Design and Development of Heavy Duty Diesel Engines: A Handbook [1st ed. 2020] 978-981-15-0969-8, 978-981-15-0970-4

Table of contents :
Front Matter ....Pages i-xvi
Introduction (P. A. Lakshminarayanan, Avinash Kumar Agarwal)....Pages 1-13
Front Matter ....Pages 15-15
Modern Diesel Combustion (Walter Knecht, P. A. Lakshminarayanan)....Pages 17-65
Supercharging (Walter Knecht, P. A. Lakshminarayanan)....Pages 67-83
Introduction to Turbocharging—A Perspective on Air Management System (D. A. Subramani, R. Dhinagaran, V. R. Prasanth)....Pages 85-193
Topics on Selective Catalyst Reduction (P. Kumar)....Pages 195-236
Strategies to Control Emissions from Off-Road Diesel Engines (M. V. Ganesh Prasad)....Pages 237-274
External Exhaust Gas Recirculation (P. Sahaya Surendira Babu, P. Kumar)....Pages 275-311
Diesel Particulate Filter (K. C. Vora, Kartik E. Gurnule, S. Venkatesh)....Pages 313-339
Conversion of Diesel Engines for CNG Fuel Operation (G. Jeevan Dass, P. A. Lakshminarayanan)....Pages 341-392
Simulation of Gas Flow Through Engine (Neelkanth V. Marathe, Sukrut S. Thipse, Nagesh H. Walke, Sushil S. Ramdasi)....Pages 393-403
Front Matter ....Pages 405-405
Development of Ports of Four Stroke Diesel Engines (Nagaraj S. Nayak, P. A. Lakshminarayanan)....Pages 407-425
Design and Analysis Aspects of Medium and Heavy-Duty Engine Crankcase (Swapnil Thigale, M. N. Kumar, Yogesh Aghav, Nitin Gokhale, Uday Gokhale)....Pages 427-465
Connecting Rod (Prakash R. Wani)....Pages 467-508
Critical Fasteners, Highly Loaded Bolted Joints (Prakash R. Wani)....Pages 509-523
Crankshaft (Prakash R. Wani)....Pages 525-573
Gaskets (Osamu Aizawa)....Pages 575-599
Design of Valve Train for Heavy Duty Application (Aniket Basu, Nitin Gokhale, Yogesh Aghav, M. N. Kumar)....Pages 601-636
Engine Retarders (M. V. Ganesh Prasad)....Pages 637-678
Engine Gear Train Design (Vishal Bhat, M. N. Kumar, Yogesh Aghav, Nitin Gokhale)....Pages 679-729
Piston and Rings for Diesel Engines (Subrata Neogy, Vikas Ramchandra Umbare, Vineet Ahluwalia, P. A. Lakshminarayanan)....Pages 731-761
Cooling, Coolants, and Water Pump and Oil Pump (S. Seetharaman, P. A. Lakshminarayanan)....Pages 763-793
Design of Electronic Control for Diesel Engines (M. Leelakumar)....Pages 795-830
Front Matter ....Pages 831-831
Study of Noise and Vibration Problems Related to Heavy Duty Diesel Engines (P. A. Lakshminarayanan)....Pages 833-884
Front Matter ....Pages 885-885
Future Diesel Engines (Z. Gerald Liu, Achuth Munnannur)....Pages 887-914

Citation preview

Energy, Environment, and Sustainability Series Editor: Avinash Kumar Agarwal

P. A. Lakshminarayanan Avinash Kumar Agarwal Editors

Design and Development of Heavy Duty Diesel Engines A Handbook

Energy, Environment, and Sustainability Series Editor Avinash Kumar Agarwal, Department of Mechanical Engineering, Indian Institute of Technology Kanpur, Kanpur, Uttar Pradesh, India

This books series publishes cutting edge monographs and professional books focused on all aspects of energy and environmental sustainability, especially as it relates to energy concerns. The Series is published in partnership with the International Society for Energy, Environment, and Sustainability. The books in these series are edited or authored by top researchers and professional across the globe. The series aims at publishing state-of-the-art research and development in areas including, but not limited to: • • • • • • • • • •

Renewable Energy Alternative Fuels Engines and Locomotives Combustion and Propulsion Fossil Fuels Carbon Capture Control and Automation for Energy Environmental Pollution Waste Management Transportation Sustainability

More information about this series at http://www.springer.com/series/15901

P. A. Lakshminarayanan Avinash Kumar Agarwal •

Editors

Design and Development of Heavy Duty Diesel Engines A Handbook

123

Editors P. A. Lakshminarayanan Formerly with Simpson and Co. Ltd. Chennai, Tamil Nadu, India

Avinash Kumar Agarwal Department of Mechanical Engineering Indian Institute of Technology Kanpur Kanpur, Uttar Pradesh, India

ISSN 2522-8366 ISSN 2522-8374 (electronic) Energy, Environment, and Sustainability ISBN 978-981-15-0969-8 ISBN 978-981-15-0970-4 (eBook) https://doi.org/10.1007/978-981-15-0970-4 © Springer Nature Singapore Pte Ltd. 2020 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

Foreword

It might appear ironic that a new book on internal combustion engine (ICE) is being brought out, just when the very future of ICE is being widely debated. Technologies do not suffer natural death. They die because research focus and innovative ideas tend to move away from mature technologies into emerging technologies. This book is sufficient evidence that several practising technologists and academic scholars continue to remain focussed on pushing the envelope of ICE capabilities with innovative ideas and deeper research. Such a focus is essential to offer a competitive benchmark to emerging technologies, such as electric power, as well as to smoothen disruptions which will be unavoidable, when there is a generational change in technologies. Recent innovations in technology of ICE have mostly focused on emission reduction, in our quest to retrieve the environment. Whilst these have come about mostly as a result of the threat of regulatory mandates, it is remarkable that these innovations have often resulted in breakthrough results. In the process, collateral benefits to materials science, electronics and fuel chemistry have also been seen. This is not surprising as technology solutions are almost always interwoven with inputs from different branches of science. Research and innovation are not always motivated by threats. An inspirational goal, a moon shot, could be equally effective. It might be necessary to envision an audacious design of ICE, meant and relevant for the transportation needs of the twenty-first century, and direct innovation towards that goal. The book has been edited by two outstanding ICE technologists of our country, Dr. P. A. Lakshminarayanan and Prof. Avinash Kumar Agarwal. I was fortunate to receive my tutorials on ICE from Dr. Lakshminarayanan, who was my valuable colleague at Ashok Leyland. I remain grateful to him.

v

vi

Foreword

This book is a compulsory read for all those who are engaged directly or indirectly with internal combustion engines. R. Seshasayee Former Managing Director Ashok Leyland Ltd. Chennai, India

Preface

Energy demand has been rising remarkably due to increasing population and urbanization. Global economy and society are significantly dependent on the energy availability because it touches every facet of human life and activities. Transportation and power generation are two major examples. Without the transportation by millions of personalized and mass transport vehicles and availability of 247 power, human civilization would not have reached contemporary living standards. The International Society for Energy, Environment and Sustainability (ISEES) was founded at Indian Institute of Technology Kanpur (IIT Kanpur), India in January 2014 with an aim to spread knowledge/awareness and catalyse research activities in the fields of Energy, Environment, Sustainability and Combustion. The Society’s goal is to contribute to the development of clean, affordable and secure energy resources and a sustainable environment for the society and to spread knowledge in the above-mentioned areas and create awareness about the environmental challenges, which the world is facing today. The unique way adopted by the society was to break the conventional silos of specializations (Engineering, science, environment, agriculture, biotechnology, materials, fuels, etc.) to tackle the problems related to energy, environment and sustainability in a holistic manner. This is quite evident by the participation of experts from all fields to resolve these issues. The ISEES is involved in various activities such as conducting workshops, seminars, conferences, etc. in the domains of its interests. The society also recognizes the outstanding works done by the young scientists and engineers for their contributions in these fields by conferring them awards under various categories. Second International Conference on ‘Sustainable Energy and Environmental Challenges’ (SEEC-2018) was organized under the auspices of ISEES from 31 December 2017 to 3 January 2018 at J N Tata Auditorium, Indian Institute of Science Bangalore. This conference provided a platform for discussions between eminent scientists and engineers from various countries including India, USA, South Korea, Norway, Finland, Malaysia, Austria, Saudi Arabia and Australia. In this conference, eminent speakers from all over the world presented their views related to different aspects of energy, combustion, emissions and alternative energy vii

viii

Preface

resource for sustainable development and cleaner environment. The conference presented five high voltage plenary talks from globally renowned experts on topical themes, namely, “Is it really the end of combustion engines and petroleum?” by Prof. Gautam AlphaTIG, Saudi Aramco; “Energy Sustainability in India: Challenges and opportunities”, by Prof. Baldev Raj, NIAS Bangalore; “Methanol Economy: An option for sustainable energy and environmental challenges”, by Dr. Vijay Kumar Saraswat, Hon. Member (S and T) NITI Aayog, Government of India; “Supercritical carbon dioxide Brayton cycle for power generation” by Prof. Pradip Dutta, IISc Bangalore and “Role of Nuclear Fusion for Environmental Sustainability of Energy in Future” by Prof. J. S. Rao, Altair Engineering. The conference included 27 technical sessions on topics related to energy and environmental sustainability including 5 plenary talks, 40 keynote talks and 18 invited talks from prominent scientists, in addition to 142 contributed talks, and 74 poster presentations by students and researchers. The technical sessions in the conference included Advances in IC Engines: SI Engines, Solar Energy: Storage, Fundamentals of Combustion, Environmental Protection and Sustainability, Environmental Biotechnology, Coal and Biomass Combustion/Gasification, Air Pollution and Control, Biomass to Fuels/Chemicals: Clean Fuels, Advances in I.C. Engines: CI Engines, Solar Energy: Performance, Biomass to Fuels/Chemicals: Production, Advances in I.C. Engines: Fuels, Energy Sustainability, Environmental Biotechnology, Atomization and Sprays, Combustion/Gas Turbines/Fluid Flow/Sprays, Biomass to Fuels/Chemicals, Advances in I.C. Engines: New Concepts, Energy Sustainability, Waste-to-Wealth, Conventional and Alternate Fuels, Solar Energy, Waste Water Remediation and Air Pollution. One of the highlights of the conference was the Rapid-Fire Poster Sessions in (i) Energy Engineering, (ii) Environment and Sustainability, and (iii) Biotechnology, where more than 75 students participated with great enthusiasm and won many prizes in a fiercely competitive environment. 200+ participants and speakers attended this 4-day conference, which also hosting Dr. Vijay Kumar Saraswat, Hon. Member (S and T) NITI Aayog, Government of India as the chief guest for the book release ceremony, where 16 ISEES books published by Springer, under a special dedicated series “Energy, environment and sustainability”, were released. This was the first time that such significant and high-quality outcome has been achieved by any society in India. The conference concluded with a panel discussion on “Challenges, Opportunities and Directions for Future Transportation Systems”, where the panelists were Prof. Gautam Kalghatgi, Saudi Aramco; Dr. Ravi Prashanth, Caterpillar Inc.; Dr. Shankar Venugopal, Mahindra and Mahindra; and Dr. Bharat Bhargava, DG, ONGC Energy Center; and Dr. Umamaheshwar, GE Transportation, Bangalore. The panel discussion was moderated by Prof. Ashok Pandey, Chairman, ISEES. This conference laid out the roadmap for technology development, opportunities and challenges in Energy, Environment and Sustainability domain. All these topics are very relevant for the country and the world in present context. We acknowledge the support received from various funding agencies and organizations for the successful conduct of the Second ISEES conference SEEC-2018, where these books germinated. We would therefore like to acknowledge SERB,

Preface

ix

Government of India (Special thanks to Dr. Rajeev Sharma, Secretary); ONGC Energy Center (Special thanks to Dr. Bharat Bhargava); TAFE (Special thanks to Sh. Ananda Patil); Caterpillar (Special thanks to Dr. Ravi Prashanth); Progress Rail, TSI, India (Special thanks to Dr. Deepak Sharma); Tesscorn, India (Special thanks to Sh. Satyanarayana); GAIL, Volvo; and our publishing partner Springer (Special thanks to Swati Meherishi). The editors would like to express their sincere gratitude to large number of authors from all over the world for submitting their high-quality work in a timely manner and revising it appropriately at a short notice. We would like to express our special thanks to Dr. Sunil Kumar Pandey, Assistant GM and Mr. A. Sundaresan, VP of Ashok Leyland, Prof. Ms. Dhende, Assistant Professor and Dr. Suhas Deshmukh, Associate Professor of Government College of Engineering, Karad, Mr. Shriram K., Lead Engineer, NVH and Mr. Sailesh H., Lead Engineer, Aerodynamics of TurboEnergy Ltd., Prof. P. Mohanan, Prof. L. M. Das, Dr. Thipse and Mr. Rairikar of Automotive Research Association of India (ARAI), and automotive consultants, Dr. A. Gopinath and Mr. Bonthala Ranga Rao who reviewed various chapters of this book and provided very valuable suggestions to the authors to improve their manuscript. The main drivers for the diesel engine design have been (a) to satisfy progressively tightening emission standards for nitric oxides, particulate matter and hydrocarbon plus (b) the push to reduce the carbon dioxide emissions. Substantial improvement in fuel consumption by optimizing combustion, reducing friction, tuning the engine to the end application and packing more power in small envelopes are the techniques. The book covers diesel engine design for heavy duty application by authors who have specialized in the design of important subsystems and have rich experience in matching them to the engine as well as the intended vehicle application, for 15 to 40 years. The introduction of fast digital computation and high-level algorithms enabled fine and live control of vehicular engines using principles already well known in electricity generating application and marine application, wherein the load and speed change is not as fast as in vehicular engines, for example, selective catalytic reduction and high injection pressure. Also, technology hitherto confined to very large engines could be scaled down successfully because of improvement in consistent manufacturing, e.g. pistons with steel top, highly loaded bearings, two stage turbocharging, high pressure injection systems and steel gaskets. In addition, breakthrough in injection systems like common rail system, opened up possibilities in combustion, not imagined some two decades back. Still, the engine technology is evolutionary and hence it is important to have sound knowledge of constructing the power train components, viz., the crankcase, connecting rod, piston, cooling system and oil pump. We thank SAE, DieselNet, VDI, AVL, Bosch, Shriram Pistons Ltd., Drawfolio, TurboEnergy Ltd., Sulzer, Ashok Leyland Ltd., ARAI, CTS Corporation, BICERA, Emitec, Ashok Leyland, Simpson and Company Ltd. and Kirloskar Oil Engines Ltd. to refer to their valuable research and data in the many chapters of this book on Design and Development of Heavy-Duty Diesel Engines. The editors invited the authors of profound experience and knowledge for sharing their expertise in design

x

Preface

and development of heavy-duty diesel engines. Thus, the book is intended as a support to engine designers by sharing with them several important procedures used in the industry. This book assembles vital information spread in several books and papers in different languages, concisely in one place in English. The references are mentioned duly in the text as well as listed at the end of the chapters. Chennai, India Kanpur, India

P. A. Lakshminarayanan Avinash Kumar Agarwal

Introduction

The book is on design of diesel engines. The engine design is motivated by advanced emission standards, and yearning for fuel economy as well as downsizing to achieve lightweight. The book develops the subject gradually by discussing various topics so that it is useful to a practising engineer and a student intending to develop a career in engine design. The book is broadly divided into five parts after introduction. The introduction chapter summarizes the book after giving some thoughts on estimating the power, maximum torque and shape of the torque curve to satisfy vehicles on- and off-road. Keywords Thermodynamics
Combustion
Gas Exchange
Modern Diesel Combustion
Supercharging
Turbocharging
Air Management
Selective Catalytic Reduction, Emissions
External Exhaust Gas Recirculation Systems
CNG Fuel Operation
Simulation
Ports
Crankcase
Connecting Rod
Critical Fasteners
Highly Loaded Bolted Joints
Crankshaft
Gaskets
Valve Train
Heavy Duty Application
Engine Retarders
Engine Gear Train
Piston And Rings
Cooling
Coolants
And Water Pump And Oil Pump
Electronic Control
Noise and Vibration
Future

Part 1: Thermodynamics, Combustion, Gas Exchange, Emissions At the heart is the modern diesel combustion which uses an optimally matched turbocharger. The most efficient way to clean up the exhaust of nitric oxides and hydrocarbons as well as some particulate matter (PM) is to use the selective catalytic reduction and use exhaust gas recirculation. To control PM in the engine, the most effective technique is by increasing the injection pressure and boosting the cylinder pressure. Chapters are dedicated to these subjects. Also, a chapter

xi

xii

Introduction

discusses converting a diesel engine to advanced spark ignited gas engine with examples. Simulation plays an important role in interpreting the experimental results as well as predicting the parameters useful to the designer.

Part 2: Design of Engine Build Separate chapters are written for the design of the five C’s, namely, the Cylinder head, Crankcase, Connecting rod, Pistons, Camshaft and Crankshaft. Critical fasteners holding the engine parts together are given under a heading. The success of the gaskets is the success of an engine. Engine valve train and gear train are treated separately. The most important attributes to the combustion phenomenon are by the fuel injection equipment. Both mechanical and electronic fuel injection equipment are discussed in detail in a chapter. Electronic injection with the help of computers enables control of emissions and reduces noise, and it is treated in a separate chapter.

Part 3: Noise and Vibration The section on noise discusses noise power, ranking of noise emitting parts at steady state, silencer and transient noise when vehicle passes by. Theory of rigid body vibration and transmission of forces to the feet are the topics related to vibration. Typical examples of a three cylinder without an internal balancer and a four-cylinder engine with internal balancer make the difficult subject interesting.

Part 4: Future Curtain is raised to have a glimpse of the future diesel engines that are in the precincts of laboratory for closing the book on the design of diesel engines. P. A. Lakshminarayanan Avinash kumar Agarwal Walter Knecht

Contents

1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P. A. Lakshminarayanan and Avinash Kumar Agarwal

Part I

1

Thermodynamics, Combustion, Gas Exchange, Emissions

2

Modern Diesel Combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Walter Knecht and P. A. Lakshminarayanan

17

3

Supercharging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Walter Knecht and P. A. Lakshminarayanan

67

4

Introduction to Turbocharging—A Perspective on Air Management System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. A. Subramani, R. Dhinagaran and V. R. Prasanth

85

5

Topics on Selective Catalyst Reduction . . . . . . . . . . . . . . . . . . . . . . 195 P. Kumar

6

Strategies to Control Emissions from Off-Road Diesel Engines . . . . 237 M. V. Ganesh Prasad

7

External Exhaust Gas Recirculation . . . . . . . . . . . . . . . . . . . . . . . . 275 P. Sahaya Surendira Babu and P. Kumar

8

Diesel Particulate Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313 K. C. Vora, Kartik E. Gurnule and S. Venkatesh

9

Conversion of Diesel Engines for CNG Fuel Operation . . . . . . . . . 341 G. Jeevan Dass and P. A. Lakshminarayanan

10 Simulation of Gas Flow Through Engine . . . . . . . . . . . . . . . . . . . . 393 Neelkanth V. Marathe, Sukrut S. Thipse, Nagesh H. Walke and Sushil S. Ramdasi

xiii

xiv

Part II

Contents

Design of Engine Build

11 Development of Ports of Four Stroke Diesel Engines . . . . . . . . . . . 407 Nagaraj S. Nayak and P. A. Lakshminarayanan 12 Design and Analysis Aspects of Medium and Heavy-Duty Engine Crankcase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427 Swapnil Thigale, M. N. Kumar, Yogesh Aghav, Nitin Gokhale and Uday Gokhale 13 Connecting Rod . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467 Prakash R. Wani 14 Critical Fasteners, Highly Loaded Bolted Joints . . . . . . . . . . . . . . . 509 Prakash R. Wani 15 Crankshaft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525 Prakash R. Wani 16 Gaskets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575 Osamu Aizawa 17 Design of Valve Train for Heavy Duty Application . . . . . . . . . . . . 601 Aniket Basu, Nitin Gokhale, Yogesh Aghav and M. N. Kumar 18 Engine Retarders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 637 M. V. Ganesh Prasad 19 Engine Gear Train Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 Vishal Bhat, M. N. Kumar, Yogesh Aghav and Nitin Gokhale 20 Piston and Rings for Diesel Engines . . . . . . . . . . . . . . . . . . . . . . . . 731 Subrata Neogy, Vikas Ramchandra Umbare, Vineet Ahluwalia and P. A. Lakshminarayanan 21 Cooling, Coolants, and Water Pump and Oil Pump . . . . . . . . . . . . 763 S. Seetharaman and P. A. Lakshminarayanan 22 Design of Electronic Control for Diesel Engines . . . . . . . . . . . . . . . 795 M. Leelakumar Part III

Noise and Vibration

23 Study of Noise and Vibration Problems Related to Heavy Duty Diesel Engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833 P. A. Lakshminarayanan Part IV

Future

24 Future Diesel Engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887 Z. Gerald Liu and Achuth Munnannur

About the Editors

P. A. Lakshminarayanan was Chief Technology Officer at Simpson and Co Ltd, Chennai, India till 1-Jan-2019, and before 2011 was the head of the Engine R&D at Ashok Leyland, India. After completing his B. Tech, MS and PhD from Indian Institute of Technology, Madras, he worked as a research associate in University of Technology, Loughborough, UK for four years. He has more than 35 years industry experience, having worked in various capacities in Kirloskar Oil Engines Limited, Ashok Leyland and Simpson and Co Ltd. Dr. Lakshminarayanan is a fellow of SAE, ISEES, and Indian National Academy of Engineering and a member of Combustion Institute. He has co-authored two books and more than 50 research papers and holds seven patents. Prof. Avinash Kumar Agarwal is a Professor in the Department of Mechanical Engineering in Indian Institute of Technology Kanpur. His areas of interest are IC engines, combustion, alternative fuels, conventional fuels, optical diagnostics, laser ignition, HCCI, emission and particulate control, and large bore engines. He has published 24 books and 230+ international journal and conference papers. Prof. Agarwal is a Fellow of SAE (2012), ASME (2013), ISEES (2015) and INAE (2015). He received several awards such as Prestigious Shanti Swarup Bhatnagar Award-2016 in Engineering

xv

xvi

About the Editors

Sciences, Rajib Goyal prize-2015, NASI-Reliance Industries Platinum Jubilee Award-2012; INAE Silver Jubilee Young Engineer Award-2012; SAE International’s Ralph R. Teetor Educational Award-2008; INSA Young Scientist Award-2007; UICT Young Scientist Award2007; INAE Young Engineer Award-2005.

Chapter 1

Introduction P. A. Lakshminarayanan and Avinash Kumar Agarwal

Abstract The book is on design of diesel engines. The engine design is motivated by advanced emission standards, and yearning for fuel economy as well as downsizing to achieve light weight. The book develops the subject gradually by discussing various topics so that it is useful to a practising engineer and a student intending to develop a career in engine design. The book is broadly divided into five parts after introduction. The introduction chapter summarises the book after giving some thoughts on estimating the power, maximum torque and shape of the torque curve to satisfy vehicles on- and off-road.

1.1

Improved Diesel, to Ban Or Not to Ban

The diesel engine replaced the steam engine as the work horse of modern civilization for the past hundred years. It has been powering almost every hard-working machinery like ships, trains, agricultural tractors, buses and for the last fifty years passenger cars. The advantage of very high fuel efficiency, high back up torque enabled very wide application. The engine could accept widely varying quality of fuels. However, the stigma attached to the engine, of nitric oxides (NOx) and high particulate matter (PM) emission by virtue of smoke and unburnt oil as well as noise, could not be easily erased from the attention of ever vigilant society. It appeared the end of the road had been announced for the diesel engine: an old-fashioned technology and the root cause of disproportionate levels of pollution in cities and countryside, giving gave credibility to the headlines and commentary in the media. A number of improvements have taken place in the last twenty-five years to improve these difficulties successfully. Most importantly, diesel offers the P. A. Lakshminarayanan (&) Formerly with Simpson and Co. Ltd., Chennai, India e-mail: [email protected] A. K. Agarwal Indian Institute of Technology, Kanpur, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_1

1

2

P. A. Lakshminarayanan and A. K. Agarwal

efficient and flexible use of fuels with high energy density, and excellent storage and distribution options. A mix of technologies is needed in the future as the role the modern diesel engine has to play and hence, innovation instead of bans is the right direction (Anton Andres, Daimler 2018). The engine manufacturers are clearly aware of the responsibility towards climate, clean air as well as the demands of individual mobility: • The advantage in respect to CO2 over comparable gasoline engines is undisputed. • The NOx emissions of many vehicles on the road reduced by up to 80%, through software updates relatively easily because of the rich experience and knowledge about the selective catalytic reduction (SCR). • The new generation engines satisfying Euro-6 standards are having a huge market traction and the emissions are already far less than mandated limit values. • In countries with strict emission control, the NOx pollution has dropped by more than 70% from 1990, even though the real driving cycle emissions (RDE) are higher the laboratory measurements, and in the next five years the drop is expected to be 90% from today. • In cities like Delhi and Stuttgart it has been found the contribution by automobiles to PM is about 5% only. PM10 and PM2.5 from engines are drastically reduced with the introduction of Euro-6 norms with a Diesel Particulate Filter (DPF). • With the new generation fuel injection systems working in conjunction with SCR and mild-hybrid options like start-stop, e-turbocharger offer substantial improvement to fuel consumption as well as emissions. To do away with the diesel at this point in time would be a big mistake, for both environmental and economic reasons (Anton Andres, Daimler 2018). It may not be out of place to remind ourselves that reduction in well-to-wheel CO2 emission is possible in case of electrical vehicles, only when the electricity is generated using non-coal-based power plants (Hofacker 2017).

1.2

Application of Diesel Engines

Diesel engines are developed in laboratories to satisfy emission standards, fuel efficiency demand and cost. Application of right diesel engines to the end application like a vehicle is important. Right choice of the size, power rating, weight and other parameters at the right cost decides the success of the engine in the market. Front loading the work to the simulation community is expected to give some relief to the designers and reduce the anxiety (Fancher 1979; Northcote 2006).

1 Introduction

1.3

3

Duty of Diesel and Its Sizing

Interestingly one size of diesel does not fit all the nations, conditions and users. Whereas 400 hp-650 hp may be common for trucks in Europe with fast and safe roads at an average speed of 100 km/h, in an emerging county like India an average 50 km/h for a truck is quite common. Similarly, off road the rating is determined the initial expenditure as well as the size of the land holding. Therefore, the rating of the engine is nearly half in India since the speed or the land area determine the power. In addition, the average duty of a truck engine itself is about 50% in developed countries with the result of large excess power available for overtaking even at 150 km/h.

1.4 1.4.1

Notes on Weight, Fuel Consumption and Optimum Power for a Given GVW Fuel Consumption

For pulling 50 tonnes at 60 km/h 152 kW (max) are needed. For pulling it at 80 km/h you need 225 kW (max), Fig. 1.1. We can estimate the fuel consumption as given in Table 1.1. Therefore, the increase in fuel consumption for achieving higher velocity alone is 12.5%. Higher speed and hence higher rating therefore do not pay in terms of fuel consumption, unless the vehicle power train is 12.5% (equivalent to an improvement in specific fuel consumption, SFC by 25 g/kW h) more efficient when scaled up. The fuel consumption of engines improves with brake mean effective pressure (bmep). However, the difference between bmeps 14 and 20 bar is about 1.5–2%. For example, the minimum SFC of 5.7 l Euro-4 is 206 g/kW h and 8 l Euro-4 is

300 250

Demand Power, kW

Fig. 1.1 Power demand from the prime mover as a function of road speed for different Gross Vehicle Weight (GVW)

200 150 100 50 0 20

30

40

50

60

70

80

90

100

veh. Speed, km/h 15 t, GVW

20 t, GVW

25 t, GVW

30 t, GVW

35 t, GVW

40 t, GVW

45 t, GVW

50 t, GVW

4 Table 1.1 kW, kmph, hours to travel 1000 km and energy required

P. A. Lakshminarayanan and A. K. Agarwal kW

km/h

Hours to travel 1000 km

Energy, kW h/ 1000 km

150 225

60 80

16.7 12.5

2500 2813

203 g/kW h rated at 204 hp at 2100 rpm (15.3 bar bmep) and 306 hp at 1750 rpm (19.7 bar bmep) respectively corresponding to the optimum piston speed of 8 m/s for fuel consumption. A plot is made of the energy consumption against rated power, kW Fig. 1.2, for different gross vehicle weight, GVW. • Fuel consumption is minimum at around 50 km/h, • Energy required to travel 1000 km at 3 different speeds namely 50, 70 and 90 kmph for different tonnages is plotted in dotted lines. • Speeds in the range of 50 and 60 kmph offer minimum fuel consumption. Figure 1.2 is plotted differently in Fig. 1.3 showing the penalty in fuel consumption if the speed is different from 50 kmph for different GVW. Thus, the optimum range of power required for minimum fuel consumption can be found, for example when the penalty is 1%, Fig. 1.4. This is shown in Table 1.2. In other words, rating high power engines at lower power can lead to improvement in fuel economy. For grading and acceleration, some reserve has to be maintained. Usually, a 20% reserve is sufficient and this pushes the maximum speed capability to 75 kmph on level road at full load. Such an approach is used in most of the successful vehicles (Fig. 1.4) in India where speeds above 80 km/h at full load are not desired. In Europe, on the other hands, customers wish 120 kmph at the maximum and deviate substantially from the power required to run at 50–60 camphor example, a 35

3500 3000

Energy, kW h

Fig. 1.2 Energy to travel 1000 km as a function of rated power for minimum time of travel

2500 2000 1500 1000 500 0

0

50

100

150

200

250

Rated power, kW GVW 15 t

GVW 20 t

GVW 25 t

GVW 30 t

GVW 40 t

GVW 50 t

50 km/h

70 km/h

90 km/h

300

1 Introduction 45%

Increase in energy to travel 1000 km w r t 50 km/h

Fig. 1.3 Penalty in fuel consumption if the speed is different from 50 kmph as a function of road speed

5

35%

50 t 40 t

25%

35 t 25 t

15%

20 t 15 t

5% -5%

0

20

40

60

80

100

speed, km/h

250

Rated Power, kW

Fig. 1.4 Relation of minimum and maximum of the range of optimum power rating against GVW (t) for a penalty of 1% with respect to the gross vehicle weight

200 150

Series5 min, optimum

100

max, optimum with 20% margin

50

successful in India

0 0

10

20

30

40

50

60

GVW, t

Table 1.2 Optimum range in power for minimum fuel consumption

Tonne

min, kW

max, kW

50 132 161 45 120 145 40 109 132 35 95 118 30 83 104 25 72 90 20 59 75 15 48 62 a margin for acceleration, gradient etc.

With 20% margina, kW 194 175 159 142 125 108 90 75

tonner would be successful in India using 120 kW (160 hp) engine for reasons of fuel economy, whereas a 16 tonner would be designed with 240 kW (340 hp) in Europe to achieve the maximum speed of 140 kmph on level ground.

6

1.5

P. A. Lakshminarayanan and A. K. Agarwal

Typical Design Parameters

Piston speed Type

Cylinders

Litre

Speed, rpm

C

4 5.3 2170 6 8 2170 A, B 4 3.8 2200 6 5.7 2200 a All are turbocharged intercooled engine

Power, hp

Bmep, bar

Mass balancer, kg

Mass of EGR circuit

306 459 153 230

23.8 23.8 16.5 16.5

13 0 9 0

13 20 0 0

The mean piston speeds of some four and six-cylinder engines given in Table 1.3, are plotted in Fig. 1.5. The economic speed for best fuel consumption for these diesel engines is 8 m/s. Taking advantage of turbocharging, it is usual to design the modern engines at as low operating speed of the piston as possible. Nine m/s is usually the mean piston speed at the rated speed. Maximum cylinder pressures Figures 1.6 and 1.7 show the cylinder pressure of different engines estimated empirically (see example calculation below) for the engines listed in Table 1.3. Table 1.3 Empirical formula The number of cylinders

z

Remark

Values

Weight power train alone per cylinder excluding gear train and flywheel, housing Peak pressure capability

A

Proportional to bore2
wall thickness i.e., A ¼ k1
b2
t Proportional to thickness2/ bore2 i.e., p ¼ k2
t2 =b2

k1 = 0.0014

Gear casing, flywheel housing, flywheel, gears Oil and water Mass balancer Gadgetry e.g., EGR Bore, mm = b qffiffiffi Thickness, mm, t ¼ b kp2

p

B

Roughly the same

C D E

According to capacity

k2 = 65 200 kg cm−2 (for grey iron) k2 = 100 000 kg cm−2 (for CG iron) 170–175 kg

Table 3 Table 3

Weight of the power train alone per cylinder, kg, A ¼ k1 b2 t Weight of engines, kg, W ¼ zA þ B þ C þ D þ E Weight of power train alone (excluding gear and flywheel sides) per maximum cylinder pressure, Fig. 1.8

1 Introduction

7

mean piston speed, m/s

11.0 10.0 9.0 8.0 7.0 6.0 5.0 A6

A4

B6

B4

C4

C6

Fig. 1.5 Mean piston speeds at rated speed for different engines listed in the table

300

max. cyl. Pressure, bare

280 260 240

15% EGR

220 bar limit

220 200 180

180 bar limit

160 140

SCR

120 100 10

12

14

16

18

20

22

24

bmep, bar (~power/swept vol./speed) B4 150 hp, B6 225 hp

B6 207 hp

A6 184 hp, A4 122 hp

A6 166 hp

B4 129 hp

Fig. 1.6 Cylinder pressures, bmep and maximum pressure capabilities for SCR engines: B4 (4-cylinder, 3.7 l), B6 (6 cylinder, 5.7 l)

Example: Calculation of maximum cylinder pressure kW

Given

hp rpm Litre Power/litre

hp = kW/0.735 Given Given

165 224 2500 5.7 28.9 (continued)

8

P. A. Lakshminarayanan and A. K. Agarwal

(continued) kW

Given

165

Bmep at rated speed Boost Motoring pressure Peak pressure SCR Euro-4 or Euro-3, bar Peak pressure 15% EGR, bar Peak pressure 30% EGR, bar

900
hp/(rpm
litre) Bmep/5.5 compression ratio1.4
boost 1.05
motoring pressure 1.15
peak pressure SCR 1.30
peak pressure 30% EGR

14.2 2.6 136.1 142.9 164.4 185.8

Weights An empirical model for calculating Weight of engines depending on bore and peak pressure capability can be arrived at as explained below. The predicted weight of vehicles in Fig. 1.8 using the formula seems to be working out within 2% accuracy for the six types of engines, Fig. 1.9. Importance of backup torque For a given vehicle, the maximum road speed is a function of only the rated power. For acceleration, it is the backup torque or the maximum torque is important. For appreciating this Fig. 1.10 is useful. Here, let Tr be the engine friction when declutched and vehicle reaction at the wheel with friction added when the vehicle is moving. Similarly, I be the engine inertia alone when declutched and the reflected vehicle inertia is added when the vehicle is moving. The maximum engine torque is shown by the black continuous line and the friction is shown by green line. When the engine is accelerated from 300

max. cyl. Pressure, bare

280 260 240

15% EGR

220 bar limit

220 200 180

180 bar limit

160 140

SCR

120 100 10

12

14

16

18

20

22

24

bmep, bar (~power/swept vol./speed) C6, 245 hp C6,327 hp

C6, 272 hp C4, 231 hp

C4, 218 hp C6, 422 hp

Fig. 1.7 Cylinder pressures, bmep and maximum pressure capabilities for EGR engines: C4 (4cylinder 5.3 l), C6 (6-cylinder 8-litre)

1 Introduction

9

Weight of power train excluding gear side and FW side / max cylinder pressure capability, kg/bar

0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30

A6

A4

B6

B4

C4

C6

Fig. 1.8 Ratio of weight of power train to maximum cylinder pressure of different vehicles

Predicted weight

1200 1000 y = 1.0221x R² = 0.9492

800 600 400 200 0 0

200

400

600

800

1000

1200

actual weight Fig. 1.9 Comparison of predicted weight and actual weight of engines

160 140

Engine torque, Te

Torque

120 100

final pt., B 80 (Te-Tr) α accleration

60

Initial pt., A

40

Resisting torque, Tr

20 0

0

500

1000

1500

2000

2500

Speed, rpm

Fig. 1.10 The process path when an engine is accelerated from speed A to B for an external resistance which increases parabolically with engine speed (obtained from vehicle speed)

10

P. A. Lakshminarayanan and A. K. Agarwal

low idle to high idle speed, for example, from point A to point B, the path taken by the engine is as shown by the dashed line. It can be seen that the excess torque given by (Te − Tr) contributes to acceleration. Applying Newton’ second law, Acceleration ¼ ðTe  Tr Þ=I time taken to accelerate to final speed; B / I=ðTe  Tr Þ This model is validated over a number of similar agricultural tractors with the bare engine rated in the range 55–80 hp, Fig. 1.11. Assuming the friction torque is a small fraction of the engine torque, the acceleration time is inversely proportional to torque at low end. In Fig. 1.12, the Fig. 1.11 is plotted differently, to show the excellent correlation. System stability The operating point of a diesel engine is determined by the solution of the curves of the engine torque versus speed and the resisting load torque versus engine speed. The operating point is stable only when the engine curve is having a negative slope and the resistance curve is positively sloped at the operating point. Only then, any

Fig. 1.11 Time to accelerate a tractor, for different settings of torque at low speed end

1 Introduction

11 12

1000/Engine Torque in Nm

Fig. 1.12 Correlation of acceleration time (Fig. 1.11) with respect to the reciprocal of maximum engine torque at low speed end

10 8 6 4 2 0

0

5

10

15

20

acceleration time on load, s

disturbance increasing the speed will result in a net braking torque to bring the engine to the solution point; similarly, any decrease in speed will be restored by a net accelerating torque. Therefore, the engine should operate as in Fig. 1.13a and not as in Fig. 1.13b. Engine governing is designed to avoid the latter (Fig. 1.13b). Along the maximum torque line, it may not be always avoidable especially where the manifold pressure compensator cuts in. Manifold pressure compensation A single stage turbocharger is not able to pump enough air in the low speed range into the engine. This becomes critical when the rated bmep of the engine is very

(a)

(b) Road load (resistance)

Road load (resistance)

Torque

Torque

Prime mover (Engine)

Engine speed

Fig. 1.13 a Stable operation, b unstable operation

Engine speed

Prime mover (Engine)

12

P. A. Lakshminarayanan and A. K. Agarwal

high. Therefore, a manifold pressure compensation is integrated with the fuel injection system to reduce fuelling at low speeds to the level of a naturally aspirated engine. By this, in a small speed range, the torque drops precipitously as the speed decreases. The dropped torque is nearly equal to the maximum torque a naturally aspirated engine would produce. Now, when the vehicle is accelerated under load following a typical saw-tooth shaped curve of engine speed with respect to time; during the gear change after declutching, the engine speed drops naturally due to internal friction. The manifold pressure compensator would reduce the fuelling and hence the torque if the speed falls below the critical speed and the engine would hesitate to accelerate. Solutions are arrived at by using e-turbocharger or R2S turbocharger to avoid the severe drop in torque in the range of speeds envisaged during gear change.

1.6

Secrets of Fuel Economy

Even with the best designed and developed engine the important parameter of fuel economy may elude the vehicle customer for many reasons. Tyres have the highest influence on in fuel economy below 80 km/h, whereas aerodynamics is the most important factor above 80 km/I. The “rock solid” rules are listed in the table below (Guide, Cummins MPG 2003; van Dam et al. 2009): Cause

% change in fuel consumption

2% reduction in aerodynamic drag Each 2 km/h increase, above 90 km/h Worn tyres Ribbed tyres Efficient driver Avoiding idle time of one hour in long haul Running in *50000 km Well formulated oil

−1 +2 −7 −3 −20 −1 – −2

During service, the cause of excessive fuel consumption could be due to a number of factors (Guide, Cummins MPG 2003; Sturm and Hausberger 2005): • • • • • •

Engine factors Vehicle factors and specifications Environmental factors Driver technique and operating practices Fuel system factors Low power or driveability problems

1 Introduction

13

References Anton Andres, Daimler (2018) Improving the diesel is better than banning it. https://www. autoindustriya.com/auto-industry-news/daimler-improving-diesels-is-better-than-banning-it. html Hofacker A (2017) The future of the combustion engine let’s get down to business. ATZ Worldw 119(9):8–13 Fancher PS (1979) Simulation of the directional response characteristics of tractor-semitrailer vehicles. Final report Northcote NM (2006) The modelling and control of an automotive drivetrain. PhD dissertation, University of Stellenbosch, Stellenbosch Guide, Cummins MPG (2003) Secrets of better fuel economy: the physics of MPG. Cummins Engine Company, Columbus, Indiana van Dam W, Kleijwegt P, Torreman M, Parsons G (2009) The lubricant contribution to improved fuel economy in heavy duty diesel engines. No. 2009-01-2856. SAE Technical paper Sturm PJ, Hausberger S (2005) Energy and fuel consumption from heavy duty vehicles. COST

Part I

Thermodynamics, Combustion, Gas Exchange, Emissions

Chapter 2

Modern Diesel Combustion Walter Knecht and P. A. Lakshminarayanan

Abstract The chapter begins with the historical development of diesel combustion and explains engine Combustion Mechanisms followed by Fuel Injection and Supercharging. It explains the important step in the evolution of engine design to reduce NOx by Exhaust Gas Recirculation. Apart from the popular systems there are also Alternative Diesel Combustion Systems to remember. The evolution of the diesel engine is undoubtedly driven by the legislative standards for the Emissions of Internal Combustion Engines. The real concern regarding Global Climate Change imposes the limit to carbon dioxide and hence indirectly the fuel consumption. The particulate affective animal breathing systems as well as global warming is controlled tightly. This calls for accurate measurement of Particle Number. Apart from gaseous emissions, there is also a cap on the noise emissions. Al these requirements are satisfied to a large extent by the use of sophisticated air flow and fuel injection inside the engine. However, the advanced emission norms are satisfied only if there is Exhaust Aftertreatment to abate Nitric oxides and particulates. Hard working large engines have heat recuperation systems to consume the exhaust and coolant heat usefully. For carrying out this work, the losses have to be estimated correctly. To save cost and noise, many times the diesel engines are converted to operation on neat gas. For development of a country’s infrastructure the diesel engine is being further developed to compete with other power sources as the engine is advantageous regarding logistics, storage, efficiency and compactness.

W. Knecht Knecht Engine Consulting GmbH, Kesswil, Switzerland P. A. Lakshminarayanan (&) Formerly with Simpson and Co. Ltd., Chennai, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_2

17

18

2.1

W. Knecht and P. A. Lakshminarayanan

Historical Development of Diesel Combustion

The Diesel Engine was invented by the German Engineer Rudolf Diesel (1858– 1913) with a patent application on February 28, 1892 which led to the German patent number 62207. The diesel engine (compression ignition engine) is an internal combustion engine where the heat of compression is used to initiate the ignition and burn the fuel which has been injected into the combustion chamber. This is in considerable contrast to the spark-ignition engine which employs a spark plug to ignite the air-fuel mixture. The thermal efficiency of the diesel engine is very high due to the use of a high compression ratio, Fig. 2.1.

2.1.1

Piston Lower Side Used for Compression of Charge Air

In the course of the further engine development a variety of split combustion chambers have been investigated until in the 1930s the direct injection engine has been adopted Figs. 2.2 and 2.3 (Knecht 1993).

Fig. 2.1 Diesels first engine, 1896–97, Bore = 250 mm, Stroke = 400 mm

2 Modern Diesel Combustion

19

(a) Two Prechamber engines of early days and Saurer engine, harbinger of later day direct injection engines

Swirl Chamber engine

Pre chamber engine

(b) Popular Indirect Injection Engines

(c) Direct Injection engine

Fig. 2.2 Diesel combustion systems

2.2

Engine Combustion Mechanisms

Combustion is defined as the burning of a fuel and oxidant to produce heat and/or work. It is a chemically oxidized exothermic reaction process with a rapid oxidation generating heat. The initial slow oxidation accompanied by relatively little heat is

20

W. Knecht and P. A. Lakshminarayanan

Fig. 2.3 High speed Saurer PD diesel engine with direct injection, 1936. Bore = 80 mm, stroke = 120 mm, cylinder capacity = 3,617 dm3. 54 kW at 3000 rpm (Heldt 1936)

termed as ignition. Combustion occurs either in flame or non-flame mode whereby the flames are categorized as premixed flames or diffusion (non-premixed) flames. The requirements of combustion are two fluids (fuel and oxidizer) and both fluids must be in a gaseous phase and mixed within the flammability limits. Ignition occurs at a temperature equal or greater than the ignition temperature. It is the major energy release mechanism on the Earth and key to mankind’s existence. Combustion comprises thermal, hydrodynamic and chemical processes. It begins with the mixing of fuel and oxygen in air. The fuel may be gaseous, liquid or solid and the mixture is ignited by means of a heat source. When ignited, fuel and oxidant chemically react and the consequent heat release is a process that sustains itself. The combustion produces heat, light, chemical species, pollutants, mechanical work as well as plasma. Sometimes, a low-grade fuel, e.g., coal, biomass or coke can be burned partially (gasification) for producing higher-grade fuel, e.g., methane. Furnaces, combustors, boilers, reactors and engines are designed to utilize combustion heat, chemical species and work. In modern diesel engines efforts are undertaken to optimise the compression ratio, the piston bowl geometry, the port induced intake air motion, the fuel injector nozzle configuration (number of orifices, size of orifices, the injection angle relative to the piston bowl, the nozzle protrusion), the number and quantity of fuel injection events and especially the supercharging system, in order to realise the engine-out emissions and the power.

2 Modern Diesel Combustion

2.3

21

Fuel Injection

An important area of diesel combustion represents the fuel injection. Initially this was mainly accompanied by air assisted injection of the fuel, Fig. 2.4. Later followed the so-called ‘solid injection’ or ‘air-less injection’. As mentioned, considerable efforts have been directed towards improved Diesel fuel injection systems with higher fuel injection pressures, permitting to influence the rate and time of injection and with regard to injector-to-injector and shot-to-shot accuracy of the injection system. The mainly used injection systems are: • • • • •

in-line pump Distributor pump Unit pump Unit injectors Common rail injection.

The duties of the Diesel fuel system are the injection of fuel into the combustion chamber with an optimal rate of injection (dQ/dt), a good fuel atomisation and ensuring optimal injection timing. A flexible control of these parameters in the whole operating area by the control system is very essential. Fig. 2.4 Injection valve with pressure air assistance in diesel engine 1895

22

W. Knecht and P. A. Lakshminarayanan

The general requirements on fuel injection systems of heavy-duty diesel engines in view of low fuel consumption and low emissions are: • Accurate fuel metering injector-by-injector and shot-by-shot; smaller production tolerances. • Increased injection pressures also in the engine’s low speed range. • Flexible injection rate control in the whole engine operating range including the adoption of multiple injection events. • Reduction in nozzle hole diameters e.g. from approximately 0.18–0.10 mm • Increased number of nozzle holes in conjunction with a low port-induced in-cylinder air motion • Flexible control system • Diagnostic system. In view of EURO VI legislation for heavy duty road vehicles, manufacturers push to even higher injection pressures up to 3000–4000 bar (Delphi, Scania). The major limitations on injection pressure are related to the strength of the nozzle tip (the high pressure in the nozzle tip can cause fatigue). The metal between the holes may suffer crack propagation from crack initiation sites within the holes. Another aspect that could lead to failures is the strength of the injection pipes, the cam stresses and the ability to accept the pump drive torque. Currently all modern diesel engines which need to comply with stringent emission legislation are equipped with the common rail injection.

2.3.1

Unit Injector

In this system the injection pump and the fuel injector form one unit. The unit injector is driven by the engine’s camshaft either directly via a push rod or indirectly via a rocker-arm assembly. Due to the fact that no high-pressure tubes are required, higher injection pressures can be realised, Fig. 2.5. Cam-operated unit injectors do not allow an engine speed-independent choice of the injection pressure: the maximum pressure is available at rated speed, but the pressure is reduced at low speeds due to the design of the cam. Novel systems with forced control of the needle permit a flexible adjustment of the injection pressure at needle opening. Two important limitations of electronically controlled unit injectors are their inability to adequately decouple injection pressure from engine speed and their inability to allow sufficient multiple injection events.

2 Modern Diesel Combustion

23

Name Plate

Rocker Foot

Electrical Connector Roll Pin

Slide Plate Plunger Retention Clip

Connector O-Ring Insulators

Follower

Stator Pin O-Ring

Follower Retainer Plunger

Nozzle Nut O-Rings

Follower Spring Body Body O-Ring DMV20

Needle Spring Seat Stop Plate

Nozzle Nut Spring Cage Nozzle Spring Shim Nozzle Spring Nozzle Needle Nozzle Body

Fig. 2.5 Electronic unit injector (Bosch)

2.3.2

Common Rail Injection System

The common rail injection system was in principle invented in the early years of the 20th century (Knecht 2004). This system permits a choice of injection pressure independently of the engine speed. The injection begin and end can be relative freely chosen. Injection pressure is generated by a high-pressure pump and is available in a rail for all cylinders. Electrically operated injectors are used in place of pressure-controlled fuel injectors. In a common rail injection system, the in-line pump and the distributor pumps are being replaced by more flexible fuel injection systems which offer a higher-pressure capability. Mainly used in HD diesel engines were unit injectors and since 1999 more and more common rail systems are employed. The common rail system is widely used in passenger car diesel engines, in light duty vehicles (LDV) and now also in Heavy Duty engines. Particularly this injection system permits multiple injections, Fig. 2.6 (Imarisio and Rossi Sebastiano 2000; Egger et al. 2000; Krieger 2000; Dingle and Lai 2005): • Very advanced pilot injection in order to achieve a partially homogeneous combustion or to improve cold operation • A pilot injection with a transition to a boot injection for combustion noise- and smoke-reduction • A main injection or a split main injection for stable rich operation

24

W. Knecht and P. A. Lakshminarayanan

Fig. 2.6 Multiple fuel injection

• A post injection which leads to additional kinetic energy for a particulate matter (PM)-reduction • Late post injection as a help for particulate trap regeneration or to increase NOx-conversion in SCR-systems or in NOx absorber catalysts for hydrocarbon addition. In common rail injectors the conventional magnetic valves are being replaced by piezo-electric actuators which are faster than current solenoids and are also lighter, Fig. 2.7. The amplifier piston common rail system is suited for high injection pressure despite a relative low rail pressure and permits flexible rate shaping, Fig. 2.8. Ultimately a closed loop Lambda-control can help to reduce emissions. Regarding fuel injection equipment, the following areas are currently in development: • injection rate shaping aiming to choose within the whole operating area of an engine from a square to a ramp and to a boot injection • multiple injection • Increase of fuel injection pressures which helps to improve the mixture formation due to the higher turbulent kinetic energy and smaller fuel droplets and, hence, higher droplet surface which leads to an improved combustion process in the low-medium engine speed range. Around the rated power, however, an increased injection pressure may be needed to ensure acceptable injection duration. • Currently advanced combustion-feedback based control strategies are considered a means to provide low fuel consumption and reduced emissions. Such a control can be realised with permanently in each individual cylinder installed pressure sensors.

2 Modern Diesel Combustion

25

Fig. 2.7 Bosch piezo injector

Fig. 2.8 Amplifier piston common rail system (APCRS), Bosch

• The envisaged advantages can be: reduction of dispersion in emission cycles, improved emission stability over lifetime, compensating for tolerances, enable highest power density, improved diagnosis, more stable diesel particulate filter regeneration, improved cold startability, cylinder selective control, enabling alternative combustion concepts and others.

26

2.3.3

W. Knecht and P. A. Lakshminarayanan

Digital Valve Technology

Sturman Industries has (Steward et al. 2008) developed an interesting Digital Valve technology platform which has been applied in diesel fuel injectors in form of a hydraulically intensified fuel injector capable of multiple injection events as well as in fully flexible valve trains for cam-less engines. The key element of the Sturman technology is the Digital Valve. By energising one coil moves the armature in the valve body. When the armature reaches the end of the stroke and the current to the coil is stopped, the residual magnetism holds the armature against the valve housing, securing the valve either open or closed without any need for an external continuous power. Hence, a Digital Valve does not require a spring or any energy to operate. The special electromagnetic system takes only a transitory signal to open and another to close it. Sturman successfully employed Digital Valves in hydraulically intensified fuel injectors with intelligent controls not requiring a high-pressure pump while still offering very high-pressure capabilities. Furthermore, a control of a hydraulic fluid with a Digital Valve was employed in a cam-less engine to actuate the intake and exhaust valves in a fully flexible manner. Sturman has further developed an interesting combustion system whereby air is injected into the combustion chamber before, during and after the ignition. All the air used during the combustion enters the cylinder through the air injector. The start of combustion is triggered by the sudden presence of oxygen. A certain disadvantage may be the need of high-pressure air before the air injector, Fig. 2.9.

2.3.4

Injector Nozzle

Of great importance are the nozzle protrusion and the injector nozzle characteristics. Especially the inflow conditions into the nozzle holes are very essential. With an experimental nozzle with variable nozzle geometry significant improvements of the combustion process and the emission formation can be achieved, Fig. 2.10. An important aspect in the development of new combustion systems is the design of the injection system and especially the shape and geometry of the nozzle. This is of importance with regard to the microstructure of spray holes, spray characteristics and atomisation. Advanced technologies especially for manufacturing nozzle spray holes are employed such as electro discharge machining and hydro-grinding or laser machining. The flow through the spray holes depends on a few important characteristics such as: • the ratio of spray hole length/diameter; the rate of taper and spray inlet rounding geometry (macro geometry) • boundary layer/surface condition (micro geometry) • flow seeping/detachment

2 Modern Diesel Combustion

27

Fig. 2.9 Sturman digital valve, diesel fuel injector with pressure intensifier and cam-less valve actuation with digital valve

• • • • •

Reynolds number These characteristics determine the main spray patterns as: cone angle penetration and momentum droplet size and distribution.

28

W. Knecht and P. A. Lakshminarayanan

Fig. 2.10 Vario nozzle and possible injection configurations (Bosch)

2.4

Supercharging

Today in most industrialised countries mainly road vehicles with turbocharged engines are on the road. Turbocharging of diesel engines affects important parameters which are relevant for power, torque characteristic, pressures and temperatures prior and after the manifolds and hence the peak gas temperature in the cylinders. This influences also the fuel consumption and emissions. The characteristics of the turbocharging system affect also the gas exchange work and the

2 Modern Diesel Combustion

29

dynamic behaviour of the engine. Furthermore, if turbocharging is employed for high power the injection system has to be matched for the higher fuel mass injection and care should be taken for elevated thermal and mechanical loading of the relevant engine components and the dynamic built-up of boost pressure. For more information on supercharging see Chap. 4.

2.5

Exhaust Gas Recirculation

Exhaust gas recirculation is for many years known to be very effective for a reduction of nitrogen oxides (NOx) formation. This is due to the replacement of air by inert combustion products and in addition exhaust gas possesses a higher specific heat than air and both leads to reduce in-cylinder oxygen concentration and reduced temperatures of the cylinder contents and, hence, to a reduced NOx formation. Generally, with increased EGR-rates the NOx-emissions are reduced, however, the use of EGR leads to an increase in emissions of particulate matter since EGR results usually in a reduction of the air/fuel-ratio. A full compensation by means of turbocharging is in most cases not possible. It must be differentiated between ‘internal’ and ‘external’ EGR. While in the first case hot combustion products are retained by interventions on the valve. External EGR can be done over a ‘high pressure loop’ (HPL) or in form of a ‘low pressure loop’ (LPL), Fig. 2.11. Control of the EGR-rates by the pressure difference between exhaust p3 and intake p2 (HPL) is sometimes difficult and another possibility represents the use of an EGR-pump with a flexible speed control. Measures such as exhaust throttling or the use of a venturi tube on the inlet side are needed to create an additional pressure difference. All these measures lead, however, to a deterioration of the gas exchange and an increase in fuel consumption. Using a reed valve in the EGR pipe and utilising the pressure pulses in the exhaust gas appears to be a more appropriate approach. A positive pressure difference p3 – p2 is achievable by the VGT-control. Another approach with the so-called Pulse EGR system, the exhaust valve is slightly opened during the intake stroke and leads to a flow of exhaust gas into the cylinder. With this system no throttle valve is needed in the exhaust stream. An alternative EGR-system—active High-EGR concept—has been presented by IAV (Rohrssen and Höffeler 2011). To increase the circulation a Lysholm twin-screw compressor with a rotary-slide control is placed between the exhaust gas turbocharger and the engine intake. The compressor draws cooled exhaust gas from beyond the oxidation catalyst and mixes it in a controlled manner with fresh charge air. The system employs an engine-driven and electrically assisted compressor to keep the EGR-ratio at an optimal level. The EGR-ratio is controlled actively by two rotary valves.

30

W. Knecht and P. A. Lakshminarayanan

Fig. 2.11 Exhaust gas recirculation systems (EGR)

The EGR-rate is defined as: xEGR ¼ where m_ EGR ¼ EGR flow rate and m_ A ¼ Air flow rate

m_ EGR m_ EGR þ m_ A

2 Modern Diesel Combustion

31

Fig. 2.12 Cooling system with air cooled low temperature EGR cooling (Flik 2009)

Cooled EGR has a higher effectiveness in NOx-reduction than uncooled EGR. Several basic EGR systems can be considered for heavy duty diesel engines. There are several concerns against the application of EGR in heavy diesel engines such as: • Increased peak cylinder pressure if the air/fuel-ratio should be maintained. • new turbocharger match for an increased compressor pressure ratio • Cooled EGR results in a substantial increase in heat in the coolant and necessitates possibly a modification of the vehicle cooling system. • elevated engine wear. With cooled EGR, the heat rejection to the coolant increases and in most cases a change of the cooling systems is not avoidable. In Fig. 2.12 a system with a two stage EGR-cooling is presented. In the first stage the cooling is achieved via the main cooling circuit (t approx. 150 °C). In the second stage through direct air cooling, the EGR is cooled to a temperature similar to that of the charge air (Flik et al. 2009). Indirect charge air cooling is a step towards intake air temperature control. The coolant is cooled by a low temperature coolant to air radiator, which replaces the direct charge air cooler in the vehicle cooling module, Fig. 2.13.

2.6

Alternative Diesel Combustion Systems

In recent years efforts have been directed towards advanced novel Diesel combustion systems using homogenisation effects. Thereby the fuel droplets in the cylinder charge should be evenly distributed and autoignition is to occur at many places around fuel droplets while a high local air/fuel ratio is essential to avoiding soot formation and leading to a lower local temperature and herewith reducing the formation of nitrogen oxides, Fig. 2.14.

32

W. Knecht and P. A. Lakshminarayanan

Fig. 2.13 Cooling system with indirect charge air cooling (Flik et al. 2009)

A number of combustion strategies have been explored in the past years. These approaches rely on the use of cooled EGR to reduce NOx- and PM-formation by controlling the combustion temperature, Fig. 2.15. Some of these combustion modes are referred to as low temperature combustion (LTC) and include premixed charge compression ignition (PCCI), smokeless rich combustion and homogeneous charge compression ignition (HCCI). Two primary factors that influence the combustion efficiency are peak combustion temperature and the mixing characteristics of fuel and charge (fresh air and EGR). It is differentiated between two types of PCCI combustion: early and late PCCI. A combination of early (BMEP 5–10 bar) and late PCCI (beyond BMEP 10 bar) has been used to extent LTC over the complete operating range of a HD diesel engine. A better solution appears the use of early PCCI and advanced diffusion-controlled combustion. Diffusion controlled combustion occurs when the fuel burning rate is controlled by the rate of fuel injection and subsequent mixing with the charge. Engine efficiency and NOx-emissions of different combustion strategies are given in Figs. 2.15 and 2.16. The various considered combustion systems are: • • • • •

HCCI Homogeneous Charge Compression Ignition HCLI Homogeneous Charge Late Injection LTC Low Temperature Combustion HPLI Highly Premixed Late Injection PCCI Premixed Charge Compression Ignition (early, late).

DCCS Dilution Controlled Combustion System The improvement and flexibility in operation of the above alternative diesel combustion systems was only partially favourable and difficulties were noted in an operation at high loads which is important for heavy duty vehicles. To date no novel combustion system has led to advantages which would justify a large volume application.

2 Modern Diesel Combustion

33

(a) Heterogenous Combustion

(b) Homogeneous combustion Fig. 2.14 Comparison of heterogeneous and homogeneous combustion

2.7

Emissions of Internal Combustion Engines

Air pollution existed in prehistory and originated from volcanic eruptions, wildfires and dust-storms with associated sulphur dioxide, carbon monoxide and particulate emissions. In the middle Ages, coal burning was outlawed in London while Parliament was in session. However, the phenomenon of air pollution accelerated dramatically due to the increase in emissions since the Industrial Revolution. Air

34

W. Knecht and P. A. Lakshminarayanan

Engine Brake Thermal Efficiency

60% 50%

Early PCCI

Advanced Diffusion control

40% 30%

Smokeless Rich

Conven onal Diesel

20% 10%

Late PCCI 0% 0.1

1

EPA 2010

10

EPA 2004

EPA 1998

Engine Out NOx

Fig. 2.15 Fuel efficiency and NOx emission capability of different combustion strategies (Stanton 2013) Fig. 2.16 Diesel combustion systems for low NOx/low soot formation (Moser AVL)

pollutants are dispersed in the atmosphere through convective and turbulent movement. In addition, some pollutants undergo chemical reaction with others particularly in the presence of sunlight, where photochemical oxidants are formed by combination of certain hydrocarbon molecules with oxides of nitrogen. Air pollution emissions are mainly man-made and can be divided into several categories: • Automotive machines like, cars, trucks, aircrafts, trains, ships etc. which mostly burn mineral fuels • Agricultural sources like livestock, rice farming, etc., which produce methane (from) and particles from cultivation. • Industrial sources generating chemical waste gases as by-products • Construction activity that produce particles and fossil fuel combustion products • Decaying organic matter in landfills produces methane.

2 Modern Diesel Combustion

2.8

35

Global Climate Change

The potential effect of greenhouse gas (GHG) emissions on the global climate change is much discussed. Greenhouse gases effectively absorb thermal infrared radiation emitted by the Earth’s surface, by the atmosphere and by clouds. Therefore, greenhouse gases trap heat within the surface-troposphere system and this is called Greenhouse Effect. There are several gases which contribute to the so-called ‘greenhouse effect’, namely: Carbon dioxide CO2, methane CH4, nitrous oxide N2O, groups of partially halogenated organic species such as chlorofluorocarbons (CFC), hydrochlorofluorocarbons (HCFC), hydrofluorocarbons (HFC) and several others. The predominant greenhouse gas for the retention of heat on the Earth’s surface is the atmospheric water vapour. Carbon dioxide (CO2) is the principal anthropogenic greenhouse gas that affects the Earth’s radiation balance. According to the IEA (International Energy Agency) the global CO2 emissions was in 2010 approximately 30.6 gigatonnes. The specific CO2-emissions vary considerably: OECD-states 10 tonnes/person; China 5.8 and India 1.5 tonnes per person. In order to achieve the target of not exceeding an increase in surface temperature by 2 °C, in the year 2020 not more than 32 gigatonnes CO2 should be emitted into the atmosphere. Methane is a potent GHG. The global estimates of methane emissions are not very reliable. According to the FAO (UN Food and Agriculture Organisation) it was stipulated that globally animal breeding contributes 18% of by humans contributed GHG-emissions. Cattle breeding: A grown-up cow releases per day 0.4 kg methane and 8 kg CO2. It has been found that with fewer animals but increased milk production (milk production has increased from 5000 in 1978 to 7300 kg milk per cow in the year 2005) and, hence, the specific methane emission per kg milk was reduced by 35%. Tropic forests emit according to new studies significant quantities of methane (Frankenberg et al. 2005; Keppler et al. 2006; Ferretti et al. 2005; Lowe 2006). It was estimated that 10–30% of the annual methane emissions originate from plants. Nitrous oxide N2O is besides CO2 and methane a very important GHG and a significant source represents the ground in forests (Ambus et al. 2006). Aerosols show an important effect on the radiation balance of the earth. Very small suspended particles (solid or liquid) can reflect the sun light and as condensation kernels contribute to the cloud formation and hence, their effect is in tendency ‘to reduce’ the global surface temperature. Global warming and increased GHG-concentrations due to human activities are generally not disputed. The global surface temperature has increased. The concentrations of GHG in the troposphere (from the Earth’s surface to approximately 10 km altitude) have considerably increased.

36

2.9

W. Knecht and P. A. Lakshminarayanan

Emission Legislation

A detailed description of Legislation was not included in this book since The Worldwide Emission Standards are given and updated in http://delphi.com/pdf/ emissions/Delphi-Passenger-Car-Light-Duty-Truck-Emissions-Brochure-20132014.pdf, http://delphi.com/pdf/emissions/Delphi-Heavy-Duty-Emissions-Brochure -2013-2014.pdf. The continuous tightening of the emission standards and test procedures were a key driver of Diesel Combustion development (Emission, carbon dioxide, noise). The higher the emission reduction requirements became, the more it was realised that the targets could not solely be achieved by combustion optimisation but that exhaust gas aftertreatment was needed also for diesel engines. The achievement in terms of emissions which have been realised since the 1970s is considerable, Fig. 2.17, and have been realised without a significant change in fuel consumption. For each step, in emission standards, appropriate strategies were considered, Fig. 2.18. With the enforcement of the EURO 3 emission standards, all European diesel engines were equipped with fuel injection systems with Electronic Diesel Control (EDC). However, mechanical governors are still widely used in developing countries where the market is not yet ready to accept EDC.

120

NOx PM

100

-21%

Verbrauch Lkw

Prozent

80 60

-98%

40 20 0 Euro 0 1982

Euro I 1992

Euro II 1996

Euro III 2000

Fig. 2.17 Emission reduction as function of time

Euro IV 2005

Euro V 2008

Euro VI 2013

2 Modern Diesel Combustion

37

Technical solution for Stage IV 1 stage TC

Fuel consumption

2 stage TC SCR

EGR 40

Euro III

Combustion

0

0

0

0

SCR

EGR

DPF Euro V

0,01 Euro VI

50%

PM (g/kWh)

EGR Rate

SCR efficiency

AdBlue consumption

10%

add. Cooling power

100

2.0

0.46

3.5

5.0

NOx (g/kWh)

FPT Industrial

19

Technical solution for Euro VI 1 stage TC

Fuel consumption

2 stage TC

0

EGR 10%

Euro III

Combustion

EGR Rate

40

AdBlue consumption

SCR efficiency

100

0

0

DPF SCR 0,01 Euro VI

Euro V 0.46

2.0

3.5 FPT Industrial

Fig. 2.18 Emission reduction strategies

5.0

NOx (g/kWh)

20

50%

add. Cooling power

SCR

0

38

2.10

W. Knecht and P. A. Lakshminarayanan

Measurement of Particle Number

An important aspect is related to a Particle Number Counter. This device measures the non-volatile particle number concentration. In order to do this, an appropriate measurement system as described in ECE R49, Amendment Nr. 6 is required. The system consists of a VPR (volatile particle remover), the VPR (particle number counter). The VPR consists of a first particle diluter 1, an evaporation tube (ET) and of a second evaporation tube (particle number diluter 2). The PNC is an n-butanol based condensation core particle counter (Giechaskiel et al. 2012), Figs. 2.19 and 2.20.

2.11

Noise Emissions

In the second half of the 20th century the disturbances due to noise emissions of any kind (rail, road, aircrafts, building machinery, etc.) became not just a nuisance but also affect the health of human beings. Noise is recognised as a major environmental problem and is the cause of increasing public concern. As a consequence, the legislative regulations regarding noise produced by vehicles and all types of mobile and stationary machinery are becoming increasingly stringent. First vehicle noise standards were issued in the European Union in 1981 (EC Directive 81/334). Over the past 30 years noise emissions from commercial vehicles have been steadily reduced. In 1980 the regulated drive by noise for trucks was 90 dB (A), today the level is 78 dB (A). This reduction of 12 dB in noise corresponds to 16 times less noise. That is 16 trucks of today together emit the same noise as 1 model year 1980 truck. For road transport the vision for the year 2020 is to halve the perceived level of noise. The most important areas where improved solutions and approaches are needed are: • Reduction of rolling noise by low noise tyres and quiet maintainable road surfaces. This is essential since rolling noise of vehicles has become a dominant noise element in many traffic situations at medium to higher vehicle velocities. • Reduction of propulsion noise (engine, transmission, intake and exhaust noise) which is also a significant element in vehicle accelerations. • Improved traffic management in order to allow even traffic flux and for congestion avoidance. Disturbing noises from powertrains are often equated with insufficient product quality. Therefore, especially for engines it is not sufficient to reduce the sound pressure level for achieving the requested sound quality. The engine noise reduction can be approached in three ways: • reduction of the noise excitation, for example by optimising the cylinder pressure development

2 Modern Diesel Combustion

Partial Flow

Full Flow Fig. 2.19 Particle sampling systems for particle concentration measurement

39

40

W. Knecht and P. A. Lakshminarayanan

Fig. 2.20 Particle number measurement (Giechaskiel et al. 2012): It consists of a volatile particle remover (VPR) and a particle number counter (PNC); The VPR removes volatile particles and dilutes the sample. The PNC measures the number concentration of particles >23 nm

• reduction of solid body noise transmission, for example by stiffening components • reduction of the sound radiation, for example by partial encapsulation. The sources of noise from a road vehicle can be separated into four main areas: the engine, the exhaust, the air intake and the noise produced as a result of the tyre rolling in contact with the road. The noise that is radiated from the tyre surface is produced by several mechanisms: • vibrations of the tyre surface • vibrations of the thread blocks • resonances of the air cavities in the contact patch between the tyre and the road surface. The strategy needed to achieve significant reductions in engine noise is different between the classes of engine. The basic rules are, however, common. A holistic approach is necessary and the ability to follow the engine development from the concept phase through to the start of production is imperative. Perhaps the most important point is the ability to influence the design of a new engine at the concept phase of the design. The major parts of the engine structure, the block, head and crank train are the first parts of the design that start to solidify

2 Modern Diesel Combustion

41

during the concept design. These components are traditionally some of the loudest parts of the engine and low noise concepts must be integrated into these major components. Development of analytical tools to calculated noise radiation has been an important part of this process. Another more efficient method to find noise sources is by microphones on moved by a robot across the engine and now by means of an acoustic camera, Fig. 2.21. With this equipment the noise problems can be more precise and differentiated be shown. A special arrangement of microphones (array) is coupled with a digital camera. Based on the measured audio-visual data the computer calculates a detailed noise map. The acoustic camera permits a visualisation of noise. The measured data can also be used to calculate the acoustic power of the whole engine; this is an integration of measured sound intensity vectors over the surface and is given in Watt. Subdividing the measuring surface into panels and calculating the acoustic power per panel can be a useful way to identify and rank the importance of particular areas or components on the engine. Analysis of combustion noise is an important part of noise development (Scarth and Ortiz 2004). Combustion noise analysis is a way of quantifying the harshness of the combustion process and to make a separation between mechanical and combustion noise. This is done by making a frequency analysis of the pressure measured in the cylinder and passing this through a standard structural attenuation filter. The difference between this value and the value measured at the microphone is the mechanical noise of the engine at this operation point, Fig. 2.22. Combustion noise analysis has become more important with the introduction of the common rail fuel injection system. With electronic control of every point in the load/speed map there is a great potential for noise optimisation, Fig. 2.23. Recent work has shown the importance of rail pressure as well as start of injection in controlling combustion noise and the uses of pilot injection has had a great effect at lower engine speeds and low idle. Noise quality is becoming an important aspect of noise development for commercial vehicles. This is where the type of noise is important and not simply the overall level. Values such as roughness, sharpness and tonality are calculated and compared between engines. Thus, to understand and quantify exactly what people find unpleasant in a noise and try to rectify it. To achieve the very low noise targets expected of state-of-the-art engine designs there are no simple solutions according to a general recipe possible. The importance of a sound base design cannot be overstated, but the more significant noise reductions are obtained by gathering many small effects. Modifications to covers and supports, isolation of covers and oil sumps, changes to combustion parameters, and reduction of tolerances are some of the areas, which need to be optimised. Here the analysis tools are used again to propose design modifications to solve noise problems. Working in parallel between testing and analysis, design solutions can be tested before tooling changes are made. Using this approach in recent engine developments, reductions of 6 dB(A) have been obtained between current production and the new engines starting production now.

42

W. Knecht and P. A. Lakshminarayanan

Fig. 2.21 Anechoic test bed with microphone robot (left) and Acoustic camera setup in anechoic test room (right)

2 Modern Diesel Combustion

43

Fig. 2.22 Separation of mechanical and combustion noise

2.12

Exhaust Aftertreatment

2.12.1 Particulate Matter Reduction Particulate filter developments started in the 1980s. At that time busses were equipped with burner regenerated particulate traps. This was no success, because of durability problems and the reliability of sensors of the control system (Meinrad

44

W. Knecht and P. A. Lakshminarayanan

Fig. 2.23 Combustion noise meter as a standard equipment for development test beds

and Giorgio 1989). Also, a large-scale fleet test in Germany in which many vehicle manufacturers participated did not lead to a breakthrough in PM-trap usage (Becker and Heine 1999). Several filter media can be used with high filtration efficiencies and different types of particulate filters have been evaluated. Basically, it can be differentiated between deep bed filters and wall flow filters. Different filter types are: • • • • •

Wall flow filters (Cordierite, silicon carbide) Sinter metal filters Knitted ceramic or metallic fibres Ceramic or metallic foams Bypass filter. Critical factors of particulates traps that need optimisation are:

• • • • • • •

soot filtration efficiency ash capacity thermal-mechanical robustness pressure drop—loaded and unloaded with soot regeneration fuel penalty cost impact on performance.

2 Modern Diesel Combustion

45

Critical for the material are thermal stresses in the trap material due to local peaks in temperature during spatially limited particulate combustion. If too much particulate matter is in the trap at the commencement of regeneration, the substrate material can be damaged if the energy released cannot be dissipated. Therefore, a material with a high heat capacity and a high thermal conductivity is an advantage. The effectiveness of different PM filter types was evaluated and It appears that the different filters all can reach a similar good PM mass-reduction. Particulate filter regeneration The oxidation of soot (carbon) requires in oxygen-rich exhaust gas a certain minimal temperature of around 550–600 °C. In vehicles the regeneration conditions (temperatures) are mission- and filter position-dependent. Active regeneration Burner systems, electrical heating systems, microwave systems and engine measures to increase exhaust gas temperatures (EGR, late post injection, reduced boost pressure and others). Passive regeneration These are approaches without any external energy addition. The Continuously Regenerating Trap (CRT), Fig. 2.24, uses a Pt-based oxidation catalyst before the filter media in order to convert nitrogen monoxide NO (90– 95% in engine raw emission) into nitrogen dioxide NO2. NO2 has been found to oxidise soot at a lower temperature than oxygen (Allansson 2000). The conversion of NO into NO2 is dependent on the temperature and the fuel sulphur content. That latter is due to the preferential oxidation of SO2 to SO3 over the Pt-catalyst rather than the oxidation of NO to NO2 and the conversion is substantially reduced with increasing sulphur levels (Warren et al. 1998). Furthermore, NO2 oxidises SO2 into SO3 (homogeneous catalytic gas phase reaction) and, therefore, the NO2-concentration is also reduced in the presence of high SO2-concentrations. Based on the reaction 2NO2 + C = CO2 + 2NO the theoretical NO2/soot-ratio is 7, 7 and for the reaction NO2 + C = CO + NO the NO2/soot-ratio is 3, 8 by weight. A basic problem is, that excessive nitrogen dioxide may be formed and passes through the filter and reaches the atmosphere and NO2 is not only toxic (MAK = max. permissible workplace concentration = 5 ppm vol.) but also an odorous brownish gas. Soot oxidation by NO2 is a ‘slow process’ and regenerations take considerably longer than with oxygen. Due to sulphate generation at elevated fuel sulphur contents (above 15–25 ppm sulphur), particulate matter standards of 0.02/0.03 g/kWh can even with a particulate trap not be achieved. The Catalysed Soot Filter (CSF) uses a reactive catalyst to facilitate soot oxidation. Its basic disadvantage is the inferior low temperature performance when compared to the CRT. Other approaches such as applying ultrasonic technology in order to increase particle sizes by ultrasonic agglomeration via a standing wave field and subsequent inertia separation are merely in an exploratory phase.

46

W. Knecht and P. A. Lakshminarayanan

Fig. 2.24 Continuously regenerating trap (CRT) with ceramic wall flow filter

Fuel-borne catalysts are mixed to the fuel and result in a reduced soot oxidation temperature. Different additive types have become available such as Satacen (Fe), Eolys (Ce), Octimax (Fe/Sr), Platinum Plus (Pt/Ce) and others (Cu, K, Na, Mn). Particulate traps and fuel catalysts can be assessed by evaluating the ‘balance point temperature’ where the back-pressure gradient is zero (the temperature at which the rate of oxidation of particulate matter and the rate of removal are equivalent so that there is no net build-up of soot). The ‘balance point temperature’evaluation has to be done with the same engine, since data from different engines are not accurately comparable, Fig. 2.25. The regeneration of PM-traps can be assisted by a number of engine measures which increase the exhaust gas temperature.

460

200

440

180

420

Exhaust Temperature [°C]

47

160

Engine Speed = 1960 rpm

400

140

380

120

360

100

T

340

Balance Point

= ~ 375°C

80

320

60

300

40 Temp. bef. CSF [°C]

280

20 dp CSF [mbar]

260

0 50

1,

Pressure drop over filter [mbar]

2 Modern Diesel Combustion

0 00

2,

0 50

2,

3,

0 00

3,

0 50

4,

0 00

4,

0 50

5,

0 00

5,

0 50

0 00

6,

0 50

6,

7,

0 00

7,

0 50

0 00

8,

8,

0 50

9,

0 00

0

9,

0 50

0 00

,

10

Time [s]

PM filter type: Fuel type:

CSF IMF-BP5 (2 ppm S) + Octimax

Soot loading before test: 80 g

Fig. 2.25 Evaluation of balance point temperature

Fuel effects The fuel characteristics and notably, its sulphur content is very important in the context of exhaust gas aftertreatment and exhibit a significant effect on the particulate trap operation. Not only the fuel sulphur but also the sulphur of the lube oil can be relevant to the operation of a particulate trap system. It has been estimated that the sulphur from the lube oil can be equivalent to a fuel sulphur content of approximately 10 ppm. Ash accumulation in particulate traps The accumulation of ash originating from the lube oil and wear particles in the particulates filter does lead to a steady increase in back pressure as function of mileage and, hence, to an increase in fuel consumption. The lube oil ash consists mainly of calcium sulphate and zinc phosphate. Particularly, high-quality oils used in order to lengthen the oil change intervals have a large number of additives and thus produce more oil ash. This necessitates the need of additional maintenance in that the trap is to be periodically cleaned. This can become even worse if an engine has an elevated lube oil consumption. In fact, the lube oil consumption in modern diesel engine has been dramatically reduced. Particle filter cleaning systems for wall flow filters can be quite complicated and expensive. After a thermal regeneration and cool-down, the filter must be washed and subsequently dried. A cleaning process may take—depending on type—a few minutes up to one hour, Fig. 2.26.

48

W. Knecht and P. A. Lakshminarayanan

Fig. 2.26 Particulate filter cleaning equipment (FSX)

2.12.2 Nitrogen Oxides Reduction ‘Lean NOx Catalysts’ or ‘HC-DeNOx Catalysts’ A ‘lean NOx catalyst’ is an oxidation catalyst which contains additionally porous zeolites which have a HC storage capacity. Zeolites are sieves for molecules consisting of minerals (silicon, aluminium and oxygen) with a crystal structure with many cavities. In a zeolite—depending on the pore diameter—only special molecules are retained. ‘Lean NOx catalysts’ reduce NOx with hydrocarbons as reducing agent. Two basic catalysts types exist: • low temperature active Pt catalysts • high temperature active catalysts on the basis of ion-exchanged zeolites, typically Cu/ZSM-5. ‘Lean NOx catalysts’ have been employed as: • active systems: with diesel fuel addition as reducing agent • passive systems: without fuel addition the hydrocarbon requirements are relatively high:

2 Modern Diesel Combustion

49

For high temperature catalysts 12 00  CH2 00 þ NO þ 14:5 O2 ! 12 CO2 þ 12 H2 O þ 0:5 N2 For low temperature catalysts 14 00  CH2 00 þ NO þ 20:6 O2 ! 14 CO2 þ 14 H2 O þ 0:33 N2 þ 0:17 N2 O Low temperature catalysts lead to a nitrous oxide formation (N2O) which contributes about 5% to global warming and N2O is also reducing the ozone layer in the stratosphere. N2O is an undesirable secondary emission and a GHG of three-way catalysts of spark ignition car engines. ‘Lean NOx catalysts’ exhibit an elevated NOx-conversion usually in a relatively small temperature window, Fig. 2.27. This means that in a road vehicle the location of the catalyst is critical (and not freely selectable) in order to ensure the existence of the required exhaust gas temperatures. Modern diesel engines have usually very small HC-concentrations in the exhaust gas stream (max. 150 ppm) and, therefore, substantial quantities of diesel fuel have to be injected before the catalyst. Recent tests with an active ‘lean NOx catalyst’ system in a HD diesel engine have resulted in the ESC-test in max. 10% NOx-reduction, but 3–5% increased fuel consumption. NOx Absorbers This catalyst type is also known as ‘NOx storage catalyst’ or ‘Lean NOx-trap (LNT)’. Unlike catalysts, which continuously convert NOx to N2, LNT operation comprises several steps that occur as the engine cycles between lean and rich operating modes. During the lean-burn phase of the cycle, platinum promotes the oxidation of NO to NO2, which is needed to successfully store NOx until the fuel-rich part of the cycle. The NOx storage medium, usually an alkaline oxide such as barium oxide, is converted to nitrate by reaction with the NOx.

Fig. 2.27 NOx-conversion of ‘lean NOx-catalysts’

50

W. Knecht and P. A. Lakshminarayanan

The storage capacity of the oxide is not unlimited, so after the engine runs for some time, it becomes necessary to empty the NOx store. The engine rich is run rich for some seconds for removing the NOx, for decomposing the nitrate and releasing of the NOx. This also results in higher levels of carbon monoxide and hydrocarbons in the exhaust stream. These two compounds react with the released NOx, reducing it to N2. Then, engine runs lean again and the cycle repeats. Therefore, NOx absorbers work in a discontinuously. They are sensitive to sulphur which is trapped as a sulphate on the NOx storage sites. The procedure for desulphurisation of catalyst, however, needs high temperatures in the region of 700 °C for a period of a few minutes. Desulphurisation is never complete and, therefore, the NOx conversion reduces with operating time (Geckler et al. 2001). The different steps of the reaction mechanism are presented in Fig. 2.28. • oxidation of NO to NO2 (2NO + O2 ! 2NO2) on precious metal, such as Pt. The majority of NOx emitted from the engine is in the form of NO. NO2 is more easily adsorbed than NO. • adsorption of NO2 at k > 1 • desorption in short phases at k < 1 in which the nitrates are split • reduction of NO2 at k < 1 in the presence of hydrocarbons and carbon monoxide at the rhodium of the catalyst leading to the formation of N2, CO2 and H2O.

Fig. 2.28 Operating-principle of lean NOx-trap

2 Modern Diesel Combustion

51

A fuel with less than 10 ppm sulphur is, therefore, essential. The durability of a storage catalyst depends heavily on the used diesel fuel characteristics. NOx-absorber catalysts exhibit in steady state conditions conversion rates up to 90% and are effective in a temperature range (250–450 °C). A NOx-absorber has to be integrated into the engine management system in order to account for the actually stored NOx-mass. The adsorbed NOx depends on several factors such as NOx-flow rate from the engine, duration of adsorption, exhaust gas mass flow rate, the exhaust gas temperature, the air/fuel ratio and others. The engine management system is also responsible for the periodic change-over from the oxidation/adsorption-phase to the desorption/reduction-phase in dependence to the stored NOx-mass, the exhaust gas temperature, the exhaust gas mass flow rate, etc. The desorption/reduction-phase can be initiated with a common rail injection system by means of a late post-injection. The latter may be problematic in larger engines since it may lead to wall-wetting by the late injected fuel and subsequently affect the lube oil life negatively. With other injection systems including unit injectors, HC-addition in form of diesel fuel injection must be done anywhere in the exhaust pipe before the catalyst. In such applications, a special low-pressure fuel system was employed. Selective Catalytic Reduction (SCR) Selective Catalytic Reduction (SCR) has for some years been used in power plants (where heavy fuels with high sulphur content are used) and it has in the past years been demonstrated to work also in road vehicles (Fränkle et al. 1997). Although high NOx-conversion can be achieved in a continuous manner, negative for this system are so far the following points: • need of an additional liquid (urea/water solution = AdBlue). • weight and volume requirement on a commercial vehicle • before appropriate sensors became available only an open loop control was available • high cost of a system, and • urea infrastructure to be built-up. Selective catalytic reduction (SCR) using ammonia as a reducing agent can be considered a solution applicable in HDV. Liquid or solid ammonia precursors can be used since they can decompose in the vehicle’s exhaust system to form ammonia. Ammonia is toxic and for safety reasons considered not usable for on-board storage in a road vehicle. Therefore, a urea/water solution (named AdBlue) is used in road vehicles. Urea is widely available and used and is non-toxic and, therefore, a suitable reducing agent. Nevertheless, a number of other reducing agents have been assessed such as: Methanamide, Ammonium formate, Guanidinium formate, etc. (Jacob 2006). This was done in order to find a possible substance with a lower freezing point, Table 2.1. SCR has also been investigated for applications, both in cars and in LDV.

52

W. Knecht and P. A. Lakshminarayanan

Table 2.1 Properties of ammonia and ammonia-forming substances Ammonia precursor compound

Freezing point (°C)

Composition (by weight-%) HS

AdBlue (HS) Ammonium formate (AF) Denoxium (HS + AF) Methanamide (MA) Ammonium carbamate (AC) Urea (HS), globule, 2 mm dia.

−11 −35

32.5

−26/−30

20

−28

AF

MA

AC

40

77.5 60

0.20 0.13

0.22 0.14

26

54

0.20

0.22

20

0.30 0.44

0.33

0.57

0.42

80 >99

133

>99

H2O

Content of active ammonia Weight Volume (kg/kg) (kg/dm3)

For some years, a consortium consisting of IVECO, MAN, Daimler and Siemens has been engaged in the development of an SCR-system, especially for HDV applications. In this system, a controlled quantity of urea/water-solution is injected into the exhaust gas stream before the catalyst. The latter consists of two sections: First a hydrolysis catalyst in which the urea/water solution is converted into ammonia and secondly, the SCR-catalyst where the reduction of NOx occurs. Very important for the application of such systems in road vehicles is the accurate reducing agent dosage in relation to the actual NOx-flow rate. Thereby, a slippage of ammonia must absolutely be avoided under all operating conditions— particularly in transient conditions—of the engine. Ammonia in the exhaust gas after the catalyst can also be avoided by a third catalyst section consisting of an oxidation catalyst (slip catalyst). Early systems employed an extruded, full-ceramic cell catalyst (100, 200, 300 cells per square inch), made on the basis of mixed oxides and it is characterised by a high activity and selectivity. The material consisted mainly of titanium dioxide and wolfram trioxide. The vanadium pentoxide content affects mainly the temperature dependence of the NOx-reduction and the sulphur dioxide oxidation. The substrate consists of the catalyst material in contrast to other types where an active catalyst material layer is applied on the metallic or ceramic substrate. Three types of SCR catalysts are being evaluated: • Vanadium based SCR catalysts with exhaust gas temperatures below 650 °C. • Iron zeolite SCR catalysts • Copper zeolite SCR catalysts. In actual HDV operation, the engine copes with rapid load- and speed-changes. The exhaust- and NOx-mass flow rates as well as the exhaust gas temperature, therefore, vary continuously. This has to be accounted for at all times in the dosage of the reducing agent. A special electronic control unit (ECU) was developed,

2 Modern Diesel Combustion

53

which accounts for the actual state of engine and catalyst and affects—based on complex computations—the urea dosage. The SCR-ECU communicates with the ECU of the fuel injection system via a CAN-Bus (Controller Area Network). Until suitable sensors (ammonia, NOx) become available, the control is open-loop. The injection of urea into the exhaust gas stream was initially supported by pressurised air from the vehicle air system. The injection occurs via a multi-hole nozzle and the aim is to obtain a homogeneous distribution. Today a direct liquid AdBlue injection without compressed air is used. A further development includes a pre-catalyst prior to the hydrolysis catalyst in order to convert NO into NO2 and then the reaction in the SCR-catalyst is much faster, Fig. 2.29. This aims at a volume reduction of the catalyst. The effectiveness of an SCR system in the different points of an ESC test is presented in Fig. 2.30. For EURO 5 and especially EURO 6 legislation for the higher vehicle weight classes, an exhaust gas aftertreatment appears not avoidable. While with ‘lean-NOx-traps’ the main problems are the achievement of a long-time stable calibration of the ‘rich operation’ and an improvement of the ageing effects (especially due to sulphur), SCR does offer a higher conversion rate and no significant technical problems. The reduction of nitrogen oxides was very considerable. Additionally, it was noted that the hydrocarbon (HC) emissions are also substantially reduced, while this was accompanied by a carbon monoxide (CO) increase. The fuel consumption was not negatively affected by the use of a SCR-system.

EXHAUST GAS

HY

SCR 2NH 3 + 2NO + ½O

2

→ 2N 2 + 3H 2O

CO(NH 2 )2 + H 2O → 2NH 3 + CO 2

UREA HY

SCR

OX 2NH 3 + 1½O

2

→ N 2 + 3H 2O

2NH 3 + NO + NO OX

HY

SCR

→ 2N 2 + 3H 2O

OX NO + ½O

OX

2

PM TRAP

SCR

2

→ NO 2 OX

C + 2NO 2 → CO 2 + 2NO NO + ½O

2

→ NO 2

Fig. 2.29 Different SCR-system configurations. HY = hydrolysis catalyst; SCR = SCR catalyst; OX = oxidation catalyst

54

W. Knecht and P. A. Lakshminarayanan

Fig. 2.30 Emissions before/after SCR-system in ESC test

The SCR-system functions not only in steady state conditions, as shown in Fig. 2.31, but also results in significant NOx-reductions in transient tests and on the road. Considerable improvements in NOx-conversion with SCR-systems have been achieved with a model-based urea dosing strategy as shown in Fig. 2.32. In actual road operation, the consumption of urea/water-solution is approximately 3–5% of the diesel fuel consumption. Since the crystallisation point of urea/ water solution is −11.5 °C, a viable solution for an operation at very low ambient temperatures is needed. Substantial NOx-reductions in the ETC- test (European Transient Cycle) and in vehicle tests on the road have been noted.

Fig. 2.31 NOx-emissions before/after SCR-system; European transient cycle

2 Modern Diesel Combustion

55

Fig. 2.32 High-e SCR dosing algorithm of model-based urea dosing strategy

A novel ammonia storage was suggested by Amminex of Denmark. Thereby, ammonia is absorbed in a solid storage material which is called AdAmmine. Therefore, ammonia is not stored in gaseous form. Ammonia is released either with electric heating or engine coolant from the cartridge containing the AdAmmine. A simple dosage and direct injection of ammonia into the exhaust pipe is employed. This system does not suffer from the fact that the reductant does freeze at −11 °C as in case of AdBlue. An AdBlue-tank requires due to its low active ammonia content of approximately 0.22 kg ammonia/dm3 a significantly larger tank volume. Using ammonia leads to an improvement of the low temperature NOx-conversion which is still a disadvantage using AdBlue with SCR systems. The system consists of an ammonia storage container for a fast cold-start ammonia availability and a control unit which is accounting on a flow sensor, a pressure sensor and a valve. SCR System using ammonia carbamat as an ammonia source. Another development aims at the use of a solid reductant, such as ammonia carbamat Fig. 2.33. Such a system has been tested in a city bus in Germany (Krüger et al. 2003). The use of a solid reductant could be attractive for cars and LDV’s. The advantages of a solid reductant are: reduced volume of ammonia carbamat of 28% relative to AdBlue (=100%), no freezing of ammonia carbamat, improved mixing of the gaseous, ammonia/CO2-gas mixture with the exhaust gas prior to the catalyst, no hydrolysis needed, no ammonia storage on-board, because the evaporation is reversible.

56

W. Knecht and P. A. Lakshminarayanan

Diesel Engine

cooling water

ECU NH 3 control valve

raw exhaust gas

NH 3

Ammonia Carbamat (powder )

cooling water

Oxidation Catalyst

SCR -System

cleaned exhaust gas

Fig. 2.33 SCR-system employing ammonia carbamat

Combined Aftertreatment Systems For EURO VI additionally a PM-trap, possibly EGR and an SCR-system might be required. A combined system consists of a PM-trap and an SCR-system. Several combined aftertreatment systems consisting of a diesel particulate trap and an SCR-catalyst have been proposed, Fig. 2.34. One method to reduce the total volume of catalysts in the diesel exhaust system is to combine the SCR and DPF catalyst technology by coating SCR catalyst technology on particulate filters. SCR catalysts coated on filters must have a high thermal durability to withstand repeated soot regenerations. Cu-zeolite catalysts have demonstrated durability up to 900 °C (Ballinger et al. 2009; Fedeyko et al. 2009). In Japan Toyota has developed the DPNR (Diesel Particulate NOx Reduction) which is a combination of PM-trap and ‘lean NOx trap’ with one substrate. Eaton has proposed a non-urea diesel emission aftertreatment system. As a reductant is diesel fuel employed. The ‘lean NOx trap’ (LNT) breaks down NOx into nitrogen and ammonia. Initially fuel is injected into the exhaust gas stream by an electronically controlled dosing system. The injected fuel and the exhaust gas pass a reformer which produces hydrogen (needed for the subsequent ammonia NH3 production) and CO. NOx is reduced in the LNT as well as in the SCR-catalyst. In principle the system offers several advantages: no need of AdBlue inclusive AdBlue-quality sensor, the tank and reductant heating system. The first SCR trucks were delivered in Europe in 2006. More than 3 million of these medium and heavy-duty vehicles are forecast to be in use by the year 2020. More than 24 million passenger cars and 6 million light-duty vehicles are likely to use SCR technology by 2020. All over Europe and Japan a significant number of AdBlue distribution stations have in the meantime become available.

2 Modern Diesel Combustion

57

Fig. 2.34 Combined exhaust gas aftertreatment systems (EMINOX)

2.13

Heat Recuperation

Still a considerable portion (approximately 30%) of the energy input in the fuel is wasted and leaves the internal combustion engine via the exhaust gas stream. Therefore, some means of better utilising the exhaust energy would be extremely beneficial. One possibility is the adoption of a turbocompound system where a power turbine after the turbine of the turbocharger is used to recover more energy from the exhaust gas and the power extracted is fed via a mechanical drive system back to the crankshaft, Figs. 2.35 and 2.36. Thereby an integrated reduction gearbox is employed. The gear drive results in reaction forces for which the bearing system of the power turbine has to be designed. Typically, approximately 15% of the total engine power can be provided by the power turbine. In real vehicle application, instead of a mechanical turbocompound system, the power turbine can be linked with a high-speed generator to produce electricity which can be used to cover the on-board requirements or it can be fed back to the crankshaft via a flywheel motor. Another approach which is not new but was always considered to be too expensive and technically involved represents the addition of a ‘steam cycle’ in the exhaust, Fig. 2.37. Thereby, an evaporation and condensation process are used to regain exhaust gas and EGR-cooler heat and transfer it into mechanical energy. As a working medium a low-boiling liquid such as water, alcohols and others can be

58

Fig. 2.35 Compound turbocharging

Fig. 2.36 Mechanical turbocompound system

W. Knecht and P. A. Lakshminarayanan

2 Modern Diesel Combustion

59

Cooling Air

Charge Air Cooler

Diesel Radiator

Rankine Radiator

Water pump

Diesel Engine

C

T

Generator

Diesel Oxidation Catalyst (DOC)

Expander

EGR Cooler

Exhaust

Diesel Particulate Filter (DPF)

Rankine Heat Exchanger SCR Catalyst Feeder Pump

Fig. 2.37 Heat recuperation system (Hountalas et al. 2012)

employed. This working medium will be heated in a heat exchanger by the exhaust gas stream such that they are transferred to steam. The steam can subsequently be used in a steam engine or steam turbine to generate mechanical energy and electricity. After that, the steam is in a condensation brought back to the liquid state and in a closed loop recirculated to the heat exchanger. In a commercial vehicle, simulations have shown that in long-distance haulage considerable gains and even in urban missions ‘some’ gains in fuel consumption are feasible. The largest improvement in fuel consumption was found to be in the organic Rankine cycle with approximately 11.3% and the corresponding value for steam was approximately 9%. In both cases also the EGR- and CAC-heat was utilised (Hountalas et al. 2007, 2012).

2.14

Flexible Engine Systems

Apart from a flexible control of the fuel injection system and the air flow to the engine with a turbocharger with variable turbine entry geometry, other flexibly controllable systems could exhibit in some applications advantages: • variable valve actuation has been shown to lead to the following advantages: internal exhaust gas recirculation, lower fuel consumption (gas exchange), improved transient response, increased low-speed torque, improved startability

60

W. Knecht and P. A. Lakshminarayanan

and warm-up, temperature management for exhaust gas aftertreatment, increased exhaust brake power, • Variable compression ratio (VCR) is expected to improve cold startability, emissions and fuel consumption. • Variable swirl can lead to a better compromise between low engine speed (increased swirl to reduce smoke or increase BMEP) and at high engine speed (over-swirling). • Variable cooling system or thermal management systems allow a control of the cylinder temperatures over the full range of running conditions. This can be realised by replacing a mechanically driven water pump with an electrically driven water pump in conjunction with an electric valve and fan which are all electronically controlled. Nucleate boiling is seen as the ultimate goal of engine thermal management as the thermal capacity of the coolant during the initial part of the phase change (prior to film boiling) allows the engine temperature difference to be further reduced. This point is achieved with an average flow rate of approximately 20% of that required on a non-nucleate boiling system and reduces the pump requirements greatly. • Even the oil pump can be replaced by a smaller oil pump, and an electrically driven oil pump which provides increased flow under extreme conditions. An air-hybrid (pneumatic hybrid engine) has been suggested, Fig. 2.38. This technology connects each cylinder via a fully variable charge valve to a shared pressure tank. During vehicle braking, the fuel is cut off and the engine takes in air, pumping it to the pressure tank. The compressed air is later used for starting or driving the engine without fuel. Hence very low fuel consumptions are expected (Voser et al. 2012; Moser et al. 2012).

2.15

Losses

Losses occur in various processes and engine components and they should be reduced as much as possible. In a modern HD diesel engine, the mechanical losses might be distributed as follows, Fig. 2.39.

2.16

Fuels

The characteristics of the diesel fuel properties are in compression ignition engine of great importance, Table 2.2. For example, Cetane number, low sulphur content. Due to the move away from nuclear energy and the potential shortage of fossil fuel such as oil, it is likely that for road vehicles renewable fuels will be more often be

2 Modern Diesel Combustion

61

Fig. 2.38 Pneumatic hybrid system proposal (ETH Eidgenössische Technische Hochschule, Zürich)

40% 30% 20% 10% 0%

Fig. 2.39 Distribution of mechanical losses of a HD diesel engine

used. Already now natural gas and coal derived fuels such as Fischer-Tropsch fuel are more often used. Natural gas and liquid fuels as derived from fracking from slate are potentially more frequently used. Other renewable fuel candidates are: bio fuels, dimethyl ether,

–

– °C

– kg/kg

Kin. Viscosity at 20 °C Kin. Viscosity at 40 °C Carbon content) Hydrogen content Oxygen content

Evaporation heat (kJ/kg) Ignitability ()

38 220

MJ/kg

Lower calorific value Cetane number Autoignition temperature, Octane number Theoretical Air requirement Boiling point,

33.1

mm2/s

% weight % weight % weight

78.7

vol. % gas in air mm2/s

°C

0.90– 0.93 35.0

kg/dm3

Density liquid

Chemical formula

Rape seed oil

Table 2.2 Properties of alternative fuels

76.85 12.4

4.9

6.8–7.8

3.5–15

334–347

– 12.5

49 153

37.7

0.88

Rape seed methyl ester (RME)

52.2 13.0 34.8

3.4–18

38 12 50

5.5–26

1110

65

−20 460

111 6.5

5 450

19.5

– 9.0

>55 235

27.6

CH3–OH

CH3–O–CH3 0.79

Methanol

Dimethyl-ether (DME)

52 13 35

3.5–15

904

78

108 9.0

8 420

25

CH3– CH2–OH 0.79

Ethanol

86 14 0

3.2

0.6– 6.5 3.08

180– 370 250

– 14.6

40–55 250

42.5

0.842

Diesel fuel

84.9 15.1 0

3.57

210–338

–

>74

0.77

Fischer Tropsch fuel

62 W. Knecht and P. A. Lakshminarayanan

2 Modern Diesel Combustion

2.17

63

Gas Engines

Conversions of diesel engines are commonly changed to spark ignition engines for lean burn or stoichiometric combustion systems. Fuel admission is mainly manifold injected by single or multipoint injection. Also, direct injection of the gaseous fuel into the combustion chamber with spark ignition has been investigated. Recently new results with dual fuel have been presented whereby a small quantity of diesel fuel is injected in order to start ignition of the gaseous fuel (Rohrssen and Höffeler 2011).

2.18

Future Developments

Despite some promising cars, LDV’s and HDV with fuel cell propulsion exist, a large introduction of fuel cell powered cars is unlikely to exist before two to three decades. This is due to the fact that a hydrogen production and infrastructure has to be built up. Therefore, for road transport, the diesel engine has to be further improved. For some applications some renewable fuels and gas engines might be used. A key question is related to the legislation: Should the currently usual emissions and the CO2-emission (respectively the fuel consumption) be further reduced or should mainly the CO2-emissions be reduced. For road vehicles the matching of the powertrain and the vehicle characteristics is of great significance since it affects fuel consumption and emissions. Considering EURO VI engines as state of art, the engine is characterised by: • SCR to reduce nitric oxides • High EGR, of the order of 30% in conjunction with – High pressure common rail, of the order of 2000 bar • Multiple injection for reducing noise, nitric oxides and smoke • Diesel Particulate Filter to control particulate number • High cylinder pressures of the order of 180–200 bar. What will in future be required technologies can only partially be answered, without reliable emission targets The logical areas are: • Hybridisation with electric prime movers for city and low speed running when torque demand is low • Injection pressures rising to 3000 bar • Cylinder pressures rising to 220 bar • More integration of vehicle operation with engine • New homogeneous combustion systems to limit nitric oxides and particulates in cylinder • Alternative or new fuels like dimethyl ether, natural gas with related hardware.

64

W. Knecht and P. A. Lakshminarayanan

References Knecht W (1993) Geschichte der Verbrennungsmotoren-Entwicklung in der Schweiz, Verlag Oskar Baldinger, 5222 Umiken. ISBN 3-905 129-06-10 Heldt PM (1936) Saurer diesel for light vehicles, automotive industries, 26 September 1936 Knecht W (2004) Some historical steps in the development of the common rail injection system. Newcom Soc. 74:89–107 Imarisio R, Rossi Sebastiano GM (2000) Potential of future common rail diesel engines. In: International symposium ‘the future of diesel engine technology for passenger cars, Porto Cervo/Sardegna, Italy, 12–13 October 2000 Egger K et al (2000) Piezo—electric common rail fuel injection systems: state of the art. In: International symposium ‘the future of diesel engine technology for passenger cars, Porto Cervo/Sardegna, Italy, 12–13 October 2000 Krieger K (2000) Diesel fuel injection technology—an essential contribution towards environment friendly powerful diesel engine. In: SAE 2000, India mobility conference, Delhi, 13–14 January 2000 Dingle PJG, Lai M-CD (2005) Diesel common rail and advanced fuel injection systems. SAE Steward J et al (2008) The digital engine—combustion control using hydraulic valve actuation an air injection, Comodia 2008, Sapporo, Japan, 28–31 July 2008. www.sturmanindustries.com Flik M et al (2009) Emission reduction in commercial vehicles via thermomanagement. In: 30th international Vienna engine symposium Stanton DW (2013) Systematic development of highly efficient a clean engines to meet future commercial vehicle greenhouse gas regulations. In: SAE International 2013 L. Rey Buckendale Lecture Frankenberg C, Meirink JF, van Weele M, Platt U, Wagner T (2005) Science 308:1010–1014 Keppler F, Hamilton JT, Braß M, Röckmann T (2006) Nature 439:187–191 Ferretti DF, Miller JB, White JW, Etheridge DM, Lassey KR, Lowe DC, Meure CM, Dreier MF, Trudinger CM, Van Ommen TD, Langenfelds RL (2005) Science 309:1714–1717 Lowe DC (2006) Nature 439:148–149 Ambus P, Zechmeister-Boltenstern S, Butterbach-Bahl K (2006) Biogeosciences 3:135–145 http://delphi.com/pdf/emissions/Delphi-Passenger-Car-Light-Duty-Truck-Emissions-Brochure2013-2014.pdf http://delphi.com/pdf/emissions/Delphi-Heavy-Duty-Emissions-Brochure-2013-2014.pdf Giechaskiel B et al (2012) Measurement of automotive nonvolatile particle number emissions within the european legislative framework: a review. Aerosol Sci Technol 46:719–749 Scarth P, Ortiz D (2004) Idle noise reduction of light, medium and heavy-duty diesel engines. In: Thiesel 2004 Conference, Valencia, Spain, 7–10 September 2004 Meinrad S, Giorgio C (1989) Laboratory results in particulate trap technology. SAE-paper 890’170 Becker K, Heine P (1999) Konzept und Ergebnisse des Russfilter-Grossversuchs der Bundesrepublik Deutschland, TÜV-Bericht 12/99 Allansson R et al (2000) The use of the continuously regenerating trap (CRTTM and SCRTTM) to meet future emission legislation, 21. In: Wiener Motorensymposium, 4–5 May 2000 Warren JP et al Effects on after-treatment on particulate matter when using the continuously regenerating trap (CRTTM). IMechE seminar diesel engines—particulate control, London, 23 November 1998 Geckler, Sam, Dean Tomazic, Volker Scholz, Margaret V. Whalen, Dale McKinnon, John Orban, Robert A. Gorse, Owen Bailey, and James C. Hoelzer (2001) Development of a Desulfurization Strategy for a NO x Adsorber Catalyst System. SAE Transactions: 406–414 Fränkle G et al (1997) SINOx—Ein Abgasreinigungssystem für Nutzfahrzeuge. In: Wiener Motorensymposium Jacob E (2006) Perspectives on Mobile SCR Technology, 15. Aachener Fahrzeug- und Motorentechnik

2 Modern Diesel Combustion

65

Krüger M et al (2003) Ein kompaktes Feststoff-SCR-System, MTZ 6/2003, Jahrgang 64 Ballinger T et al Evaluation of SCR catalyst technology on diesel particulate filters. SAE 2009-01-0910 Fedeyko JM et al Development of thermally durable Cu/SCR catalysts. SAE 2009-01-0899 Hountalas DT, Mavropoulos GC, Katsanos C, Knecht W (2012) Improvement of bottoming cycle efficiency and heat rejection for HD truck applications by utilization of EGR and CAC heat. Energy Convers Manag 53:19–32 Hountalas DT et al (2007) Study of available exhaust gas heat recovery technologies for HD diesel engine applications. Int J Altern Propuls 1(2/3):228–249 Hountalas DT et al (2012) Improvement of bottoming cycle efficiency and heat rejection for HD truck application by utilization of EGR & CAC heat. Energy Convers Manag 53(1):19–32 Voser C et al (2012) In-cylinder boosting of turbocharged spark-ignited engines. Part 1: model-based design of the charge valve. Proc IMechE. Part D: J Automob Eng 226:1408–1418 Moser M et al (2012) Design methodology of camshaft driven valves for pneumatic engine starts. In: IFAC workshop on engine and powertrain control, simulation and modelling Rohrssen K, Höffeler G (2011) The IAV active High-EGR concept. MTZ Worldw eMagazine 72 (1):22–27

Chapter 3

Supercharging Walter Knecht and P. A. Lakshminarayanan

Abstract Historically, diesel engine was supercharged to increase power. The charger was run by mechanically by the engine crankshaft. This enabled airflow as a function of engine speed, ideal for a piston engine. However, as it would, it wasted a lot of exhaust energy and turbocharging was invented where the turbine driven by the exhaust gases drives the compressor. This has its own problems like mismatch in air flow in terms of lower airflow than required at lower speed range and higher airflow at higher speed range. This was resolved by various techniques. The easiest, but not the most efficient, method is waste gating the turbocharger. Variable turbine entry turbocharger, two stage turbocharger and other ingenious solutions increased the overall efficiency of the engine and improved the power to weight ratio of the diesel engine. Surging, wheel speed and choking are the boundaries of operation of a turbo compressor. The boundaries are stretched by new designs of the compressor wheel. With the advent of electronics intelligent control of turbocharger enabled enhancing the airflow in the low speed range and hence the engine torque. Charge air cooling plays an important role in increasing the air flow, reducing nitric oxide emissions as well as the power. The development of turbochargers is continuing with new concepts that enhance power, engine efficiency as well as emissions.

W. Knecht Knecht Engine Consulting GmbH, Kesswil, Switzerland P. A. Lakshminarayanan (&) Formerly with Simpson and Co. Ltd., Chennai, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_3

67

68

3.1

W. Knecht and P. A. Lakshminarayanan

Historical Development of Supercharging

The power of an internal combustion engine can be increased with three possibilities: – Increase of the cylinder capacity – Increase of the engine speed – by an improved filling of the cylinders with air (supercharging) Already Rudolf Diesel mentioned supercharging in his patent dated February, 28 1892 whereby pressurized air was led into the cylinder before start of compression. The basic examples of exhaust gas turbocharging are based on Swiss patent 35259 of 1905. This patent was conceived by the Swiss engineer Alfred Büchi (1879–1959), Fig. 3.1 (Buchi 1953).

Compressor

Air

Charge AirCooling

Turbine

Piston

Exhaust Gas

Fig. 3.1 Turbocharged diesel engine according to Büchi’s Swiss patent 35259 from 1905

3 Supercharging

3.2

69

Purpose of Supercharging

Supercharging with/without charge cooling permits to realise a higher charge air density in the intake manifold and it is widely used to increase the engine power and to increase the specific power per unit cylinder capacity, i.e. downsizing of the engines. Pe pecm ¼ Vh 40 s where, Pe Vh pe cm s

power, kW cylinder capacity, dm3 brake mean effective pressure, bar mean piston speed, m/s piston stroke, m

3.3

Supercharging Methods

Supercharging can be achieved by a mechanically from the engine via a belt, a chain, gears or a shaft driven compressor or with an exhaust gas driven turbine: – – – – – – –

Sliding vane supercharger Roots rotating lobe supercharger Screw type supercharger (Lysholm compressor) G-oscillating spiral displacer supercharger centrifugal compressor pressure wave supercharger (Comprex) turbocharger

The pressure wave supercharger Comprex was developed over many years by ASEA (formerly Brown Boveri), Fig. 3.2. It is a very clever machine which has been tested in cars as well as in commercial vehicles. The energy to compress the fresh air is due to short time direct contact of the exhaust gas stream and fresh air transmitted by pressure waves. The energy exchange happens in open straight channels of a rotor which are periodically closed. Some superchargers have a lower adiabatic efficiency than turbochargers. The adiabatic efficiency is a measure of a compressors ability to compress air without adding excess heat to that air. In the 1950s several manufacturers produced engines with mechanically driven compressors. The power required to drive the compressor led to an increased engine fuel consumption. In on-road operation, however, the vehicle fuel consumption was not always significantly higher than of a turbocharged engine. Mechanically driven

70

W. Knecht and P. A. Lakshminarayanan

Fig. 3.2 Comprex pressure wave supercharger; A = gas housing, B = rotor, C = belt; D = air housing; 1 = suction air; 2 = charge air; 3 = exhaust gas from engine; 4 = exhaust

superchargers had a favourable lag (the time needed for the exhaust system and supercharger to generate the required boost pressure) and the exhaust brake performance was in vehicles very good.

3.4

Turbocharging

The main components Fig. 3.3 of a turbocharger are: – an exhaust gas driven turbine (mostly radial inflow turbine) and housing – a compressor wheel (mostly centrifugal compressors) and housing – An interconnecting support spindle mounted on a pair of fully floating plain bearings which are themselves encased in a central bearing housing made from Ni-cast iron. Compressor and turbine matching are essential to achieve best performance of the engine. Exhaust gas turbochargers have in recent years been dramatically improved offering an increased compressor pressure ratio and a certain control of the air flow to the engine. Some features such as waste gates are standard and in small engines turbochargers with variable turbine entry geometry are widely used. In heavy duty truck diesel engines production started already in 1998 with turbochargers with variable turbine entry geometry. Since a general trend to downsizing engines for commercial vehicles is noted, novel turbochargers and turbocharging system development is in progress. Single stage turbochargers with axial-radial compressor wheels might be one solution. Another costlier approach is the use of two-stage turbocharging.

3 Supercharging

71

Turbocharging

Fig. 3.3 Turbocharger design

From an engine point of view, the degree of supercharging is limited by the maximum permissible mechanical and thermal loading of the engine components. There appears to be a tendency to increase the maximum cylinder pressure from 160 bars towards 250 bars. A limitation is also given by the piston: While currently used piston alloy can tolerate 0.45 kW/mm2, with new developments of piston alloys (steel) a loading of 0.55 kW/mm2 appears feasible. The progress in turbocharger design has been such that durability, bearings and heat resistance have been continuously improved. The increases in achievable engine power are linked to improvements in compressor performance. Increased specific powers also affect the thermal loading of the materials of the compressor wheel and the combustion chamber. The desired torque characteristic for heavy duty diesel engines allows that over a wide speed range and optimisation of fuel economy and driveability is essential and this translates in high boost pressure, wide compressor operating characteristics and some form of air handling control. Parallel to aerodynamic and mechanical design, the development in materials and manufacturing techniques are essential. Today’s turbochargers already use high-grade aerospace alloys in turbine and compressor and still they operate more and more at the limits of their capabilities. For high boost pressures, compressor wheels made from cast titanium are already in production (Cummins Turbocharger Technology formerly Holset) (Tennant 2001). The design and material choice of compressor and turbine is important to account for stresses induced by resonant blade vibrations and alternating stresses induced by speed cycling of the turbocharger. Another area where materials are essential is turbine and compressor housings to provide better stability and resistance to thermal fatigue and weight reduction.

72

W. Knecht and P. A. Lakshminarayanan

The obvious limitation of a fixed geometry turbocharger is that it only can produce power in a relationship unique to the engine speed and air flow rate; i.e. the matching is always a compromise leading to excessively high boost pressures at rated conditions and a low boost pressure in the low engine speed range as well as to an unacceptable transient response. In road vehicles a good transient response of a supercharging device is of considerable importance with regard to dynamic emission tests. Turbochargers need to offer an electronic control of the air handling system within the whole engine operating range. In order to realise a control of the turbocharging system wastegate, blow-off valves and variable geometry can be employed. At turbocharged engines the exhaust manifold arrangements with respect to cylinder combinations are to be carefully designed. This has to be done for either a constant pressure operation or for a pulse charging system. On the intake side a combined charging system (Cser 1971; Anisits and Spinnler 1978; Knecht and Signer 1981) can be beneficial. Thereby, tuned intake manifolds after the compressor are used, Fig. 3.4. Optimised for low engine speeds the combined turbocharging system offers increased air mass flow rates in the low engine speed range and reduced air flow rates at higher engine speeds.

Fig. 3.4 Combined turbocharging system of Saurer D4KT-B diesel engine; A = compensation volume; B = resonance tube; C = resonance volume; 1 = resonance tube; 2 = compensation volume; 3 = connecting pipe; 4 = resonance volume; 5 = suction pipe

3 Supercharging

3.4.1

73

Wastegate

One approach is the wastegate or bypass turbine. In bypassing the turbine wheel with some of the exhaust gas at high engine speed, a smaller turbine housing can be used and this leads to higher low speed boost pressures and an improved transient response. A wastegate also avoids over-speeding of the turbocharger. Undesired is the negative difference of boost pressure and pressure before turbine in the high-speed range, since this affects the fuel consumption negatively (gas exchange) and, therefore, the base turbine housing size and the required bypass level must be carefully matched. The wastegate opening is usually controlled by boost pressure acting on a spring biased actuator. A further step represents the development of a system permitting an electronically controlled modulation of the boost pressure. Furthermore, an electric actuation of the wastegate has been developed.

3.4.2

Variable Turbine Entry Turbocharger

The limitations of any waste gate system are primarily due to the relationship between increased bypass flow and decreasing apparent turbine efficiency. Contrary a variable geometry turbine (VGT) offers a range of turbine flow areas at high efficiency. A variable turbine entry area allows the turbine to be matched to the available energy in the exhaust gas. In order to vary the turbine entry area either swing vanes or an axially movable fixed vane nozzle ring (moving across the turbine inlet passage) can be adopted, Figs. 3.5, 3.6, 3.7. The availability of a wider turbine flow range without compromising efficiency allows the engine manufacturer to increase the engine speed range over which high boost pressure can be achieved. The ability to vary the turbine area during transient operation enables a faster engine response by a more rapid boost pressure build-up. The actuation of the area variation device involves a pneumatic actuator using electronically modulated high-pressure air taken from the vehicle’s air system. The pneumatic actuation has been changed to an electric actuation consisting of a brushless motor and a control electronics system, Fig. 3.8. The variable geometry turbocharger can also be beneficially used in order to increase the engine brake power, Fig. 3.9. A high braking performance which reaches the rated engine output is desired in commercial vehicles. A suitable system represents a decompression brake, matched with the VGT in which the pressure within the cylinder is released by a small opening of the exhaust valve shortly before TDC. The braking work of a conventional decompression brake is simply related to the engine displacement and exhaust back pressure while with a VGT more air is fed into the engine at higher boost pressure and, thus increasing the compression work.

74

W. Knecht and P. A. Lakshminarayanan

Fig. 3.5 Turbocharger with variable turbine entry geometry (Holset, Cummins Turbosystems)

Max.... Open

3/4 Open

Closed

Fig. 3.6 Turbine entry variation with moveable fixed vane nozzle ring moving across the turbine inlet passage

3 Supercharging

75

Fig. 3.7 Sliding vanes in a turbocharger with variable turbine entry geometry

Fig. 3.8 FPT Industrial Cursor 10 engine with turbocharger with variable turbine entry geometry (Biaggini and Knecht 2000)

76

W. Knecht and P. A. Lakshminarayanan

Fig. 3.9 VGT-advantage shown in (a with regard to the decompression brake (shown in (b)) for commercial vehicles

3.4.3

Two Stage Turbocharging

For very highly rated diesel engines two-stage turbocharging may be needed. Apart from higher boost pressure ratio, two-stage turbocharging leads to an increased air flow range and efficiency as well as an improved high-altitude

3 Supercharging

77

capability. The installation of a two-stage turbocharging system with an inter-stage cooling may, however, lead to installation problems in a road vehicle. Special development is needed to ensure that a two-stage turbocharging system has a good transient response and at high BMEP’s a low fuel consumption. If two turbochargers are in a series arrangement, the total pressure ratio is pTOT = pC1  pC2. Apart from conventional two-stage turbocharging also arrangements with high-pressure (HP) compressor and turbine bypass lines are feasible. These so-called ‘modulated two-stage turbocharging systems’ allow a control of the pressure on the exhaust side and with a large exhaust bypass flow the system reverts essentially to a single-stage operation, Fig. 3.10. Such a system offers optimised turbocharging across a wider range of engine speeds than conventional two-stage turbocharging. A single stage compressor will have a maximum pressure ratio capability limited by the compressor speed and the compressor outlet temperature in dependence on the compressor material. Thus, for highly rated engines a single stage compressor may not be able to give enough boost pressure to satisfy the air/fuel-ratio requirement. The maximum pressure ratio is limited by the compressor speed and the compressor outlet temperature which is depending on the compressor material. Furthermore, the sea level pressure ratio is limited by the required altitude margin and the application limit by the engine duty cycle and the life requirement. In these cases, two turbochargers in series arrangement may be the answer i.e. 2 stage turbocharging. Intercooling can be used between the compressor stages for added performance benefit with added complexity for ducting and installation. The benefits of 2 stage turbocharging are: high pressure ratios with each turbocharger running at relatively low rotational speeds; wide flow range; improved compressor efficiency at the operating point of each compressor; if intercooled high apparent compression efficiency. Potential problems of 2 stage turbocharging: additional high temperature aftercooling may be necessary to reduce temperature before conventional aftercooler; high temperature HP compressor wheel and HP compressor housing needed; twice the blowby flow into the crankcase; packaging complexity; oil system has to be capable of feeding two turbochargers; thermal cracking on HP stage more likely. Two stage turbocharging with modulated or switched HP stage: achieve a wide flow range using two stage turbocharging permits modulation or switch the HP turbocharger using compressor and turbine bypass valves; if bypasses closed the system behaves like conventional two stage turbocharging resulting in high pressure ratios at low and medium engine speeds; at high engine speeds the HP turbine and compressor are too small so they have to be progressively bypassed. Apart from two-stage turbocharging also sequential turbocharging has been tested. Thereby two different sized single-stage turbochargers are employed: the smaller turbocharger is ‘active’ at low engine speeds and the second larger turbocharger with its higher inertia rotor is in use at higher engine speeds. With this

78

W. Knecht and P. A. Lakshminarayanan

Fig. 3.10 a Two-stage turbocharging (above); b Modulated 2-stage turbocharging (below)

3 Supercharging

79

system no multiplication of the pressure ratios occurs. Problematic may be the transition phase from one turbocharger to another.

3.4.4

New Compressor Wheel Design

Increasing pressure ratios in single-stage compressors invariably means that the compressor wheel has to rotate at higher speeds and has to withstand higher temperatures. Advances in design, manufacturing processes and materials technology have made significant improvements possible. A possible alternative to conventional two-stage turbocharging represents to a certain degree the use of axial/radial compressor wheels which do permit higher compressor pressure ratios, Fig. 3.11. A new car turbocharger for speeds up to 240,000 rpm and the use of a magnetic bearing has been tested. Because of no physical contact there would be less friction and no need for lubricating oil which also might reduce the turbocharger cost.

Fig. 3.11 Axial/Radial compressor wheel

80

3.5

W. Knecht and P. A. Lakshminarayanan

Boosters and Electric Turbochargers

Other systems such as the hydraulically or electric driven turbo compressors (booster) permit an improved assisted transient response. Thereby, the booster is arranged prior to the compressor of the turbocharger and the device is usually only a few seconds after the start of acceleration in operation. Another system comprises an electric motor/generator on the turbocharger shaft between the compressor and turbine wheels, Figs. 3.12 and 3.13. In motored operation, the engine benefits from acceleration assistance as in case of the electric booster and in generator mode electricity can be produced. The electrical system is made compatible with the engine/vehicle electrical system. This becomes particularly attractive as vehicle electrical systems move towards higher voltages. The turbocharger feeds excess exhaust energy—converted to electricity—into the system at high speeds and draws electrical energy to supply ‘air on demand’. All above applications require the availability of high speed electrical machines allowing speeds in excess of 100,000 rpm and the appropriate power electronics.

Fig. 3.12 Turbocharger with electric motor/generator

3 Supercharging

81

120000 110000

Speed, rpm

100000 90000

Base turbocharger, Iner a 0.00174 kg m2

80000 70000

5 kW motor assist, Iner a 0.00209 kg m2

60000

10 kW motor assist, Iner a 0.00209 kg m2

50000 40000

0

1

2

3

4

Time, s

Fig. 3.13 Turbocharger speed versus time with base turbocharger and with 5 and 10 kW motor assist

3.6

Charge Air Cooling

Intercoolers provide a means of reducing the charge inlet temperature between compressor outlet and the engine’s inlet ports and this is associated with several advantages: • Reduction of components thermal stresses since the cylinder head temperature is even under high load low. • Increase the charge mass in each cylinder which permits an elevated power • It reduces nitrogen oxides and soot emissions due to a lower combustion temperature and it reduces black smoke at low engine speeds and high loads due to lower combustion temperatures In road vehicles, mainly air-to-air intercooling are employed due to its higher effectiveness.

3.7

Future Developments

With tighter emission standards for engines applied to marine, on-road or off-road, almost all the engines will be turbocharged with little elbow room for naturally aspirated diesel engines. The turbochargers will be two-stage for obtaining very large boost pressure. Low end boost will be increased by e-boost that would come on when required (BorgWarner 2017) or during acceleration. This will enable lower smoke and hence lower particulates during transient operation by at least 20%. New ideas like the BorgWarner R2S (Sweetland and Schmitt 2004), Fig. 3.14 are interesting as it solves the problem by combining two waste-gated turbochargers—

82

W. Knecht and P. A. Lakshminarayanan

Fig. 3.14 R2S two stage turbocharger (BorgWarner)

one with smaller turbine housing for the low speed range and the other with larger turbine housing for the high-speed range—that shift the responsibility at some mid-speed range. R2S can be implemented more easily than variable geometry turbocharger (VGT). However, the flexibility of the VGT cannot be ruled out. The problems of bearings of the variable geometry facing heat and pressure are solved by continuous improvement, e.g., by simply supporting the vanes instead of at one end or by using axially moving ring as discussed in the chapter.

References Anisits F, Spinnler F (1978) Entwicklung der kombinierten Aufladung an neuen Saurer-Fahrzeugdieselmotor D4KT. MTZ 39:10 Biaggini G, Knecht W (2000) The advanced iveco cursor 10 heavy duty truck diesel engine. No. 2000-05-0071. SAE Technical Paper BorgWarner (2017) Looking ahead into the future of turbocharging, BorgWarner Knowledge Library 2017, [email protected] Buchi AJ (1953) Exhaust turbocharging of internal combustion engines: its origin evolution, present state of development, and future potentialities. J Franklin Inst

3 Supercharging

83

Cser G (1971) Ein neuartiges Verfahren zur Verbesserung der Abgasturboaufladung. MTZ 32 (10):368–373 Knecht W, Signer M (1981) Development of the Saurer D4KT-B diesel engine. No. 810342. SAE Technical Paper Sweetland P, Schmitt F (2004) Regulated 2-stage (R2S) charging systems for future diesel applications. In: 2004 diesel engine emissions reduction conference Tennant H (2001) Heavy duty air handling. Engine Technol Int

Chapter 4

Introduction to Turbocharging—A Perspective on Air Management System D. A. Subramani, R. Dhinagaran and V. R. Prasanth

Abstract This chapter provides an introduction to air management system with an emphasis on turbocharging, its role in emissions, fuel economy and performance. A typical turbocharger has a radial turbine run by the exhaust gas (thus extracting useful energy) which in turn spins a compressor that compresses air drawn from the atmosphere through the intake system. Sending compressed air into the engine allows more fuel to be burnt within the same volume, increasing the efficiency of the engine (lower surface area means lower frictional and heat transfer losses). In other words, a turbocharger could be used to “downsize” the internal combustion engine, with reduced losses. It is also accompanied by reduced emissions since the air flow could be increased as desired with a turbocharger.

4.1

Introduction

Turbocharger comes in various types—one such classification is based on the aerodynamics: Fixed geometry, waste gate, Variable Geometry, multi-stage turbochargers. Innovative ideas also come up now and then in terms of replacing the radial turbine with an axial one and so on. Turbochargers come up with different bearing systems—journal bearing, which could be fully floating or semi-floating and ball or air bearings. Selecting the right Turbocharger is the starting point of a new design. Since airflow and pressure requirements vary over the operating range it is important to pick the right overall size of the turbocharger at this stage. Turbo matching, as it is called involves deciding the overall size, bearing system, and the size of the compressor and turbine stages. This requires an understanding of the aerodynamics, rotor-dynamics and materials of the turbocharger components. A brief overview of the aerodynamics of the turbocharger is provided. The basic working principles of a turbocharger are then discussed. Current design practice D. A. Subramani
R. Dhinagaran (&)
V. R. Prasanth Turbo Energy Private Limited (TEL), Chennai, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_4

85

86

D. A. Subramani et al.

relies heavily on simulations. Physics behind simulations both one-dimensional and three-dimensional are described. Simulations are used for aerodynamic design as well as for ensuring durability. Material involved in the design of Compressor, turbine, housings and bearings are briefly touched upon. Noise, vibration and lubrication and a brief overview of the innovations happening in the field of turbocharging are discussed.

4.2

Need for Air Management

Global warming has been recognized as a major threat to our existence and in order to mitigate the same world leaders from 195 countries adopted the first-ever, legally binding global climate pact in December 2015 (called the Paris Accord) and developed an action plan to limit global warming to below 2 °C, come 2050. To meet these objectives, Green House Gas (GHG) emissions have to be cut by 85% within the next 33 years. This means reducing CO2 emissions by 2.6% per year on average, or 0.6 Gt per year in absolute terms. The transportation sector accounts for fourteen percent of the total global GHG emissions and the automotive industry, in particular, is under pressure to step-up emissions reductions. So far, due to aggressive emissions targets being set (107 g CO2/km by 2025 in the US) by regulators in the major markets, the industry has managed to achieve significant cuts in emissions (Fig. 4.1). Apart from CO2 targets countries have norms for other harmful emissions from combustion such as NOx, CO and particulate matter. These norms are dependent on the vehicle type and Table 4.1 provides one such for the passenger car segment. These are aggressive, for example moving from Euro V to Euro VI the NOx reduction is more than 50% for a diesel engine. Turbocharging is one of the ways to achieve emission targets, at least so far.

Fig. 4.1 Target CO2 reduction over time (http://www.acea.be)

4

Introduction to Turbocharging …

Table 4.1 European emission norms for passenger cars (ICCT, 2016)

4.3

Emission norm Diesel Euro 1 Euro 2 Euro 3 Euro 4 Euro 5a Euro 5b Euro 6 Gasoline Euro 1 Euro 2 Euro 3 Euro 4 Euro 5 Euro 6

87 Date

CO

NOx

PM

Jan-92 Jan-96 Jan-00 Jan-05 Sep-09 Sep-11 Sep-14

2.72 1.00 0.66 0.50 0.50 0.50 0.50

– – 0.50 0.25 0.18 0.18 0.08

0.14 0.080 0.050 0.025 0.005 0.005 0.005

Jan-92 Jan-96 Jan-00 Jan-05 Sep-09 Sep-14

2.72 2.20 2.30 1.00 1.00 1.00

– – 0.15 0.08 0.06 0.06

– – – – 0.005 0.005

Turbocharger Basics

The history of turbocharging is tied to that of the internal combustion (IC) engine. Gottlieb Daimler and Rudolf Diesel investigated increasing the power output and reducing the fuel consumption of IC engines by compressing the combustion air, in the late 1800 s. In 1925, Swiss engineer Alfred Büchi was the first to successfully exploit exhaust gas turbocharging, and achieved a power increase of more than 40%. This was the beginning of the introduction of turbocharging into the automotive industry. The real breakthrough in passenger car turbocharging came in 1978 with the introduction of the first turbocharged diesel engine passenger car in the Mercedes-Benz 300 SD, which was followed by the VW Golf Turbodiesel in 1981. By means of the turbocharger, efficiency of the diesel engine and driveability were improved, and the emissions significantly reduced (Watson and Janota 1982). Today, IC engines are turbocharged not just from the performance perspective but for reducing fuel consumption and, consequently, environmental pollution on account of lower carbon dioxide (CO2) emissions (Baines 2005). The cut-section of a typical Turbocharger is shown in Fig. 4.2. In the cut section is shown 1. Compressor housing, 2. Turbine housing, 3. Compressor wheel 4. Turbine wheel and the shaft. The air and exhaust flow directions are indicated as well. The turbocharger is different from a supercharger in its ability to drive a compressor using the exhaust gas from the engine which otherwise would be wasted. A supercharger gets its energy from engine itself and hence will take away a part of the engine energy. The energy from the exhaust comes from the temperature and

88

D. A. Subramani et al.

Fig. 4.2 Cut section of a typical turbocharger

pressure of the exhaust. This energy is transformed into shaft power by a turbine which is connected to the compressor with a shaft. The compressor pushes more air into the engine by compressing it. Typically, radial compressor and turbine are used although some have tried axial as well as mixed-flow versions. Figure 4.3 shows the direction of air flow through the compressor and engine exhaust gas flow through the turbine. Air enters axially and leaves nearly radially from the impeller and then tangentially from the compressor housing. In the turbine hot exhaust gas enters tangentially to the housing and nearly radially into the turbine wheel and leaves axially into the exhaust outlet. The increase in air flow to the engine allows more fuel to be burned which increases power as illustrated by the torque curves in Fig. 4.4. The advantages of turbocharging are shown in Fig. 4.5 which compares a 1.6 L turbocharged engine with a naturally aspirated engine of similar power and has a displacement of 2.8 L. Compared with a naturally aspirated engine of identical power output, the fuel consumption of a turbocharger engine is lower, since some of the exhaust energy, which otherwise would have been wasted contributes to the engine’s efficiency. Thus, turbocharging helps in “downsizing” the engine. Due to the lower volumetric displacement of the turbocharged engine, frictional and thermal losses are less. Therefore, the power-to-weight ratio, i.e. kilowatt (power output)/kilograms (engine weight), of a turbocharged engine is much better than that of the naturally aspirated engine. The turbocharged engine’s installation space

4

Introduction to Turbocharging …

89

Fig. 4.3 Direction of air flow in a turbocharger compressor and turbine

Fig. 4.4 Torque characteristics of a turbocharged engine compared with that of a non-turbocharged engine

requirement is also smaller than that of a naturally aspirated engine for the same power output. Another advantage of turbocharging is altitude compensation. A naturally aspirated (non-turbocharged) engine designed for best performance at sea level will not perform as well at higher elevations since air density decreases with altitude. This reduction in air density reduces the amount of air available for combustion. In order to maintain the desired air–fuel ratio for combustion of approximately 14.6–1 it is necessary to reduce the fuel as altitude increases, which in turn reduces the

90

D. A. Subramani et al.

Fig. 4.5 Comparison of a naturally aspirated engine with a turbocharged engine

Fig. 4.6 Altitude compensation due to turbocharging

power approximately at the rate of 3 percent per 1,000 feet of altitude for a naturally aspirated engine (Fig. 4.6). A turbocharged engine achieves nearly the same power at altitude as sea level because the inlet air to the engine is compressed. Even though the inlet air pressure and density are reduced, the pressure at turbine exhaust is also reduced, which results in a greater pressure difference between turbine inlet and exhaust. The mass flow of air supplied to the engine at altitude will therefore be nearly the same as sea level allowing the same amount of fuel to be burned.

4

Introduction to Turbocharging …

91

This is similar to why a jet engine is able to fly at higher altitudes than a propeller driven engine (Cumpsty 2003). The other benefits come from reduction in emissions especially NOx and particulates and the improved fuel economy due to engine downsizing. A turbocharged engine is also expected to cost less (weighs less) compared to an equivalent naturally aspirated engine. Since the outer boundary of the turbocharged engine is smaller the sound radiating surface is reduced resulting in lower noise than a naturally aspirated engine with identical output. Turbocharged engines find applications ranging from aerospace propulsion to marine engines, while the volume of applications lie in the passenger vehicles and commercial vehicles. These applications represent a wide variety of requirements that can include several factors such as high low-speed torque, improved transient response, good fuel economy, and low emissions. The compressed air discharged from the turbocharger has a higher temperature which limits the density of the airflow. By cooling the compressed air further increasing is possible, which is achieved by a heat exchanger that is also known as an “after-cooler”. The flow will however incur a pressure drop in the after-cooler as it traverses through the heat exchanger (Fig. 4.7). Turbochargers come in different types (Fig. 4.8). The simplest is a fixed-geometry turbo charger. In contrast we have a variable geometry turbine-based turbocharger which as the name indicates has a variable area nozzle in-front of the radial turbine. The efficiency of a turbine wheel is optimized for a particular speed and flow rate and hence the geometry of the radial turbine (particularly inflow angle of the turbine blades) is dictated by these conditions, and hence the turbine will be less than optimal at other flow rates and speeds. The variable nozzle allows effective utilization of the exhaust gas over a range of flow rates and turbine speeds. Some turbochargers include what is known as a “waste-gate”, which essentially bypasses the exhaust gas flow beyond a certain flow. With this arrangement, the turbine can be designed for smaller flow and hence flow above a certain set rate is bypassed into the exhaust.

Fig. 4.7 Turbocharger and engine configurations without and with a charge air cooler

92

D. A. Subramani et al.

Fig. 4.8 a Fixed geometry, b waste gated, c variable geometry turbocharger

Fig. 4.9 Typical compressor performance (map)

Performance of a typical compressor (called a compressor map) is shown in Fig. 4.9. Key characteristics of the map are: 1. Speed lines (red) which show the pressure rise vs flow rate for different turbocharger speeds 2. Boundary limit (green) on the left-hand side (lower flow) for each speed line is the surge limit—below this limit, the flow in the compressor is dominated by reverse or recirculating structures. Presence of surge also introduces noise and is detrimental to the flow stability and mechanical durability of the system. Surge must therefore be avoided. 3. On the right-hand side (higher flow) is the “choke limit”—for any particular turbo speed there is a highest flow possible called the choke flow, beyond which

4

Introduction to Turbocharging …

93

Fig. 4.10 Typical turbine performance (map)

the flow inside the impeller is marked by the presence of shock waves (Zucker and Biblarz 2002), which introduces further losses and hence reduces the efficiency 4. The highest turbo speed is determined by the structural considerations. Compressors are typically made of Aluminium alloys and in some cases Titanium. High turbo speeds introduce centrifugal stresses in the wheel, which ought to be limited to the material yield limits. 5. On the map constant efficiency lines are overlaid, forming closed regions called efficiency islands. The engine flow requirements (operating line) when plotted on the compressor map should lie within these limits. The turbine performance (Fig. 4.10) is indicated by corrected mass flow vs expansion ratio lines. Here again we have choke and speed limits. It is customary to show the efficiency vs expansion ratio curves alongside (bottom part). The efficiency as measured from the gas stand tests which have continuous flow of exhaust gas is lower at the lower flow rates compared to that of the ones obtained from engine tests. This is due to the effect of the pulsating engine exhaust flow. The mass flow or volume flow shown in the performance maps are normalized/corrected to some inlet conditions in order to avoid variability arising from test inlet conditions.

4.4

Overview of Turbocharger Matching

The turbocharger and engine are interdependent systems and in order for the overall system to be efficient it is important to design (or select) the appropriate turbocharger size and type. The turbocharger compressor delivers air to the engine and the engine uses the air to burn fuel. Fuel is added in proportion to the air available. The energy in the engine exhaust is dependent upon the fuel burnt and the “waste” energy drives the turbocharger turbine. The turbocharger turbine drives the

94

D. A. Subramani et al.

compressor. The turbocharger and engine thus reach a mutually balanced condition of energy and flow. This balancing of energy and flow is known as turbocharger/engine matching (Nguyen-Schafer 2015). A best-matched turbocharger will give the best fuel economy and engine/vehicle performance. During matching the safe operating limits of the turbocharger considering the altitude at which the application will be used is ensured. Matching is in general done using numerical prediction tools as opposed to expensive engine tests. The turbocharger compressor must match engine air flow requirements—this in general means optimal efficiency along the engine operating curve. Different applications require the optimization to be carried out at different speeds, for example, a passenger car will require high efficiency at lower speed and part load while a commercial truck will require higher efficiency in the mid-speed range. This will also be dependent on the market where the application will be used. The turbine size is then selected to give adequate power to drive the compressor. Higher size will mean increased efficiency but also increased inertia which will reduce the transient performance and hence the just right size is selected. Figure 4.11 shows a schematic of the compressor selection process illustrating an incorrect and a correct selection. In order to match the right turbocharger certain engine performance parameters are needed such as the following: Parameter

Unit

Engine power Engine torque BMEP brake mean effective pressure Fuel input BSFC brake specific fuel consumption BSAC brake specific air consumption Volumetric efficiency Air/fuel ratio Charge cooler effectiveness

kW Nm MPa mm3/stroke g/kW-h kg/kW-h % – %

Fig. 4.11 Turbocharger compressor selection (left—incorrect, right—correct)

4

Introduction to Turbocharging …

95

The basic principles of behind matching prediction are the following: 1. Compressor-to-turbine energy balance: Work done in compressor equals work done in turbine minus mechanical losses 2. Compressor-to-turbine flow balance: Flow through turbine equals flow through compressor plus fuel flow 3. Compressor-to-turbine speed balance: Speed of compressor and turbine wheels must be equal, since they are connected by a shaft.

4.5

Matching Process in Detail

Step 1: Collect engine data 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Engine displacement Rated speed Speed at peak torque Maximum boost pressure Standard conditions of pressure and temperature Power target at rated speed BSFC target at rated speed and peak torque A/F at cleaner pressure loss at rated speed Muffler restriction at rated speed Engine volumetric efficiency at rated and peak torque Engine ΔT vs A/F ratio

Step 2: Calculate the pressure and temperature at the engine intake manifold, P1E and T1E , at the desired (Fig. 4.12) conditions.

Fig. 4.12 Turbocharger without and with an after-cooler showing the locations

96

D. A. Subramani et al.

1. To begin, the inlet conditions to the turbocharger must be determined which come from the following equation: P1C ¼ Pamb  DPair filter

ð4:1Þ

where, P1C is the pressure at the inlet to the compressor, Pamb is the ambient pressure at the inlet to the air cleaner, and DPair filter is the pressure drop across the air filter Typically, a pressure drop, (ΔP) of 3.5 kPa is used for the air cleaner. 2. Next, the inlet conditions at the engine intake manifold are calculated. For a turbocharged and non after-cooled engine, the inlet conditions are simply: P1E ¼ P2C

ð4:2Þ

where: P1E is the intake manifold pressure and represents the boost pressure required by the customer and P2C is the pressure at the turbocharger compressor discharge. T1E is equal to T2C which equals the compressor outlet temperature obtained from assuming a certain efficiency (say 70%). For a turbocharged and after-cooled engine the following differences apply: T1E ¼ T2C  eðT2C  Tamb Þ

ð4:3Þ

where Tamb ¼ temperature of cooling air at inlet to charge air cooler; e ¼ effectiveness of charge air cooler ðtypical value is 80%Þ P1E ¼ P2C  DPaftercooler

ð4:4Þ

DPaftercooler ¼ intercooler pressure drop ðtypical value is 14 kPaÞ Step 3: Continuing with step 2 the calculation of engine airflow will be as follows (Baines 2005): m_ E ¼ C

P1E N ðVd Þ gvol ; T1E 2

ð4:5Þ

4

Introduction to Turbocharging …

97

where • • • • • • •

m_ E ¼ engine airflow, kg/sec: C ¼ 5:81  105 is a function of Rðgas constant for air Þ P1E is engine intake manifold pressure in kPa T1E is intake manifold temperature in degrees Kelvin. Vd is displacement of engine in litre. N is the number of revolutions per minute for a 4-stroke engine. gvol is the engine volumetric efficiency in percentage.

Using this equation, engine airflow is calculated—at least for three points including rated speed. These steps are illustrated in the examples below. Example 1 Calculate the air flow for a 1.8 L naturally aspirated engine running at 3500 rpm where the ambient pressure is 1 bar, the air cleaner pressure drop is 357.5 mm H2O, and the ambient temperature is 25 °C. Take volumetric efficiency to be 90%. Answer: • • • •

Vd ¼ 1:8 L Þð9:8Þ DPair filter ¼ ð357:5 ¼ 3:5kPa 1000 P1E ¼ Pamb  DPair filter ¼ 100  3:5 ¼ 96:5 kPa 3500 m_ E ¼ 5:81  105 ð2596:5 þ 273Þ ð1:8Þ 2 ð0:9Þ ¼ 0:0533 kg/s

Example 2 For the engine given above calculate the mass flow, compressor pressure ratio and compressor outlet temperature for a boost requirement of 150 kPa. Also, calculate mass flow and pressure ratio for 70 and 40% of rated speed (3500 rpm), where the boost requirements are 130 kPa and 40 kPa respectively. Assume compressor efficiencies of 70%, 70% and 65% at 100%, 70% and 40% of rated speeds. Answers: 1. At rated speed of 3500 rpm For the turbocharged engine the manifold pressure is given by the boost P1E ¼ Pamb þ Boost ¼ 100 þ 150 ¼ 250 kpa Compressor total-total efficiency given by gc;tt

c1 ðP2C =P1C Þð c Þ 1 ¼ ðT2C =T1C Þ  1

Will be used to obtain compressor outlet total temperature T2C Here for an engine without a charge air cooler P2C ¼ P1E ¼ 250 kPa and

ð4:6Þ

98

D. A. Subramani et al.

P1C ¼ Pamb  DPair filter ¼ 100  3:5 ¼ 96:5 kPa Therefore c1 1:41 ðP2C =P1C Þð c Þ 1 ð250=96:5Þð 1:4 Þ 1 ¼ 1:447 ðT2C =T1C Þ ¼ 1 þ ¼ 1þ gc;tt 0:70

Or T1E ¼ T2C ¼ ð1:447Þð298Þ ¼ 431:1K ¼ 158:1  C And m_ E ¼ 5:81  105

250 3500 ð1:8Þ ð0:9Þ ¼ 0:0955 kg/s ð431:1Þ 2

2. At 70% rated speed (2450 RPM) similarly P1E ¼ Pamb þ Boost ¼ 100 þ 130 ¼ 230 kpa ðT2C =T1C Þ ¼ 1 þ

ð230=96:5Þð 0:70

1:41 1:4

Þ 1

¼ 1:402; and

T1E ¼ ð1:402Þð298Þ ¼ 417:9 K ¼ 144:9  C m_ E ¼ 5:81  105

230 2450 ð1:8Þ ð0:9Þ ¼ 0:0635 kg/s ð417:9Þ 2

3. And at 40% rated speed (1400 RPM) we have P1E ¼ Pamb þ Boost ¼ 100 þ 40 ¼ 140 kpa ð140=96:5Þð ðT2C =T1C Þ ¼ 1 þ 0:65

1:41 1:4

Þ 1

¼ 1:173; and

T1E ¼ ð1:173Þð298Þ ¼ 349:4 K ¼ 76:4  C m_ E ¼ 5:81  105

140 1400 ð1:8Þ ð0:9Þ ¼ 0:0264 kg/s ð349:4Þ 2

With initial calculations of air flow requirements completed, step 3 is to select the compressor map. Before airflow data can be plotted on the compressor map it must be corrected to the map conditions of temperature and pressure.

4

Introduction to Turbocharging …

99

Air flow is corrected to map conditions using the equation shown here where: m_ cor

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi T1C =Tref ¼ m_ E P1C =Pref

ð4:7Þ

where m_ cor is corrected compressor flow in kg/s. Here Tref and Pref are reference values of Temperature and Pressure respectively (298 K and 100 kPa here). After correcting the engine flow to the compressor map conditions, it is then plotted on the map as is shown in Fig. 4.13. These points include rated speed, 70% rated speed, and 40% rated speed. With the data plotted on the map, the points can be compared against the matching targets. These targets usually include good efficiency at all operating conditions and a specified value of say 70% at rated speed or peak torque. There should be sufficient altitude margin. That is, the maximum sea level speed should be such that the increase in speed that comes with altitude can be accommodated within the map or customer supplied target speed. Operating too close to surge line may result in instability and noise. It could also lead to turbocharger failure. Therefor adequate surge margin is needed. If all of these targets are not met, then a smaller or larger trim should be selected or another compressor map should be used. Step 4: Once a compressor has been selected the next step is to find a turbine to drive the compressor. Turbine selection is based on the requirement that the turbine power be equal to the power required to drive the compressor. This is represented by the equations for turbine power and compressor power. PC ¼ m_ air Cpair ðT02  T01 Þ

ð4:8Þ

And from compressor efficiency we have Þ  ðc1 c

gC ¼

P02 P01

1

ðT02 =T01 Þ  1

ð4:9Þ

Therefore, from Eqs. 4.8 and 4.9 m_ air Cp;air T01 PC ¼ gC And likewise, for the turbine we have

" ðc1Þ # P02 c 1 P01

ð4:10Þ

100

D. A. Subramani et al.

Fig. 4.13 Compressor map with the required flow and pressure plotted

" PT ¼ m_ exhaust Cp;exhaust T03



P4 1 P03

Þ# ðk1 k

ð4:11Þ

gT

Applying the constraints based on conservation of mass and energy Mass flow of air þ fuel mass flow ¼ engine exhaust flow Compressor Power þ mechanical losses ¼ Turbine power Which is typically written as PC ¼ PT gmech

ð4:12Þ

i.e. m_ air Cp;air T01 gC

"

P02 P01

Þ ðc1 c

"

# 1 ¼ gT gmech m_ exhaust Cp;exhaust T03



P4 1 P03

Þ# ðk1 k

ð4:13Þ

Turbine expansion ratio pT can be obtained with the following result, "

ð1=pT Þ

ðk1Þ k

m_ air Cp;air T01 ¼ 1 gC gT gmech m_ exhaust Cp;exhaust T03

( ðc1Þ )# P02 c 1 P01

ð4:14Þ

Before calculating the turbine expansion ratio, it is first necessary to determine the turbine inlet gas temperature and turbine efficiency. Turbine inlet temperature,

4

Introduction to Turbocharging …

101

T1T can be obtained either from DTE curves for the engine or from previous database. For turbine efficiency, gT values of apparent efficiency from engine data should be used, especially at expansion ratios less than 1.6, engine pulse effect is very high. Once T1T and gT have been obtained, the expansion ratio can now be calculated using the equation described above. A calculation of turbine flow is now required. From engine data on actual air fuel ratio mass flow of fuel can be obtained and hence the exhaust mass flow. Mass flow of air + fuel mass flow = engine exhaust flow For the three operating points selected (100, 70, 40 of rated speed) we have the following mass flow parameter and expansion ratios. 100% speed Input: gC ¼ 0:7; gT ¼ 0:6; gmech ¼ 0:9; exhaust gas specific heat ratioðk ¼ 1:36Þ AFR ¼ 23; T03 ¼ 873; Output: Expansion Ratio ðpT Þ ¼ 3:03; m_ exhaust ¼ 0:0997 kgs ; pffiffiffiffiffiffiffiffiffiffiffiffiffi MFP ¼ 0:971kg=s K=bar 70% speed Input: gC ¼ 0:7; gT ¼ 0:7; gmech ¼ 0:9; exhaust gas specific heat ratioðk ¼ 1:36Þ AFR ¼ 21; T03 ¼ 720; Output: Expansion RatioðpT Þ ¼ 2:78; m_ exhaust ¼ 0:0665 pffiffiffiffiffiffiffiffiffiffiffiffiffi MFP ¼ 0:615 kg=s K=bar

kg s ;

40% speed Input: gC ¼ 0:65; gT ¼ 0:85; gmech ¼ 0:9; exhaust gas specific heat ratioðk ¼ 1:36Þ AFR ¼ 19; T03 ¼ 600;

102

D. A. Subramani et al.

Fig. 4.14 Turbine map with the required flow and pressure plotted

Output: Expansion RatioðpT Þ ¼ 1:49; m_ exhaust ¼ 0:0278 kgs ; pffiffiffiffiffiffiffiffiffiffiffiffiffi MFP ¼ 0:457 kg=s K=bar Turbine flow is corrected differently and the corrected flow (called mass flow parameter) is expressed as MFP ¼ mexhaust

  kg pffiffiffiffiffiffiffiffiffiffiffiffiffiffi T03 ðK Þ=P03 ðbar Þ s

ð4:15Þ

The calculated mass flow parameter (MFP) and expansion ratio for various speeds are plotted on the turbine map (Fig. 4.14). If the calculated points fit, then the match has been achieved. If the points do not match, then either a smaller or larger A/R turbine housing must be used. If there is still no match, then another turbine must be selected. In this case the 40% and 70% speed points lie nicely on the engine map. However, the 100% speed point is outside, indicating a potential to include waste gated turbine.

4.5.1

Compressor Features

Typically, a compressor stage is developed in such a manner that it can be used over a range of flow and pressure ratios by making minor changes to the geometry by machining. Following are some of the key compressor wheel and housing features that control the performance (Fig. 4.15).

4

Introduction to Turbocharging …

103

Fig. 4.15 Typical parameters that influence compressor performance

• Flow and speed are a function of the wheel tip diameter • Flow is also a function of the wheel trim, which is defined as the inducer diameter divided by the tip diameter and multiplying by 100. Larger trim will provide higher flow for a given wheel diameter. • Diffuser width and height affect compressor performance. • Volute is characterised by area (A) at a section near outlet and its radius (R). Compressor efficiency is also a function of A/R. • Mechanical considerations for the compressor include bore stress, blade stress and blade frequency.

4.5.2

Turbine Features

Similarly, turbine performance depends on the following key geometry features (Fig. 4.16): • turbine wheel tip diameter affects the flow and speed • exducer diameter controls the flow • The trim which is calculated by dividing the exducer diameter by the tip diameter and multiplying by 100.

104

D. A. Subramani et al.

Fig. 4.16 Typical parameters that influence turbine performance

– Turbine flow is a function of trim, but in the case of turbine is it also a function of A/R. Turbine wheel trim as well as the housing A/R can be changed to increase or decrease the flow capacity of the turbine. – A/R also has an effect on turbine efficiency and there is an optimum for a wheel size and trim. • Mechanical limits for the turbine are based on hub stress, blade stress, blade frequency and temperature.

4.5.3

Review of Matching

The main objective of turbocharger matching is to select the best turbocharger and engine combination so that each work together to provide best performance. For this, it is important to select a compressor that will provide adequate air for combustion and meet the following requirements. The compressor selected must have good efficiency at low speed, rated speed, and peak torque. It must have good altitude margin, that is sufficient speed margin to allow operation at altitude, and it must have adequate surge margin to prevent operation in areas of compressor instability. Then a turbine must be selected to drive the compressor. The turbine must have a good balance between high power at low speed and high efficiency at rated power.

4

Introduction to Turbocharging …

4.5.4

105

Operating Limits

Once a compressor and turbine have been selected that meet the airflow requirements of the match the mechanical limits of the turbocharger are verified. All mechanical limits must be met with altitude margin to assure that operation at the targeted altitude can be achieved. Mechanical limits are influenced by: • The maximum shaft speed allowed by the bearing system (by shaft motion tests to be described later) • The maximum compressor speed is the lowest of either the 4th order vibration frequency or the stress speed limit. The stress speed limit is dependent on the highest stress in the wheel which can be in the bore, blade, or back disk. • The maximum turbine speed is the 5th order vibration frequency or the hub, blade, or back disk stress. • The maximum expected turbine inlet temperature. Material selection is based on this. There are several material options available—ferrous alloys containing Nickel, Chromium and Molybdenum capable of withstanding high temperatures (up to 900 °C for Diesel and 1050 °C for Gasoline engines). • Bearing housing cooling, essential on gasoline applications, is also beneficial on some higher temperature diesel applications. • Check sealing capability: operation within the area of positive seal ΔP will avoid oil leakage. • Verify containment capability: this ensures that in the event of a blade failure the broken pieces are contained within the housing and does not fall out on the road. To reiterate the objectives of matching, there are specific requirements, which must be met. The best match meets the following requirements. – – – – – – – –

Good low speed (low flow) torque Good rated speed brake specific fuel consumption Low smoke, emissions Good drivability Does not exceed shaft speed or blade frequency limits at altitude, and Does not exceed the temperature limit at altitude. Altitude margin should ensure that shaft speed limits are not exceeded at altitude Mechanical requirements such as stresses, blade frequency, and temperature limits are met for all conditions. In addition, wheel burst, containment and sealing requirements must be satisfied.

106

4.5.5

D. A. Subramani et al.

Review of the Matching Process

Figure 4.17 shows an overview of the steps involved in matching that will be discussed in more detail on the following slides. • To get started, step 1 is to collect data from the customer. • After collecting the data, engine air flow is calculated at the required conditions. • Next, step 2 is to correct the engine air flow to the compressor map conditions and plot on the compressor map. • Step 3 is the turbine selection. Here the required turbine is determined based on a power balance calculation. The calculated match point is used to select the best turbine with the help of the turbine maps. • Step 4 is the final step in the matching process—where compliance to mechanical limits are ensured.

4.6 4.6.1

Turbocharger Aerodynamics Basic Principles

The radial compressor and turbine that are present in a turbocharger come under the category of turbomachines, in general. A turbomachine is a device in which energy is transferred to or from a continuously flowing fluid by the dynamic action of one or more of the moving blade rows (Dixon 1998). Turbomachines are classified according to the geometry of the flow path: axial—if it is mainly parallel to the axis and radial if it is perpendicular to the axis of rotation. When the main flow has

Fig. 4.17 Performance prediction overview

4

Introduction to Turbocharging …

107

Fig. 4.18 Axial, radial and mixed flow turbomachines and selection criterion (Lewis 1996)

components in the axial and radial direction at the outlet of the rotor it is a mixed flow machine. The choice of the type of turbomachine to be used in a turbocharger comes from all overall non-dimensional parameters called specific speed (Fig. 4.18).

4.6.2

Compressible Flow Basics

A review of basic principles of gas dynamics can be found in a highly readable book on this topic such as Zucker and Biblarz (2002). The fundamental governing equations are based on the principles of conservation of mass, momentum, and energy. Conservation of mass applied to a section in a fluid flow such as shown in Fig. 4.19 becomes m_ ¼ q1 A1 V1 ¼ q2 A2 V2

ð4:16Þ

Conservation of energy applied to a system of fixed mass is stated (First law of thermodynamics) as Q ¼ W þ DE

ð4:17Þ

where Q is the total head added, W is the work done by the system and DE is the net change in energy of the system. For a flowing fluid the same principle results in (for the example control volume shown in Fig. 4.19

108

D. A. Subramani et al.

Fig. 4.19 Control volume for applying conservation of mass and energy

h1 þ

V12 g V2 g þ z1 þ q ¼ h2 þ 2 þ z2 þ ws 2gc gc 2gc gc

ð4:18Þ

where h is the stagnation enthalpy defined as h = u + pv. (u: internal energy, v: specific volume, p: pressure) V, velocity of flow q, heat added and ws, the shaft work going out of the control volume. Example 3 Exhaust gas at 720 °C enters a turbocharger turbine at the rate of 0.08 kg/sec. The gas expands through a pressure ratio of 2.9 and leaves at 590 °C. Velocities entering and leaving are negligible and there is no heat transfer (Fig. 4.20). Calculate the power output of the turbine. T1 ¼ 720  C T2 ¼ 590  C _ ¼ 0:08 kg=sec m

V1  0 V2  0 Q¼0

Applying the steady flow energy equation: h1 þ

V12 g V2 g þ z1 þ q ¼ h2 þ 2 þ z2 þ ws 2gc gc 2gc gc

Neglecting the variation in potential, velocities and heat transfer we have h1 þ

V12 g V2 þ ðz1  z2 Þ þ q ¼ h2 þ 2 þ ws 2gc gc 2gc ws ¼ h1  h2 ¼ cp ðT1  T2 Þ ¼ ð1063Þð720  590Þ ¼ 138:2 kJ=kg _ s ¼ ð0:08Þð138:2Þ ¼ 11:1 kW Power ¼ mw

4

Introduction to Turbocharging …

109

Fig. 4.20 Illustration for example 3

4.6.3

The Concept of Stagnation Properties

The static properties are those that would be measured if you moved with the fluid. Stagnation state is a reference state defined as that thermodynamic state which would exist if the fluid were brought to zero velocity and zero potential. The stagnation state must be reached. 1. Without any energy exchange (Q = W = 0) and 2. Without losses. The stagnation process is isentropic! Consider a fluid that is flowing (Fig. 4.21) and has the static properties as shown at section (a). At location (b) the fluid has been brought to zero velocity and zero potential under the conditions of no energy exchange and losses. If we apply the energy equation to the control volume indicated for steady one-dimensional flow, we have

110

D. A. Subramani et al.

Fig. 4.21 Illustration of the stagnation concept

ha þ

Va2 g V2 g þ za þ q ¼ hb þ b þ zb þ ws 2gc gc 2gc gc

ð4:19Þ

V2

Which leads to ha þ 2gac þ ggc za ¼ hb . Introduction of the stagnation (or total) enthalpy makes it possible to write equation in a more compact form. For example, the one-dimensional steady-flow energy equation V12 g V2 g þ z1 þ q ¼ h2 þ 2 þ z2 þ ws 2gc gc 2gc gc

ð4:20Þ

ht1 þ q ¼ ht2 þ ws

ð4:21Þ

becomes

But condition (b) represents the stagnation state corresponding to the static state (a). Thus, we call hb the stagnation or total enthalpy corresponding to state (a) and designate it as hta: Thus hta ¼ ha þ

V22 g þ z2 2gc gc

ð4:22Þ

4

Introduction to Turbocharging …

111

Or for any state, we have in general, ht ¼ h þ

V2 g þ z 2gc gc

ð4:23Þ

When dealing with gases, potential changes are mostly negligible, and hence we write ht ¼ h þ

V2 2gc

ð4:24Þ

Velocity of sound (a) in a fluid (found based on applying conservation principles on a small control volume) is given by pffiffiffiffiffiffiffiffiffiffiffiffiffi ð4:25Þ a ¼ cgc RT where, c is the ratio of specific heat capacity of the fluid (= 1.4 for air and 1.33 for exhaust gas, approximately). gc is the proportionality constant in the Newton’s second law (gc = 1, numerically if SI units are used). R is the gas constant and T, the static temperature. A quantity of importance in compressible flow is the concept of Mach number which is defined as M¼

V a

ð4:26Þ

where V is the velocity of the flow and a, the velocity of sound in the medium. For a perfect gas the relation between stagnation and static quantities can be expressed in terms of Mach number ht ¼ hð1 þ

c1 2 2 M Þ

ð4:27Þ

Tt ¼ Tð1 þ

c1 2 2 M Þ

ð4:28Þ

 c1 2 c=ðc1Þ 2 M

ð4:29Þ

rffiffiffiffiffiffiffi cgc ¼ const m_ ¼ pAM RT

ð4:30Þ

 pt ¼ p 1 þ

4.6.4

Coordinate Frame and Frame of Reference

In turbomachinery, cylindrical coordinates are used extensively (Lewis 1996). In addition, turbomachinery flows are visualized in planes that are somewhat unique.

112

D. A. Subramani et al.

Fig. 4.22 Different views used in turbomachinery. a meridional view, b blade-to-blade view and c three-dimensional view showing a spanwise cut section, for a mixed flow machine (Lewis 1996)

Figure 4.22 shows the different views of a mixed flow compressor or pump— meridional view (top left), “blade to blade” view (right) and a span-wise (along the height of the blade) section (bottom left) showing the location at which the blade-to-blade view is taken. Meridional view shows the flow path and the blade and is obtained by taking a projection of the blade on a constant h plane (for example h ¼ 0Þ—this is not same as a side view although it appears to be so. Blade-to-blade view is obtained by taking the conical section (shown in bottom left) and un-wrapping it to obtain a flat two-dimensional surface. The x-axis of the blade-to-blade view is normalized with radius so that the same coordinate transformation could be used for all machine types, axial, radial or mixed flow. Not only that the flow inside a turbomachine is visualized in these views, the blades are designed in these views as well. The advantage is that the blades appear to be classical aerofoils in blade-to-blade view, irrespective of their complexity in three dimensions. In turbomachinery rotating frame of reference is routinely used to understand the flow. The advantage is that the flow appears to be steady in a frame rotating along with the rotor. This requires an understanding of the relationship between relative and absolute frame velocities. To illustrate this point consider the scene in Fig. 4.23. A passenger in the train is moving with a velocity ðvPT Þ of 2 m/s with reference to the moving train or the passengers seated in it. The train itself is moving with a velocity ðvTG Þ of 9 m/s with respect to an observer standing in the platform and in the same direction as that of the moving passenger. The velocity of the moving passenger with respect to the observer on the ground ðvPG Þ would therefore be

4

Introduction to Turbocharging …

113

Fig. 4.23 Relationship between relative and absolute velocities

Fig. 4.24 Velocity triangle at the exit of a centrifugal impeller

vPG ¼ vPT þ vTG ¼ 11m=s: Or in a general sense we could write vabsolute ¼ vrelative þ vframe

ð4:31Þ

In a centrifugal impeller if the observer is stationed on the impeller (as depicted in Fig. 4.24) the flow in the frame of the observer will try to follow the blade passage and hence the relative velocity W will follow the exit blade angle. The impeller itself has a linear velocity of U and therefore the absolute velocity will be V ¼ UþW Velocity triangle is used heavily in turbomachinery.

ð4:32Þ

114

4.6.5

D. A. Subramani et al.

Euler Equation

Applying Newton’s second law to an elemental stream tube (Fig. 4.25) going through a turbomachine we have Torque ¼ rate of change of angular momentum ¼ rate of change of moment of linear momentum Expressed analytically s¼

d ðmrCh Þ dt

_ 2 Ch2  r1 Ch1 Þ ¼ mðr

ð4:33Þ ð4:34Þ

Power input is then obtained by multiplying torque by the angular velocity i.e., _ 2 Ch2  r1 Ch1 Þ ¼ mðU _ 2 Ch2  U1 Ch1 Þ P ¼ W_ ¼ sX ¼ Xmðr

ð4:35Þ

where X is the angular velocity of the rotor and U is the blade speed. From energy equation neglecting heat transfer the adiabatic work done by the shaft is ht2  ht1 ¼ ðU2 Ch2  U1 Ch1 Þ

ð4:36Þ

Which is known as the Euler’s energy equation for a turbomachine. This equation is applicable to both compressors as well as turbines. In the Euler’s equation, Ch2 and Ch1 are the circumferential flow angles and can be related to the blade angles at the inlet and the outlet. Using the blade angles, one could obtain approximate values for velocities at the exit of the impeller. The velocities at the impeller exit in practice will deviate slightly from the blade angle and this is termed as slip. With the velocities thus determined one could compute the work input/output using the Euler equation.

Fig. 4.25 Elemental steam tube with velocities (Lewis 1996)

4

Introduction to Turbocharging …

115

Fig. 4.26 Cut section of a typical turbocharger compressor (BWTS 2014)

4.6.6

Compressor Aerodynamics

Figure 4.26 shows a cut-away of a turbocharger compressor which illustrates the relationship of the impeller or wheel to the other components included in a turbocharger compressor. Flow is shown entering from the left. Looking down from the top, the outflow is seen as exiting radially but from the front, the flow is seen as exiting in the tangential direction. Figure 4.27 provides additional detail about the features of a turbocharger compressor. On the left is an axial or front view through the housing. On the right is a side or meridional view section though the axis-of-rotation. Starting at the inlet (front view) is the wheel. The forward portion of the wheel is called the inducer and

Fig. 4.27 Front and Meridional views of a centrifugal compressor

116

D. A. Subramani et al.

the latter section is the impeller. The wheel has blades for turning the flow and imparting momentum to it. Partial blades which start at a distance from the leading of the full blade are called splitters. The blades typically lean backwards, opposite to the direction of wheel rotation. This is called back-sweep and is an important design parameter. After leaving the wheel, the flow passes through a diffusing passage. Some turbochargers have vanes in the diffuser. A diffuser with vanes has a higher peak efficiency but limited flow range and therefore is not commonly used in turbochargers. The flow from the diffuser is discharged into the housing flow passage called volute. The volute flow area progressively increases, clockwise from the tongue. The tongue separates the volute from the discharge section of the housing. An imaginary radial line tangent to the tongue, called the T-T section, divides the volute and the discharge section. In the side view, the flow passage in the wheel is formed by the hub contour and the shroud contour, and the blade meridional passage. A small clearance is required between the shroud contour and the wheel. The amount of clearance has a significant impact on the aerodynamic performance of the compressor. The commonly used performance parameter in turbomachinery is efficiency. It relates actual energy input to the energy input in an ideal situation. In a compressor efficiency is a measure of how effective, the compressor is in converting the energy to a pressure rise. Temperature rise inevitably accompanies pressure rise and the purpose of the compressor is to effectively pressurize the air, with the least increase in temperature. Efficiency is computed by dividing the ideal energy input required to increase the inlet total pressure to a specified exit total pressure by the actual energy input to accomplish the same task. A useful way to visualize efficiency is with a T-s diagram (Fig. 4.28) which shows lines of constant total pressure, in a temperature-entropy chart. Considering the impeller, the actual compression process proceeds from the inlet to the wheel exit. Stagnation or total pressure goes from P01 to P02 accompanied by a temperature rise from T01 to T02 at the wheel exit. In an ideal compression, there are no losses so the process proceeds along a vertical (no entropy generated) and crosses the P02 line at an ideal temperature of T02s. This temperature (T02s) is not known but the ideal energy input can be rewritten in terms of total-to-total pressure ratio. This pressure ratio can be measured and hence the total-to-total efficiency. gC ¼

gC ¼

Cp ðT02s  T01 Þ Ideal work input ¼ Actual work input Cp ðT02  T01 Þ Þ  ðc1 c P02 1 P01

ð4:37Þ

T02 =T01  1

An ideal compressor is one which compresses the air without excess temperature addition. It is important to understand the flow process in other components—in

4

Introduction to Turbocharging …

117

Fig. 4.28 T-s diagram for a compressor (Dixon 1998)

order to facilitate this a schematic with station labels is shown (Fig. 4.29). The total temperature remains same from station 0 to station 1, the impeller inlet. However, there would be a small pressure loss reflecting in a reduced total pressure from station 0 to station 1. From the impeller exit to the diffuser exit the static pressure increases but the diffusion process results in a total pressure drop resulting in P03 < P02. In the volute and further a small diffusion occurs with a rise in static pressure but with a drop in total pressure. The total temperature on the other hand remains constant from impeller exit to stage exit (station 5), for an adiabatic compression process.

4.6.7

Compressor Design

The next subject is compressor aerodynamic design. Aerodynamic design can be described as the science of arriving at the geometric parameters for the various components such as the intake, impeller, diffuser, volute and discharge satisfying the often-conflicting requirements of range, efficiency and size. Empirical correlations for losses and performance of each of the components (intake, impeller, diffuser, volute, discharge) are obtained from experiments focused on detailed measurements. These correlations are incorporated in some form of computer model and are used heavily during the initial sizing (preliminary design) of the compressor stage (Xu and Amano 2008). The compressor designer faces four major, highly interdependent challenges. The first three of these, are the demand for greater flow range, efficiency and

118

D. A. Subramani et al.

Fig. 4.29 Compressor stage meridional and axial views showing station numbers

4

Introduction to Turbocharging …

119

Fig. 4.30 Compressor design process

pressure ratio. The fourth is to accomplish the above three in minimum (package) space. The turbocharger must fit onto the engine and then into the compartment where space is typically at a premium. In some cases, aerodynamic performance is compromised to accommodate packaging constraints. Aerodynamic design involves two phases, Preliminary Design (PD) Phase and Detailed Design (DD) Phase (Fig. 4.30). The requirements comprise meeting a certain pressure ratio, flow and efficiency target at design point and off-design points. Design point might correspond to the engine rated speed. Performance targets will also be specified along the engine operating line. Mechanical specifications are also required such as life requirement for a certain operating cycle (duty cycle). The preliminary design phase utilizes one-dimensional analysis, where variations in flow properties are considered only in the direction of flow, not across the flow path. Therefore, at a given station the flow is represented by a single temperature, pressure and velocity value. The Preliminary Design (PD) analysis establishes the major dimensions of the components in the compressor and wheel inlet/exit blade angles. The geometry thus obtained from the PD phase is used in predicting the performance at the off-design flow and pressure ratio. Typically, there will be iterations between the design point and off-design points to establish a design that has an acceptable performance trade-off. The Detail Design Phase (DDP) is where the detailed profiles for the wheel, diffuser and housing are established. Today, 3-D computational fluid dynamics (CFD) programs are available for this to optimize these components. The DDP also includes mechanical optimization of the blade thicknesses to keep these to a minimum while satisfying life requirements and casting/machining requirements. Figure 4.31 is a schematic of the 1-D compressor prediction model used in the PD phase. This modelling is done using a computer program and the purpose is to optimize a large number of inter-related variables for the wheel, diffuser and housing. A number of the dimensions selected during the analysis is shown in Fig. 4.31. These are typically dimensions, that define the cross-sectional areas at each of the

120

D. A. Subramani et al.

Fig. 4.31 Schematic of preliminary design models

station locations described in Fig. 4.29. For example, for the wheel this would include the hub and tip radii at the blade leading edge, wheel diameter and the blade width at the exit, b2. For the diffuser, this would include diffuser exit diameter and the diffuser exit width, b3. Performance predictions are based on empirical and analytical methods specifically applicable to each component. For the wheel, the 1-D wheel design procedure makes use of the so-called jet-wake model which recognizes that the presence of two zones in the impeller flow: jet on one side of the flow passage and a stagnant wake on the other side. One of the correlations for the wheel involves the ratio of the Mach number at the inlet of the wheel to the Mach number of the jet at the wheel exit. Another correlation relates the direction of the exit flow to the blade in terms of what is called the slip factor. Diffuser modelling makes use of an empirical skin friction coefficient in the total pressure loss calculation. Losses in the housing are based on again empirical models. An important consideration in designing the wheel is the relationship between the incoming air and the leading edge of the blades. This relationship is illustrated by the velocity diagram at the impeller inlet (Fig. 4.32). The relative velocity is what a person would see who stands on the hub of the wheel, rotating around with the wheel and looking at the incoming flow. In many ways, this is the most important velocity vector because the angle between it and the blade is a major consideration in designing the blade shape.

4

Introduction to Turbocharging …

121

Fig. 4.32 Velocity diagram at the inlet to the impeller

The blade leading edge angle is a major design feature because it is one factor that influences the surge flow, and hence range, at low speeds and flows. Figure 4.33 shows the relative air velocity vector and the inlet blade angle. The angle that this vector makes with respect to the axial direction is called the inlet flow angle ðb1;air Þ. The angle blade (at inlet) makes relative to axial direction is called the blade angle at inlet ðb1;blade Þ. The difference between these angles is the incidence angle and is a major design factor. High positive incidence can lead to flow separation (right side picture of Fig. 4.33) and lead to compressor surge at lower speeds and flows.

Fig. 4.33 Influence of angle of incidence at inlet to impeller

122

D. A. Subramani et al.

Fig. 4.34 Velocity diagram at the impeller exit

A velocity diagram at the wheel exit (Fig. 4.34) shows the relationship between the wheel and the velocity vectors. In an ideal case where the flow follows strictly the blade geometry, the flow angle at the exit will be equal to the blade angle. In reality, however the flow deviates from the blade geometry and the difference is called “deviation or slip”. For instance, for a radial wheel if the flow exactly followed the blade then the relative velocity vector would also be entirely in the radial direction. Absolute velocity is what one would observe standing on the ground near the compressor. The absolute velocity can be resolved into its radial and tangential components. The tangential component of the absolute velocity is involved in the ubiquitous “Euler equation” of turbomachinery, which will be seen now. The Euler equation is essentially a statement of Newton’s second law applied to the wheel. It starts from Torque ¼ rate of change of angular momentum; and Power ðEnergy per unit timeÞ ¼ Torque  Angular velocity sðper unit mass flowÞ ¼ DðrVh Þ

ð4:38Þ

Energyðper unit mass flowÞ ¼ sx ¼ ðr2 Vh2  r1 Vh1 Þx ¼ U2 Vh2  U1 Vh1 ð4:39Þ

4

Introduction to Turbocharging …

123

Fig. 4.35 Impeller geometry parameters defined in the preliminary design phase

Since rx ¼ U; blade velocity S Slip factor is defined as Kslip ¼ V2h =U2 and ¼ 1; for ideal case

ð4:40Þ

) DEEuler ¼ U2 Vh2  U1 Vh1 equals U22 for the case of no slip and V1h ¼ 0

ð4:41Þ

It says that the energy added to the flow (per unit mass flow) is equal to the tip speed times the tangential component of the absolute velocity, at the wheel exit, minus the same quantity at the wheel inlet. Typically, the flow enters the wheel axially (tangential component of the inlet velocity is ‘zero’), the second portion of this equation drops out leaving energy input equal to U2, tip speed, times the tangential component of the absolute velocity ðV2h Þ. Kslip is a parameter used to quantify the ratio between V2h and tip speed. In effect, it quantifies slip or deviation. Consider an ideal case where there is no slip, i.e., Kslip equals 1. Substituting this into the Euler equation, it is seen that energy input is equal only to the tip speed squared. This is the maximum amount of energy a wheel could add to the flow. Figure 4.35 provides a summary of all the major geometric dimensions and features of a compressor wheel that are selected in the PD phase. Starting at the wheel inlet, the radius to the inducer tip, r1t, is selected in such a way to minimize the inlet relative velocity or Mach number, but providing enough area in the leading-edge portion of the wheel, or throat area, to provide acceptable choke flow capacity. The ratio of this radius, divided by the wheel exit radius, r2, is called the trim, often used to characterize the flow capacity of a wheel for its specified exit radius or diameter. Another key parameter is the exit axial dimension, b2, or ‘b’ width. This influences the area ratio between the wheel inlet and exit and is a key parameter in optimizing the wheel to perform best at a particular pressure ratio. Most turbocharger compressor wheels have full blades with leading edges near the front extreme of the wheel and splitter or partial blades, which have leading edges that are well behind the full blade leading edges. Splitter blades are used to add energy to the flow as are full blades but have the leading edges downstream of the full

124

D. A. Subramani et al.

Fig. 4.36 Diffuser geometry parameters defined in the preliminary design phase

blades to minimize blade blockage in the throat formed between the full blades which would reduce choke flow rate. The next flow path element is the diffuser (Fig. 4.36) which forms the flow path between the wheel exit and the housing volute. Static pressure increases in the diffuser due to increase in area. Too high a diffusion will result in flow separation in the diffuser. Vaneless diffusers are common in turbochargers due to the efficiency over a wide flow range. Diffuser width and height (defined by r3) together determine the level of diffusion. The purpose of the housing (Fig. 4.37) is to collect the swirling flow at the diffuser exit and to direct it to a common location which is the discharge duct. Clockwise from the tongue, the area in the volute is shown to be very small. Progressing further clockwise the area increases. This is necessary because as more and more mass flow collects in the volute from the diffuser, in order to maintain constant velocity in the volute, the area must increase. If the area distribution were not designed in this way, there would be a static pressure variation around the diffuser, which could both reduce flow range and efficiency. Geometric parameters specified in the preliminary design phase include the dynamic centre, R, at the T-T section is found using the ‘idealized’ flow equation shown at the right. The values for b3 and b3 are known from the diffuser analysis and A4 is set to achieve a small amount of velocity reduction between the diffuser exit and the T-T section. The area A4 divided by the dynamic centre R is the common housing parameter, A/R. The housing exit area, A5, and radius to the centre of this area, r5, are set to achieve a controlled amount of velocity reduction in the discharge section and to satisfy any attachment dimensions requested by the engine manufacturer.

4

Introduction to Turbocharging …

125

Fig. 4.37 Volute geometry parameters defined in the preliminary design phase

Fig. 4.38 Preliminary design model of a compressor stage in the computer software AxSTREAM (SoftInWay)

Fig. 4.39 Results from the preliminary design model of a typical compressor (Dhinagaran et al. 2017)

Computer programs such as AxSTREAM (SoftInway Inc) are available to do an optimum selection of various geometric parameters. A typical preliminary design model and the results for a compressor stage are shown in Figs. 4.38 and 4.39. The prediction accuracy of the 1-D models in this case appears to be rather good.

126

D. A. Subramani et al.

Fig. 4.40 CFD Mesh of a turbo compressor

Fig. 4.41 Comparison of test data (gas stand) with CFD results (Dhinagaran et al. 2017)

Once the preliminary design phase is completed, detailed blade shape is designed. At this stage three-dimensional Computational Fluid Dynamics (CFD) software such as ANSYS-CFX (Ansys Inc) or Fine/Turbo (Numeca Inc) are used. Blade design is carried out in a special software wherein the blade details such as the meridional shape, blade-to-blade angle distributions as well as initial thickness distributions are prescribed. Thickness distributions and fillet dimensions are optimized later in structural design stage. A typical CFD mesh for impeller along with the housing is shown in Fig. 4.40. With the availability of fast computing these calculations for obtaining a full map (which involves running the calculations for a number of operating points) takes only a few days, comparable with the time required for testing. Results in terms of computed pressure ratio and efficiency are compared with test data in Fig. 4.41. The prediction accuracy is very high. With the high fidelity of numerical prediction tools, the number of iterations in testing is drastically reduced.

4

Introduction to Turbocharging …

4.6.8

127

Turbine Aerodynamics

A turbine extracts energy from the flow whereas the compressor adds energy and it is in some sense a reverse of a compressor. There are three types of turbines that are used in turbochargers. These are defined according to the direction of inlet flow as seen in the meridional views (Fig. 4.42). In a radial turbine the flow enters radially, i.e., perpendicular to the axis-of-rotation while in the axial turbine the flow enters parallel to the axis-of rotation. In between these is the mixed-flow turbine where flow enters at an angle of less than 90°. to the axis of rotation. However, in all the three types the flow exits axially. For small sizes (lower flow rates) radial turbine is the preferred option while axial turbine is efficient for applications demanding larger flow rates. The function of the turbine housing (Fig. 4.43) is to collect the engine exhaust flow and distribute it uniformly around the turbine wheel. There are two general types of housings, (1) housings with no moving parts, also known as fixed geometry and, (2) housings with moving parts, i.e., variable geometry, including the waste-gate housing (Feneley et al. 2017). The housing types include the single scroll housing, the divided or twin-scroll housing and the variable geometry nozzle housing. For a turbine (Fig. 4.44) at the wheel inlet, the absolute velocity is to a large extent in the tangential direction or in the direction the wheel is rotating. The flow leaving the wheel (exit absolute velocity) is illustrated by the blue arrow. In the velocity diagram at the wheel exit (top view) the relative velocity is following the

Fig. 4.42 Different types of turbines used in Turbochargers

Fig. 4.43 Flow in a turbine housing with and without variable geometry nozzle

128

D. A. Subramani et al.

Fig. 4.44 Velocity diagrams at the inlet and outlet in a fixed geometry turbine

4

Introduction to Turbocharging …

129

blade angle while the absolute velocity is nearly axial. Based on the Euler equation, Energy is extracted from the flow when this change in flow direction occurs. Euler’s Equation for a turbine (for the wheel): DE ¼ Actual energy extracted per unit mass ¼ U5 V5h  U6 V6h

ð4:42Þ

DE ¼ Actual energy extracted per unit mass ¼ U5 V5h  U6 V6h Power by turbine ¼ turbine mass flow rate  DE The most important turbine performance parameter is the adiabatic efficiency, which means that there is no energy lost from the turbine due to heat transfer to its surroundings. The efficiency defined is the total-to-static adiabatic efficiency, gTs which is the actual energy extracted from the flow divided by the ideal energy extracted. gTS ¼ Turbine Total to Static Efficiency ¼

DE DEideal

Which from the T-S diagram (Fig. 4.45) can also be expressed as

gTS

  cp T05  T06;actual  ¼  cp T05  T6;ideal   cp T05  T06;actual   ¼ c1 c cp T05 1  1=pRTS

ð4:43Þ

Turbine power is mass flow times actual energy extracted and can be expressed in terms of turbine efficiency, mass flow rate, inlet total temperature and total-to-static pressure ratio as follows:

Fig. 4.45 Temperature-entropy diagram for a turbine

130

D. A. Subramani et al.

  c1 c Power by turbine ¼ gTS mturbine cp T05 1  1=pRTS gTM ¼ turbine mechanical efficiency ¼ gTS gmechanical   c1 c Power by compressor ¼ gTM mturbine cp T01 1  1=pRTS

ð4:44Þ ð4:45Þ ð4:46Þ

This is called the power balance equation because it equates the power it takes to drive the compressor with the power produced by the turbine, taking into account the losses in the bearings and seals. This is a very significant relationship. It illustrates, for example, that for a given compressor power requirement, as the turbine-mechanical efficiency goes up, the turbine pressure ratio must go down. This is why it is important to develop turbines with higher efficiency, i.e., to reduce the required turbine pressure ratio to drive the compressor and therefore, to reduce the back-pressure, or pumping work, required by the engine. The power balance equation is also the basis for measuring turbine-mechanical efficiency on the gas-stand or engine. The compressor power is measured, along with the other parameters on the right side of the equation, and turbine-mechanical efficiency, is then computed.

4.6.9

Turbine Design

Similar to the compressor the turbine aerodynamic design includes a Preliminary Design (PD) Phase and a Detail Design (DD) Phase. To start with, let us assume that a compressor that meets the specifications has been selected for which the map is available (Fig. 4.13 chosen). The steady state engine operating line (called the lug line) comprising several operating points is plotted on the compressor map (Fig. 4.13). We need to determine at each of these points the flow rate, the power to drive the compressor, and shaft speed. Once air-fuel ratio used in the engine is known turbine mass flow and inlet temperature can be determined. The turbine exit pressure which equals atmospheric pressure plus the exhaust system losses can be determined knowing the exhaust system losses (catalytic converter etc.). The turbine shaft speed is known since it is the same as for the compressor. Mechanical considerations such as life required for a duty cycle, wheel inertia and packaging constraints are also set. Aerodynamic Design The objective of aero design is to arrive at the over-all geometric features that produce the inlet/outlet velocity diagrams which provide the required turbine power and meets has the highest efficiency (or meets to minimize pressure ratio (engine back-pressure). Preliminary design uses 1-D computer codes to establish key geometric arameters.

4

Introduction to Turbocharging …

131

Fig. 4.46 Geometry parameters defined in the preliminary design stage of a turbine

The major over-all dimensions and features selected in the Preliminary Design phase are shown (Fig. 4.46) in a meridional view to the left and an axial view to the right. Starting at the inlet is the flange inlet area, then the T-T section area (A), and the radius R to the centre of the T-T section. The key parameter to decide in the turbine volute is the A/R ratio. The ratio A/R is used to increase or decrease the turbine flow, in addition to the trim. Turbine wheel inlet diameter, outlet dimeter, axial blade length, blade width at the inlet and hub diameter are all part of the dimensions set in the PD stage. Blade angles at inlet and outlet at various spanwise locations from hub are also part of the PD calculations. Initial blade angle distribution and thickness distribution are also set in PD phase. Final blade thickness and angle distributions, hub and shroud contour are all arrived at the detail design phase. As in the case of compressor the detail design is carried out using 3D CFD tools. A typical CFD mesh is shown in Fig. 4.47. Inputs that go into the flow solver include the geometry, the shaft speed, inlet total pressure and exit static pressure. Fig. 4.47 A CFD model of a turbine stage

132

D. A. Subramani et al.

Fig. 4.48 Typical results from CFD simulation

Flow rate will be determined by the CFD solver. In addition, the solver also computes local aerodynamic parameters such as pressure, temperature and velocities at each node. Results from a 3D CFD calculation is visualized in terms of vectors, contours and streamlines (Fig. 4.48). These details give insight into the flow field and help us to address any areas of high losses. The flow angle required in Euler equation could easily be computed from averaging the velocities shown. The results from the flow solver are then visualized in terms of different flow parameters (Fig. 4.48). The streamlines show the overall flow and identifies flow structures such as swirl. The contours of Mach number shown in the blade to blade view as well as an axial cut section shows areas of maximum and minimum Mach numbers. Mach numbers higher than unity cause “shocks” and associated losses. Velocity vector plot in a cross section shows the flow angles. By studying various combinations of such plots, the aerodynamic designer can identify areas where there are defects in the flow field that will reduce efficiency and then the designer

4

Introduction to Turbocharging …

133

Fig. 4.49 Relative contribution to polar moment of inertia of the rotating system

can make housing and/or wheel flow path changes to correct these before submitting the designs for fabrication. An important consideration that comes into play both in the aerodynamic design and the mechanical design of a wheel is the polar-moment-of inertia. This is because of the impact it has on turbocharger and ultimately, engine and vehicle transient response. The turbine wheel is a much more important factor than the compressor in this consideration because approximately 70% of the inertia for a rotating assembly is in the turbine wheel (Fig. 4.49). The final values of blade thickness and fillet radii and back disk dimensions are established in a mechanical analysis to be sure that stress and vibration criteria are met. After the design is completed the next step is to fabricate a few development parts for aerodynamic testing. Turbine wheels are in general made out of a high temperature nickel chromium alloy and is done using an investment casting process using wax patterns. These patterns are made in a tool with metal inserts that fill the region between the blades. Wax is injected between the inserts into cavities which model the blades themselves. The inserts are then slid radially outboard to release the wax pattern. The wax wheel patterns are mounted on a stem (like leaves on a stem) and are coated with a ceramic material which is sintered and hardened. Subsequently the wax is melted and, in its place, molten metal is poured resulting in a wheel which has the intended geometry. The next step after fabricating hardware is to run a gas-stand test. Figure 4.50 is a photo of a typical test set up in the lab. One major difference between a gas-stand test and a turbocharger test on an engine is that on a gas-stand, the turbine flow rate

134

D. A. Subramani et al.

Fig. 4.50 Typical gas stand setup

is provided by a laboratory compressor and thus is independent of the flow rate through the turbocharger compressor. The total turbine flow rate is the air flow from the laboratory compressor plus the fuel flow rate. In the gas stand turbine inlet total temperature, inlet static pressure, and exit static pressure are measured besides turbine shaft speed and mass flow rate. The power balance equation described above (Eq. 4.46) is used to compute the turbine-mechanical efficiency.

4

Introduction to Turbocharging …

4.6.10

135

Introduction to Noise Vibration and Harshness

Noise, vibration and harness (NVH) has risen to be one of the key parameters driving vehicle level comfort, and small turbochargers park themselves in the high frequency regime—1 to 10 kHz—readily influencing customer sound perception owing to human auditory sensitivity and low background masking noise levels (Muszynska 2005). Various types of aero and vibro-acoustic sources exist within an automotive turbocharger such as rotor unbalance related vibrations, aerodynamic pulsations, and oil film self-excited oscillations. These sources sympathetically interact with the system level environment i.e. connected ductwork, support brackets, and heat shields to radiate objectionable noise levels if proper sensitivity and transfer function analysis to take corrective actions is not done in the early stages of the program evolution. Clear understanding of rotordynamics and tribological relationships is essential in the product development phase of a turbocharger. Rotordynamics presents itself as an interdisciplinary work covering fluid and bearing interactions, and involves studying structural, aerodynamics and thermal transport mechanisms (Nguyen-Schafer 2015). Insights are provided on these aspects, along with influences of manufacturing processes. Dominant NVH sources and their propagation paths at both the turbocharger and vehicle system level channels are discussed. The challenges in product development on arriving at high performing turbocharger in terms of low airborne and structure-borne noises, good rotordynamics stability with minimal wear and friction are covered.

4.6.11

Turbocharger Rotordynamics

A systematic analysis of turbocharger rotor and supporting bearing starts with separate analysis of the structural and fluid elements. Later, fluid-structure interaction (FSI) is evaluated through non-linear analysis methods. The following sections discuss the linear analysis considering the rotor as rigid and flexible structures. Later, the flexible rotor is coupled with the fluid film journal bearing and the complete FSI is discussed along with key influential parameters and a sensitivity analysis done on an application case is presented. Figure 4.51 shows the rotor structure coupled to the radial and axial bearings. The fluid domain constitutes of the oil coming through a casing (not shown) into the radial bearing oil feed holes and to the axial thrust bearing.

136

D. A. Subramani et al.

Fig. 4.51 Turbocharger Rotor and Bearing System

4.6.12

Introduction to Radial Bearing

The exhaust gas driven turbine energy must transmit to the compressor stage with minimal losses through the bearing system. Among the two types of bearing systems shown in Fig. 4.52a Semi-Floating Ring Bearings (SFRB) and (b) Fully Floating Ring Bearings (FRB)—the latter that is suspended in the lubricant, by virtue of its rotational speed, brings down the circumferential speed gradient. The ring rotates in the range of 0.25–0.4 times the shaft speed, and as a result has a positive effect on lubricant drag. On the other hand, this bearing system now has two spinning oil films, and have to sustain the dual self-excited oil whirl vibrations in the inner and outer clearances that have a reciprocal influence on each other. A thorough understanding of rotordynamics of bearing systems is required, and related aspects will be reviewed in brief in the coming sections.

Fig. 4.52 a Semi-floating bearing (left), b fully floating bearing (right)

4

Introduction to Turbocharging …

4.6.13

137

Linear Rotordynamics

In linear rotordynamics, structural components like turbine wheel and shaft assembly, flinger sleeve, thrust ring, compressor wheel and shaft nut are considered. Oil film and journal bush are not included here. Lateral bending related vibration of the rotor assembly is a major focus area, and the influence of torsional vibration of the rotor system is assumed negligible. Brief discussion is presented on the influence of rigid and flexible behaviour of the rotor on vibration, self-centring due to unbalance phase, and finally gyroscopic effect of compressor and turbine disks on rotor Eigen frequencies are discussed.

4.6.14

Rotor System Eigen Frequencies

At lower rotational speed, rotor system behaves like a rigid system. During first critical speed, rotor has conical mode shape where compressor and turbine wheel ends deflect in opposite directions as shown in Fig. 4.53. Rotor system has cylindrical mode shape at second mode where compressor and turbine ends deflect in the same direction as shown in Fig. 4.54. Figure 4.55 shows the measurement done with the small rotor hung close to free-free condition with a rubber band, and the impact response is captured by a non-contact microphone. The 2D frequency plot is shown in Fig. 4.56 with the first and second peaks correlating well with the simulated modal results. Figure 4.57 shows the first and second bending mode shapes obtained from the numerical simulation.

Fig. 4.53 Conical mode

Fig. 4.54 Cylindrical mode

138

Fig. 4.55 Measurement setup

Fig. 4.56 Rigid rotor—measurement results

D. A. Subramani et al.

Introduction to Turbocharging …

4

139

Fig. 4.57 Bending mode shapes

4.6.15

Unbalance Location Versus Rotor Axis

Below first critical speed, the rotor force increases linearly with speed, and so does the deflection. The unbalance force faces away from the rotation axis, and is in-phase with the deflection as shown in Fig. 4.58. The deflection direction will be perpendicular to the unbalance force as it approaches critical speed. Balance is maintained between the rotor stiffness and inertial components. Damping force is the key to keep a check on amplitude that raises asymptotically at critical speeds as is seen in Fig. 4.59. At higher speeds greater than the critical speed, the unbalance force and deflection remain out of phase, and as a result, the deflection drops as shown in Fig. 4.60.

4.6.16

Unbalance Location Versus Deflection Locus

Very similar to the previous section observations, Fig. 4.61 shows the unbalance location (dark spot) in the polar coordinates close to the deflection direction

Deflection / mm

Shaft Spin

10000

Sub Critical Speed

20000

30000

40000

50000

Speed - rpm Unbalance

Fig. 4.58 Sub-critical speed

140

D. A. Subramani et al.

Shaft Spin

Deflection / mm

Critical

Critical Speed

0

20000

40000

60000

Speed / rpm

80000 100000

Unbalance

Fig. 4.59 Critical speed

Deflection / mm

Shaft Spin

Super Critical Speed

Super Critical

0

50000

100000

Speed / rpm

150000

Unbalance

Fig. 4.60 Super-critical speed

(Nguyen-Schafer 2015). As the rotor goes close to the critical speed, the deflection locus gains 90° phase. Further increase in speed results in the shaft deflection positioning itself 180° to unbalance, and as a result dragged closer to the unbalance radius, which is the self-centring phenomenon.

4

Introduction to Turbocharging …

141

Fig. 4.61 Shaft deflection locus

4.6.17

Gyroscopic Effect

In a turbocharger, the compressor and turbine wheels have much higher diameters compared to the shaft diameter, and hence their gyroscopic effects must be taken into consideration as the turbocharger speeds shoot up. The wheel’s polar moment of inertia (mass times squared radius) and rotor speed gives the gyroscopic effect as shown in Eq. 4.1. Based on the angular momentum theory, the gyroscopic forces are created by virtue of the tilt of the rotating shaft from its axis of rotation. A positive gyroscopic force causes a stiffening effect, and a negative force a softening effect. This phenomenon causes the eigen frequencies to shift with speeds as shown in Fig. 4.62.   :  :  MGyro ¼ Ip X hy i  Ip X hx J

4.6.18

ð4:47Þ

Forward and Backward Whirl

When the spinning direction of the shaft coincides with the whirling direction, the resulting motion is called forward whirl as shown in Fig. 4.63a. Turbocharger operation in forward whirls leads to an increase in stiffness due to the previously described gyroscopic effect. Hence, the critical speed of the rotor increases with respect to the rotor speed. Forward whirls can be usually excited by unbalance.

142

D. A. Subramani et al.

Fig. 4.62 Gyroscopic and whirl effect—Campbell diagram

(a)

Shaft Spin

Forward Whirl Self-rotation of the shaft and Rotation of the shaft at inside of the bearing are in same direction

(b)

Shaft Spin

Backward Whirl Self-rotation of the shaft and Rotation of the shaft at inside of the bearing are in opposite direction

Fig. 4.63 Forward and backward whirl

Similarly, when the shaft spinning direction is opposite to the whirling direction, the resulting motion is called backward whirl as shown Fig. 4.63b. Turbocharger operation in backward whirl leads to a loss in rotor stiffness due to the gyroscopic

4

Introduction to Turbocharging …

143

Fig. 4.64 Different fluid system

effect. Critical speeds of the rotor hence decrease with respect to the rotor speed. Backward whirls usually do not occur during turbocharger operation.

4.6.19

Fluid Bearings

A thin fluid film between two solid surfaces that are either stationary or moving support the turbocharger rotor loads. The load bearing capacity is derived from the reactive pressure force that is internally developed or supplied externally. Below are few types of fluid bearings and their pressure sources. • Hydrostatic Bearing Hydrostatic bearing has static fluid along the surface of the shaft and the journal where fluid will be supplied externally and the pressure is maintained with the help of external pump as shown in Fig. 4.64a. • Hydrodynamic Bearing In hydrodynamic bearings, the pressure is developed either by two surfaces moving closer or away causing a squeeze action, or by a converging fluid that develops between two rotating surfaces or one rotating and one stationary surface. This is shown in Fig. 4.64b. The thickness of these films is in the order of micrometres. The convergence creates pressures normal to the surfaces that are in contact, and leads to the reactionary forces that define the load bearing capacity of the film. • Hybrid Fluid Bearing Hybrid fluid bearings are a combination of both hydrostatic and hydrodynamic bearing. This is observed in a typical turbocharger. Pressurized oil is supplied to the bearing system continuously. In addition to that, shaft spins at very high speed, which develops the oil pressure further at the converging wedge as shown in Fig. 4.64c.

144

4.6.20

D. A. Subramani et al.

Non-linear Rotordynamics

The pressure rise described in the previous section varies significantly in the journal bearing clearances. Pressure distribution, hydrodynamic bearing forces and lubrication regimes are influenced by the rotor speed, geometries, lubricant quality, surrounding thermal conditions, surface roughness and mechanical alignment. The two types of journal bearing configurations shown in Fig. 4.52a, b are most commonly used bearing configurations, though variants exist in terms of an integrated axial and radial bearing. The SFRB has a squeeze film in the outer clearance between the bearing and the casing, and a spinning oil film in the inner clearance between the shaft and the bearing. On the other hand, the fully-floating configuration has two spinning films and presents itself as a complicated rotordynamics problem. The rotor shaft always has an eccentricity with respect to the bearing centre. This causes development of a convergent wedge for the incoming oil in the clearances between the solids. These spinning films by virtue of the residual oil flow (Nguyen-Schafer 2015) in the convergent wedge exhibit a self-excited oil whirling phenomenon. The circumferential velocity (Muszynska 2005) of this tangential reaction force—often termed as cross-coupled stiffness—decides the onset of instability and the corresponding amplitude. With the onset of this instability, the rotor attains equilibrium in a larger orbit with this follower force constantly pushing the rotor in the tangential direction as shown in Fig. 4.65. This once of instability generally varies from 0.1X to 0.5X of the rotation speeds, and ones this whirl frequency locks into a rotor eigen mode it is termed as oil whip. Certain amount of damping must be ensured to keep a check on this instability as it could end up going beyond the critical minimum oil film thickness requirement and cause the rotor to touch the bearing walls. Fig. 4.65 Stable orbit

4

Introduction to Turbocharging …

145

Fig. 4.66 Unstable orbit

Figure 4.66 shows the orbit plot of a journal bearing with some number of outbursts. Such bearings could end up with boundary contact with any internal or external triggers during transient operations.

4.6.21

Oil Whirl Behaviour

• Sub Synchronous Motion 1 (Sub1 oil Whirl): Sub1 is due to the oil whirl synchronization of the inner gap (oil film between journal bearing and shaft). This occurs at 200–500 Hz at lower turbocharger speed. This causes Conical Motion of the rotor (1st mode) as shown in Fig. 4.67. • Sub Synchronous Motion 2 (Sub2 oil Whirl): Sub2 is due to the oil whirl synchronization of the inner gap (oil film between journal bearing and shaft). This occurs at 600–1000 Hz higher turbocharger speed. This Causes Cylindrical Motion of the Rotor (2nd Mode) as shown in Fig. 4.67. • Sub Synchronous Motion 3 (Sub3 oil Whirl): Sub3 is due to the oil whirl synchronization of the outer gap (oil film between journal bearing and bearing housing). This self-excited vibration can occur together with sub2 in the inner oil film. Sub3 shows the conical movement of the shaft (1st Mode) with bending of the shaft as shown in Fig. 4.67. The “sub3” usually dominates.

146

D. A. Subramani et al.

Fig. 4.67 Sub-synchronous oil whirls

The oil whirling frequency (Nguyen-Schafer 2015) of the inner oil film is given by: xi ¼ ki ðX þ XR Þ ¼ ki Xð1 þ RSRÞ  Q0

ð4:48Þ

where, • k is the fluid circumferential average velocity ratio of the inner oil film. • RSR ¼ XR =X is the ring speed ratio (the ratio of bearing ring speed to the rotor ring speed) • Q′ is the axial oil flow From Eq. 4.48 it is clear that the RSR and axial flow components influence the oil whirl frequency and amplitude. Mechanisms to reduce the circumferential oil velocity like radial grooves and reduced clearances, and increasing the axial oil flow by means of spiral or axial slots tend to mitigate the oil whirl effect. It must be noted that these solutions could have negative effect on friction loss, and in case of fully floating bearings, a solution for whirl in the inner clearance can have a reverse effect on the whirl in the outer clearance. The RSR for the bearing is derived to be: RSR ffi

1 1þ

g0 ðT0 Þ L0 c1 gi ðTi Þ Li c2

 3 D0 Di

ð4:49Þ

4

Introduction to Turbocharging …

4.6.22

147

Most Influential Parameters

Key ratios that need to be given careful consideration during the development of bearings for an application to meet bearing friction, effective thermal transport, noise and rotor stability, are discussed below. They are the typical ratios of— • Do/Di—Bearing Diameter (Do) to Bearing Inner Diameter (Di)—larger the diameter ratio Do/Di at a given inner bearing diameter Di better the damping coefficient of the bearing acting on the rotor. This has a positive effect on noise. On the other hand, the RSR drops, friction increases, and bearing contact wear could occur due to increased stiffness in the outer clearance. • Lo/Li—Bearing Outer Width (Lo) to Bearing Inner Width (Li)—larger width ratio with higher Lo has similar effect as the diameter ratios but not in the same magnitude. Per Eqs. 4.1 and 4.2, whirl varies linearly with Lo but cubically with Do. • C1/C2—Bearing clearance (C2) to Inner Bearing Clearance (C1)—larger the ratio with reduced C2, better is the stability due to lower rotor response from the conical mode. It must be noted that all the above ratios have to be considered together, as they have a direct influence on the rotor behaviour. Also, the exact effect on a given rotor system should be evaluated on an application basis considering oil characteristics, operating speeds, and engine conditions.

4.6.23

Thermal Transport and Frictional Losses

A key function of the lubricant is to efficiently carry out the heat from the TC, and at the same time ensure minimal wear and frictional losses. Figure 4.68 shows the flow of thermal energy (San Andres 2012) for an SFRB at a shaft speed of 200 rpm. The turbine side of the rotor is subjected to temperatures as high as 900 °C, and the compressor side to 200 °C. The shaft conducts heat from the turbine side, and the inner film lubricant then carries away a major portion of it while convecting it to the bearing rings and to the outer film lubricant. Thermal analysis done on the bearing system shows that the contribution of frictional drag being at only 30% of the overall temperature rise in the oil films. This study has clearly shown the need to design adequate flow paths in the bearing stage to carry out the heat that is generated or conducted during the operation of a turbocharger. Figure 4.69 shows the overlaid plots of frictional power losses from three types of bearings under no axial force (Vanhaelst et al. 2016). As expected, the power loss levels increase with speeds, and the semi-floating journal bearing showed

148

D. A. Subramani et al.

Fig. 4.68 Thermal energy flow in the rotor bearing system

highest power loss throughout the speed range of evaluation. The fully-floating bearing, by virtue of lesser relative circumferential lubricant velocity, shows lesser shear stress, torque and hence lower power loss compared to the semi-floating bearings. On the other hand, with no inner gap lubricant between shaft and bearing, and with ball elements inducing lower friction, the ball bearings showed best low-end-friction behaviour compared to the journal bearing types.

4.6.24

Balancing

Turbine wheels and compressor wheels see slight shifts in their mass centres compared to the geometrical centres due to the production process. Furthermore, after the turbine wheel and shaft welding process, the polar mass inertia axis will end up at an angle with respect to the rotation axis as shown in Fig. 4.70. These aspects will lead to unbalance force that varies proportionally with squared rotational speed, and apart from generating noise could cause bearing wear and rubbing at extreme levels.

4

Introduction to Turbocharging …

149

Fig. 4.69 Bearing systems—power loss curves (Vanhaelst et al. 2016)

Fig. 4.70 Unbalance effect

4.6.25

Methods of Balancing the Rotor

Low-speed balancing Low-speed component balancing is used to balance compressor and turbine wheels at speeds of around 5000 rpm where they can be considered as rigid structures. It is generally carried out by balancing in two planes—the nose and back-face of the individual component as shown in Fig. 4.71. In both compressor and turbine wheel shaft assembly, material is removed at nose and back face as shown in Fig. 4.71.

150

D. A. Subramani et al.

Fig. 4.71 Component level balancing

Component Level to Core balancing Chaining The part level higher Incoming unbalance of shaft and wheel assembly which involves higher correction time and depth, lower productivity in balancer, higher rejection rate in part level and core balancing. Possible root causes for higher Incoming unbalance are casting variability and misalignment of shaft and wheel assembly. Here we have studied the misalignment of shaft and wheel assembly to reach considerable reduction in rejection with respect to TW incoming unbalance and as end result on part level balancing and core balancing. Results from the balancer for baseline and modified TW assembly are shown in Fig. 4.72.

4.6.26

High-Speed Core Balancing

High-speed core balancing is applicable to flexible rotors running at high speeds up to at least 75% of their maximum operational speed. The core unbalance results from mounting the compressor wheel on the turbine shaft and wheel and as well due to flexible deformations occurring at higher rotor speeds while passing through the critical modes. Unbalance correction planes are defined generally towards the compressor side for ease of access and material removal. Influence Coefficient Method Balancing (ICM) is the popular approach for detection and correction for high speeds and at two planes as shown in Fig. 4.73. Accelerometer fixed on the compressor housing measures the acceleration of the rotor system. Machine calculates the unbalance in the rotor from the measured acceleration by influence co-efficient method, and material is removed between

4

Introduction to Turbocharging …

151

Chaining

Component

Rejection Trend - Baseline

Rejection Trend - Modified Core

6226

Total Qty

8308

Total Qty RejecƟons

RejecƟons

848 14%

473 5.6%

Fig. 4.72 Component level to core balancing chaining

Fig. 4.73 High speed core balancing planes

compressor wheel blades and shaft nut of the compressor housing with two different milling cutters as shown in Fig. 4.74. Once the material is removed, again unbalance of the core assembly is measured. Final unbalance related amplitude and phase curves are shown in Fig. 4.75. In few NVH sensitive applications, sorting is also done using compressor discharge pulsations as is illustrated in Fig. 4.76. Here, 1X and 2X stand for first and second order pulsations.

152

Fig. 4.74 High speed balancing setup

Fig. 4.75 Final unbalance graph

Fig. 4.76 Compressor wheel discharger pulsations

D. A. Subramani et al.

4

Introduction to Turbocharging …

4.7

153

Introduction to Thrust (Axial) Bearing

A thrust bearing is designed to support high axial loads. In turbocharger, exhaust gas enters inside turbine housing. drives the turbine wheel and shaft assembly, and goes out to the atmosphere. Compressor wheel, assembled at another end of the shaft. compresses the atmosphere air and delivers the pressurized air to engine. Due to the pressure difference between compressor and turbine side, rotor system will try to move axially. This axial movement of the rotor system is controlled and supported by the thrust bearing. Complete setup of the rotor and bearing system is shown in Fig. 4.77. Nomenclature of the Thrust Bearing Nomenclature of the thrust bearing are listed below and the same is mentioned in Fig. 4.78. • • • • • • • • • • • • • •

Number of Pads (n) Inner Diameter of Journal (Di) Outer diameter (Do) Mean Diameter (Dm) Bearing Width (W) Thrust Inner dam width (di) Thrust Outer dam width (do) Overall length (l) or Overall angle (Ɵ) Tapered length (lwed) or Wedge angle (Ɵwed) Wedge height or Taper (cwed) Oil Groove length (loil) or oil groove angle (Ɵoil) Surface roughness of Taper and flat (Rz, thrust) Surface roughness of Thrust ring (Rz, ring) Surface roughness of Flinger sleeve (Rz, FS)

Fig. 4.77 Rotor and bearing system

154

D. A. Subramani et al.

Fig. 4.78 Nomenclature of axial bearing

Working Principle During the turbocharger operation, a thrust wedge surface is formed in the thrust plate and the thrust plate is fixed in the bearing housing. Thrust ring and flinger sleeve are assembled with the shaft in such a way that the thrust plate will be between thrust ring and flinger sleeve as shown in Fig. 4.79. Oil holes are formed in the thrust plate to supply the fresh oil to the thrust bearing wedge face from bearing oil inlet port. During the operating condition, shaft assembly will rotate the oil radially, and this will form hydrodynamic oil film between thrust face and thrust ring. Similarly on the other side, between the thrust plate and flinger sleeve. A reactionary bearing force is generated by the oil film acting upon the thrust ring due to the wedge and squeeze film actions as shown in Fig. 4.79. This bearing force, to ensure stable operation, supports the axial load in the system.

Fig. 4.79 Working principle of thrust bearing

4

Introduction to Turbocharging …

155

Table 4.2 Axial bearing sensitivity analysis

Influential Parameters of Thrust Bearings Various parameter ratios have been studied in the literature (Nguyen-Schafer 2015), and some of the key influential ones are listed below: Ratio W/L While designing thrust bearing if W/L ratio is larger, this leads to effective oil temperature raise which reduces the oil viscosity. Drop in oil viscosity lowers oil film thickness. On the other hand, smaller W/L ratio leads to improper pressure build up in the thrust bearing due to a short wedge. In addition to that, short wedges creates restriction to the radial flow which increases effective oil temperature. This leads to reduction in oil film thickness. Ratio lwed/L If the ratio of Lwed/L is too small it will affect the oil pressure build up due to short hydrodynamic wedge effect. This reduces the oil film thickness which will create addition friction loss in the axial bearing and increases effective oil temperature drastically. Meanwhile if the Lwed/L is too large, this will increase the oil pressure build up due to large wedge effect, which increases the oil flow rate. This pressure builds up drops suddenly at the flat region, and sudden drop in pressure will lead to poor load carrying capacity of the bearing system. Ratio lwed/Cwed Smaller lwed/Cwed will affect the oil pressure build up, which reduces the oil film thickness. This will increase the effective oil temperature. Similarly, if the lwed/Cwed is large it increases oil pressure and oil film thickness. This increases the oil flow rate to higher levels.

156

D. A. Subramani et al.

Mean Diameter Mean diameter is the average diameter of inner and outer diameter of the thrust bearing. Axial bearing friction power loss is equal to square mean velocity at the mean diameter of the system. Larger mean diameter will lead to higher friction loss in the bearing. If the mean diameter is too short this reduces the load carrying capacity of the bearing. Sensitivity analysis Table 4.2 shows the sensitivity analysis done on a thrust bearing system using commercially available tool. Such comprehensive analysis needs to be done during the early stages of product development to select the best suitable thrust bearing configuration.

4.8 4.8.1

Turbocharger NVH Introduction to Noise

A pressure fluctuation that propagates through a medium and has a unique impression with respect to audible frequency and amplitude when it reaches the receiver is called sound. The speed at which these sound waves propagate is readily influenced by the medium it travels in, like gas, liquid, or solid. Humans can only hear these pressure fluctuations when the source frequency lies between 20 and 20 kHz. Waves with frequencies above and below this audible frequency zone are called ultrasound and infrasound waves respectively. It is appropriate to define noise at this point as an unwanted sound. As shown in Fig. 4.80, there are two major classifications of noises -structure-borne and airborne- and difference lies in the transmission medium.

Fig. 4.80 Types of noise

4

Introduction to Turbocharging …

157

The unit of measurement for describing sound is decibel (“dB”), and the preferred analysis domain is frequency (Hertz) or number of cycles per second. In the frequency domain, acoustic emissions can be classified as tonal or narrow band noises with single to few frequency components, and broadband noises with multiple frequency components spread over hundreds or thousands of hertz. A classical approach adopted in NVH study is to discretize the problem into source, path and receiver categories. Each category needs to be thoroughly understood in terms frequency content, magnitude, and weightage of influence. This approach gives flexibility in arriving at an optimal solution during the product development process based on the design space available at a component, sub-system or system level.

4.8.2

Types of Turbocharger Noises

Turbocharger contains in it a multitude of noise emission possibilities that can be broadly classified into structure-borne journal bearing and unbalance noises, and airborne pulsation, blade passage, and compressor stage flow noises. In addition, noises that could come from waste gate rattle or flow passage disturbances also remain areas of annoyance. The schematic of the same is shown in Fig. 4.81. Predominantly turbine side transmits structure-borne and compressor side airborne noises.

Fig. 4.81 Turbocharger noise source generation and transmission

158

4.8.3

D. A. Subramani et al.

Structure Borne Noises

Unbalance noise After low and high-speed unbalance corrections, the residual unbalance in the rotor causes unwanted vibrational behaviour. This results in structure-borne noise from the turbocharger. Balancing of rotor assembly is done up to 70–75% of the maximum application speed using the high-speed core balancer. Thermal migration of the unbalance from initial unbalance level to a used core unbalance level will occur due to stress relieving at the turbine shaft weld zones and at the compressor bore regions. This along with the repeatability of the balancing machines have to be taken into account to arrive at a suitable mass production unbalance limit. Generally, vehicle in-cabinet noise is measured with such limit samples to ensure the production limit is at or below the threshold of audibility. The transfer path of the rotor unbalance is through the journal bearing oil film into the bearing housing. From the bearing housing, the vibration is transmitted to all the connecting components and their surface radiation efficiencies will result in noise. Large surface area vibrating structures will create the most noise. The general audible frequency range of this unbalance noise is between 500 and 4500 Hz tracking the turbocharger speeds, and at engine speeds ranging between 1200 and 3500 rpm. Figure 4.82 illustrates the unbalance noise complaint measured on a vehicle. The left side spectral map shows the microphone measured inside the vehicle cabin and the right side shows the vibration levels measured on oil drain. The vertical axis is

In-cabin Noise

Fig. 4.82 Spectral map showing unbalance noise and vibration

Vibration

4

Introduction to Turbocharging …

159

frequency (hertz), horizontal is time (seconds), and amplitude is the last axis. The turbocharger spools up to a maximum speed of 285,000 rpm, and then decelerates. Based on the spectrum results, the 1X vibration levels (1st order) correlate to the unbalance noise which is audible on the vehicle. Sub-synchronous noise: All journal bearing turbochargers show oil film related self-excited vibrations that do not follow linear dynamics. They generally occur between 0.1X to 0.5X, and are termed as Sub1, Sub2 and Sub3 and higher harmonics of the same. Among all these, only Sub2 that occurs as the inner oil whirl falling between 600 and 1100 Hz translates to annoying noise both objectively and subjectively in the vehicle interior. On the other hand, the outer oil whirl (Sub3) has low frequency signatures that get masked in the engine background noise and are not deemed as objectionable. The Fig. 4.83 illustrates the constant tone noise complaint measured on a vehicle. The left side spectral map shows the microphone measurement on the vehicle in cabin and the right side shows the vibration levels measured on oil drain. The vertical axis is frequency (hertz), horizontal is time (seconds), and amplitude is the last axis. The turbocharger spools up to a maximum speed of 150,000 rpm, and then decelerates. Based on the spectrum results, the constant tone vibration levels (1X) correlate to the no of turbocharger blades i.e. the above noise spectrum has 7th order blade pass noise which relate to the 7 splitter blades in the compressor wheel. Case Study: Investigation of System level BPF Noise Interaction and Propagation using advanced tool—An Acoustic Camera: This investigation is on engine bed to bed variations in acoustic emissions related to the blade passes noise frequency from the compressor side and intake ducting. An

4

Introduction to Turbocharging …

165

Fig. 4.91 Spectral map shows blade pass noise on vehicle

advanced tool in the form of acoustic camera has been used to identify the noise generation or emitting region corresponding to the audible frequency. Acoustic Camera Signal Interface Group’s (SIG) ACAM 100 is a microphone array for locating and analysing sound sources in real time. Combined with OptiNav’s BeamformX software it makes a powerful compact system. Figure 4.92 shows the acoustic camera with microphone arrays. It uses SIG ACAM 100 hardware combined with 40 digital MEMS microphones and five-megapixel optical camera in a planar, 40  40 cm array. High frequency blade pass noise with respect to 6th order of compressor wheel was clearly audible and objected by the customer on a particular engine bed. Initially, acoustic camera was kept facing the turbocharger compressor and intake ducting side because it relates to the compressor side noise source and then engine run-ups were performed. The audible region from the 6th order compressor blade pass noise is replayed on the beamforming software to identify the hot spots. A few hot spots were identified as shown in Fig. 4.93a. Hot spots pointed to possible radiation or leaks at the intake ducting pipe connections from air filter to the turbocharger inlet. To check if it were a leak, the intake ducting from the other engine bed was installed in the current engine bed, and the complaint noise levels and corresponding spatial hot spot zones came down to acceptable levels as shown in Fig. 4.93b. In summary, the discussions presented turbocharger NVH investigative challenges with system level interactions and a methodology to tackle them efficiently with the usage of an acoustic camera beyond the conventional accelerometers and microphone pickup sensors.

166

Fig. 4.92 Acoustic camera

Fig. 4.93 a Baseline measurements, b modified measurements

D. A. Subramani et al.

4

Introduction to Turbocharging …

167

(c) Flow Noise There are two major types of flow noises occurring in the compressor stage. Namely, • Tip-in—Flow noise due to TC operation close to the surge limit line during vehicle acceleration. • Tip-out—Flow noise due to sudden drop in compressor discharge demand, during vehicle deceleration. Flow noise is incepted by the flow reversals of the charge air in the compressor wheel by virtue of running it at off-design points on the compressor maps. Marginal flow separation at the inlet side near the blade causes vortices that result in the tip-in noise. It manifests as noise in a broad frequency range between 1000 to 4000 Hz. There is another tonal noise phenomenon called rotating stall, where in, the stall jumps from one cell to the other at a fraction of the rotating speed of the rotor. On the other hand, deep surge condition in the compressor wheel where the charge air totally recirculates from the compressor outlet to compressor inlet is the reason behind the tip-out noise. Figure 4.94 shows a typical vehicle drive pattern that triggers both the tip-in and tip-out noise. Tip-in noise occurs due to close proximity of the compressor operation close to surge limit line during the ramp-up drive cycle from a to b as shown in Fig. 4.94. Tip-out noise occurs when the driver suddenly releases the acceleration pedal to decelerate the vehicle from b to c as shown in Fig. 4.94. At this moment, the required high-pressure mass flow rate for the engine is reduced drastically as the engine throttle valve begins to close. But the turbocharger is still running at high

Fig. 4.94 Engine lug line in vehicle dynamic condition: a–b: acceleration, b–c–a: deceleration, b– c: compressor stall

168

D. A. Subramani et al.

Fig. 4.95 Compressor noise map on gas stand

speed due to its inertia, leading to pressure increase at the compressor outlet. This moves engine lug line towards left side of compressor map causing mild or deep surge in the compressor working condition, consequently leading to flow noise from the turbocharger peripherals like the air intake and exhaust ducts. Figure 4.95 shows TC level noise map from a gas stand cell at each operating point of the compressor obtained using a dynamic pressure sensor on the compressor inlet. The colour contours represent pressure levels. An engine lug line in yellow is overlaid on this noise map. Noise levels are higher when compressor operates near surge region. On the right-hand side, vehicle level 3D spectral map recording is shown in Fig. 4.96 that has amplitude, time, and frequency distribution. It shows TC spool up and down cycles for every five seconds interval. Similar high compressor noise at both tip-in and tip-out conditions was observed as seen at the TC level gas stand tests. Control Measures for Flow Noise: Active Solutions: Compressor trim change, inlet ported shroud Passive Solutions: Mode filter, resonator, anti-surge algorithm Active Solutions: 1. Compressor Wheel Trim Change The Compressor trim describes the ratio of the inlet diameter of the wheel to the nominal outlet diameter D1/D2. Since the inlet area is directly proportional to the inlet diameter, the trim directly influences the flow capacity of the wheel. Hence changing the trim of the wheel is often used for adapting the map of the compressor to a certain flow range.

4

Introduction to Turbocharging …

169

Fig. 4.96 Stall noise in vehicle

As trim is reduced, the compressor map moves toward left allowing high boost pressure at low volume flow rate. Hence, compressor operates away from surge zone as shown in Fig. 4.97. 2. Compressor Housing—Ported Shroud The ported shroud allows air recirculation in the stalled zones in the compressor housing, which extends the compressor stage map width. The Compressor recirculation is formed by a slot between ported shroud and compressor housing as shown in Fig. 4.98 which connects the inlet channel of the compressor. The compressor inlet stall is restrained as long as possible by diverting the reversed flow

Anti-surge line

Fig. 4.97 Anti-surge line

170

D. A. Subramani et al.

Fig. 4.98 Compressor housing with ported shroud

through the recirculation channel and leading it back into the main flow of inlet volume. Due to this recirculation, the flow is stabilized and the surge line is shifted to lower volume flow rates. On the other hand, it also allows more air to pass through the compressor at high flow rates, which extend the choke margin. Figure 4.99 show compressor noise map for ported shroud and non- ported shroud. Acoustic behaviour is improved with the help of recirculation especially near surge region. 3. Vehicle Level Anti-Surge algorithm During operation of a turbocharged engine, the turbine extracts energy from the exhaust and drives the compressor. When load is suddenly reduced, the exhaust energy drops, and the turbine no longer has the power to drive the compressor to keep the speed up, thus the flow rate through the compressor drops faster than the pressure on the outlet drops, resulting in a surge event. Figure 4.100 shows the engine control unit outputs overlaid at the time of a vehicle pedal release instant. With the anti-surge algorithm functional, at the pedal release time stamp, an instantaneous command is sent to open the EGR and turbocharger VTG vanes. The compressor outlet pressure (P2) passes via EGR valve to the exhaust, thus reducing the compressor discharge pressure. Also, VTG opening at the same instant will help to release the exhaust energy (P3) in the turbine stage, thus reducing the turbo speed as shown in Fig. 4.100. Passive Solution: Modal Filter Turbo compressor internal passages and piping systems define a half wave resonant frequency and a quarter wave resonant frequency. In addition to acoustic length resonances, there exists another type of resonance—radial acoustic resonance

4

Introduction to Turbocharging …

Fig. 4.99 Compressor noise map for ported and non-ported shroud

Fig. 4.100 Anti-surge logic: EGR and VTG Opens as pedal released

171

172

D. A. Subramani et al.

Fig. 4.101 Nodal patterns

involving a three-dimensional acoustic resonance with standing acoustic pressure wave patterns that are also perpendicular to piping and flow passage axes. These resonances involve Bessel function (alpha) solutions for the cylindrical wave equation and are defined by: fmn ¼ amn

a d

ð4:50Þ

Nodal patterns are presented in Fig. 4.101. Radial acoustic cross-mode resonance standing waves not only rotate, but will propagate up and down the piping as a helical wave. Table 4.3 shows the radial acoustic resonances corresponding to compressor inlet and outlet duct diameters. The frequencies change corresponding to the nodal patterns. The above Fig. 4.102 shows the fundamental and harmonic of compressor inlet hose is eliminated by disruption of azimuthal mode using a nodal pattern ring f (1,0). Typical surge noise (frequency = 0.5X) around 1200 Hz was also suppressed by 10dBA. This is further illustrated in Fig. 4.103, where during the deceleration zone between 40 and 43 s, the strong surge noise is eliminated by the introduction of the modal filter.

4

Introduction to Turbocharging …

173

Table 4.3 Radial acoustic resonance Duct nodal patterns

Compressor outlet

Compressor inlet

Comp inlet pipe

Air filter connecting Pipe

Diameter (m) f1,0 f2,0 f3,0 f0,1 f2,1

0.024 8718.2375 14461.475 19892.3375 18143.0375 31752.175

0.032 6538.678 10846.11 14919.25 13607.28 23814.13

0.046 4548.646 7545.117 10378.61 9465.933 16566.35

0.07 2989.11 4958.22 6820.23 6220.47 10886.46

Fig. 4.102 Improvement in harmonic frequency

Fig. 4.103 Effect of mode filter at compressor inlet for surge noise

174

4.9

D. A. Subramani et al.

Testing Methodologies

Data Acquisition Selection: Previous sections introduced various turbocharger noises and their frequency spreads. To conduct NVH measurement activities on turbocharger, we need suitable data acquisition systems (DAQs) with certain salient features that make for effective resolution and analysis. Proper vibration analysis requires acquirement of accurate time varying signal from an accelerometer. Thus, the selection parameters play a vital role in achieving this objective. A data acquisition system that can record and playback the recorded data should be used with stand-alone or front-end capabilities. The Fig. 4.104 shows a data acquisition system used for illustration. This DAQ has eight channels with 5 V input voltage as well as binaural recording with replay options and is also capable of rpm measurement channel called tacho input, apart from CAN and GPS provisions. Depending on the frequency of interest, it is necessary to decide the number of samples and sample rate during data logging usually kept at 48 kHz. Time histories thus acquired are converted to frequency domain through in-built Fast Fourier Transform (FFT) algorithms, FFT is a process (FFT Wiki [Wikipedia.org/Fast Fourier Transform]) by which the time-varying input sample is divided into its individual frequency components. The important parameters in defining FFT are the resolution, frequency range, window type and averaging type. The most advanced data acquisition system has the capability of FFT based real time analyses and adjustable filter modes with in-built battery that enhance long hours of standalone capacity (up to 6 h) which makes it robust during measurement.

Fig. 4.104 Data acquisition system

4

Introduction to Turbocharging …

175

Fig. 4.105 Data acquisition system

The acoustic and vibration sensors (say accelerometers) are connected along with binaural headset to the data acquisition device through appropriate low noise cables as shown in Fig. 4.105. The data recording is monitored through a laptop connected to data acquisition system.

4.9.1

Analysing Methodology

Vibration analysis starts with time varying input signal from the accelerometer which is pass through certain filters and digital converters to be processed by FFT which converts the time domain data to frequency domain with corresponding phase and amplitude.

When we speak about FFT which computes DFT (Discrete Fourier transform) and produce exactly the same result as evaluating the DFT definition directly, the most important difference is that an FFT is much faster. XK ¼

N 1 X

xn e

i2pkn N

ð4:51Þ

n¼0

The most important parameters in FFT are resolution, aliasing and sampling rate. This FFT resolution defines the number of lines that appear as information in the FFT plot. Noise and Vibration spectrum are plotted with respect to time and frequency. The Turbocharger operating frequency is identified from the acceleration/ Vibration spectrum. The recorded noise is played in the software and heard by the analysis engineer using binaural headsets with various combinations of noise filters applied to filter out other noise. The Noise due to Turbocharger is identified corresponding to the TC frequency. The noise duration, amplitude and frequency at specific instances is marked on the spectrum. The Turbocharger noise spectrum and

176

D. A. Subramani et al.

Table 4.4 10-Point NVH subjective rating sheet 10

Outstanding

…not perceptible

9

Excellent

8

Very good

7

Good

6

Satisfying

5 4

Dissatisfying borderline Insufficient

3

Poor

2

Unacceptable

1

Safety risk

…hardly perceptible by specialist in special driving cycle Not audible for end customer …Slightly audible by specialist in special driving cycle Hard perceptible for end customer …Slightly audible in a few customers relevant driving cycle A very critical end customer could complain …quietly audible in a customer relevant driving cycle A critical end customer could complain. As borderline sample with small probability of occurrence (not every TC over lifetime) barely acceptable …audible in customer relevant driving cycles An average end customer could complain …Clearly audible in customer relevant driving cycle Every end customer could complain …loudly audible in many customers relevant driving cycle Every end customer will complain and worry about damages. …Very loudly audible in all driving cycle End customer will complain and put vehicle for …extremely loud audible in all driving cycles End customer stops operation of vehicle due to fear damage

the subjective assessment are correlated and the final inference/conclusion on the existence of Turbocharger and other noise and the possible cause is reported. Initially turbocharger noise and vibration are evaluated using cold test conditions to evaluate 1X synchronous noise and constant tone noise. The vehicle is driven for a short period till the engine and oil temperature is warmed up. Further tests like WOT (Wide open throttle) and Tip-in and Tip-out are conducted under geared operation. Apart from recording the data, the noise levels are subjectively assessed and rated to the 10-Point scale (1 being bad and 10 being good). The subjective assessment is mutually agreed and signed by test engineer and the customer. The below Table 4.4 shows the 10-point rating scale sheet which has been followed internally.

4.10

Vehicle NVH: A Turbocharger Perspective

Vehicle NVH refinement is done throughout the product development cycle with target in-cab noise levels derived from experience or through benchmarking. Corresponding target acoustic transfer functions are cascaded down from system to component levels, and a typical distribution of noise sources (Laurent 2012) types, and their frequency contents are depicted in Fig. 4.106.

4

Introduction to Turbocharging …

177

Fig. 4.106 Structural and air borne noise radiating frequencies

One of the unique noise behaviour of a small passenger car turbocharger is that it spins up to speeds as high as 300,000 rpm (or 5000 Hz), and has a structure-borne noise source in the form of unbalance moment related excitation to such high frequencies. Per Fig. 4.106, typical NVH refinement for structure borne noise is limited to 1000 Hz, and as a result, end up having sensitive peripheral component amplifiers such as heat shields and exhaust pipes not refined to higher frequencies. This puts great onus on the turbocharger supplier to reduce the residual unbalance forces at these high speeds that could have a direct impact on production scrap and throughput times for high volume applications.

4.11

System Level Counter Measures for Turbocharger Noise

Few applications are cited that show the influence of the system level transfer path in the acoustic emission. Transmission of Sub2 Oil Whirl Noise through support bracket The structure-borne constant tone source excitation created by the rotating components is transferred to acoustic sensitive parts (radiated components) through sensitive transmission path components like mounting brackets, support brackets etc. In the transmission path NVH development approach, first one is to attack the excitation sources and minimize the excitation and overall noise as much as

178

D. A. Subramani et al.

Fig. 4.107 Bracket influence on tonal noise

possible and the next one is to address the noise transmission path components through decoupled studies. Turbocharger tonal noise contribution through the mounting paths was studied and results shown in Fig. 4.107. The bracket had a resonance in the Sub2 oil whirl frequency and was transmitting this energy to the engine block. Ones the bracket was removed, Fig. 4.107 shows improvement in the vehicle interior noise and subjective evaluations showed a positive outcome. This case study, shows that a bearing noise could be addressed through system level investigations as well. Exhaust Decouplers Decoupling elements (Fig. 4.108) are installed at various positions of the exhaust system. These flexible connections are used to avoid direct contact between the exhaust system and the turbine housing. Exhaust system components are the

4

Introduction to Turbocharging …

179

Fig. 4.108 Decoupler between cat-con and turbocharger

radiators for turbocharger unbalance (structure borne vibration) noise. Flexible connections are restraining the vibration transfer between turbo and exhaust system and result in the reduction of unbalance noise. Figure 4.108 shows the colour plot comparison of the baseline measurement and with the decoupled cat-con evaluation on the vehicle, the turbocharger 1X synchronous noise level has been improved by 2 rating subjectively with the decoupled cat-con which is shown in the in cabin microphone and also on the near field microphone on the cat-con side. External Insulation External insulators are used for protection against the highest temperatures and noise radiation due to structural excitation from the turbocharger. External insulators consist of a single carrier sheet, insulating core and a cover. The core can be built up from several layers of embossed aluminium foils which are sandwiched together as shown in Fig. 4.109. It can also be an insulating mat (ceramic felt or glass fibre mat). This decrease or damped the amplitude in the medium to high frequency noise radiation of the components and results in improved acoustics within the passenger section.

180

External Insulation of Cat-Con

D. A. Subramani et al.

Acoustic Holograph Hot spot: Cat Con

Fig. 4.109 External insulation illustration on exhaust catalytic convertor

Heat Shields The influence of assembled heat shield systems upon the acoustic radiation behaviour of the turbocharger is due to both the large surface area and installation. Parameters such as the component’s surface, the thickness of the associated insulation material as well as the number and positioning of the utilized bolt joints play a key role in the vibration transmission. Figure 4.110 shows heat shield contribution to noise. Countermeasures for Pulsation Noise—Compressor Outlet Duct Radiation One of the countermeasures to reduce pulsation induced turbo whistling noise is to modify the wall thickness or stiffness material of compressor outlet ducts. This dictates the frequency dependant radiation efficiencies of the ducts. The pipe between the turbocharger compressor and the intercooler connection to the engine plays a major role in air borne noise transmission. When the metal compressor outlet duct is used, the pulsation whistling noise strongly reduces due to the damped vibration amplitude. The effectiveness of the metal compressor outlet duct is tested

4

Introduction to Turbocharging …

181

Noise Spectrum with/without external insulation of cat-con

Fig. 4.110 Exhaust heat shield contribution to noise

on a vehicle by measuring the acoustic pressure pulsations on turbo compressor outlet, and corresponding radiated noise in the near field as well as in-cab noise levels as shown in Fig. 4.111.

182

D. A. Subramani et al.

Fig. 4.111 Compressor outlet duct acoustic radiation

Resonator in compressor ducts Pressure pulsation is also caused by the non-uniform flow or pressure fluctuations in compressor outlet. One of the countermeasures to reduce pulsation induced turbo whistling noise is to directly install a resonator on compressor outlet or an integrated resonator in compressor housing to attenuate the pressure fluctuations. Resonator reduces the pulsation whistling noise by reflecting acoustic energy back to where it comes from or by converting incident acoustic energy into heat or by a combination of both. Figure 4.112 shows the colour plot comparison of the baseline

Fig. 4.112 Resonator influence at compressor outlet

4

Introduction to Turbocharging …

183

(without resonator) and after resonator. The plot of the left side shows the baseline measurement of the in-cabin noise and the plot on the right side shows the improvement in pulsation noise on the in-cabin noise with resonator.

4.12

Innovations in Turbochargers

Over the past several decades the basic turbocharger configuration has gone through several changes. Innovations in this area can be categorized into ideas that improve the efficiency, operating limits (surge, choke, and speed limit), utilization of engine pulse energy, durability, manufacturing ease and so on. A brief description of a few of these ideas is provided here.

4.12.1

Compressor with Splitter Blades

The maximum mass flow capability for a given speed for a compressor is limited by the section of minimum area which happens to be near the inlet of the impeller. The presence of blades adds blockage to the flow. While more blades would be useful from an efficiency point of view (guide the flow between) it is detrimental to the choke flow. One way of circumventing this problem is to add blades (Fig. 4.113) which start away from the minimum area—so that the blockage does not reduce the minimum area at the same time guiding the flow near the exit. These are known as splitter blades.

4.12.1.1

3D Blade Design

For reasons of ease of manufacturing in general the blade geometry is defined as a ruled surface—this helps to machine using a process called flank milling fairly Fig. 4.113 Centrifugal impeller with a splitter blade

184

D. A. Subramani et al.

Fig. 4.114 Three-dimensional blade design

rapidly. However, defining the blade taking advantage of the benefits of more optimal aerodynamic setting (called 3D blade design, Fig. 4.114) allows compressor efficiency to increase by as high as 3%.

4.12.2

Boreless Compressor Wheel

The maximum stresses occur usually near the bore in a typical compressor wheel which has a through-bore. This bore is needed to connect the compressor wheel to the turbine wheel and shaft. By removing the bore the maximum stresses reduce drastically to an extent the fatigue life is doubled (Fig. 4.115).

Fig. 4.115 Boreless compressor wheel shown alongside a through-bore wheel with zone of maximum stress (Jean-Luc et al. 1991)

4

Introduction to Turbocharging …

4.12.3

185

Twin Scroll Turbochargers

The twin scroll turbine design (Fig. 4.116) is characterized by a volute that is divided meridionally by a wall with two parallel inlets. Each inlet is feeding the nozzle-less turbine through the entire rotor circumference. The divided volute itself is fed through an engine exhaust manifold which has separated volumes. The pulse energy is utilized for improved transient engine performance while minimizing engine pumping losses, in this configuration. Twin scroll turbochargers are commonly used in engines where the ‘low-end’ performance, i.e. providing relatively high engine torque at lower engine speed is of paramount importance. In Fig. 4.116 for example, one of the scrolls is connected to cylinders 1 and 4 while the other is connected to cylinders 2 and 3.

4.12.4

Double Scroll Turbocharger with VGT

In this configuration the scroll is divided radially. The big advantage of double scroll technology is in the separation level between both volutes which enables the pulse energy of each cylinder is transferred in an efficient way to the turbine wheel. This arrangement improves the scavenging behaviour significantly in addition to improving the torque at lower engine speeds (Fig. 4.117).

4.12.5

Variable Outlet Turbine

The variability of this device is based on outlet area changes as opposed to the more common systems that are based on inlet turbine geometry changes. The new variable turbine type is termed variable outlet turbine (VOT). The flow variability is achieved by variation of the flow cross section at the turbine outlet using an axial displacement of a sliding sleeve over the exducer and provides a simple solution for flow variability (Fig. 4.118). High Performance Turbocharger with Water-Cooled Casing Using water cooling for the casing and compressor impeller limits the engine’s surface temperature making the turbochargers thermally very durable (Fig. 4.119). But, the drawback of this technology is that the turbine outlet temperature will be lower and it affects the after-treatment process. In order to have good after treatment process to reduce the emissions the turbine outlet temperature should be high.

186

Fig. 4.116 Turbocharger with a twin volute turbine housing

D. A. Subramani et al.

4

Introduction to Turbocharging …

Fig. 4.117 Turbocharger with a double scroll turbine housing (BWTS 2014)

Fig. 4.118 Turbocharger with a sliding sleeve in the turbine housing

187

188

D. A. Subramani et al.

Fig. 4.119 Turbocharger with a water-cooling circuit in bearing and compressor housings (BWTS 2014)

4.12.6

Water-Cooled Compressor Housing

Water-cooled compressor housings are designed for those who need high specific engine power outputs. The extreme boost pressure ratio however, will lead to a high compressor outlet temperature, which causes coking layer in the compressor and finally reduces power output. Therefore, a cooling system is integrated into the low-pressure stage compressor housing to reduce the compressor outlet temperature and ultimately to prevent the coking of oil in the compressor volute and the charge air cooler. A water-cooled compressor housing assists downsizing to achieve a maximum level of specific power and a maximum transient response with large improvements for fuel consumption. Water cooled compressor housing will have an efficiency advantage against that without water cooling (Fig. 4.120).

4.12.7

Recirculation Groove

At low flow rates a part of the flow entering the impeller is recirculated back into the inlet creating an apparent increase in the inflow. This increases the surge side stability. The position of the recirculation port is critical in achieving the improvement without affecting the efficiency. In principle choke margin could also be increased in a well-designed ported shroud design (Fig. 4.121).

4

Introduction to Turbocharging …

189

Fig. 4.120 Turbocharger with a water-cooled compressor housing (BWTS 2014)

Fig. 4.121 a Compressor housing with a ported shroud, b map width improvement and c flow in a ported shroud turbocharger

190

D. A. Subramani et al.

Fig. 4.122 Electric booster (BorgWarner turbo systems)

4.12.8

Electric Booster

In the conventional turbochargers, it takes a few seconds for the engine to respond and the turbo to start spinning which is known as the turbo lag. The e-booster eliminates it by augmenting the turbocharger with an electric motor. An electric supercharger can spool fully in as little as 0.5 s through the direct connection with the throttle, giving virtually instant maximum boost. An interesting visual of eBooster can be seen in https://www.youtube.com/watch?v=iEGisoOnewg (Fig. 4.122).

4.12.9

Kinetic Energy Recovery System

The turbocharger can be used to drive a generator when excess power is available and when run in motor mode can augment the power required for the compressor. Here the exhaust turbine, compressor wheel, and generator are all mounted on a common shaft. The electric assisted compressor can thus reduce the “turbo-lag” at lower speeds while at the same time offering energy recovery higher speeds (Fig. 4.123).

4.12.10

Regulated Two-Stage Turbocharger

Two stage turbochargers essentially have more than one turbo. One such concept is to use two stages one for lower flow and the other for higher flow allowing the benefits of the small and larger compressor maps to be used effectively in different parts of operation (Fig. 4.124).

4

Introduction to Turbocharging …

191

Fig. 4.123 Electric assisted turbo charger (BWTS 2014)

Fig. 4.124 Two-stage turbo charger (BWTS 2014)

The intake air mass flow goes through compression in the low pressure (LP) stage and, further compression takes place in the high-pressure stage. As a result of the pre-compression in the LP stage the relatively small HP compressor operates at a higher-pressure level, so that the required air mass flow throughput can be obtained. At low engine speeds and low exhaust gas mass flows, the bypass

192

D. A. Subramani et al.

remains closed and the entire exhaust gas expands through the HP turbine. This results in a very quick and high boost pressure rise. With increased engine speed, the bypass valve is opened, progressively shifting more of the expansion work to the LP turbine. Major benefits of a regulated two-stage turbocharging are: • high torque at lowest engine speeds resulting in improved transient response • power increase at rated speed and hence potential to reduce rated speed • reduced fuel consumption, smoke and NOx emissions.

4.12.11

Bearing Systems

Most of the bearings used in turbochargers today employ the floating ring journal bearings. With the journal bearings, the proportion of bearing friction increases when the turbocharger speed is low. Introduction of ball bearings reduce the power loss significantly providing improved transient response as well as fuel economy benefits (Christopher 2014). The ball bearings although cost need lesser oil supply and can offer improved thrust bearing capacity and hence is likely to see more applications in the future!

References ANSYS-CFX V18.2, Ansys Inc AxSTREAM, SoftInway Inc Baines NC (2005) Fundamentals of turbocharging, concepts NREC Christopher M (2014) Ball bearing turbocharger – technology development. In: Schaeffler Symposium, USA Cumpsty N (2003) Jet propulsion, 2nd edn. Cambridge University Press Dhinagaran R, Balamurugan M, Seran K, Gopalakrishnan M, Vasudevan R, Swathi CL (2017) Development of efficient compressors for turbochargers. In: ASME GT2017-64359, Proceedings of the turbo expo turbomachinery technical conference & exposition, June 26– 30, Charlotte, NC USA Dixon SL (1998) Fluid mechanics and thermodynamics of turbomachinery, , 5th edn. Elsevier Butterworth–Heinemann eBooster, BorgWarner turbo systems. https://www.youtube.com/watch?v=iEGisoOnewg Feneley AJ, Pesiridisa A, Andwaria AM (2017) Variable geometry turbocharger technologies for exhaust energy recovery and boosting-a review. Renew Sustain Energy Rev 71:959–975 Fine/Turbo, Numeca Inc https://www.acea.be/industry-topics/tag/category/co2-from-cars-and-vans ICCT Briefing (2016) A technical summary of Euro6/VI vehicle emission standards. www.icct.org Jean-Luc PF, Jerome WT, Voytek K (1991) Turbocharger compressor wheel assembly with boreless hub compressor wheel. US Patent US4986733A Laurent G (2012) Vehicle NVH design. Lecture notes Lewis RI (1996) Turbomachinery performance analysis, 1st edn. Butterworth-Heinemann

4

Introduction to Turbocharging …

193

Muszynska A (2005) Rotordynamics, vol 188. CRC Nguyen-Schafer H (2015) Rotordynamics of automotive turbochargers, 2nd edn. Springer Nowald1 G, Boyaci1 A, Schmoll1 R, Koutsovasilis P, Schweizer B (2015) Influence of circumferential grooves on the non-linear oscillations of turbocharger rotors in floating ring bearings. In: The 14th IFTOMM world congress, Taipei, Taiwan, Oct 25–30 Powerful turbocharging system for passenger car diesel engines (2014) BorgWarner Knowledge Library, BorgWarner Turbo Systems (BWTS) Gmbh San Andres L (2012) On the effect of thermal energy transport to the performance of (semi) floating ring bearing systems for automotive. J Eng Gas Turbine Power 134 Vanhaelst R, Kheir A, Czajka J (2016) A systematic analysis of the friction losses on bearings of modern turbocharger. Combust Engines 164(1):22–31 Watson N, Janota MS (1982) Turbocharging the internal combustion engine. MacMillan Publishers Ltd Wikipedia.org/Fast Fourier transform Xu C, Amano RS (2008) Design and optimization of turbo compressors. WIT Trans State Art Sci Eng 42 Zucker RD, Biblarz O (2002) Fundamentals of gas dynamics, 2nd edn. Wiley

Chapter 5

Topics on Selective Catalyst Reduction P. Kumar

Abstract SCR concept is used effectively in large electric power stations to abate nitric oxides for the last fifty years. Here, simpler controls to inject ammonia in the engine exhaust over a catalyst were enough since the change in emission was only with respect to load and slow. However, application of this technique to truck and car engines was challenging and invited control and thermal problems. This chapter, after introducing the chemical technology goes in depth, explaining the engine optimization for SCR technology, the trade-off of NOx and soot, and rail pressure and timing. The principle of NOx reduction is explained using NOx model with maps of exhaust flow, storage of ammonia in the catalyst and dosing ratio. A temperature model is essential for precise control of nitric oxides. Also, a model for hydrolysis of urea to ammonia and accelerated reaction rates in the presence of nitrogen dioxide is given to consider upstream diesel oxidation catalyst that not only increases the concentration of nitrogen dioxide but also oxidizes carbon monoxide as well as hydrocarbons. Understanding the specification and salient properties of urea solution is important for the success of the SCR technology. In the highly transient thermal environment, there are various types of potential failures due to creep, thermal stress and flow stress. Further, the hydrolysis of urea just after injection, nozzle clogging, crystallization in the catalyst, ammonium di sulphide plugging, active catalyst surface plugging, poisoning due to high sulphur in the fuel as well as potassium and other alkali metals from lubricating oil and other flow phenomenon are studied for the longevity of the SCR system in the engine with the optimally located urea injector. Airless injection system does away with the need for compressed air which is necessary for air assisted injection system; however, distribution of urea and its hydrolysis in the flow is more complex and hence more detailed design analysis must be carried out. Continuous development and tightening of emission limits call for Cu or Ze based catalyst that lights off at a lower temperature than the cost-effective and sulphur tolerant vanadium-based

P. Kumar (&) VE Commercial Vehicles Ltd., Indore, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_5

195

196

P. Kumar

catalyst. Choice of metallic or ceramic catalyst substrate creates an intense dilemma between cost, manufacture, thermal response, and life. Coated or extruded catalyst is again a design trade-off. The high-speed electronic control or the dosing control unit communicates with engine management system. The entire application exercise involves in modelling the storage, reaction and slip of ammonia over the catalyst for transient cycles. Therefore, some companies have preferred exhaust gas recirculation which is less fuel efficient than SCR. SCR in emerging markets has its own challenges and advantages. There are challenges in the field regarding dosing system failure, HC poisoning, catalyst wash out, reliability of NOx sensors. Application of SCR calls for long duration field trials at varying loads, at high altitudes and temperatures which are either not envisaged in the engine laboratory or not possible to simulate easily.

5.1

Introduction

In Euro-2 and Euro-3 NOx is reduced by retarding the injection timings and lower rail pressures at the cost of fuel economy and particulates (PM) emissions. Moving towards Euro 4/5 norms, the NOx standards for mobile heavy-duty diesel engines are becoming tighter. This chapter presents NOx Selective Catalytic Reduction (SCR) technology for Euro 4–6. The influence of SCR system on fuel economy and CO2 levels are demonstrated. Engine optimisation technique adapted through major PM reduction is carried out through combustion optimization resulting in increased NOx. NOx is reduced in the exhaust catalytic convertor by injection (32.5%) urea solution in water with SCR system in the engine exhaust using a dosing system assisted by air at pressure. The results obtained for the Euro-4 engine with SCR show 6% improvement in fuel economy over base BS-III engine, 50% reduction in PM and 70–90% reduction in the NOx was found at the tailpipe, in European Stationary Cycle (ESC) and European Transient Cycle (ETC) across SCR based on the Injection quantity for Euro 4 and Euro 5 Norm. The SCR system is rugged and efficient over the long engine life. The SCR catalyst is not affected by high sulphur in fuel, unlike Diesel Particulate Filters (DPF). The cost of urea over a cycle is offset by the gain in fuel economy in on-road operating conditions. SCR technology is attractive and feasible for moving towards the Euro4 and Euro5 standards, and beyond, with minimum modifications to the structure of the engine. Selective Catalytic Reductions (SCR), Exhaust Gas Recirculation (EGR) plus Diesel Oxidation Catalyst (DOC) in conjunction with Particulate Oxidation Catalyst (POC) are the routes for meeting Euro 4 and for beyond Euro 4 EGR, SCR and Diesel Particulate Filter (DPF) are recommended. After-treatment devices of SCR are more attractive to the Heavy-Duty (HDV) and Medium Duty Vehicles (MDV), for fuel economy, minimum hardware changes, reliability, and attractive operating cost (Johnson 2009, 2010). Activities and benefits from the engine optimised with SCR after-treatment device are discussed in this chapter.

5 Topics on Selective Catalyst Reduction

5.2

Terminology

SCR Urea solution UDS ACU ECU Mixer Open loop Closed loop ASC OBD EGR PM DOC EOE SOE DPF HC POC

5.3 CO CO2 NH3 Urea NOx H2O SO2 SO3 NO NO2 N2 HCNO V2O5 TiO2 W2O3 Fe-Ze Cu-Ze

197

Selective catalyst reduction Aqueous solution of urea with water 32.5% Urea dosing system After treatment control unit Engine/electronic control unit Component for creating turbulence for mixing urea with exhaust gas No NOx feedback provided to ACU/ECU for the NOx conversion correction NOx feedback provided to ACU/ECU for the NOx conversion correction Ammonia slips catalyst On-board diagnostics Exhaust gas recirculation Particulate matter Diesel oxidation catalyst Engine out emission System out emission Diesel particulate filter Hydrocarbon Particulate oxidation catalyst

Chemical Formulae Carbon monoxide Carbon-dioxide Ammonia CO (NH2)2 Oxides of nitrogen (NO + NO2) Water Sulphur dioxide Sulphur trioxide Nitric oxide Nitrogen dioxide Nitrogen Isocyanic acid Vanadium pent oxide Titanium diOxide Tungsten oxide Iron zeolite Copper zeolite

198

5.4 5.4.1

P. Kumar

Emission Trends Emission Norm and Engine Technology

Referring to the NOx Fig. 5.2 PM trade-off chart, there are two ways of achieving Euro 4 or Euro 5 (Fig. 5.1): (a) reducing NOx in cylinder and PM by after-treatment. This occurs in the EGR engines. Here, a part of the exhaust gas is circulated back to the engine and NOx reduction happens due to the slow and inert combustion resulting in higher soot, which will be converted to CO2 in aftertreatment device like DPF or POC or PM catalyst (b) In SCR technology engines, PM is reduced in cylinder by conventional engine combustion with advanced injection and by improving spray quality with high injection pressure; NOx is reduced in the aftertreatment device like SCR.

5.4.2

Engine Optimization for SCR Technology

Rail Pressure and Timing, NOx and Soot Trade-Off NOx and soot trade-off are to be understood for the various ratings of an engine. This depends on the rail pressure and timing for the common rail system where the nozzle configuration is chosen for the required engine power rating and emission level (Table 5.1). Soot reduction is achieved with higher injection pressures and SFC improvement achieved through advancing the timing. Optimization of rail pressure, timing, injection system, combustion bowl and air management system are done to meet the engine out emission. Figure 5.3 provides soot, NOx and BSFC trade-off for various rail pressures and temperatures. This needs to be plotted for various steady state

Table 5.1 Emission legislative limit

Emission CO HC NMHC CH4 NOx PM NH3 Smoke (ELR)

Unit g/kWh

ppm m−1

Euro4 ESC 1.5 0.5 – – 3.5 0.02 25 0.5

ETC 4 – 0.6 1.1 3.5 0.03 25

Euro5 ESC 1.5 0.5 – – 2 0.02 25 0.5

ETC 4 – 0.6 1.1 2 0.03 25

5 Topics on Selective Catalyst Reduction

199

0.10

3 Current Engine Technology

Parculate, g/kW h

0.08

0.06

0.04

SCR 0.02

4

5 6

0.00 0

2

4

6

8

10

12

14

NOX, g/kW h

Fig. 5.1 Emission norm and engine technology for meeting, Euro-3, 4, 5 and 6 boundaries shown as circles

0.12

Bsoot, g/kW h

0.10

In-cylinder Tradeoff

0.08

EGR System

0.06 EGR +DOC + POC

0.04 0.02

SCR System

0.00 0

2

4

6 8 BSNOx, g/kW h

10

12

14

Fig. 5.2 Soot versus NOx versus SFC trade-off

modes from lower speed to higher speed and higher load to lower load. Interval is based on the accuracy and model requirements if any. Initial SCR efficiency is browsed to get the maximum possible NOx engine out emission to meet the required emission level. Engine optimization is carried out to meet the PM and NOx standards. SCR is introduced to meet the NOx system-out emission limits. Additionally, SOF conversion is also expected with SCR to meet the PM emission targets.

200

P. Kumar

Fig. 5.3 Rail pressure, injection timing trade-off for NOx and soot

5.4.3

SCR System History

Selective catalytic reduction of NOX using NH3 as the reducing gas was patented in the U.S. by Engelhard Corporation in 1957. Researchers in Japan invented vanadium/titanium catalysts for SCR, which are successful in operation. This combination forms the basis of current SCR catalyst technology. Several other countries in Western Europe have enacted stringent NO emission regulations that mandated the installation of SCR leading to extensive development of catalysts. Selective catalytic reduction of NOx using ammonia as the reducing agent was patented in the United States by the Engelhard Corporation in 1957 (Jensen-Holm et al. 2012). Urea based SCR technology was chosen by a number of manufacturers for meeting the NOx limit of 2 g/kWh according to Euro V (2008) and the JP 2005 standards, in Europe and Japan. As early as in November 2004 Nissan in Japan and in early 2005 Daimler introduced SCR in commercial diesel trucks ahead of the incoming norms. In 2010, SCR was widely used by the manufacturers of heavy-duty engines in the United States, to meet the stringent NOx limit of 0.27 g/kW h. Though SCR was introduced in some light duty vehicles to satisfy US EPA Tier2, many manufacturers preferred NOx adsorbers. However, by about 2012–2015, most of the Tier 2 vehicles have been changed to SCR using urea. In Europe, SCR has found a much wider application in Euro 6 vehicles. To meet the US Tier 4i/EU Stage IIIB emission standards, SCR has become inescapable in non-road diesels.

5 Topics on Selective Catalyst Reduction

201

Fig. 5.4 SCR reduction mechanism (2009-01-173)

5.5

Principle of NOx Reduction Mechanism

Urea solution injected into the exhaust at more than 200 °C, undergoes evaporation, thermolysis and hydrolysis in the exhaust stream to form NH3. The NH3 is absorbed in the surface of the catalyst, reacts with NO and NO2 to form harmless N2. 4NH3 þ 4NO þ O2 ! 4N2 þ 6 H2 O ðPredominant reactionÞ 2NH3 þ NO þ NO2 ! 2N2 þ 3H2 O ðTen times faster reactionÞ 4NH3 þ 2NO2 þ O2 ! 3N2 þ 6H2 O ðSlow reactionÞ Source of NH3 is urea; urea is solid substance and therefore, aqueous urea of 32.5% (eutectic) composition is used. COðNH2 Þ2 ! HCNO þ NH3 : As illustrated in the Fig. 5.4, the reaction of NH3 with NO is not instantaneous as NH3 has to find place in the catalyst pores. NO reacts with NH3 and leaves as N2. The formulation of catalyst, V2O5 improves the reaction at all temperatures even when the volume of the catalyst is low.

5.6 5.6.1

Models in SCR Calibration NOx Model/Map

NOx model provides the NOx (g/h) for various engine speed and load operation in the vehicle. NOx (ppm): f (Speed, Load), is corrected for humidity, cylinder temperature.

202

P. Kumar

Fig. 5.5 NOx flow estimation by control unit

HCNO þ H2 O ! NH3 þ CO2 Model based calibration provides engine-out NOx as a function of engine speed, quantity, rail pressure and timing. This does not use map for NOx calibration and is useful for all strategies meeting the warm-up, high altitude and defeat strategies, high altitude and EGR (if any). Model based engine-out NOx is used in advanced engines with high ECU capacity to perform NOx and torque output. This will be extremely advanced to meet dynamic and transient conditions of engine (Fig. 5.5).

5.6.2

Exhaust Mass Flow Rate

Exhaust mass flow which is a summation of the air flow and fuel flow is calculated based on the boost pressure, temperature and volumetric efficiency or air mass flow is transmitted from the engine EMS, Fig. 5.6 shows the methodology adapted for exhaust mass flow prediction (Fig. 5.7).

Fig. 5.6 Exhaust mass flow model

5 Topics on Selective Catalyst Reduction

203

Fig. 5.7 Catalyst efficiency model & dosing quantity model

5.6.3

Efficiency Model

It is pre-calibrated maps based on the exhaust temp, mass flow rate and NH3 storage. Based on the efficiency model the NH3 left out in the catalyst is calculated to find the urea dosing rate. Conversion efficiency is calibrated for a given dosing system, muffler passage, mixing, catalyst size and catalyst coating. The efficiency depends on NH3 stored, catalyst temperature and exhaust mass flow rate. Efficiency of vanadium-based catalyst is directly proportional to the catalyst temperatures up to 500 °C and drops above this temperature, due to nil ammonia storage at high temperatures and possibilities of NH3 oxidation above this.

5.6.4

Storage Model

The storage model provides the information of maximum NH3 stored in the catalyst as a function of exhaust mass flow rate and exhaust temperature. Storage exceeding the limit leads to the NH3 slip from the catalyst. Continuous monitoring NH3 stored in the catalyst is modelled through the efficiency and engine-out NOx. NH3 storing capacity ¼ f ðexhaust temp; exhaust mass flowÞ NH3 stored ¼ f ðConversion efficiency; NOx engine outÞ

5.6.5

Dosing Ratio

The dosing ratio is a pre-fed map as a function of exhaust temperature, flow rate, speed and load. The dosing ratio is tuned for the emission requirement based on the conversion efficiency requirements. Ideally, in steady state, the dosing ratio is equal to the NOx conversion efficiency. Dosing ratio ranges from 60 to 130% based on the requirement of conversion efficiency and the operating points to fill catalyst with NH3 at low temperature.

204

5.6.6

P. Kumar

Temperature Model

All models of SCR depend on catalyst surface temperature model. This will be either the average temperature of the SCR-in and SCR-out or the ratio of SCR-into out temperature base on the catalyst location. A few SCR configurations are without SCR-out temperature sensor and SCR-out temperature is predicted by the model based on warm-up, heat transfer, thermal inertia of the brick. This involves a few tuning constants such as convection coefficient, catalyst heat-up coefficient and ambient convection coefficient. Coefficient tuning tests are conducted at various conditions, namely steady state, transient, warm-up and cold start to ensure the average and outlet temperatures of the catalyst are predicted well to match with the actual. The model is used for cost sensitive market, but the output from a sensor is more effective and accurate especially when no ammonia slip catalyst is used.

5.6.7

OBD

NOx monitoring, NOx control monitoring, sensor monitoring and plausibility checks form the major activity in the OBD. The threshold limit for NOx for mal-function indication lamp on (MIL-on) and torque reduction vary for Euro-4 and Euro-5. NOx sensor is a complicated system in SCR system; electronics is important up to tail pipe for safe and reliable running of vehicle in field. NOx sensor start-up procedure is calibrated and checked to have effective and reliable output from NOx sensor. NOx sensor is not started for operation till engine management system sends the dew point signal to NOx sensor control unit. This is to ensure the NOx sensor is safely operated at temperatures higher than 200 °C since condensation of water in NOx sensor should be avoided since zirconia present in NOx sensor cracks with water dew on it.

5.7

SCR System Logic and Structure

See Table 5.2.

5.7.1

SCR Operation and Concepts

A SCR system comprises of an ACU, a catalyst housing, a urea injector and urea dosing system (UDS). The UDS gets command from ACU for dosing quantity. The quantity calculation is done with engine-out NOx, efficiency requirement at operating point and maximum possible conversion efficiency.

5 Topics on Selective Catalyst Reduction

205

Table 5.2 SCR system logic and structure NOx determination

SCR models

Urea solution flow models

SCR OBD models

NOx model Or NOx maps Or NOx sensor

SCR catalyst model NH3 storage control Speed-load strategy Dosing strategy NOx tail-out prediction Urea consumption strategy SCR-out temperature model

Pump swept rate, pipe length compensation Pressure difference across injector Flow temperature, pressure compensation Injection pipe model Tank level estimation

NOx SOE model Time-scheduled checks Sensor plausibility models Torque control activation model

Engine-out NOx is calculated based on NOx model and the equivalent amount of urea injected based on the dosing ratio map is estimated. The efficiency of the NOx conversion is calculated based on the efficiency model. NH3 stored in the system is calculated with the engine-out NOx and conversion efficiency. Further, the dosing of the urea solution is based on the NH3 stored and the maximum NH3 storage limit. If the NH3 stored is higher than the maximum storage capacity, the dosing bit is disabled. 20–30% margin for the NH3 storage is maintained to avoid slip during rapid heating-up of the catalyst and transient operations.

5.7.2

Types of SCR System

See Table 5.3.

5.8

Reductant (Urea Solution) Specification

See Table 5.4.

206

P. Kumar

Table 5.3 Types of SCR system SCR system components

Types

Remark

UDS

Airless system, air-assisted system

SCR catalyst substrate SCR catalyst material Catalyst configuration

Metallic, ceramic

Air assisted for better optimization Airless high-pressure system for low cost Metallic for fast heat-up Ceramic for cost and good thermal inertia V based for S tolerant Ze for better low temperature conversions Pre-oxidation catalyst for better low temperature and high conversion efficiency. ASC for lower slip, for E5 requirements Modular—low back pressure Liner—single brick/space constrain applications Mixer for reducing the mixing length for compact design Single—low cost, high EMS load for temperature average modelling Integrated for compact Individual for variant applications Electronic for air less system Mechanical for air assisted Depends on standardisation of control system Master-SCR ACU calculates urea quantity Slave-SCR ACU controls only dosing, calculation Close loop for better control Sensor based system is accurate, expensive Model based system uses high CPU capacity for calculation Map based system is less complicated and less accurate for all applications

SCR catalyst arrangement

Cu-Ze, Fe-Ze, V-based Combination of pre-oxidation, SCR catalyst, ASC. Hydrolysis catalyst-uncommon Modular, linear

Urea mixing

Natural mixing, mixer hardware

Thermocouple

Double, single

Muffler

SCR integrated, individual

Injectors/ nozzle Control system ACU

Electronic, mechanical nozzle

Feedback NOx prediction

Open loop, closed loop Sensor based, map based, model based

ECU integrated, ACU-independent Master-SCR, slave-SCR

Table 5.4 Reductant (urea solution) specification as per DIN 70070 Chemical formula (urea)

CO (NH2)2

Molecular weight Urea proportion Appearance Hazards Solution density at 20 °C Quality standard Testing method Handling, transportation and storing standard

60.06 32.5% by mass in water Clear and colourless Non-toxic, non-explosive 1087–1092 kg/m3 ISO 22241-1:2006 ISO 22241-2:2006 ISO 22241-3:2008

5 Topics on Selective Catalyst Reduction

5.9 5.9.1

207

Salient Properties of Urea Solution Temperature and Ageing

A logic is required to compensate for temperature and concentration variations during SCR operation to have controlled NOx and NH3 slip for Euro 6/US10. The effect of variations of temperature and concentration on the tail pipe NOx. emissions is insignificant, in case of Euro 4/5. The major component of urea solution is water which has less effect on expansion and compressibility during operation. Normally, every 2–3 diesel filling, urea solution is filled since urea consumption is usually 4–5% of fuel consumption. Accordingly, urea tank is sized.

5.9.2

Freezing and Material Stress

• Level sensor shape and tank structure to be such that to meet the freezing phenomenon of the urea solution. • Tank size and shape must consider sufficient expansion volume and accommodate freezing. • Sensor element must be positioned close to heating source and suction point.

5.9.3

Creeping Behaviour

• SCR system design should be with less sealing and connection points • Integrated solution with one access point and mounted on the top of the urea tank • Points of failure must not be in contact with urea solution, e.g. electronics.

5.9.4

Harsh Environment

Tank materials are chosen to last the lifetime of the truck in urea solution at the operating temperature and sustain operational vibration. A suitable shield must be employed around the tank to ensure no breakage due to vibration, temperature and stone hit.

5.10

Exhaust Layout and General Parts of SCR System

DOC + Air Assisted Injection + Mixer + SCR Catalyst + ASC + NOx sensor (Fig. 5.8).

208

P. Kumar

Fig. 5.8 SCR catalyst arrangements in exhaust system

1 1 1 1

g g g g

of of of of

urea produces 0.57 g of NH3 urea solution produces 0.185 g of NH3 NO requires 3.1 g of urea solution NO2 requires 4.2 g of urea solution

All the above requirements are based on the molar balancing.

5.10.1 Hydrolysis Pipe To accommodate the urea injection nozzle and hydrolysis/thermolysis (urea to NH3), the hydrolysis pipe should have sufficient length. The nozzle should be mounted in such a way that wall wetting or crystallization is avoided in the vehicle application. Any urea droplet on wall may lead to late evaporation leading to discrepancy in calibration and optimisation for system-out NOx. The thermocouple is so positioned that the urea spray does not touch the thermocouple sensing point to allow the thermocouple read the correct exhaust temperature.

5.10.2 Urea Hydrolysis and Mixing When aqueous urea is injected into hot gas, following physical and chemical reactions take place: • Evaporation of the water, after which solid urea particles are suspended in exhaust gas • Urea melts and evaporates • Urea hydrolysis • Mixing of reagent with exhaust gas

5 Topics on Selective Catalyst Reduction

209

Urea Pump ! Urea Nozzle ! Exhaust pipe The urea pump meters and deliver required amount of urea in case of air assisted system. The urea nozzle mixes air with urea and sprays into the exhaust pipe. In an airless system, the electronically controlled urea injector meters and injects the urea solution into exhaust gas. Urea Solution ! Urea ! 1 Ammonia þ 1 Isocynic acid ! 2 Ammonia COðNH2 Þ2 ! NH3 þ HCNO Thermolysis ð[ 200  CÞ HCNO þ H2 O ! NH3 þ CO2 Hydrolysis

Pre-Oxidation Catalyst (DOC) Exhaust stream from the turbocharger flows out through a diesel oxidation catalyst (DOC), a mixer, a SCR catalyst and an ammonia slip catalyst (ASC) and leaves to atmosphere with reduced NOx and some excess ammonia that slips the reactions. The ratio of nitric oxide (NO) to nitrogen dioxide (NO2) in the diesel exhaust is approximately 95:5. Precious metal i.e., platinum is the catalyst (10–15 g/ft3) in DOC increases the NO2 fraction by oxidising NO at low temperatures (200–300 °C). 2NO þ O2 ! 2NO2 The activation energy for the above reaction reduces with high catalyst loading. However, the fuel must be free from sulphur for efficient and durable reaction throughout the life time of the vehicle. A 50–50 NO–NO2 mixture increases the

Fig. 5.9 NO ! NO2 conversion in DOC with various catalysts

70

NO to NO2 conversion, %

60 50 40 30 20 10 0 100

200

300

400

Temperature, deg. C

500

210

P. Kumar

reaction rate of NOx reduction in SCR catalyst even at low temperatures, thus improving conversion efficiency at low temperature. NO þ NO2 þ 2NH3 ! 4 N2 þ 3H2 O The ratio of NO2/NO: 1 provides a good conversion in the SCR catalyst even at low temperatures 200–300 °C. The catalyst action in the DOC and SCR must be tuned for the city duty cycle to have conversion efficiency above the OBD threshold limit. A mixer mounted downstream of the urea injector increases the uniform urea distribution and turbulence. The uniformity increases the NH3 formation rate in the exhaust and avoids deposition of urea in the piping and the catalyst. The turbulence in the exhaust enhances the heat and mass transfer of urea into exhaust. The temperature profile in SCR catalyst during European Transient Cycle (ETC) is lower than during European Steady Cycle (ESC). The first 0–1200 s of ETC exhibits higher temperatures and consequently higher conversion rates than the later 1200–1800 s (Fig. 5.9). Closed loop SCR provides higher conversion efficiency at low and high temperatures. Though the presence of DOC (NO ! NO2) catalyst further increases conversion efficiency in temperature below 300 °C, no difference is observed at high temperature. The DOC enables design of smaller SCR volumes even when the catalyst is aged. Higher platinum loading is not needed and the enhanced target for SCR is reached even with 25–40% of NO2 at about 250–300 °C in typical test cycles or in running conditions of trucks. The application at very low temperatures (buses, distribution trucks in cities) may need higher loading to keep SCR activity at 220– 300 °C. An example of the effect of loading on NO2 formation with typical spatial velocity (SV) of 100,000 h−1 for heavy-duty vehicle is shown in Fig. 5.10.

1.0

100%

0.8

80% open loop without NO to NO2 cat

60% wih NO to NO2 cat

ESC

0.6 0.4

40% ETC

0.2

20% US Transient

0% 0

100

200

300

400

Catalyst temperature, deg C

Fig. 5.10 SCR conversion and catalyst temperature of HD engine cycle

0.0 500

temperature frequency

NOX conversion efficiency

closed loop

5 Topics on Selective Catalyst Reduction

211

A critical factor is the stability of NO2 formation in long-term during the life time of the vehicle. Passive regeneration of diesel particulate filter (DPF) might require higher precious metal loading; however, active regeneration is also implemented in a Euro-VI vehicle, where the role of DOC is just to extend regeneration intervals by passive regeneration at higher temperatures and to oxidize injected HC efficiently before DPF and improve the fuel economy. The formation of sulphate is the main reason for keeping platinum loading low in a DOC. Even with the low-sulphur fuels ( 1 or ɸ < 1. Diesel engines are best operated with excess air, with lambda ranging from 1.5 to 2.5, so that small particles that form when the reaction was with only a stoichiometric amount of air are burnt out. By limiting the lambda to 2.5, excessive generation of nitric oxides formed due to higher availability of air is kept under control. mair mfuel AFRactual k¼ AFTstoichiometry FARactual /¼ FARstoichiometric A=F ¼

The maximum temperature during combustion, termed as adiabatic flame temperature, is estimated from the first law of thermodynamics assuming the combustion process as adiabatic, where no heat crosses the system boundary. This temperature is a function of fuel’s calorific value, Air fuel ratio, specific heat capacity of reactants, and the inlet temperatures of fuel and air. The adiabatic temperature is higher for fuels with higher calorific value, with higher inlet air and fuel temperatures, and with combustion at stoichiometric air-fuel ratio. For oils the adiabatic combustion temperature is around 2,150 °C (3,902 °F) for oil (ISO 8178-4:1996; AVL gas exchange and combustion analysis V4.1). At high combustion temperatures inside the engine, due to the reaction between nitrogen and oxygen present in the air, nitrogen oxides consisting of Nitric oxide (NO), Nitrous oxide (N2O), Nitrogen di oxide (NO2), di nitrogen tri oxide (N2O3)

6 Strategies to Control Emissions from Off-Road Diesel Engines

245

and unstable nitrogen pentoxide (N2O5) commonly referred to as NOx, are formed at high temperatures. The oxidation of nitrogen compounds in fuel which is in meagre amounts is an additional source. NO and NO2 emissions are of prime concern to engine emissions. Nitric oxide is a colourless, odourless gas while NO2 is a reddish-brown gas with pungent smell formed by further oxidation of NO and their formation mechanisms are shown in Eqs. 6.2 and 6.3. N2 þ O2 ! 2NO

ð6:2Þ

2NO þ O2 ! 2NO2

ð6:3Þ

The popular extended Zeldovich mechanism for NOx prediction, involving oxygen (O) and hydroxyl (OH) free radicals, has been improved by considering N2O formation during combustion and the principal governing reactions are as shown in Eqs. 6.4–6.11. NOx species (predominantly NO with small amounts of NO2) appear in significant amounts beyond 2800 °F (1540 °C) and the rate of its generation depends on the excess air ratio. It is widely accepted that NO species is formed in the combustion chamber at near stoichiometric fuel-air mixtures. Therefore, for reducing NOx emissions, diesel engines run lean, an equivalence ratio less than unity. O2 ! 2O

ð6:4Þ

O þ N2 ! NO þ N

ð6:5Þ

N2 þ O ! N2 O

ð6:6Þ

N þ O2 ! NO þ O

ð6:7Þ

N þ OH ! NO þ H

ð6:8Þ

N2 O þ H ! N2 þ OH

ð6:9Þ

O þ N2 O ! N2 þ O2

ð6:10Þ

O þ N2 O ! 2NO

ð6:11Þ

The rate constants for NO formation mechanism are given in Table 6.1. The chemical kinetic mechanisms for Nitric oxide species’ formation described in the C–O–H systems apply to CI as well as SI engines. However, for CI engines, fuel injection into the engine cylinder is just before the start of combustion, and heterogeneous species and temperature distribution inside combustion chamber is attributed to the non-uniform fuel distribution during combustion. Ignition delay is defined as the time lag between the start of fuel injection and the first identifiable pressure increase. During this time period, fuel evaporates and forms a mixture with surrounding air that is flammable and that can ignite at many sites simultaneously.

246

M. V. Ganesh Prasad

Table 6.1 Rate constants for the NO formation mechanism (Rizvi Syed 1999) Rate ki ¼ k0i T e a

R1 R2 R3 R4 R5 R6

N2 þ O ! NO þ N O2 þ N ! NO þ O N þ OH ! NO þ H N2 O þ O ! NO þ NO O2 þ N2 ! N2 O þ O OH þ N2 ! N2 O þ H

r1 r2 r3 r4 r5 r6

TA T

¼ k1 cN2 cO ¼ k2 cO2 cN ¼ k3 cOH cN ¼ k4 cN2O cO ¼ k5 cO2 cN2 ¼ k6 cOH cN2

k0, cm3 mol s

a

TA, K

4.93e13 1.48e8 4.22e13 4.58e13 2.25e10 9.14e07

0.0472 1.5 0.0 0.0 0.825 1.148

38048.01 2859.01 0.0 12130.6 50569.7 36190.66

This delay is a function of fuel’s molecular structure, physical properties such as volatility, viscosity, surface-tension, and engine characteristics such as compression ratio, injection timing, and pressure. There are two regimes in diesel fuel combustion (after ignition delay), premixed phase followed by diffusion-controlled combustion phase. During “premixed” combustion regime, the fuel-air mixture around stoichiometric composition spontaneously burns and propagates the flame, while during the diffusion-controlled phase, species interaction between already burnt gases, air, and unburnt fuel vapour-air mixture, leads to random heterogeneous mixtures. During pre-mixed phase, which occurs early in the combustion process, when the mixture is at a higher temperature, leading to higher NO formation and rising cylinder pressure. After reaching peak cylinder pressure, combustion gas temperatures decrease as the gases tend to expand and the charge tends to be dilute, NO chemistry gets frozen. Figure 6.5 indicates the concentration of major species measured in a quiescent direct injection diesel engine using a quick response sampling valve (1 ms open time) (Aoyagi et al. 1980). It can be seen that the instantaneous NO concentration increases rapidly after the start of combustion, up to an inflexion peak point where the instantaneous equivalence ratio drops from rich to lean. At the point of peak NO species concentration, majority of the bowl is filled with flame and incidentally, CO concentration is also maximum. NOx formation is also influenced by the air fuel A/F ratio and engine design factors such as compression ratio, fuel injection timing, and cooling losses. Soot and other deposits in the combustion chamber insulate the heat and thus avoiding the drop of combustion temperatures. Particulate matter is one of the components of diesel exhaust emissions that include solid, semi-solid or liquid particles. They are formed from the high boiling fraction of the fuel when its hydrocarbon molecules lose hydrogen through cracking and thereby become charged. These particles have a solid core of elemental carbon, in an agglomerated form, with other substances like ash formed from the oxidation of the lube oil and lube oil additives, metallic abrasion particles, sulphates, nitrates, silicates, and ultrafine aerosols present in liquid and solid phase, attached to the surface of the core carbon. Liquid particulates consisting of the fuel and the low

247

15.0%

900

10.0%

600

5.0%

300

NOx ppm

CO2 O2 CH4, mole fracon, %

6 Strategies to Control Emissions from Off-Road Diesel Engines

0.0%

0 -20

CO2

0 CO,

20 O2

40

CH4, mole fracon, %

60 NOx ppm

C6H4, Mole fracon, %

1.5%

1.0%

0.5%

0.0% 0

20

40

Temperature K

C6H4, Mole fracon, %

60

Soot

2400

80

2200

70

2000

60

1800

50

1600

40

1400

30

1200

20

1000

10

800

Cylinder Pressure bar

-20

0 -20

0

20 Crank angle, degree

Temperature K

40

60

Cylinder Pressure bar

Fig. 6.5 History of emissions formation in a thermodynamic cycle (Aoyagi et al. 1980)

248

M. V. Ganesh Prasad

boiling fractions of the lubricating oil appear as white smoke during cold starting, idling, and at low loads. As the white smoke tends to cease with the increase in load, soot or black smoke, which is much more persistent, is of great concern. The sulphur derived particulates are directly proportional to the fuel sulphur content. At high combustion temperatures, fuel sulphur is converted to SO3 and thereby to SO4 and built together with the bounded water the sulphur fraction. A sulphur content of 0.05% leads to a sulphur derived particulate emission of about 0.01 g/kW h which is 50% of the Euro-IV PM limit value. The particle number and size distribution were also used to assess the environment impact of the diesel fuel exhaust gas emissions in addition to the particle mass. The size of these particles is widely varying and can be in the order of nanometre till millimetre. At low loads, soot forms during the initial stage of combustion both in the lean flame zone and the spray core and diminishes as the combustion proceeds. However, at high loads soot formation in the spray core is persistent and is due to the pyrolysis reactions occurring because of oxygen deficiency. Fuel deposition on the cylinder walls can also result in soot for the same reason. Soot is not a problem if it oxidizes during combustion to carbon dioxide. However, if the rate of its formation is greater than the rate of its oxidation, it can be released into the environment as black smoke. This situation arises when higher output is desired and the fuel is injected faster. While particulate matter can lead to irritation of human breathing system, it is the particles of smaller size that are generally invisible to the naked eye, that cause serious health effects. Fine particulates can also cause haze, thereby reducing visibility.

6.7

Impact of Engine Design and Operating Parameters on Emissions

Engine design and operating variables that impact emissions of a diesel engine include the following: – – – – – – –

Air fuel Ratio Swirl ratio Turbo charging—Intercooling Intake Air Temperature Humidity Fuel properties Piston design

– – – – –

Crankcase ventilation Fuel Injection characteristics Exhaust back pressure Compression ratio Coolant temperature

6 Strategies to Control Emissions from Off-Road Diesel Engines

6.7.1

249

Air-Fuel Ratio

The effect of the air-fuel ratio on hydrocarbon emissions is a function of the engine torque. At low loads and no-load speeds, increasing air fuel ratio leads to a rise in hydrocarbons emissions, because fuel concentration in the core is small and most of the fuel is in the lean flame out region while under moderate loads, the HC and oxygenates concentration decrease as there is more heat and fuel burns more efficiently. If the fuel air mixture is increased further owing to the higher loads, the combustion of the fuel in the spray core becomes deficient in oxygen leading to higher HC emissions. At such conditions, NOx begins occur in spray core and at positions within the cylinder where air fuel ratio is rich, even if it is lean flame region. Overall, the concentration of NOx in this region is higher than in the spray core where there is a greater possibility for NOx to undergo reduction reaction due to the presence of fuel hydrocarbons. It is observed from real time data that the rate of NO formation is higher for the fuel rich mixtures than for the stoichiometric or the lean mixtures due to higher temperatures and pressures. Figure 6.6 illustrates the impact of the air fuel mixtures of varying composition on emissions. The CI engines use leaner mixtures with air fuel ratio of 25:1 and they generate high combustion temperatures, which is primarily because of the higher combustion pressures. Due to the difficulties in modifying the piston engine design to improve the emissions quality, a number of alternative techniques are used to alter the combustion process itself. Diesel emissions contain HC, CO, NOx, aldehydes, particulate matter PM. The presence of both HC and formaldehyde is related to the incomplete oxidation of the fuel at high loads, which require burning

Fig. 6.6 Impact of air-fuel ratio on emissions (Rizvi Syed 1999)

250

M. V. Ganesh Prasad

Fig. 6.7 The effect of air fuel ratio on MEP and SFC (Rizvi Syed 1999)

rich combustion mixtures to generate power. The smoke particulates are also an outcome of the inefficient combustion. Their formation is hard to control, even by using lean combustion mixtures AFR of 30:1. Thus, the points of maximum power a function of MEP and economy are shifted toward the higher AF ratios (23:1) as depicted in Fig. 6.7. For a 16:1 compression ratio engine, an AF ratio of 22:1 or 23:1 appears to be optimal with respect to smoke and fuel consumption.

6.7.2

Swirl Ratio

Swirl ratio is a characteristic of the engine as the ratio of rotational speed of air, measured using a paddle wheel, to that of the engine speed during suction stroke. Intake port design induces swirl motion in the air during suction stroke that improves combustion (Arcoumannis et al. 1993) by generating turbulence at compression TDC due to swirl–squish interaction. This interaction is a strong function of combustion bowl geometry and it is intense in re-entrant bowl. Near to compression TDC, swirl and turbulence further increase while the in-cylinder air is squeezed into the bowl while conserving angular momentum, Fig. 6.8. Additionally, at this point, increase in turbulence intensity can also be due to squish and reverse squish motions. Consequently, two peaks in turbulence intensity, before and after TDC, are observed (Thundil Karuppa Raj and Manimaran 2012; Kondoh et al. 1985) and these will improve combustion efficiency, increase NOx emissions and reduce soot and HC emissions (Dent 1979; Saito et al. 1986). HC emissions reduce with swirl till a threshold value beyond which they increase due to

6 Strategies to Control Emissions from Off-Road Diesel Engines

251

Fig. 6.8 Swirl and Tumble at different crank angles (Thundil Karuppa Raj and Manimaran 2012)

Fig. 6.9 Peak NOx and Soot emissions at different swirl ratios (Thundil Karuppa Raj and Manimaran 2012)

growth in lean flame out region. The peak cylinder pressure, temperature, and heat release rate also increase with swirl ratio due to increase in turbulent intensity, which reduces ignition delay thereby decreasing soot emissions and increasing NOx emissions as shown in Fig. 6.9.

6.7.3

Turbocharging—Inter Cooling

A turbocharger is a device that uses untapped enthalpy of exhaust gases to drive a turbine which in turn operates a compressor to pump more air into the combustion chamber by increasing density of the intake air and thereby improving volumetric efficiency. Fuel injection quantity can be proportionately increased to generate more power output and improve combustion efficiency, overall fuel economy and power to weight ratio. In general, greater the amount of air, the greater is the power output. Since work is done on the intake air by these turbochargers, the charge temperature increases thereby leading to a decrease in air density and increase in the amount of NOx emissions while encountering a power drop. This problem can be overcome by

252

M. V. Ganesh Prasad

cooling the intake air in a heat exchanger, intercooler, before allowing it to enter the engine intake manifold. Turbo charging however has no significant effect on HC and CO emissions.

6.7.4

Intake Air Temperature

As mentioned earlier, higher intake air temperatures will result in higher combustion temperatures and NOx emissions while having drop in power. For fuels with high volatility, spray penetration reduces with increase in air temperature forming fuel rich zones near the injector nozzle and the decomposition reactions at high temperatures accelerate the formation of particulate emissions. On the other hand, low volatile fuels have rapid oxidation reactions resulting in lower PM emissions. Low air intake temperatures will hinder fuel from mixing with air and leading to misfiring, which results in the formation of HC, CO, and the particulate emissions. A temperature controller can assist in regulating the temperature of the intake air for best engine performance.

6.7.5

Humidity

Its effect on the amount of NOx relates to its effect on the flame temperature. High humidity lowers this temperature and hence the amount of NOx formation. The results show that an increase in humidity reduces NOx emissions by 3–14% under test conditions. Conversely, humidity does not have any impact on smoke, CO and HC emissions (Asad et al. 2013).

6.7.6

Diesel Fuel Properties

Fuel related parameters that impact diesel vehicle emissions include the fuel’s cetane number (Index), density, volatility, distillation range, viscosity, the concentration of aromatics and sulphur, and the additive usage for avoiding deposits. The impact of each of the above said properties is discussed below.

6.7.6.1

Cetane Number

Cetane number, CN is a property of the diesel fuel that indicates the speed of the combustion within the cylinder and is inversely proportional to the ignition delay. CN of a fuel varies between a scale of zero and 100 with methylnaphthalene at ‘0’ and cetane hexadecane at 100.

6 Strategies to Control Emissions from Off-Road Diesel Engines

253

Fig. 6.10 Variation in cetane number of different fuels (Rizvi Syed 1999)

The higher the cetane number, the shorter is its ignition delay and hence the fuel is more easily combustible leading to higher combustion efficiency and low hydrocarbon emissions. However, due to high cylinder pressures favouring fuel cracking, particulate matter emissions can increase (Cataluña and da Silva 2012). An increase in 5% of CN can lead to an increase in PM emissions by 35–40%. Low cetane number fuels lead to misfiring, undesirable engine deposits, high HC emissions, and rough running. Modern diesel fuels have a minimum cetane index of 40 as per EN590 or ASTM D975. The cetane number improvers can be used however, they do well if the natural cetane number of the fuel is low, see Fig. 6.10. Fuel with low natural cetane number, has the smallest response to cetane improving additives.

6.7.6.2

Density and Aromatics

Fuel density is a property of the fuel that is related to its energy content entering the engine. A reduction in the density of fuel can lead to can lead to a decrease in NOx emissions in mechanical fuel injection engines but not in the modern engines using electronic fuel injection. Reducing the concentration of aromatics in fuel can greatly assist in the reduction of NOx and PM10. The EPA recommends a maximum aromatics content of 10% in the fuel.

6.7.6.3

Volatility

A typical diesel fuel volatility curve, Fig. 6.11, indicates the temperature at various boiling fractions, which affect different performance parameters such as flash point, startability, white smoking tendency, and cold flow properties, except the cetane number. T95, the temperature where 95% of distillation is achieved under specified test conditions (ASTMD86), when reduced can decrease NOx emissions, while increasing hydrocarbon and CO emissions, and without affecting PM10.

254

M. V. Ganesh Prasad

Fig. 6.11 Diesel fuel volatility (Rizvi Syed 1999)

6.7.6.4

Diesel Sulphur

The sulphur present in the diesel fuel converts to sulphates in the exhaust, it can lead an increase in the PM emissions. There is a strong correlation between the fuel sulphur and the PM emissions and hence concentration limits in diesel for sulphur content has been progressively decreasing since 1982 (Lakshminarayanan and Nagaraj Nayak 2011). In the current scenario, the allowable sulphur concentration across the world varies from 10 to 500 ppm. At 300 ppm sulphur, the sulphate particles make up to one-tenth of the PM emissions from an engine releasing 0.15 g/kWh of particulate emissions. Reducing the concentration of sulphur in diesel fuel to ultra-low levels can lead to other issues such as drop in lubricity, compatibility issues with softs seals, reduced oxidation stability, low-temperature operability, and drop in conductivity which needs to be addressed by introducing new additives in diesel.

6.7.7

Deposit Control Additives/Cleanliness Agents

Fuel and/or engine oil can cause deposits on the engine components and the amount of deposits depends on the engine design, composition of fuel and oil, and the operating conditions. These deposits can form on critical component like injector which can adversely affect the fuel spray pattern that in turn can lead to an increase in emissions and loss of fuel economy. Detergents are used as additives in fuel and oil to maintain a desired level of cleanliness of the critical components.

6.7.8

Piston Design

Pistons with low crevice volume and smaller first lands have low lubricant consumption resulting in lower HC and particulate emissions (Lakshminarayanan and

6 Strategies to Control Emissions from Off-Road Diesel Engines

255

Nagaraj Nayak 2011). Similar results are seen on high speed engines with articulated pistons where the skirt is of aluminium and the crown is of iron or a ferrous alloy which allows the clearance to be small.

6.7.9

Crankcase Ventilation

This involves directing the blow by gases, that leaked off through the piston rings and containing volatile hydrocarbons, from the crankcase back into the combustion chamber as shown in Fig. 6.12. These systems use air intake depression or pressure drop to direct hydrocarbons into the combustion chamber. In this process, unburnt hydrocarbons burn away reducing emissions released and improving the fuel efficiency.

6.7.10 Fuel Injection Strategies As described earlier, in diesel engines, fuel is injected inside hot compressed air as the piston reaches to compression TDC. In order to burn, fuel must mix thoroughly with the surrounding air. Conventional diesel engines operate with an excess air

Fig. 6.12 Closed crankcase ventilation (www.dieselnet. com)

256

M. V. Ganesh Prasad

factor since proper fuel air mixing is difficult and limit the formation of the particulates emissions. As a result of this strategy, the engine’s power output is limited and injection variables viz., high injection pressures, more number of sprays, optimized orientation of the nozzle in the cylinder head, and high swirl ratios can compensate and support in rapid fuel air mixing. The mixture composition, injection timing and sequence, spray pattern, degree of the fuel’s atomization, and turbulent intensity are important from the emissions perspective. The combustion mixture composition is a function of the air fuel ratio, injection pressures, spray pattern, number of sprays, and the injector tip design. At high injector pressures the fuel is in a highly atomized state mixing with air more thoroughly, increasing the homogeneity of the air fuel mixture, and is less likely to form soot and thereby improving combustion efficiency. The resultant mixture burns rapidly and lessening hydro carbon, black smoke, and particulate emissions substantially. However, this process results in an increase of NOx emissions. Carbon deposits on the injector tip will damage the degree of atomization increasing the soot emissions. Deposit control additives can be used to overcome this concern based on the requirement. High rates of injection tend to increase combustion temperature and eliminate soot emissions. Since the formation of the soot is also related to the oxygen deficiency, an improper spray design or pattern can aggravate the situation. Injection timing also termed as the start of injection (SOI) needs to be precisely controlled since a small change in injection timing has a large impact on emissions. Advancing the injection timing in general helps in reducing hydrocarbon and particulate matter emissions while increasing NOx formation due to higher temperatures during combustion and the longer fuel residence time. Other disadvantages of this are excessive combustion noise and an increase in thermal and mechanical stresses. Excessive advancing increases the ignition delay and the fuel vapour and droplets have more time to interact in the swirling air thereby falling in the lean flame out region and leading to higher HC emissions. On the other hand, retarding the SOI reduces NOx by lowering the peak temperatures while increasing smoke, particulate matter, hydrocarbon emissions and fuel consumption. Distinct injection strategies such as pilot injection and rate shaping injection are used to further refine the quality of combustion. Pilot injection is similar to the stratified charge principle used in the gasoline engines wherein a small quantity of fuel is injected into the cylinder prior to the main injection event. Pilot injection would be between −10 and 0°, to initiate combustion, followed by the main fuel charge between 0 and 10°. Rate shaping injector helps in controlling the flow by adjusting the needle valve lift positively using piezo actuators in combination with injection pressure control to achieve the desired rate. Control of the injection rate shape is well known for its effectiveness in exhaust emissions reduction, fuel consumption, and combustion noise (Tanabe et al. 2005) with up to 25% NOx reduction. As this type of flow is continuous, the engine experiences less hesitation and noise. After injection can happen due to leaks at needle valve fuel after the main injection has taken place. The lower the amount of the after injection, the lower is the amount of smoke.

6 Strategies to Control Emissions from Off-Road Diesel Engines

257

All the above injection strategies usage can be easier with the use of electronic control systems, which use sensors to monitor the functioning of the various components and automatically correct in case of malfunctioning. Common monitoring parameters such as engine speed, load, coolant and exhaust gas temperature, intake air temperature, and pressure are to be monitored. A microprocessor is placed to interpret the sensor signals and control the electronic fuel injectors. Injector sensors based on movement of the injector needle during injection are very precise and hence can improve the combustion quality. This involves the control of the fuel quantity, timing, injection rate, and preciseness of the fuel delivery during each engine cycle for the instantaneous engine conditions. Fuel leakage, dribble or after injection, which directly contributes to HC emissions, is usually avoided by keeping the volume of the fuel in the injector nozzle low. On board diagnostics also help reduce in-use emissions resulting from malfunction by early detection and signalling repair. Injection process in diesel engines can be direct injection (DI) or indirect injection (IDI) as shown in Fig. 6.13. In direct injection engines, fuel is directly injected into the combustion chamber above the piston after the compression stroke followed by fuel ignition. In indirect injection engines, fuel is injected into a pre-chamber, which is connected to the cylinder through a narrow passage. During compression stroke, hot air which is forced through this passage generates a strong swirl inside the pre-chamber, where the fuel is injected and the ignition occurs. The rapid expansion of gases forces them into the cylinder where further burning takes place. IDI engines suffer from harder starting and lower efficiency, up to 20%, but run at high engine speeds and hence chosen for smaller automotive and light duty truck applications. In IDI engines compression ratios are higher (23:1) than DI engines (17:1) and hence generate higher combustion temperatures, and therefore produce more NOx.

Fig. 6.13 a Direct injection process; b indirect injection process (Rizvi Syed 1999)

258

M. V. Ganesh Prasad

6.7.11 Exhaust Back Pressure It is the pressure exerted by the exhaust gases in the exhaust system on gases exiting the engine. The extent of the back pressure determines the amount of the exhaust gas retained in the combustion chamber. If the back pressure is low and does not lead to substantial dilution of the combustion mixture, combustion enhancement occurs, thereby lowering the HC emissions. This is because the tail end exhaust gases, primarily responsible for the back pressure, are rich in HC and burn with the fresh charge. If, however, the back pressure is high and charge dilution occurs, inefficient combustion will result, which will lead to higher HC emissions. The charge dilution, on the other hand, will decrease the amount of NOx. Valve timing events are timed in an internal combustion engine so as to obtain maximum power. Conceptually, the intake valve should open when the piston is at top centre and close when the piston is at bottom centre. The exhaust valve should open when the piston is at bottom centre and close when it is at top centre. For high-speed engines, exhaust valve closing and intake valve opening is overlapped to obtain a higher power output. Overlapping involves two valves to be open simultaneously. This strategy takes the advantage of fluid dynamics to scavenge the residual gases more effectively, allowing a greater amount of fresh charge to enter the cylinder, therefore resulting in increased output (Benajes et al. 1996). The degree of overlap varies among engines and its effect on emissions is analogous to that of the exhaust back pressure; that is, it primarily affects HC emissions. It lowers them at very low overlap 2°, beyond which they increase. It has no effect on CO emissions, unless the combustion mixture becomes richer at which time it leads to increased emissions.

6.7.12 Compression Ratio It is the ratio of instantaneous cylinder volumes at the beginning and end of the compression stroke. As the compression ratio increases, the air temperature at the end of the compression stroke also increases. In general, higher values of compression ratios lead to higher thermal efficiency and thus improved fuel economy. While in diesel engines greater compression ratios are preferred for supporting auto ignition, it leads to knocking and hence not desirable in gasoline engines. The side effect of too high a compression ratio in a diesel engine is the increased HC emissions and the increased formation of NOx. Compression ratio affects emissions by influencing the surface to volume ratio (s/v). A decrease in the compression ratio leads to a large decrease in the s/v ratio, hence a decrease in HC emissions, but at the expense of lowering thermal efficiency and engine power. While the lower thermal efficiency is undesired, it is beneficial in lowering the HC emissions by increasing the exhaust temperature that facilitates oxidation of HC to water and carbon dioxide in the exhaust system (www.dieselnet.com). One strategy that is being pursued to boost the output of an engine, while maintaining the decrease in HC, is to alter the compression ratio

6 Strategies to Control Emissions from Off-Road Diesel Engines

259

according to the operating situations. Combining this strategy with super-charging or turbo charging has the advantage of even a greater increase in output.

6.7.13 Coolant Temperature The coolant’s influence on the amount of HC, CO, and NOx arises from its ability to affect combustion temperature by way of the combustion mixture. High coolant temperatures lead to lower HC and CO emissions but higher NOx. Low coolant temperatures, on the other hand, will cool the cylinder walls and the combustion mixture more effectively than the high coolant temperatures; hence they will result in lower NOx.

6.8

NOx Control Strategies

With the understanding on the phenomenon of NOx formation and the effect of engine design parameters, it can be realized that optimising engine operating parameters and fuel composition has its limits in complying with increasingly stringent emission standards. However, further research is going on with the use of electronic fuel management and exhaust after treatment technology (thermal afterburning and catalyst assisted reduction of pollutants). Thermal afterburning involves holding of the exhaust gases in the presence of excess air at high temperatures to complete their oxidation to CO2 and H2O and with no effect on NOx. Catalyst assisted reduction of pollutants, as the name indicates, needs the usage of catalyst for the chemical reactions that occur beyond a temperature to reduce pollutants. However, during engine warm up or cold start, catalytic conversion is ineffective and can be assisted by thermal after burning technology to control HC, CO and NOx emissions. Metal hydride cold start heaters using hydrogen gas can help attaining suitable operating temperatures rapidly.

6.9

SCR

Modern diesel engines use two popular NOx control strategies viz., Selective Catalytic reduction (SCR) and exhaust gas circulation (EGR) combined with better fuel delivery timing and injection rate control, increased injector and cylinder pressures, advanced combustion chamber designs, and particulate filters. SCR is an approach involving the reduction of NOx within the exhaust system, rather than in the combustion chamber, wherein a nitrogen compound with N–H bonds, a gaseous reductant typically ammonia or urea, is injected in the form a fine spray into the exhaust system, in a controlled manner such that its quantity is proportional to the

260

M. V. Ganesh Prasad

Fig. 6.14 Schematic diagram of SCR mechanism

amount of NOx. A schematic of exhaust system with SCR is illustrated in Fig. 6.14. There are storage issues with anhydrous ammonia while aqueous ammonia is relatively safer. Urea is one of the safest reductants to store, but it must be converted to ammonia for NOx reduction process. When urea is injected into the tail pipe, due to the high exhaust temperature, it decomposes to isocyanic (HNCO) acid which reacts with water vapour to form ammonia (NH3) and carbon dioxide. This reaction is followed by a reduction reaction and supported by a catalyst, leading to the formation of diatomic nitrogen, (N2), and water, (H2O) (Kumar et al. 2009), as shown in Eqs. (6.12–6.15). These catalysts are typically oxides of base metals (vanadium, molybdenum and tungsten), zeolites, or other precious metals. Zeolite catalysts can operate at higher temperatures, have more thermal stability and are expensive as compared to Vanadium and Tungsten catalysts. Though the most favourable range of temperature for the above reaction is 630 to 720 K, the reaction can be acceptable between 500 to 720 K but with longer residing time. The minimum temperature for an effective chemical reaction is a function of the fuel, exhaust gas composition, and the catalyst geometry. 4NO þ 4NH3 þ O2 ! 4N2 þ 6H2 O

ð6:12Þ

2NO2 þ 4NH3 þ O2 ! 3N2 þ 6H2 O

ð6:13Þ

NO þ NO2 þ 2NH3 ! 2N2 þ 3H2 O

ð6:14Þ

The reaction for urea instead of ammonia is: 4NO þ 2ðNH2 Þ2 CO þ O2 ! 4N2 þ 4H2 O þ 2CO2

ð6:15Þ

SCR system involves high initial cost with a tank that typically has level, temperature, and urea quality sensors designed to endure freezing and thawing environment. Urea injection quality, injection temperature and mixing are critical to ensure proper functioning. Injecting urea at a temperature 40% of kms/L of gasoline). Considering the dry combustion of gas, it is essential to use

9 Conversion of Diesel Engines for CNG Fuel Operation

343

high temperature low wear alloy material for valve seat in case of bi-fuel engine to avoid valve seat wear. This is not possible in aftermarket retro-fitment, which may end up in valve seat failure when operated in gas mode, after covering certain kms.

9.3

Conversion of Diesel Engine for CNG Fuel Operation

The diesel engines are compression ignition engines. The temperature reached by the air inducted at the end of compression stroke will be higher than the self-ignition temperature of the diesel fuel, which will ensure the combustion of the well atomized diesel fuel as injected into the cylinder. The excess un-throttled air inducted into the engine is preferred for complete combustion. The power requirement variations of the engine achieved only by adjusting the diesel injected quantity. For the CNG operation, it is necessary to convert the compression ignition engine to a spark ignition engine, since the natural gas has self-ignition temperature much higher than diesel fuel. The properties of natural gas compared with gasoline and diesel are given in Table 9.1. A high energy spark is essential for initiating the ignition of the gas air mixture. The octane rating of the natural gas is higher than gasoline; hence the engines can operate with higher compression ratios and can deliver better thermal efficiency compared to gasoline engines but always lower than diesel engines. The higher self-ignition temperature of the natural gas needs a high energy spark for starting a sustainable flame knurl in the air fuel mixture. In spark ignition engines, knocking can be avoided by selecting suitable compression Table 9.1 Comparison of fuel properties CNG versus gasoline and diesel Fuel property

CNG

Gasoline

Diesel

Composition

Mostly methane + CO2 + N2 + gaseous hydro-carbons 0.651 (w.r.to air)

Mixture of hydro-carbons

Mixture of Heavier hydro-carbons

0.72–0.78

0.823–0.826

49.4

44.00

42.5

– 120–130

– 91–95

>51 –

540

228–447

Approx. 300

5.3–15 16.9–17.2

1–7.8 14.7

0.6–6.5 14.6

0.24

0.29

–

Density Kg/m3 @ 20 °C Lower heating value (MJ/Kg) Cetane number Research octane number Auto ignition temperature (°C) Flammable limits Stoichiometric A/F Min ign. energy (mJ)

344

G. Jeevan Dass and P. A. Lakshminarayanan

ratio based on the octane number of the fuel. Also, the composition variation of the natural gas from well to well or even from the same petroleum well with respect to time, may require careful study of suitable compression ratio which will avoid knocking even with lowest octane rated gas, this requirement arises from the presence of heavier hydrocarbons in the fuel. Since the octane number of natural gas is around 120–130, it is possible to keep the compression ratio up to 12:1 in case of naturally aspirated engines and 11.5:1 in case of turbocharged engines. The conversion of the diesel engine to CNG engine will eliminate the diesel fuel systems hence it is necessary to control the engine fly-up rpm within the designed limits of the particular diesel engine.

9.4

Combustion Strategy

Lean Burn For operating a CNG engine in lean burn, the excess air factor (lambda) will be kept at 1.3–1.6, the excess air available will reduce the NOx emission, due to lower combustion temperature, Refer Fig. 9.1. However, it may require an oxidation catalyst to reduce the hydrocarbon and CO emissions. This combustion strategy can be useful when the legislative norms do not demand lowest NOx levels. Whereas for the future emission compliance with lowest NOx emissions, it will be necessary to introduce a Selective Catalytic Reaction (SCR) catalyst for meeting the Euro V and Euro VI emissions. Also, it is necessary to take care of faster mixing of air fuel mixture and quicker propagation of combustion, which can be achieved using complicated combustion chamber designs, to compensate the slow burning characteristics of the natural gas. The ignition energy required will be higher to retain consistent spark kernel from the spark plug, adaptation of high energy ignition systems. Considering the requirement of oxygen catalyst, along with SCR and necessity of high energy ignition system with special combustion bowls to ensure the complete combustion makes it difficult to keep the lean burn engines in the low emission mandated markets. As the emission compliance becomes stringent, the complexity of the system along with cost will be increasing exponentially. Lean burn naturally aspirated engines cannot deliver power output equivalent to the base diesel engine whereas it is possible to attain equivalent power output in a turbocharged engine. The thermal efficiency of the engine will be comparable with the base diesel engine with lower engine thermal load as well the exhaust temperature. The close loop control of the gas induction with a feedback from the wide band oxygen sensor helps to maintain the consistency of air fuel ratio in lean burn engines.

9 Conversion of Diesel Engines for CNG Fuel Operation

345

Fig. 9.1 Combustion strategy: lean burn and stoichiometric

9.4.1

Stoichiometric

In the stoichiometric combustion, the engine operated at chemically correct air fuel ratio, Refer Fig. 9.1. The combustion produced pollutants can be converted to harmless emission by using a three-way catalytic converter, which can convert the pollutants up to 95%. The absence of oxygen in the exhaust can aid the reduction of NOx to N2 and O2 and the oxygen can oxidise the CO and HC to harmless pollutants with a three-way catalytic converter. The combustion with chemically correct air fuel ratio, with limited adequate air quantity for the combustion has no excess air for cooling the combustion gases, will lead to higher operational temperatures, means higher thermal load on the engine components. The higher temperature operation will require increased cooling requirements and the higher exhaust temperature will help in the catalytic converter effectiveness but will require special consideration in selecting the engine components’ materials and as well heat shields introduction to keep the under-bonnet temperature within acceptable limits. In stoichiometric combustion the air fuel ratio must be maintained within certain limits (k = 0.98–1.02) to achieve the best possible three-way catalytic converter efficiency. The fine tuning of the gas induction is achieved by a close loop control system operated by an ECU which adjusts gas induction as per the oxygen availability in the exhaust. With the three-way catalytic converter, it is possible to effectively achieve ultralow emission levels. The thermal load of the engine can be controlled using exhaust gas recirculation. The EGR may enhance NOx reduction, but in stoichiometric engine the lower NOx is taken care of by the three-way catalytic converter. Also, EGR will improve the efficiency of the engine by

346

G. Jeevan Dass and P. A. Lakshminarayanan

reducing the pumping losses, which can be used for fine tuning the engine efficiency further. The stoichiometric engine can tolerate gas composition variations; hence it can be used with varying gas compositions. The stoichiometric engines do not require any fast burning aids in combustion chamber, a simple bath tub combustion bowl is sufficient for achieving the required power with ultralow emission compliance. The correct air fuel ratio for combustion can be initiated with normal ignition systems. The transient response of stoichiometric engines will be much better than the lean burn engines, will be at par with modern diesel engines. The lean burn and stoichiometric combustion can be adopted using any of the gas induction systems either carburetion/gas injection.

9.5

Gas Induction System

For achieving sustained combustion with required power output, definite quantity of the air fuel mixture needs to be inducted into the combustion chamber. The gas induction into the engine can be carried out in using a simple venturi or injecting in the manifold or even with in-cylinder direct injection. The venturi-based carburettor system has become obsolete, due to the absence of close control on the air fuel ratio to meet the required emission levels.

9.5.1

Carburetion

The simplest of way to introduce natural gas into the engine is carburetion using a venturi, which will mix the air and the gas, schematically represented in Fig. 9.2. The way of introduction is similar to age old gasoline engines, with advantages of instant homogenous mixture formation due to gaseous form of fuel and HC reduction during deceleration compared to gasoline due to absence of wall wetting and subsequent gasoline evaporation. The simple carburetted gas engines emissions can be controlled by closely monitoring the gas quantity adjustment using a close loop system as represented in Fig. 9.3. By introducing an oxygen sensor (also called lambda sensor) in the exhaust, the gas induction quantity adjusted to maintain required air fuel ratio within required level so that a suitable after treatment device used to achieve required emission levels. The simpler venturi carburettor system has its own limitations in meeting stringent emission norms with transient cycle requirement, due to sluggish response of mechanical parts.

9 Conversion of Diesel Engines for CNG Fuel Operation

347

Fig. 9.2 Carburetted gas engines

Fig. 9.3 Carburetted gas engines with close loop control system

9.5.2

Gas Injection

The gas introduction can be by injection either in single point or multipoint either at manifold or port, which can replace the simple venturi carburettor system. The gas injection system will be electronically controlled and can be more accurate than

348

G. Jeevan Dass and P. A. Lakshminarayanan

(for bifuel application) Fuel selector Switch

Single stage Pressure regulator Filter

Fuse and Relay Unit

Engine Control Multipoint Injection Multi cylinderEngine

Unit

Vehicle communication Link

CNG Tank

Fig. 9.4 Multi point injection gas system

venturi mixing, refer Fig. 9.4. The added advantage of better transient response with complete fuel cut off during deceleration and faster correction of the close loop control will satisfy the modern requirements of emission norms with transient cycles.

9.5.3

Gas Injected Engines with Close Loop Control and On-Board Diagnosis

The addition of on board diagnosis to curb the emissions in case of any components or system failures without any manual interruption can be achieved by calibrating the engine to limit the power output in case of any failure or malfunction of any gas and ignition systems components. The schematic system of sensors and actuators as well as the electronic control unit is shown in Fig. 9.5. The emission compliance can be met with single point or multi point gas injection systems.

9 Conversion of Diesel Engines for CNG Fuel Operation

349

Fig. 9.5 Complete multipoint gas injection system with on board diagnosis

9.6

Ignition System

9.6.1

Distributor Type Ignition System

9.6.1.1

Ignition in the CNG SI Engine

The ignition system for a CNG engine should provide enough high voltage for starting the flame front. Considering the high self-ignition temperature of natural gas, the high voltage requirement is more than a normal gasoline engine. The lean burn natural gas engine will require further increase in the secondary voltage to ensure reliable flame front knurl development. The design of the ignition system for any SI engine varies according to • Ignition trigger • Timing adjustment • Distribution of high voltage

9.6.2

Ignition Trigger

The ignition trigger, normally known as ignition timing for a natural gas engine requires adjustments with respect to speed and load of the engine. As the speed increases the timing advancement requirement increases due to reduction in absolute time scale. Even though this is a requirement of any SI engine, the magnitude varies in natural gas engine due to slower burning characteristics of natural gas. The

350

G. Jeevan Dass and P. A. Lakshminarayanan

load dependent of ignition timing is almost same as other SI engines, the lighter loads requires advancement of ignition timing. In natural gas engine other parameter which influences the ignition timing is the gas composition. Since the natural gas composition may vary time to time and well to well the ignition timing adjustments also vary. The presence of higher hydrocarbons like propane and butane (even though the percentage is low) may influence the ignition timing requirement. In the similar way the higher presence of inert gases like CO2 (particularly in coal bed methane) and N2 will require ignition timing adaptation. For finalising the ignition timing of a natural gas engine, it is essential to test the engine with different gas composition fuels, in particular with the best gas (highest methane content with lower higher hydrocarbons and inert gases) and worst gas (lowest methane content and higher content of higher HCs and inert gases) possible in a country and if necessary to adopt two ignition timings for the engine. To overcome this there is an international reference in Europe G21 and G25 for finalising the ignition timing mapping during engine development.

9.6.3

Timing Adjustment

The ignition timing adjustment is possible in an engine and can be incorporated with respect to speed of the engine using a centrifugal advance mechanism in case of mechanical system. The adjustment with respect to load of the engine is possible using a vacuum control unit; the inlet manifold vacuum provides an accurate index for the current engine load in case of simple mechanical systems. In case of electronic system, the same functions can be brought into effect by means of signal from the engine speed sensor located in the flywheel housing, provides the instantaneous and accurate input. For timing adjustments with respect to load can be incorporated using a manifold absolute pressure sensor. It is possible to further improve the accuracy of ignition timing with respect to gas air mixer in case of electronic unit by means of combination sensor for manifold pressure and temperature sensor. The electronic adjustment of ignition timing is more accurate and faster transient response is feasible compared to mechanical systems. This will help an engine designer for achieving perfect combustion and hence lower emission possibilities. The timing adjustments required for the stoichiometric engines may be equivalent to any other SI engines whereas for the lean burn engine the accuracy of timing adjustment and advance of ignition timing varies due to lean mixtures. Also, the operation of the lean burn engine in the vicinity of misfiring limit requires accurate calibration of the engine for ignition timing. Also, with the electronic ignition timing adjustment it is possible to bring in the knocking avoidance. The knocking of any cylinder can be detected using a knock seismic mass sensor and the particular cylinder timing can be retarded by a definite crank-angles and once the knock free combustion ensured, it is possible to advance the timing to achieve the thermal efficient operation point—which will be executed automatically by the engine control unit.

9 Conversion of Diesel Engines for CNG Fuel Operation

9.6.4

351

Distribution of High Voltage

For creation of a high voltage an ignition coil is used. It is necessary to fully charge the coil before the high voltage generation. The time required for charging the ignition coil is called the dwell period. The dwell period is constant in case of conventional ignition systems with contact breaker, where the dwell is decided by the cam profile of the ignition distributor. In electronic systems the dwell can be varied and it can be further enhanced by use of individual ignition coil for each cylinder. The energy is stored in an inductive device in rare cases with a capacitance device. The high voltage for creating a spark to initiate the combustion knurl is carried out by disconnecting the primary inductor from the power supply followed by transformation. The disconnection of the primary voltage in the ignition coil will induce a high voltage in the secondary coil, which will be distributed to the cylinder which is in combustion stroke. The distribution of high voltage is possible using a distributor in the conventional systems. The distributor will be driven half the speed of the engine and the secondary voltage from the ignition coil will be communicated by a HT cable to the distributor which will be distributed to the individual cylinder by High tension cables. In case of electronic system, the signal from the crank or cam sensor will provide the position signal of the engine to the ECU, and individual ignition coils will be activated (primary voltage disconnection) and the high voltage will be given to the appropriate spark plug. In case of natural gas engines, the conventional contact breaker distributor system should be avoided, since the high voltage transition between contacts may create a spark and any damaged distributor cap can lead to fire accident possibility in case of gas leakage. Only the distributors with electronic trigger box to be used. The requirement of 3D ignition timing with respect to speed and load of the engine for achieving desired thermal efficiency and lowest emission possible, it is necessary to use electronic ignition systems in future natural gas engines.

9.6.4.1

Contact Breaker Type Ignition System

The simplest conventional mechanical coil-ignition system is triggered by a contact breaker. The current flowing through the ignition coil is interrupted by a mechanical contact, located in the distributor. This type of ignition systems has become obsolete, hence only a brief description is provided.

9.6.4.2

Operation of the Contact Breaker Ignition System

Synchronisation of the secondary voltage with the position of the combustion TDC of the individual cylinder is ensured by the mechanical coupling of the ignition distributor with the cam shaft or any shaft rotating at half the engine speed. The

352

G. Jeevan Dass and P. A. Lakshminarayanan

desired ignition timing can be attained by rotation of the ignition distributor. In case of natural gas engines derived from diesel engine, the drive for fuel injection pump can be used for the location and drive for the distributor. To avoid misuse of the rotation of the distributor by field operators, it is possible to limit the distributor rotation by providing proper limiting clamps in the distributor. The high voltage distribution ensured by the permanently coupled rotor located in the upper section of the distributor shaft, through the high-tension cables. The direction of rotation of the rotor shaft is to be considered, while connecting the high-tension cables, to the individual cylinders as per the firing order, refer Fig. 9.6. As the ignition switch is turned on, the battery voltage is supplied to the primary winding of the ignition coil. When the contact breaker is closed, the current flows through the primary winding of ignition coil to ground. This flow of current through the primary winding of the ignition coil builds up magnetic field in the secondary winding of the ignition coil, thus storing the ignition energy. The current rise is exponential owing to the inductance and the higher resistance of the primary winding. The charging time depends on the dwell angle, which is determined by the design of the cam which activates the contact breaker via the cam follower. At the end of the dwell period, the ignition distributor cam opens the ignition contact and interrupts the coil current. The number of windings in the secondary circuit, the current flowing through the primary circuit, the charging time (dwell angle) of the primary circuit decides the secondary voltage developed by the primary circuit interruption. It is necessary to have a capacitor connected in parallel to the contact breaker to avoid arcing at the ignition contacts. The high-tension voltage generated by the ignition coil transmitted to the centre point of the ignition distributor. The rotor located in the top portion of the distributor communicates the high tension to the outer electrode which will be

Fig. 9.6 The conventional ignition system

9 Conversion of Diesel Engines for CNG Fuel Operation

353

provided to the individual spark plug through a high-tension cable. The high-tension voltage causes a disruptive discharge at the spark plug as a spark. The spark voltage of a conventional ignition system will be 400 V and the spark duration will be 1–2 ms. For a CNG engine the conventional contact breaker ignition system is not suitable due to chance of sparking during operation of the contact breaker. Any crack in the distributor cap may lead to chance of sparking in the circuit and this will become a fire hazard in case of any gas leakage from the gas induction system. Also, the high secondary voltage requirement of natural gas, particularly with the lean burn engine, necessitates the assurance of secondary voltage reserve. The open circuit voltage should be much higher than the operational voltage requirement for sustained combustion at all operating condition of the engine. Any wrong selection of the ignition system will lead to misfiring and finally end up with catalytic converter failure. The continuous operation of the contact breaker ignition system will lead to wear of the contact points, due to constant contact of the follower with the cam, which needs attention and adjustments. In operation, the ignition timing of the engine will be constantly varying due to wear of the contacts. Considering the variation in ignition timing in operation, frequent adjustments and possible spark during operation, this system is not recommended for the natural gas engines.

9.6.4.3

Spark Advance Mechanism

For a given quantity of gas air mixture (in an engine cylinder), the time required for combustion remains almost constant. As the engine speed increases the requirement of advancing of ignition timing to match the combustion time requirement is eminent. For this purpose, a centrifugal advance mechanism can be introduced in the distributor which can advance ignition timing with respect to engine speed, by rotating the cam, refer Fig. 9.7. Even though the desired ignition timing advance can be achieved by selection of springs and design of counter weights, it is not possible to vary the advance curve as per our requirement due to limitation of the mechanical system. The advance curve will be of progressive nature with respect to speed, as shown in Fig. 9.8. Whereas in the electronic system without mechanical distributor the advance curve can be decided at individual points of engine operation conditions, hence the optimised combustion can be achieved effectively. The natural gas engines require advanced ignition timings and the advance curve requirement will be more compared to gasoline, due to slower propagation of flame front. The ignition timing of the natural gas engine for a specific operational point will be advanced compared to gasoline engine and the counter weights and low-tension springs may require for the desired advance curve.

354 Fig. 9.7 Centrifugal advance mechanism—no advance and in operation

Fig. 9.8 Timing control system with respect to engine speed

G. Jeevan Dass and P. A. Lakshminarayanan

9 Conversion of Diesel Engines for CNG Fuel Operation

9.6.4.4

355

Vacuum Advance Mechanism

The vacuum advance mechanism of a contact breaker type conventional ignition system can adjust the ignition timing with respect to the engine power or engine load. The negative pressure (with respect to atmospheric pressure) in the intake manifold near the throttle can serve as the measure for the vacuum advance requirement. The low load engine operating conditions require advanced ignition timings since the gas air mixture burns slowly. At low loads the ignition timing can be advanced using a diaphragm type mechanism, as depicted in Fig. 9.9. As the engine load increases the ignition timing can be retarded and will be matched to the required level at wide open throttle conditions. The conventional vacuum advance mechanism can provide only definite advance curves with respect to load variation and the diaphragm used will have its own hysteresis which also limits the ignition timing advance accuracy with respect to engine loads. The vacuum advance mechanisms found to be effective in case of small engines where the intake manifold vacuum variation is higher, whereas in the larger engines the vacuum advance mechanism found to be ineffective and will not provide any advance with respect to the lower engine loads. In case of the electronic system, the ignition timing advance with respect to lower loads can be effectively incorporated with the intake manifold absolute pressure sensor. The electronic system can provide a three-dimensional ignition timing map as indicated in Fig. 9.11, which ensures ignition timing adjustments to the optimised combustion which will help to improve the engine efficiency with reduction in emission.

Fig. 9.9 Vacuum advance mechanism of conventional system

356

9.6.4.5

G. Jeevan Dass and P. A. Lakshminarayanan

Breaker Triggered Semi Electronic Ignition System

The semi electronic system with transistor is identical to conventional ignition system, whereas the contact breaker can be replaced by an ignition transistor. The transistor provides the switching function of the primary circuit of the ignition coil. When the contact of the breaker trigger is closed, a control current flow through the base of transistor makes emitter and collector of the transistor electronically conductive. The layout of the system is shown in Fig. 9.10. In this condition the current flows through the primary winding of the ignition coil. Once the breaker is open, there is no control current to the base and the transistor becomes non-conductive. This corresponds to the switch off position of the conventional system. The usage of transistorised system can extend the service life of the distributor and as well the primary current can be increased. The higher primary current can help in increasing the ignition coil which will be leading to higher secondary voltage, spark duration and spark current. In this system the contact chatter at high engine speeds can be avoided but the problem of contact breaking spark which is detrimental in natural gas engines cannot be solved.

Fig. 9.10 Breaker triggered ignition system

9 Conversion of Diesel Engines for CNG Fuel Operation

9.6.5

357

Distributor-Less Ignition

The conventional ignition system with mechanical centrifugal and vacuum advance mechanisms can provide simple advance characteristics, which can barely meet the requirement of engine operation. The highly demanding power output, ability to achieve lower emissions do require more flexibility in spark advance for the present and future SI engines. The advanced electronically controlled distributor-less ignition system, a pulse generator signal in the form of engine speed signal is sufficient to trigger the ignition. The manifold absolute pressure signal provides the load signal and a microcomputer computes the required ignition point adjustment and modify the output signal issued to the trigger box accordingly. It is possible not only to add other control parameters, but also can achieve good starting behaviour, fine-tuned idle control along with knock control. It is possible to achieve a 3-dimensional advance curve for the engine operating range, refer Fig. 9.11, which can be optimised on each and every operating point for best power output, fuel economy, lowest emission with sufficient knock margin.

9.6.6

Spark Plug Trials on the Engine

9.6.6.1

Spark Plug Selection

The spark plug suitability for an engine can be decided based on the spark plug trials, which need to be conducted using instrumented spark plugs. The special spark plugs used for the spark plug selection will be of the approximate heat range and configuration suitable for an engine. The operating range of the spark plug decided by the temperature measurement of the insulator tip extended into the combustion chamber. The spark plug electrode temperature should not drop below

Fig. 9.11 Possible 3-dimensional ignition timing with electronic system

358

G. Jeevan Dass and P. A. Lakshminarayanan

500 °C, for attaining the self-cleaning of the electrodes—which is normally checked at 25% of the wide-open throttle load of the engine. This is more critical in case of gasoline engines, due to carbon deposition on the electrodes. In case of the natural gas engines, this is not critical. The maximum temperature of the electrodes should not be more than 850 °C; this will be checked at the wide-open throttle load, at rated rpm of the engine, in the hottest cylinder in case of a multi cylinder engine (Table 9.2). This feature is to avoid auto-ignition of the charge.

9.6.6.2

Operating Temperature of Spark Plug

The spark plug tip electrodes experience the combustion heat during the power stroke and also cooled by the incoming air gas mixture in the induction stroke. Whereas the spark plug outer shell will have the temperature equivalent to cylinder head. It is necessary that the heat absorbed by the electrodes should be dissipated to the outer shell and further to the cylinder head. The operating temperature of the spark plug in a gas engine will be higher in natural gas engines due to the following reasons. • For heavy duty engines, normally the diesel engines converted for the gas operation with minimum modifications in the cylinder head, replacing the injector bore with spark plug threading. The spark plugs in such engines will be located inside the cylinder heads, which requires colder spark plugs—which can dissipate the heat at the earliest to the cylinder head. • The higher operating temperature of the stoichiometric natural gas engines will heat up the spark plug electrodes more, hence will require a colder spark plug. • The operating speed and specific rated load of the engine decides the suitable heat range of the spark plug. The above consideration of the engine operating conditions and engine configuration decides the required heat range of the engine.

9.6.6.3

Selection of Correct Heat Range of Spark Plug

The thermal load withstanding capability of the spark plug is decided by the heat range of the spark plug. Based upon engine design, particularly the effecting heat dissipation from the spark plug and working condition of the engine decides the spark plug requirement of a natural gas engine. Selection of a hotter spark plug will end up in melting of the glass seal between the electrodes in the spark plug and further operation may lead to gas air mixture escaping the cylinder without combustion and may lead to fire hazard. Also, it may lead to auto ignition of the air gas charge, leading to misfire and damage to the catalytic converter. Selection of a colder spark plug will end up with lower tip electrode temperature, which is detrimental in attaining sustained spark in the low

9 Conversion of Diesel Engines for CNG Fuel Operation

359

load particularly cold starting/operating conditions. The deposit of combustion product leading to fouling of electrodes is not a concern in natural gas engines. However, the selection of correct heat range should be of greatest importance to ensure reliable operation of a NG engine.

9.6.6.4

Heat Range Selection for Natural Gas Engines

To maintain the sparkplug operating temperature within safe limits, it is necessary to select a spark plug which will dissipate the combustion heat at the same time maintains the operating temperature of the spark plug. The heat range specification of a spark plug is decided by the surface area of the insulator nose. The spark plugs with longer insulator nose can absorb more heat and dissipate less heat—hotter plugs, whereas the shorter nose can dissipate the heat faster and hence run cooler. The heat range of the spark plug is indicated by the numerical provided in the nomenclature of the spark plug. The lower number indicates colder plugs and higher indicates hotter plugs. The spark plug selection to be carried out by the engine manufacturer and the spark plug supplier with the measurement of the electrode temperatures in the operating engines using the instrumented spark plugs. The thermocouples fitted special spark plugs in the earth and centre electrode will be used to check the self-cleaning temperature at the part load condition of the engine in the coldest cylinder. The maximum temperature will be checked in wide open throttle condition at rated rpm in the hottest cylinder. Also, the spark plug supplier can conduct the ionic current measurement of the spark plug to decide the correct spark plug heat range for the particular engine.

9.6.6.5

Ionic Current Measurement Method for Spark Plug Selection

The ionisation effect of the presence of flames makes it possible to detect the presence of mixture combustion by means of an oscilloscope by measuring the conductivity across the spark plug electrode gap. The time curve of the initial Table 9.2 Spark plug temperature (°C) at full load with thermocouple in earth electrode Spark plug FR7KC Cylinder

Engine speed (rpm) 1400

1600

1800

2000

1 673 691 737 772 2 686 695 746 772 3 683 703 749 739 4 694 711 764 802 5 706 737 789 831 6 765 775 803 835 Cylinder 6 is found to be comparatively hotter than the other cylinders

2200

2800

752 797 801 817 843 859

794 805 821 840 862 856

360

G. Jeevan Dass and P. A. Lakshminarayanan

combustion of the air gas mixture shows characteristic changes referring to the thermal load of a spark plug. By this method it is possible to detect auto ignition irrespective of the temperature which is a characteristic of the engine and the selected spark plug. By this method it is possible to decide the presence of pre-ignition and post-ignition of the engine. Also, by suppressing the ignition spark at certain intervals, it is possible to trace post-ignition and its percentage share in relation to suppression rate, as the combustion temperature raise by advancing the ignition timing deliberately. A change in the ionic current trace in the oscilloscope provides the precise determination, even without suppressing the ignition spark, the transition from post ignition to the beginning of pre-ignition. By this it is possible to evaluate the heat range of the spark plug for a particular engine design. The pre-ignition occurs only when a hotter spark plug is used. The absence of pre and post ignition provides the indication of correct spark plug heat selection for a particular engine application. A typical ionic current measurement of a natural gas engine is given Table 9.3.

9.6.6.6

Designation of Spark Plug Types

Each spark plug manufacturer has special codes for spark plug designation. Normally any designation of the spark plug provides the following data • Type of seat and thread: This specifies the spark plug tightening nut across flat: 14/16/17.5/20.8 mm and the thread size, namely M18
1.5/M14
1.25/ M12
1.25/M10
1. • In case any diesel engine conversion the suitable spark plug size to be selected based upon the space available in the cylinder head. • Version of the spark plug: water proof, surface gap, half thread, motor sports, suppression resistor presence, low power engines. In case of NG engines, it is necessary to select a spark plug with appropriate suppression resistor for meeting the EMI requirements. • Heat range: Ranging from 2 to 13. Normally for a NG engine, it is necessary to select a colder spark plug, whereas the spark plug evaluation with the spark plug supplier is a must. Table 9.3 Ionic current measurement (%), pre/post-ignition, measurements made in cylinder 6 at full load Spark plug FR3K1332-1 FR3K1332-1 FR3K1332-2 a Slight knocking

Advance

Engine speed (rpm) 1400 1600 1800

2000

2200

2800

+0.0a +6.0a +6.0a

0 0 0

0 0 0

0 0 0

0 0 0

0 0 0

0 0 0

9 Conversion of Diesel Engines for CNG Fuel Operation

361

• Thread length: There are varieties of thread lengths available with spark plug manufacturer. The correct thread length to be adopted based on the engine cylinder head design. While converting a diesel engine to NG operation, it may be needed to match the thread length of the spark plug with necessary changes in the machining of the cylinder head. • Electrode versions: Single, double, triple and four electrodes versions are possible. The number of electrodes will not only provide the assured spark in the worn-out conditions, it also increases the spark plug life for a particular engine. • Electrode material: The electrode material mostly decides the life of the spark plug. The copper electrode will provide the lowest kilometres of operation; 15,000 km in case of NG engines whereas the Yttrium electrodes can provide a life of 90,000 km. • Gap specification: The spark plug manufacturers provide a range of spark plug gaps, which is decided based upon fuel used and the electrode material. The copper electrode will require a gap of 0.7 mm for a sustained spark whereas the yttrium electrode spark plugs with high energy ignition system, can operate with spark plug gap as low as 0.35 mm.

9.6.6.7

Sparkplug Threading Specification in the Cylinder Head

While converting the diesel engines for CNG application, it is necessary to replace the diesel injector with a spark plug. Normally it is necessary to consider the wall thickness between the spark plug and the exhaust valve seat area in the combustion face of the cylinder head. Lower valve bridge thickness between the spark plug bore and the valve seats will lead to cylinder head valve bridge crack which will result in either engine overheating or catalytic converter failure. The threading of the spark plug in the cylinder head is to be made as per the standard. The fit must match to the specification. Normally the usage of low sac nozzles in the diesel engines have reduced the diameter of the injectors to the lowest extend. While locating a spark plug in that location, it may need changes in the water jacket core of the cylinder head. Also, it is mandatory to ensure adequate cooling of the spark plug bore for proper heat dissipation and to maintain the thermal load of the spark plug to the required level. Normally if the spark plug can be located away from the rocker cover it is beneficial and if not possible due to the inherent design of the base diesel engine, it is necessary to take care of the lubricating oil entry avoidance with proper sealing into the spark plug bore.

9.6.6.8

Fitting of a Spark Plug

Based upon the cylinder head material, thread size of the spark plug, type of the seat of the spark plug the tightening torque will vary, the recommendations specified by the spark plug manufacturer should be followed. Over tightening of the spark plug will lead to seat failure and also the sealing failure due to crack in the insulation material of

362

G. Jeevan Dass and P. A. Lakshminarayanan

the spark plug. Lower tightening will lead to compression loss and misfiring of the engine. Proper care is a must while fitting the spark plugs in a diesel converted NG engines since the spark plugs are located inside the cylinder head, the spark plug must be inserted slowly so that the spark plug gap between the electrodes should not get disturbed by dropping the spark plug. In case of plug on coils (individual ignition coils located just above the spark plugs), it is necessary to ensure that the spark plug bore diameter is larger enough to accommodate the seal provided in the individual plug on coils. Any deviation to the required dimensions should be sorted out jointly with the coil manufacturer and engine manufacturer and if necessary it may require special seal development for the NG engines, this is purely engine specific.

9.6.6.9

Long Life Spark Plugs

Normally with the copper spark plug electrodes the spark plug life expectancy limited to around 15,000 km. Using the higher-grade material like Nickel–Yttrium it is possible to extend the life up to 30,000–45,000 km. With iridium electrodes it is possible to extend the life further up to 90,000 km. The life expectancy of the spark plug will be lesser than the gasoline application due to higher operational temperature in case of a NG engine. It is necessary to do the field validation with the spark plug manufacturer to specify the spark plug life in a NG engine. Non-adherence of spark plug change period will lead to misfiring due to higher spark plug gap which will end up in catalytic converter failure.

9.6.6.10

Mistakes and Consequences

Incorrect Heat Range Hotter spark plugs will lead to pre-ignition and also the blowing of electrode, which may lead to gas air mixture escaping the cylinder resulting in fire hazard. Colder spark plugs will lead to misfiring at low loads and cold operating conditions, resulting in higher emissions and catalytic converter failures. Incorrect Thread Length Spark plug protruding more into the combustion chamber will lead to expansion of the spark plug fitting threads and it will become difficult to remove the spark plugs. Any forceful removal of the spark plug will lead to cylinder head damage. Lesser protrusion of spark plug into the combustion chamber will lead to misfiring, non-uniform spark and flame front propagation leading to knocking, pre-ignition and misfiring. All will end in engine damage and catalytic converter failure. Normally it is advisable not to use any extra washer under the spin down washer available in the spark plug. Any washer introduction will lead to poor heat dissipation, leading to hotter operation of the correct plug, finally end with engine failures.

9 Conversion of Diesel Engines for CNG Fuel Operation

9.7 9.7.1

363

Engine Modifications Piston

The Direct injection diesel pistons normally have the combustion chamber located in the piston with a compression ratio ranging from 16 to 18.5:1, which is essential for attaining the compressed air temperature above the self-ignition temperature of the diesel. In case of natural gas engines, the compression ratio must be reduced to 11–12:1, to avoid knocking. Normally the reduction in compression ratio can be achieved by increasing the bowl volume. In case of lean burn engines, for fast burning requirements, it may necessary to have complex bowl profiles, whereas in stoichiometric engines, it is possible to achieve the required combustion with normal bath-tub or cylindrical combustion bowl itself. For Naturally aspirated engines (Diesel naturally aspirated engines cannot meet the present emission requirements but CNG can comply with emission norms) the compression ratio can be as high as 12:1 and for the turbocharged engines, the compression ratios normally limited in the range of 11–11.5:1. Typical CNG combustion bowl for stoichiometric engine is given in Fig. 9.12, with the diesel bowl as reference

9.7.2

Cylinder Head

The self-ignition temperature of natural gas is higher than the gasoline and requires spark for initiation of combustion. For locating the spark plug, it is necessary to do the necessary modifications in the cylinder head. It is possible to locate the spark plug in place of the diesel injector in the cylinder head. The recent highly optimized diesel engines are designed with injectors of lowest possible diameters, to increase the intake and exhaust valve sizes. This may pose a problem in locating the spark plug in the cylinder head. Considering the smallest possible size of spark plugs, nowadays the spark plug manufacturers are introducing with M10 threading and long reach (lengthy threaded area) spark plugs. Also, to accommodate the spark plug tightening tool within

Diesel Piston bowl

CNG Piston Bowl

Fig. 9.12 Pistons for diesel and CNG engines—Bowl configurations

364

G. Jeevan Dass and P. A. Lakshminarayanan

Diesel cylinder head with injector

Introduction of spark replacing injector

plug –

Fig. 9.13 Cylinder head modification—spark plug replacing diesel injector

smallest diameter, the hexagon sizes also reduced using 12 faces with reduced across flats. Such spark plug options to be looked into completely for effective modification of the cylinder head. Normally for achieving the equivalent power and torque rating of the diesel engines with stoichiometric combustion, no change required in the port flow characteristics as well the cooling jackets of the cylinder head. But avoidance of the helical intake port with possible higher flow directional port can improve the breathing of the CNG engine and will help to overcome the lower air inlet depression and in turn higher fuel induction energy. The cylinder head with diesel injector and with the spark plug replacing is shown in Fig. 9.13.

9.7.3

Intake Manifold

The intake manifold of the diesel engine normally has no special requirements, in most of the cases it will be a simple box type. The major requirement of swirl can be created in the intake port. But in case of the natural gas engines, it is better to have intake port with necessary profile to ensure proper filling of the cylinder and to mix the gas with air in case of gas injection system—both single point and multi point. In case of carburettor engines also it may necessary to modify the intake manifold for locating the venturi in place. The single point as well as the carburetted engines are filled with air fuel mixture in the intake manifold and higher valve overlap (of the base diesel engine) combined with wrong ignition timing will magnify the chances of misfire. This is completely avoided in case of multi point injection, but it is necessary to avoid cross flow of gas between cylinders can be avoided by extending the gas injection

9 Conversion of Diesel Engines for CNG Fuel Operation

4-cylinder MPFI intake manifold

365

4-cylinder SPFI intake manifold

Fig. 9.14 Intake manifold for MPFI and SPFI CNG engines

near to intake valve. These characteristics become more pronounced while meeting stringent emission norms compliance. The typical MPFI and SPFI intake manifolds for a 4-cylinder CNG engine are shown in Fig. 9.14.

9.7.4

Exhaust Valve and Valve Seat

The valve seat and valve material for the natural gas engines have to be different from the normal diesel engines since the requirement is high temperature wear resistant alloy. The absence of the carbon in the combustion compared to diesel engines will leave the valve seat to wear fast unless used with a wear resistant alloy. Also, the exhaust temperature of the diesel engines normally never cross more than 650 °C normally, whereas in any CNG engines including the naturally aspirated engines, the exhaust temperatures will reach up to 800 °C which is more predominant in turbocharged and intercooled CNG engines. Hence it is necessary to use a special high temperature wear resistant alloy for the exhaust valve seat and valve for the CNG engines. The interference fit of the valve seat and special wear resistant coating of the exhaust valve will help in extending the reliability of the valve train equivalent to diesel engine.

9.7.5

Exhaust Manifold

The thermal loading on the exhaust manifold also will be higher in CNG engines, even in idling the exhaust temperatures will be of the range of 500 °C. The maximum exhaust temperature will normally reach above 750 °C and in a higher power density engine, it may reach even 825 °C. It is necessary to ensure that the exhaust manifold material can withstand these elevated temperatures. Wrong selection of material will lead to not only a component failure, it may cause fire hazard also. Higher molybdenum percentage in the cast iron can resolve this issue without much increase in the cost of exhaust manifold.

366

9.7.6

G. Jeevan Dass and P. A. Lakshminarayanan

Introduction of Gas System in the Engine

The Single point injection system may require some provision in the intake manifold to mount the gas injectors with the throttle body. Normally the gas injectors available are having a definite flow rate only and in case of SPFI it is necessary to increase the number of injectors as per the requirement of gas quantity for the engine. The intake manifold may require only a simple flange change in case of compact engines, may require a new intake manifold with different profile itself. In case of multi point injection, it will be necessary to introduce the definite boss in the intake manifold for mounting the gas injectors. Care must be taken to ensure that the flow from the injectors is always tangential flow to the air inducted into the engine. The perpendicular flow of the injector may obstruct the air flow and may lead to further lower volumetric efficiency, more than the gas displaced air induction. Also, it is necessary in both the cases, the injectors are mounted as per the installation guide lines of the manufacturer and the injector should be able rotate in the axis at the same time it is not having any vertical motion. This will ensure stress free O ring for the injector. The layout has to be prepared for the total gas system and location of the low-pressure systems is decided in such a way that there is no obstruction to the gas flow into the engine.

9.7.7

Introduction to Ignition System in the Engine

The distributor-less, electronically controlled ignition systems are used in the current CNG engines. The ECU will break the circuit of the primary coil of the individual ignition coil and the secondary voltage induced and will be given to the spark plug, using a high-tension cable, based on the input from crank, cam and manifold absolute pressure sensors signals. Further use of plug top coil can avoid the high-tension cable and the ignition coils located on the top of the spark plug, which are connected to the boot, will receive the primary coil circuit breaking signal from the ECU and will provide the high tension developed to the spark plug. The spark ignition timing will be decided during calibration of the engine, in steady state and transient stages. With the electronically controlled ignition timing it is possible to choose the optimum timing for the engine to obtain the maximum power, lowest fuel consumption and emission. The correction of the ignition timing based on the transient operation, atmospheric conditions, engine operating conditions etc. is possible which can be incorporated in the software to ensure lowest fuel consumption and lowest emissions in all operating conditions. With the advanced ignition systems, it is possible to adjust the ignition timing of an individual cylinder of a multi cylinder engine independently. Whenever knocking is sensed in one cylinder of a multi cylinder engine, ECU can immediately retard that particular cylinder to a defined degree and will advance the timing

9 Conversion of Diesel Engines for CNG Fuel Operation

367

once the knocking is absent continuously and can attain the desired ignition timing in that cylinder also. This can help us not only to operate at the optimum ignition timing but also to safe guard the engine from detrimental knocking. Electronic control unit: The electronic control unit, which is a microprocessor which receives signal from various sensors and based on calibration decides the throttle opening based on the demand and provides the gas into the engine by actuating the gas injector and provides correct ignition point to start the combustion (Table 9.4). During deceleration the dump valve will be operated in the turbocharged engines to safeguard the electronic throttle. The following system, refer Fig. 9.15 which provides a generic input to the ECU and outputs provided to the actuators (Table 9.5).

9.7.8

Cooling System

The cooling system of the normal diesel engines comprises of the coolant passages in the engine core components, coolant circulation pump, fan and radiator for the engine coolant. While the diesel lean burn engine always has excess air for combustion and hence the heat dissipated to the engine core components like cylinder liner, crankcase, cylinder head is as per the design parameters. Whereas if we are converting the same engine for the stoichiometric gas operation, the absence of excess air for combustion, increases the thermal loading of the engine to a higher limit, which need to be compensated by additional coolant circulation either by water jacket modifications in the cylinder head or by increasing the circulation by

Fig. 9.15 Function of ECU—inputs (signals) and outputs (actuators)

368

G. Jeevan Dass and P. A. Lakshminarayanan

Table 9.4 Common description of the sensors and mounting location of on the engine No.

Input sensor

Function

Controlling parameter

Installation location

1

Air mass e.g. hot wire anemometer

Gas quantity determination for A/F ratio determination

In the air intake between air cleaner and TC compressor inlet

2

Oxygen sensor Wide band/ narrow band

For close loop control on gas induction As well pre and Post cat provides the health of cat con

In the exhaust pipe, prior to catalytic converter (cat con) and post cat con

3

Coolant temperature sensor

To provide the instantaneous air quantity entering the engine during suction stroke Input regarding the oxygen availability in the exhaust— wideband provides amount of O2 in exhaust whereas narrow band just provides the O2 availability/absence Engine coolant temperature indication

In crank case/ cylinder head in coolant passage

4

Gas pressure and temp

Gas injection quantity and ignition timing correction based on engine warm up, crucial for idling stability etc. Gas injection quantity correction and leak detection diagnosis

5

Boost pressure (TC)/TMAP (NA) Electronic accelerator pedal Knock

For injection quantity calculation/correction

In the intake manifold

Demand indication to ECU

Accelerator pedal

Provide the input to ECU to adjust the ignition timing

In the crankcase

Gas injection timing and spark timing— determination/ correction

Cylinder head

Base input for all operational parameters determination

Flywheel housing

6

7

8

Phase (Cam)

9

Speed sensor

Used to measure the pressure and temperature of inlet gas Measures the air inlet pressure and Temperature in the manifold Converts the driver demand to measurable input to ECU Seismic mass: measuring vibration in crankcase to characterise knocking of individual cylinder Provides signal regarding the position of the engine and in conjunction with speed sensor for locating the first cylinder TDC To sense the position and/or rotational speed (RPM) of the crank

In the gas rail

9 Conversion of Diesel Engines for CNG Fuel Operation

369

Table 9.5 Common description of the actuators and mounting location on the engine No.

Output actuator

Function

Controlling parameter

Installation location

1

Gas injector (SPFI/ MPFI)

To introduce the required gas quantity decided by the ECU

ECU signal to gas injector opening

2

Ignition coil

ECU signal to primary coil

3

Electronic throttle body

4

Dump valve

To induce secondary voltage for starting ignition at the appropriate cylinder through spark plug To introduce the air quantity as per ECU requirement based on the torque demand and corrections as per operating conditions To release the boost pressure during sudden deceleration by relieving the TC compressor out to low pressure side of intake

On intake manifold entry (SPFI) and individual runner of intake manifold (MPFI) On the cylinder head (with high tension lead) or on the spark plug (plug-on coil) In the intake manifold

ECU signal to actuating circuit of throttle body ECU signal to dump valve

Between the pipe connecting the high and low-pressure sides of intake

water pump capacity increase. The coolant circulation increase should be supported by higher heat dissipation in the radiator, with higher heat dissipation capacity and also with increased fan size for better air circulation through the radiator. If the increase in size of the radiator is difficult in the heat dissipating surface area, it can be adjusted by increase in number of rows in the radiator. Whenever the coolant heat dissipation has been taken care, one more requirement of the heat dissipation of the lubricating oil also need consideration, to maintain the lubrication oil temperature within the specified limits, in most of the cases, the oil cooling system may require higher number of plates for this purpose. Normally in the turbocharged gas engines, it is required to use water cooled bearing housing turbocharger. The coolant circulated in the turbocharger also to be considered while taking care of the cooling system design. The only consoling factor in the gas vehicle is the coolant which is used for circulating in the first stage regulator, for avoiding the Joule Thompson effect of gas expansion, will provide some reduction in the cooling system, which will be very meagre.

9.7.9

Catalytic Converter

The after-treatment requirement of a CNG engine will be comparable with the gasoline engine rather to the diesel engine. The stoichiometric CNG engine will require a three-way catalytic converter for treating the exhaust gases. The lean burn

370

G. Jeevan Dass and P. A. Lakshminarayanan

engine will require the oxidation catalyst for HC and CO conversion but the stringent NOx requirement may need a Selective Catalytic Reactor like the diesel engine with ammonia. The stoichiometric CNG engine uses the three-way catalytic conversion and is always the lambda value for this, kept as close as possible to the chemically correct air fuel ratio. It may require a marginal rich bias may help in attaining the required NOx emission. The loading of the precious metal platinum, palladium and rhodium (Pt: Pd: Rh) will decide the conversion of the harmful gases. Methane is a difficult gas to oxidise and if it is needed to meet the total hydrocarbon (THC), including methane, it may need to use heavily loaded catalytic converter. Also, the need for heating the catalytic converter in the cold stage of emission to be taken care for oxidising the methane content in the exhaust. The substrate (metallic/silicon) along with the wash coat and coating of precious metal is most important for stable and durable life of catalytic converter. Any misfiring/combustion irregularities will damage the catalytic converter substrate, which can melt when the temperature reaches 1100 °C whereas efficient conversion will require the temperature range of 750–800 °C. It is necessary to obtain support from the catalytic converter manufacturers to select the correct size and technology. By providing the raw emission data and the required emission level to be met, it is possible to select the substrate size and loading of precious metals. It is necessary to position the catalytic converter as near as to the engine exhaust manifold, to ensure the heating up and attaining the light off temperature at the earliest.

9.7.10 Turbocharger The turbocharger for a stoichiometric CNG engine will be different from the diesel turbocharger due to the variation in the air quantity requirement of a stoichiometric engine to a lean burn diesel engine. The absence of excess air in the CNG engine will lead to excess exhaust temperature. The higher exhaust temperature will need a material change in the turbine housing. Where ever a normal Silicon Molybdenum (Si–Mo) material turbine housing used in the diesel engine may need either a GJV material which can withstand up to 825 °C continuous operation and higher than this temperature may require Ni-resist turbine housing material. Ni-resist material will be with almost 30% Nickel with obvious increased turbocharger cost. The turbochargers for the stoichiometric CNG engine preferably with water cooled turbine housing. The water-cooled turbine housing can reduce the operational temperature of the bearings of the shaft and will provide extended life for the turbocharger. The turbocharged CNG engine will require a dump valve either operated mechanically or with electronic controlled by ECU. The dump valve or bye-pass valve will protect the throttle of the engine, from damage while sudden deceleration occurred in the regular operation of the engine.

9 Conversion of Diesel Engines for CNG Fuel Operation

9.8 9.8.1

371

CNG System Vehicle Adaptation Storage Cylinders

The required CNG for the vehicle operation is stored in cylinders at a higher pressure in the range of 200–250 bars, to ensure the energy density for the regular operation. The cylinders are classified into four categories. The type 1 cylinders which are made of completely steel will impose maximum weight penalty. The type 2 cylinders will be steel liners backed with carbon fibre composite material in which the gas is stored at the same pressure of the type 1 cylinders, with lesser weight penalty. The cylinder liners made of Aluminium with carbon fibre composite material can reduce the weight penalty further, called as type 3 cylinders. The type 4 cylinders are complete composite, where the weight penalty is the lowest. The restriction of using type 4 cylinders in all vehicles is the cost of the cylinders. The CNG storage cylinders normally need to be inspected at definite interval as the explosive department’s recommendations of the concerned country. Any violation of this requirement will lead to major accidents. When a greater number of cylinders is used for storing CNG, it is necessary to ensure that the cylinders laid out considering the safety requirements. If the cylinders are positioned within the vehicle compartments it is necessary to provide necessary venting arrangements to clear any gas leakage from the cylinder valves or joints.

9.8.2

High Pressure Tubes

The high-pressure tubes which are normally used between the storage cylinder valves and the pressure regulator should capable of withstanding at least one and half times the maximum storage pressure. Normally stainless-steel tubes and for low cost applications, cold drawn steel tubes with necessary environment protection is being used in vehicles. When the high-pressure tubes are laid out from the storage cylinder up to the high-pressure regulator, it is necessary to ensure proper routing of the pipes, clamping of the pipes and also providing pig-tails of U bends in the pipes between joints for ensuring minor expansions to accommodate the relative movement of the CNG components. Also, avoidance of pipes near any heat source or providing heat shield for the pipe from the heat source is essential for safe operation of the CNG vehicle in a long run.

9.8.3

Cylinder Valves

The cylinder valves which are normally mounted on the cylinders facilitate the filling of cylinder and as well supplying the stored gas to the pressure regulator

372

G. Jeevan Dass and P. A. Lakshminarayanan

while the engine is operational. The cylinder valves normally provided with provision of connecting gas pipes and also provisions closing manually in some cases with solenoid valves. The cylinder valves also provided with pressure relief devices which can vent out gas whenever pressure or temperature is above the set limits.

9.8.4

Pressure Regulator

For covering more range, the natural gas stored at a high pressure normally at 200 bars, whereas for inducting into the engine with air, it is necessary to reduce the pressure almost to atmospheric in case of carburettor or to a pressure around 5– 8 bar in case of gas injection. The pressure reduction is achieved using pressure regulators, which are operated with springs and diaphragms in some cases with expanding pistons. The reduction in pressure will lead to lowering the temperature of the gas due to ‘Joule Thomson’ effect, which may lead to gas supply blockage when the engine operates at higher loads and speeds. To avoid this reduction in temperature, normally the engine coolant will be passed through the pressure regulator. Normally a nominal flow of one LPM is more than sufficient to avoid this reduction in temperature.

9.8.5

Gas Filter

The gas supplied from the compressor of the filling station may contain oil, which is introduced at the compressor of the gas and some debris including silicon particles. These contaminants are detrimental for the gas injection system. Hence it is necessary to use either a high-pressure filter or in the low-pressure filter of the pressure regulator. It is a necessity to use the filter to avoid debris of gas well (silica) and also the lubrication oil carried away by the gas from the high-pressure compressor. Normally coalescent filters are used, in the low-pressure side with consideration of the pressure drop across the filter.

9.8.6

Layout

The general layout for a single storage cylinder is given below and whenever multi cylinders are used, it is necessary to use close loop multi point filling for reduced filling time. Also, care must be taken for venting out the gas by connecting the pressure relief devices. It is advisable to locate the cylinders in such a way that any possible leakage should not enter the passenger compartment. If located in the boot space of the vehicle without partition from passenger compartment, it is necessary

9 Conversion of Diesel Engines for CNG Fuel Operation

373

to provide special provisions for air circulation in such way that any possible gas leakage will be vent out to the atmosphere.

9.9 9.9.1

Improving Thermal Efficiency Helical Intake Port

The Diesel engines have the swirl creating helical inlet ports for providing swirl motion for the inducted air. The main feature of this swirl is for better mixing of the diesel and air to ensure proper combustion. The swirl creating helical profile of the inlet port by virtue poses flow reduction, the restriction introduced for the better mixing of diesel particulate with air is an unwanted and rather volumetric reducing factor for the Natural Gas engines. The avoidance of the swirl creating helix in the intake port while converting the diesel engine for natural gas operation will improve the volumetric efficiency. The directional port with possible maximum flow coefficient will ensure induction of higher air quantity, which will improve the volumetric efficiency of the engine.

9.9.2

Higher Throttle Opening

With the improved directional intake port, it is possible to go for a wider throttle opening of the most suitable throttle diameter for the engine. Higher the throttle opening will reduce the negative work required for the gas exchange process of the P-V diagram. When the air can be inducted with lesser restriction and also with wider throttle opening will improve the breathing of the engine.

9.9.3

Limited EGR

Due to the negative pressure created in a normal spark ignition engine with the restriction created by the throttle will lead to higher negative work to be spent for the suction of the air into the engine. The negative pressure of the suction stroke in the P-V diagram can be reduced by limited EGR introduction. Care should be taken to limit the quantity of the EGR so that the effect is only to avoid the higher negative pressure creation in the intake stroke and not to deteriorate the combustion. The first two parameters will improve the air quantity inducted into the engine and the third one will improve the negative work spent for suction stroke, will improve the overall thermal efficiency of the engine.

374

9.10

G. Jeevan Dass and P. A. Lakshminarayanan

Simulation and Development of a MPFI CNG Engine

It is possible to simulate a CNG engine prior to actual development. The experience of thermodynamic simulation to develop a turbocharged, multipoint fuel injection engine to meet required emission levels with simple orifice controlled EGR is explained as an example. The thermodynamic model used to optimize the ignition timing and predict the engine performance over the operating range of speed and load while allowing stoichiometric combustion. Optimum turbocharger, diameter of the EGR-pipe and the throttle body size were selected based on thermodynamic simulation. The 3-way catalytic converter and the silencer at the turbine outlet were also simulated. The real engine was developed based on the simulation and the engine performance. Power output and Specific Fuel Consumption, NOx and heat release rate from the model were found to be agreeing well with the experiments on the real engine. Considerable time was saved at the engine test bed because the thermodynamics and gas exchange could be simulated accurately.

9.10.1 Model Description The simulation software solves one-dimensional equations of thermodynamics and flow. The variation in properties like velocity, pressure and temperature along the length of the pipe can be accurately predicted. The circuit diagram for the simulation on the graphic user interface is shown in Fig. 9.16. The air enters into the air-cleaner, CL1 through the system boundary, SB1. The air then gets compressed in the compressor, C of turbocharger, TC1 and goes into the intake manifold, PL1 via air cooler, CO1. A throttle, R1 restricts the airflow according to the demand torque and the stoichiometric air fuel ratio is maintained by automatically adjusting gas injection quantity. Multi-point fuel injection technology (I1, I2, I3, I4, I5, and I6) is used to inject the fuel into individual inlet ports of the manifold. The fuel mixes with the incoming air and enters the cylinders (C1, C2, C3, C4, C5, and C6). After combustion, the exhaust gas enters the turbine, T of the turbocharger, TC1. A part of the exhaust gas taken from exhaust manifold is recirculated back through pipes 30 and 33 into the air intake pipe controlled by an orifice connected to the inlet of the compressor, C. This is called Hybrid EGR technology. The remaining quantity goes through the turbine, T and finally escapes into the environment through system boundary SB2. A catalytic converter CAT1 and a plenum chamber PL2 were connected for maintaining backpressure after the turbine T. MP1, MP2 etc. are the points where the thermodynamic properties are monitored. Combustion Modelling—The homogenous combustion taking place inside an SI engine (Lakshminarayanan 1974) is modelled using Vibe 2-zone model as it can calculate Nitric Oxide (NOx) also. The calculation of ignition delay and burn duration is explained in detail below.

9 Conversion of Diesel Engines for CNG Fuel Operation MP8 3 2

PL1

MP28 6

MP27

R1

375

7

8

9

10

11

MP9 I1

MP1 0 I2

I4

I3

12

13

MP11

15

14

C1

C2

C3

16 C4

2 0 MP26

26

21

25

J3

J2

23

27

33

MP14

MP13

MP23

22

MP24

2 MP6

MP7

C6

J4

30 C O1

17 C5

24

J1

3

I6

MP12

19

18

I5

MP1 5

MP5

TC 1 MP4

MP16

31 J5 MP22

MP2 5

34 J6

CL 1

1

29

SB1 MP1

MP2

CAT1 MP1 8

4 MP17

5

MP19 PL2 MP20

MP3

28 MP21 SB2

Fig. 9.16 Schematic diagram of a six-cylinder CNG engine

Ignition delay and Burn duration—Ignition delay and burn duration of the 135-kW turbocharged engine were calculated using the correlation (Lammle 2005) given below. The ignition delay simulated was found to be matching well with the calculated values as can be seen from Fig. 9.17. Though the burn duration was not comparing well with the calculated values it served as the starting point for the simulation.  rpm 0:3965 rpm0
 BD rpm 0:3663 ¼ IB rpm0

ID ¼ ID0




376

G. Jeevan Dass and P. A. Lakshminarayanan

Fig. 9.17 Ignition delay versus speed

where, rpm o ID BD

Desired speed, rpm Reference state Ignition delay Combustion duration.

Vibe 2-zone model: For engines with external mixture preparation, the selection of a two-zone model where the first law of thermodynamics is applied to the burned zone and unburned zone respectively is possible. The two zones are separated by a flame front. The advantage of this model is that it can also calculate the NOx. The two-zone (Vibe 1970) model is used in conjunction with the Vibe heat release rate function. Start of combustion, duration of combustion and shape parameter are the Vibe Parameters required for modelling combustion using Vibe 2-zone algorithm. The first law of thermodynamics is applied to the unburned and burned zones as follows. dmb ub dVb dQF X dQWb dmb dmBB;b ¼ pc þ  þ hu  hBB;b da da da da da da dmu uu dVu X dQWu dmb dmBB;u ¼ pc  þ hu  hBB;u da da da da da m u p V Q QW a b U BB

Mass fraction Internal energy Cylinder pressure Volume of gas Heat transfer Wall Heat transfer Crank angle Burned zone Unburned zone Blow by.

9 Conversion of Diesel Engines for CNG Fuel Operation

377

The above two equations are the energy balance equations for burned and b burned zones. The term, hu ¼ dm da covers the enthalpy flow from the unburned to the burned zone due to the conversion of a fresh charge to combustion products. Heat flux between the two zones is neglected. In addition, the sum of the volume changes must be equal to the cylinder volume change and the sum of the zone volumes must be equal to the cylinder volume. Vb þ Vu ¼ V dVb dVu dV da þ da ¼ da Solving the above two energy equations using some elementary algebra involving substitutions, an equation for the temperature derivative of the burned zone and unburned zone versus crank angle is obtained. The amount of mixture burned at each time step is obtained from the Vibe function specified by the user. Knowing the temperature derivatives and mass fraction the heat transfer in the burned and unburned zone for each crank angle is calculated and added to get the total heat release rate at each crank angle. For all other terms, like wall heat losses etc., models similar to the single zone models with an appropriate distribution on the two zones are used. Heat Transfer—For the gases burning at high temperatures the main heat transfer would be to the piston, cylinder head and cylinder liner. Heat transfer to the walls is given by: Q ¼ hADT h A D T

Heat transfer coefficient: [J/m2 K] Surface area, m2 Difference Temperature.

The heat transfer coefficient is calculated by the correlation developed by Woschni (1967). The Woschni model published in 1978 for the high-pressure cycle is summarized as follows:    0:8 VD Tc;1  0:2 0:8 0:53 hw ¼ 130D pc Tc C1 cm þ C2 pc  pc;0 pc;1 Vc;1 C1 C2 D cm cu VD

2.28 + 0.308 cu/cm = 0.00324 for DI engines = 0.00622 for IDI engines Cylinder bore Mean piston speed Circumferential velocity Displacement per cylinder

378

G. Jeevan Dass and P. A. Lakshminarayanan

pc,0 Cylinder pressure of the motored engine Tc,1 Temperature in the cylinder at intake valve closing (IVC) pc,1 Pressure in the cylinder at IVC. The modified Woschni heat transfer model (Woschni 1991) published in 1990 aimed at a more accurate prediction of the heat transfer at part load operation: " 0:53 hw ¼ 130 D0:2 p0:8 C1 cm c Tc

(

)#0:8
 VTDC 2 1þ2 IMEP0:2 V

VTDC V IMEP T

TDC volume in the cylinder Actual volume in the cylinder Indicated mean effective pressure Temperature.   2 V T  In the case that C2 pc;1D Vc;1 pc  pc;0  2 VTDC IMEP0:2 the heat transfer coefV c;1 ficient is calculated according to the formula published in 1978. For the gas exchange process, both Woschni models use the same equation for the heat transfer coefficient, which is as follows: 0:53 hw ¼ 130 D0:2 p0:8 ½C3 cm 0:8 c Tc

C3 D cm cu Tc pc

6.18 + 0.417 cu/cm Cylinder bore Mean piston speed Circumferential velocity Gas temperature in the cylinder Pressure in the cylinder.

Later AVL (BOOST, User manual for AVL BOOST, AVL List GmbH) has modified the correlation considering the strong influence of heat transfer during the gas exchange on volumetric efficiencies of the engine, especially for low engine speeds. The heat transfer coefficient is calculated as follows. " h ¼ max

c4 p T din min

0:53 hWoschni ; 130 D0:2 p0:8 c Tc

14 Pressure (Pa) Temperature (K) Pipe diameter connected to intake pipe (m) Intake port velocity (m/s).

(
 )# din 2 c4 jvin j D

9 Conversion of Diesel Engines for CNG Fuel Operation

379

The expression having c4, d, din, min terms in the heat transfer equation takes care of the effect of volumetric efficiency of the engine. The maximum of the two heat transfer coefficients calculated during the gas exchange is taken for calculation of wall heat transfer. Turbocharger: A simple turbocharger model was used for steady state engine performance where the dynamics of the turbocharger (i.e. the variation of the turbocharger speed) were not considered. Furthermore, the turbocharger efficiency is kept constant during the engine cycle. As many test calculations have proven, this model provides good accuracy for steady state engine calculations. It is very convenient to work with this model, as only the mean values for the compressor efficiency, the turbine efficiency, and the mechanical efficiency of the turbocharger must be specified. This reduces the required input dramatically. "

#
k1 p4 k PT ¼ mT :gm;TC :gs;T :cp T3 1  p3 For steady state engine operation, the performance of the turbocharger is determined by the energy balance or the first law of thermodynamics. The mean power consumption of the compressor must be equal to the mean power provided by the turbine: PC ¼ PT and "
k1 # p2 k cp T 1 1 PC ¼ mC gs;C p1 1

where PC PT mc ηs,C cp T1 p2/p1 mT ηm,TC ηs,T T3

Compressor power Turbine power Mass flow rate in compressor Compressor efficiency Specific heat at constant pressure Compressor inlet temperature Compressor pressure ratio Turbine mass flow Mechanical efficiency of the turbocharger Turbine efficiency Turbine inlet temperature.

Incorporating all the models the simulation was run for a period of 40 cycles to attain steady state (Kumar et al. 2008). Initially a full throttle performance at 8 speeds ranging from 2400 to 1000 rpm was simulated. Pressure drops across

380

G. Jeevan Dass and P. A. Lakshminarayanan

air-cleaner, air cooler and silencer were obtained from bench experiments and used in the simulation. A cell size of 30 mm was chosen along the axis of all the ducts. The importance of this engine simulation arises from the fact that the turbocharger, EGR requirement (EGR pipe diameter) and throttle body diameter selection were done based on the simulation results.

9.10.1.1

Selection of Vibe Parameters

To start with, the ignition timing values of a similar naturally aspirated engine were considered as base for the new turbocharged engine simulations. Start of Combustion (SOC) and Combustion duration of NA engine were obtained from simulation. Ignition delay calculated from the difference of Ignition timing and SOC of the NA engine is validated against the model (Lakshminarayanan 1974). Ignition delay hence obtained was added to the Start of Combustion of turbocharged engine obtained from simulation to get the ignition timing for turbocharged engine. With the increase in speed the timing was advanced in order to achieve minimum BSFC. As CNG has high ignition delay (high octane fuel) there is no danger of knocking due to ignition advance. However, a knock sensing and control is used in the real engine developed. The variation of start of combustion with respect to speed is shown in Fig. 9.18.

9.10.2 Selection of Turbocharger The performance of the turbocharger influences the power and BSFC of the engine. In the turbocharger algorithm the equivalent turbine discharge coefficient, compressor pressure ratio, overall efficiency of the turbocharger, compressor efficiency Fig. 9.18 Start of combustion versus speed

9 Conversion of Diesel Engines for CNG Fuel Operation

381

and mechanical efficiency are the inputs required to model the turbocharger. The equivalent turbine discharge coefficient (ETDC) is a function of turbine pressure ratio and exhaust gas flow rate and hence decides the size of the turbine. Pressure ratio across the compressor decides the size of the compressor. The equivalent discharge coefficient of the turbine and compressor ratio specifies the turbocharger. ETDC ¼

Aeff Ageo

Compressor: The airflow required to maintain a stoichiometric A/F ratio of 17.2 is calculated from BSFC, Power and fuel flow rate. So, the pressure ratio required to obtain the calculated airflow is derived from simulation. In this way design of the compressor part of the turbocharger is frozen. The pressure ratio vs flow graph plotted at all operating points is shown in Fig. 9.19. An efficient compressor should be selected such that the whole band will lie in the best efficiency zone. Aeff

pffiffiffiffi rffiffiffi m T R 1 w ¼ 2 p

Aeff Effective area Ageo Geometrical area pffiffiffi m T Mass flow parameter p

R w k dout

Gas constant = 287 J/kg K Pressure function Ratio of specific heats Diameter of turbine outlet pipe.

1.7

Pressure ratio

1.6

1.5

1.4

1.3

1.2

1.1 0

20

40

60

80

Air flow, g/s

Fig. 9.19 Pressure ratio across compressor versus air flow

100

120

140

160

382

G. Jeevan Dass and P. A. Lakshminarayanan

Turbine: The procedure followed for designing a compressor cannot be adopted for turbine, as there is an extra parameter namely the flow through the waste gate, which increases the complexity of the design. For this purpose, the method of turbine design is slightly changed by considering three different turbine housing sizes, 4.92, 5.92 and 6.92 for simulation. Any turbine is designated as “x.yz”, where x is the volute cross section area in cm2 at which the variation in properties is zero, y tells us whether the turbine is waste gated or non-waste gated and single entry or double entry, z tells us about the trim level. Based on the changes in the three parameters the turbine maps will change accordingly. A typical turbine map is shown in Fig. 9.20. sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2 c þ 1 ffi c c c w ¼ c1 pp43  pp43 2 Ageo ¼ p4 dout

In the present study the equivalent turbine discharge coefficient for all the housing sizes were calculated at all points in the map by the formula shown above. The equivalent turbine discharge coefficient hence calculated is substituted and varied to get the assumed waste gate flow. The procedure is followed for all the speeds. The effective area calculated is plotted against turbine pressure ratio in Fig. 9.21. From the simulation it was observed that the Turbocharger 2 with 5.92 housing was suiting well because of its effective turbine inlet area. The 5.92 housing turbocharger is satisfying all the requirements that were defined at the beginning of the simulation. Turbocharger 1 with 6.92 housing was not giving the required torque at lower speeds due to lesser airflow at low speeds and with 4.92 housing turbocharger (Turbocharger 3) the exhaust temperatures were observed to be exceeding 800 °C at lower speeds and BSFC was poor at higher speeds due to high pumping losses. Fig. 9.20 Turbine map, flow versus pressure ratio

9 Conversion of Diesel Engines for CNG Fuel Operation

383

Fig. 9.21 Effective area of turbine versus pressure ratio across the turbine

The variation of boost pressure and turbine inlet pressure with respect to speed are shown in Figs. 9.22 and 9.23 for the selected turbocharger.

9.10.3 Optimization of EGR Pipe Diameter EGR technique is used in this engine mainly to control the thermal loading unlike in diesel engines where it is used to reduce NOx values. In diesel engines exhaust gas replaces part of the incoming air, Fig. 9.24. In a CNG engine it adds to the incoming air. In the present engine the EGR flow is determined by the diameter of the EGR pipe (Fig. 9.25) connecting the exhaust manifold to the inlet of the compressor and the pressure difference across the pipe.

Fig. 9.22 Boost pressure versus speed

384 Fig. 9.23 Turbine inlet pressure versus speed

Fig. 9.24 Exhaust gas temperature before turbine with and without EGR

Fig. 9.25 Brake specific fuel consumption versus speed for different EGR pipe diameters

G. Jeevan Dass and P. A. Lakshminarayanan

9 Conversion of Diesel Engines for CNG Fuel Operation

385

Fig. 9.26 Power versus speed for different EGR pipe diameters

A cost-effective turbine made of cast iron withstands temperatures less than 800 °C. In order to maintain the exhaust temperature below the specified value of 750 ° C, suitable pipe orifice diameter should be chosen. In the present case four different sizes 6.5, 7, 8, 10 were considered and 6.5 mm diameter was finalized among them based on the BSFC, exhaust temperature and Power. The variation in BSFC and Power with respect to EGR pipe diameter is shown in Figs. 9.25 and 9.26. It can be seen from the figure that SFC with 6.5 mm is optimum and the power obtained is higher because of lesser EGR flow compared to other pipes of different diameters. The exhaust gas temperature reduced by 50° compared to without EGR as can be seen from Fig. 9.24. With increase in EGR pipe diameter the exhaust temperature reduces further but at the expense of Power and BSFC. In order to reduce the pressure pulses in the pipes two 6.5 mm pipes were connected in parallel. A Percentage EGR flow of 7–7.5% was observed with 6.5 mm diameter pipes at all speeds. The variation of exhaust temperatures with without EGR pipe diameter is also plotted in Fig. 9.24. The comparison between simulation and experimental values of EGR flow is given in a later section. If the pipe diameter is reduced less than 6.5 mm then exhaust gas temperature comes closer to 800 °C that is not safe for the turbine housing.

9.10.4 Selection of Throttle Body Diameter The throttle body regulates the amount of air flow required based on the signal from ECU. The diameter of the throttle body selection is the critical base parameter for achieving required power output from the engine in its design. The throttle body diameter is decided based on engine’s response to throttle opening. As the throttle body diameter decreases engine’s response improves i.e. variation of power at lower openings would be satisfactory. The constraint is that the diameter could not be less than a particular value as that would result in power loss at higher speeds

386

G. Jeevan Dass and P. A. Lakshminarayanan

Fig. 9.27 Variation of power for different throttle diameters

and when throttle is wide open. For simulation we have considered four diameters 45, 53, 57, 63 mm (among these 53 mm is not readily available). The variation of power with flow coefficient is shown in Fig. 9.27. Throttle body with 45 mm diameter or less will result in loss of Power and hence is ruled out. Throttle body with 53 mm would be the ideal value but it is not commonly available. As the volume of the CNG gas engines produced is less, to develop a new throttle body is expensive. Hence among the remaining variants, throttle body of diameter 57 mm was finalized. But with 57 mm throttle body diameter the power obtained would be more than 135 kW at rated speed. By limiting the throttle opening in the ECU dataset the power is reduced to 135 kW. This was found to be cost effective solution.

9.10.5 Engine Development The turbocharged engine was developed and experiments were carried using the turbocharger, EGR pipe diameter and throttle body diameter selected by using the thermodynamic model. The ignition timing optimised using the model is entered in the ECU dataset. The simulated results were compared with the experimental results obtained. Power, torque, BSFC, exhaust gas temperatures and EGR % were plotted at various speeds. Pressure and heat release rate for various speeds were also plotted against crank angle. As can be seen from Figs. 9.28 and 9.29, the curves matched well with the experimental curves. The maximum PFP obtained was observed at the rated speed and decreasing with decrease in speed. The important observation is that, even though the ignition timing starts at 27° before TDC at 2400 rpm, the peak firing pressure occurs after TDC, which shows the high ignition delay of the fuel. The pressure curve clearly shows that there is no knocking in the combustion

9 Conversion of Diesel Engines for CNG Fuel Operation

387

Fig. 9.28 Comparison of experimental and simulated cylinder pressure versus crank angle diagrams at 1400, 1600, 2000, and 2400 rpm

Fig. 9.29 Comparison of experimental and simulated rate of heat release versus crank angle diagrams at 1400, 1800, 2200 and 2400 rpm

388

G. Jeevan Dass and P. A. Lakshminarayanan

Fig. 9.30 Comparison of experimental and simulated power versus speed

The variation of power with speed is shown in Fig. 9.30. Though the maximum power obtained was 146 kW it is possible to obtain 135 kW by limiting the electronic throttle opening at higher speeds. The torque curve plotted at various speeds is shown in Fig. 9.31. The maximum torque obtained was 606 Nm. The torque curve is flat with maximum torque occurring over a range of speeds from 1400 to 2000 rpm. The torque and Power curves were closely matching with the experimental results. The BSFC curve plotted at various speeds at full throttle is shown in Fig. 9.32. The exhaust temperature by simulation is fairly close to the experimentally observed temperature (Fig. 9.33). The minimum BSFC obtained was 200 g/kWh at 1800 rpm. The maximum deviation between the simulation and experimental values is within 2%. The EGR flow through the two 6.5 mm parallel pipes is calculated and the EGR flow percentage is plotted for various speeds in Fig. 9.34. The percentage flow through the EGR pipe seems to be almost constant at all speeds with minimum and maximum being 6.8% and 7.5% respectively. The variation in the experimental and

Fig. 9.31 Comparison of experimental and simulated torque versus speed

9 Conversion of Diesel Engines for CNG Fuel Operation

389

Fig. 9.32 Comparison of experimental and simulated brake specific fuel consumption versus speed

Fig. 9.33 Comparison of experimental and simulated exhaust temperature versus speed

Fig. 9.34 Comparison of experimental and simulated EGR % versus speed

simulated values may be due to waste gate action as only one of the two limbs of the TC bypasses exhaust gas resulting in an unequal flow through the EGR pipes. This may be corrected by suitably differentiating the two pipe diameters to ensure equal flow through the EGR pipes.

390

G. Jeevan Dass and P. A. Lakshminarayanan

Fig. 9.35 Comparison of experimental and simulated NOx values at typical points

The exhaust temperature at turbine inlet was measured both in simulation as well as in experiments and was plotted in Fig. 9.33 at various speeds at full throttle. The maximum exhaust temperature obtained in the simulation was 760 °C that is less than the maximum limit of 800 °C. Percentage EGR flow ¼

EGR flow EGR flow þ Air flow

EMISSIONS—The Vibe 2-zone model can calculate the NOx values in the cylinder. The NOx values were calculated at the three emission speeds 2400, 1900 and 1400 rpm. A comparison between the simulated and experimental values is shown in Fig. 9.35. The NOx value can be adjusted using a parameter called NOx production constant. The NOx values were simulated before the catalytic converter. With a conversion efficiency of 95%, the catalytic converter can reduce the NOx quantity to the required limits. Comparison of SFC Between Simulated and Experimental Values—A part throttle performance was taken to compare the SFC values at all loads and speeds. It can be seen from Fig. 9.36 that the simulation is agreeing reasonably well with experimental values. The maximum deviation observed is 2.5%. As conclusion, it is possible to simulate the engine performance as well select the turbocharger, throttle body and EGR orifice diameter for meeting the required emission levels. This can reduce the development and engine evaluation time which can shorten the development time of a new rating or a new engine, Appendix.

9 Conversion of Diesel Engines for CNG Fuel Operation

391

700 ........ Experimental ____ Simulation

201 202

203

203

202

600

200 201

205

20 20

202

202

202

204

06 05

204

204

204

206

206

208

206

204 208

208

210

202

202

206

206

500

Torque, Nm

204

208

210

210 210

220 220

220 220

210 210

400

300

220 220

230 230

230 230

230 230

250 250

250 250

250 250

200 300 300

300

100 1200

320

320

320

1400

1600

1800

2000

2200

2400

Speed, rpm

Fig. 9.36 Part throttle SFC comparison between simulation and experiment Acknowledgements We thank Mr. Walter Knecht for his valuable consultations on various aspects of the engine.

Appendix A table below at a typical point of 146.5 kW at 2400 rpm demonstrates the success of the thermodynamic model for a turbocharged engine model using simple Vibe parameters derived from a similar naturally aspirated engine. At, Boost pressure = 1.61 bar, Lambda = 1, Bar Pressure = 750 mm of Hg

Power (kW) Torque (Nm) BSFC (g/kWh) Peak pressure (bar) Air flow (kg/h) Fuel flow (kg/h) EGR flow (kg/h) Exhaust temp before turbine (°C) Exhaust temp after turbine (°C) NOx (ppm)

Experimental

Simulation

146 582 207 100 525 31 38 768 683 1866

146 582 205 97 514 30 42 760 664 1851

392

G. Jeevan Dass and P. A. Lakshminarayanan

References BOOST, User manual for AVL BOOST, AVL List GmbH Kumar MLV, Mavi MS, Lakshminarayanan PA, Jeevan Dass G, Gopal G (2008) Thermodynamic simulation of turbocharged intercooled stoichiometric gas engine. SAE Technical Paper 2008-01-2510 Lakshminarayanan PA (1974) A mathematical model for a 4-stroke spark ignition engine. CSIR Ind J Technol 15 Lammle C (2005) Numerical and experimental study of flame propagation and knock in a compressed natural gas engine. PhD Thesis, Swiss Federal Institute of Technology, Zurich Vibe II (1970) Brennverlauf und Kreisprocess von Verbennungsmotoren. VEB Verlag Technik Berlin (German translation of the Russian original) Woschni G (1967) A universally applicable equation for the instantaneous heat transfer coefficient in the internal combustion engine. SAE Paper 670931 Woschni G (1991) Einfluß von Rußablagerungen auf den Wärmeübergang zwischen Arbeitsgas und Wand im Dieselmotor. In: proceedings to. Der Arbeitsprozeß des Verbrennungsmotors. Graz

Chapter 10

Simulation of Gas Flow Through Engine Neelkanth V. Marathe, Sukrut S. Thipse, Nagesh H. Walke and Sushil S. Ramdasi

Abstract Modern diesel engine has to meet both legislative and customer requirements simultaneously. Engine sub-systems are required to be precisely matched considering their interactive effect. It is a costly and time-consuming task if done experimentally. This task is becoming more complex due to flexibility in the advanced subsystems, which makes optimum matching of the subsystems practically becomes impossible. Use of simulation tools is very effective to handle this complexity and has advantage of significantly lower development time and cost. 3D CFD and 1D thermodynamic simulation tools are effectively used to analyse and tune engine subsystems performance. This gives opportunity to tune sub-system performance with their interactive effects, without a need of making physical prototypes. Considering the execution time, 1D thermodynamic simulation tool is more useful in the early stage development of the engine.

10.1

1D Thermodynamic Model

It is necessary to validate the formulated 1D thermodynamic model with test data. Model well validated for the required output response, can be used for predictive study (Adolph et al. 2009). Validate a quasidimensional model for predictive combustion and emissions demands considerable data and efforts. Hence for study of processes expect combustion and emissions, models with non-predictive combustion are useful, which are relatively easy to validate. To prepare 1D model engine geometrical data like bore, stroke, CR, manifold dimensions, valve timings, valve lift profiles, port flow coefficients, turbocharger maps, intercooler and EGR cooler effectiveness, etc. are required. The charge air cooler and EGR cooler are calibrated separately for required pressure and temperature drop (Heywood 1988). Figure 10.1 shows such 1D model prepared for a 3-cylinder turbocharged engine, whereas Fig. 10.2 shows 1D model prepared for a 4-cylinder turbocharged engine. N. V. Marathe
S. S. Thipse (&)
N. H. Walke
S. S. Ramdasi Automotive Research Association of India, Pune, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_10

393

394

Fig. 10.1 1D model prepared for a 3-cylinder turbocharged engine

Fig. 10.2 1D model prepared for a 4-cylinder turbocharged engine

N. V. Marathe et al.

10

Simulation of Gas Flow Through Engine

10.2

395

Gas Exchange Process

The basic purpose of Gas Exchange process optimization is to tune the burned gas removal and fresh gas intake process as per target requirements (Rakopoulos and Giakoumis 2009). Gas exchange process is characterized by parameters like volumetric efficiency, trapping ratio, residual mass fraction and pumping losses. These parameters depend on design of intake system from inlet air cleaner to intake port, valve train design including valve sizes and design of exhaust system. Table 10.1 gives list of the major system design variables which can be varied to optimize the system output parameters. Intake valve opening phase is extended beyond suction stroke to use inertia of the filling air. However, if inlet valve closes after start of effective compression stroke, a reverse flow of charge air into intake system is possible. This is a speed dependent phenomenon and IVC is required to be selected such that volumetric efficiency is better at higher operating speed, at the same time there is minimum backflow at the lower speeds. Figure 10.3 shows example of reduction in fresh charge reverse flow during suction stroke due to change in IVC timing. Graphs in Fig. 10.4 show effect of IVC on volumetric efficiency at manifold condition, PMEP and BSFC. Pumping loss reduces with late IVC, however due to reduction in volumetric efficiency BSFC is deteriorating at all the speeds but at different rates. Hence IVC selection is finding out optimum location which is according to the performance targets (Rocafort et al. 2005). The Valve Overlap (VOL) influences the airflow though the engine and the pumping work. An increase in VOL potentially reduces the pumping losses at higher engine speeds by reducing residual gas compression at gas exchange TDC. If the VOL is chosen too large, it may have a negative effect of reverse flow of burned mass from cylinder to inlet manifold as well as reverse flow from exhaust manifold to cylinder. The gas exchange during VOL also depends on the interfering blow down pulse of other cylinder. Figure 10.5 shows example of reverse flow during

Table 10.1 List of the major system design variables system output parameters System variables

Tuning parameters

Valve timings IVC, EVO, valve overlap Valve lift profiles Gas exchange path Port parameters Valve sizes Intake and exhaust system Intake/exhaust manifold dimensions Secondary runner

Maximum volumetric efficiency Speed zone of maximum volumetric efficiency Internal EGR Trapping ratio Residual fraction Fuel consumption PMEP Air-fuel ratios Miller cycle effect Packaging

396

Fig. 10.3 Reduction in reverse flow in suction stroke

Fig. 10.4 Effect of IVC on engine parameters

N. V. Marathe et al.

10

Simulation of Gas Flow Through Engine

397

Fig. 10.5 Reverse flow during VOL

VOL. Many times this reverse flow is deliberately kept which is termed as internal EGR as a cost effective method for NOx control (Schögl et al. 2009). Figure 10.6 shows effect of VOL, in terms of variation in IVO at constant EVC, on volumetric efficiency and BSFC at intermediate speed. Section of IVO and EVC is compromise between the two parameters volumetric efficiency and BSFC due to pumping losses. Also these trends will be different at different operating speeds which also needs to be considered while selecting VOL timings. Exhaust Valve Opening (EVO) has its main influence on fuel consumption (Schwarz and Spicher 2003). Retarded EVO increases the expansion stroke and therefore improves the high pressure cycle efficiency. However, if the timing of EVO is chosen too late, less time is available to blow down the cylinder charge, which leads to increase in the pumping work. Also with late EVO energy available to turbocharger also reduces. For each engine speed an optimum timing of EVO exists, where the balance between long expansion stroke and sufficient time for the blow down leading to best BSFC. There could be effect of EVO on volumetric efficiency due to interference of the blow down pulse to gas exchange of other cylinder. Figure 10.7 shows effect of EVO on volumetric efficiency, BSFC and PMEP at different operating speeds. The inlet flow sets-up expansion waves in the intake system. These waves are reflected from inlet manifold open end converting into pressure waves. The wave frequency and amplitude depend on the inlet system dimensions, especially inlet manifold dimensions. Intake manifold dimensions are required to be selected such that volumetric efficiency is maximized in the required speed zone (Teng 2011). Figure 10.8 shows effect of primary pipe length on volumetric efficiency. The volumetric efficiency also depends on the valve sizes and port flow coefficients. Figure 10.9 shows improvement in inlet mass flow rate due to improvement in inlet port flow coefficient characteristics.

398

N. V. Marathe et al.

Fig. 10.6 Effect of VOL on volumetric efficiency and BSFC

The amount of EGR with VOL adjustment is low in magnitude. Also, excessive VOL need measures to avoid valve piston interference (Xu et al. 2002). Hence another option of having secondary valve lift is exercised. The intake and exhaust valves are opened as required to provide secondary lift (Fig. 10.10). Figure 10.11 shows pressures and mass flow rate with secondary exhaust at 2200 and 1400 rpm. The necessary condition for exhaust gas flow is that exhaust manifold pressure is greater than cylinder pressure, when the secondary valve lift is active, otherwise there will be reverse flow of the charge from cylinder to exhaust manifold. The mass flow rate becomes very sensitive to secondary valve opening

10

Simulation of Gas Flow Through Engine

Fig. 10.7 Effect of EVO on volumetric efficiency, BSFC and PMEP

399

400

N. V. Marathe et al.

Fig. 10.8 Effect of primary pipe length on volumetric efficiency

Fig. 10.9 Effect of inlet port flow coefficients on inlet mass flow rate

position (Zücker et al. 2001). This configuration also has limitation to increase EGR fraction at lower speeds. Another approach is opening secondary inlet valve in blow-down or exhaust stroke as shown in Fig. 10.12. In this case cylinder pressure should be more than inlet manifold pressure during the active secondary intake valve lift. If the secondary valve lift timings are retarded, then instantaneous cylinder pressure may fall below inlet manifold pressure leading to flow reversal, as shown in Fig. 10.12. In both the configurations, control of internal EGR at all speed load points is not possible and the timings are required to be selected to have compromise between different operating conditions. Internal EGR gives benefit of simplicity and less cost, however the major disadvantages include no deactivation possible during cold operating conditions and also the hot EGR is recirculated

10

Simulation of Gas Flow Through Engine

Exhaust valve lift in Suction Stroke

401

Intake Valve Lift in Exhaust Stroke

Fig. 10.10 Valve lift for exhaust and inlet

(a). Pressures in cylinder, inlet and exhaust; valve lifts. Speed: 2200 rpm

(b). Pressures in cylinder, inlet and exhaust; valve lifts. Speed: 1400 rpm

(c). Mass flow rates throughexhaust and inlet; valve lifts; speed: 2200 rpm

(d). Mass flow rates through exhaust and inlet; valve lifts; speed: 1400 rpm

Fig. 10.11 Pressures (a, b) and mass flow rate (c, d) with secondary exhaust valve lift

402

N. V. Marathe et al.

(a). Pressures in cylinder, inlet and exhaust; valve lifts. Speed: 2200 rpm

(c). Mass flow rates through exhaust and inlet; valve lifts; speed: 2200 rpm

(b). Pressures in cylinder, inlet and exhaust; valve lifts. Speed: 1400 rpm

(d). Mass flow rates through exhaust and inlet; valve lifts; speed: 1400 rpm

Fig. 10.12 Pressures (a, b) and mass flow rate (c, d) with secondary intake valve lift

reducing benefits of cooled EGR. In case of secondary valve lift, valve dynamics also becomes critical and has to be taken care with precise simulation analysis.

10.3

Conclusion

Simulation of combustion and mass flows of air and gases is demonstrated by using one-dimensional models. The simulation is eminently suitable for arriving at valve timings at different speeds and loads, volumetric efficiency, internal EGR or sizing a turbocharger for the given application. With high speed computation is economically possible nowadays, such a tool is also used online either in the engine laboratory or vehicle trials during optimisation of fuel injection equipment or after-treatment systems, to obtain physical parameters that are very difficult to measure, but necessary for tuning the engine. Such a tool can be used in conjunction with computational fluid dynamics to get more detailed insights into combustion and flow phenomena.

10

Simulation of Gas Flow Through Engine

403

References Adolph D, Rezaei R, Pischinger S, Adomeit P et al (2009) Gas exchange optimization and the impact on emission reduction for HSDI diesel engines. SAE Technical Paper 2009-01-0653 Heywood JB (1988) Internal combustion engine fundamentals. McGraw Hill Inc., pp 205–248, 779 Rakopoulos CD, Giakoumis EG (2009) Diesel engine transient operation. Springer, London, pp 217–225 Rocafort J, Andreae M, Green W, Cheng W et al (2005) A modeling investigation into the optimal intake and exhaust valve event duration and timing for a homogenous charge compression ignition engine. SAE Technical Paper 2005-01-3746 Schögl O, Schmidt S, Abart M, Kirchberger R, Fitl M, Gschwantner P (2009) Early stage development of a 4-stroke gas exchange process by the use of a coupled 1D/3D simulation strategy. SAE Paper 2009-32-0101 Schwarz F, Spicher U (2003) Determination of residual gas fraction in IC engines. SAE Technical Paper 2003-01-3148 Teng H (2011) A thermodynamic model for a single cylinder engine with its intake/exhaust systems simulating a turbo-charged V8 diesel engine. SAE Int J Engines 4(1):1385–1392 Xu H, Fu H, Williams H, Shilling I (2002) Modelling study of combustion and gas exchange in a HCCI (CAI) engine. SAE Technical Paper 2002-01-0114 Zücker W, Maly R, Wagner S (2001) Evolution-strategy based, fully automatic, numerical optimization of gas-exchange systems for IC engines. SAE Technical Paper 2001-01-0577

Part II

Design of Engine Build

Chapter 11

Development of Ports of Four Stroke Diesel Engines Nagaraj S. Nayak and P. A. Lakshminarayanan

Abstract Fresh air is breathed in and products of combustion are exhaled by the inlet and exhaust ports. The efficiency of air flow is given by the flow coefficient. The flow is turbulent for most of the period and avoiding recirculation zones improves the flow. The energy for the flow is imparted by the piston during the intake stroke and only partly in the exhaust stroke as blowdown is a significant contributor during exhaust process. The energy during intake is partitioned between components contributing to flow and to swirl, which is important to support combustion in a direct injection engine. Of all the types of intake ports producing swirl in the cylinder, helical port is amenable to theoretical treatment with less empiricism; also, a helical port is stable in production and highly efficient. Helical port design considers free vortex with a correction for friction at cylinder liner surfaces. The optimum helical port is one which minimises the variation of exit velocity about the periphery of the valve seat. The exhaust port is designed to accelerate the flow at zones where there is a tendency to separate and then, a diffuser is designed to gain pressure. The definitions for the flow coefficient and swirl number vary from institution to institution. The definitions followed in this chapter were pioneered by AVL; these are widely used in many countries as de facto standard. The engine swirl can be reproduced at steady state rig for benchmarking or tuning the ports designed based on theory.

11.1

Introduction

A naturally aspirated engine takes in air from the atmosphere past the intake valve and through the port. By virtue of resistance to the flow the pressure drops and the pressure in the cylinder is lower than the atmosphere. The pressure below the N. S. Nayak (&) Caledonian College of Engineering, Glasgow Caledonian University, Seeb, Oman e-mail: [email protected] P. A. Lakshminarayanan Formerly with Simpson and Co. Ltd., Chennai, India © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_11

407

408

N. S. Nayak and P. A. Lakshminarayanan

piston, however, is nearly atmospheric and hence the piston works against the pressure difference, Fig. 11.1. Similarly, when the spent gases are sent past the exhaust valve, the flow resistance demands piston to perform work to push the exhaust gases and hence the pressure above the piston is higher than atmospheric. The pumping work done during the intake and exhaust strokes is irreversible as it is converted to low grade heat. The loss can be as much as 5% of useful work at full load and higher at part loads. In case of a turbocharged engine, clever choice of cam profile as well as the pressures in the inlet and exhaust manifold can reduce the pumping work. In a throttle-controlled gasoline engine, the pumping work is high at part loads at which a passenger vehicle would operate most of the time. Hence, in operating conditions, a gasoline engine pales compared to a diesel engine. Since the resistance is a function of the square of flow velocity which scales with engine speed, the pumping work is a strong function of engine speed. The proportionality constant of the relationship is dependent on how well the flow follows the profile of the port and the valve. However, the flow would depart from the actual profile of the port or the manifold locally in zones where the pressure gradient is severely adverse and a recirculation or dead zone is formed to reduce the effective area of the flow. Therefore, the shapes of inlet and exhaust ports must be designed so that such dead pockets are not formed and the resistance to the flow is minimised. The contortion of the ports in the head is challenging to a designer wanting to avoid recirculating zone. The inlet port of a direct injection diesel engine has to support combustion by providing swirl at the top dead centre so that air taken inside is used as fully as possible. In case of a direct injection engine, at the best 75% of air taken in can be used, because of heavy stratification of fuel even with swirl support. As the number of sprays increases the swirl can be lesser (Dani et al. 1990). Since, the number is finite, the swirl cannot be zero. The energy for the swirl is by the flow velocity created by the suction action of the piston. This flow velocity is decomposed perpendicular to the flow area and

Fig. 11.1 PV diagram of intake and exhaust strokes

11

Development of Ports of Four Stroke Diesel Engines

409

along the area. The former contributing to the flow and the latter to the swirl. That is the energy given by the piston to the flow is partitioned between swirl energy and flow energy. Therefore, the flow is compromised when the swirl demand is increased.

11.2

Characterization of Flow Properties of Valve Ports

11.2.1 Resistance to Flow Fluid Flow in a Closed Conduit Through an Orifice (British Standard 1983) The flow, Q_ through an orifice of geometrical area, aT in a closed conduit of a fluid density, q for a pressure difference, Dp is given by, sffiffiffiffiffiffiffiffiffi 2Dp Q_ ¼ laT q Analogy of Flows Through a Port and a Closed Conduit with an Orifice The valve seats on the port along a 45° or a 30° frustum of a cone. The geometry of the surface has inner diameter, d and an outer diameter. The flow through the ports can be analogous to the orifice flow. The choice of characteristic area of flow for a pipe (conduit) with simple geometrical section like a circle, a square or a rectangle is straight forward. However, for a complicated inlet and exhaust port the inner seat diameter is considered as representative dimension of the conduit, while characterising a given port. The shortest geometrical area between valve and the seat is equivalent to area, aT. The true flow area is obtained by the product of the discharge coefficient, µ and the geometrical area, a. In orifice flows, when the ratio of orifice area to the pipe area is tending to zero, the flow coefficient is near to unity. Analogously, when the valve lift is very small, the flow fills this area and the discharge coefficient with respect to the actual flow area, aT is nearly unity. As the valve lifts flow tends to separate from the seat and a contraction in flow is observed and hence there is a drop in the discharge coefficient. At large lifts, the geometrical area loses its significance and the port geometry controls the flow. Hence, at higher lifts the discharge coefficient drops precipitously. The analogy brings in complicated calculations of geometrical area and extremely discontinuous variation in discharge coefficient of flow with respect to the lift. An elegant solution is proposed by Thien (1965). He defined a gross flow coefficient with respect to the geometrical area of the conduit which is the product of contraction of the flow and the discharge coefficient of the true geometrical area as follows.

410

N. S. Nayak and P. A. Lakshminarayanan

The discharge coefficient µ which is related to the area a. Then, nominal contraction coefficient; r ¼

a a ¼ clear area pd4 2

where d inner valve seat diameter. effective air flow area ¼ lr

pd 2 4

The net flow coefficient, lr is used while simulating the engine or characterising the ports. If the effective air flow area is referred to the piston area, effective air flow area ¼ lrb

pD2 4

where b valve seat throat area/the piston area. The coefficient, lr increases marginally with lift and saturates when the port geometry takes over. For comparison of relative quality of various ports, stationary flow tests at a fixed pressure differential are completely satisfactory. Stationary flow tests corroborated with the engine tests.

11.2.2 Mean Flow Coefficient For the average piston velocity, Cm ¼

Sn ; 30

the average velocity of incompressible air flow through the port of an engine with the characteristic ratio, b is given by the following equation. Wm ¼

Sn 30b

where S is the engine stroke and n the engine speed, rpm. It can be shown that – – – –

for the same initial conditions, the same heat transfer from the port wall to the gas and for the same connecting rod to crank radius ratio, for crank angle, a. The cylinder pressure in the cylinder can be written as follows.

11

Development of Ports of Four Stroke Diesel Engines

411

pressure ¼ f ðWm ; lrÞ  f ðaÞ The effective air flow area is the integral of instantaneous effective air flow area during the intake event. ðlrÞm ¼

1 p

Z

BDC

lr da

TDC

The average gas velocity, Wm ¼

Wm : ðlrm Þ

The average drop in pressure, Dpm for intake or exhaust, can be obtained with the actual effective flow area assuming incompressible flow of mass density, q between top and bottom dead centres. q Dpm ¼ Wm2 and 2 V¼V Z h 1 q 2 w dV Dpm ¼ Vh 2 V¼0

where, w instantaneous air velocity through port / piston velocity. pD2 SZðaÞ 4 where Vh ¼ cylinder volume

air charge volume; V ¼

Substituting, 1 1 ffi ðlrÞm ¼ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi p R p dzðaÞ3 1 da 0 da ðlrÞ2 Here, z(a) is the displacement function of the crank mechanism with stroke, S = 1.

412

N. S. Nayak and P. A. Lakshminarayanan

11.2.3 Rotation of the Air in the Cylinder The swirl is not uniform over the length and radius of the cylinder either during the intake stroke or at the end of it near the bottom dead centre. The friction at the wall and in the flow helps in creating some order in the flow especially in the tangential direction. Experiments show (Dent and Derham 1974) that the flow in this direction is nearly solid body type rotational flow. The swirl, nr(a) of incremental volume, dV, enters the cylinder and helps in augmenting the swirl of the body of air in the cylinder. The instantaneous swirl is obtained from the stationary rig to be explained later in the chapter. The infinitesimal swirl is integrated from the top dead centre position of the piston to the bottom dead centre. Thien (1965) defines the ratio of the rotational speed of the cylinder charge to engine speed, called the average swirl number. Now, n  r

n

m

¼

1 nVh

Z

Vh

nr ðaÞ dV

0

11.2.4 Swirl Inducing Intake Ports While the radial component of the velocity of air through the gap between the valve and valve seat in the port, contributes to filling in of the cylinder, the tangential component contributes to swirl. In general, the tangential velocity is not uniform over the periphery of the valve seat. For a tangential port, it is highly non-uniform and for a helical port it is usually designed to be uniform. The uniformity of the tangential velocity leads to improved flow coefficient since the losses in the port are minimised. The tangential component that introduces swirl is characterised by the angular momentum of the air flow. In the absence of wall friction, the momentum will be conserved. For obtaining the angular momentum at the bottom dead centre of intake stroke or at the end of compression, the calculation is first performed assuming no losses and then correction is made to account for the losses. The correction, x is in the range of 0.35–0.4. The angular momentum L, with respect to the centreline, of the air entering the cylinder within a time interval, consists of the angular momentum of the flow leaving the port, LP and the angular momentum of the exit impulse, LA, Fig. 11.2. L ¼ LP þ LA

11

Development of Ports of Four Stroke Diesel Engines

413

Fig. 11.2 Velocity, impulse and angular momentum of air flowing through the valve opening into the cylinder

where, Z LP ¼ qhr 2

2p

Vr Vu du Z LA ¼ JA e sin uA ¼ qhre 0

2p

Vr Vu sinðu  xÞdu

0

JA is the component of exit impulse perpendicular to the axis; e is the eccentricity of the valve and uA is the direction of exit impulse. Figure 11.3 shows schematically the most popular methods of introducing swirl in the cylinder. The masked valve, masked valve seat or tangential port creates a high discharge impulse by a forced one-sided flow through the valve port. On the contrary, high swirl momentum is created inside a helical port. For too long, masked valve was used because the engine can be easily tuned for swirl, in spite of the high expense and problems with guiding the valve precisely as well as frequent valve pounding observed in the engines; the severe loss of flow performance leading to poor volumetric efficiency is a major disadvantage. The effect of the masked valve-seat is similar to that of a masked valve, but asks for deeply recessed valve seat in the cylinder head, though it is less problematic than a masked valve. This design is often being used combining with the tangential

414

N. S. Nayak and P. A. Lakshminarayanan

Fig. 11.3 Methods to produce rotation of air change

port, though at low lifts it is less effective and less stable. The one-sided discharge is arranged in the port by a sharp reduction of the port above the valve seat. In practice, with all these designs, heavy scatter in flow performance and swirl is noticed in production. Therefore, these strategies are limited to designs where the required swirl is relatively low. By comparison, a helical port producing swirl inside the port is the only design amenable to calculations and accurate design, as will be seen later in the chapter; the swirl depends exclusively on the port shape and the core shift affects the characteristics very little. Perpendicular to valve centreline, the component of the exit impulse is given by JA (Fig. 11.2). Thus, the angular momentum has two components, namely the swirl produced by the port as in the case of a helical port and in the case of a tangential port or the case with a masked valve, by the non-symmetrical flow. It is apparent from these expressions that the rotation of the cylinder charge can be induced both by the port-flow producing swirl and non-symmetrical discharge from the valve port. In the case of swirl producing port flow, the cylinder swirl depends only on the magnitude of air rotation in the port. For swirl production by means of discharge impulses the valve eccentricity e and the direction angle uA of this impulse are important.

11

Development of Ports of Four Stroke Diesel Engines

415

Fig. 11.4 Scatter band of swirl producing intake ports (Thien 1965)

If a helical port is carefully designed, a mean flow coefficient of 0.5 can be achieved for medium range swirl numbers. Tangential ports do not come close to these values. It is possible to estimate the extent to which the rotation of the cylinder charge depends on the moment of momentum of the impulse of a tangential port by varying the angle, a to the line connecting valve and cylinder centrelines and the eccentricity, e of a swirl producing port. It can be seen that the swirl is fairly independent of the two variables and that only the discharge angular momentum is paramount. Figure 11.4 presents a scatter of flow characteristics of masked valves, tangential and helical ports against lift values non-dimensionalised by the characteristic inner valve seat diameter. Usually, the non-dimensional valve lift by cam is 0.25 at its maximum. The valve seat angle and shape to an extent affect the flow at low lifts. The tangent to the flow coefficient graph at these lifts can be precisely calculated. However, at higher lifts the flow inside the ports take over as explained already. The higher curves correspond to helical ports and the lower range correspond to other ports. Therefore, almost all modern diesel engines use a helical intake port for generating swirl. In case of four valve arrangement with two inlet ports, one is helical and the other neutral (zero swirl) so that the latter does not interfere much with the swirl produced by the former.

11.3

Design of Helical Ports

For a perfect helical port, the cross-sectional areas are arranged in the form of a helix and the assumption of potential vortex flow is not out of the ordinary. The vortex is “free” because it is not created forcefully, for example by an impeller inside a centrifugal pump. Free vortex satisfies the law r  Vu ¼ constant , where Vu is the tangential velocity at any radius, r. Theoretically, the swirl number nr/n for

416

N. S. Nayak and P. A. Lakshminarayanan

such a port is independent of the valve lift, and it is observed to be true in experiments, to a large extent. With some empirical correction to the wall friction, it is possible to design the port for a swirl number required by the designer of engine combustion.

11.3.1 Development of Rotation Inside the Cylinder If the wall friction is neglected, then the angular momentum of the air flow inside the cylinder is conserved. In this case the wall forces are directed radially to the cylinder centre. The rate of angular momentum, L in relation to the cylinder axis into the cylinder consists of the LP exiting the port flow and due to the impulse, LA. L ¼ LP þ LA In case of cylinder flow, the latter is negligible; in case of tangential flow the former with more importance attached to the eccentricity, e of the valve and the direction of this momentum angle, uA. In other words, when a helical port is used the angular momentum in the cylinder is independent from the position of the intake port. Angular Momentum and the Cross-Sectional Contour of the Helix The flow inside the port is generated by the suction caused by the piston and hence the energy is transferred to the air flow from the piston. Since, by the absence of an impeller there is no stirring effect the flow and the vortex inside the port are “free”. The area can be changed gradually in any manner as a designer would like resulting in different profiles for the discharge velocity at the valve seat, for example. By calculus of variation it can be proved that there exists a profile which consumes minimum energy maximising the flow coefficient. Without going for the rigorous proof, it can be said here that the “minimum” profile is when the exit velocity at the valve seat is uniform. Then, the cross-sectional areas of the helical port are designed: (a) Flow velocity obeys the law of potential vortex i.e. r  Vu ¼ k ¼ a constant. The flow rate, Qu flowing through the cross-section, Fu of the helix and the dimensions of the cross-sections, see Fig. 11.5. Z Qu ¼ k

r1 r2

y dr r

The cross-section Fu is at u from the end of the helix, y is the height of the cross-section in the distance, r from the valve axis. (b) Since the total quantity Q should leave the helix uniformly, i.e.

11

Development of Ports of Four Stroke Diesel Engines

417

Fig. 11.5 Horizontal sectional view of a helical port and velocity profile

Qu ¼ Q

u 360

Hence, for two arbitrary cross-sections F1 and F2 the relation is obtained as follows. hR ro y

i dr r1 r u hR i F1 ¼ 1 ro y u2 r2 r dr F2

By choosing one cross-section namely, the entrance cross-section FI, the size of all other cross-sections is fixed. The angular momentum of the flow at the exit of the port of an ideal helical port is given by, LP ¼ QqVu r ¼ Qqk

418

N. S. Nayak and P. A. Lakshminarayanan

At the entry section I, the flow, Z

ro

QI ¼ k ri

y dr r

 cross section I

Therefore, LP ¼ q hR

Q2 i

ro y ri r dr cross section I

A typical helical port designed for minimum losses is shown in Fig. 11.6. The cross-section I is the beginning of the helix. The lines I to VIII, are sections of the helix and the outline is curved when looked from the top. The area of any section is bounded at the bottom, by the line 1. Below the surface of revolution, the free vortex forms a sink continuously at the valve seats in the valve and the cylinder head. At the end of the helix, the section tends to the line 1. A lip is formed between lead of the inlet port to section I and the upper face of the last section. It ends between the cross-sections VIII and I. The designer is free to create the sections if they are described above. The part of port, which leads the air from the manifold to the helix plays no role in angular momentum; the only care to be taken is to have a gradual change in area offering a streamlined flow, avoiding severe bends. A steadily accelerating flow in the lead would help in providing a favourable pressure gradient and avoid separation at bends. The helical port once designed is first tested at the rig described elsewhere in the chapter and later in an actual engine. The combustion engineer asks for either increase or decrease in swirl during combustion optimisation. It is easy to change the swirl in a helical port than in any other ports: by shortening of the lip of the helix which causes a decrease of rotation. Therefore, it is usual to design a port of

Fig. 11.6 Helical port

11

Development of Ports of Four Stroke Diesel Engines

419

swirl slightly higher than the intent so that it is easy to reduce it later. By shortening lip by Du, the beginning of the helix is shifted by Du. However, the flow rate is greater than for the original helix, because the circumferential components are dependent on the quantity of air flowing through the sections. Swirl in the Cylinder, the Angular Momentum at the Port Exit and its Size (Dani et al. 1990) The rate of angular momentum the air flowing at z, through a section of the cylinder is given by the relation, Z

2p

Lz ¼ q 0

Z

D 2

ca cu r 2 dr du

0

where ca is the axial and cu the tangential component of the velocity at some axial distance and D is the bore of the liner. Lz is a strong function of the axial distance and is less than the angular momentum of the flow leaving the port. What is more, it is extremely difficult by simple calculus to estimate the integral. Instead, in practice, the speed, nr of the paddle wheel like vane in a steady state rig is measured, the vane itself scaled to the bore of the engine. By assuming that the flow inside the engine liner is a rigid body rotation, Lz ¼ G

D2 nr p 8 30

where G is the rate of mass flow. Since, in case of a helical port the angular momentum is created only in the port LP of the port flow, then a simple relation between the swirl number, nr nr qVh ¼ n G 30 and the sectional areas of the helix may be found. For the ideal helical port with no flow losses Vr, Vu and Va are independent of the angle of the helix, the angular momentum, LP with respect to the first section and the air flow, G can be arrived at. G2 i LP ¼ hR ro y q ri r dr

cross section I

The volume flow rate is related to the mass flow rate by the relationship: G ¼ qQ. Then, considering the factor, x for loss of momentum to the cylinder walls and errors in measurements, Lz ¼ xLP

420

N. S. Nayak and P. A. Lakshminarayanan

2 r 3 Zo y 4 dr 5 r ri

¼

8Vh x pD2 nnr

cross section I

Substituting Vh ¼ p4 SD2 , Z

ro

ri

y dr r



x ¼ 2S nr  cross section I

n

It can be seen from the equation above, that for an ideal helical port, the swirl number nr/n is constant and it is independent of the valve lift. To characterise of the swirl of the cylinder charge the mean swirl number (nD/ n)m can be used. It is the ratio of the speed of the cylinder charge assumed as rigid body of the diameter D and the length equal to the stroke S to the engine speed. The momentum Lcyl of the cylinder is the integral of the angular momentum flowing to the cylinder during the intake stroke with a ranging from TDC to BDC. Z BDC Lcyl ¼ Lz da TDC

Using the equations of flow already derived in the beginning of the chapter, the mean rotational speed of the air in the cylinder can be calculated.

Z n  1 BDC nr c 2 r ¼ da n m p TDC n cm If k ¼ rl, where, r is crank radius and l is connecting rod length, then by substituting ratio of instantaneous to mean piston velocity n  r

n

m



1 k cos a ffi sin a 1 þ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi p 1  k2 sin2 a

¼

The mean swirl number relationship becomes, nr  nr  1 ¼    n n m 1 R BDC c 2 p TDC

cm

da

Simplifying, Z ri

ro

y dr r

 ¼ cross section I

 2 R 1 BDC c p TDC cm da n  2Sx r n m

11

Development of Ports of Four Stroke Diesel Engines

421

With known values of x, the stroke and the connecting rod length the helix can be easily sized by the relationship above. Usually, k lies in the range of 0.20–0.28, the numerator varies in a narrow range between 1.24 and 1.26. For the design, the useful to know that with normal length of lip and medium swirl numbers, x ¼ 0:5 for ports with extremely shortened lip, x ¼ 0:35:

11.4

Design of Exhaust Port (Scheiterlein 2013)

The exhaust port has a role to lead the gases with minimum losses to the manifold so that the pumping work is reduced. Usually, the exhaust valve is open before the BDC when the pressure is substantially higher than the manifold pressure and hence there is a blowdown during which substantial gases escape. For this reason, exhaust valves are made smaller, about 80% of the inlet (about 32% of the engine bore) valve giving more space to provide bigger inlet valves (about 40% of the engine bore) in the head because inlet port struggles to breath air in, with a substantial part of the energy given away to create swirl. At BDC the piston velocity is zero and does not help in pushing out the gases. With progress in crank angle the piston gains speed and drives the gases out effectively. However, it does not mean that the exhaust port is carelessly designed. Care must be taken to avoid dead recirculation zones that can block the effective flow area especially at sharp bend near the valve guide. Figure 11.7 shows a typical exhaust port. The characteristic dimension of the port is the inner seat diameter of the valve seat in the head. The area gradually increased up to section II by 15% of the cross section at the inner seat, just ahead of the valve guide area. The separation of flow would happen if the same velocity is continued at the bend beyond this point because of the adverse pressure gradient formed by the changing direction of the flow. To provide a favourable pressure gradient is a proven method; the area of the flow section is decreased to accelerate the flow, by reducing the area at section III by about 15% from section II. This would call for a serious reduction in area, if section II were not gradually increased as described above. Up to section V the area is maintained nearly constant. After that some pressure is gained by designing a diffuser up to section VI. This would enable a lower pressure at the entry the exhaust port and hence better evacuation. The area at the manifold side is about 50% more than section V or the area at the inner seat (Fig. 11.8).

422

N. S. Nayak and P. A. Lakshminarayanan

Fig. 11.7 Typical exhaust port and sections

Fig. 11.8 The profile of the exhaust port to avoid separation in the port shown in Fig. 11.7

11.5

Port Flow Rig

Figure 11.9 shows the test rig for studying the port in the laboratory for benchmarking, optimising or improving. A blower suitably sized for providing the flow through the ports of choice at a pressure differential of 250 mm Water at least when the valve is fully lifted. The blower is connected to the surge tank that would remove the turbulent fluctuations and stabilise the flow to or from the measuring unit that is interposed between the blower and the tank. The flow itself is measured using an orifice meter selected as per the standard. Two dampers are installed at the blower outlet to finely vary the flow from the tank. Figure 11.9 shows the head and a cylinder liner fitted above the tank for measuring the inlet port by sucking the air from the tank. The blower can be rearranged so that the air can be pumped into the tank when the exhaust port is under study. A paddle wheel is installed in the cylinder liner at a standard distance of 1.75 D and its speed is measured using a toothed wheel connected to the shaft of the vane. The housing containing the wheel is kept far away from the bottom of the liner lest it affects the flow in the cylinder. The steady flow results of flow coefficient, lr is plotted against the dimensionless valve lift obtained by dividing by the inner seat diameter.

11

Development of Ports of Four Stroke Diesel Engines

423

Fig. 11.9 Steady flow rig and paddle wheel for study of ports

Intake Port Flow Tests The mass of air of density, q flowing without losses through the unobstructed valve seat area Fv ¼ p4 d 2 can be calculated. Gmax

sffiffiffiffiffiffiffiffiffi Dp ¼ qFv 2 qm

qm is the average of densities at the entrance and the exit. The ratio of the effective to the geometrical flow area is obtained from the mass air flow measured, G using the orifice meter. lr ¼

G Gmax

To obtain the swirl number useful to the designer, the engine speed corresponding to the measured airflow, G is arrived at by considering a virtual engine of 100% volumetric efficiency.

424

N. S. Nayak and P. A. Lakshminarayanan

n¼

30G qVh

In conjunction with the measured data of nr and G, as well as the speed of a virtual engine, the swirl number can be obtained and plotted against the non-dimensional valve lift to characterise the inlet port. nr nr qVh ¼ n G 30 With the valve lift curve and the stationary bench test results the mean swirl number can be obtained. n  r

n

m

Z ¼p 0

p



dzðaÞ 2 nr da da n

Both the swirl numbers are proportional to the stroke to diameter ratio S/D. It is useful to mention here that ports of different engines are benchmarked considering a virtual engine of stroke equal to bore, S/D = 1. A Note on Paddle Wheel There are a few assumptions when a paddle wheel is used to measure the swirl. • The air rotates as a solid body, i.e., the tangential velocity is linearly proportional to the radius from the cylinder axis • The wall friction is negligible • There is no friction in the bearings that support the shaft of the paddle wheel • The paddle itself is very light. Though all these are true to some extent, it would be ideal to measure the angular momentum of the airflow directly. Tippelmann (1977) introduced a concept of flowing the swirling air through a straight honeycomb structure to rectify the flow. The angular momentum is absorbed by the structure held stationary on trunnion bearings. The holding torque is measured precisely by a torque meter. By the third law of Newton, this torque is exactly equal to the angular momentum of the swirl, but for the friction in the bearings which is negligibly small. Thus, this device is rid of most of the points against the paddle wheel. However, because the volume of legacy data on ports with industries and the institutions is so large, an equivalent paddle speed is calculated from the Tippelmann torque for use by the designers. Exhaust Port Flow Tests The arrangement for stationary flow tests of exhaust ports is almost the same as shown in Fig. 11.9, but for the missing paddle meter and the reversed flow direction. Usually, the exhaust ports work with a higher pressure drop than the intake ports. For this reason, for stationary tests a higher-pressure differential of 500-kPa/m2 is practised. The flow coefficient is calculated as for the inlet port keeping in mind the entrance conditions are those prevailing in the cylinder liner.

11

Development of Ports of Four Stroke Diesel Engines

425

References British Standard (1983) Measurement of fluid flow in closed conduits. BS-1042 Dani AD, Nagpurkar UP, Lakshminarayanan PA (1990) Universal mixing correlations for the performance and emission of open chamber diesel combustion supported by air swirl. No. 900446. SAE Technical Paper Dent JC, Derham JA (1974) Air motion in a four-stroke direct injection diesel engine. Proc Inst Mech Eng 188(1):269–280 Scheiterlein (2013) Der Aufbau der raschlaufenden Verbrennungskraftmaschine, vol 11. Springer Thien GE (1965) Development work on intake and exhaust ports of four stroke diesel engines. Österreichische Ingenieur-Zeitschrift 9 Tippelmann G (1977) A new method of investigation of swirl ports. SAE Trans 1745–1757

Chapter 12

Design and Analysis Aspects of Medium and Heavy-Duty Engine Crankcase Swapnil Thigale, M. N. Kumar, Yogesh Aghav, Nitin Gokhale and Uday Gokhale

Abstract The crankcase is the base part to which all important parts of an engine are assembled inside and on to it. The structure though is apparently stationary, it transmits or receives highly fluctuating loads at high frequency, from the piston, crankshaft and different pumps and gears inside it, and through bolt holes, its walls, ribs and surfaces mating with the other components. It enables the piston movement and wears out over the period of its maintenance life. The subject on crankcase is vast and an attempt is made in this chapter to describe the design and development of this basic part, with sufficient peer references. To construct a crank case of a heavy-duty diesel engine, materials of choice are grey cast (GJL), vermicular (CJL) and ductile irons (GJS). The physical and mechanical characteristics of grey, vermicular and ductile cast irons are tabled for use in various calculations. The strength of the material, basically affected by alloying elements and their effect on phase transformation and material properties is given with reference to the basic iron carbon equilibrium diagram. The resulting material properties specific to application like shock, vibration, fatigue and heat transfer are given. The cylinder liner may be integral with the crankcase or separate depending on the philosophy of design and application. The liners are classified as dry liner and wet liner; the latter can be either having a stop at the top or at the middle. The wear of the liner especially by the high contact pressure of the rings at the top and bottom dead centres is controlled by providing sufficient oil film thickness without much carryover past the piston to the combustion chamber. The surface is carefully honed where the type of honing is a choice after balancing the cost and required performance. The liner thickness is designed by not only considering the strength but also stiffness against cavitation. The design of a liner, in general, ponders over the failure modes like liner fillet cracking, bore distortions, bore polishing and, in case of wet liner, cavitation. The functionalities of the bays like crankcase top deck, between the cylinders, crankcase bottom and main bearing cap, crankcase front end and crankcase rear end are taken care of while designing the crankcase. While laying out the top deck, the following important parameters are studied: gasket S. Thigale (&)
M. N. Kumar
Y. Aghav
N. Gokhale
U. Gokhale Kirloskar Oil Engines Ltd., Pune, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_12

427

428

S. Thigale et al.

sealing, crankcase top stiffness, cylinder head bolts and cone of compression, deck cooling by CFD analysis, brinelling and indentation, fretting, oil-hole management, and bore distortion. Similarly, while designing bays between the cylinders, the parameters to be considered are coolant heat transfer, flow velocity, cavitation, crankcase ventilation, piston secondary motion and NVH. The important parameters while planning the bottom and main bearing caps are the assembly aspects, strength against firing and inertia loads, high cycle fatigue of crankcase and main bearing caps, main and side bolt fatigue, fretting at crankcase at the bearing cap interface and NVH aspects. The front end is constructed by considering the mounting components, and NVH by carrying out modal analysis; the bearing crush, radial loads, sealing methods, main and side bolt placement are some of the important aspects borne in mind. On the flywheel end of the crankcase, locating the flywheel housing, oil seal housing design as well as NVH are the aspects to take care. The design is validated experimentally at the hydro-pulsation rig. Finally, the success of the design is very dependent on the quality of production.

12.1

Introduction

The cylinder block or crankcase referred interchangeably is one of the major components of the reciprocating internal combustion engine which houses piston, cam shaft, main bearings, crankshafts, crank train, lubrication and cooling system. It forms an interface for other mounting components such as cylinder head, front cover, flywheel housing, water and oil pump, sump etc. Crankcase is the heaviest component in the engine and accounts for approximately 1/3rd of an engine mass. The cylinder block is subjected to static bolt assembly loads, dynamic gas pressure and heat loading and resulting vibration loading under engine operating conditions. With varieties of loading conditions, the complex path within crankcase over which these loads are distributed makes design of crankcase the most challenging task. Based on a type of engine and its application there are many verities of cylinder crankcase configurations. In the present chapter attention is placed on diesel engine cylinder block for power generation and heavy-duty mobile applications. Typical diesel engine cylinder block can be sectioned into different bays considering the loading types and functionalities these sections/bays have to fulfil. The upper deck or fire-deck experiences highest heat loading and gasket compression loading is critical to the functioning of cylinder head gasket. The region between the cylinders has important task to perform on thermal management within crankcase. This section houses coolant water jacket which takes out about 10–13% of total heat generated in the engine and responsible to maintain thermal distortions of the cylinder bore within limits which is directly linked to engine performance and emission aspects. The bottom section includes a main bearing caps and crankcase skirt. The main bearing cap receives dynamic gas pressure loads and inertia forces from moving masses in the engine and has important task to last

12

Design and Analysis Aspects of Medium and Heavy-Duty …

429

Fig. 12.1 Crankcase bays identified in three-cylinder inline engines

against these severe loadings for multiple engine overhauls and promised life to the customer of heavy-duty engines. The skirt receives vibration loading from crankcase walls and being highest flat area in the crankcase act as both emitter and transmitter of noise to the surroundings. On the front end of the crankcase a gear train is secured and has space for various mountings such water, oil and fuel pump. The gear cover protecting ingress of foreign particles to the gearing system is also one of the major sources of engine/gear noise. On the rear end of crankcase flywheel housing is mounted which is designed to resist bending moments and torque from application. The Finite Element Analysis (FEA) methods are used to analyse and design a various section of crankcase under complex loading conditions to ensure first time right design for durability validations. Figure 12.1 identifies the different bays in the crankcase briefed above. Considering the increased demand on lower fuel consumption, requirements are floated to reduce the overall weight of engine. Crankcase being heaviest component in the engine is always the target from weight reduction front. One of the easiest options to decrease the weight is to use material with lowest density and higher elastic properties. Traditionally grey cast iron was the preferred choice for crankcase material and continues to do so because of its easy machinability, higher compressive strength and good thermal conductivity. For large bore engines compacted graphite iron or vermicular iron is gaining the traction because of its higher tensile strength. The details of grey, vermicular and ductile cast iron are explained at the start of the chapter. The liner or bore of the crankcase is the central part of engine where chemical energy of fuel is converted to heat energy. It also houses piston on which gas

430

S. Thigale et al.

pressure acts and heat energy is converted to mechanical energy through connecting rod and crank-train assembly. The liner is surrounded by coolant jacket or crankcase wall. Depending on the relative position of cooling jacket with reference to liner and performance/cost requirements; liner is built into different configurations viz. dry, parent bore/integral and wet liners. The liner section explains the difference between different liner types, its performance requirements, and design and analysis aspects. The latter sections of the chapter explain the design and analysis aspects of different bays (identified in above paragraphs).

12.2

Block and Liner Materials

Over the years grey cast iron is the preferred choice for the crankcase material. As there is increased demand on the lower fuel consumption and reduced weight of the engine assembly, many automotive engines adapted Aluminium as the preferred material for crankcase. However, in case of heavy-duty engines grey cast iron still dominates because of its wide availability ease of casting and machining, comparatively good conductivity over ductile cast iron. In heavy duty engine, head is bolted to the block by cylinder head bolts. Under thermal operating environment head expansion is constrained by cylinder block resulting in tensile stresses near the fire-deck. Further at the bottom bay, the bearing caps are bolted to the crankcase and subjected to tensile loading from peak gas loading resulting in tensile stresses in all sections of the crankcase. Further firing forces from adjacent cylinders changes the stress sign and this continues over the lifetime of the engine. This requires higher fatigue strength of a crankcase material. The traditional grey cast iron has lower tensile proportional strength (approx. 169 MPa) and endurance strength of 65 MPa. This requires higher cross sections and additional ribbing of the crankcase made of grey cast iron to meet the durability requirements and limit the weight reduction and lower specific fuel consumption potential of the engine. The vermicular or compacted cast iron which has tensile strength between grey cast iron and ductile cast iron gaining the traction as alternate crankcase material which can help in decreasing the section sizes of the block structure. Ductile iron even with highest tensile and fatigue strengths are least preferred material because of its lower conductivity and higher cost.

12.2.1 Gray Cast Iron The carbon percentage in cast iron varies from 2 to 5%. Even with such a high concentration of carbon, the grey cast iron is soft and has lower tensile yield which enables ease in machining. At eutectic temperature the austenitic iron is converted to a combination of ferrite + pearlite phases and with good control over solidification process graphite flakes in the grey iron can be ensured. The solidification of

12

Design and Analysis Aspects of Medium and Heavy-Duty …

431

Fig. 12.2 Micro structure of grey cast iron

ferrite and pearlite iron should be completed before melt reaching carbide eutectic temperature which inhibits precipitation of graphite flakes and instead carbide may precipitate out of ferrite + pearlite resulting in increased strength and difficulty in machining of grey iron. Thus, it is paramount of importance to complete the solidification of grey iron within specified temperature limits. Ensuring right solidification time/temperature carbon precipitates out of ferrite and pearlite to graphite flakes. The right mix of silicon in alloy enables precipitation sites for graphite. Any increase in silicon results in enhancing of graphite flake distribution which helps in overall strength improvement of the grey cast iron. The graphite flakes are responsible for crack initiation sites in the grey cast iron and limits its tensile strength. However, the graphite flakes resist the shear forces between the iron crystals, and responsible for very high compressive strength. Figure 12.2 shows the crystal structure of grey cast iron. In previous paragraph effect of solidification time and temperature of casting is briefly explained. In complex castings with thin and thick section the graphite and carbide phases may simultaneously exists. Alloying elements such as copper, nickel, and cobalt helps in managing the solidification time/temperature and improve the cast ability. Controlled copper concentration is often a preferred choice, as high copper content degrades the fatigue strength of grey cast iron. Other alloying elements such as chromium and molybdenum are added to improve the

432

S. Thigale et al.

tensile and fatigue strength of grey cast iron. However, this alloying element limits the temperature window for graphite formation and makes material difficult to cast.

12.2.2 Compacted Graphite or Vermicular Cast Iron The compacted graphite iron is making advances in becoming the material of choice by designers for the usage of engine crankcases. The compacted graphite iron scores over grey cast iron because of its higher mechanical strength, stiffness and wear resistance. The compacted graphite iron is alternatively referred as vermicular cast iron owing to its worm like/vermicular graphite shapes. The graphite flakes are absent in the compacted graphite cast iron which decreases the stress concentration when subjected to loading. The microstructure of the vermicular cast iron is shown in Fig. 12.3. The graphite in the compacted iron can achieve 20% spheroidal structure. The vermicular cast iron can be obtained by close control over casting process with measured addition of magnesium material. With improved mechanical strength over grey cast iron much thinner parts can be made with compacted graphite cast iron. With compacted iron the section thickness of crankcase may fall from commonly observed 5 to 6–3 mm. However, casting process limits the section thickness on lower end. The real benefits of weight reduction can be realised by innovation in casting process to be able cast the Fig. 12.3 Micro structure of vermicular cast iron

12

Design and Analysis Aspects of Medium and Heavy-Duty …

433

sections of smaller thickness. This cast iron is difficult to machine and to achieve targeted machining rates simultaneous machining of part with multiple machining heads is required.

12.2.3 Ductile Cast Iron Ductile iron commonly referred as spheroidal or nodular cast iron owing to presence of graphite in spheroidal or nodular shape. The presence of spheroidal graphite significantly changes the material properties from grey cast iron. It contains carbon in the range of 3–3.6%. The nodular graphite shape is obtained by adding alloying elements as magnesium, cerium and tellurium etc. The microstructure of ductile cast iron is shown in Fig. 12.4. The ductile cast iron offers unique properties as increased strength, elasticity and stiffness over grey cast iron. Thanks to its enhanced properties over grey cast iron, several parts of high-performance heavy-duty diesel engine such as connecting rods, exhaust manifold, piston pin, main bearing caps are made from ductile cast iron.

Fig. 12.4 Micro structure of ductile cast iron

434

12.3

S. Thigale et al.

Cylinder Liners

The cylinder liner is a hollow cylinder shell in which combustion takes place. It is subjected to cylinder pressures which lead to hoop stresses in the liner. The thickness of liner should be selected such that adequate safety margins are available against these stresses. Further liner sees the considerable thermal gradients between inner and outer surface of liner which introduces the thermal stresses and liner has to be designed to withstand thermal stresses as well. The liner is also victim to the piston slap force because of secondary motion. This secondary motion excites the liner and has influence on the overall engine noise levels and cavitation in the coolant jacket leads to the damage of the liner. The detail on cavitation in the coolant jacket and liner failure modes are explained in subsequent paragraphs. One of the functions of the liner is to offer good wear surface and cylindricity for piston rings to glide on without piston ring leakages through piston ring—liner interface. The liner bore distortion from initial machining, clamping loads, engine operating thermal disturbs the liner—piston ring interface and may result in higher blowby, higher oil consumption and friction. Given the stringent requirements of strength, anti-corrosion and wear property; cast irons (because of its lubricating properties) is the widely used material for liners. The cast iron is porous in nature and helps in minimizing the risk of piston seizure and good solution for piston galling. However, cast iron is not one of the strongest materials which requires it is not be used in basic form and some level of alloying is ensured. The alloying elements generally used are chromium, nickel and copper with concentrations not more than 5%. The details of materials are presented in material section. Liner is classified into dry and wet liner concepts depending on whether coolant is in direct contact with liner outer surface. Following section provide the design construction, peculiarities and advantages and disadvantages of different liner types.

12.3.1 Wet Liner Design In wet liner concept the coolant is in direct contact with the liner outer surface. The wet liners are produced from sand casting and inserted into the block castings between the block bulkheads. Liner is flanged at the top end to provide location and secure in the block casting. The top surface of liner is in contact with combustion bead of cylinder head gasket. The bottom face of the liner flange contacts the block and together form enclosure for the coolant called coolant jacket. The bottom end of liner houses grooves for rubber O rings which provides sealing for coolant and avoid mixing of oil from crankcase with coolant. The direct contact of coolant with liner ensures lower metal temperatures in the liner. The fire-deck of block is also cooled from the coolant. With lower temperatures, the liner bore distortions under operating thermals are smaller compared to dry liner engines. The wet liners are further classified into top stop and mid stop design options. In the top stop design

12

Design and Analysis Aspects of Medium and Heavy-Duty …

435

the coolant is present over entire coolant jacket path and even cooling is achieved. In mid stop liners, the cooling is achieved for only top portion of the liner and fire-deck of the block. However, in both the design combustion gasket sealing is achieved at the fire-deck of the block. In wet liner construction machining of only inner surface of the liner is critical; better outer surface finish ensures better heat conduction and uniformity of cooling; however no specific machining requirements are placed on the outer surface of wet liners. The wet liners can be easily replaced in case of engine overhauls. The wet liner design can’t be changed to dry liner in later stage of engine development or engine upgrade programs. In wet liner construction, to accommodate width of liner flanges at top deck the bore spacing is generally on higher side. Further manufacturing tolerances cause different sleeve protrusion heights at the cylinder deck; the cylinder head gasket should be designed to compensate this in an engine operation. The alternate option is to machine the top of liner after assembly to maintain same sleeve protrusion heights. The cost of wet liner engines is higher than parent bore construction. The wet liner construction is the preferred choice for most heavy-duty diesel engines. Figure 12.5 shows the wet liner construction in the engine.

Fig. 12.5 Wet liner design construction (a top stop, b mid stop)

436

S. Thigale et al.

12.3.2 Dry Cylinder Liner A dry cylinder liner is in the form of barrel shape with flange at top to secure it in position in the block. In dry liners heavy flanges at the top decks are absent which can benefit in reduction of the bore spacing. The dry liners are not in direct contact with coolant which increases the overall temperature in the dry liner concept. The material of dry liner can be different from that of block casting; the liner material can be hardened for protection against wear. The entire outer surface of dry liner is in close contact with block casting which requires accurate machining of both inside and outer surface of dry liners. No separate sealing at the bottom of the liner is required in dry liner concept design. The open and closed deck crankcase can utilise this liner type. They are press fitted into the crankcase. The interference selected for press fit should be matched to operating temperature such that no loosening or additional stressing occurs under operating thermals. The loosening results in clearance between block and liner which results in decrease of heat transfer between coolant and liner. It is recommended to ensure right contact between the block casting and liner to avoid shooting up of the liner temperatures. Too tight contact between liner and block casting is also dangerous as this may results in block distortions, liner cracking and scuffing on inner surface of liner. With dry liner construction temperature control is very difficult, may result in higher temperature at the piston rings. Further excessive liner bore distortion can occur from the raised temperature. The dry liner can be made of iron with alloying elements, steel etc. The material choice is based on the lubrication, wear requirements. Figure 12.6 shows the dry liner construction.

12.3.3 Integral Liner/Parent Bore The most commonly used liner construction is of parent bore or integral type wherein cylinder liner walls are cast integrally with the crankcase. The common usage of this type of liner construction is found in a closed deck type of the crankcase. The cast integral liner is surrounded by cooling jacket. On the inner side; coolant is in contact with liner outer surface and on the outer side coolant is in contact with crankcase outer wall. This construction helps in minimizing the cylinder bore spacing, and decreases the width and total axial length of the crankcase. With this liner construction heat transfer from coolant to the cylinder top deck surface can be enhanced. Since parent bore is cast simultaneously with block, the outer surface of parent bore can’t be machined and casting surface with uneven thickness and surface finish may result in uneven cooling of the parent bore. This may have impact on bore distortion under thermal loads. An integral cylinder liner is cost effective to manufacture, but does not offer flexibility for engine overhaul or rebuild. Figure 12.7 shows the parent bore liner construction.

12

Design and Analysis Aspects of Medium and Heavy-Duty …

437

Fig. 12.6 Dry liner design construction a Press fit b Inserted type

12.3.4 Liner Bore Distortion and Honing Process The liner bore distortion can be defined as deviation of liner bore from its ideal round shape under the influence of manufacturing clearance, bolted assembly loads and hot thermal and peak cylinder pressure operating conditions. The important function of the liner is to allow piston ring to slide over without causing wear or scuffing on the liner surface. It is required for piston ring to conform to liner bore shape to avoid leakage of combustion gases past the liner bore—ring interface. At the same time oil in the crankcase should not pass this interface into the combustion zone. The higher bore distortion may result in incapability of ring to conform cylinder bore and light seal between ring—bore and ring—piston groove gets disturbed. This leads to higher blow by, increased oil consumption and emission of the engine. Further heat transfer between ring and liner bore is also affected resulting in higher temperature of the ring. The higher temperature on the piston ring may damage ring coatings. So, it is important to limit the liner bore distortions

438

S. Thigale et al.

Fig. 12.7 Parent bore liner design construction

of the bore to avoid ill effects on engine performance and emissions. The liner bore distortion in the assembly loading to some extent can be controlled by avoiding excess preloads, stiffening of cylinder fire-deck etc. However, these remedies are not enough to control overall distortions. Generally, torque plate honing with cylinder block assembly and torque plate having equivalent head stiffness and production gasket is performed such that bore profile is inverted. When production engine is built under bolt loading this inverted profile becomes round. This is widely accepted procedure to get rid of assembly level liner bore distortions. However, the bore distortions during operating thermal loadings are several magnitudes higher than bore distortions under the assembly loads. Further bore shapes also changes substantially during engine operating conditions. However, the bore distortions during engine operating thermals can’t be measured so it is not possible to map the liner distortion profiles during torque plate honing. The CAE analysis simulating assembly and operating thermals and pressures can get the information of bore distortions. The similar distortions can be achieved with hot honing by passing coolant of specific temperatures around liner/in the cooling jacket. Further with advanced machining processes called light or professional form honing

12

Design and Analysis Aspects of Medium and Heavy-Duty …

439

developed by Gehring the desired shape of bore can be achieved (Flores 2015). Based on the CAE calculations and the distortion measurement from hot coolant supply strategy can be set up to decide to follow form honing light or form honing professional approach for bore machining. The form honing light can form bore in bottle shape where top bore is kept small to avoid amplification of piston slap forces and bottom bore is gradually increased to limit the frictional power loss. If bore distortions of cylinders under operating thermals can’t fit to ‘form honing light’ approach more advanced option of ‘form honing professional’ can be utilized to compensate thermal bore distortions to achieve running conditions round and straight bores. The SAE paper compiles the information on line bore distortions during various assembly and operating conditions. Further liner bore distortion results with torque plate honing and hot honing are also studied. The effect of material type on line bore distortion with influence of mounting components is also reviewed. Figure 12.8 compares the bore distortions under engine assembly and different operating conditions. Figure 12.9 shows bore distortion shapes with torque plates and mounting components. Figure 12.10 shows the progression of liner bore distortion from room temperature assembly through hot honing to room temperature bore distortion shape. Figure 12.11 shows the bore profiles/shapes manufactured from form honing light process. Figure 12.12 shows the capability of form honing professional in achieving the complex bore profiles which under running conditions can form round and straight profiles.

Fig. 12.8 Bore distortion under assembly and different engine running conditions

440

S. Thigale et al.

Fig. 12.9 Bore distortion shapes with torque plate and different mountings

Fig. 12.10 Bore distortion shapes progression from before and after honing

12.3.5 Liner Cavitation The liner cavitation/corrosion problem goes unnoticed during initial durability tests. It appears after several hours of engine operation (e.g. more than thousand hours). The discovery of liner cavitation problem can be seen when oil mixing with coolant is observed. This happens when cavitation corrosion breaks the integrity of liner material. The SAE paper on liner cavitation intensity measurement (Green and

12

Design and Analysis Aspects of Medium and Heavy-Duty …

441

Fig. 12.11 Different bore shapes with Form Light Honing (Flores 2015)

Fig. 12.12 Free shape profile from Form Professional Honing (Flores 2015)

Engelstad 1993) provides the macro and micro aspects of liner cavitation. The macro aspects include liner transverse vibrations because of piston secondary motion resulting in dynamic cycles of expansion and compression waves in the coolant resulting in formation of bubbles cavities. They start to grow with

442

S. Thigale et al.

accumulation of energy, until coolant pressure starts to increase due to compression phase. At this time bubble becomes unstable and implodes abruptly. When a bubble collapse near the liner, it collapses in unsymmetrical manner forming micro-jet of fluid directed towards the liner with speed as high as 600 m/s and results in local pressure of 1000 kPa. The micro-jets directly attack the liner surface which is already weakened by corrosion. The damage of liner occurs when pressure acting on the surface because of micro-jet dynamic effects is higher than the surface strength of liner material. Figure 12.13 shows the impact location of piston on to liner bore because of piston secondary motion. In Green and Engelstad (1993) test set up for engine cavitation intensity measurement is explained. The effect of coolant type, temperature, pressure and flow on cavitation is studied. The findings of the measurement suggest that coolant back pressure and temperature affect the cavitation. The critical cycle for cavitation varies from engine to engine. Another SAE paper (Hosny et al. 1996) attempts to investigate the liner vibration and pressure waves in the coolant jacket by modelling fluid structure interactions and thereby predicting the cavitation. This approach helps in prediction of liner cavitation. The paper also discusses the results of different parametric studies (impact magnitudes, coolant jacket clearances etc.) to obtain the coolant pressure and liner surface velocity profiles. Fig. 12.13 Location of piston impact with cylinder bore (Hosny et al. 1996)

12

Design and Analysis Aspects of Medium and Heavy-Duty …

12.4

443

Engine Cooling and Coolant Jacket

The middle bay of cylinder needs to perform important function of thermal management in the engine. This bay houses coolant jacket which takes away about 10– 13% of total heat produced in the engine. The coolant liquid used for engine cooling is 50/50 mixture of water and ethylene/propylene glycol. This liquid is effective in taking out heat without phase change over wide temperature range from −57 to 120 °C depending on pressure developed by the water pump. Generally, coolant pump is mounted at the front of block and discharge from coolant pump first goes to the front cylinder and progress towards the back cylinder gaining the heat along the way. The relatively hot coolant enters to the back-cylinder head and comes out from the front cylinder head. The coolant path in the block is shown in Fig. 12.14. The coolant path causes temperature imbalance explained above; to counter this small portion of coolant from block about mid cylinders is transferred to head. The additional holes are drilled on the block for this purpose. To minimise the temperature difference between cylinders, a cross-flow cooling circuit shown in Fig. 12.15 can be utilized. For this purpose, additional headers are cast in the block castings. These headers facilitate the coolant flow transfer to coolant jacket of individual cylinders. From there coolant is transferred to cylinder heads and output of cylinder is transferred to low pressure header. For multi bank cylinders the coolant circuits of one bank are replicated for other bank of cylinders. While designing the coolant paths it should be ensured that stagnant zones should be avoided at the same time no high temperature region of the block experience coolant starvation. The region/surfaces in the block/liner close to fire-deck are above the boiling point of the coolant. The coolant on the surface or

Fig. 12.14 Typical cooling path in engine

444

S. Thigale et al.

Fig. 12.15 Cross flow cooling path to compensate temperature imbalance

very close surface has zero velocity and coolant near to higher temperature surface boils. The coolant flow velocities should be maintained in the range of 1 to 1. 5 m/s. If the velocities of bulk fluid are not maintained the boiling coolant changes phase and higher temperature surfaces will be covered with superheated vapour’s which hinder the effective heat transfer to the coolant. The process of formation of vapour around the surface is called as film boiling and, in any case, should be avoided as this affects the cooling efficiency and surface temperatures may be raised to unacceptable levels causing excessive thermal stresses and fatigue failure. The computational fluid dynamics (CFD) analysis procedures should be followed for optimizing the coolant jacket and coolant flow around the cylinders. With this hot stagnate spots can be identified and corrected in the engine development stage. In closed deck engines, coolant is available all around the liner and block walls. This ensures metal surface temperatures are maintained within allowable limits. However, in case of open deck engines with siamesed bores the bridge region is not supplied with coolant resulting in higher temperature at these locations. To ensure some coolant to reach at the bridge the slit is machined between cylinders which help in partial cooling of the bridge. Figure 12.16 shows the siamesed block with slit for partial cooling at the bridge location. The peak thermal loading on the block can be observed at cylinder top deck, and at cylinder bridge location. The depth of the coolant jacket can range from 50 to 90% of the stroke for a cast iron crankcase. From packaging perspective, the smaller depth of coolant jacket is preferred, it also helps in maintaining coolant velocity higher enough to avoid film boiling. The smaller height of coolant minimises the engine warm up time during start up. The deck temperature along with position of top ring reversal also dictates the height of coolant water jacket.

12

Design and Analysis Aspects of Medium and Heavy-Duty …

445

Fig. 12.16 Siamesed block with slit for bridge region cooling

12.5

Block Fire Deck

The cylinder block fire deck supports the cylinder head gasket; the combustion beads which surrounds cylinder is supported from liner/block. It has several openings like coolant holes, oil holes and oil return passages etc. which are surrounded by gasket back land/low pressure beads. The block also has openings for bolt clearance holes. The stiffness of block fire-deck is important for the successful functioning of cylinder head gasket under assembly and engine operating loading conditions. The stiffness of block fire deck also plays important role in limiting the liner bore distortions. The cylinder block fire-deck and block walls are cooled by cooling jackets which limits the cylinder block fire deck temperature much smaller than the cylinder head fire-deck. Under thermal operating conditions the head with higher temperature expands towards the block/liner, the block deck with lower temperature arrest the thermal expansion of the head. This results in compressive load on the cylinder head fire deck, and introduces the tensile forces next to block fire deck. The additional contact pressure at block—gasket during operating thermal may sometimes cross the yield point of the block material and leads to brinelling failure of the block/head fire-deck. In engines with a single cylinder head assembled with block the compressive stresses linearly increase from the end bore to the central cylinders. Individual

446

S. Thigale et al.

heads significantly reduce these stresses. Modular head construction (individual heads) helps in reduction of these stresses. In engines with longer lengths having single cylinder head construction, head is cut partially at the block top deck between bores to enable stress relaxation. The peak cylinder pressure and piston inertial loads apply radial loads on the crankcase walls, resulting in deflection of the block fire deck. This displacement of cylinder walls may cause cylinder head gasket fretting, which can result in leakages or gasket culet cracking issues. Additional clamping load or supporting for the block walls may be required avoid this failure mode. This phenomenon is prevalent in lower duty engines. The crankcase walls can be made stronger to compensate the thrust loads and minimize the deflection of cylinders. Another important aspect in designing the block fire deck to ensure full proof combustion sealing lies in a selection and placement of dowels at the top deck. Generally, cylinder head owing to its higher temperature expands more than cylinder block. The expansion of cylinder head cause to and fro movement of the cylinder head on the gasket which may result in gasket fretting and leakage issues. To avoid uncontrolled head movements dowels are provided on the two ends of cylinder block. The dowel at one end of the block should be of solid construction and at the other end spring type dowel is employed. This is done to avoid interaction of cylinder head with two dowels when clearances between dowel and head are overcome under thermal expansion. If cylinder head contacts with both dowels, head bows between dowels; may result in uncompressing the cylinder head gaskets and possible leakages. The spring type dowel compensates part of expansion and safeguards the head from bowing and consequent leakages from the cylinder head gasket.

12.6

Design and Analysis Aspects of Crankcase Bottom Bay and Main Bearing Caps

The crankcase bottom bay consists of bulkhead, crankcase skirt and main bearing caps. The bottom of the crankcase also houses the main oil galleries, oil rifles to distribute oil to the main, cam shaft and con rod bearings. The bulkhead forms crankcase walls and partially supports the crankshaft through main bearings and depending on the cam shaft location in the crankcase, supports cam shaft as well. The bulkhead provides rigidity to the crankcase and distributes the weight, loadings of the mounting components. There is a series of bulkheads in the engine depending on engine configuration e.g. 4-cylinder inline engines have five bulkheads; likewise, a 10-cylinder V-configuration has six bulkheads. In inline engine configuration, loads from peak cylinder pressures are approximately equally distributed on to each bulkhead/main bearing caps. In V cylinder configuration each bulkhead/bearing cap experience pulsating combustion pressure loads from left and right bank firing. The flywheel end bulkhead is

12

Design and Analysis Aspects of Medium and Heavy-Duty …

447

subjected to additional inertia loads because of flywheel mass apart from peak pressure loads from combustion within cylinders. The bulkhead has to meet the engine durability requirement under application of above explained loads. Further it has important function to perform in overall engine structure born noise reduction. Figure 12.17 shows the bulkhead section. The detailed discussion of design and analysis aspect of bulkhead is presented in the following section.

Fig. 12.17 Typical bulkhead section

448

S. Thigale et al.

The main bearing cap houses bottom half of main bearing and like bulkhead supports the combustion loads. The main bearing caps are connected to the bulkheads through main bolts (aligned in cylinder axis). Apart from bolt assembly loads it is also subjected to radial hoop/tangential loads from bearing assembly/ interference. In its basic form, main bearing caps represents circular beam subjected to bending loads from peak cylinder pressure. The cylinder pressure load changes over the thermodynamic cycle; thus, nature of loads on main bearing caps are of pulsating type. This requires main bearing caps to be designed against fatigue failure. Depending on the engine configuration (inline/V engine), the main bearing cap assembly with bulkhead may have different arrangement like assembly with main bolt alone, additional side bolt arrangement or inclined bolt arrangement etc. Further acoustics design of crankcase also results in different configurations of main bearing caps. The below section explains the different design/assembly configurations, analysis requirements for main bearing caps corresponding to engine types/ functions expected. The skirt of the engine owing to its flat surface; is one of the prominent (structure born) noise emission source. The skirt is not a load bearing section and thus not influenced by the combustion/inertia loads generated in the engine. Depending on engine configuration/type there can be different design versions of the skirt. The overview of design details of skirt and its effect on overall noise radiation from the engine is explained in separate section of crankcase skirt.

12.6.1 Design Aspects of Main Bearing (MB) Cap The crankshaft is supported half in the bearing cap and other half in the bearing walls/bulkhead. The support in the bulkhead is stiffer compared to main bearing cap, thus main bearing cap has to sustain equal load as bulkhead with comparatively smaller cross section. This calls for higher strength and stiffer material for a main bearing caps compared to the bulkhead. In general, the bearing caps are made of spheroid graphite iron or GJS grade cast iron. The main bearing cap in simplest form can be visualized as curved beam which has to resist the shear and bending forces resulted from assembly (bolt and bearing crush) and combustion pressure. In general, MB caps are found in either complete I section or half-moon shape. However, depending on layout/packaging constraints or analysis recommendations other shapes are also implemented. The main bearing caps are secured to the bulkheads by main bolts. The bolts are taken deep in the crankcase to avoid the negative influence of it on the bearing housing distortions. However, in some engines presence of crossing oil galleries limits the height of main bolts. Figure 12.18 shows the crossing oil galleries and limitations put on main bolt heights. In the across direction (perpendicular to crankshaft axis) if main bolts are away from main bearing bore, the distortions under bolt assembly will be smaller. However, they are placed as close possible to main bearing bore to ensure enough contact pressure at the block—MB cap interface near to bore and avoid joint

12

Design and Analysis Aspects of Medium and Heavy-Duty …

449

Fig. 12.18 Crossing oil galleries

opening under application of firing loads. Generally, width of MB—block interface near to main bearing bore is 1/3rd of interface width past bolt hole. The above explained interface locations are identified in Fig. 12.19.

450

S. Thigale et al.

Fig. 12.19 Main bearing spigot and MB-block interface

In inline engine configuration, the combustion forces/peak cylinder pressures act vertically (spread over MB area such that horizontal component is negligible) and are reacted by main bolts and MB cap and bulkhead. In some exceptional cases such as misfiring, there exist some horizontal components of forces which are rare events and can be dealt with main bolts alone. However, in case of V engine configuration there exists substantial magnitude of horizontal component of peak cylinder pressure which calls for cross bolts to avoid fretting failure at the block— MB cap interface. For V engine configuration with wider V angle (90°) side forces can be of larger magnitude and design provision shall be made for extra row of cross bolts. The cross-bolting configuration also improves the rigidity of main bearing cap and helps in improving the acoustic aspects of the skirt at the same time additional load path for noise transmission from the skirt. The assembly sequence of main bearing cap with individual bulkhead is followed as follows: The main bearing cap is pushed towards bulkhead against the interference about spigot (shown in Fig. 12.19) and main bolts are tightened to ensure assembly of main bearing cap with bulkhead. This interference about spigot is critical from the stresses at fillet location adjacent to MB—bulkhead interface. Any additional

12

Design and Analysis Aspects of Medium and Heavy-Duty …

451

interference will generate higher mean stresses in the fillet and fatigue durability at this location can be compromised. Post assembly of main bearing cap with bulkhead, bearing housing is distorted. To ensure perfect round shape of bore for bearing assembly; the bulkhead and main bearing cap bore are machined together. This assembly of bearing shell in the bearing housing generates the radial stresses in the bearing housing which adds means stresses to the main bearing cap and affect fatigue lifetime. The bearing housing diameter should be at least 1. 5 times the inner diameter to generate enough contact pressure between bearing shell and housing. If bearing housing is not properly sized there is possibility of relative motion between the bearing shell and bearing housing. This can result into fretting damage at the back of bearing shell. Generally housing inner diameter is laser treated to improve the wear resistance and improve the friction between bearing shell and housing. Later cross bolts are assembled between skirt and main bearing caps. The skirt being less stiff structure may introduce bending in the cross bolt. The skirt/main bearing cap interface shall be designed such that bending of cross bolt is limited. After assembly of main bearing cap with bulkheads the entire joint is subjected to firing and inertia loads from engine operation. Considering multiple interfaces between main bearing and bulkhead/skirt and its effect on stresses in the joint, careful tolerance stack up of main bearing cap and block assembly should be performed. The surface irregularities and flatness values at the interfaces should be maintained on the lower side. The failure of the main bearing cap in the field can be costly affair considering that main bearing caps alone can’t be replaced as crankcase has to be taken out for machining with new bearing caps and old bearing shells replaced with thicker bearing shells. With amount of effort and cost involved in case of failure of the main bearing caps; extra care should be exercised while designing/analysis bearing caps.

12.6.2 Design Aspects of Bulkhead and Skirt As mentioned earlier there are series of bulkheads between cylinders supporting engine main bearings/crankshaft. The end bulkhead forms mating interfaces for front and rear side mountings on the block. On the front side bulkhead supports front/gear cover, oil/water pump whereas rear bulkhead supports flywheel housing, starting alternator etc. One of the important considerations in engine design is of placement of thrust washer. In earlier days thrust washer was placed at the rear end of the bearing/bulkhead. However, considering premature wear and damage of these axial bearings and with increased peak cylinder pressures and bigger dampers the correct position of thrust bearing is at the centre of the engine. Further possible bending of crank throw under generator/rotor mass can also cause additional wear of the thrust bearings. The bulkheads are connected to each other by outside left/ right block walls generally called skirt of the engine. Similar to main bearing caps the bulkhead is subjected assembly stresses from bolting of main bearing cap and bearing shell. It is also subjected to peak pressure loading from firing of cylinders

452

S. Thigale et al.

and inertia loading from rotating/reciprocating masses. Further each bulkhead experiences pulsating loads from firing of adjacent cylinders. This requires bulkhead to be designed for fatigue loading. The bulkhead strengthening/ribbing is designed from FEA analysis and fatigue safety factor evaluations considering various factors such as shape, size, temperature, loading, reliability, stress gradient etc. for above mentioned loads. The ribbing/strengthening of bulkhead also plays important role in NVH optimization of the block. The material addition in the bulkhead near main bearing cap helps in increasing the fundamental frequency of axial vibration mode of main bearing. Sometime complete filling of material in the bulkhead enhances the NVH behaviour of the block. While adding the material in the bulkhead appropriate relief shall be provided for crankshaft counterweights and liners especially in the V engine configuration. The most stressed location in the bulkhead is fillet near to Mb cap-bulkhead interface. The assembly stresses at these locations are developed from interference about spigot and main bearing crush. Operating pulsating loads makes this location critical from the perspective of fatigue failure. Under assembly and operating loads, the bulkhead or most part of block is under tensile loading. The tensile strength of grey cast iron from which block is made is smaller compared to steel or ductile cast iron. However grey cast iron has exceptional compressive strength. To take advantage of this, some smaller engines are provided design concept with through bolts which clamps head with block and main bearing by single bolt along the height of crankcase. Figure 12.20 shows the through bolt concept for Ford DIATA 1. 2 l diesel engine. There are various types of skirt design prevalent in engines. They are primarily classified as short and long skirt type. In short skirts design the main bearing cap and oil sump interface location to the block at same elevation (main bearing centreline). This helps in manufacturing/machining ease as same surface of block is used for main bearing cap and oil sump mounting. Additionally, in short skirt design as the skirt height is lowered, form stiff bottom structure. However, option of additional strengthening of main bearing cap through cross bolting is not possible. In heavy duty diesel engines long skirt design is preferred as main bearing rigidity with additional cross bolts can be enhanced. Further long skirt design facilitates installation of ladder frame/bed plate to improve NVH design aspect of the block. In the long skirt design entire bottom surface of skirt is available for oil pan mounting which helps in avoiding complexity in machining and sealing of the oil pan interface. The Doctoral thesis on control of diesel engine noise from Agren (1994) explains the critical excitation paths in engine crankcase and suggests various remedial design versions to control the noise from engine crankcase. The improvement/reduction in the noise level for mentioned design options is also provided. Figure 12.21 shows the design features of short and long skirt design. The detailed discussions of NVH aspects of engine block are presented in following paragraphs.

12

Design and Analysis Aspects of Medium and Heavy-Duty …

Fig. 12.20 Through bolt design for Ford DIATA 1. 2 l diesel engine

Fig. 12.21 Design features of short and long skirt design

453

454

S. Thigale et al.

12.6.3 Analysis Aspects of Engine Bottom Bay 12.6.3.1

Structural Durability

The bulkhead and main bearing cap being loaded from assembly loads, firing pulsations and inertia loads from reciprocating and rotating masses subject to failure from fatigue phenomenon. Further the interface of MB cap—bulkhead is under the threat of fretting damage. The bearing shells also need to maintain certain minimum contact pressure with housing to avoid separation of the joint. The bolts connecting MB cap and bulkhead are also prone to stretching and relaxation under engine operation and are target for fatigue failure. Considering above loading and possible failure modes the bulkhead and main bearing cap are analysed in details by FEA method to assess the stresses and corresponding stress amplitudes, joint tightness/ opening, sliding/micro movements to ensure durability of the same during physical test trials and field operation. In the concept design stage, only single bulkhead is generally considered for FE calculations. The bulkhead and MB cap are coupled together by bonded contacts and pulsating peak bearing forces/loads derived from hand calculations/engine 1D analysis program such as AVL Excite are applied to assess the stress amplitude levels in the main bearing and bulkhead. Generally, stress-based optimization analysis is carried out to automatically decide on the ribbing pattern in the bulkhead. In Londhe and Sen (2010) concept level stress-based design optimization strategy and analysis procedure of bulkhead is explained. The loads and boundary conditions for simple FE analysis are shown in Fig. 12.22. Once the basic design optimization is complete, more detailed analysis including effect of assembly sequence, interferences and detailed transient dynamic loading of gas pressure and inertia is applied to virtually validate that the stress amplitudes and fatigue safety factors are met as per durability requirements. The CAE process for this step is quite involved. To obtain the detailed dynamic bearing loads over engine cycle, EHD (Elastic—Hydrodynamic) analysis with multi-body analysis to simulate the motion of the engine parts is utilized. The stiffness of components such as

Fig. 12.22 Loads and boundary conditions for bulkhead stress optimization setup

12

Design and Analysis Aspects of Medium and Heavy-Duty …

455

block, MB cap and crankshaft which has direct influence on the EHD analysis is modelled as flexible body. The P – Ɵ/Force – Ɵ is supplied as input for this analysis. The obtained transient loading on bearing cap along with assembly sequence is simulated using FE calculations. The crankcase FE model of at least three cylinders including rear end cylinder is used for this analysis. The detailed simulation flow chart for durability analysis is presented in Fig. 12.23. The multi-body engine specific tool; AVL Excite is generally used to estimate the main bearing loads. The FE analysis can be performed by commercial FEA solvers such Abaqus, Ansys, MSC Nastran etc. The fatigue safety factor calculations can be performed with commercially available tools such as FEMFAT/Fe-Safe etc. Figure 12.24 shows some of the stress plots on bulkhead and main bearing caps. The fatigue safety factor plots are shown in Fig. 12.25. The bearing contact pressure variation over engine operation is shown in Fig. 12.26. 12.6.3.2

NVH Analysis

In SAE paper on NVH optimization of 16-cylinder engine (Czerny et al. 1993) the vibration transfer from combustion chamber to outer surfaces of the engine is identified into two paths: • The outer vibration transfer path where excitation is transmitted via upper cylinder block and the cylinder head • The inner vibration path where the excitation is transmitted to the sound emitting surfaces through piston, connecting rod, crankshaft to crankcase skirt. Figure 12.27 shows the vibration transfer path in the engine. The inner path consists of components with clearances which results in impact and further generates additional excitation forces. The upper part of cylinder block and cylinder head relatively stiff and lower part of the block because of open skirt is weak and important from NVH standpoint. The (Czerny et al. 1993) provides the guideline on frequency range for acoustical importance. For automotive application acoustical frequency range starts from 600 to 800 Hz while for large marine engine frequency range starts at 300 Hz. The first step in the NVH analysis is to perform modal analysis to identify the natural modes of vibration of the component. With this possible resonance with engine excitation orders can be studied and appropriate design changes can be undertaken to avoid the resonance. The most important modes of engine crankcase are torsion, bending, main bearing cap axial and skirt lateral bending. Figure 12.28 shows the fundamental mode shapes of main bearing caps and skirt. In the initial stage of design effort is always directed towards the modification of crankcase bottom structure to increase the frequency from original levels. With these changes the response peaks at original frequencies can be avoided. The Doctoral thesis on control of diesel engine noise from Agren (1994) presents the design ideas to increase the stiffness of bottom part of crankcase. The comparison of noise level reduction for evaluations carried out with concept ideas for 6-cylinder diesel engine

456

S. Thigale et al.

Excite EHD: Inputs: 1. Coarse FE Model of Bulkhead and MB cap 2. Bearing and crankpin Geometry data 3. Loading Input (Force vs

crank angle)

Outputs: 1. Gas forces and Inertia Load history over 720degree cycle on bearing inner circumference

Fatigue with FEMAT TransMax Inputs: 1. Fine meshed model including bolts, bearing etc. 2. Load history at fatigue critical nodes mapped on bearing inner Outputs: 1. Displacements and Stress

Fatigue with FEMAT TransMax Inputs: 1. Abaqus ODB with transient load history from Excite 2. Material Property Outputs: 1. Critical nodes against endurance 2. Stress and loading history at critical nodes

Safety factor > 1.5

2. Contact Pressure, slip etc.

FEMFAT Fatigue: Inputs: 1. Abaqus ODB 2. Material properties Output: 1. Endurance safety factors

OK

FE Calculation for Fretting Inputs: A. Fine meshed model including bolts , bearings etc. B. Minimum preload both for main and side bolts C. Load history at critical nodes from transient dynamic analysis using Excite D. Material and Boundary conditions

Effective delta slip < 15 micron and contact pressure < 35 MPa

Outputs: A. Effective delta-slip between gas and inertia loading B. Contact pressure and opening ok

Fig. 12.23 Simulation flow chart for bulkhead durability analysis

12

Design and Analysis Aspects of Medium and Heavy-Duty …

Fig. 12.24 Stress plots of bulkhead and main bearing cap

Fig. 12.25 Fatigue safety factor plots of bulkhead and main bearing cap

457

458

S. Thigale et al.

Fig. 12.26 Bearing contact pressure variation from assembly to engine operation

Fig. 12.27 Vibration transfer path

is also presented. Increasing the stiffness at bearing caps by adding axial, torsion stiffeners and lateral stiffness at the crankcase skirt is shown to be effective in several papers (Kanda et al. 1990; Selmane et al. 2004; Viersbach et al. 1995; Okamura and Arai 2001; Busch et al. 1991). Common design options to increase the stiffness and frequency at the bottom end of crankcase are; introduction of bed plate between block and oil pan, ladder frame, combinations of ladder frame and bed plate, main bearing with single or double beam arrangement along with ladder frame, cross bolting to connect skirt and main bearing caps etc. The various design options of bottom bay of the crankcase discussed in Sect. 12.6.3 helps in enhancing the overall NVH performance of the engine.

12

Design and Analysis Aspects of Medium and Heavy-Duty …

459

The design options for bedplates ranges large castings constructions to thin steel sheet of 6–10 mm thick. The bedplates are cross bolted to the skirt at main bearing caps. The usage of bedplate alone may help in frequency shifts into higher frequency range, but radiation efficiency increase at higher frequency range may diminish the noise level improvements (Agren 1994). In SAE paper from Yanmar Diesel Engine (Kanda et al. 1990), it is shown that bearing beams bolted (ideally in pair) to the main bearing caps or integrated with main bearing cap can achieve the noise level reductions in the range of 6 dB (A). Figure 12.29 shows the chart of noise level improvement over speed spectrum of the engine for original and modified design, at idle (zero mean effective pressure) and a mean effective pressure of 0.7 MPa. The noise from the original design was varying from 70 to 83 dBA when speed increased from 700 to 2500 rpm when the engine was running idle and at higher mean effective pressure it was varying from 88 to 91 dBA.

Fig. 12.28 Mode shapes of main bearing cap and skirt

460

S. Thigale et al.

Fig. 12.29 Noise level comparison between original and modified design

The NVH analysis of 9 L diesel engine in Agren (1994) indicates that, the highest noise level reductions are achieved with ladder frame design. The peak sound power reduction of 9 dB was possible for 1/3rd octave bands. The measurements indicate that the stiffness of ladder frame is from the struts for 3rd order opening and closing modes of the crankcase at frequency range of 1. 6–2. 2 kHz. The noise levels of block can be further reduced by implementing isolated oil sump design. In SAE paper (Selmane et al. 2004) practical method of radiated noise CAE analysis is explained. One of the most commonly used analytical process for radiated noise evaluation is the surface velocity level (SVL) method. In this method post modal analysis, the frequency response analysis of engine block structure with excitation from main bearing and piston side is performed to evaluate the surface velocities of the radiating surfaces for frequency range of interest. These surface velocities are then squared, area weighted and summed to arrive at input sound power level of block structure for a particular frequency. This input sound power is then integrated over frequency range to estimate overall input sound power level of block. These input sound power levels are compared with baseline to judge on the improvement or degradation of radiated noise for design iterations. The problem with this approach is that, the improvement indicated in component level analysis may not reflect in engine system level analysis. In Selmane et al. (2004) case study of 5. 4 l diesel engine cylinder block is presented. The noise level calculated for modified structure of block with beam attached to bearing cap from component level SVL analysis did not match with engine system level estimation of noise levels. Figure 12.30 compare the FE model for component and system level ‘SVL’ analysis. Figure 12.31 compares the results of component level and engine system level analysis. To overcome the limitations of SVL method, another method called bolted joint acceleration method (BCA) is employed wherein the acceleration computed at block flanges are computed and compared to design iterations. Here

12

Design and Analysis Aspects of Medium and Heavy-Duty …

461

Fig. 12.30 FE model for component level and system level ‘SVL’ analysis

Fig. 12.31 Comparison of noise levels between component level and engine system level analysis

the objective is to reduce the accelerations at cylinder block bolted joints in order to reduce the overall engine radiated noise. Figure 12.32 compares the acceleration results of component level and engine system level analysis. The result correlates to the engine system level noise predictions. The detailed analysis procedure for estimation of sound radiation potential is explained in Schneider et al. (2002). This is the most accurate method to calculate the noise radiation. Further entire noise generation path can be simulated. In this complex model of all the components attached to engine is prepared. The analysis approach involves multi-body simulation of engine with elastic hydro-dynamic model of bearing to calculate the bearing reaction forces and piston slap forces for entire engine speed spectrum. This is combined with FEA to calculate the harmonic response of entire model. Coupled acoustic analysis using Boundary Element Method (BEM) helps in estimation of noise radiation potential of entire engine assembly. Generally, this analysis is performed as a final check and before start release of engine components for proto development.

462

S. Thigale et al.

Fig. 12.32 Acceleration levels for component level analysis

12.7

Design Aspects of Block Front and Rear End

The front and rear end of block forms mounting surface for many of the engine accessories including water and oil pump, fuel injection pump, camshaft, gearing system, flywheel housing, starting alternator etc. It has to perform the task of resisting the mounting loads, moments from mounting components and loads/ moments arising from the application. Following sections briefly explains the design aspects of front and rear end of crankcase.

12.7.1 Design Aspects of Block Rear End The rear end of the crankcase is the interface for PTO. The flywheel housing covering flywheel is mounted on the rear end of the crankcase. Depending on application, the flywheel housing is subjected to verities of loads which are transferred to crankcase through housing. Following are the important loads flywheel housing has to deal with: In industrial/heavy duty mobile application, mostly the flywheel housing is in the suspended state and transmission is directly mounted on the flywheel housing. Thus, during application, flywheel housing is subjected to large vertical bending moments. These loads cause significantly higher stresses on the top and bottom end of flywheel housing. The top end of the moments can be navigated by placing the top row of bolting of the flywheel housing on to crankcase as close to the flywheel housing flange as possible; shown in Fig. 12.33. This requires rear end of the crankcase stiff enough to compensate these loadings. Additionally, movement of first row of bolts close to flywheel housing top flange may introduce additional bore distortions in the active part of bore for parent bore construction. The tolerance stack-up of the bolts boss and crankcase bore can also become critical; shown in Fig. 12.34. To counter vertical bending moment at the bottom part, the crankcase

12

Design and Analysis Aspects of Medium and Heavy-Duty …

463

Fig. 12.33 Flywheel housing bolting to counter bending loads

Fig. 12.34 Flywheel housing and bore section indicating closeness of upper row bolting

and flywheel housing shall be made as stiff as possible. Additional support/bolting from sump also has to be explored. The flywheel housing is also subjected to horizontal bending loads because of vehicle cornering. The bosses on the flywheel housing and crankcase rear part should be designed to counter these loads. The flywheel housing is also subjected to short circuit torque from alternator which is essentially compensated by bolts. The rear end of the crankcase should provide

464

S. Thigale et al.

enough rigidity to the flywheel housing to shift its fundamental frequencies to higher order to avoid its resonance with dominant engine excitation orders. Another important component on the rear end of the crankcase is rear seal; which is mounted in behind the flywheel or just outside the rear main bearing. The rear seal keeps the oil leaking out where crankshaft exits the engine. On the outer periphery, rear seal avoids the oil leakage from crankcase/oil pan. This requires careful machining of the crankcase diameter (where rear seal mounts) for the superior functioning of the seal.

12.7.2 Design Aspects of Block Front End The gearing system, belt drives are mounted on the front of the crankcase. Thus, naturally most of the internal parts driven from engine power are mounted on this end. Further to avoid ingress of foreign particles and leakage of oil from the rear end; front cover is also employed at the front end. The front cover also may sometime accommodate the mounting of fuel injection pump. The front cover is one of the major noise emission surfaces in the engine. The gear cover receives excitation from crankcase forms the path for structural noise transmission; further it also transmits the noise generated from gear rattle (gear mesh impact). This requires proper stiffening and facilitates additional mounting points on the crankcase to push the gear cover frequencies to the higher order to avoid the resonances at engine excitation and gear meshing frequencies to lower the response magnitudes.

References Agren A (1994) On measurement, assessment and control of diesel engine noise. Doctoral thesis, ISSN 0348 – 8373 Busch G, Maurell R, Meyer J, Vorwerk C (1991) Investigation on influence of engine block design features on noise and vibrations. SAE Paper No. 911071 Czerny L, Schwaderlapp M, Wagner T (1993) NVH optimization of a 16-cylinder diesel engine. SAE Paper No. 932492 Flores GK (2015) Graded freeform machining of cylinder bores using form honing. SAE paper No. 2015-01-1725 Green GW, Engelstad RL (1993) A technique for the analysis of cylinder liner vibrations and cavitation. SAE Paper No. 930582 Hosny DM, Tibbetts D, Luenz R (1996) Cavitation intensity measurements for internal combustion engines. SAE Paper No. 960884 Kanda H, Okubo M, Yonezawa T (1990) Analysis of noise sources and their transfer paths in diesel engines. SAE Paper No. 900014 Londhe A, Sen A (2010) A systematic approach for design of engine crankcase through stress optimization. SAE Paper No. 2010-01-0500 Okamura H, Arai S (2001) Okamura NVH laboratory, experimental modal analysis for cylinder block-crankshaft substructure systems of six-cylinder in-line diesel engines. SAE Paper No. 2001-01-1421

12

Design and Analysis Aspects of Medium and Heavy-Duty …

465

Schneider M, Lahey HP, Steffens C, Sonntag HD (2002) CAE Process to eliminate powertrain noise and vibration. SAE Paper No. 2002-01-0459 Selmane A, Felice M, Li Y (2004) Engine cylinder blocks and heads NVH improvements: bolt accelerations computation methodology. SAE Paper No. 2004-01-0990 Viersbach U, Maurell R, Guisset P, Rossion JP (1995) Engine noise radiation—prediction and test comparison. SAE Paper No. 951342

Chapter 13

Connecting Rod Prakash R. Wani

Abstract The connecting rod converts the reciprocating motion of piston into the rotating motion of the crankshaft. Generally, it can be seen in three parts, i.e., small end, shank and big end. The connecting rod motion is complex as the small end is reciprocating along cylinder axis and big end is rotating along with the crankpin. The Loads on a connecting rod are categorized as three types namely, Firing load, Inertia load and other loads. The analysis of loads on Connecting Rod by classical method must be carried out for sizing and shaping before going for detailed analysis using the finite element method for both static and dynamic loads. The examples for the classical method are available in the appendix. The analysis for the four load cases namely, Bolt Preload and Bearing and Bush Interference, Gas Pressure Loading, Inertia Loading and Combined Loading is presented. Enhancing the yield strength and fatigue strength is achieved by choice of Materials and heat treatment. Some practical aspects during design like Weight grouping of connecting rods, Push-out force test and Testing of the connecting rod are given. The fracture splitting method for connecting rods is becoming popular as an exercise in cost reduction. The manufacturing process of connecting rod is described in brief. At the end of the chapter various failure modes are described which are borne in mind while designing the connecting rod.

13.1

Introduction

The connecting rod is an important element of engine that joins the piston and the crankshaft. It is used to achieve the rotating motion of the crankshaft from the reciprocating motion of piston. Normally it consists of small end that reciprocates with piston, big end which is attached at crankpin of a crankshaft and the shank portion that joins small and big ends.

P. R. Wani (&) Government College of Engineering, Karad, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_13

467

468

P. R. Wani

Depending on the fixation of piston pin the construction of small end can vary. In floating piston arrangement, there is a clearance between the piston and piston pin. The piston pin is fixed at the small end. This arrangement results in compact design of the piston pin. In case of floating piston pin, the connecting rod at small end is provided with the bush. The piston pin is axially stopped by two circlips with the help of grooves provided in the piston. For highly loaded engines the use of floating pins is popular. The cross section of the shank of the connecting rod is normally of I-shaped to resist buckling. The big end of the connecting rod is normally split so that the connecting rod and the cap are formed to facilitate the assembly and dismantling the connecting rod from the crankshaft. The split can be avoided if the construction of the crankshaft is of built-type, i.e., the crankshaft is now split having parts like crankpin and webs with bore to suit the crankpin assembly. The former arrangement is popularly used in case of assembly and maintenance point of view. The split of the connecting rod may be perpendicular to the connecting rod axis or inclined. The inclined split connecting rods permit increased big end bore diameter and still assembly is possible through the cylinder bore. In split connecting rod and cap, it is important that these two parts are located properly during operation of the engine. If the alignment is not maintained properly the bearing life and performance is hampered. Also, fretting failure may take place at the joint of the connecting rod and cap. To locate the cap with respect to the connecting rod the arrangements like serrations, tongue and groove type of construction are used along with dowel pin or fitted bolts. It is important that the mating faces of the connecting rod and the cap are in contact properly. The tendency of fretting is increased if the area in contact is reduced. Nowadays, some of the connecting rods are split at big end by fracture technology. The notching is done at the inner diameter by laser and the oversize mandrel with taper is pressed into the semi-finished big end bore. This results into a fracture split connecting rod, where almost 100% area of the contact is assured. However, in such cases the material of the connecting rod should be chosen with care. With this process, the manufacturing cost is appreciably reduced. However, the initial investment in machinery is high. To locate the bearing shell in rod and cap a notch or dowel is provided. It should be ensured that the bearing is not fouling with the rotating part of crankpin or fillet. In case of highly loaded engines the pistons require cooling, which is provided by the oil. The drilling through the connecting rod big end or shank should leave sufficient wall thickness around this drilled hole. Sometimes, pressurized oil is provided to the small end bearing through the drilling made inside the connecting rod.

13

Connecting Rod

13.2

469

Loads on a Connecting Rod

Firing load During combustion, the piston is subjected to the firing pressure. The forces acting on the piston are transferred to the piston pin and then to the connecting rod. The maximum load on the connecting rod due to firing pressure, p maximum load on piston ¼ D2 pfp 4 where pfp = peak firing pressure and D = piston diameter Inertia load Inertia force is induced due to the motion of the piston along the cylinder axis. It acts along the cylinder axis. The magnitude of inertia force is the product of the reciprocating mass and acceleration. The mass of the piston assembly and some fraction of the connecting rod which can be assumed as the reciprocating mass needs to be considered for this calculation. The portion of this reciprocating mass and the acceleration of the reciprocating mass can be evaluated as discussed in Appendix 1 Distribution of connecting rod mass at small end and big end can be shown as follows. Wassly ¼ Wrecip þ Wrot
 L1 þ L2 Wassly ¼ Wrecip L2 Or
Wrecip ¼ Wassly

L2 L1 þ L2



The displacement x of piston assembly from the TDC at the time when the crank angle is h, can be shown to be ) x ¼ ðL þ RÞ  ðL cos u þ R cos hÞ

470

P. R. Wani

Or ¼ 1 þ k4  cosh  k4 cos 2h  velocity; v ¼ x_ ¼ xR sin h þ k2 cos 2h dv dh dv acceleration; a ¼ v_ ¼ €x ¼ dh dt ¼ dh x 2 cos 2h ¼ Rx cos h þ n x R

where n¼

1 k


 cos 2h )Inertiaload ¼ mrecip  acceleration ¼ mrecip Rx2 cos h þ n This force is varying with respect to crank angle h. The first part involving cosine h term is called the primary inertia force and the second part involving cosine 2h is called the secondary inertia force. The frequency of the secondary force is twice that of the primary inertia force.
maximum inertia force ¼ mrecip Rx

2

1 1þ n



where h ¼ 0; p; 2p; . . . It is to be noted that the maximum firing load and the maximum inertia load are not acting at the same instance. Generally, at the start of suction stroke for a four-stroke cycle engine, the inertia force may be predominant while at the start of the power stroke, firing load may be predominant. Thus, in a working cycle the load is fluctuating. As the inertia load is opposing the firing load at firing TDC position, the resultant load is reduced to that extent. Other loads The dimensions of cross section of the shank are based on the buckling load. The permissible buckling load as per Rankine-Gordon formula for a beam under compression is Fxx ¼

fc A  2 1 þ a Rlxx

13

Connecting Rod

471

where Fxx = critical load with respect to xx axis of the cross section Fc = allowable compressive stress A = minimum cross section area of the shank a = constant depending on material   l Rxx

= slenderness ratio

L = length of the connecting rod from small end to big end Rxx is the radius of gyration of cross section about xx axis The whipping stress acting on the cross section is calculated as follows. The average maximum whipping force acting due to the weight of the connecting rod is Fi ¼

qAl 2 Rx 2g

where q = weight density of connecting rod material A = cross section area R = crank throw or crank radius l = length of connecting rod Fi = Force on connecting rod due to rotation Maximum bending moment, 2Fi l Mmax ¼ pffiffiffi g 3 Maximum whipping stress, rb ¼

Mmax Mmax ¼  I Z y

where I = area inertia of the cross section y = distance of outermost fibre from the neutral centre of I section The dimensions of small end bearing side diameter and length can be estimated by application of bearing stress criterion, i.e.

472

P. R. Wani

Allow able bearing stress ¼ ¼

Fmax on small end projected bearing area Fmax diameter of small end  length of bearing at small end

A similar criterion can be applied for the big end. As the small end bush or big end bearing should be fitted in such a way to avoid the rotation relative to the housing, the radial and hoop stresses may be calculated by applying the thick and thin composite cylinder theory.

13.3

Load Analysis of Connecting Rod by Classical Method

In general, for the connecting rod analysis the I-section calculations are performed to confirm that sufficient margin is available against the critical buckling load. The section with minimum cross section area is chosen for such calculations. The increase in stresses due to stress concentration factor at the junctions is either taken approximately or factor of safety requirements are increased accordingly. As a guideline I section dimensions with proportion of 4T-5T-T are usually assumed. The details of reasoning are explained in the Appendix 2. The small end and big end calculations are done with some approximations by applying the curved beam theory. Please refer Appendix 3 for details (Bremi 1971). The interference fit between the small end bush and the connecting rod is treated as the composite cylinder so that radial and hoop stresses can be calculated. Similar treatment can be given at big end. At the big end, if the connecting rod is split and bolting is used, it is necessary to consider additional load due to bolt clamping. The calculation of big end bolt can be performed based on the standard VDI 2230 (VDI 2230). The sample calculation for a typical case is shown in Appendix 4. Whipping stresses are combined with the stresses due to firing loads and inertia loads. The maximum compressive stress and maximum tensile stress are taken for approximate calculation of fatigue stress. It is then compared with the endurance stress of the material chosen for the connecting rod. Additional aspects like manufacturing variations should be given due consideration so that variation of dimensions and mechanical properties are covered while finalising the design.

13

Connecting Rod

13.4

473

Load Analysis of Connecting Rod by Finite Element Method

Analysis Details It is observed that peak firing pressure in diesel engines occurs after the TDC (top dead centre) position. However, for initial simulation it can assumed as occurring at TDC. Maximum inertia load will act at TDC position. An axis-symmetric model of Connecting Rod Subassembly is modelled. The model is meshed, assembled and material properties are applied (Pravardhan 2004). Figure 13.1 shows some critical locations for stress calculations in a connecting rod. Load Cases: Case 1: Bolt Preload and Bearing and Bush Interference In case of the split connecting rods the cap and rod are fastened together with nut and bolt type of arrangement. The tightening of the bolt with the nut or inside the tapping of the connecting rod is required to hold these parts including the big end bearing. This assembly induces the compressive loads in the connecting rod and cap. Due to tightening the bolt gets stretched and is subjected to the tensile load. The bearings at the small end and at the big end are subjected to the loads caused by the interference fit. The stresses due to such assembly conditions are in fact present in the parts even before the engine is in operation. This case is simulated and treated as the preload on the assembly. Case 2: Gas Pressure Loading The contact condition of the piston pin and the bush is considered as clearance fit. Similarly, at the crank pin bearing, clearance is considered. Crank pin is modelled as rigid surface and constrained in all 6 degrees of freedom. The bolt is included in the model, but not preloaded. One fourth of this maximum gas force is applied on

Fig. 13.1 A typical stress distribution at some locations of a connecting rod

474

P. R. Wani

the piston pin. For application of this force reference point is created and then it is coupled to the piston pin upper surface. Case 3: Inertia Loading The inertia force is calculated at maximum over-speed condition. The bolt is included in the model, but not preloaded. One fourth of this maximum inertia force is applied on the piston pin considering symmetric model. For application of this force a reference point is created and then it is coupled to the piston pin lower surface. In this case the possibility of split line separation is also evaluated during inertia loading of the connecting rod. Case 4: Combined Loading In this case the two combined loadings are simulated. Combined load1 ¼ bolt load þ bearing bust interference load þ gas pressure and Combined load2 ¼ bolt load þ bearing bust interference load þ inertia load Most of points on shank subjected to more compressive stresses and the points at outer side of big end and small end are subjected to more tensile stresses. The points at the junctions of small end to I section and at big end to I section are subjected to multi axial state of stress. The equivalent stress amplitude can be calculated based on Von Mises criterion, as follows:

Sqa ¼

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2 u 2  2 u 2 2 þ t2 þ t2 t Sax  Say þ Say  Saz þ ðSaz  Sax Þ þ 6 taxy ayz azx 2

The equivalent mean stress is calculated as follows: Sqm ¼ Smx þ Smy þ Smz It had been observed that mean shear stress had no effect on cyclic bending or cyclic torsion fatigue limits. After obtaining the equivalent mean stress and stress amplitude, the equivalent stress amplitude at R = -1 (corresponds to SNf) was obtained by using the commonly used modified Goodman equation: Sqa Sqm þ ¼1 SNf Su

13

Connecting Rod

475

Fig. 13.2 A model and typical stress distribution at some locations of a connecting rod

The Fig. 13.2 shows the model of rod and the stresses at big end. Initial rough calculations can be performed on 2D FE models as it can be completed relatively quickly. Once the trend is known 3D FE models can be used for more detailed analysis. Various factors can be considered for optimization of the connecting rod. Generally, cost optimization is main objective to deal with the competition in the market. However, in some applications like defence, the packaging dimensions may have more importance. The limited life, high cost material may be allowed in such cases to obtain the compact equipment. To be on the conservative side the maximum tensile load is taken as the load at maximum speed and assumed to be occurring at TDC. Also, for the compressive load calculations peak firing pressure is considered. The manufacturing constraints need to be given a proper thought at the design stage. Depending on the material selected and the mechanical properties sufficient factor of safety should be assured against the buckling load. Fatigue strength is the most significant factor (i.e., design driving factor) in the design and optimization of the connecting rod. To design for fatigue, modified Goodman equation may be used. The fatigue stress amplitude and the maximum stress value should be less than the allowable values. Additional constraints imposed during the optimization process may be maintaining the forge ability as well as interchange ability of the connecting rod with the existing one. Cost may be reduced by changing the material to crackable forged steel in place of regular steel forging.

476

13.5

P. R. Wani

Materials and Heat Treatment

For the connecting rods of engines with low brake mean effective pressure generally the low-cost material is preferred. Hardenable medium carbon materials like CK 45, En9, and En8D are used in normalised, quenched and tempered condition. For high BMEP engines generally materials like En16, 42CrMo4 are used in NQT condition. The failure of the connecting rod may result into breakage of surrounding parts. In such cases the elongation of the steel material reduces the damage to some extent. Spheroidal Graphite iron material of grade 600 or 700 is sometimes used for low duty connecting rods. However, to remove the bend or twist of such rods is very difficult during the manufacturing process. For fracture split connecting rods the crackable steel material C-70 is used. It is expected in this case that the brittle crack is formed at the big end of the connecting rod at the time of splitting process.

13.6

Some Practical Aspects During Design

The connecting rod experiences the combination of different loads. The varying cyclic load acting on the rod makes it critical from durability aspect. Being one of the central parts of the engine, deep thought is required during design stage itself. If there is a change in dimensions of the connecting rod, the other parts in the vicinity may require the alterations those may be difficult to accommodate. The changes in connecting rod processing are also costly and time consuming. For these reasons the study of latest production technologies, finite element modelling and simulations, optimization techniques are usually applied. Application of new materials is also the subject of research for making of the connecting rods. If the connecting rod is split perpendicular to the connecting rod axis, the nut and the bolt arrangement may be used to clamp the cap with the rod. In such cases the bolt head is special so that it locks the bolt rotation while tightening the nut. When the connecting rod split is inclined the bolt is used with tapping on the connecting rod. In such a case no nut is required. Normally, one of the tapping is blind and hence the minimum wall thickness around the tapped hole must be critically ensured. In vee engines, if the connecting rods are placed side by side, the size of the inside chamfer of the big end bore can be reduced the two connecting-rod mating side. This can be possible as there is no question of crank pin fillet riding on the bearing. If compact size of engine is desired in vee engines, a pair of articulated master and slave rod is used. In such a case it is like a radial engine with a hole at big end on the master rod that accommodates the big end of the slave rod. For such engines, the stroke of the two piston rod assemblies is slightly different. It must be ensured from the layouts that Connecting rod is not interfering with other parts like crankcase, cylinder liner, piston, piston cooling nozzles, balance weights etc.

13

Connecting Rod

477

Dimensional and form tolerances on big end bore diameter, small end bore, centre to centre distance between the big end and small end, bend and twist should be kept as minimum as possible. The bolt-hole axis should be maintained square to the mating faces. In case of the split connecting rods, the tightening of the bolts is an important aspect. These bolts are used for holding the rod, cap and big end bearings together. The torque on the bolt induces the preload on the bolt and other mating parts. The systematic calculations for the bolted joint can be performed by referring to VDI 2230 standard (VDI 2230). If the split of the connecting rod is inclined, the bolts must also resist the side force. The magnitude of the side force Q can be evaluated based on equations in Appendix 3. Usually for this purpose the connecting rods are serrated at the parting faces. Some connecting rods may have tongue and groove type of arrangement. With serrations or tongue-groove combination it is possible to restrict the motion of cap relative to connecting rod in the plane of their parting surfaces.

13.6.1 Weight Grouping of Connecting Rods As the inertia force due to connecting rod mass is varying in magnitude during engine running, it creates varying force on the foundation for a single cylinder engine. To have the consistent behaviour of the engine, the weight of the connecting rod and the piston assembly should be in a closed interval and not with wide spread. For multiple cylinder engines, it becomes necessary to control the weights of the connecting rods in the same engine to maintain the inertia force magnitude at equal level. In case of high-speed engines, the control on distribution of weight is important. Close tolerance on weight distribution can keep vibrations are under control. The weight of the connecting rod can be adjusted by removal of material that is provided additionally on the big end and or small end side. This removal should not impair the strength against the operational loading on the connecting rod.

13.6.2 Push-Out Force Test The interference force between the bush and the connecting rod resists the rotation of small end bush. This resistance is proportional to the push out force. The push out force is the force required to push the bush from its assembled position in the connecting rod. Generally, a suitable fixture is made to find this force and it can be estimated by a load cell or on a universal testing machine. The push out force is dependent on the interference between the mating parts. It also depends on the modulus of elasticity of the assembled components, surface finish of the components and the assembly process e.g., use of hand press or liquid nitrogen etc. (Wani et al. 2005).

478

P. R. Wani

13.6.3 Testing of the Connecting Rod The design calculations are generally based on the material properties of the machined and ground specimen whether they are for endurance limit or the ultimate strength. The fatigue strength is function of material, component size, notch sensitivity, surface roughness etc. Hence the component needs to be tested. To account for the process variation number of components are tested for endurance. For the connecting rod under study it may be a costly affair to build the complete engine to study the reliability. The components like connecting rods are tested for fatigue by staircase method. In staircase method, specimens are tested sequentially, with first one being tested at a load level equal to the estimated median fatigue limit, using preliminary information to define this value. If first specimen withstands the load level, second specimen is tested at higher load level. If the first specimen fails before reaching pre-decided number of load cycles, second specimen is tested a lower load level. Load levels thus jump up or down, depending on whether current test specimen fails or survives. The fixture should be made strong enough to sustain the higher fatigue loads. The fixture is not supposed to fail earlier than the sample connecting rods. To facilitate the evaluation in both applications viz., power generation and industrial, by a single series of fatigue test, the two most severe cases are generally selected for the basis of test loading. The maximum tensile and compressive forces in the operating conditions are estimated. The safety margin factor of say 1.4 is applied at the start of the stair case method with an incremental step of 0.2. Mean safety factor and standard deviation are calculated statistically. Figure 13.3 shows Safety Coefficient for connecting rods by staircase method (Wani et al. 2005). During testing on the rig, the specimen under test must be produced with due care to represent the final components. The assembly needs to be done carefully so that no additional loading is induced because of the misalignment.

Fig. 13.3 Safety Coefficient for connecting rods by staircase method (Wani et al. 2005)

13

Connecting Rod

13.7

479

Improvement in Fatigue Strength

Fatigue strength of the connecting rod can be improved by shot peening process. In shot-peening, the steel balls are bombarded with help of nozzles on the rod. The shot-peening may be done only on un-machined part of the connecting rod. Due to balls impacting on the rod a layer of compressive stress is created. The improvement in the fatigue strength is a function of intensity of shot peening and the depth of peening surface. The fatigue strength of the forged surface of the rod is lower than that of the machined smooth surface. Shot peening on forged surface help to improve the fatigue strength. The impact of decarburization on fatigue life was investigated by Ilia and Chernenkoff (2001). They concluded that decarburized layer depths equal to or higher than 0.4 mm decrease the fatigue life.

13.8

Connecting Rods Produced by Fracture Splitting

The big end of a steel connecting rod is split by a method known as “Fracture splitting” method. It is separated by producing a fine crack from two notches using hydraulic force. For the split surfaces to be precise and to retain the shape and dimensions of the connecting rod during the process of splitting, the fracture should be brittle in nature. Normally forged steel connecting rod undergoes plastic deformation before actual fracture and hence is not suitable for this process. In case of carburised low carbon steel, the core of the fracture will not be brittle. Generally, C70 steel is used because of its crackability, higher strength and lower cost. It may be possible to further reduce the weight at the small and big ends using other facture crackable materials like micro-alloyed steels. Weight of the shank region cannot be easily reduced due to limitations of manufacturing constraints. To avoid the multiple fracture surfaces in-depth study is done from non-linear simulations (Kubota et al. 2004). Some of the process steps in the conventional manufacturing sequence are eliminated like machining of the mating faces of the cap and rod, and drilling and reaming for dowels. About 25% reduction in production cost and 15% reduction in overall cost are achieved by this process. The distinct surface of the cracked surface restricts the relative movement of the rod and cap, providing firmness and stiffness to the joint. In addition, the stresses at critical locations at the split are reduced.

480

P. R. Wani

13.8.1 Process Sequence for Connecting Rod by Fracture Split • • • • • • • • • • • • •

Rough Grinding Rod Small End (S.E.) Drilling and Boring Bolt Seat milling Bolt Hole Drilling and Tapping Big End (B.E.) semi finish Boring Laser notching Fracturing Assembly of Rod and Cap. B.E. Facing and Chamfer B.E. Finish and S.E. Bush Bore Notch Milling B.E. Honing Deburring and Cleaning

It has been seen that laser notching exhibited best fracture splitting results, when compared with broached and wire cut notches. Research on development of lightweight connecting rods based on fatigue resistance analysis of micro alloyed steel was conducted by Kuratomi et al. (1990). The study found that the micro alloyed steel exhibits lower fatigue strength than the quenched and tempered steel for smooth specimens, but equivalent or higher fatigue strength for notched specimens. High fatigue strength free machining micro alloyed steel was developed and used for connecting rods by Nakamura et al. (1993). The influence of alloy elements such as C, Mn, Cr, V, S, Pb, and Ca, and their impact on fatigue strength and machinability were discussed. A 0.33%C-1.05%Mn-0.5%Cr-0.12%V-0.55% S-0.20%Pb–Ca composition was found to be the best composition to improve fatigue strength. For the cap splitting process, the connecting rods had 45° notches about 0.5 mm deep, machined on the sides of the bore and on the edges of the rod in the same plane as the bore notches.

13.8.2 Costing Comparison Despite the substantially lower weight of the material used, however, the cost of the powder forged rough stock could be higher than that for the conventional hot drop-forged rough stock, because of additional operations of powder formation, pre-form formation, pre-sintering, and sintering (Afzal and Fatemi 2004). With recent introduction of new materials such as C-70 splittable steel, this key advantage of powder metal connecting rods no longer exists, as machining of matching

13

Connecting Rod

481

surfaces of splittable steel are no longer required. As mentioned in the literature review, fracture-split technology as applied to forged steel connecting rod cuts the total cost by 25%, compared to the conventional forged steel connecting rod. Dipak et al. (2010), Kubota et al. (2004), Ashley (1991). The C-70 steel has higher strength than the powder metal material. Other alternative steels are also being developed with higher fatigue strength. Micro-alloyed 36 MnVS4 steel shows better fatigue strength than C-70 steel. For this new micro-alloyed steel, the component tests on materials showed 15–20% increase in fatigue strength (Pravardhan and Ali 2005). For fracture split connecting rod it is important to have the parting at the expected plane, typically like the conventional parting line made by slitting process. One notch machined on the rod before fracture may not give this expected result. The crack does not propagate along a straight plane and it makes the section weak. The crack is shown in Fig. 13.4. Additional notches on the top and bottom surface are sometimes tried as shown in red, Fig. 13.5.

Fig. 13.4 Crack formation during fracture splitting of the connecting rod, by providing notches on the interior surface of the big end bore

Fig. 13.5 Additional notches on the big end, on the top and bottom surface

482

13.9

P. R. Wani

Failure Modes

Figures 13.6, 13.7 and 13.8 show the modes of failure of the connecting rods. Generally, connecting rod failure is due to fatigue and often it is seen at the junction of small end to I-section. The rod in Fig. 13.5 seems to have run for number of hours even after the breakage. The rods in Figs. 13.5 and 13.6 once again appear to be failing at the small end due to fatigue and high compressive load. The failure of connecting rod may be caused sometimes due to the fold defect during forging process. Hydraulic shock due to leakage of water in the cylinder chamber can sometimes create a very high pressure resulting into severe bending of a connecting rod.

Fig. 13.6 A failure of a connecting rod at junction of small end

Fig. 13.7 A failure of a connecting rod at small end (Wani et al. 2005)

13

Connecting Rod

483

Fig. 13.8 A failure of a connecting rod at small end

Acknowledgements Authors acknowledge with thanks SAE, Society of Automotive Engineers for granting the permission to use Figs. 13.3 and 13.7 from reference (Wani et al. 2005) through Copyright Clearance Centre, www.copyright.com. They are thankful to “Sulzer Tech. Review and Sulzer Management Limited, Switzerland” for permitting to use the Figs. 13.13, 13.14, 13.15, 13.16 and 13.17 along with the equations in Annexure III from reference (Bremi 1971).

Appendix 1 Distribution of connecting rod mass at small end and big end (refer Figs. 13.9 and 13.10) Let L ¼ centre distance between Small end and big end L1 ¼ distance of the Centre of gravity ðCGÞ from the small end L2 ¼ distance of CG from big end L ¼ L1 þ L2

Fig. 13.9 Location of C.G. for the connecting rod

484

P. R. Wani

Fig. 13.10 Distribution of reciprocating and rotating masses for con rod

Wassly ¼ weight of connecting rod assembly Wrecip ¼ reciprocating portion of the connecting rod assembly Wrot ¼ rotating portion of the connecting rod assembly Wassly ¼ Wrecip þ Wrot Also, the moment of forces at CG is zero. Wrecip  L1  Wrot  L2 ¼ 0 Therefore, Wrot Wassly


 L1 ¼ Wrecip L2

 L 1 þ L2 L2 ¼ Wrecip Or Wrecip ¼ Wassly L2 L1 þ L2

L ¼ centre distance between the small end and the big end R ¼ crank radius h ¼ crank angle u ¼ obliquity angle x ¼ displacement of piston assembly from the TDC at h From the Fig. 13.11, L þ R ¼ x þ L cos u þ R cos h ) x ¼ ðL þ RÞ  ðL cos u þ R cos hÞ

13

Connecting Rod

485

Fig. 13.11 Sketch of connecting rod at crank angle h

But, L sin u ¼ R sin h x ¼ R


 ffi 1 1 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1þ  cos h  1  k2 sin2 h k k

Where k¼

R L

Fourier development of the equation gives x A2 A4 A6 ¼ A0 þ A1 cos h  cos 2h þ cos 4h  cos 6h þ    R 4 16 36 Getting the values of A0, A1, A2… in terms of k and neglecting higher order terms of k, k being less than 1, we get,

486

P. R. Wani

i.e. x k k ¼ 1 þ  cos h  cos 2h R 4
4  k velocity; v ¼ x_ ¼ xR sin h þ cos 2h 2 dv dh dv acceleration; a ¼ v_ ¼ €x ¼ ¼ x dh dt dh 
cos 2h ¼ Rx2 cos h þ n Where n¼

1 k




)Inertiaload ¼ mrecip  acceleration ¼ mrecip Rx

2

cos 2h cos h þ n



Appendix 2 Inertia of I-section The connecting rod small end and big end form a hinged joint in one plane where bend is tested. The other plane where twist is tested, the rod has more resistance to bending against the compressive load. In the latter case, the equivalent length is given below calculating the slenderness ratio. l¼

lactual 2

Hence to have the connecting rod equally strong about the both the axes, the critical buckling load, Fi should be such that F¼

fc A fc A  2 ¼  2 1 þ a Rlxx 1 þ a 2Rl yy

i.e.


1 Rxx

or Ixx ¼ Iyy since I ¼ AR2 we get

2


¼

1 Ryy

2

13

Connecting Rod

487

Fig. 13.12 Typical dimensions of I section

Ixx ¼ 4Iyy With the I-section defined by 4 t. 5 t. t as shown in the Fig. 13.12. This selection of cross section is taken as a starting guide line.

Appendix 3 Calculation of the connecting rod Big end (Bremi 1971) Stresses at the big end is arrived at by the theory of curved beams. The big end is pulled by the inertia forces. Half connecting rod split by the plane passing through the centre line of the cylinder can be imagined as a hook. The shear and normal forces and the bending moment at the split. Since the split is in equilibrium the two halves of the connecting rod the respective forces on the faces are equal. The reader is referred to the work done by P. Bremi (Bremi 1971). The formulae involved in brief are shown below. Refer Figs. 13.13, 13.14, 13.15 and 13.16 for simplified connecting rod model.

Fig. 13.13 Big end of connecting rod as a curved beam (Bremi 1971)

488 Fig. 13.14 Simplified rectangular cross section (Bremi 1971)

Fig. 13.15 Location of general cross section (Bremi 1971)

Fig. 13.16 Sketch showing various parameters of big end (Bremi 1971)

P. R. Wani

13

Connecting Rod

489

Bending stress, rðzÞ ¼

N M M z þ þ F rF rkF r þ z

ð13:1Þ

where Z

z2

kF ¼ 

z bðzÞdz rþz

z1

IðzÞðr þ zÞbðzÞ ¼ 

DkF Qþ kF

Z

z

I ðz0 ÞbðzÞdz0

ð13:2Þ ð13:3Þ

z1

where Z

z1

z0 bðz0 Þdz0 r þ z0

Z

z

z

DkF ¼ 

ð13:4Þ

If DF ¼

bðz0 Þdz0

ð13:5Þ

z1

Then

Q DF DkF  IðzÞ ¼ bðzÞðr þ zÞ F kF

ð13:6Þ

Deformation energy of the curved beam U¼ Z U¼

/2

/1

Z

z2 z1

r2 2E

r2 ðr þ zÞbðzÞdzd/ 2E

ð13:7Þ ð13:8Þ

Substituting from (13.1) Z U¼

/2

/1




1 Nr2 NM M 2 1 þ þ2 1þ d/ 2E F F k rF IðzÞ ¼


 Q h2 2  z 6h3 4

ð13:9Þ ð13:10Þ

490

P. R. Wani

For comparison, a still simpler equation Q bh

ð13:11Þ

J2 ; 2G

ð13:12Þ

J 0 ðzÞ ¼ UQ ¼ where G = shear modulus G¼

E ; 2ð1 þ cÞ

ð13:13Þ

where c = Poisson ratio Deformation energy of a beam piece of length, l UQ ¼

Z lZ

h=2

J2 b dz dx h=2 2G

0

ð13:14Þ

Substituting (10) and (11) in the above equation, Z

l

Deformation energy due to Q; UQ ¼ 0

3 Q2 dx 5 Gbh

ð13:15Þ

Or UQ0 ¼

Z 0

l

1 Q2 dx 2 Gbh

ð13:16Þ

When the beam is loaded with moment M, Z

l

Deformation energy due to M; UM ¼ 0

6M 2 dx Ebh3

ð13:17Þ

If the beam is loaded at the end with a force P, Q¼P

ð13:18Þ

M ¼ Px

ð13:19Þ

and

U¼



2P2 l3 3 h2 ð1 þ mÞ 1 þ 5 Ebh3 l2

ð13:20Þ

13

Connecting Rod

491

Or

P 2 l3 1 h2 U ¼2 1 þ ð1 þ mÞ 2 2 Ebh3 l 0

ð13:21Þ

In general case of a connecting rod, If 1 and l ¼ 3h 3

m¼ Then, U¼

2P2 l3 ½1 þ 0:09 Ebh3

ð13:22Þ

And U0 ¼ 2

P 2 l3 ½1 þ 0:06 Ebh3

ð13:23Þ

The shear stress is not negligible since its contribution to the distortion energy is about 10% of that of the normal stress. Since Uis within 1% of U’ is only 1% Eq. 13.11 is used without loss of accuracy, instead of Eq. 13.10 while calculating the distortion energy. For a bent beam, JðzÞ ¼ Z UQ ¼

/2 /1

Q F

1 Q2 rdu 2G F

ð13:24Þ ð13:25Þ

This energy can have the energy as Eq. 13.9 superimposed on it so that the following formula is valid for the deformation energy of the bent beam. Z U¼

r2

r1



1 þ 1k 2 1 E rN 2 þ 2NM þ m þ r Q2 d/ 2EF G r

Displacement due to Nl = nl Displacement due to Ql = ql Displacement due to Ml = ml Ref Fig. 13.16

ð13:26Þ

492

P. R. Wani

We define a¼pb

ð13:27Þ

n o p p K/ ¼ 0 for 0  /  ; K for  /  a 2 2

ð13:28Þ

And

for the split beam shown in figure. In the split beam, there will be no load N, u and C, moment created in the section of the connecting rod. The following relationships are obtained. N/ ¼ N cos / þ Q sin /  K/ cos /

ð13:29Þ

Q/ ¼ N sin / þ Q cos / þ K/ sin /

ð13:30Þ

Moment of inertia of all forces at the centre of connecting rod,   M/ ¼ N r0  r/ cos /  Qr/ sin / þ K/ sin /

ð13:31Þ

Substituting in Eq. 13.26, the equations for deformation energy of half big end are as follows: U¼

1 aNN N 2 þ aNQ NQ þ aNM NM þ aNK NK þ aQN QN þ aQQ Q2 2 þ aQM QM þ aQK QK þ aMN MN þ aMQ MQ þ aMQ MQ þ aMM M 2 þ aMK MK þ aKN KN þ aKQ KQ þ akM KM þ aKK K 2 ð13:32Þ "

2  2
r/ r/ r0 1 r/ r cos2 / þ 2 ¼  cos / þ 1 þ  cos / k r 2 r/ r r/ 0 EF

E þ sin2 / d/ G Z

aNN

a

ð13:33Þ aQQ

"

#


 2 r/ r/ 2 1 r/ 2 E 2 sin /  2 sin / þ 1 þ ¼ sin / þ cos2 / d/ ð13:34Þ 2 k G EF r r 0 Za

aMM

Za

r/ ¼ EF 0

"


#  2 1 r/ 1 1þ  d/ k r 2 r/2

ð13:35Þ

13

Connecting Rod

493

" #
 2 r/ r/ 1 1 r/ 1 sin /  1 þ ¼ sin / d/ r/ k r2 r/ EF r 0

ð13:36Þ

"  #
 2
r/ r/ 1 1 r/ r 1 cos / þ 1 þ ¼  cos/ d/ 2 r/ k r r/ r/ EF r 0

ð13:37Þ

Za

aQM ¼ aMQ

aMN ¼ aNM

Za


 r/ r/ r/ r cos / sin /  cos / sin / þ sin /  cos / r/ r r 0 EF #
 2
 r 1 / r E  1þ  cos/ sin /  sin / cos / d/ k r 2 r/ G Z

aNQ ¼ aQN ¼

a


 r/ K / r/ r/ r  cos2 / þ cos2 /  cos /  cos / r/ r r 0 EF K #
 2
 r 1 / r E  1þ  cos/ cos /  sin2 / d/ k r 2 r/ G Z

aNK ¼ aKN ¼

a

Z

aKK

ð13:39Þ

r/ K/ h r/  sin / cos / þ sin / cos / EF K r 0 #
 2 1 r/ E  1þ sin / cos /  sin / cos / d/ k r2 G

ð13:40Þ

" #
 2 r/ K / r/ 1 1 r/ 1  cos / þ 1 þ ¼ cos / d/ r/ k r2 r/ EF K r 0

ð13:41Þ

aQK ¼ aKQ ¼

aMK ¼ aKM

ð13:38Þ

a

Za

" #
 2 2 r/ K/ r/ 1 r/ E 2 2 2 2 ¼ cos /  2 cos / þ 1 þ cos / þ sin / d/ k r2 G EF K 2 r 0 Za

ð13:42Þ Stresses on the big end n¼

@U ¼ aNN N þ aNQ Q þ aNM M þ aNK K @N

ð13:43Þ

q¼

@U ¼ aQN N þ aQQ Q þ aQM M þ aQK K @Q

ð13:44Þ

m¼

@U ¼ aMN N þ aMQ Q þ aMM M þ aMK K @M

ð13:45Þ

To these we add the following quantity

494

P. R. Wani

k¼

@U ¼ aKN N þ aKQ Q þ aKM M þ aKK K @K

ð13:46Þ

For the above calculations, the connecting rod is assumed to be split. In actual case, it is not. Now, Nl ¼ Nr

ð13:47Þ

Ql þ Qr ¼ P

ð13:48Þ

Ml ¼ Mr

ð13:49Þ

nl ¼ nr

ð13:50Þ

ql ¼ qr

ð13:51Þ

ml ¼ mr

ð13:52Þ

A set of linear equations for the six unknowns Nl, Ql, ml, Nr, Qr and Mr is obtained using Eqs. 43–45 for each half of the big end. This system can be easily converted back to the system 

     alNN þ arNN N l þ alNQ  arNQ Ql þ alNM þ arNM M l ¼ arNQ P  alNK K l  arNK K r ð13:53Þ

      alQN  arQN N l þ alQQ þ arNQ Ql þ alQM  arNM M l ¼ arQQ P  alQK K l þ arQK K r ð13:54Þ    l    aMN þ arMN N l þ alMQ  arMQ Ql þ alMM þ arMM M l ¼ arMQ P  alMK K l  arMK K r

ð13:55Þ

This selection of cross section is taken as a starting guide line. The calculations are iterated for the modified configuration, depending on the magnitude of the stresses. A typical stress distribution is shown in Fig. 13.17.

13

Connecting Rod

495

Fig. 13.17 Stresses in big end of the connecting rod (Bremi 1971)

Appendix 4 Connecting rod Big end bolt sample calculation The calculations are performed for a connecting rod having two bolts. The side force Q is not considered as the rod assumed with horizontal split. Bolt sizeM12  1.5 Bolt Quality10.9 mrecip. 1.720 kg. mrod without cap1.305 kg. r ¼ crank radius ¼ 0:054 m 2pn where n ¼ 3150 rpm 330 rad/sec: -¼ 60 L ¼ con: rod centre distance 0:187 m  r Fmax ¼ ðm recip þ m rodÞr-2 1 þ 22925 N L Bolt torque10 kgm with oil (80% yield) Preload as per VDI 2230 (VDI 2230) is estimated as 54004 N per bolt Total preload108008 N The force Finterference due to bearing overstandAssuming 25 kg/mm2 stress on bearing,

496

P. R. Wani

The dimensions of bearings: 2 mm thick  23 mm length Area on one side ¼ 46 mm2 Force on each side 25  46 ¼ 1150 kg Force on both sides ¼ 2300 kg Force on each bolt ¼ 1150 kg )Finterference ¼ 11280 N )Total load on bolt ¼ Fmax þ Finterference ¼ 22925 þ 11280 N ¼ 34205 N 108008 )Cover factor ¼ 34205 ¼ 3:16 [ 2 Hence safe Connecting rod Big end bolt detailed calculation as per VDI 2230 The connecting rod big end joint as shown in Fig. 13.18 is example of eccentrically clamped eccentrically loaded joint. The analysis is to be done for 12.9 grade M9  1 bolt and grade-12 nut. The bolts are tightened with a high precision tightening spindle. C45 was selected as the material for the clamped parts. For the rated engine speed (n = 4000 rpm), we have the following initial parameters. Axial force at the interface Fa ¼ 3:7  103 N

Fig. 13.18 Forces acting on the interface of a connecting rod bearing cap bolted joint

13

Connecting Rod

497

Dw min Da max D1 Dh Dha max Dhw min

13.8 9.9 7.9 9.43 10 14

dhw min dha max d1 dh max dw min dc

14 9.65 9.25 9.0 13.2 13.4

Fig. 13.19 Dimensions(mm) of the clamping and clamped parts of the connecting rod bearing cap bolted joint and the nut and head bearing surfaces

498

P. R. Wani

Bending moment at the interface MbA ¼ 30 Nm Transverse force at the interface FQ ¼ 420 N From the bending moment MbA and the axial load FA, the lever arm of the eccentric load application is determined as a¼

MbA ¼ 8:1 mm FA

Assuming for simplicity that the bending moment is constant over the clamping length lK Calculation procedure The calculations are made following the calculations steps R1 to R10 given in Sect. 4.1 VDI 2230. R1Rough determination of the bolt diameter d and the clamping length ratio lK/d. The bolt diameter is given as 9 mm from the design shown in Fig. 13.19. The clamping length ration is lK 41:5 ¼ 4:6 ¼ 9 d Rough determination of the surface pressure under the bolt head: p¼

FlA 0:9

AP

 PG

Measurement of screw in mm Place

l1

d1

2.5 3.0 6.5 10.0 15.8 3.7

8.7 9.2 9.0 9.2 8.35 7.68

13

Connecting Rod

499

Other Measurements Thread d2 d3 d P Ad3

Clamped parts 8.27 7.66 9 7 46 mm2

DA db lg g a u v w n

Nut bearing 15.5 9.25 41.5 0.3 8.1 5.3 6.8 1.4 17.6

Dw, min Db, max D1 dh Dh0, max Dh0, min

Head bearing 13.8 9.9 7.9 9.43 10 14

dhw, min dho, min dh df db, max dw, min dp

14 9.65 9.25 8.7 9 13.2 13.4

FM ¼ 42:6  103 N for lG ¼ 0:12 from Table 3 of VDI 2230. With  p 2 dW;min  dh2 a;max 4 ¼ 64 mm2

AP head;min ¼

and  p 2 Dw;min  D2ha;max 4 ¼ 71 mm2

APnut;min ¼

follows Pmax ¼

FM N ¼ 740 [ PG mm2 0:9 APhead;min

where PG = 700 N/mm2 from Table 39 of VDI2230. Further examination in R10. R2 Determination of the tightening factor aA The bolt is tightened using a high precision tightening spindle, which has been adjusted by measuring the elongation of the bolt (after pre-calibrating the bolt as the force measuring element). The tightening factor aA = 1.6 according to Table 8of VDI 2230 (for large angles of rotation, fine threads and resilient joints). R3 Determination of the required minimum clamp load, FK erf 1. The requirement for friction grip at the interface (lTr = 0.12) is: FK erf 1 ¼ lF0 ¼ 3:5  103 N Tr 2. To avoid one-sided opening at the rated speed of the engine, FK erf2 is calculated (from Eq. 3.52 of VDI 2230) using the dimensions given in Fig. 13.19.

500

P. R. Wani

FK erf2 ¼

ða  sÞu FA ¼ 8:1  103 N IBT AD þ su

Where AD ¼ 121 mm2 IBT ¼ 2474 mm4 R4 Determination of the load factor, Uen The transmission of bending moments and normal forces at the interface leads to a bolted joint with eccentric load application. Moreover, the load is not introduced under the bolt head and the nut but inside the clamped parts. Since the greater part of the connecting rod joint surrounding the bolt can be considered to be a clamping sleeve, n = 1/3 is estimated in this case. With Aers ¼ 122 mm2 Using Eq. (3.34 of VDI 2230)   dP 1 þ aIsBAersers   Uen ¼ n 2 dS þ dP 1 þ sIBAersers The resilience of the bolt is calculated using Eq. 3.8 of VDI 2230 dS ¼ dK þ d1 þ d2 þ    þ dGM Its components are found as follows 1. Thread elasticity in the nut: dG ¼

0:5d ¼ 0:477  106 mm/N ES A3

Where ES ¼ 205  103 N/mm2 2. Elasticity due to the displacement of the nut: dM ¼

lM ¼ 0:276  106 mm/N ES AN

13

Connecting Rod

501

Where lM ¼ 0:4 d 3. Elasticity of the free and loaded part of the thread: d6 ¼

l6 ¼ 0:392  106 mm/N ES A3

4. Elasticity of the shank: d15 ¼

X

li E A i¼15 S N

¼ 3:064  106 mm/N 5. Elasticity of the head: dK ¼

0:4 d ¼ 0:276  106 mm/N ES AN

We thus have dS ¼ dG þ dM þ d6 þ d15 þ dK ¼ 4:49  106 mm/N Resilience of the clamped parts dP: For the determination of the resilience of the clamped parts, the small eccentricity of the bolt (s = 0.3) is not allowed for. Thus, dP will be determined instead of d*P. The assembly preload which causes embedding acts concentrically. For a cross-section study of the clamped parts of a substitution body Fig. 13.5 of VDI 2230 should be referred. Since the joint has a clearance hole, with dimensions DA and dh, and because dw  DA  dw þ lK We find with Eq. (3.17 of VDI 2230) Aers ¼ 122 mm2 Where DA ¼ 15:5 mm; dw ¼ 13:2 mm; lK ¼ 41:5 mm; dh ¼ 9:25 mm

502

P. R. Wani

With EP ¼ 205  103 N/mm2 We find dP ¼

lK ¼ 1:66  106 mm/N EP Aers

Thus, with IB ers = 2833 mm4 because Ders = 15.5 mm determined from Aers and dh = 9.25 mm we obtain from Eq. (3.43) of VDI 2230 Uen ¼ 0:118 R5Determination of the loss of preload due to FZ embedding From Fig. 50 of VDI 2230 we find the amount of embedding fZ = 6.1  10−3 mm (rounded), for lK/d = 4.6. Thus, the loss of preload due to embedding, FZ ¼ fZ

1 ¼ 0:976  103 N dP þ dS

R6 Determination of the required bolt size The maximum assembly preload is calculated from:  FM max ¼ aA FK erf þ ð1  Uen ÞFA þ FZ However, for the case of a bearing cap bolted joint, FM min must be further increased by the amount FL required for the elastic and plastic deformation which occurs in the shell bearing because of over sizing. To compensate for over-sizing of the shell bearings, an axial load FL of 6.2  103 N per bolt is required. Thus, in this case we have  FM max ¼ aA FK erf þ ð1  Uen ÞFA þ FZ þ FL With FK erf ¼ FK erf 2 ; since FK erf 2 [ FK erf 1 FM max ¼ 29:7  103 N Assuming a coefficient of friction lG = 0.12 (Table 5 of VDI 2230), we obtain from Eq. (3.26 of VDI 2230) an assembly preload of

13

Connecting Rod

503

FM ¼ rM A0 ¼ 41:6  103 N for the bolt M9  1 of strength grade 12.9. Therefore, A0= As= 49.8 mm2 and mRp 0:2 min rM ¼ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi h  i ffi 1þ3

2d2 d0

P pd2

þ 1:155lG

2

N mm2 d0 ¼ ds ¼ 7:963 mm N Rp 0:2 ¼ 1100 mm2 ¼ 835

And with m ¼ 0:9, follows FM [ FM max The dimensions of the bolt are correct. With lK ¼ 0:12, we have a tightening torque of
 Dk m l MA ¼ FM 0:16P þ lG  0:58d2 þ 2 K ¼ 60:3 Nm where Dk m Dw min þ Dha max ¼ ¼ 5:95 2 4 lG ¼ 0:12 R7 The check for lK/d and U can be omitted since these values are already determined exactly. R8 The check that the maximum permissible bolt force is not exceeded Uen FA  0:1 Rp0:2 min A0 440 N\5480 N R9 Determination of the alternating stress endurance of the bolt. Because of the eccentric clamping and loading, the bolt is subject to tensile stresses and to bending stresses.

504

P. R. Wani

From Eq. (3.74 b of VDI 2230), rSA b




1 s lK Es a pd33 Uen FA ¼ 1þ  Uen a lers EP 8I B ers Ad3

It follows from Uen ¼ 0:118;

lK Es pd 4 ¼ 1; FA ¼ 3:7  103 N; ¼ 1; I B ers ¼ IB ers  h lers Ep 64

¼ 2474 mm4 That on the side in tension N mm2 N N ¼ 9:5 þ 46:5 mm2 mm2 N ¼ 56 mm2

rSA b ¼ ð1 þ 4:88Þ  9:49

With strain gauges attached to the bolt shank in the vicinity of the interface, a tensile stress variation of 52 N/mm2 and a compressive stress variation of –32 N/ mm2 relative to the pre-load were measured, after applying an axial force of FA to a joint with minimum bolt pre-load FV erf ¼ FK erf 2 þ FL . From this we obtain a tensile stress component of 10 N/mm2 (9.5 N/mm2 was calculated) and a bending stress component of 42 N/mm2 (46.5 N/mm2 was calculated). Figure 13.20 gives a comparison between the calculated and measured stress distribution in the interface plane of the bolt.

Fig. 13.20 Stress distribution in the interface plane of the bolt

13

Connecting Rod

505

On the tensile side, the bolt alternating stress is rB ¼

rSA b ¼ 28 N=mm2 2

Figure 48 of VDI 2230 gives rA ¼ 54 N=mm2 Thus, we have rB \rA R10 Checking calculation of the surface pressure under the head bearing The nut bearing is not examined. AP nut min [ AP head min; See R1 FM þ Uen FA N ¼ 660 mm2 AP head min 2 p\pG ¼ 700 N/mm

p¼

pG value is taken from Table 9 of VDI 2230. The checking calculation shows that the function of the bolted joint, under strength stipulations is fulfilled. Symbols and notations (VDI 2230) A

Cross sectional area

AD Aers

Interface area minus the area of the hole for the bolt Equivalent cross section area of a hollow cylinder with the same elastic resilience of the clamped parts bearing area of the bolt head or nut Outside diameter of the clamped sleeve- for the interface surfaces of a varying circular form (dA = the diameter of the inner circle) Inner diameter of bearing surface of the nut Inner diameter of the bearing area of the clamped parts under the nut (at the start of the fillet of the clamped parts) Outside diameter of the bearing area of the clamped parts under the nut (at the thread or at the start of the fillet of the outside diameter) Outside diameter of bearing area of the nut (at the start of the fillet of the outside of the nut) minor diameter of the thread in the nut Young’s modulus of the parts clamped Young’s modulus of the bolt Axial force calculated along the bolt axis or the components for a given working load, FB (continued)

AP DA Da Dha DhW DW Df EP Es FA

506

P. R. Wani

(continued) A FB FK

Cross sectional area erf

FM FM max

FM min

FQ FV FV erf

FZ IB ers I B ers IBT Mb P Rp a d da

0.2

dc dh dha dhW dW d0 d2 d3 F2

Working load in a joint in any direction required clamping load for sealing functions, friction grip and prevention of one-sided opening at the interface Initial clamping load (assembly preload); The values in tables of VDI 2230 are calculated with a 90% of the elastic limit using rred Initial clamping load for which the bolt is designed considering lack of precision in the tightening technique the expected bedding in during operation, the minimum required clamping load Minimum initial clamping load established at FM max because of lack of precision in the tightening technique Transverse force normal to bolt axis Preload Minimum preload required to ensure sealing functions, friction grip and one sided opening at the interface by loss of the force at the interface Loss of preload due to bedding in during operation Equivalent area moment of inertia IBers minus area moment of inertia for the bolt hole Area moment of inertia for the interface Bending moment at the bolting point due to the axial loads, FA and FS applied eccentrically thread pitch 0.2% proof stress as per DIN ISO 898 Part 1 Distance at which the load acts from the axis of the surface AB Bolt diameter = outside diameter of the thread (nominal diameter) Inner diameter of the face of the bolt head (at the inlet of the transition radius of the shank) Core diameter of the face of the bolt head Bore diameter of the clamped parts; inner diameter of the equivalent cylinder Inner diameter of the bearing surface of the bolt head (from the start of the fillet the bore) Outer diameter of the bearing surface of the clamped parts with the bolt head (from the start of the fillet of the outer diameter of the clamped parts) Outside diameter of the plane head bearing surface (at the inlet of the transition radius of the head) Diameter at the smallest cross section of the bolt shank Pitch diameter of the bolt thread Minor diameter of bolt thread Plastic deformation by bedding in (continued)

13

Connecting Rod

507

(continued) A

Cross sectional area

l li lt n,  n

Length, in general Length of the bolt Clamping length A factor given by the ratio of the thickness of sections of clamped parts unloaded by the axial load, FA to the clamping length lK Surface pressure maximum permissible pressure under a bolt head Distance between the bolt and the axis of the surface AB Distance between the edge of clamped part (prism) from the axis of the surface AB (in the direction opposite to A-A) Distance between the edge in clamped part (prism) from the axis of the surface As (in the direction opposite to A-A) Tightening factor, FMmax =FMmin Elastic resilience Resilience of engaged thread Resilience of any part, i Resilience of bolt head Resilience of the clamped parts for the concentric bolting and concentric loading Resilience of the clamped parts for eccentric loading Resilience of the clamped parts for eccentric bolting and eccentric loading Resilience of the bolt Coefficient of friction in the thread Coefficient of friction for bolt head A fraction of the yield load to which the bolt is tightened Stress amplitude at the endurance limit Alternating stress acting on the bolt Tensile stress due to FM Equivalent stress, (relative stress) tensile stress (due to bending) in bolt thread caused by the axial load FSA and the bending moment Mb due to eccentric application of load Same as rSAb , but compression stress due to bending Load factor, FSA/FA Load factor for eccentric application of axial load FA Load factor, /e for introduction of load through the clamped parts Load factor for introduction of axial load FA concentrically at a distance = n lK

p pG s u v aA d dG di dK dp dp dp ds lG lK c rA ra rM rred rSA b rSA d / /e /en /n

508

P. R. Wani

References Afzal A, Fatemi A (2004) A comparative study of fatigue behaviour and life predictions of forged steel and PM connecting rods. No. 2004-01-1529. SAE Technical Paper Ashley Steven (1991) Connecting rods that crack by design. Publication, Mechanical Engineering-CIME Bremi P (1971) Calculation of the stresses and most important deformations on a connecting rod big end with the help of a computer. Sulzer Tech Rev, Switzerland 1:59–64 Dipak S, Khan AM, Jaipuria A (2010) Dynamic load analysis and optimization of a fracture-split connecting rod Kubota T, Iwasaki S, Isobe T, Koike T (2004) Development of fracture splitting method for case hardened connecting rods. No. 2004-32-0064. SAE Technical Paper Pravardhan SS (2004) Dynamic load analysis and optimization of connecting rod. PhD diss., The University of Toledo Pravardhan SS, Ali F (2005) Connecting rod optimization for weight and cost reduction. No. 2005-01-0987. SAE Technical Paper VDI 2230: Systematic calculation of heavy duty bolted joints: joints with one cylindrical bolt Wani PR, Dani AD, Reddy PV (2005) Study of connecting rod for high BMEP engines. No. 2005–26-003. SAE Technical Paper Ilia E, Chernenkoff RA (2001) Impact of decarburization on the fatigue life of powder metal forged connecting rods. No. 2001-01-0403. SAE Technical Paper Kuratomi H, Uchino M, Kurebayashi Y, Namiki K, Sugiura S (1990) Development of lightweight connecting rod based on fatigue resistance analysis of microalloyed steel. SAE transactions 487–491 Nakamura S, Mizuno K, Matsubara T, Sato Y (1993) Development of high fatigue strength free machining microalloyed steel for connecting rods. SAE Transactions 858–865 Fatemi A, Zoroufi M, Shenoy P, Afzal A (2005) Comparative durability study of competing manufacturing process technologies. Mechanical, industrial and manufacturing engineering department, The university of toledo, toledo, ohio

Chapter 14

Critical Fasteners, Highly Loaded Bolted Joints Prakash R. Wani

Abstract Due to clamping of the parts, the compressive stress is induced at the joint surface. The compressive stress keeps the joint away from opening out. The reliability of the joint depends on what is the reliability of the fasteners geometry, correct material selection and heat treatment, reliability of load analysis and reliability in achieving expected joint clamping load during assembly of the relevant parts. In this chapter, the preload on critical fasteners is explained using the force diagram of a bolted joint and the settling force, pre-loading on the bolts and tightening by torque-controlled tightening, angle-controlled tightening and other methods is described in detail. Important practical aspects like thread engagement length, limiting surface pressure and fatigue loading are dealt with at length. The influence of temperature especially in cylinder head bolts is important to consider while designing. The high temperature fasteners form a special category in design and development. Various methods to improve the fatigue strength of the joints are given. The time-honoured design process as per VDI 2230 is explained and examples are provided in the appendix. Also, the typical failure modes for which the design work is carried out are provided.

14.1

Introduction

We come across the bolted joints in various forms like single bolted joints or multi bolted joints. The joint may clamp the cylindrical bodies or circular flanges or rectangular shaped parts. Due to clamping of the parts, the compressive stress is induced at the joint surface. The compressive stress keeps the joint away from opening out. The reliability of the joint depends on what is the reliability of the fasteners geometry, correct material selection and heat treatment, reliability of load analysis and reliability in achieving expected joint clamping load during assembly of the relevant parts (VDI 2230; Radzimovsky 1952). P. R. Wani (&) Government College of Engineering, Karad, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_14

509

510

P. R. Wani

Fig. 14.1 General sketch of a bolted joint

The bolts can be classified depending on the shank viz. Straight shank bolts and reduced shank- or necked down- bolts. Bolt may be with coarse pitch or fine pitch, high strength or normal, metric or imperial dimensions, with vee threads; square threads or with other thread forms. Figure 14.1 shows the general sketch of a bolted joint. Various international standards are available to know the preferred sizes for the bolts. These standards give the information on the strength of the nuts, bolts, and major, minor and pitch diameter data, useful cross-sectional area carrying the load, bolt head height, across flat and across corner dimensions etc.

14.2

Basic Working Principle and Joint Diagram

If we treat a bolt as a beam with circular cross section as indicated in Fig. 14.2, loaded with an axial load F, Force F ¼ Crosssection area A Change in length Dl ¼ Strain ¼ Original length l Stress F l ¼ E¼ Strain A Dl Stress ¼

14

Critical Fasteners, Highly Loaded Bolted Joints

511

Fig. 14.2 Rod elongation due to force F

Now, Force F ¼ Deflection Dl EA ) stiffness ¼ l

Stiffness ¼

In general, the cross section of clamped parts is much more than that of the bolt and hence the plot force-deflection diagram to get the stiffness is as given in Fig. 14.3. Consider the bolted joint when the bolt is tightened with the help of a nut. In this tightening process (preloading of the bolt), the external load is still not in the picture. During preloading, the bolt is stretched (positive deformation) and the same force is acting on the clamped parts and these parts are getting compressed (negative deformation). If we represent this on force-deflection diagram (Fig. 14.4) with different lines like Line 1—for positive force, positive deflection Line 2—mirror image of line 1 about y axis giving positive force and negative deflection Line 3—mirror image of line 2 about x axis giving negative force and negative deflection Line 4—parallel to line 2 Line 2 represents the clamped part if the compressive force were applied to get the negative deflection. Fig. 14.3 Stiffness of clamped pars > stiffness of bolt

512

P. R. Wani

Fig. 14.4 Nut and bolt stiffness

Line 2 is for representation only. At preloading condition, the bolt is getting stretched. The same force results in compression of clamped parts that is shown by line 2. Line 4 is parallel to line 2 such that at preload, F the bolt elongation is db and the clamped parts compression is dc. Now consider the external force F1 that acts on the assembly. It can be noted that this force F1 will be shared by the bolt and clamped parts. Thus, considering the bolt and the clamped parts as two springs in parallel, deflection is the same for both. db ¼ dcl Fb Fcl ¼ kb kcl Also, Fb þ Fcl ¼ F ¼ external force Fb ¼ force shared by the bolt Fcl ¼ force shared by the clamped parts Solving the above equations, we get, Fb ¼

kb F kb þ kcl

It shows that if the stiffness of the bolt is less than that of the clamped parts, a smaller fraction of the external load is shared by the bolt. This load sharing can be shown in the joint diagram, Fig. 14.5. We can see that Fpr is the preload in the joint and this force is creating compression dc. When the external force acts on the joint the bolt gets further stretched and the compressive force on the clamped parts reduces b the amount Fj. The net force on the clamped parts,

14

Critical Fasteners, Highly Loaded Bolted Joints

513

Fig. 14.5 Bolt and clamped parts stiffness’s in parallel

Fj ¼ Fpr  Fcl This force keeps the joint away from separation. We can understand from the joint diagram (Fig. 14.6) that Fj is the joint force which is responsible to keep the joint away from failure. To get more and more of Fj increase of preload is helpful for a joint. This increase results in shifting of the point A towards right hand side. However, one cannot neglect that this preload increases bolt stress even before any application of the external load that the joint must withstand. The external load shared by the bolt and the preload during the assembly combined place the limit on the force on the bolt. Thus, to have the effect joint capacity we should have proper design to allow maximum external load, maximum preload still below the allowable load (coming from the bolt strength point of view). The force Fj remaining in the joint also depends on other factor such as settling force. Settling force is caused by the permanent deformation of the bolt or the clamped parts.

Fig. 14.6 Bolt joint diagram

514

P. R. Wani

Fig. 14.7 Settling force, Fz

14.3

Settling Force Fz

If the bolt is deformed plastically by a tensile force F, its deformation line curves up to the point K as in Fig. 14.7. The deformation line for unloading follows the straight-line KB and intersects the base line at B, With OB as the plastic deformation. Further when the joint is subjected to an external load F the joint diagram DBCK applies, in which the preload has decreased by Fz compared with the initial preload at the assembly stage. An analogous situation arises if the clamped parts are plastically deformed by a compressive force. The loss of preload due to plastic deformation is given by Fz ¼

deformation OB reslience of bolt þ resilience of clamped parts

where resilience is the reciprocal of stiffness, i.e. deflection per unit force. Fz can be treated as the loss of preload due to embedding of parts. In general, 75–80% of the load corresponding to the proof load of bolt is applied as the preload on the joint, where the proof load is approximately the load in the bolt corresponding to the yield stress.

14.4

Pre-loading on the Bolts and Tightening Methods

The essential preload induced at the joint faces is the most important factor to arrive at the dimensions of the bolt. The friction between the threads and mating parts plays a major role while tightening the bolt. As the determination of coefficient of friction is not so easy to estimate, the margin of safety should be kept to be on the

14

Critical Fasteners, Highly Loaded Bolted Joints

515

Table 14.1 Guidance values of tightening factor, aa Guidance value, aa

Tightening technique

1.0 1.2–1.6 1.4–1.6 1.7–2.5 2.5–4.0

Yield controlled tightening, angle-controlled tightening Hydraulic tightening Nut runner with high precision Nut runner with spindle Impact wrench

conservative side. Depending on the method adopted for tightening, the margin factor i.e. tightening factor aa should be selected as per guidelines given in Table 14.1.

14.5

Torque Controlled Tightening

Torque wrenches are used for tightening the bolts at many places. The wrenches are indicating or controlling type. The preload in the bolt depends on the tightening torque and the friction at various joint surfaces. 

d Deff T ¼ F tanðu þ l0th Þ þ l 2 2 1



where T ¼ applied torque to bolt F ¼ preload in the bolt d2 ¼ pitch diameter of the bolt l tan u ¼ th pd2 lth 0 lth ¼ cosða=2Þ lth ¼ coefficient of friction in the thread a ¼ included angle of the bolt thread or thread flank angle Deff ¼ effective friction diameter in bolt head or nut bearing surface l1 ¼ coefficient of friction for bolt head bearing surface If a = thread flank angle = 60°, the equation can be written as

516

P. R. Wani



Deff l T ¼ F 0:16 pitch þ 0:58 d2 lth þ 2 1



The preload is function of the torque applied, thread dimensions and the friction. As the torque increases, the preload force increases. The preload force and the torque induces the stress in the bolt before actual load is applied. This stress is generally limited to about 80 to 90% of the yield stress so that the bolt can withstand further load during operation.

14.6

Angle Controlled Tightening and Other Methods

In angle controlled tightening method, the angle through which the nut has rotated with respect to the bolt after establishing the initial contact is measured. The clamped parts are compressed, and the bolt is elongated during the tightening. The axial elongation of the bolt will be equal to the pitch value if the nut is rotated by 360° or in the proportion to the angle of nut rotation. With this method the effect of variation in friction has less effect on the prediction of the clamping load. If the tightening is done in the plastic region, the experimental values for the stretch should be used in the calculations of the clamping loads. Yield-controlled tightening is another method where the preload can be increased to a better value due to identification of the yield point of the fastener during tightening. It is done by measuring the rate of elongation while tightening. When the yield point is reached, the rate of elongation increases. The work of measurement and monitoring is done with help of sophisticated electronic instruments. With predetermined shut off value of this rate, the torque is controlled. With yield controlled tightening method, the effect of friction is minimised. If the plastic deformation of the bolt does not take place the bolts can be reused many times. However, during tightening of the bolts if the plastic deformation takes, the permanent elongation of the bolt results into change of dimensions and mechanical properties of the bolt. Due to the change of dimensions, the reusability of the bolt is affected. After a certain permanent elongation of the bolt it must be replaced by a proper one.

14.7

Some Practical Aspects

While designing the joint, the sufficient allowances must be kept permitting different variations like material strength, surface properties of the components, the shape of bolts and nuts etc.

14

Critical Fasteners, Highly Loaded Bolted Joints

517

The bolt should be designed in such a way that combined stress during preloading and working should be less than the permissible yield stress of the material. With yield-controlled tightening, slightly more load can be taken by the bolt if it is having good ductility and toughness. The scatter in the tightening process should be relatively low.

14.7.1 Thread Engagement Length In working condition, the bolt is subjected to tensile force. Due to nut and bolt thread engagement the load is distributed over the threads. If this engagement length is not sufficient it may lead to the shear of the weaker threads out of the two. In normal cases where bolt and nut arrangement is used, bolt thread length is larger than the nut height. So it has to be ensured that the nut height is adequate to provide the required tensile strength of the joint. The minimum thread engagement length should be decided considering the mating component’s strength and stiffness, the size of bolt, fine or coarse pitch and nature of surfaces contributing to friction. It should also be noted that excessive thread engagement does not add to the proportionate strength to the joint. The joints where soft material like Aluminium is used for components, the thread engagement length should be comparatively more than that for the materials like alloy steels.

14.7.2 Limiting Surface Pressure The tightening of the bolt results into a compressive stress being developed at the joint faces as well as at the bolt head and nut mating face areas. Though it is necessary to have sufficient compressive stresses at these surfaces to maintain the contact, it should not exceed the allowable compressive bearing stress at any of the surfaces in contact. If the surface pressure at the bolt head or nut bearing is higher than the compressive yield point of the clamped materials, the creep process (time dependent plastic flow) occurs. This will reduce the preload on the joint and it is not desirable. The excess of the surface pressure should be avoided at the time of preload and at the time of service load application, i.e. preload or the maximum service load contact are a near the nut or bolt head surface pressure\compressive yield stress of the clamped parts

surface pressure ¼

Guiding limiting surface pressure values are given in Table 14.2 for different joint materials.

518

P. R. Wani

Table 14.2 Limiting surface pressure guiding values Material

Tensile strength, N/mm2

Limiting surface pressure, N/mm2

St37 St50 C45 42CrMo4 GG15 GG25 GG35 GG40

370 500 800 1000 150 250 350 400

260 420 700 850 600 800 900 1100

To reduce the compressive stress sometimes the washers are used. Generally these washers are machined, hardened and ground so that they are stronger than the materials of the clamped parts.

14.7.3 Fatigue Loading If the external load on the joint is cyclically varying, the bolts also experience the fatigue load, Fig. 14.8. For fatigue testing of the fasteners, the S-N curve is obtained for the actual bolts rather than the specimen. The uniaxial push pull loading is normally applied through the hydraulic servomotors at the predetermined frequency. With such tests the effect of manufacturing process can be ensured, and it helps in correct determination of the bolt endurance strength.

Fig. 14.8 Fatigue load on the bolt

14

Critical Fasteners, Highly Loaded Bolted Joints

14.8

519

High Temperature Fasteners

As the requirement of high power with minimal volume capacity of engine is emphasized due to competition, turbo charging of the engine is common now a day. Due to use of turbochargers the working temperature of the parts of engines like exhaust manifold bolts and turbocharger clamping fasteners is at elevated level. Normal steel fasteners are operating up to the temperature of 370 °C without degradation in properties. The temperature of turbo charger fasteners may reach around 650°. In such cases the fasteners are made from suitable alloy materials. These alloys are precipitation hardenable. ASTM –A-453 standard “standard specification for high temperature bolting with expansion coefficient comparable to Austenitic stainless steel” can be referred while selecting the material. Even after the operation at elevated temperature the materials do not lose the properties after returning to room temperature. They have good corrosion resistance. Typically, the composition of high temperature bolting steel is as follows (Mcmann et al. 2011). Ni Cr Ti Mg Mo Si Al V C

18–20% 13.5–16% 2.2–2.8% 2% max 1% 1% 0.5% 0.5% 0.08%

High percentage of Nickel tends to increase the cost of the fasteners.

14.9

Improvement in the Fatigue Strength of the Joints

If possible, the following solutions are tried: Increase preload on the bolt that will increase the residual clamping forces. Thread rolling after the heat treatment can improves the endurance limit of the threaded joints. The endurance strength of the bolts reduces as the thread nominal diameter increases. For the same configuration if the thread diameter can be reduced, it results into increase of the fatigue strength. Limited improvement is possible if the friction between the threaded parts is minimised. The notch effect is dominating in case of fine pitch bolts having high strength material. In such cases, the increase in pitch of the bolt can result into improvement of the endurance strength.

520

P. R. Wani

If the modulus of elasticity of nut material is reduced, then the bending resilience of the thread teeth of the nut is increased. It results into more even distribution of the load in the bolt. Bolt with high elastic resilience and clamped parts with low elastic resilience result into low portion of the load to be carried by the bolt. This can relatively increase the joint capacity.

14.10

Design Process as Per VDI 2230 (VDI 2230)

As per VDI 2230 following steps are followed while designing the critical fasteners. With the help of charts given by, VDI 2230 the initial estimation of the bolt size and grade can be done. The ratio (lk/d) and limiting surface pressure values are also considered for this estimation. In the next step the tightening factor alpha is determined that is function of lubrication and surface conditions. The calculation of minimum clamping force required to keep the joint intact is done. In this step the basis is to avoid the one-sided opening of the joint. In the next step the load factor is determined. It is dependent on the stiffness of the various parts of the joint under consideration. The loss of preload due to embedding or settling force is the determined. Now the required bolt size is calculated and the tightening torque with the specific lubricating conditions is determined. Depending on the change required to make the joint safe the number of iterations may be performed. The maximum force on the bolt due to loads in operation should be limited so that the resultant load, i.e. pre-load and the load on bolt during operation combined do not exceed the bolt strength. If the loads are of varying cyclically, the bolt is also subjected to the periodically varying stress. In such cases the alternating stress on the bolt must be less than the permissible fatigue strength for chosen bolt. The influence of such factors as the kind of thread, form of nut, material, production method, and size of bolts must be investigated. Stiffness of connected parts in bolted assemblies also has not been sufficiently investigated and must be studied both experimentally and theoretically.

14.11

Failure Modes of Threaded Fasteners

The cost of the specific fastener which fails may not be very high but the total loss to the system may be very high in many cases. For this reason, the critical fasteners are given a lot of importance right from design stage. The joint failure really starts when the contact faces start opening out. Figure 14.9 shows the failure of the joint in which the contact of the mating faces is

14

Critical Fasteners, Highly Loaded Bolted Joints

521

Fig. 14.9 Loss of contact of mating faces of the joint

lost. One end of the joint is open which will lead to the frequent opening and closing of the contact between mating faces during engine operation. This will finally result into failure of the bolt. In reduced shank bolt in which normally the shank diameter is less than the minor diameter of the thread. In reduced shank bolts, due to more resilience it carries less fraction of the external load. In such cases, the bolt should not be over tightened. If the bolt is excessively tightened it leads to failure in the shank zone. Sometimes the failure starts at the first thread where along with the load; the stress concentration factor at the root of thread amplifies the stresses. The bolt failure can also occur due to thread stripping. For low tensile materials like aluminium tapping is done. If the shear strength of the bolt or of the material where tapping is done is less, the thread can strip from its place leading to failure. Thread stripping is a gradual process where in the radial deformation of the thread takes place due to varying shear strength of nut and bolt materials. The force acting on the nut produces bending effect on the thread if relative radial deformation takes place between the two materials. The situation becomes worse if the tapping done has produced tapered hole. This taper reduces the effective shear area of the threads. In case the thread engagement length is not sufficient, the bolt failure may occur in presence of above defect.

522

P. R. Wani

Appendix: Some Critical Fasteners in Engines Connecting rod bolts For critical fasteners sometimes, it may be necessary to analyse the stresses with great details. For finite element analysis the connecting rod assembly is considered. The bolt threads and nut threads are modelled considering the diametral interferences. The inertia and gas firing loads are simulated on connecting rod and effect at the threads is studied. Calculation example is given in the chapter of connecting rod. Similar treatment can be adopted for main bearing cap bolts. Main bearing cap bolts: sample example for a multi cylinder inline engine p Peak firing force ¼ ð96 mmÞ2 ð85 bar Þ 4 ¼ 55987N 1 Force on each cap ¼  55987 2 ¼ 27993 N Force on each bolt with cover factor of 2 ¼ 27993 N The selected bolt is M14  2, 10.9 quality. With 140 Nm torque, the preload will be 69876 N at 80% of yield. Since 69876 N > 27993 N i.e., the bolts selected should be able to take the abuse of fatigue loading. Main bearing load maximum ¼ 28757 Nðfor the third bearing at 1000 rpmÞ Maximum load on each bolt ¼ 14379 N Interference fit force due to bearing: Assuming 25 kg/mm2 stress and 2  23 mm size as we have for connecting rod also, the force on each bearing cap bolt ¼ 11280 N Total force on bearing ¼ 14379 þ 11280 N ¼ 25659N Factor of safety ¼ ¼ 2:72

69876 25659

14

Critical Fasteners, Highly Loaded Bolted Joints

523

Cylinder head bolts Number of bolts; n ¼ 6 Pressurized diameter ¼ 104 mm Minimum allowed cover factor ¼ 3 p pmax Maximum gas load per bolt; FGbolt ¼ d 2 4 n p 8:5 MPa ¼ ð150 mmÞ2  4 6 ¼ 12034 N Minimum preload required per bolt ¼ ¼ cover factor  Maximum gas load ¼ 3  12034 ¼ 36102 N min:preload required per bolt ¼ 36102 N Six bolts: M10  1.25, 10.9, 125 long on 137 pitch circle diameter, With 80% yield; torque ¼ 6:26 kgm with oil Preload ¼ 37500 N [ 36102 N Hence safe.

References Mcmann F et al. (2011) Carpenter 286-LNI alloy: a lower cost option for high temperature auto and truck fasteners. Carp Prod Radzimovsky EL (1952) Bolt design for repeated loading. Mach Des VDI 2230: Systematic calculation of heavy duty bolted joints: joints with one cylindrical bolt

Chapter 15

Crankshaft Prakash R. Wani

Abstract A crankshaft is used to convert reciprocating motion of the piston into rotary motion. The crankshaft in an engine is probably the most complex of all the shafts used in any machinery, and, as the name implies, it is far from being anywhere near a straight shaft. With the help of examples, crankshafts are classified depending on the type of supports as overhung and centre crankshaft or based on the number of throws as single throw or multi-throw shafts. The procedure for design the crankshafts is explained in detail using calculations of the crankshaft strength and stress. The factors affecting the fatigue strength are deliberated. Plots of oil film thickness explain the wear pattern. Inherently, single-, two-, three- and four- cylinder in line engines are not fully balanced for inertia forces or couples; if the cost permits, counter rotating balance shafts are designed to neutralize these forces. Otherwise, these forces are left unbalanced; they cause rigid body vibration of the engine and also are transmitted to the vehicle and to other parts. Vee engine shafts are treated slightly differently from the inline engine shafts. The designed shaft when made does not have mass distribution as per the design because of manufacturing tolerances to forging and machining. This imbalance is removed at a balancing machine within the limits specified in standards. The inertias of individual throws, piston, and connecting rod as well as the flywheel with the stiffness of the shaft result in multiple natural frequencies in the rotational direction. In case of long shafts as in the case of a six-cylinder engine, the torsional vibration can have a resonance frequency in the operating range of speeds and can induce fatigue usually starting from the oil hole in crank pin. Therefore, calculations of moments of inertia, equivalent stiffness and the natural frequencies as well as the amplitude of vibration are important. If the amplitude is sufficiently high, the torsional stress can exceed the fatigue strength limit leading to failure. To avoid the natural frequency in the speed range of operation an oscillator in the form of a torsional vibration damper is added. The new system not only shifts the frequencies but al-so reduces the dynamic magnifier to reduce the torsional stresses. Various parameters like characteristic frequency at fixed points, damping ratio, unit tuning ratio, optimum P. R. Wani (&) Government College of Engineering, Karad, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_15

525

526

P. R. Wani

tuning ratio are introduced and with the aid of a characteristic help-graph, the tuning ratio of a rubber or spring damper can be selected. While rubber is relatively less expensive, heat dissipation in the damper must be carefully predicted to estimate the temperature of the rubber in the damper as the properties of rubber are highly dependent on temperature. The type of rubber is properly selected and manufactured with great care to avoid aging of rubber at the operating temperature. When the heat may not be easily dissipated fluid dampers are useful. Such dampers without a spring only dampen the vibration to save the shaft but do not play any major role in shifting the natural frequency of the shaft system. The engine load can be quickly simulated at specific rigs to estimate the bending fatigue or torsional fatigue. Finally, the importance of the design of bolts and the applied tightening torque are important to hold the flywheel, the crank pulley, connecting rods, bearing caps together, cannot be understated.

15.1

Introduction

A crankshaft converts reciprocating motion of the piston into rotary motion. The crankshaft mainly consists of crankpin, webs, journals and end parts. The crankpin is connected to connecting rod whereas web connects the crankpin and a journal. The power is transmitted from piston through connecting to crankshaft at the time of fuel combustion and from crankshaft to piston at other instances. The crankshaft is an important part carrying the power to the driven machinery. It has to be carefully designed to withstand the fluctuating bending and shear loads without change of shape and size. As the crankshaft length mainly decides the engine length and crank radius decides the engine height, optimization of these parameters to get minimal weight is focused by the designers. Tribological aspects must be satisfied to take care of the lubrication at main journal and crankpin bearings.

15.2

Types of Crankshaft

The crankshaft can be of different construction depending on the number of cylinders i.e. single throw crankshaft or multi throw crankshaft, for inline engine or Vee engine. Built in crankshafts are built by joining the separate webs and crankpins. Mostly one piece crankshafts are used for heavy duty engines. Multi cylinder crankshafts with one piece type generally require the bearings to be split into two to facilitate the assembly and dismantling. For heavy loaded crankshafts there isa single throw between two main journals. For lightly loaded crankshafts however there are two or more throws between main journals.

15

Crankshaft

527

Fig. 15.1 a. A five throw crank shaft for 10 cylinder V-engine. b. A five throw crank shaft for 10-cylinder V-engine- simplified model

Figures 15.1a, b show a five throw crank shaft for 10-cylinder V-engine Oil Holes Drilled in Crankshafts For lubrication of journals and pins, the oil is provided under pressure. The oil has to reach the journal. This is done by providing the suitable gallery the crankcase or bearing housing. The oil holes are to be drilled through journal and webs to supply the oil to crankpin (Fig. 15.2). In some cases, the holes opening into webs are plugged.

Fig. 15.2 Oil holes drilled in the crankshaft

528

P. R. Wani

At the ends there may be a provision to mount the flywheel or damper and or pulley. This arrangement makes it possible to drive water pump, fan or alternator depending on the application and construction.

15.3

Crank Shaft Strength Calculations

For the satisfactory working of a crankshaft the stresses should be limited to the permissible values and should have sufficient margin of safety. The bending stresses induced due to firing loads, inertia loads due to reciprocating masses, shear stresses due to torque transmission need to be considered. Due to fluctuating torque torsional vibrations need to be kept within the permissible limits. The centrifugal forces due to rotation of the eccentric masses of crankshaft system should be considered those create the centrifugal forces. The bearing stresses induced on crankpins and main journal bearings should be well below the allowable bearing stresses for the chosen materials. As the loads are varying during the working cycle of the engine the stresses on the crankshaft are also cyclically varying. Hence the design should be done based on fatigue strength in bending and torsion. The crank webs should be joined to the crank pin and main journals with generous fillet radii to minimize the stress concentration factor. To account for the variations in production, sufficient margin should be kept with statistical consideration. Major Steps during Design • Starting from the piston loads and further resolution of loads along and across the connecting rod axis, the loads on crankpin should be determined. • Considering the benchmark initial dimensional model of the crankshaft should be generated. • Assuming the crankshaft as a beam with supports at bearings the bearing reactions can be computed for different crank angles for total working cycle. • The bending moments and the stresses should be computed for various crank positions. The stress concentration factor in bending and torsion for crank pins and journals should be estimated. This factor is dependent on crankshaft dimensions such as fillet radii, journal and pin diameters and the width and thickness of the crank web. Also the oil holes increase the stress concentration factor. • Assuming the material and its mechanical properties, equivalent stresses and factor of safety should be determined. The dynamic analysis should be performed to understand the behaviour of the crankshaft. • If necessary the iterations should be done to arrive at the required factor of safety. • These calculations should encompass all worst possible working conditions of the engine.

15

Crankshaft

529

Fig. 15.3 Finite element model of crankshaft throw

• For the single cylinder engines the crankshaft can be assumed as a simply supported beam. The additional loads due to overhang can be superimposed. For multi cylinder engine crankshaft, the continuous beam is assumed to account the effect of bending moments. The crankpin and journal diameters should have sufficient strength to withstand the bending, shear, vibrations (amplitude and stress) along with the provision for bearing area. Calculations of crankshaft Analytical strength calculations for the crank web can be performed with help of various international standards such as Det Norske Veritas (2002), Germanischer Lloyds (1992) etc. Nowadays, commercial softwares are available with help of which detailed analysis is can be done. Figure 15.3 shows the finite element model of one half of a crank throw. Figure 15.4 shows the stress distribution in the throw and typically at the fillet radius. Dynamic stress calculations A combination of nonlinear Multi-body System Simulation (MSS) with linear Finite Element Analysis (FEA) is done to get good reliability of results with reduced time requirement for analysis (Robert et al. 2000). The various effects which can be included like non-linear analysis, gyroscopic effects and interaction due to deflection of engine block structure. Figure 15.5 shows the model of multi cylinder crank shaft system used by Loibnegger and Thomas (2001) with customized software. Various parts like piston, connecting rod crankshaft bearings etc. are shown schematically and their data is stored in separate files. Once the stresses are evaluated at different locations and at various crank positions over the cycle, the stress history is used to calculate the safety margin against fatigue. The fatigue safety margin figures for a particular node at the fillet are shown in Fig. 15.6. Modified Goodman diagram can also be used for manual calculations where the maximum and minimum stress levels in the working cycle are found (Loibnegger and Thomas 2001).

530

P. R. Wani

Fig. 15.4 a. Finite element model of crankshaft showing stress distribution. b. Finite element model of crankshaft showing stress distribution

Parameters considered in the Model of recent customized software Concluding the features of model and excitation, the following effects are considered in the crankshaft simulation: • The 3D crankshaft implicitly considers the coupled axial, bending and torsional vibration. • Gas and mass forces excite the vibration of the structure. • Due to rotation, secondary inertia forces and moments are taken into account.

15

Crankshaft

531

Fig. 15.5 The multibody model for a typical crank shaft system (Loibnegger and Thomas 2001)

• The actual stiffness of the oil film is considered and is affected by the bearing geometry, the oil viscosity and the bearing clearance. • Additionally, the EHD simulation in radial bearings considers the local pressure build up in the oil film being influenced by the elastic clearance distribution and the oil flow in the bearings. Thus, misalignment and edge loading in the bearings is considered. Furthermore, the assessment of engine friction and oil flow is supported. • Damping effects are caused in the bearings, in the damper and in the shaft structure. • The dynamic interactions of flywheel and shaft are taken into account.

15.4

Tribology Aspect for Pin and Main Bearings

The bearings are taking the direct loads acting from the rotating components and then transferring the same to the static housing. To get the better life of the bearings, the calculated bearing pressure must be less than the allowable pressure for the chosen bearing material. Table 15.1 shows the maximum allowable bearing pressures for commonly used bearing materials.For more detailed calculations and

532

P. R. Wani

Fig. 15.6 Fatigue Smith diagram for a node in the crankshaft model (Loibnegger and Thomas 2001)

saving of computation time, the customised softwares are available to compute the maximum oil film pressure and minimum oil film thickness for the given geometry of bearings, cylinder pressure– crank angle diagram, engine speed and lubricating oil properties. The minimum oil film thickness in the bearings and maximum oil film pressure are calculated these days with the help of software. Based on these predictions and numerous measurements and the mean allowable pressures the bearings are finalized. Journal and bearing dimensions are thus decided at the same time. Crankshaft wear and undersizing of journals During the engine operation, the force is acting on the top side of the crankpin and at the bottom side of the main journals. This force is of varying nature and is acting in every working cycle. This leads to wear of the crankpin and the journals. The wear pattern is such that the journals lose their cylindrical form. Excessive wear should not be permitted as it may cause adverse effect on engine performance or sometimes breaking of the material of the bearing layer. As the crankshaft is an expensive component, instead of replacing repair is carried out in case of wear beyond permissible limit. Before repair, it is necessary to check the crankshaft

15

Crankshaft

533

Table 15.1 Design data for bearing (Mahadevan and Balaveera Reddy 1989)

1

2

3

Machinery

Bearing

Maximum p** MN/m2

Suitable Z**, Ns/m2

Minimum (ZN/p)*

c'/r

l/d

1 Automobile and aircraft engines

2 Main

3 5. 5–12. 0

4 0. 007–0. 008

5 2. 18

6 –

Crank pin Wrist pin Main

10. 0–24. 0

1. 45

–

16. 0–34. 5

1. 16

–

2. 9

0. 001 900 mg/ kWh Reagent consumption deviation ±50%

(continued)

22

Design of Electronic Control for Diesel Engines

825

(continued) Bharat stage

VI-2

SCR OBD requirements monitoring failures Reagent delivery Reagent availability Proper consumption Regent quality SCR conversion efficiency Electrical component monitoring failures

OBD limits (mg/kWh)

NOx control requirements

NOx control limits

NOx > 1200

Activation of driver warning system/ inducement system in case of Poor reagent quality Reagent consumption Tampering Low reagent level

Reagent quality limit: NOx > 460 mg/ kWh Reagent consumption deviation ±50%

Bharat stage

PM monitoring—OBD threshold limits (mg/kWh)

DPF monitoring

VI-I

25

VI-II

25

Performance monitoring Delta Pressure > = 60% between deteriorated and a new DPF Emission threshold limit (25 mg/kWh)

In Use Performance Ratio (IUPR) (AIS 137, part 4, ch. 8) The In-use performance Ratio (IUPR) for specific monitor is calculated as mentioned below IUPRm ¼ Numerator=Denominator where ‘Numerator’ is incremented as a counter value whenever the number of times the specific monitoring conditions are satisfied in the vehicle running condition. ‘Denominator’ is incremented as a counter value whenever specific monitor related vehicle operating condition which is satisfied (refer example below). Group Monitor The In-Use performance ratio (IUPRg) for a group monitors are calculated based on the average value of various specific monitors IUPRg ¼ Numeratorg =Denominatorg where ‘Numeratorg’ is an incremented counter value which has lowest in use performance of the specific monitoring in the group of monitors.

826

M. Leelakumar

‘Denominatorg’ is also an incremented counter value of the specific monitor which has lowest in use performance ratio within the group of monitors. The vehicle operating condition where this monitoring conditions are satisfied for denominator, are: – – – – – –

Denominator is incremented within 10 s, if defined criteria is fulfilled. i.e. 600 s since engine start 55 °C < Ambient < −7 °C Elevation =1150 rpm for minimum 300 s Accelerator Pedal not pressed and vehicle Speed < 1.6 kmph/Engine Speed is less than idling + 200 rpm.

22.10

ECU Development Process, V-Cycle

The world has migrated from the traditional legacy-based hand-written C code to model based approach for agile process development in embedded domain. The project management is handled through the various software tools for e.g. SVN/ IDoor/PTC Integrity. This approach enables the original equipment manufacturers (OEM) to reduce the development time, cost and improve the quality of the software code. The typical V cycle development process is shown in Fig. 22.29 and the details are explained below.

Fig. 22.29 ECU V cycle process for embedded development (Jaikamal and Zurawka 2010)

22

Design of Electronic Control for Diesel Engines

827

System Requirement: This covers the details of the complete interface of the components involved in the entire system. It provides a skeletal view for the complete system along with the target application. The components of the system are elaborated below. Functional Requirement: This provides the information about the function with details of input and output where the model-based approach or tables, curves and empirical calculations are defined. This also gives the brief objective of the function in the software development. Detailed Requirement Engineering (Software Development Phase): Algorithms and software functions using MATLAB and Simulink are developed along with software interfaces. The base layer for code generation uses Simulink coder or target link coder to C-code, specific to the rapid prototyping. Unit Testing and Hardware-In-Loop Testing (Software Validation Phase) Software-In-Loop (SIL) is an alternative method for validation of the embedded software in day-to-day development. This requires a power computing PC for running the plant model with C-code, generated through auto-coding for verification and iteration of the functions before carrying out the validation in the real application. The legacy codes are more commonly tested the with the plant model before validation in the vehicle environment. This approach does not require the physical hardware and can reduce the time and costs of the new software development considerably. Also, the reuse is very attractive and improves the quality and reliability of the codes before release to vehicle environment. The software can be easily extrapolated for various other functions. SIL is carried out separately by means of unit testing and debugging for verification of the C-code. Hardware-In-Loop (HIL) is the process of validating the ECU software along with the ECU and vehicle. The software is validated for intended function along with HIL system. HIL system does not require an engine and vehicle for testing it with similar application. This powerful method validates the software repeatedly, automatically and over a wide range of operating conditions without carrying the risks of the vehicle environment. Verification on Engine Test Bed Finally, the software validation is carried out in the real environment such as an engine test bench, after verification and validation on the unit testing and hardware-in-loop environment for checking the engine performance and emission at the engine dynamometer. Validation on Vehicle for Series Production Afterwards, this software is migrated for higher level of validation in terms of system verification and reliability, in roadworthy testing conditions, before moving to series production of a larger volume of vehicles.

828

M. Leelakumar

Fig. 22.30 Rapid prototype tool chain for ECU software auto coding

22.10.1

Advantages of Rapid Prototyping

To gain confidence in the development strategy, use of rapid prototyping is explored through the pilot project approach. The main advantage for this activity was set by an initial effort to define the tool chain for the development which accelerates the development of control algorithms for engine initially, and vehicle subsequently. This is a very vital approach followed to verify the software and achieve the performance of engine or vehicle before moving towards testing the intended product. Previously, the hand-written code was used in control development, which requires high development time and intense verification and validation process. With the model-based rapid prototype approaches, the engineers are able to understand the implication of the control function on systems before moving to the real-time environment. The auto-code generated software is loaded in the ECU for checking the performance of the vehicles, Fig. 22.30. There are various tool chains available for this purpose: TargetLink or MATLAB based auto code generation approach for the rapid prototype hardware from DSpace, ETAS, Woodward, NI LabVIEW, PI-Innovo etc.

22.11

Production Code Generation

The production code generation is the next step in rapid prototype development. This is associated with efficient code generation and burns the controller directly for series production hardware. This requires changes in the base software layer along with low level driver interfaces. The driver development depends on the controller and all the tool chains for the development, considering the hardware. Moreover, the base-layer software developed in hand-written C-code requires the wrapper for interfacing with the application software which is developed by model-based approach. The process of the software development is similar to the approach followed in the rapid prototyping method, except for the specific association with the hardware processor.

22

Design of Electronic Control for Diesel Engines

829

To avoid this wrapper, the interfacing is defined as blocks in the Simulink graphical development environment for all the base software requirement. Thus, the complete software can be developed using the model-based approaches.

22.12

Closed Loop PID Control Mechanism

Conventional mechanical governors are found to be very inefficient in precisely controlling the speed. Further, a turbocharged engine has several components and it increases the complexity for working with conventional mechanical governing. The conventional governor property changes over time due to wear and tear of the components. Due to this, it may not be able to provide optimum control in various operating conditions such as ambient temperature and pressure and so on. Therefore, the electronic governor is now widely used with the PID controller concept. The PID control is a closed loop control mechanism which has many advantages in terms of meeting the response with high precision. This provides additional features like automatic tuning, gain scheduling, and continuous adaptation. The closed loop controller (PID) developed is in series with plant model for altering the plant response meeting the requirements. In Fig. 22.31, y is the measured process variable, r is the reference variable, u is the control signal. The control effort, eðtÞ ¼ yðtÞ  r ðtÞ is derived from the feedback, y(t) as well as the required speed control such as reference variable r(t) before tuning the following PID parameters for attaining the stable plant response: Kp ¼ K Ki ¼ K=Ti Kd ¼ K=Td

Proportional control gain Integral control gain Differential control gain.

The integral, proportional and derivative parts can be estimated as control actions based on the past, the present and the future using first order step response output in the Simulink. The equation of the controller in discrete domain for tuning the gain constant to attain the required response is C ðsÞ ¼ K p þ K i

Fig. 22.31 Closed loop feedback

1 þ Kd s s

830

M. Leelakumar

Also, the equation is expressed in the time domain as shown below (

1 Zt deðtÞ uð t Þ ¼ K e ð t Þ þ eðsÞds þ Td Ti 0 dt

)

The simple model is sufficiently reliable to provide satisfactory PID gains. The closed loop algorithm and tuning of the gain constant for stable response is achieved in different vehicle platforms. This methodology automates the calibration constant of idle speed controllers while satisfying requirements on stability, performance, and robustness of the closed-loop system. Tuning method for speed response is started with gain, Kp to attain the set speed with this proportional variation. After attaining the values of the required speed with proportional gain as well as stability of the response, Ki is tuned for attaining the set speed within the required tolerance for speed, within the desired time period, at steady state. Any change in response from a high-speed to the low speed or intermittent speed can cause undershoot and engine may shut off. For transient control, Kd is tuned to have smooth behaviour from one steady state to set speed point.

References AIS137 (part 4, ch. 8), Test method, Testing equipment and Related Procedures for Type Approval and Conformity of Production (COP), Testing of M and N category vehicles having GVW exceeding 3500 kg for Bharat Stage VI (BSVI) Emission Norms as per CMV Rules 115, 116 and 126 Bosch R (2014) Bosch automotive electrics and automotive electronics: systems and components, networking and hybrid drive Gerhardt J, Hönninger H, Bischof H (1998) A new approach to functional and software structure for engine management systems-BOSCH ME7. No. 980801. SAE Technical Paper Jaikamal V, Zurawka T (2010) Advanced techniques for simulating ECU C-code on the PC. No. 2010-01-0431. SAE Technical Paper Reif K (ed) (2014) Diesel engine management. Springer Vieweg, Wiesbaden. (Reference Book) Thate JM, Kendrick LE, Nadarajah S (2004) Caterpillar automatic code generation. No. 2004-01-0894. SAE Technical Paper

Part III

Noise and Vibration

Chapter 23

Study of Noise and Vibration Problems Related to Heavy Duty Diesel Engines P. A. Lakshminarayanan

Abstract Noise can be described as what is heard, and vibration as what is felt by a person. The pulsation of air particles in contact with vibrating structure or fluctuating flow causes noise. Noise from the engine due to exhaust and inlet flows are predominant at lower frequencies and at high frequencies it is usually structure related. Noise from the cooling fan is not negligible. Noise at steady state is studied but more annoying is transient noise. The pass-by and interior noise are the net effect of the noise from the engine, tires, vehicle panels, the fan and the exhaust silencer. Exhaust noise and engine noise are the loudest components of the pass-by noise from diesel trucks and buses. Vibration on the other hand, is mainly due to inertia forces left unbalanced by design as well as inevitable manufacturing tolerances to the masses and linear dimensions. Further, the side thrust to the cylinder walls and the forces at the bearings induce a vibrating torque about the crankshaft axis, proportional to the brake mean effective pressure. This chapter is written under three subtitles namely, noise at steady state, transient noise and engine vibrations.

23.1

Noise at Steady State

The sources of noise in a diesel vehicle make an interesting study for their effect on the interior and exterior of the vehicle. Mainly, the noise from the silencer and the radiator fan, and vibrations of the surfaces of the engine and the vehicle structures contribute to the total noise. The noise from the silencer is measured as passby noise according to standards (Indian Standard 1998) and can be estimated by a full thermodynamic model of the engine coupled with the silencer (Lakshminarayanan 2013) with reasonable accuracy. Though, it is possible to predict the noise emitted by the engine structure by virtual prototyping—a worthwhile goal, the current state-of-the-art generally allows such approaches to be used for trend analysis only

P. A. Lakshminarayanan (&) Formerly with Simpson and Co. Ltd., Chennai, Tamil Nadu, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. A. Lakshminarayanan and A. K. Agarwal (eds.), Design and Development of Heavy Duty Diesel Engines, Energy, Environment, and Sustainability, https://doi.org/10.1007/978-981-15-0970-4_23

833

834

P. A. Lakshminarayanan

rather than for bypassing a physical prototype stage altogether. Therefore, it is necessary to rank the noise by measurements and try methods to reduce them. In this paper, a case study of ranking the noise by the vibrating surfaces of the engine structure using of pulse B&K equipment (Bruel and Kjaer) in ranking the noise and steps to reduce the radiation of the noise energy is discussed.

23.1.1 Sound Intensity and Energy Sound intensity is the rate of flow of sound energy through a unit area perpendicular to the direction the flow. It is defined as,
I ¼ p
v Here, p is the instantaneous pressure and
v is the particle velocity. The versatile B&K Pulse equipment is used for measuring noise and vibration. Here two microphones kept at 12 mm measure the sound pressure. The amplitude remaining nearly the same, there is a small phase difference along the distance of separation, because sound takes definite time to travel the distance between the microphones. Using the two pressures, particle velocity at the midpoint between the two microphones is calculated (Raleigh 1896, 1945) and hence the sound intensity. The surface integral of the intensity on the entire closed area, S enclosing the engine gives the net outflow of the sound power is obtained. ZZ Sound power ¼ 
I:d
S S

The sound intensity meter measures the noise intensity of the surrounding neighbourhood as well. However, their net effect on the integral is zero as these sources are not enclosed by the closed surface according to the Gauss–Ostrogradsky divergence theorem (Piskunov 1965). In other words, any noise vector originating outside the closed surface is either reflected by the hard surfaces of the engine or simply passes through the air if not intercepted by the engine surfaces. When the measurement surface is close to the engine surface, the near field noise is evaluated. The ratio of total sound power radiated by the engine (excluding the inlet and exhaust gas flow effects) to the mechanical shaft power is defined as the noise radiating efficiency. Figure 23.1 shows the radiating efficiency is a strong function of the engine speed for many direct injection diesel engines of bores from one meter to as small as 7.5 cm. The design of the engines and the brake mean effective pressure cause the scatter.

23

Study of Noise and Vibration Problems …

835

Noise radia ng efficiency, dBA

-45 -50 -55

maximum

-60 -65

minimum -70 -75 400

900

1400

1900

2400

2900

3400

Engine speed, rpm

Fig. 23.1 Engine noise radiating efficiency

23.1.2 Ranking Engine Surfaces for Noise Emission 23.1.2.1

Test Set up

The basic engine without the fan is loaded at a dynamometer at a constant speed, Fig. 23.2a. A cuboid shaped net is created around the engine on the area, S. The sound intensity is measured at the midpoints of the individual eye of the net and its contours are plotted. Also, the integrals over the rectangular areas in the near field of each critical part, e.g. engine oil sump, are evaluated individually for the sound energy radiated by the part, assuming the area is nearly parallel to the radiating surface. The contours of the sound intensity indicate hot spots and the integrals give its contribution to the total sound energy. For example, the contours on the front side surface are shown in Fig. 23.2b. Plots of all the other sides are shown later.

23.1.2.2

Results of Baseline Measurement

Engine test and ranking First, the base line measurement is taken and the sound energy obtained for different surfaces covering critical components is listed and ranked, Table 23.1 and the components were identified for noise reduction. Figure 23.3 gives the actual contribution of each component to the noise energy flow. This list excludes the noise generated by the air and gas flows at the inlet and the exhaust.

836

P. A. Lakshminarayanan

Fig. 23.2 a Engine set up, b Noise intensity contours of the base engine on the front side surface

Table 23.1 Ranking of parts for noise energy

23.1.2.3

Rank

Engine component

Rank

Engine component

1 2 3 4 5 6 7

Crank case Inlet manifold Sump Rocker cover Damper Fuel injection pump Timing gear case

8 9 10 11 12 13 14

Flywheel housing Injectors Starter motor Air compressor Alternator Oil cooler Exhaust manifold

Vehicle

Transfer path technique (Plunt 2005) was applied to the vehicle to rank various sources contributing to passby noise in air, when the fan was not engaged. The silencer, crank case and the sump were identified as the top three components to be treated.

Study of Noise and Vibration Problems …

Noise Power, dBA (ref= 1 pico W)

23

837

120 115 110 105 100 95 90

Fig. 23.3 Noise energy flow from different components

23.1.2.4

Techniques to Reduce Noise Radiation from Different Parts of the Engine

The techniques adopted for various surfaces of the top five components, Table 23.2. Crank case: Fuel pump and exhaust sides In an actual vehicle, the sides can be covered by hanging noise absorbing sheets, in the vehicle, Fig. 23.4. For the experiment at the engine test cell, the hot spots of these sides are covered by noise absorbing sheets. Sump A sump made of viscoelastic material of very small thickness sandwiched between two steel sheets each half as thick as the regular sump, Fig. 23.5. Thus, this layout does not call for any change in the assembly line or the engine layout. The Table 23.2 Five top ranked components 1 2 3 4 5

Component

Method

Crank case (fuel pump and exhaust sides) Sump Timing gear casing Valve rocker cover Pulley/belt

Noise suppression by shielding Reduction at source by material change Reduction at source Reduction at source by material change Belt type selection for reducing flutter, rigid pulleys with spokes

838

P. A. Lakshminarayanan

Fig. 23.4 Typical (experimental) arrangement in a vehicle to shield noise from the crankcase sides

Fig. 23.5 Viscoelastic material sandwiched between steel sheets

propagation of vibrations to the metallic layers generates micro-movements between them. The viscoelastic interlayer damps these movements (Andersson 2012). The sump was tested independently by external excitation at 95 °C (courtesy Trelleborg) by attaching accelerometers at different antinodes of the sump, Fig. 23.6. Figure 23.7 shows that there is a reduction in vibration mobility by at least 6 dB at all frequencies on all the sides compared to the standard sump for an excitation force of one newton, when the sump is made of sandwich material. However, the grey curve shows only the marginal improvement if on a regular sump with damping material is patched up at typical antinodes. Plastic sump Also, a thermoplastic sump was developed for the engine which is quiet throughout the frequency range of operation. Figure 23.8 shows the design and the actual sump fitted to the engine. For the study of the noise reduction, the plastic sump was used. The mobility of the plastic sump is lower than oil sump made of sandwich material. However, the assembly and gasket are specially designed. For demonstrating the noise reduction capability, the plastic sump was used. Valve Rocker cover A rocker cover of SMC (sheet moulded compound) was developed and used for trials, Fig. 23.9. However, in a vehicle, the bonnet is covered by noise absorbing material and the noise emitted by the valve rocker cover may not be radiated to the interior of the vehicle or to the passer by.

Study of Noise and Vibration Problems …

23

839

Mobility, dB [Ref: 0 dB = 1m s-1 N -1 ]

Fig. 23.6 Arrangement to study the damping behaviour of the engine oil sump made of viscoelastic sandwich material

-60 -70 -80 -90 -100 -110

standard

patching

Viscous laminar

-120 0

1000

2000

3000

4000

5000

6000

Frequency, Hz

Mobility, dB [Ref: 0 dB = 1m s-1 N -1 ]

Mobility on the front side of the sump for excitation by one newton force -60 -70 -80 -90 -100 -110

standard

patching

Viscous laminar

-120 0

1000

2000

3000

4000

5000

6000

Frequency, Hz Mobility on the bottom side of the sump for excitation by one newton force

Fig. 23.7 Comparison of mobility of oil sumps: standard, viscoelastic lamina and standard sump patched with damping material

840

P. A. Lakshminarayanan

Fig. 23.8 Plastic sump- CAD model and the actual sump

Fig. 23.9 Valve rocker cover made of SMC

Table 23.3 Improvement in noise

Side

% reduction in noise power

Fuel injection pump side Exhaust side Front side Top side Flywheel side Overall

51 26 26 61 28 39

Timing gear casing: Front side The cogged vee belts have large surface area and tend to flap to produce noise. A poly vee belt (apart from its other advantages like long life and ability to use smaller pulleys) reduces the flap and other types of noise substantially. Results Table 23.3 shows that about 40% of the noise power (Watt) emitted by different surfaces of the engine could be reduced and Fig. 23.10 shows the substantial shrinking of contours corresponding to hot spots because of the improvements.

Front side 102 dBA contour has been replaced by 100 dBA Exhaust side 100 dBA contour has shrunk and 97 dBA is in its place

Fuel pump side 102 dBA contour has shrunk to 100 dBA

Study of Noise and Vibration Problems …

Fig. 23.10 Shrinking of hot spots after improvements

99 dBA has given way t0 95 dBA

23 841

842

P. A. Lakshminarayanan

23.2

Transient Noise from the Silencer

Limits to noise emission are laid down in the Indian Central Motor Vehicles Rules (CMVR) for different categories of vehicles (Indian Standard 1998). Simulation using Linear Acoustic Model in AVL Boost is helpful to the designers to check their concepts virtually for relative improvements and avoid time consuming experiments. The linear acoustics model is made up of a source, a system and a termination, no matter how complex or how many elements are used in the modelling. Thus, noise parameters like insertion loss and transmission loss can be obtained during the design stage itself. Further, the calculations can be used to validate the measurement data in an actual experiment (Lakshminarayanan 2013). Standard procedure and guidelines for measurement of noise are laid down in the national standards. The levelled test site consists of a central acceleration section surrounded by a substantially flat dry track to maintain the rolling noise low. There should be no obstacle that affects the sound field within the vicinity of the microphone and the sound source. The measurements are not to be done in adverse weather conditions such as rain, storm and the heavy winds that would affect noise, instruments and test results. For measurements, the A-weighted sound level of the background noise should be at least 10 dBA below the sound level produced by the vehicle.

23.2.1 Simulation of Transient Operation For the transient cycle simulation of the vehicle, the following important inputs are required: • • • • • •

Model of the silencer Driving cycle (velocity versus time) Full model of the turbocharger and combustion Engine control unit (ECU): for defining the fuel quantity, rate of injection maps. Vehicle gear ratios Inertial values for gear trains, engine and the vehicle.

23.2.1.1

Silencer Model

The layout of the experimental silencer is given in Fig. 23.11. The silencer is split into several sections to prepare a linear model, Fig. 23.12 (Bodén 1995; Fairbrother et al. 2005). The three pipes 7, 8 and 9 in Fig. 23.11 are represented by a single pipe of equal cross-sectional area and equal porous area. This enabled representing this complex geometry elegantly as a pipe-in-pipe (PIP2) model. Two pipe-in-pipe

23

Study of Noise and Vibration Problems …

843

Fig. 23.11 Silencer layout

Fig. 23.12 Representation of silencer in Fig. 23.11

(PIP1, PIP2) sections and one perforation (PER1) are introduced. The interconnections are made using pipes 29, 37, 32, 33, 35, 40 and 43. The pipe 33 jutting into the central chamber is suitably specified. The rectangular cross section of the outer can is converted to a circular section of equal cross-sectional area. Therefore pipes

844

P. A. Lakshminarayanan

Fig. 23.13 Linear model of the representation given in Fig. 23.12

30, 41, 42, 39, 34, 35 and 36 are represented by equivalent circular pipes. One of the ends (SB2, SB6, SB3 and SB4) of each pipe 30, 39, 41, 42, 34 and 36 has a boundary with zero flow coefficients to simulate a wall. Junctions J10, J9 and J11 are used to combine two or more pipes. Pipe 43 corresponds to the duct from the silencer to the atmosphere. The end of 43 open to the surroundings is terminated by SB5 with a flow coefficient equal to 1 in both the directions. The microphone MIC1 is connected to the end of pipe 43 by specifying its x, y and z positions with respect to the outlet of the pipe and its position with respect to the ground. This system is entered into the graphic user interface (GUI) of the AVL-Boost as shown in Fig. 23.13.

23.2.2 Engine Model The circuit diagram for the simulation on the graphic user interface is shown in Fig. 23.14 for a 6-cylinder engine. The air enters into the air-cleaner, CL1 through the system boundary, SB1. The air is compressed in the compressor, C of turbocharger, TC1 and goes into the intake manifold, PL1 via air cooler, CO1. The fuel injected in the individual cylinders (C1, C2, C3, C4, C5 and C6) mixes with the incoming air. After combustion, the exhaust gas expands in the turbine, T of the turbocharger (TC1) and finally escapes into the environment through system boundary SB2. The thermodynamic properties are monitored at the measurement points MP1, MP2 etc. An ECU1 and an engine interface EI1 are connected to give necessary inputs to the cylinder and the turbocharger. The ECU scheme is shown in Fig. 23.15 and dealt with later in Table 23.4. A monitor MNT1 is connected to monitor some important variables like turbocharger speed, compressor pressure ratio etc. The grey lines are the electrical connections and the black lines are the pipe connections.

23

Study of Noise and Vibration Problems …

845

Fig. 23.14 Layout of the engine and the silencer for simulation using AVL boost

23.2.3 Combustion Model AVL mixing controlled combustion (MCC) model [2, 9, and 10] is one of the accurate models available in AVL Boost for prediction of heat release rate for the given speed and fuelling. This model was further improved in reference (Lakshminarayanan et al. 2002) for more realistic prediction. However, for the simulation, the basic MCC model (Chmela and Orthaber 1999) is used. Some of the inputs (Figs. 23.16, 23.17, 23.18 and 23.19) for predicting heat release rate are • • • • • •

Injector nozzle hole size Number of holes in injector nozzle Rail pressure Rate of Injection curve Combustion parameter Fueling at different loads.

846

P. A. Lakshminarayanan

Fig. 23.15 Flow chart for ECU x, y = sensor channels Output value = actuator channel (AVL Boost User’s guide, version 4.0 (May 2002)

Table 23.4 Maps for the ECU and engine interface in Fig. 23.15 input for simulation Description

Map

Input x y

Map

Output z

Baseline map

1 2

n n

L f

Fuelling Rate of injection against crank angle

3 4

n n

f –

Figure 23.16 Figure 23.19, typical at 1500 rpm Figure 23.18 Figure 23.17

–

–

Table 23.5

MCC combustion parameter Common rail injection pressure (input to MCC combustion model) –

–

–

none

–

–

–

none

–

n

p

–

Fuelling = 0

n

p

e.g., Table 23.6

Fuelling

Steady state 1 correction Acceleration – correction Deceleration – correction Minimum 1 correction Maximum 1 correction p = Pressure ratio across n = speed L = Torque/load f = fuelling

the turbo compressor

Study of Noise and Vibration Problems …

Fig. 23.16 Variation of fuelling with load signal for various speeds

Fuelling, mg/stroke/cylinder

23

847

100

800 rpm 1500 rpm 2000 rpm 2500 rpm

80 60 40 20 0 0.0

0.2

0.4

0.6

0.8

1.0

Load signal 1600

rail pressure, bar

Fig. 23.17 Variation of rail pressure with speed

1200 800 400 0 0

1000

2000

3000

speed, rpm 1.2 1.0

MCC Parameter

Fig. 23.18 Variation of MCC combustion parameter with fuelling for various speeds

0.8 0.6

800 rpm 1500 rpm 2000 rpm 2500 rpm

0.4 0.2 0.0 0

20

40

60

Fuelling, mg/stroke Rail pressure and rate of injection curve are defined in the ECU for the corresponding speed and fuelling, and several important constants like the MCC combustion parameter constant. The indicated work done on the piston is calculated

848

P. A. Lakshminarayanan 0.00006

8.3 mg 25.8 mg 42.4 mg

Rate of injecƟon

0.00005

14.4 mg 31.6 mg 48.9 mg

20.1 mg 37 mg 55.4 mg

0.00004 0.00003 0.00002 0.00001 0.00000 -10

0

10 crank angle, degree

20

30

Fig. 23.19 Rate of injection curves for different fuelling at 1500 rpm

applying the first law of thermodynamics to the heat release rate and heat losses to the walls. Frictional losses are subtracted from the indicated work to obtain the brake power for a given speed and fuelling. Electronic Control Unit In most modern engines, many functions of the engine are controlled by an electronic engine management system. It is necessary to model such a control device especially for the simulation of transients. In the ECU, the same input to the unit produces different outputs depending on the state of the unit. The engine control model features three states: steady state, engine acceleration and engine deceleration. An ECU can be created for simulation of mechanical engines as well, as amply demonstrated in reference (Lakshminarayanan 2013). The transition from steady state to the state of engine acceleration is triggered if the gradient of the load signal versus time exceeds a threshold specified by the user. The transition to engine deceleration is triggered the same way when the negative gradient of the load signal exceeds the user specified threshold. A baseline steady state value is taken from the baseline map. This value is subjected to corrections by adding values from correction maps or by multiplying it by factors from correction maps. In the case of acceleration or deceleration, other corrections are applied to the steady state value from the corresponding acceleration or deceleration correction maps. Then it is checked whether the output is within predefined bounds which themselves are defined as maps. If it is within the maximum map or minimum map, then the same value is considered as the output value. If it is outside the minimum or maximum maps, then the corresponding minimum or maximum value is taken from the maps. ECU data for the 165 kw engine given for simulation is shown in Tables 23.5 and 23.6, and Figs. 23.16, 23.17, 23.18 and 23.19. During simulation, based on

23

Study of Noise and Vibration Problems …

849

Table 23.5 Data for steady state correction map Steady state correction

Parameter

On time

Off time

Positive load changes

Lag time for correction Time constant for correction Lag time for correction Time constant for correction

1 1 1 1

1 1 1 1

Negative load changes

s s s s

s s s s

engine speed and load, the ECU selects the corresponding fuelling required per cylinder from the baseline map plotted as graph in Fig. 23.16. To the base value, the steady state correction values are applied based on the load changes. For brevity, the correction applied is same for all, irrespective of the load change. The fuelling obtained is compared against the maximum and minimum maps. The minimum quantity of fuelling is specified zero (in an actual engine a very meagre amount of fuel is injected to keep the injector tip cool) and the maximum quantity of fuelling depends on speed and boost pressure, e.g., as shown in Table 23.6. If the fuelling exceeds the maximum or minimum value, then the corresponding maximum or minimum value would be taken. Otherwise, the fuelling that is obtained after the correction maps is considered as the final value for the cylinders. Corresponding to the fuelling and speed, the MCC combustion parameter and rate of injection curve were decided from the maps shown in Figs. 23.18 and 23.19. The rail pressure is selected based on the engine speed alone as shown in Fig. 23.17. All these three inputs are required for simulating the MCC combustion model. With speed changing every cycle, the importance of ECU is evident from the above graphs. Either the load signal or a desired engine speed can be selected as the guiding input signal of the ECU. If desired engine speed is selected as the guiding input signal, then ECU calculates load signal based on the current operating engine speed and desired engine speed and also using its proportional, integral and differential values. The calculated load signal is taken by engine interface. The fuelling and rate of injection curves are decided based on the speed and corresponding load signal. Zt

ls ¼ p½ndes  n þ i ½ndes  ndt þ d 0

Table 23.6 Maximum values map for fuelling (mm3 stroke−1 cylinder−1)

d ½ndes  n dt

Boost pressure (kg cm−2)

Speed (rpm) 500 1000

1500

2000

2500

0.0 1.0 1.5 2.0 2.5

35 35 52 75 75

35 35 52 85 85

35 35 52 85 85

35 35 52 75 75

35 35 52 75 75

850

P. A. Lakshminarayanan

Table 23.7 p, i, d constants used in the simulation ECU to control the actual speed close to the desired speed

Turbocharger

ls p i d ndes n

Proportional constant (p) Integral constant (i) Differentiation constant (d) Inertia

Units

Value

1 rev min−1 rev−1 min s−1 s rev−1 min kg mm2

0.005 0.005 0 22

load signal proportional control gain (rpm−1) integral control gain (rpm−1) differential control gain (s rpm−1) desired vehicle speed reduced to crankshaft speed (rpm) engine speed (rpm)

The constants p, i and d for the two cases are given in Table 23.7. Engine Interface The Engine Interface Element is used to supply data to elements in a computer model which are connected by wires. In the Boost version 4.0, the link to external applications via the Engine Interface is not available. Actuators are served with data from Data Sets only. Required input for a data set definition is its name, the unit of the evaluated values and the type of the data base which can be either a constant value or one of the following. • • • • •

Table List of Tables Regular Map Cyclic updated Table Map of Tables.

Fuelling maps, rate of injection maps, rail pressure maps, combustion parameter maps are specified in the Engine interface element (EI). Turbocharger For accurate prediction of turbocharger speeds, the entire turbocharger data should be given as input. Also, turbocharger speeds are predicted when the option of full model of the turbocharger is invoked in the software. The inputs are the compressor maps surge line data, the turbine map and the moment of inertia of turbocharger, Fig. 23.20.

23

Study of Noise and Vibration Problems …

Fig. 23.20 Typical compressor and turbine map for the 165-kW engine

851

1,39,000 rpm 70%

Pressure raƟo

73%

65%

75% 76%

77% 60%

Volume flow

pressure ra o efficiency

Volume flow

Pressure ratio, mass flow, turbocharger speed and efficiencies are extracted from the maps of the compressor and the turbine. During simulation, the corresponding pressure ratio and mass flow are selected based on engine speed and load.

23.2.3.1

Driving Cycle

The vehicle details are given below, Table 23.8. The performance of the engine for a driving cycle depends on the parameters like the nature of track (road), gear ratios, rear axle ratio, driver and vehicle Table 23.8 Vehicle details

Parameter

Value

Width of truck Vehicle weight Friction coefficient Friction force Frontal area

2.5908 49000 0.0022 1057 14

m kg – N m2

852

P. A. Lakshminarayanan

Table 23.9 Gear steps Gear step

Total ratio

Gear step

Total ratio

1 2 3 4 5

74.2 51.5 36.6 27.0 20.3

6 7 8 9

14.8 10.6 7.8 5.8

tonnage. For the two cases chosen here, Table 23.9 gives the gear ratios considering the rear axle ratio. Gear shifting happens if the engine speed falls below the minimum gear shifting speed of the engine or rises above the maximum gear shifting speed. If the engine speed falls below the minimum gear shifting speed of the engine, the program shifts the gear from higher to lower and if the engine speed rises above maximum gear shifting speed the computer program shifts the gear from lower to higher. The minimum and maximum gear shifting speed is given by the user. Gear ratio ¼

23.2.3.2

Speed of the engine shaft Speed of the wheel at the rear axle

Load on the Vehicle

There are two types of loads acting on the vehicle namely, the rolling resistance and the drag force. Suitable constants should be given to define these loads. The constants can be defined with respect to time also. F = a + b v2 where the parameters affecting the load on the vehicle, F are given in Table 23.10.

Table 23.10 Parameters determining the load on the vehicle F v a b Cr Mv g qa CD Av

Total load on a vehicle Vehicle speed Rolling resistance, Cr  Mv  g ½ qa CD Av Coefficient of rolling resistance, Table 23.11 Mass of vehicle corresponding to the gross vehicle weight (GVW) Acceleration due to gravity Air density Drag coefficient, Table 23.12 Frontal area of the vehicle

N m s−1 N N s2 m−2 kg m s−2 kg m−3 m2

23

Study of Noise and Vibration Problems …

853

Table 23.11 Rolling resistance coefficient, Cr (Clark and Dodge 1979) Cr

Description

0.0002– 0.0010

Railroad steel wheel on steel rail Hardened steel ball bearings on steel production bicycle tires at 120 psi (8.3 bar) and 50 km h−1, measured on rollers Special Michelin solar car/eco-marathon tires Tram rails standard dirty with straights and curves Typical BMX bicycle tires used for solar cars Car tire measurements Ordinary car tires on concrete Ordinary car tires on grass, mud, and sand Ordinary car tires on sand

0.0022–0.005 0.0025 0.005 0.0055 0.0062–0.015 0.010–0.015 0.055–0.065 0.3

Table 23.12 Vehicle drag coefficient CD (Nnamani and Atiqullah 2017)

Vehicle type

Drag coefficient, CD Low Medium

High

Experimental Sports Performance 60s Muscle Sedan Motorcycle Truck Tractor-Trailer

0.17 0.27 0.32 0.38 0.34 0.50 0.60 0.60

0.23 0.38 0.38 0.50 0.50 1.00 1.00 1.20

0.21 0.31 0.34 0.44 0.39 0.90 0.90 0.77

Rolling resistance Rolling resistance; a ¼ GVW  g  Cr ¼ 49000  9:81  0:0022 ¼ 1057 N Aerodynamic Drag The aerodynamic feature of a vehicle in motion is given by the drag coefficient, Table 23.12. Lower the coefficient, lower is the air resistance. Calculation of Frontal Area Frontal area represents the area of the silhouette of the vehicle from the front. Coefficient b ¼ 0:5  air density  Drag coefficient; CD  Frontal area ¼ 0:5  1:2  1  14 ¼ 8:4 N sm2 Various vehicle parameters are given in Table 23.13.

854

P. A. Lakshminarayanan

Table 23.13 List of inputs required exclusively for simulating 49-ton vehicle

Vehicle mass Maximum clutch torque to avoid clutch slip Tire radius Rolling resistance constant (a) Drag force constant (b) Minimum gear shifting speed Maximum gear shifting speed Simulation time

49000 1500 0.5 6000 30 1500 2500 11

kg Nm m

rpm rpm min

23.2.4 Measurement for Passby Noise 23.2.4.1

Standard Procedure: Heavy Duty Vehicles with Engine Power Less Than 225 KW

See Fig. 23.21a. The heavy-duty vehicle with an engine power less than 225 kW (Indian Standard, IS3028, 1998) is driven to approach the line AA′ at a speed of 50 km h−1 or at the speed corresponding to an engine speed equal to three quarters of the speed, S at which the engine develops its maximum rated power, whichever is lower. The heavy vehicle in which the total number of forward gear ratio is x (including the ratios obtained by means of an auxiliary gear box or a multi-ratio drive axle) is first tested for the passby noise using a gear ratio equal to or higher than x/n, where n = 2. If x/n does not correspond to a whole number, the nearest higher gear ratio is used. When the front end of the vehicle with above approach speed and one gear ratio reached the line AA′, the accelerator control is fully operated as rapidly and smoothly as possible and held in the fully opened position until the rear of the vehicle reaches the line BB′; the accelerator control is released as quickly as possible. The trailers of articulated vehicles which cannot be uncoupled are not be considered with regard to crossing of line BB′. Then the experiment is repeated by up-shifting the gear from x/n and terminated in the gear X where the engine develops its rated maximum power speed, S while passing the line BB′. The test result is the passby noise obtained from the gear ratio producing maximum sound level.

23.2.4.2

Experimental Switching Technique for Indoor Measurement

Instead of driving the test-vehicle past two stationary microphones as stated by the standard, indoor measurement setups use two rows of microphones placed alongside the test vehicle (Fleszar et al. 2001), Fig. 23.21b. The vehicle is run on a chassis dynamometer and accelerated in the same way as stated in the standard, however for the equivalent distance travelled by the periphery of the wheels. Time

23

Study of Noise and Vibration Problems …

855

Fig. 23.21 a Schematic diagram of the test track and b the arrangement of eleven microphones fixed with respect to the engine in simulation

histories are measured by the microphones in parallel with vehicle parameters and the speed of the periphery of the wheel. Using the information from the dynamometer, a sophisticated algorithm calculates the position of the vehicle relative to the microphones as a function of time. The calculated position and the time histories of noise measured by different microphones are used to extract the contributions to the noise at the position of the standard microphone (Indian Standard 1998), had the vehicle been moving. The extraction considers the decay of noise with distance and the Doppler Effect. A synchronized single time history is created by stitching these time history sections together and interpolating across the boundaries of the segments. The set up can repeat testing conditions precisely and is available throughout the year irrespective of the weather.

856

P. A. Lakshminarayanan

23.2.5 Simulation of Passby Noise Switching technique is employed in the present simulation work as the microphones are fixed with respect to the engine in AVL-Boost software.

23.2.5.1

Switching Technique

Eleven microphones are put at 6 m distance perpendicular to vehicle exhaust (7.5 m from the vehicle centre line) and 1 m vertically from the outlet and laterally from −10 m to +10 m at an interval of 2 m. The silencer outlet itself is at 0.25 m above the ground. The microphones are at a height of 1.2 ± 0.1 m as specified in the standard. In the simulation, the effect of the reflection by the hard ground is considered. Therefore, the microphones are on a Lagrangian frame when simulated using the AVL-Boost software but in the standard procedure (Indian Standard 1998) it should be Eulerian. Noise measured by all the microphones in the simulation is post processed using simple transformation from Lagrangian to Eulerian frame to obtain the passby noise. However, the Doppler Effect is neglected during this transformation. For example, if the vehicle is running at less than 50 km h−1, the frequency shift of noise will be less than 4%. In other words, when A-filtered and transformed, shifting of frequency spectrum and its effect or attenuated in the atmosphere are not considered.

23.2.5.2

Calculation Procedure to Satisfy the Indian Standard 3028

First, the vehicle approaches at ¾th of the rated speed of the engine (*1900 rpm) at a gear > x/n (>9/2 = 5th gear). The approach speed is 17.6 km h−1 (