surfaces for succeeding machining operations. Cluster diagram categorizes the similar machining operations in clusters for ease in tooling management and generation of machining sequencing. At the outset, the computer-aided geometric part model helps in spatial visualization of the different stages of machining operations and thereby helps in determining the machining process sequence of the targeted component.

698

G. V. S. S. Sharma et al.

The procedure traced in this paper can be horizontally deployed across the product manufacturing industries and process engineers to generate process sequence across cross-functional team for any product manufacturing process sequencing during the process design stage.

References 1. Niu, Z., Martin, R.R., Langbein, F.C., Sabin, M.A.: Rapidly finding CAD features using database optimization. Comput. Aided Des. 69, 35–50 (2015). https://doi.org/10.1016/j.cad. 2015.08.001 2. Chu, X., Tso, S., Tu, Y.: A novel methodology for computer-aided process planning. Int. J. Adv. Manuf. Technol. 16(10), 714–719 (2000). https://doi.org/10.1007/s001700070023 3. Sadaiah, M., Yadav, D., Mohanram, P., Radhakrishnan, P.: A generative computer-aided process planning system for prismatic components. Int. J. Adv. Manuf. Technol. 20(10), 709–719 (2002). https://doi.org/10.1007/s001700200228 4. Sharma, G., Dumpala, R.: Teaching of mechanical engineering concepts through threedimensional geometric modeling. Int. J. Mech. Eng. Educ. 0306419015603013 (2015). https:// doi.org/10.1177/0306419015603013 5. Deja, M., Siemiatkowski, M.S.: Feature-based generation of machining process plans for optimised parts manufacture. J. Intell. Manuf. 24(4), 831–846 (2013). https://doi.org/10.1007/ s10845-012-0633-x 6. Subrahmanyam, S., Wozny, M.: An overview of automatic feature recognition techniques for computer-aided process planning. Comput. Ind. 26(1), 1–21 (1995). https://doi.org/10.1016/ 0166-3615(95)80003-4 7. Li, W., Ong, S.-K., Fuh, J.Y., Wong, Y., Lu, Y., Nee, A.Y.: Feature-based design in a distributed and collaborative environment. Comput. Aided Des. 36(9), 775–797 (2004). https://doi.org/ 10.1016/j.cad.2003.09.005 8. Babic, B., Nesic, N., Miljkovic, Z.: A review of automated feature recognition with rule-based pattern recognition. Comput. Ind. 59(4), 321–337 (2008). https://doi.org/10.1016/j.compind. 2007.09.001 9. Whybrew, K., Britton, G., Robinson, D., Sermsuti-Anuwat, Y.: A graph-theoretic approach to tolerance charting. Int. J. Adv. Manuf. Technol. 5(2), 175–183 (1990). https://doi.org/10.1007/ BF02601605 10. Sharma, G., Rao, P.S., Babu, B.S.: Process-based tolerance assessment of connecting rod machining process. J. Ind. Eng. Int. 1–10 (2016). https://doi.org/10.1007/s40092-015-0138-2

Chapter 59

Monitoring the Dynamics and Tracking of a Vehicle Using Internet of Things (IoT) Sri Harsha Dorapudi , Vivek Varma Buddharaju , V. V. Vimal Varma and R. Ramesh Abstract The demand for Internet use was not limited to day-to-day activities, also its extended use of creating a global platform for machines to make objects smarter which can communicate, dialog, compute, and coordinate. Identifying these advantages of multiple tasking of the machines by the Internet to make the work faster and smoother. This innovative idea of creating a platform for machines using the Internet and to make them smart, they are enabled by embedding electronics into the objects which seamlessly integrate globally resulting cyber-physical infrastructure. The present investigation centralizes the idea of performing dynamic analysis while designing a vehicle by physically embedding the load cell strain gauge sensors at the critical points of stress concentrations on the vehicle that are identified by the software analysis (ANSYS). The strains and the corresponding stresses are measured from the resistance/voltage change of the strain gauges when deformed, using a programmable microcontroller called Arduino. The retrieved data is monitored through the Internet-enabled device by connecting to the cloud using a Wi-Fi modem. These experimental results of the vehicle tracking system and some experiences on practical implementations can be used to perform stress and strain analysis for developing an automobile and also to secure the vehicle from theft, a Global Positioning System (GPS) tracking device is designed using the Internet of things (IoT) with BLYNK, a mobile/PC application. Keywords Internet of things · Cyber-physical systems · Arduino · Load cell strain gauge sensor · Global positioning system sensor

S. H. Dorapudi (B) · V. V. Buddharaju · V. V. Vimal Varma · R. Ramesh Department of Mechanical Engineering, MVGR College of Engineering (A), Vizianagaram, Andhra Pradesh 535005, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_59

699

700

S. H. Dorapudi et al.

59.1 Introduction The Internet of things (IoT) has emerged as a concept over the last decade and relies on the vision of everyday objects having an end-to-end IP connection and providing with the ability to transfer data over a network through the Internet [1]. The objects which are involved with a wide range in IoT encompass several related systems, such as Radio Frequency Identification (RFID), Machine-to-Machine (M2M), and Wireless Sensor Networks (WSNs). Many business applications based on the IoT have already turned out, notably smart grid or smart infrastructures. IoT is expected to connect diverse technologies to facilitate a wide range of new applications by connecting physical objects together in support of intelligent decision-making [2]. IoT projects use the Arduino as a microcontroller because of the flexibility of the programming and its use [3]. Due to this capability and availability of large sensors and libraries to the program, it can be used to a wide range of applications and a software model (ASIP) which was developed in order to manage the solutions for advanced computational resources as well for the better communication with the clients [4]. The location information which is useful in tracing out a vehicle through IoT which popularizes the use of smartphones with a powerful sensor ensuring security and which in turn provides feasibility for location-based projects. It is also clear that the geographical coordinates retrieved through the use of a GPS module are important in locating the vehicle [5]. In the use of GPS, a well-established technology Global System for Mobile communication (GSM) which makes an efficient source for vehicle monitoring and anti-theft mechanism which has Intel Galileo Gen 2 development board, a quad-band GPS/GPRS/GSM module, Vehicle speed sensor, LCD display, and an open-source Cloud Server to obtain the required data [6]. With the advent of embedment of strain gauges on to the objects which exhibit good linearity and high sensitivity, a detailed study of the effect of different electric parameters of strain gauges need to be done for the measurement [7]. To make such use of strain gauges, the information basing on the characteristics of the strain gauge to be identified for the right selection for different applications. It is also critical to understand the orientation and size for high strain gradient measurements [8]. As well, the bonding made to the structural members to test under different conditions is considered important to safeguard the system from erroneous errors [9]. The importance of electrical load cells for measuring large loads which is suitable for dynamic conditions which in turn could use the magnetic amplifiers and transistors together with automatic selfbalancing bridges for high accuracy. The load cells are extended to many fields of interest like robotics, industrial machinery, etc. for different measurements. In order to make these measurements the basic knowledge of the load cell is necessary [10]. In addition, the selection and use of the Android application are also important for the access and control of devices and store the data which can be further utilized for future findings.

59 Monitoring the Dynamics and Tracking of a Vehicle …

701

59.2 Experimental Methodology 59.2.1 Monitoring the Dynamics of a Vehicle Using IoT Strain gauge load cells are used to convert a force or load into an equivalent electrical signal or digital load value. The applied force or load is translated into voltage through the change or variations in resistance of the strain gauges, which are permanently bonded to the transducer structure. Strain gauge load cells use a full-bridge configuration. The unbalanced bridge in the loaded condition, where the strain gauges are deformed causes an output voltage which is proportional to the applied load. In the present experimentation, a 3 kg straight beam load cell, which is used for recording the stress-strain values is supplied with 5 V excitation voltage from the Arduino UNO board through the E+ and E− pins of the HX711 amplifier. The voltage output of the load cell is connected to the A+ and A− pins of the HX711 amplifier, which amplifies the output to be recorded and is sent to the Arduino with possible connections as shown in the Fig. 59.1. The intention of using a load cell is because of its high sensitivity and capability of recording the small variations in the data and its resistance to fatigue failure for fully reversed cycles under different conditions. The Arduino processes the analog data obtained is converted to digital data, is sent to the cloud by connecting it to the Internet, and is monitored through a PC or any other device. The force output is converted into equivalent stresses and corresponding strains with the help of formulation encoded in the Arduino program, an open software (Arduino.cc). The standard mathematical formulation employed for writing an Arduino code and to generate the required value for the corresponding terms is: 1. Stress, σ = Weight/area = F/A where A is the cross-sectional area of the load cell = 60 × 30 mm

Fig. 59.1 Various pin connections of the circuit and load cell

702

S. H. Dorapudi et al.

2. Strain,  = σ /E where  is the strain to be measured σ is the corresponding stress measured E is Young’s modulus of the material = 200 GPa for Steel

59.2.2 Tracking of a Vehicle Using IoT The IoT-based GPS tracking system tracks the location of the vehicle in terms of latitude, longitude, speed, direction, and the number of satellites. The U BLOX NEO 6M GPS Module is used with power supply 3.3–5 V with a ceramic active antenna with size of 25 × 25 mm and module size 25 × 25 mm with a baud rate of 38,400. The GPS module with Node MCU (ESP8266 12E) is a small electronic circuit that allows connecting to available Arduino IDE software for uploading GPS tracking code to get position and altitude, as well as speed and direction. These tracking devices provide a highly accurate one pulse per second output for precise measurements. The obtained tracking results are established in the BLYNK mobile application.

59.2.2.1

Circuit Connections

The circuit pin connections between GPS Module and Node MCU (ESP8266 12E) as shown in Fig. 59.2a, b which are to be made for obtaining the required data through the Arduino IDE software are Vcc to 5 V GND to GND Rx to D1 Tx to D2 Node MCU an open-source platform with all the connections and necessary libraries like Tiny GPS plus, Software Serial, and Adafruit OLED display installed in Arduino that enables the data communication and conversion of data through the digital pins of Node MCU (ESP8266 12E).

59.2.3 BLYNK Application The android application which is a user-friendly platform creates an interface to the hardware connected to the Internet. It is used as a remote to monitor and control the IoT devices linked to it, which requires authentication to restrict the unauthorized entry and to gain access to the data. It is compatible with different boards and libraries from the application that can be imported according to the need. New projects can

59 Monitoring the Dynamics and Tracking of a Vehicle …

703

Fig. 59.2 a Node MCU connections, b connections between GPS sensor and node MCU

be created as a file and the respective data is stored as shown in the Fig. 59.3a, b. Whenever necessary, the data stored in the application is transferred in the form of a CSV file, to study and understand the information and make decisions. Further, it can also be used for the research which helps in new developments.

59.3 Results and Discussions 59.3.1 Stress-Strain Analysis Using Load Cell The stress and the strain values of interest for the vehicle are generated in the serial monitor with all possible connections of testing when the strain gauge load cell is

704

S. H. Dorapudi et al.

Fig. 59.3 a, b Project creation using BLYNK application

pressurized to a certain limit. The results obtained are as shown in Fig. 59.4a, b from the Arduino code. The strain value is obtained as 0.00 micro units as Young’s modulus of the steel load cell is of the range 200 GPa and the value is very small. The obtained experimental results show that the difference in the values is observed because of the different loading conditions. In the practical application due to different dynamics of the vehicle, the load changes are not static and even more. So the change in the values of the stress and strain will accordingly change with it. The data which is to be stored and reviewed can give relevant information needed for improving the safety of the vehicle, and necessary adjustments can be made while designing it for the future.

59.3.2 Vehicle Tracking Using GPS The GPS tracking system gets the location of the vehicle in terms of latitude, longitude, speed, direction, and the number of satellites. Which not only gives the exact location of the vehicle in terms of direction but also the other details needed for tracking which makes the task simple for the user. To establish the results on the mobile application (BLYNK) as shown in Fig. 59.5, the GPS sensor is powered through Node MCU, and the mobile is connected to the Internet. It was tested with multiple location changes for identifying the accuracy of the terms to be obtained through the application, and the data obtained by the application was recorded.

59 Monitoring the Dynamics and Tracking of a Vehicle …

705

Fig. 59.4 a, b Result of stress and strain using load cell

59.4 Conclusions This study gives a clear depiction of the output where the stresses and the corresponding strains are measured using strain gauge load cell of 3 kg using the IoT network in the dynamic condition, and the results are depicted on the serial monitor of the PC. This investigation provides direct analysis of stress and strain using sensors and IoT while developing an automobile. A priori investigations using the past data or new will help the developers to understand the difficulties in developing

706

S. H. Dorapudi et al.

Fig. 59.5 GPS location using BLYNK application

an automobile. The compact vehicle tracking system using GPS sensor and the IoT network tracks the live location of the vehicle which is visible in the BLYNK application. The location history can also be secured to identify the theft of the vehicle. In addition, the tracking application used in this investigation can be used for security and transportation purposes and can track the exact accident location as well as identify the lost vehicle in an unknown terrain. Using the data obtained through our experimentation, it also gives an understanding to avoid large factors of safety in the design of automobiles. The limitations of such study are integrating the devices and creating a real-time environment or digital twin for testing of large automobiles with a complex design of parts and an advanced software support with which a detailed code is needed for analysis which consumes a lot of time and effort. But the analyzed dynamic data can be of greater use and helps in understanding more about the difficulties and to obtain a possible solution to overcome them.

References 1. Chandrakanth, S., Venkatesh, K., Uma Mahesh, J., Naganjaneyulu, K.V.: Internet of things. Int. J. Innovations Adv. Comput. Sci. IJIACS 3(8) (2014) ISSN 2347 – 8616 2. Al-Fuqaha, A., Guizani, M., Mohammadi, M., Aledhari, M., Ayyash, M.: Internet of things: a survey on enabling technologies: protocols, and applications. IEEE Commun. Surv. Tutorials 4, fourth quarter (2015) 3. Badamasi, Y.A.: The working principle of an arduino. In: IEEE, Electronics, Computer and Computation (ICECCO) (2014)

59 Monitoring the Dynamics and Tracking of a Vehicle …

707

4. Barboa, G., Margolis, M., Palumboc, F., Raimondi, F., Weldin, N.: Taking arduino to the internet of things: the ASIP programming model. Comput. Commun. 1–15 Science Direct (2016) 5. Balbin, J.R., Garcia, R.G., Latina, M.A.E., Allanigue, M.A.S., Ammen, J.K.D., Bague, A.V.B., Jimenez, J.M.: Vehicle door latch with tracking and alert system using global positioning system technology and IoT based hardware control for visibility and security of assets. In: Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM), IEEE (2017) 6. Godavari, T., Umadevi, J.: Cloud computing based real-time vehicle tracking and speed monitoring system. Int. J. Control Theory Appl. 9(4), 1823–1830 (2016) 7. Husik, M., Kulha, P., Jakovenko, J., Vyborny, Z.: Design of strain gauge structure. In: IEEE, ASDAM 2002 Smolen Ice Castle, Slovakia (2002) 8. Younis, N.T., Kang, B.: Averaging effects of a strain gauge. J. Mech. Sci. Technol. 25(1), 163–169 (2011) 9. Tuttle, M.E.: Fundamental strain gauge technology. In: Manual on Experimental Methods for Mechanical Testing of Composites, pp. 17–26. Springer, Berlin (2012) 10. Muller, I., de Brito, R. M., Pereira, C.E., Brusamarello, V.: Load cells in force sensing analysis— theory and a novel application. IEEE Instrum. Meas. Mag. 13(1) (2010)

Chapter 60

Automated Production of Medical Screws Using Titanium Bar on Indigenous Sliding Headstock Automat Manohar Bulbule , Naveen Hosamani

and S. R. Chandramouli

Abstract This paper focuses on the development of effective methodology to machine bone screw on a 5-axis sliding head machine. Bone screw has a complex design involving special thread profiles and contours. This calls for complex machining processes such as taper threading with constant OD, slot milling, profile milling and ID threading. The threading consists of unique profile including corner radius and thread angle. Most critical operation is threading over the length of 30–120 mm. Two manufacturing concepts are envisaged initially, one being thread whirling and the other being usage of form tool. To achieve cost-effective solution, form threading is conceptualized. Total design and manufacturing, process and selection of tooling, programming and experimentation with prototype development constitute the major part of project work. Keywords Medical screws · Surgical components · Sliding headstock

60.1 Introduction Surgical screws can be difficult to produce for a couple of reasons: materials and geometry. Work materials such as titanium, one of the primary materials for bone screws, and stainless steel have unique characteristics, including being poor heat conductors and producing stringy, difficult-to-break chips. Complex geometries are also common with other medical components, including high length-to-diameter ratios, while Swiss machines have a distinct advantage such as headstock being sliding type and job continuously supported by Rotary Guide bush close to cutting area Fig. 60.1. Innovations in the medical field have driven changes in part designs and builders of Swiss-style lathes have had to innovate to keep up the technology need [1]. Sliding headstock automat also known as Swiss-type lathes was mainly designed for small precision parts manufacturing. The technology contains a bar fed by bar M. Bulbule · N. Hosamani · S. R. Chandramouli (B) Design and Development (SPG) Department, ACE Designers Pvt. Ltd., Bangalore 560058, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 M. S. Shunmugam and M. Kanthababu (eds.), Advances in Forming, Machining and Automation, Lecture Notes on Multidisciplinary Industrial Engineering, https://doi.org/10.1007/978-981-32-9417-2_60

709

710

M. Bulbule et al.

Fig. 60.1 Sliding headstock machine concept

feeder through the machine’s spindle and held be collet system. Then bar is supported by synchronous rotating guide bush. This concept helps to machine large length-todiameter ratios with high accuracy. Speciality of sliding headstock automat is that both primary and secondary operations are done in a single set-up (simultaneously operating main spindle and sub-spindle), thus providing advantage in cycle time. CNC sliding head machines are used across wide range of sectors like automotive, medical and general engineering, etc. These machines configured with 5–7 axes for complex and critical turn-mill applications, and the size ranges from Ø3.0 to Ø32 mm.

60.1.1 Structure of Medical Screw The screw is important invention in the mechanical engineering. It converts angular motion into linear motion to transmit power or to develop large forces. Screw has got a four-part construction: head, shaft, thread and tip [2]. The head contains either hexagonal, cruciate, slotted, or sockets for Philips head in design. The shank is the portion of the screw between the head and the threaded region. The thread part consists of core diameter, thread diameter, pitch and lead as shown in Fig. 60.2. The cross-sectional design is usually a buttress or V-thread Fig. 60.3. The tip of the screw is either round or self-tapping. Specialty of the most medical screws is that the threads are taper threads (ID) and outside diameter being constant. In some cases, pitch also varies along the length at a definite interval.

Fig. 60.2 Screw structure

60 Automated Production of Medical Screws …

711

Fig. 60.3 Buttress (left) and V-shape (right) thread profile

60.1.2 Process of Producing Medical Screw Machining of mono-axial medical screw having special form thread and taper core thread diameter can be produced in two methods (titanium alloy) on sliding head automat, either using thread whirling or form single-point threading tool concepts.

60.2 Methodology In this study, cutting speed, feed and depth of cut are the input variables, and cutting forces and surface roughness can be considered as the output characteristics. The following steps are involved for continuous production of medical screw.

60.2.1 Concept 1: Single-Point Form Threading Tool Concept Single-point threading, also colloquially called single-pointing (or just thread cutting when the context is implicit), is an operation that uses a single-point tool to produce

712

M. Bulbule et al.

Fig. 60.4 Single-point thread cutting

a thread form on a cylinder or cone as shown in Fig. 60.4. The tool moves linearly, while the precise rotation of the workpiece determines the lead of the thread. In external thread cutting, the workpiece can be held in either a chuck or a collet. The tool moves across the piece linearly, taking chips off the workpiece with each pass. But in case of sliding headstock automat, the tool remains stationary while the job rotates and moves linearly. Usually, 5–7 light cuts create the correct depth of the thread [3].

60.2.2 Concept 2: Thread Whirling Concept Typically performed on a Swiss-style turning centre, thread whirling consists of feeding a slowly rotating workpiece through a bolted-on whirling adaptor that contains a circular milling tool running at a high rpm as shown in Fig. 60.5. The process neutralizes many of the issues encountered with single-point threading. By positioning the cutting tool near the guide bushing, the need for a support device is eliminated and aggressive metal removal rates can be used to complete a thread with a single pass. Additionally, the achievement of helix angles by tilting the entire tool allows angles of up to 15° as shown in Fig. 60.6 [4]. Fig. 60.5 Thread whirling tool

60 Automated Production of Medical Screws …

713

Fig. 60.6 Thread whirling concept

60.3 Experimental Set-up 60.3.1 Machine Tool Set-up Rigid, high power, precision class ACE make, indigenously developed sliding headstock automat SHA20 Fig. 60.7 with specially designed accessories was used to machine the mono axial medical screw.

Fig. 60.7 Sliding headstock automat ACE SHA20

714 Table 60.1 Chemical composition

M. Bulbule et al. Element

Percentage by weight (%)

Ti

90

Al

6

V

4

C