Magnesium Technology 2019 [1st ed.] 978-3-030-05788-6, 978-3-030-05789-3

Table of contents :
Front Matter ....Pages i-xxi
Front Matter ....Pages 1-1
Magnesium Alloy Sheet for Transportation Applications (Chris Romanowski)....Pages 3-12
Magnesium for Automotive Lightweighting: Status and Challenges (Sarah Kleinbaum)....Pages 13-14
Magnesium Process and Alloy Development for Applications in the Automotive Industry (David Klaumünzer, Jose Victoria Hernandez, Sangbong Yi, Dietmar Letzig, Sang-hyun Kim, Jae Joong Kim et al.)....Pages 15-20
Thermally Activated Slip in Rare Earth Containing Mg–Mn–Ce Alloy, ME10, Compared with Traditional Mg–Al–Zn Alloy, AZ31 (Vikaas Bajikar, Jishnu J. Bhattacharyya, Nathan Peterson, Sean R. Agnew)....Pages 21-22
Front Matter ....Pages 23-23
Bimodal Casting Process of Eco-Mg Series Alloys by Vertical High-Speed Press Machine (Fabrizio D’Errico)....Pages 25-31
Investigation of the Evolution of the Microstructure in the Directionally Solidified Long-Period Stacking-Ordered (LPSO) Magnesium Alloy as a Function of the Temperature (Daria Drozdenko, Kristián Máthis, Stefanus Harjo, Wu Gong, Kazuya Aizawa, Michiaki Yamasaki)....Pages 33-36
TEM Studies of In Situ Formation of MgO and Al4C3 During Thixomolding of AZ91 Magnesium Alloy Conducted in CO2 (Ł. Rogal, L. Litynska-Dobrzynska, Bogusław Baran)....Pages 37-41
FFF of Mg-Alloys for Biomedical Application (M. Wolff, T. Mesterknecht, A. Bals, T. Ebel, R. Willumeit-Römer)....Pages 43-49
Effects of Gd/Y Ratio on the Microstructures and Mechanical Properties of Cast Mg–Gd–Y–Zr Alloys (J. L. Li, D. Wu, R. S. Chen, En-Hou Han)....Pages 51-56
Front Matter ....Pages 57-57
Evolution of Heterogeneous Microstructure of Equal-Channel Angular Pressed Magnesium (Qizhen Li)....Pages 59-63
Novel Magnesium Alloy Processing via Shear-Assisted Processing and Extrusion (ShAPE) (S. Mathaudhu, N. Overman, S. Whalen, M. Olszta, D. Catalini, K. Kruska et al.)....Pages 65-67
Effects of the Extrusion Temperature on Microstructure, Texture Evolution and Mechanical Properties of Extruded Mg–2.49Nd–1.82Gd–0.19Zn–0.4Zr Alloy (Lei Xiao, Guangyu Yang, Shifeng Luo, Wanqi Jie)....Pages 69-75
Influence of Thermomechanical Treatment on Tension–Compression Yield Asymmetry of Extruded Mg–Zn–Ca Alloy (P. Dobroň, M. Hegedüs, J. Olejňák, D. Drozdenko, K. Horváth, J. Bohlen)....Pages 77-81
Homogeneous Grain Refinement and Ductility Enhancement in AZ31B Magnesium Alloy Using Friction Stir Processing (Vivek Patel, Wenya Li, Quan Wen, Yu Su, Na Li)....Pages 83-87
Microstructure and Texture Evolution During Hot Compression of Cast and Extruded AZ80 Magnesium Alloy (Paresh Prakash, Amir Hadadzadeh, Sugrib Kumar Shaha, Mark A. Whitney, Mary A. Wells, Hamid Jahed et al.)....Pages 89-94
Experimental Investigation of Friction Coefficient of Magnesium Alloy Developed Through Friction Stir Processing with PKS Ash Powder Particles (R. S. Fono-Tamo, Esther Titilayo Akinlabi, Jen Tien-Chien)....Pages 95-99
A Review and Case Study on Mechanical Properties and Microstructure Evolution in Magnesium–Steel Friction Stir Welding (Suryakanta Sahu, Omkar Thorat, Raju Prasad Mahto, Surjya Kanta Pal, Prakash Srirangam)....Pages 101-109
Effects of Sn on Microstructures and Mechanical Properties of As-Extruded Mg−6Al−1Ca−0.5Mn Magnesium Alloy (Huajie Wu, Ruizhi Wu, Daqing Fang, Yuesheng Chai, Chao Liang)....Pages 111-117
Front Matter ....Pages 119-119
Effect of Alloying with Rare-Earth Metals on the Degradation of Magnesium Alloys Studied Using a Combination of Isothermal Calorimetry and Pressure Measurements (Lars Wadsö, Norbert Hort, Dmytro Orlov)....Pages 121-126
Effects of Li on Microstructures and Corrosion Behaviors of Mg–Li–Al Alloys (Yang Li, Tingchao Li, Qilong Wang, Yun Zou)....Pages 127-134
Galvanically Graded Interface: A Computational Model for Mitigating Galvanic Corrosion Between Magnesium and Mild Steel (Kurt A. Spies, Vilayanur V. Viswanathan, Ayoub Soulami, Yuri Hovanski, Vineet V. Joshi)....Pages 135-144
Iron Content in Relationship with Alloying Elements and Corrosion Behaviour of Mg3Al Alloys (Ha Ngoc Nguyen, Jong Il Kim, Young Min Kim, Bong Sun You)....Pages 145-150
Microstructures, Corrosion and Mechanical Properties of Mg–Si Alloys as Biodegradable Implant Materials (Weidan Wang, Ming Gao, Yuanding Huang, Lili Tan, Ke Yang, Norbert Hort)....Pages 151-157
The Influence of Temperature and Medium on Corrosion Response of ZE41 and EZ33 (M. AbdelGawad, A. U. Chaudhry, B. Mansoor)....Pages 159-167
Alloy Design Strategies of the Native Anti-corrosion Magnesium Alloy (Tao Chen, Yuan Yuan, Jiajia Wu, Tingting Liu, Xianhua Chen, Aitao Tang et al.)....Pages 169-173
Corrosion Bending Fatigue of RESOLOY® and WE43 Magnesium Alloy Wires (Petra Maier, Adam Griebel, Matthias Jahn, Maximilian Bechly, Roman Menze, Benjamin Bittner et al.)....Pages 175-181
Sacrificial Cathodic Protection of Mg Alloy AZ31B by an Mg–5Sn Surface Alloy (C. F. Glover, T. W. Cain, J. R. Scully)....Pages 183-190
Front Matter ....Pages 191-191
Evolution of the Intermetallic Particle Distribution in Thixomolded Magnesium Alloys (B. T. Anthony, B. G. Dowdell, V. M. Miller)....Pages 193-197
Revealing the Role of Combined Loading on the Tension–Compression Asymmetry in a Textured AZ31 Magnesium Alloy (C. Kale, S. Srinivasan, P. Haluai, K. N. Solanki)....Pages 199-200
An Investigation of Detwinning Behavior of In-plane Compressed E-form Mg Alloy During the In Situ Tensile Test (Jaiveer Singh, Min-Seong Kim, Seong-Eum Lee, Joo-Hee Kang, Shi-Hoon Choi)....Pages 201-206
Characterization of Staggered Twin Formation in HCP Magnesium (M. Arul Kumar, B. Leu, P. Rottmann, I. J. Beyerlein)....Pages 207-213
Dislocation Behavior and Grain Boundary Segregation of Mg–Zn Alloys (Hyo-Sun Jang, Byeong-Joo Lee)....Pages 215-218
Effect of Hot Working on the High Cycle Fatigue Behavior of WE43 Rare Earth Magnesium Alloy (Saeede Ghorbanpour, Brandon A. McWilliams, Marko Knezevic)....Pages 219-225
Effect of Solute Atoms on the Twinning Deformation in Magnesium Alloys (Jing Tang, Wentao Jiang, Xiaobao Tian, Haidong Fan)....Pages 227-230
First-Principles Investigation of the Effect of Solutes on the Ideal Shear Resistance and Electronic Properties of Magnesium (P. Garg, I. Adlakha, K. N. Solanki)....Pages 231-237
Inverse Optimization to Design Processing Paths to Tailor Formability of Mg Alloys (Wahaz Nasim, Joshua S. Herrington, Amine A. Benzerga, Jyhwen Wang, Ibrahim Karaman)....Pages 239-246
Front Matter ....Pages 247-247
Recent Progress in Development and Applications of Mg Alloy Thermodynamic Database (Rainer Schmid-Fetzer)....Pages 249-255
Hardening Effects of Precipitates with Different Shapes on the Twinning in Magnesium Alloys (Haidong Fan, Jaafar A. El-Awady)....Pages 257-261
Isometric Tilt Grain Boundaries and Solute Segregation in a Deformed Mg–Zn–Ca Alloy (Y. M. Zhu, J. F. Nie)....Pages 263-266
Metallography of Mg Alloys (Norbert Hort, Victor Floss, Sarkis Gavras, Gert Wiese, Domonkos Tolnai)....Pages 267-276
Microstructure and Mechanical Properties of High Shear Material Deposition of Rare Earth Magnesium Alloys WE43 (Z. McClelland, D. Z. Avery, M. B. Williams, C. J. T. Mason, O. G. Rivera, C. Leah et al.)....Pages 277-282
Modeling the 3D Plastic Anisotropy of a Magnesium Alloy Processed Using Severe Plastic Deformation (J. S. Herrington, Y. Madi, J. Besson, A. A. Benzerga)....Pages 283-287
Multiaxial Cyclic Response of Low Temperature Closed-Die Forged AZ31B Mg Alloy (D. Toscano, S. K. Shaha, B. Behravesh, H. Jahed, B. Williams)....Pages 289-296
Thermo-mechanical Processing of EZK Alloys in a Synchrotron Radiation Beam (D. Tolnai, M.-A. Dupont, S. Gavras, K. Mathis, K. Horvath, A. Stark et al.)....Pages 297-303
The Effect of the Orientation of Second-Order Pyramidal  Dislocations on Plastic Flow in Magnesium (Kinshuk Srivastava, Jaafar A. El-Awady)....Pages 305-310
Front Matter ....Pages 311-311
Forging of Mg–3Sn–2Ca–0.4Al Alloy Assisted by Its Processing Map and Validation Through Analytical Modeling (K. P. Rao, K. Suresh, Y. V. R. K. Prasad, C. Dharmendra, N. Hort)....Pages 313-318
Development of Manufacturing Processes for Magnesium Sheet (A. Javaid, F. Czerwinski)....Pages 319-326
Front Matter ....Pages 327-327
Incorporating an ICME Approach into Die-Cast Magnesium Alloy Component Design (J. P. Weiler)....Pages 329-330
Influences of SiC Particle Additions on the Grain Refinement of Mg–Zn Alloys (Yuanding Huang, Jiang Gu, Sihang You, Karl Ulrich Kainer, Norbert Hort)....Pages 331-338
Development, Characterization, Mechanical and Corrosion Behaviour Investigation of Multi-direction Forged Mg–Zn Alloy (Gajanan Anne, S. Ramesh, Goutham Kumar, Sandeep Sahu, M. R. Ramesh, H. Shivananda Nayaka et al.)....Pages 339-343
Electrochemical Behaviour of ECAP-Processed AM Series Magnesium Alloy (K. R. Gopi, H. Shivananda Nayaka)....Pages 345-352
Effect of Split Sleeve Cold Expansion on the Residual Stress, Texture and Fatigue Life of Rolled AZ31B Magnesium Alloy (S. Faghih, S. K. Shaha, S. B. Behravesh, H. Jahed)....Pages 353-358
A Theory for Designing Ductile Materials with Anisotropy (A. A. Benzerga)....Pages 359-362
Back Matter ....Pages 363-370

Citation preview

2019 EDITED BY

Vineet V. Joshi J. Brian Jordon Dmytro Orlov Neale R. Neelameggham

The Minerals, Metals & Materials Series

Vineet V. Joshi • J. Brian Jordon • Dmytro Orlov • Neale R. Neelameggham Editors

Magnesium Technology 2019

123

Editors Vineet V. Joshi Pacific Northwest National Laboratory Richland, WA, USA

J. Brian Jordon The University of Alabama Tuscaloosa, AL, USA

Dmytro Orlov Lund University Lund, Sweden

Neale R. Neelameggham IND LLC South Jordan, UT, USA

ISSN 2367-1181 ISSN 2367-1696 (electronic) The Minerals, Metals & Materials Series ISBN 978-3-030-05788-6 ISBN 978-3-030-05789-3 (eBook) https://doi.org/10.1007/978-3-030-05789-3 Library of Congress Control Number: 2018964033 © The Minerals, Metals & Materials Society 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express 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 Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

Magnesium and its alloys are being investigated widely for applications that require extraordinarily high strength-to-weight ratio, vibration damping, electromagnetic shielding, reduction in carbon footprint, sustainability improvements, and low toxicity. In the past decade or so, this drive has led to increased adoption of magnesium and its alloys in everyday products and it is no longer regarded as an esoteric material. The overarching goal of its widespread adoption and replacement of conventional materials such as steel and aluminum alloys has been restricted because of the cost of the material, limited sources of primary material, alloying constraints, high chemical reactivity, and challenges associated with its plasticity. Researchers/scientists, engineers, and economists from industry, government agencies/laboratories, and academic institutions alike are actively developing roadmaps for next-generation products and addressing these challenges as quickly as possible through unique and innovative methods. The TMS Magnesium Committee has been actively involved in providing a platform to these entities for disseminating the latest information, developments, and cutting-edge research and development, and showcasing the latest trends related to magnesium and its alloys through the Magnesium Technology Symposium, which takes place every year at the TMS Annual Meeting. This proceedings volume retains the essence of this longstanding tradition. The 20th volume in the series, Magnesium Technology 2019, is the proceedings of the Magnesium Technology Symposium held during the 148th TMS Annual Meeting & Exhibition in San Antonio, Texas, March 10–14, 2019. The volume captures commentaries and papers from 16 different countries. The papers have been categorized systematically based on topics pertaining to casting and solidification, thermomechanical processing, deformation mechanisms, modeling, corrosion, recycling, and applications. The symposium began with keynote sessions that featured several distinguished invited speakers from government organization, industry, and academia, who provided their perspectives on the state of the art, goals, and opportunities in magnesium alloy research and development. Dr. Sarah Kleinbaum from the US Department of Energy described the status and challenges associated with vehicle lightweighting and incorporation of magnesium components in various vehicle subsystems from 2012 through today and identified the technical challenges that currently limit their full adoption. Dr. Sean Agnew from the University of Virginia discussed the role of thermally activated slip in rare earth-containing alloys and compared it with the popular traditional alloys. This was followed by a talk on Magnesium Alloy Sheet for Transportation Applications by Dr. Chris Romanowski from Danieli FATA Hunter, who reviewed the strategies that need to be implemented for widespread adoption of magnesium alloy sheet for automotive applications. The final talk in this session was presented by Dr. David Klaumuenzer from Volkswagen AG, who presented his perspective on Magnesium Process and Alloy Development for Applications in the Automotive Industry. The Magnesium Committee also co-sponsored the TMS-DGM Symposium on Lightweight Metals: A Joint US-European Symposium on Challenges in Light Weighting the Transportation Industry. This symposium, jointly organized by TMS and DGM (the German Materials Society), covered sustainable use of lightweight metals in the design and tailoring of monolithic and composite cast, wrought, and additive materials, as well as in the v

vi

Preface

manufacturing of semi-finished and finished products, and included contributions to possible scientific and technological relevance across the different material classes. The organizers of this symposium, Eric A. Nyberg, Wilhelmus H. Sillekens (European Space Agency), Juergen Hirsch (Hydro Aluminium Rolled Products GmbH), and Norbert Hort (Helmholtz-Zentrum Geesthacht) compiled six papers pertaining to magnesium technology, which have been included in this publication. Finally, the 2018–2019 Magnesium Committee is grateful to and expresses deep appreciation for all authors for contributing to the success of the symposium; our panel of distinguished keynote speakers for sharing their valuable thoughts on the future of magnesium technology; the reviewers for their best efforts in reviewing the manuscripts; and the session chairs, judges, TMS staff members, and other volunteers for their excellent support, which allowed us to develop a successful, high-quality symposium and proceedings volume. Vineet V. Joshi Chair J. Brian Jordon Vice Chair Dmytro Orlov Past Chair Neale R. Neelameggham Advisor

Contents

Part I

Magnesium Technology 2019: Keynote Session

Magnesium Alloy Sheet for Transportation Applications . . . . . . . . . . . . . . . . . . . Chris Romanowski

3

Magnesium for Automotive Lightweighting: Status and Challenges . . . . . . . . . . . Sarah Kleinbaum

13

Magnesium Process and Alloy Development for Applications in the Automotive Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David Klaumünzer, Jose Victoria Hernandez, Sangbong Yi, Dietmar Letzig, Sang-hyun Kim, Jae Joong Kim, Min Hong Seo, and Kanghwan Ahn Thermally Activated Slip in Rare Earth Containing Mg–Mn–Ce Alloy, ME10, Compared with Traditional Mg–Al–Zn Alloy, AZ31 . . . . . . . . . . . . . . . . . . . . . . . Vikaas Bajikar, Jishnu J. Bhattacharyya, Nathan Peterson, and Sean R. Agnew Part II

15

21

Magnesium Technology 2019: Alloy Design and Casting

Bimodal Casting Process of Eco-Mg Series Alloys by Vertical High-Speed Press Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fabrizio D’Errico

25

Investigation of the Evolution of the Microstructure in the Directionally Solidified Long-Period Stacking-Ordered (LPSO) Magnesium Alloy as a Function of the Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Daria Drozdenko, Kristián Máthis, Stefanus Harjo, Wu Gong, Kazuya Aizawa, and Michiaki Yamasaki

33

TEM Studies of In Situ Formation of MgO and Al4C3 During Thixomolding of AZ91 Magnesium Alloy Conducted in CO2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ł. Rogal, L. Litynska-Dobrzynska, and Bogusław Baran

37

FFF of Mg-Alloys for Biomedical Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . M. Wolff, T. Mesterknecht, A. Bals, T. Ebel, and R. Willumeit-Römer Effects of Gd/Y Ratio on the Microstructures and Mechanical Properties of Cast Mg–Gd–Y–Zr Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. L. Li, D. Wu, R. S. Chen, and En-Hou Han Part III

43

51

Magnesium Technology 2019: Thermomechanical Processing

Evolution of Heterogeneous Microstructure of Equal-Channel Angular Pressed Magnesium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Qizhen Li

59

vii

viii

Contents

Novel Magnesium Alloy Processing via Shear-Assisted Processing and Extrusion (ShAPE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S. Mathaudhu, N. Overman, S. Whalen, M. Olszta, D. Catalini, K. Kruska, J. Darsell, V. Joshi, X. Jiang, A. Devaraj, and G. Grant Effects of the Extrusion Temperature on Microstructure, Texture Evolution and Mechanical Properties of Extruded Mg–2.49Nd–1.82Gd–0.19Zn–0.4Zr Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lei Xiao, Guangyu Yang, Shifeng Luo, and Wanqi Jie

65

69

Influence of Thermomechanical Treatment on Tension-Compression Yield Asymmetry of Extruded Mg–Zn–Ca Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P. Dobroň, M. Hegedüs, J. Olejňák, D. Drozdenko, K. Horváth, and J. Bohlen

77

Homogeneous Grain Refinement and Ductility Enhancement in AZ31B Magnesium Alloy Using Friction Stir Processing . . . . . . . . . . . . . . . . . . . . . . . . . . Vivek Patel, Wenya Li, Quan Wen, Yu Su, and Na Li

83

Microstructure and Texture Evolution During Hot Compression of Cast and Extruded AZ80 Magnesium Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Paresh Prakash, Amir Hadadzadeh, Sugrib Kumar Shaha, Mark A. Whitney, Mary A. Wells, Hamid Jahed, and Bruce W. Williams Experimental Investigation of Friction Coefficient of Magnesium Alloy Developed Through Friction Stir Processing with PKS Ash Powder Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . R. S. Fono-Tamo, Esther Titilayo Akinlabi, and Jen Tien-Chien

89

95

A Review and Case Study on Mechanical Properties and Microstructure Evolution in Magnesium–Steel Friction Stir Welding . . . . . . . . . . . . . . . . . . . . . . 101 Suryakanta Sahu, Omkar Thorat, Raju Prasad Mahto, Surjya Kanta Pal, and Prakash Srirangam Effects of Sn on Microstructures and Mechanical Properties of As-Extruded Mg−6Al−1Ca−0.5Mn Magnesium Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Huajie Wu, Ruizhi Wu, Daqing Fang, Yuesheng Chai, and Chao Liang Part IV

Magnesium Technology 2019: Corrosion and Surface Protection

Effect of Alloying with Rare-Earth Metals on the Degradation of Magnesium Alloys Studied Using a Combination of Isothermal Calorimetry and Pressure Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Lars Wadsö, Norbert Hort, and Dmytro Orlov Effects of Li on Microstructures and Corrosion Behaviors of Mg–Li–Al Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Yang Li, Tingchao Li, Qilong Wang, and Yun Zou Galvanically Graded Interface: A Computational Model for Mitigating Galvanic Corrosion Between Magnesium and Mild Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Kurt A. Spies, Vilayanur V. Viswanathan, Ayoub Soulami, Yuri Hovanski, and Vineet V. Joshi Iron Content in Relationship with Alloying Elements and Corrosion Behaviour of Mg3Al Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Ha Ngoc Nguyen, Jong Il Kim, Young Min Kim, and Bong Sun You Microstructures, Corrosion and Mechanical Properties of Mg–Si Alloys as Biodegradable Implant Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Weidan Wang, Ming Gao, Yuanding Huang, Lili Tan, Ke Yang, and Norbert Hort

Contents

ix

The Influence of Temperature and Medium on Corrosion Response of ZE41 and EZ33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 M. AbdelGawad, A. U. Chaudhry, and B. Mansoor Alloy Design Strategies of the Native Anti-corrosion Magnesium Alloy . . . . . . . . . 169 Tao Chen, Yuan Yuan, Jiajia Wu, Tingting Liu, Xianhua Chen, Aitao Tang, and Fusheng Pan Corrosion Bending Fatigue of RESOLOY® and WE43 Magnesium Alloy Wires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Petra Maier, Adam Griebel, Matthias Jahn, Maximilian Bechly, Roman Menze, Benjamin Bittner, and Jeremy Schaffer Sacrificial Cathodic Protection of Mg Alloy AZ31B by an Mg–5Sn Surface Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 C. F. Glover, T. W. Cain, and J. R. Scully Part V

Magnesium Technology 2019: Fundamentals, Mechanical Behavior, Twinning, Plasticity, Texture and Fatigue I

Evolution of the Intermetallic Particle Distribution in Thixomolded Magnesium Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 B. T. Anthony, B. G. Dowdell, and V. M. Miller Revealing the Role of Combined Loading on the Tension–Compression Asymmetry in a Textured AZ31 Magnesium Alloy . . . . . . . . . . . . . . . . . . . . . . . . 199 C. Kale, S. Srinivasan, P. Haluai, and K. N. Solanki An Investigation of Detwinning Behavior of In-plane Compressed E-form Mg Alloy During the In Situ Tensile Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 Jaiveer Singh, Min-Seong Kim, Seong-Eum Lee, Joo-Hee Kang, and Shi-Hoon Choi Characterization of Staggered Twin Formation in HCP Magnesium . . . . . . . . . . . 207 M. Arul Kumar, B. Leu, P. Rottmann, and I. J. Beyerlein Dislocation Behavior and Grain Boundary Segregation of Mg–Zn Alloys . . . . . . . 215 Hyo-Sun Jang and Byeong-Joo Lee Effect of Hot Working on the High Cycle Fatigue Behavior of WE43 Rare Earth Magnesium Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Saeede Ghorbanpour, Brandon A. McWilliams, and Marko Knezevic Effect of Solute Atoms on the Twinning Deformation in Magnesium Alloys . . . . . 227 Jing Tang, Wentao Jiang, Xiaobao Tian, and Haidong Fan First-Principles Investigation of the Effect of Solutes on the Ideal Shear Resistance and Electronic Properties of Magnesium . . . . . . . . . . . . . . . . . . . . . . . 231 P. Garg, I. Adlakha, and K. N. Solanki Inverse Optimization to Design Processing Paths to Tailor Formability of Mg Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Wahaz Nasim, Joshua S. Herrington, Amine A. Benzerga, Jyhwen Wang, and Ibrahim Karaman Part VI

Magnesium Technology 2019: Fundamentals, Mechanical Behavior, Twinning, Plasticity, Texture and Fatigue II

Recent Progress in Development and Applications of Mg Alloy Thermodynamic Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 Rainer Schmid-Fetzer

x

Contents

Hardening Effects of Precipitates with Different Shapes on the Twinning in Magnesium Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 Haidong Fan and Jaafar A. El-Awady Isometric Tilt Grain Boundaries and Solute Segregation in a Deformed Mg–Zn–Ca Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 Y. M. Zhu and J. F. Nie Metallography of Mg Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 Norbert Hort, Victor Floss, Sarkis Gavras, Gert Wiese, and Domonkos Tolnai Microstructure and Mechanical Properties of High Shear Material Deposition of Rare Earth Magnesium Alloys WE43 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277 Z. McClelland, D. Z. Avery, M. B. Williams, C. J. T. Mason, O. G. Rivera, C. Leah, P. G. Allison, J. B. Jordon, R. L. Martens, and N. Hardwick Modeling the 3D Plastic Anisotropy of a Magnesium Alloy Processed Using Severe Plastic Deformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 J. S. Herrington, Y. Madi, J. Besson, and A. A. Benzerga Multiaxial Cyclic Response of Low Temperature Closed-Die Forged AZ31B Mg Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 D. Toscano, S. K. Shaha, B. Behravesh, H. Jahed, and B. Williams Thermo-mechanical Processing of EZK Alloys in a Synchrotron Radiation Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 D. Tolnai, M.-A. Dupont, S. Gavras, K. Mathis, K. Horvath, A. Stark, and N. Schell The Effect of the Orientation of Second-Order Pyramidal Dislocations on Plastic Flow in Magnesium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 Kinshuk Srivastava and Jaafar A. El-Awady Part VII

Magnesium Technology 2019: Poster Session

Forging of Mg–3Sn–2Ca–0.4Al Alloy Assisted by Its Processing Map and Validation Through Analytical Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313 K. P. Rao, K. Suresh, Y. V. R. K. Prasad, C. Dharmendra, and N. Hort Development of Manufacturing Processes for Magnesium Sheet . . . . . . . . . . . . . . 319 A. Javaid and F. Czerwinski Part VIII

TMS-DGM Symposium on Lightweight Metals: Magnesium

Incorporating an ICME Approach into Die-Cast Magnesium Alloy Component Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 J. P. Weiler Influences of SiC Particle Additions on the Grain Refinement of Mg–Zn Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Yuanding Huang, Jiang Gu, Sihang You, Karl Ulrich Kainer, and Norbert Hort Development, Characterization, Mechanical and Corrosion Behaviour Investigation of Multi-direction Forged Mg–Zn Alloy . . . . . . . . . . . . . . . . . . . . . . 339 Gajanan Anne, S. Ramesh, Goutham Kumar, Sandeep Sahu, M. R. Ramesh, H. Shivananda Nayaka, and Shashibhushan Arya Electrochemical Behaviour of ECAP-Processed AM Series Magnesium Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 K. R. Gopi and H. Shivananda Nayaka

Contents

xi

Effect of Split Sleeve Cold Expansion on the Residual Stress, Texture and Fatigue Life of Rolled AZ31B Magnesium Alloy . . . . . . . . . . . . . . . . . . . . . . 353 S. Faghih, S. K. Shaha, S. B. Behravesh, and H. Jahed A Theory for Designing Ductile Materials with Anisotropy . . . . . . . . . . . . . . . . . . 359 A. A. Benzerga Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363 Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

About the Editors

Lead Editor Vineet V. Joshi is a Materials Scientist in the Energy and Environment Directorate at the Pacific Northwest National Laboratory in Richland, Washington, USA. He holds an M.S. in Materials Science and Engineering from Alfred University in New York, which he earned in 2010. He is currently pursuing his Ph.D. in Materials Science and Engineering from Washington State University. He currently leads several projects at the laboratory related to development of lightweight structural materials and processing of metallic nuclear fuels. He has made numerous original and important contributions through his research related to lightweight structural materials and nuclear material processing, has published more than 50 articles, has delivered as many talks at international meetings or conferences, and has five patent applications. His primary forte is correlating fundamental processing–microstructure–property–performance relationships in metallic and ceramic materials. Over the years, he helped develop a process to make high-strength titanium alloy, the results of which were published in Nature Communications. He and his team have made significant contributions in developing Shear Assisted Processing and Extrusion (ShAPETM) technique to form magnesium and other lightweight structures. Earlier, he was involved in the development of a new reactive air brazing technique, which enabled joining of mixed ionic and electronic conducting ceramics to metals in air. He serves as the Chair of the Magnesium Committee of The Minerals, Metals & Materials Society (TMS), USA. He served as the guest editor for a special issue on “Corrosion of Magnesium in Multimaterial Systems” for JOM and “Oxidation of Metals” for the journal Metals. He is the recipient of several awards at the laboratory. He has been a Member of TMS since his undergraduate days in India.

xiii

xiv

About the Editors

Magnesium Technology 2019 Editors J. Brian Jordon is an Associate Professor in the Department of Mechanical Engineering at The University of Alabama. He received his Ph.D. from Mississippi State University in 2008. His research focuses on understanding the influence of microstructure on mechanical behavior in order to model materials and structures for superior performance. His interests include fatigue and fracture, process–structure–property relationships, constitutive modeling of plasticity and damage, simulation modeling of welding and joining, and solid-state additive manufacturing. He has published over 50 refereed journal articles and over 25 conference proceedings papers, and has given over 25 invited seminars in these and related areas. His research has been supported by the Department of Energy, the Department of Defense, the state of Alabama, and private industry. Professionally, he serves as Vice Chair of the Magnesium Committee of The Minerals, Metals, & Materials Society (TMS). In addition, he currently serves on the editorial board of the Materials and Manufacturing Processes journal. In 2014, he was a recipient of the TMS Young Leaders Professional Development Award. More recently, he was a 2017 finalist for The University of Alabama President’s Faculty Research Award. Prior to coming to the University of Alabama, he was an Interim Associate Director and an Assistant Research Professor at the Center for Advanced Vehicular Systems at Mississippi State University. Dmytro Orlov is the Professor and Head of the Division of Materials Engineering at the Faculty of Engineering (LTH) in Lund University, Lund, Sweden. He is the Past Chair of a Magnesium Committee at The Minerals, Metals & Materials Society (TMS), USA. After graduating from Donetsk National Technical University and following tenure as a Research Scientist at Donetsk Institute for Physics and Engineering of the National Academy of Sciences, Ukraine, he spent 10+ years at Postdoctoral and Senior Research Scientist positions in world-renowned laboratories at Osaka, Kyoto, and Ritsumeikan universities in Japan; Monash University in Australia; and University of Nova Gorica in Slovenia. In the latter university, he also received habilitation. To date, his track record includes 20+ research projects, 7 patents, 70+ research papers and books, and approximately as many lectures at international meetings among which 20+ were invited. Dr. Orlov’s background is in the engineering of thermomechanical processing technologies for metallic materials fabrication with a core expertise in the design of deformation processing-based techniques. To date, his contributions to the field include the development of twist extrusion technique for imparting large plastic deformations to materials without changing the net shape of a workpiece, analysis of microstructure and texture evolution in metallic materials under large deformations with strain reversals as well as dependence of crystallographic orientation of surface on degradation phenomena in magnesium. The primary scope of his laboratory within LTH is the engineering of novel hybrid, composite and mono-materials

About the Editors

xv

with hierarchical architectured structures from atomic scale to macro-scale. His present research interests and ongoing research projects are focused on the design of Mg alloys for biomedical and lightweight mobility applications, multiscale architectured structures with topological control of their heterogeneity, and the development of relevant in situ characterization techniques at large-scale facilities. Neale R. Neelameggham is ‘The Guru’ at IND LLC, involved in international technology and management consulting in the field of critical metals and associated chemicals, thiometallurgy, energy technologies, soil biochemical reactor design, lithium-ion battery design, and agricultural uses of coal. He was a Visiting Expert at Beihang University of Aeronautics and Astronautics, Beijing, China, and a plenary speaker at the Light Metal Symposium in South Africa on the topic of low carbon dioxide emission processes for magnesium. He has more than 38 years of expertise in magnesium production and was involved in process development of the startup company NL Magnesium to the present US Magnesium LLC, UT, until 2011. He and Brian Davis authored the ICE-JNME award-winning (2016) article “21st Century Global Anthropogenic Warming Convective Model.” He is presently developing “stored renewable energy in coal” Agricoal™ for greening arid soils and has authored an e-book Eco-stoichiometry of Anthropogenic CO2 That Returns to Earth on a new discovery of quantification of increasing CO2 returns to Earth. He holds 16 patents and patent applications, and has published several technical papers. He has served in the Magnesium Committee of the TMS Light Metals Division (LMD) since its inception in 2000 and chaired it in 2005, and in 2007 he was made a permanent co-organizer for the Magnesium Technology Symposium. He has been a Member of the Reactive Metals Committee, Recycling Committee, and Titanium Committee, and was a Program Committee Representative for LMD. He was the inaugural chair, when in 2008 LMD and the Extraction and Processing Division created the Energy Committee, and he has been a co-editor of the Energy Technology Symposium through the present. He received the LMD Distinguished Service Award in 2010. While he was the Chair of Hydrometallurgy and Electrometallurgy Committee, he initiated the Rare Metal Technology Symposium in 2014. He is the co-editor for the 2019 symposia on Magnesium Technology, Energy Technology, Rare Metal Technology, REWAS 2019, and Solar Cell Silicon.

xvi

About the Editors

TMS-DGM Symposium on Lightweight Metals: A Joint US-European Symposium on Challenges in Light Weighting the Transportation Industry Editors Eric Nyberg is the former Director of Programmes at Brunel University London, for the Brunel Centre for Advanced Solidification Technology (BCAST). He joined BCAST in 2016 to lead the development of research programs with international partners. Prior to this, he worked for 24 years at the Pacific Northwest National Laboratory (PNNL), most recently as chief engineer, materials research and development. The US Department of Energy (DOE) recognized him with the 2016 DOE Vehicle Technologies Office Special Recognition Award. He also holds three US patents, an R&D 100 Award, and has authored or co-authored more than 50 technical publications. He earned his both bachelor’s and master’s degrees in Materials Science and Engineering from Washington State University in Pullman, Washington. A TMS member since 1990, he was awarded the Light Metals Division (LMD) Distinguished Service Award in 2012. He has been engaged in LMD and TMS functional committees, including the Magnesium Committee Chair (Chair, Vice Chair and Secretary), the LMD Council (Vice Chair), and the Program Committee (Magnesium Committee Representative.). He has also edited the Magnesium Technology Symposium proceedings and co-edited Essential Readings in Magnesium Technology (2014). Wim H. Sillekens is the Materials Science Coordinator in the Directorate of Human and Robotic Exploration Programs of the European Space Agency (ESA) and in that capacity oversees ESA’s microgravity research program on materials science using such platforms as the International Space Station. He obtained his Ph.D. from Eindhoven University of Technology, the Netherlands, on a subject relating to metal-forming technology. Since then, he has been engaged in aluminum and magnesium research, among others on hydro-mechanical forming, recycling/refining, (hydrostatic) extrusion, forging, magnesium-based biodegradable implants, and most recently on light metal matrix nanocomposites and grain-refined materials. His professional career includes positions as a post-doc researcher at his alma mater and as a research scientist/project leader at the Netherlands Organization for Applied Scientific Research (TNO). International working experience includes a placement as a research fellow at MEL (now AIST) in Tsukuba, Japan. He has (co)-authored a variety of publications (about 150 entries to date). Other professional activities include an involvement in association activities (among others, as the lead organizer of TMS Magnesium Technology 2011), international conference committees, and as a peer reviewer of research papers and proposals. In 2017, he received the TMS Light Metals Division Distinguished Service Award.

About the Editors

xvii

Jürgen Hirsch is a Senior Consultant for Hydro Aluminium Rolled Products R&D Bonn; Professor at the Institute of Physical Metallurgy & Metal Physics (IMM), Technical University RWTH Aachen; CTO of HoDforming GmbH, Düsseldorf; Lecturer IUL at Technische Universität (TU) Dortmund; Guest Professor at Central South University, Changsha, China; and Remote Researcher at Samara National Research University, Russia. He is a member of the board of the German Society for Material Science (DGM), where he served as president in 2015–2016, and is the head of FA “Aluminium”. He is also a member of the EUMAT EU platform steering committee and a member of the ICAA International Committee, where he served as president from 2004 to 2016. He completed his Dr.-Ing. in Material Science and Engineering at the IMM Institute of Physical Metallurgy and Metal Physics at the RWTH Aachen University in 1985. In 1988, he began work at Alcoa Technical Center (R&D) in Pittsburgh, USA, as a Senior Engineer. In 1991, he took a position as Senior Scientist at VAW Aluminium AG/Bonn (F&E), where he served as the Head of Department for Rolling. This organization changed to Hydro Aluminium Bonn (R&D) in 2002. In 2017, he became a consultant for “Aluminium—Metallurgy, Processing and Application”. He has authored more than 200 technical publications and articles and has edited several books on aluminum. Norbert Hort is the head of the Magnesium Processing Department at the Magnesium Innovation Centre (MagIC) within the Helmholtz-Zentrum Geesthacht, Geesthacht, Germany. Concurrently, he is a Lecturer at Leuphana University of Lüneburg, Germany. He studied Materials Sciences at Clausthal University of Technology (CUT), Germany, where he has been involved in magnesium research since the early 1990s. In 1994, he obtained his university degree in engineering. He obtained his Ph.D. in Materials Sciences in 2002 from CUT. During 1994–1999, he worked as a Researcher at the Institute of Materials Sciences (CUT) and he joined the MagIC in 2000. In MagIC, he is responsible for the development of new creep resistant alloys, biodegradable implant materials, and grain refinement and castability of magnesium alloys. This also covers in situ observations of solidification behavior using synchrotron diffraction and tomography. He is the co-author of more than 190 peer-reviewed journal papers and more than 220 contributions to conference proceedings. In recent years, he was involved in the organization of Magnesium Technology symposia at TMS Annual Meetings (2012–2014, 2018). Additionally, he was a member of the organizing committees of Magnesium Alloys and their Applications (2009, 2012, 2015, 2018). LightMat (2013, 2017), Thermec (2013, 2016, 2018), Euromat 2017, IMA annual meetings (2013, 2014), and of the conference “Light Metal Technologies 2011”. Since 2009, he has been the chairman of the technical committee “Magnesium” of the German Society of Materials (DGM).

Session Chairs

Magnesium Technology Keynote Session Vineet V. Joshi, Pacific Northwest National Laboratory J. Brian Jordon, The University of Alabama Alloy Design and Casting Mark Easton, RMIT University Wim Sillekens, European Space Agency Thermomechanical Processing Norbert Hort, Helmholtz-Zentrum Geesthacht Regine Willumeit Romer, Helmholtz-Zentrum Geesthacht Corrosion and Surface Protection J. Brian Jordon, The University of Alabama Chaitanya Kale, Arizona State University Fundamentals, Mechanical Behavior, Twinning, Plasticity, Texture and Fatigue I Sean Agnew, University of Virginia Petra Maier, Stralsund University of Applied Sciences Fundamentals, Mechanical Behavior, Twinning, Plasticity, Texture and Fatigue II Chamini Mendis, Brunel University London Domonkos Tolnai, Helmholtz-Zentrum Geesthacht

TMS-DGM Symposium on Lightweight Metals: A Joint US-European Symposium on Challenges in Light Weighting the Transportation Industry Aluminum Wim Sillekens, European Space Agency Juergen Hirsch, Hydro Aluminium Rolled Products Magnesium Eric Nyberg Norbert Hort, Helmholtz-Zentrum Geesthacht

xix

Reviewer Pool

Neale R. Neelameggham, IND LLC Chamini Mendis, Brunel University London Raymond Decker, Thixomat Inc. Vineet V. Joshi, Pacific Northwest National Laboratory Rajib Kalsar, Pacific Northwest National Laboratory Sean Agnew, University of Virginia Jishnu Bhattacharya, University of Virginia Victoria Miller, North Carolina State University Maryam Jamalian, Washington State University Kiran Solanki, Arizona State University Dmytro Orlov, Lund University Wim Sillekens, European Space Agency J. Brian Jordon, The University of Alabama Paul Allison, The University of Alabama Rogie Rodriguez, The Boeing Co. Harish Rao, National Energy Technology Laboratory Joao Moraes, Ford Motor Co. Kevin Doherty, Army Research Laboratory Eric Nyberg Jürgen Hirsch, Hydro Aluminium Rolled Products Norbert Hort, Helmholtz-Zentrum Geesthacht

xxi

Part I Magnesium Technology 2019: Keynote Session

Magnesium Alloy Sheet for Transportation Applications Chris Romanowski

Abstract

Wrought magnesium alloy sheet has a long history in the aerospace sector and more recently has become popular for personal electronic applications, but has yet to make the transition to high-volume applications in the transportation sector. While there are clear market opportunities for magnesium sheet in lightweight vehicles, the adoption of the material has been limited by the price, the limited number of suppliers and distribution channels and the silicothermic production process. Although the use of a multi-stand mill would reduce the conversion cost and thus the price of magnesium sheet, the current market volume cannot justify investment in such an expensive high capacity plant. This paper reviews these factors and describes the possibility of using twin roll casting and a novel single stand mill design as an alternate, cost-effective, method to produce low-cost magnesium alloy sheet to promote its implementation in the transportation market. Keywords










Automotive Magnesium Sheet Twin Roll Casting

Introduction Magnesium is the world’s lightest common structural metal having a high specific strength and superior damping capabilities. It is extensively used in castings, but magnesium alloy wrought sheet has only been used in aerospace, personal electronics and some high-end sports cars; it has not been widely utilized in common transportation applications such as mass-market automobiles. While there are some technical challenges to the widescale implementation of C. Romanowski (&) Danieli Fata Hunter, 600 Cranberry Woods Drive, Suite 200, Cranberry Township, PA 16066, USA e-mail: [email protected]

magnesium alloy sheet in automobiles, the main obstacle to its adoption appears to be a number of commercial factors.

Applications of Magnesium Sheet Aerospace Magnesium alloy castings have a long and successful history in aviation including leading and trailing edge flaps, control surfaces, actuators, door frames, wheels, engine gearboxes, power generation components, non-primary structural items, and other components outside of the passenger cabin. However, the recent advent and testing by the FAA of flame-resistant magnesium alloys, heralds a bright future for the metal inside the passenger cabin; applications such as aircraft seating, galley carts and cabin furniture are now foreseen. On the other hand, magnesium alloy sheet has seen very limited application in aerospace, and most of those applications were in the post-WWII years. Examples of such applications were the Convair B-36 Peacemaker that incorporated 4 tons of magnesium alloy to reduce its weight by 860 kg, thereby extending its range by 300 km and the Convair XC-99 where much of the structure and most of the skin were constructed from magnesium alloys. Magnesium alloy sheet was also used in a number of space vehicles such as the Vanguard, Jupiter, Titan 1, Polaris, Thor-able Star and the Atlas Agena rockets [1].

Personal Electronics Apart from a brief period in the 1950s when some luggage incorporated magnesium sheet (“Ultralite” Samsonite Luggage), magnesium sheet was largely absent from consumer applications until the growth of the “3C” (computers, cell phones and camera) personal electronics market (Fig. 1). The strength, rigidity, electromagnetic shielding and thermal

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_1

3

4

conductivity made magnesium the metal of choice for these applications. At first, most of these devices were assembled from die cast components, but some of the parts were thicker than desired due to the minimum wall thickness that can be achieved with die casting. The thin-wall die castings also had the disadvantage that they had inherently large runner and gate losses and high finishing costs. As the electronic manufacturers competed to make even thinner and lighter devices, they needed to incorporate wrought magnesium sheet into their designs (Fig. 1). POSCO could see the market opportunity in this trend and in 2004, working with their R&D division (RIST), POSCO launched a project to produce magnesium alloy sheet based on the magnesium twin roll casting technology originally demonstrated by Dow and Hunter in 1982 [3]. As a result of this program, POSCO completed the construction of an integrated Mg sheet plant in 2007 based on a Danieli Fata Hunter twin roll caster. This plant is in regular production making high quality, but relatively narrow, magnesium alloy sheet at a cost-effective price and is one of the world’s main suppliers of magnesium sheet for the personal electronics industry [4]. In 2010, after the success of the initial facility, POSCO invested in a much wider Danieli Fata Hunter twin roll caster, so they could explore the possibility of expanding their technology into the transportation sector. Unlike the

C. Romanowski

initial narrow facility, the coils from the wider caster needed to be shipped off-site for rolling.

Transportation The use of magnesium alloys as die castings in automobiles and light trucks is now commonplace, being used in automotive components such as steering wheels, steering column parts, instrument panel crossbeams, pedal brackets, seats, gearboxes, air intake systems, radiator supports, tailgates, and engine cradles. When used in applications such as instrument panel crossbeams, door and tailgate inners, these castings not only reduce weight when compared to steel sheet metal construction, they often offset the material cost penalty of magnesium by virtue of part consolidation. Despite some early applications of magnesium sheet in automotive bodies (Fig. 2) and the extensive use of magnesium sheet in the body of the International Metro-Lite delivery truck in the late 1950s, the use of magnesium sheet in current production automobiles is rare; the best-known examples being the roof of the latest Porsche GT3 and the panel that separates the interior from the trunk in the Renault Samsung SM7. In both cases, the sheet for these applications is sourced from the wide twin roll caster at the POSCO plant in South Korea.

Fig. 1 a Relative market share of personal electronic devices [2]. b Example of mobile phone component made from AZ61 sheet

Fig. 2 a 1952 Aero Essex magnesium bodied Allard with a body constructed from 1.3–2.0 mm thick magnesium alloy sheet that weighed only 64 kg. b 1958 Lister “Knobbly” Jaguar with magnesium alloy body. A limited run of 10 replica racers is currently being built by Jaguar, all featuring a handmade magnesium alloy body built to the

exact same specification as the original. The castings for the engine sump, clutch bellhousing, and differential housing are also being reproduced in magnesium, resulting in a finished vehicle weight of just 841 kg

Magnesium Alloy Sheet for Transportation Applications

Fig. 3 Comparison of mass savings in automotive applications for advanced materials vs. mild steel in structural panels for equivalent bending stiffness and bending strength [5]

With the notable exception of the panel in the SM7, it appears that very little has changed for automotive magnesium alloy sheet since the 1950s and that its use is still constrained to high-performance luxury automobiles. At first, this may seem surprising as magnesium alloy sheet offers vehicle lightweighting benefits that are only surpassed by fiber-reinforced plastics (Fig. 3). The remainder of this paper discusses the drivers for increased use of magnesium alloy sheet and the factors that are restraining its growth.

Market Drivers for Magnesium Sheet in Transportation Applications Expanding the demand for magnesium sheet to large tonnages requires its utilization in transportation applications such as automobiles, commercial vehicles and railway rolling stock. The highest volume and most promising market for magnesium sheet is the automotive industry, where even the use of small quantities of magnesium sheet per vehicle can rapidly amount to a significant annual consumption—for example, if each of the 591,864 Ford F150 pickup trucks sold in 2017 had incorporated just 10 kg of magnesium sheet, that would have totaled 5,919 tonnes of sheet in the vehicles (much more in coil form before stamping losses). To be successful in penetrating this highly cost-sensitive market, wrought magnesium sheet must show some clear advantages over aluminum. If magnesium is compared purely on its mechanical properties per unit weight (Fig. 3), it appears that it can only be competitive when its cost is close to that of aluminum sheet [6]. Such a comparison, however, ignores the strong regulatory mandates in many countries for improving automobile and light truck fuel economy. At the time of writing this paper, the NPRM issued by the US Office of Management and Budget for model years 2021–2026 GHG & MPG standards is proposing to freeze

5

the target at current 2020 requirements of 37 mpg (6.36 L/100 km) or 240 g CO2/mile (150 g CO2/km). Should this proposal be adopted, it may change the momentum to vehicle lightweighting in the USA, but Brazil, Canada, China, the European Union, India, Japan, Mexico, Saudi Arabia, and South Korea are maintaining their established fuel economy or GHG emission standards for light-duty vehicles [7]. Prior to the NPRM proposal, these countries with the USA were among the top 15 vehicle markets worldwide accounting for nearly 80% of new light-duty vehicles sold. Other large markets, such as Australia, Thailand, and Vietnam, are now in the process of developing similar standards. In Europe, failure by automobile manufacturers to meet European Union fleet average standards has required the manufacturers to pay an excess emissions premium for every car registered; this premium has progressively increased, and by 2021, the cost will be €95/g of exceedance. Although there are credits that can be applied for eco-innovations such as LED lights, or efficient alternators, and super-credits given for selling battery electric cars and hybrids, with the current trend to cross-overs and SUVs and the recent issues with diesel engine emissions, many of the European manufacturers will fail to meet this requirement and are facing fines that may exceed €1 billion per year. Under current EU proposals, the emissions limits will be further reduced by 15% in 2025 and 30% by 2030. The best way to achieve these fuel economy goals with a conventional powertrain, while maintaining the vehicle size, crashworthiness and performance is to reduce the vehicle weight; typically every 10% reduction in weight results in approximately 6% improvement in fuel economy. The high penalties in Europe mean that the automobile manufacturers paying the fines are willing to spend €3 to €5 for every kilogram of weight saving. If it is assumed that the vehicle is already of an aluminum-intensive construction, and the slightly higher forming costs for magnesium are ignored, this means the Europeans would pay approximately €7 per Kg for magnesium sheet. This means that there are many opportunities for magnesium sheet to substitute aluminum and steel in areas such as hood, trunk and roof inner reinforcements, interior bulkheads, door inners, and seat components. Magnesium sheet is particularly attractive in applications where it reduces mass over the front wheels and in areas near the roof of the vehicle. In the USA, even if the CAFE standards are frozen at 2020 levels, there is still an incentive to increase fuel economy by saving weight. An analysis by USCAR (United States Council for Automotive Research) has shown that with gasoline at $3/gallon (€0.69/L), magnesium sheet would need to sell at less than US$3/lbs (US$6.6/Kg) to have a benefit to the owner of the car.

6

C. Romanowski

Fig. 4 a Evolution of the global electric vehicle stock [8]. b Sales of plug-in battery electric vehicles in China [9]

Regardless of the possible freezing of US CAFE standards at 2020 levels, most of the major carmakers are now globally standardizing their passenger car and small cross-over platforms. It can thus be expected that designs and technologies developed in response to the European, Chinese and Japanese regulatory environments will permeate vehicles sold in the US market, even in the absence of similar domestic fuel economy mandates. While the above discussion has focused on fossil fuel-powered vehicles, there is also a growing trend towards electric vehicles, particularly in countries such as China (Fig. 4). The trend to electric vehicles is expected to increase as a number of countries have announced a ban on the sale of new gasoline and diesel passenger cars in coming years (e.g., Norway by 2025; Germany, India, Ireland, Israel and the Netherlands by 2030; and the UK, France and Taiwan by 2040), with some cities and regions enforcing bans even sooner. The benefits of reducing vehicle weight are equally applicable to electric vehicles, the 10% vehicle weight reduction resulting in a 6% increase in vehicle efficiency for an equivalent power to weight ratio is independent of the energy source; it applies to all vehicle powertrain types. A 2018 analysis by Peterson [10] demonstrated that if the weight of an electric vehicle can be reduced by 91 kg (5%) by the use of magnesium, the smaller battery pack necessary for equivalent range would be US$348 (€303) cheaper, and when combined with the associated mass compounding could save as much as US$1,728 (€1,503) per vehicle. Alternatively, the reduced weight of the vehicle would allow a greater range without increasing battery size. The Peterson analysis showed that with current battery pack pricing, it is less expensive to increase EV range by reducing vehicle weight than it is to add battery capacity.

the technology continues to improve, particularly for the target applications such as closure reinforcements where corrosion and surface finish are not critical and significant weight savings can be achieved, even in vehicles that already have an aluminum-intensive structure. These advancements include GM’s quick plastic forming where the magnesium sheets are quickly heated to 450 °C (842 °F), then clamped around the periphery in a die before air pressure is used to form the sheet into a panel [11]. Despite the continuously improving alloy and process technology and significant financial incentives to utilize magnesium alloy sheet as a tool for lightweighting both internal combustion and electric vehicles, magnesium alloy sheet has failed to garner traction in the market. There is now the real possibility that the automobile manufacturers may bypass it altogether and move onto the use of fiber-reinforced plastics, even though this would require a shift in manufacturing technologies and the material has some problems with end-of-life recyclability. The following sections list some of the major reasons for the failure of magnesium alloy sheet in automotive applications.

Price The current US market price for wide magnesium alloy sheet suitable for transportation applications is around US$37 to US$40 per Kg. This is significantly higher than the US$15 per Kg, EXW, China, that some Chinese suppliers are offering narrow width sheet for electronic applications. Unfortunately, at either of these prices, the use of magnesium sheet is not cost-effective in mass-market passenger cars and light trucks.

Factors Restricting the Growth of Magnesium Sheet in Transportation Applications

Availability

While there are well-known technical challenges associated with the utilization of magnesium sheet in automotive applications, many of these have largely been overcome and

While there are a number of Chinese suppliers offering narrow width sheet, there are only a very limited number of suppliers that can offer the sheet width necessary for

Magnesium Alloy Sheet for Transportation Applications Fig. 5 Example of typical closure panel inner reinforcement showing the amount of sheet lost during the initial forming and stamping step [12]

7

market develops, the magnesium sheet suppliers need to find an economically attractive way of recycling the scrap. An alternate solution would be to partner with a Tier 1 supplier that would locate their facility near the magnesium sheet plant and the Tier 1 supplier would ship finished parts to the automobile plant.

Compatibility automotive and light truck applications. The two best known are Magnesium Elektron and POSCO. It should be noted, however, that due to rolling and finishing their wide cast strip off-site, the price of the POSCO sheet does not fully reflect the cost savings gained by twin roll casting. Most major automobile manufacturers expect a reliable source of supply for the materials they (or their Tier 1 suppliers) need to manufacture their vehicles, and thus with the current limited supply options, it is unlikely that a major automobile manufacturer would risk incorporating magnesium sheet into the design of any high-volume vehicle, even if the price was competitive.

Distribution The aluminum industry has a long history of shipping coils internationally, but even they have experienced problems with shipping delays causing supply chain problems for the automakers, together with the quality of the aluminum sheet deteriorating during extended transit. For this reason, most car makers prefer to buy their aluminum sheet within the same geographic region. With the present very restricted number of wide magnesium sheet suppliers, this is currently not an option for most automobile plants.

Scrap The initial target application for magnesium sheet is expected to be closure panel inner reinforcements. In the majority of these applications, a significant volume of the incoming metal is lost as scrap during the initial forming/stamping process (Fig. 5). This not only increases the cost per unit mass incorporated into the vehicle, it also raises the issue of what to do with the scrap. The aluminum sheet suppliers enter agreements to take back the scrap from the automobile manufacturers at a pre-agreed price. The scrap is typically returned to the same mill that supplied the original coils and the aluminum company recycles it back into aluminum sheet. With the present very limited number of magnesium sheet suppliers and long transport distances, this type of arrangement would not be practical. Until the

In order for the automakers to incorporate magnesium sheet into their automobiles, it must be compatible with their current production. One aspect of this compatibility is the coating applied to the sheet at the magnesium sheet plant. The coating not only needs to protect the sheet during storage and transportation, it must also survive the forming process (particularly if the sheet is formed by a Tier 1 supplier and then shipped to the automotive plant) and it must be compatible with any joining technologies to incorporate the magnesium panels into the automobile structure (whether adhesive, or welding, or mechanical fastening). Finally, the coating applied to the magnesium must be compatible with the corrosion protection treatments and finishes that are applied to the final automobile body. Fortunately, the technology for this type of coating has already been developed [13] and has been shown to provide good protection, while being compatible with current joining technologies and without contamination of the chemical treatment baths in the automobile plant.

Sustainability and Lifecycle Assessment While wrought magnesium sheet has a significant environmental benefit in reducing fuel consumption and is readily recyclable, its lifecycle assessment can be negatively impacted by the reduction process used to produce the original ingot. Currently, the majority of Chinese magnesium ingot is produced by the silicothermic reduction of dolomite. This process is commonly known as the “Pidgeon process” and has a very low capital cost, but also has a very high carbon footprint [14–16]. The carbon emissions vary with the age of the plant and the exact choice of fuel (coal, coke oven gas, natural gas, etc.). An exception is a variation of the Pidgeon process based on Bolzano technology operated by RIMA in Brazil where they use charcoal from sustainable forestry, in combination with hydroelectric power, to significantly reduce their carbon footprint to approximately half that of fossil fuel electrolytic production and 60% lower than the Pidgeon process [17–19]. The high carbon footprint of the Pidgeon process profoundly impacts the overall lifecycle assessment of magnesium alloys in automotive applications; some analyses

8

showing that it can take 250,000 km for a magnesium casting made from fossil fuel-powered electrolytic ingot to show a net CO2 savings, while a Pidgeon process casting did not save CO2 within the lifetime of the car [20, 21]. This has discouraged some of the European carmakers from considering the incorporation of significant quantities of magnesium sheet in the structure of their cars. The sustainability requirements of the European carmakers will, however, be met by the new Qinghai smelter [22]. This smelter is located near the Qinghai Salt Lake near Golmud in China. It will use solar and hydropower to produce eco-friendly magnesium from magnesium chloride that is currently a by-product from the nearby Potash plant. The smelter is initially expected to produce 100,000 tonnes per year [23], then progressively expanding to 450,000 tonnes per year [24]. It is estimated that this plant will be able to produce magnesium ingot at US$1,670 per tonne [25].

C. Romanowski

Developing the Magnesium Sheet Market by Reducing Cost There are two main factors that influence the price of magnesium alloy sheet: the first is the cost of magnesium ingot (which can vary geographically due to import tariffs), and the second is the cost to convert the metal from ingot to magnesium alloy sheet suitable for automotive applications.

Ingot Price The price of magnesium ingot on the Chinese market in 2018 has been reasonably stable at around 15,000 Renminbi per tonne (US$2.3/kg) and it has historically maintained a price close to that of aluminum [26]. This relationship is reflected by the pricing of magnesium ingot in the European market, but due to long-standing import tariffs, the US price for magnesium has always been higher.

Promotion One of the lessons learnt by the aluminum industry, when first penetrating the automotive sheet market, is that it had to overcome decades of accumulated manufacturing experience the automakers had with steel. Not only was there a vast existing knowledge base, steel companies worked closely with automakers, embedding their personnel to constantly provide technical support and evolve their products to meet the needs of the automakers. There was also the public perception of steel being stronger and safer than aluminum. Working within the structure of the Aluminum Association, the aluminum industry cooperated to establish a group to reverse public perception and actively engage the automakers (www.drivealuminum.org). This program has already shown results, the most notable of which is the success of the new, aluminum-bodied, Ford F150 and the incorporation aluminum sheet in a growing number of vehicles.

Conversion Cost Until the advent of twin roll casting, almost all wrought magnesium alloy sheet was produced by DC casting where the metal is cast into slabs that are about 300 mm thick. The slabs were then homogenized prior to the hot and warm rolling processes. Magnesium alloys have a hexagonal close-packed (HCP) crystal structure that has very limited plasticity at room temperature and thus must be rolled at elevated temperatures to prevent cracking. A reversing hot mill is initially used to roll the slabs in a series of passes to a “reroll” coil that can be transferred to a warm rolling mill (Fig. 6). Due to the low heat capacity of magnesium and the heat losses on the hot mill run-out tables, the maximum slab size that could be rolled is limited and traditionally resulted in very small coils. These small coils have significant head and tail losses and their handling time is a significant fraction

Fig. 6 Schematic showing the process flow for the conventional production of magnesium sheet from DC casting to a “reroll” coil at the exit of the reversing hot mill that is subsequently warm rolled to finished gauge sheet

Magnesium Alloy Sheet for Transportation Applications

9

Fig. 7 Schematic showing the arrangement of a typical aluminum sheet plant hot rolling mill [27]

of the rolling time. The warm mill further rolls the reroll coil in a series of passes with inter-pass reheating to the required final thickness. This is a labor and energy-intensive method of production with high scrap losses and is a major factor in the traditionally high cost of magnesium sheet. In recent years, to reduce conversion cost, some companies have bypassed the reversing hot mill by extruding the magnesium alloy into a thin bar [2], but this technology is limited to only producing a narrow sheet that is not wide enough for most automotive applications. Although the plasticity and lower heat capacity of magnesium exacerbates the heat loss issue, the production concerns are similar to those of the aluminum industry. With aluminum, this has been addressed by using a combination of a reversing hot mill, with a multi-stand tandem mill (Fig. 7) to replace the first step shown in Fig. 6. The reversing mill rolls to a thickness where the metal starts losing heat rapidly and this plate is then fed into the tandem mill which rolls it down to a reroll coil in one pass through multiple stands. Once at “reroll” gauge, the aluminum only requires very few passes to reduce it to a sheet thickness suitable for automotive applications. A modern aluminum plant using this type of arrangement will very efficiently produce 2.5 m OD coils that have consistent properties throughout the entire length and excellent surface quality. Figure 7 shows a 3-stand tandem mill, but up to 5-stands can be used to give an exit gauge of 1.5–2 mm. At this exit gauge, a magnesium alloy coil may already be at the appropriate thickness for some transportation applications and would only require one warm rolling pass for most other applications. While the equipment arrangement shown in Fig. 7 is theoretically a very efficient way to manufacture magnesium sheet, it is commercially impractical because the cost for the plant would be very high (US$ 9 digits) and the capacity would typically be more than 500,000 tonnes per year, which is more than an order of magnitude higher than the current global magnesium sheet market.

Innovative Technology for the Production of Low-cost Magnesium Sheet The above discussion has shown that while magnesium ingot is competitively priced with aluminum, the reason for the high price of magnesium sheet is the conversion cost. In an effort to expand the magnesium sheet market, Danieli FATA Hunter has developed an innovative process

for converting ingot into high-quality, low-cost magnesium alloy sheet. The first step in the process is twin roll casting the alloy into a wide, large diameter, coil suitable for warm rolling as already industrially proven by POSCO [28]. The next step in the process is an innovative isothermal rolling technology developed in collaboration with the Oak Ridge National Laboratories and Magnesium Elektron. This technology was tested on a pilot-scale mill that combined the ability to maintain the entry and exit material at a constant temperature, with thermostatically controlled, internally heated work rolls that had the ability to maintain a constant surface temperature up to 300 °C and also an independent work roll drive system capable of producing a maximum asymmetry during rolling of 3:1 [29]. This mill demonstrated the feasibility of an industrial rolling process that not only economically reduces cast coils to the gauges required by consumer products, but also can modify the microstructure to improve the formability of the rolled sheet, while maintaining complete uniformity of properties throughout the length of the coil and a quality surface that requires minimal treatment after rolling. The preliminary results from this mill showed that when AZ31B magnesium alloy was rolled with highly asymmetric roll speeds, the strong basal texture normally found in Mg alloys is reduced and the modified microstructure improves low-temperature formability, while maintaining good surface quality [30]. This mill received a “R&D 100 Award” in 2012.

Proposed Integrated Plant for the Production of Low-cost Magnesium Transportation Sheet Based on the casting and rolling technology described above, Danieli Fata Hunter has pre-engineered a complete plant concept that has a low capital expenditure and low operating cost and is able to produce 30,000 tpa of low-cost, good-quality magnesium alloy sheet. The plant incorporates all the equipment required to convert magnesium ingot into coils of magnesium alloy sheet at final gauge that has been leveled, cleaned and given a temporary coating to prevent corrosion during in-plant storage, handling, transportation and forming. The plant comprises the following major pieces of equipment: Twin Roll Casters. In the first phase of the project, it is foreseen that only one casting machine will be installed with

10

a capacity of approximately 10,000 tpa. As the market for magnesium sheet begins to develop, it is expected that a second and a third casting machine will be installed, progressively raising the plant capacity to 30,000 tpa. Each casting line will include a dedicated furnace group consisting of two melting furnaces complete with de-stackers and drying ovens feeding a holding furnace from which the metal is pumped to the caster. If the plant is located near a magnesium smelter, the furnace arrangement could be modified to receive molten metal from the smelter. The casting lines would produce 2,000 mm wide by 2,500 mm outer diameter coils (16.6 tons) of cast strip to minimize downstream head and tail losses and improve the overall plant efficiency. Coil Reheating Oven. The coil reheating oven pre-heats the as-cast coils for delivery to the warm rolling mill. The coil reheat oven is a “pusher” style furnace where loading a cold coil on a pallet at the input end causes a hot coil to exit from the output end. The oven has upending units at oven entry and exit with the capacity to accommodate six as-cast

C. Romanowski

2,500 mm OD  2,000 mm wide coils in an eye-to-sky orientation. The pallets are automatically recirculated. Isothermal Rolling Mill. After being heated to the rolling temperature in the coil reheating oven, the magnesium coils are transferred to the 4-Hi asymmetric warm rolling mill where they are subjected to a series of rolling passes to not only reduce the thickness, but also modify and improve the microstructure of the metal making it suitable for forming into automotive parts, etc. The mill consists of a reversing 4-Hi mill stand with asymmetric drives and heated work rolls, a hot coiler positioned on either side for heating and maintaining the desired temperature of the magnesium coils and a combination loading and unloading station. This mill has been described in detail in previous publications [29]. Finishing Line. The finishing line cleans and chemically treats the as-rolled coils to prepare them for forming and subsequent painting. This line incorporates an edge trimmer and leveling section to remove any edge cracking and shape errors after rolling. The process flow of this line can be seen in Fig. 8.

Fig. 8 Process flow chart for the finishing line

Fig. 9 Layout of a 30,000 tpa integrated magnesium alloy sheet plant showing two of the three casting lines

Magnesium Alloy Sheet for Transportation Applications

Plant Layout. Figure 9 below shows a layout of the proposed plant with the auxiliary equipment such as slitting, cut-to-length and packing lines. It is expected that this type of plant could significantly reduce the conversion cost of producing magnesium sheet. It has been estimated that this conversion cost would be approximately two to three times that of aluminum sheet (currently *US$1.28/kg). This means that with an ingot price of *US$2/kg, the finished magnesium sheet could ship from this plant at *US$4.50 to US$6.00/kg.

Summary • The use of magnesium alloy sheet to significantly reduce vehicle weight has been known since the 1950s. • Most of the technological obstacles preventing the greater use of the sheet have been overcome. • The use of twin roll casting to produce low-cost, high-quality magnesium alloy narrow sheet for electronic applications is industrially proven. • In many countries, government fuel economy mandates for fossil-fueled cars incentivize the automakers to save weight and make magnesium alloy sheet attractive if the cost of wide sheet can be reduced. • Vehicle lightweighting is also attractive for electric vehicles to save battery cost. • Large-scale adoption of magnesium sheet is limited by a number of commercial and sustainability factors. • Most obstacles to the use of magnesium sheet in automobiles can be overcome by a few strategically located integrated magnesium alloy sheet plants based on twin roll casting and isothermal asymmetric rolling. Ideally, such plants would be located adjacent to electrolytic smelters using renewable power.

References 1. “Magnesium Technology,” Horst Friedrich and Barry Mordike, Springer-Verlag Berlin Heidelberg 2005, Library of Congress Control Number, 2005931991. 2. “Expansion of thermally rolled magnesium coils into mobile devices,” Kazumasa Yamazaki, Masahiko Sato, Shuji Higuchi and Takeki Matsumura, International Magnesium Association’s 71st Annual Conference, June 1–3, 2014, Munich, Germany. 3. “Innovations in the Process Technology for Manufacturing Magnesium Alloy Sheet”, E. Romano, R. Passoni, C. Romanowski, The Minerals, Metals & Materials Society (TMS) 144th Annual Meeting, Orlando, Florida, March 15–19, 2015.

11 4. “Mg Coil Production via Strip Casting and Coil Rolling Technologies,” I.H. Jung et al., Magnesium Technology 2007, The Minerals, Metals & Materials Society (TMS), pp. 85–88. 5. Private communication from Alan Luo, Ohio State University, with calculations based on “Materials Selection in Mechanical Design,” M.F. Ashby, Pergamon Press, Oxford, 1992. 6. “Developments in the Zuliani Process for Gossan Resources’ Magnesium Project”, Douglas Zuliani and Douglas Reeson, Proceedings of the 9th International Conference on Magnesium Alloy and Their Applications, July 8–12, 2012, Vancouver, BC, Canada. 7. “Impacts of World-Class Vehicle Efficiency and Emissions Regulations in Select G20 Countries,” Josh Miller, Li Du and Drew Kodjak, January 2017 Briefing Paper by the International Council on Clean Transportation (ICCT). 8. “Global EV Outlook 2018, Towards Cross Modal Electrification,” International Energy Agency (IEA), Paris France. 9. EV-Volumes.com, Trollhättan, Sweden. 10. “An Overview of Current and Emerging Industry Opportunities For Magnesium Applications,” Gregory E. Peterson, International Magnesium Association, 75th Annual Conference, 16–18 May 2018, New Orleans, USA. 11. “Quick Plastic Forming of a Decklid Inner Panel with Commercial AZ31 Magnesium Sheet,” R. Verma and J. Carter,. SAE Technical Paper 2006-01-0525, 2006. 12. DieTech NA, Roseville Michigan 48066, USA, A1LL Hood Inner. 13. “Surface Treatment of Magnesium Sheet Components in Transportation Industry,” Ilya Ostrovsky, International Magnesium Association’s 71st Annual Conference, June 1–3, 2014, Munich, Germany. 14. “Global warming impact of the magnesium produced in China using the Pidgeon process,” S. Ramakrishnan, P. Koltun, Resources, Conservation and Recycling, Volume 42, Issue 1, August 2004, pp. 49–64. 15. “Challenges for Implementation of Magnesium into more Applications,” Karl Kainer, Magnesium Technology 2016, The Minerals, Metals & Materials Society (TMS). 16. “Review of Magnesium for 75th Anniversary of the International Magnesium Association, a Personal and Subjective Discussion,” Robert E. Brown, International Magnesium Association, 75th Annual Conference, 16–18 May 2018, New Orleans, USA. 17. “Magnesium: current and alternative production routes,” Winny Wulandari, Geoffrey Brooks, Muhammad Rhamdhani and Brian Monaghan, 2010, Faculty of Engineering - Papers (Archive), University of Wollongong. 18. “LCA of Magnesium Production Technological Overview and Worldwide Estimation of Environmental Burdens,” Cherubini, F., Raugei, M. & Ulgiati, S., 2008, Resources Conservation & Recycling, vol. 52, no. 8–9, pp. 1093–1100. 19. “Assessing The Environmental Impact of Metal Production Processes,”Norgate, T.E. & Rankin, W.J. (2007), Journal of Cleaner Production, vol. 15, no. 8–9, pp. 838–848. 20. “Life Cycle Assessment of Magnesium Components in Vehicle Construction,” Simone Ehrenberger, May 2013, German Aerospace Centre e.V., Institute of Vehicle Concepts, D-70569 Stuttgart, Germany. 21. “Solutions for Next Generation Automotive Lightweight Concepts Based on Material Selection and Functional Integration,” H.E. Friedrich, E. Beeh & C.S. Roider, Magnesium Technology 2018, The Minerals, Metals & Materials Society (TMS), pp. 343–348. 22. “Development and prospect of ecological magnesium industry in Qinghai Salt Lake,” Xie Kangmin, Wang Shijun, and Hong-wei

12

23.

24.

25. 26.

27.

C. Romanowski Zhu, International Magnesium Association, 75th Annual Magnesium Conference, 16–18 May 2018, New Orleans, USA. “Qinghai Magnesium Project Overview,” Phillip W. Baker, International Magnesium Association’s 71st Annual Conference, June 1–3, 2014, Munich, Germany. Magontec Limited (Company), Executive Chairman’s address to the 2013 annual general meeting, Australian Sock Exchange Limited, 20 Bridge Street, Sydney, Australia. “The Economics of Magnesium Smelting Technology,” John Grandfield, Light Metal Age, Feb 2016, pp. 60–65. “Global Primary Magnesium Supply and Demand Balance 2013,” Alan Clark, Liping Li, Joe Zhou, John Grandfield and Sijia Sun, International Magnesium Association’s 71st Annual Conference, June 1–3, 2014, Munich, Germany. Hazelett Corporation, Colchester, Vermont 05446 USA.

28. “Wide Strip Casting Technology of Magnesium Alloys,” W. J. Park, J.J. Kim, I.J. Kim, D. Choo, Magnesium Technology 2011, The Minerals, Metals & Materials Society (TMS), pp. 143– 146. 29. “New Reversing Rolling Mill Can Make Magnesium Sheet Commercially Viable”, Roberto Passoni, Chris Romanowski, Enrico Romano, and Pier Michele Cattelino, Light Metal Age, February 2013. 30. “Shear Rolling of Magnesium Sheet for Automotive, Defense and Energy Applications”, G. Muralidharan, T.R. Muth, William H. Peter, T.R. Watkins, and Y. Wang, Oak Ridge National Laboratory, Dave Randman, Bruce Davis, Marytn Alderman, Magnesium Elektron North America, Chris Romanowski, FATA Hunter, ORNL Report TM-2012/541.

Magnesium for Automotive Lightweighting: Status and Challenges Sarah Kleinbaum

Abstract

Cast and wrought magnesium have long been identified as a key pathway to automotive lightweighting and improved energy efficiency. However, adoption in the automotive market remains low. This talk will look at the application of magnesium components into various vehicle subsystems from 2012 through today and identify the technical challenges that currently limit full adoption. Keywords

Magnesium




Lightweight




Automotive




Challenges

Methodology To determine the extent of magnesium usage in automotive components and the change in usage over time, teardown data from 38 midsize vehicles sold in the USA and available in the A2MAC1 database will be examined. The vehicles selected represent all the available teardown data for this size class and location from 2012 through 2016 and span a wide range of sales price points and curb weights within the class. Material type and weight of each component were determined visually by A2MAC1. Required component functionality (strength, stiffness, energy absorption, aesthetics, etc.) will be assessed and technical gaps identified.

Motivation Known Challenges Magnesium is widely recognized as one of the four high potential materials for achieving automotive lightweighting along with aluminum, advanced high strength steel, and polymer composites. With the lowest density of any structural metal, magnesium has twice the specific strength of mild steel. Magnesium also has great castability due to its fluidity that has led to the use of complex thin-walled magnesium die cast components, such as instrument panels, in vehicles. However, magnesium currently accounts for less than 1% of a vehicle’s weight on average due to several technical challenges. To better understand the state of magnesium penetration into the automotive industry and the need for research to increase that penetration, we will conduct an analysis of automotive components made from magnesium from 2012 through today.

S. Kleinbaum (&) Vehicle Technologies Office, Office of Energy Efficiency and Renewable Energy, Department of Energy, 1000 Independence Ave SW, Washington, DC, USA e-mail: [email protected]

Many automotive components are produced from flat sheet and stamped into their final shape. Magnesium has a strong basal texture that evolves during the rolling of ingot into sheet which limits its room temperature formability. Some alloys of magnesium can be stamped at higher temperatures, around 250 °C, due to the addition of rare earth elements but with limited draw depths and often at a higher cost. At this temperature, die lubrication also becomes a challenge as the lubricant often bakes onto the sheet and is difficult to remove prior to painting. For both wrought and cast components, corrosion resistance of magnesium is a concern. Unlike steel and aluminum, which form stable oxide layers, magnesium forms oxides that are porous and ineffective at preventing further corrosion. Paints can be applied to improve corrosion resistance but require a unique pre-treatment that requires the magnesium components to be painted individually rather than as part of a larger sub-assembly. Galvanic corrosion is also an issue when magnesium is part of a multi-material assembly. Coatings or other isolation strategies are necessary to prevent accelerated corrosion.

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_2

13

14

Welding and joining of magnesium to other materials in assemblies also present a challenge. Magnesium’s low melting temperature and immiscibility with steel prevent fusion welding, which is the first choice of the automotive industry. Fasteners such as self-piercing rivets can cause cracking of the magnesium as well as galvanic corrosion. Solid state welding techniques show some promise, but these have not yet achieved widespread adoption by the automotive community.

S. Kleinbaum

Expected Results The challenges outlined above are well-known and the subject of much academic and industrial research. The analysis to be undertaken should help highlight the progress made to overcome these challenges and where further research is needed to increase the use of magnesium to lightweight vehicles.

Magnesium Process and Alloy Development for Applications in the Automotive Industry David Klaumünzer, Jose Victoria Hernandez, Sangbong Yi, Dietmar Letzig, Sang-hyun Kim, Jae Joong Kim, Min Hong Seo, and Kanghwan Ahn

Abstract

While cast magnesium components have been extensively used in the automotive industry, only niche applications are known to exist for magnesium sheets to this date. This is believed to originate from limitations in the material property spectrum covered by existing alloys. By utilising the twin-roll casting technique, novel Mg–Zn–Ca–Zr (ZXK) alloy sheets can be developed with improved formability as a result of a texture weakening effect. More importantly, it can be shown that these new alloy sheets can be processed at industrial scale. Based on the magnesiumspecific design of a Volkswagen Passat decklid, the advantages of these new highly formable sheets in an automotive production environment can be evaluated. Keywords

Magnesium sheets Cost reduction




High formability




Applications




Introduction Despite numerous research activities in the field of magnesium alloy and process development, applications within the automotive industry remain scarce and predictions made in D. Klaumünzer (&) Volkswagen AG, Group Research, Berliner Ring 2, 38440 Wolfsburg, Germany e-mail: [email protected] J. V. Hernandez
S. Yi
D. Letzig Helmholtz-Zentrum Geesthacht, Magnesium Innovation Centre, Max-Planck-Str. 1, 21502 Geesthacht, Germany S. Kim
J. J. Kim POSCO, PosMAF, 2ro-5, Haeryong-myon, 540-856 Suncheon, South Korea M. H. Seo
K. Ahn POSCO, POSCO Global R&D Centre, Songdo-dong 180-1, Incheon, South Korea

the late 1990ies with respect to the average usage of magnesium in future passenger vehicles have not turned into reality [1]. Undoubtedly, magnesium offers a high lightweight saving potential with weight savings typically between 10 and 33% on component level when compared to aluminium. However, limitations in the processability and material performance of existing alloys restrict the overall applicability of this lightweight construction material. Currently, almost all of the automotive applications are limited to die-castings for which magnesium offers the benefit of a better castability when compared to aluminium. Prime examples for cast components are gearbox houses, interior support brackets or steering wheels. For such simple applications, any surface coating to account for the poor corrosion resistance of magnesium is not required, thus leading to no extra costs compared to the aluminium equivalent. Applications of magnesium sheets in wrought form, i.e. as sheets or extruded profiles, have so far not exceeded the prototype or near-prototype level. In this context, the only outer-skin application of magnesium sheets to date is the roof of the Porsche 911 GT3RS with very low production volumes of approximately 5000 cars over life-time [2]. In this application the high lightweight construction potential of magnesium is once again demonstrated, surpassing even that of carbon fibre reinforced plastics [2]. There are various reasons for the scarce use of magnesium sheets for high volume applications in the automotive industry [3, 4]. One of the major hurdles is the limited formability. This has two main implications. Firstly, conventional slab rolling during sheet production needs to be conducted at elevated temperatures, leading to sheet prices at least five to six times higher compared to aluminium. Additionally, any component utilising magnesium sheets also needs to be formed at elevated temperatures, which contributes to a higher complexity in a serial forming process. All of these factors lead to an unfeasible economics when comparing magnesium to aluminium components for the car body.

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_3

15

16

Recent progress made in the production and processing of magnesium sheet saw the adaption of the twin-roll casting technique, thus reducing the number of warm rolling steps and hence bearing a significant cost saving potential. While this technique provides a key component towards enhancing the economic feasibility of magnesium sheets in the automotive field, the range of alloys investigated using twin-roll casting and more importantly implemented at an industrial scale remains limited, thus still offering a very narrow property spectrum which is not competitive to aluminium alloys. This work is intended to contribute towards closing this gap. We will describe systematic research into the twin-roll casting of a Mg–Zn–Ca–Zr (ZXK) alloy, whereby we follow an upscaling approach up to industrial level. We will give a description of the linkage between alloy chemistry, processing and materials properties with particular focus on formability. We will highlight the advantages this and similar alloy systems have particularly in an automotive production context.

D. Klaumünzer et al.

done to assess the castability of this alloy system at semi-industrial and industrial scale. This is the focus of the present work. For this study, we choose an alloy composition of 0.6 wt% Zn and 0.6 wt%. Ca and add an additional amount of Zr below 0.1 wt%. This composition has not been focus of any of the previous twin-roll casting studies and ensures the formation of only low amounts of the ternary phase Mg6Zn3Ca2 which is believed to have detrimental effects on the corrosion resistance of the alloy [11]. To counterbalance the low contribution of Zn towards solid solution strengthening in such a lean alloy, particular emphasis is placed on grain refinement using Zr. Twin-roll casting trials were conducted sequentially on a semi-industrial twin-roll caster at Helmholtz-Zentrum Geesthacht and in an upscaling step using the industrial equipment at POSCO. The sheets were subsequently rolled to final gauge at elevated temperatures and subjected to a heat treatment following the rolling operation.

Alloy Selection and Texture Weakening

Semi-industrial Casting Trials and Material Characterisation

Multiple recent studies focussed on the improvement of formability by altering the strong basal texture usually found in conventional AZ31 alloy sheet [5]. The most prominent example of alloys with weak basal texture are those of the zinc-rare-earth alloy family (ZE) [6]. However, recent work has also pointed out that replacing the rare-earth elements with calcium to form zinc-calcium (ZX) alloys leads to a similar effect [6–10]. At the same time, ZX alloys have received significant attention in the biomedical field due to their excellent biocompatibility and corrosion resistance [11]. While some preliminary, lab-scale studies investigated the twin-roll casting of ZX alloys [8, 9], no work has been

Figure 1 shows the microstructure of the twin-roll cast sheet following hot rolling and a static heat treatment. By adopting this processing route, the sheet attains a homogeneous and fine grained microstructure with no signs of centre-line or reverse segregation, thus demonstrating the beneficial effects of adding a grain refiner and working with a dilute alloy system. The measured grain size lies well below 10 µm in accordance with the targets set by our alloy design concept. As shown in Fig. 2b, a texture measurement of the final sheet reveals, as expected for the ZXK alloy system, a weak basal texture with a splitting of the basal poles towards the transverse direction. Such texture stands in marked contrast

Fig. 1 Microstructure of the twin-roll cast and rolled ZXK-alloy sheet after a static heat treatment. The dilute alloy composition in addition to the grain refiner Zr leads to a homogeneous microstructure (adapted from Ref. [12])

Magnesium Process and Alloy Development for Applications …

to the strong basal texture observed for the conventional alloy AZ31 as exemplified in Fig. 2a. The benefits of such a weak texture on the formability of the sheet can be analysed quantitatively and contrasted to the conventional sheet alloy AZ31. It is clear that a weaker basal texture aids the accommodation of plastic strain in the sheet thickness direction. This effect is particularly predominant at lower temperatures at which plastic glide occurs on basal slip systems only. The propensity of a material to accommodate plastic strain in the sheet thickness direction can be expressed using the r-value. The r-value is determined in a uniaxial tensile test and corresponds to the ratio of strain in the direction of the width of the specimen relative to that in sheet thickness direction. Measuring the r-value for the new ZXK alloy and AZ31 as a function of temperature reveals both the effects of texture intensity and thermal activation of glide systems on the plastic flow behaviour of the two

Fig. 2 Comparison of the textures for the conventional alloy AZ31 and the new alloy ZXK. ZXK shows a weakened basal texture with splitting of the poles 90° to the rolling direction (RD) (adapted from Ref. [12])

Fig. 3 r-value as a function of temperature for the two magnesium sheet alloys AZ31 and ZXK taken from uniaxial test data with the rolling direction (RD) parallel to the tensile axis. A reduction in texture intensity for the ZXK alloy reduces the r-value at low temperatures

17

sheet alloys as shown in Fig. 3. While for AZ31 the strong basal texture results in a high r-value of 2.5 at room temperature, i.e. plastic flow occurs preferentially in the width-direction, a weakening of the basal texture in the ZXK alloy leads to a balancing of the strains in the two directions and an r-value close to unity. Only upon increasing the temperature, the r-value of AZ31 decreases as non-basal slip systems become progressively activated. For ZXK, a subtle increase in r-value with temperature is observed and both alloys reach the same value at 200 °C. Stretch forming is a forming operation that requires a high deformability in the sheet thickness direction, as this is the only direction from which material can be drawn to conserve volume. It is therefore not surprising that the ZXK alloy possesses a significantly improved room-temperature stretch formability compared to AZ31 as shown in Fig. 4. The degree of stretch formability can be expressed as the Erichsen index which is a measure of dome-height during a bulge test with a hemispherical punch of 20 mm in diameter. It is quite interesting to note that when compared to other sheet materials typically used in the automotive industry, the room-temperature stretch formability of the new ZXK alloy lies between that of 6000 and 7000 aluminium alloys as depicted in Fig. 4. Following the fundamental studies into the processability and properties of the ZXK alloy above using the lab-scale twin-roll casting system at Helmholtz-Zentrum Geesthacht, it was decided to perform an investigation into the upscaling of this process using the industrial equipment at POSCO. This involved a scaling of the maximum width from 650 to 2000 mm, opening the option to consider the new sheet alloy system for further investigation in prototype studies. This part of the project was successfully completed in 2017,

Fig. 4 Comparison of the Erichsen index for multiple sheet materials relevant to the automotive industry (Insert: photograph of a bulge test). While the stretch formability of AZ31 is poor, the ZXK alloy sheet shows a performance in-between that of 6000 and 7000 Al alloys. (Figure adapted from Ref. [12])

18

proving that an industrial processing of the ZXK-alloy system at a corresponding scale is indeed possible.

Magnesium Sheet Components The success in the development of new, improved magnesium sheet alloys triggered new interest in assessing the production and performance of magnesium sheet prototype components at Volkswagen with a particular focus to identify and assess the benefits that these new alloy systems entail. In an extension of this study, recent developments by POSCO were included, e.g. the recently established alloy E-Form® which shows improved formability due to a texture-weakening effect analogous to the ZXK alloy described above [13]. Figure 5 shows a car body highlighting different fields of application for magnesium sheets and identifying the individual material requirements in focus. At the current level of development, the lowest application potential for magnesium sheets lies within structural parts for which high material strength and deformability resulting in a high crash resistance are required. For structural parts, magnesium sheets compete against advanced and high strength steels—an unrealistic scenario at this point in time. In contrast, closures and doors as well as in parts outer panels offer the benefit that these components can be mounted onto the painted steel or aluminium car body, such that for the magnesium parts, separate coating processes can be followed. For such

D. Klaumünzer et al.

applications, however, corrosion resistance and a good surface quality are key. Interior parts (e.g. back-shelf, mounting plates, seat structures) are typically protected from a wet environment, thus reducing the risk of corrosion. Additionally, such parts are usually not visible such that expensive surface coatings are not required. However, for these lightweight applications, magnesium die-castings are already well established, offering in most cases a cheaper solution. Based on the above consideration, it was decided to focus prototype studies on closures and a decklid in the geometry of the current Volkswagen Passat was chosen. In a first step, sheet thicknesses as well as radii were adjusted to arrive at a magnesium specific design to account for the change in material properties compared to the steel series equivalent. In order to eliminate the risk of galvanic corrosion, a focus was placed on a full magnesium design. More precisely, this meant that all individual components (i.e. inner and outer panel as well as reinforcements, see Fig. 6 for reference) were adjusted to the use of magnesium sheet. Figure 6 shows a weight comparison between the steel series part, an equivalent aluminium construction and the new magnesium design. Compared to steel, a weight saving of 6.3 kg (*50%) is achieved. More importantly, when contrasted to aluminium, the magnesium design is lighter by about 24% or 1.8 kg, which is a sufficiently large fraction of the maximum theoretical limit of 33% based on relative densities. The individual component thicknesses range from 1.2 to 3 mm as depicted in Fig. 6. A simple comparison of the individual contribution of each component to the overall weight saving

Fig. 5 Schematic drawing of a car body highlighting the challenges for different fields of application of magnesium sheet components

Magnesium Process and Alloy Development for Applications …

19

Fig. 6 (Left) Weight comparison of a Volkswagen Passat decklid in different materials. (Right) Individual components and sheet thicknesses for a magnesium option

reveals that the outer panel possesses the highest potential for weight reduction, thus stressing the need for effective solutions for surface quality and corrosion protection for outer skin applications in magnesium. Forming trials using texture-weakened alloys, such as the ZXK alloy described above as well as E-Form®, demonstrate the benefits of the improved formability of such sheet alloys for real components. Figure 7 shows a photograph of the inner panel successfully formed at a reduced temperature of 160 °C. Forming such a complex geometry using the conventional AZ31 alloy typically requires temperatures in excess of 220 °C. It is clear that a reduction in forming temperature has substantial practical implications on the processability of magnesium sheets in an automotive production environment. Aside a saving in energy and thus costs, lower forming temperatures simplify the temperature control during forming in multiple steps in a transfer press and allow the use of a wider range of lubricants. In addition to beneficial impacts on part forming, the improved properties of the new sheet alloys help to reduce Fig. 7 Inner panel for a Volkswagen Passat decklid utilising texture weakened alloys and thus reducing the forming temperature down to 160 °C (adapted from Ref. [12])

the complexity of joining the individual components. In this context, preliminary studies have shown that the hemming operation joining the outer and inner panel can now be performed within a reduced temperature window of 150– 160 °C. As a consequence, this poses the opportunity to use conventional hemming sealers that would otherwise, i.e. at higher hemming temperatures, experience premature setting. All of these factors contribute to a simplified processing of magnesium sheets in an automotive manufacturing environment, demonstrating both the need for and potential behind improving the material properties beyond those of existing alloys.

Conclusions Due to the limited property spectrum of existing alloys, magnesium sheets have only found niche applications within the automotive industry so far. By performing alloy development on the basis of the twin-roll casting technique, a new

20

Mg–Zn–Ca–Zr sheet alloy can be shown to offer enhanced formability at lower temperature. The production of this alloy can be scaled to industrial dimensions, offering the opportunity to conduct studies at prototype component level. Based on a magnesium decklid in the Volkswagen Passat geometry, the improvement in formability helps to simplify production processes during automotive manufacturing, thus offering the chance to reduce costs. This shows how alloy development contributes towards more feasible and economic magnesium components in the automotive industry.

D. Klaumünzer et al.

6.

7.

8.

9.

References 10. 1. H. Friedrich, S. Schumann, “Research for a “new age of magnesium” in the automotive industry”, Journal of Materials Processing Technology 117 (2001) 276–281. 2. Th. Becker, “Neuer Meilenstein im Leichtbau”, mobiles 38 (2015) 6–8. 3. B.-Y. Chan, M.-S. Shim, K.S. Shin and N.J. Kim, “Current issues in magnesium sheet alloys; Where do we go from here?”, Scripta Materialia 84–85 (2014) 1–6. 4. N.J. Kim, “Critical Assessment 6: Magnesium sheet alloys: viable alternatives to steel?”, Materials Science and Technology 30 15 (2014) 1925–1928. 5. K. Hantzsche, J. Bohlen, J. Wendt, K.U. Kainer, S.B. Yi and D. Letzig, “Effect of rare earth additions on microstructure and texture

11.

12.

13.

development of magnesium alloy sheets”, Scripta Materialia 63 (2010) 725–730. J. Bohlen, J. Wendt, M. Nienaber, K.U. Kainer, L. Stutz and D. Letzig, “Calcium and zirconium as texture modifiers during rolling and annealing of magnesium-zinc alloys”, Materials Characterization 101 (2015) 144–152. Y. Chino, X. Huang, K. Suzuki and M. Mabuchi, “Enhancement of stretch formability at room temperature by addition of Ca in Mg-Zn alloy”, Materials Transactions 51 4 (2010) 818–821. T. Bhattacharjee, B.-C. Suh, T.T. Sasaki, T. Ohkubo, N.J. Kim and K. Hono, “High strength and formable Mg-6.2Zn-0.5Zr-0.2Ca alloy sheet processed by twin roll casting” Materials Science and Engineering A 609 (2014) 154–160. D.-W. Kim, B.-C. Suh, M.-S. Shim, J.H. Bae, D.H. Kim and N. J. Kim, “Texture evolution in Mg-Zn-Ca alloy sheet” Metallurgical and Materials Transactions A 44A (2013) 2950–2961. Z.R. Zeng, M.Z. Bian, S.W. Xu, C.H.J. Davies, N. Birbilis, J.F. Nie, “Texture evolution during cold rolling of dilute Mg alloys” Scripta Materialia 108 (2015) 6–10. J. Hofstetter, M. Becker, E. Martinelli, A.M. Weinberg, B. Mingler, H. Kilian, S. Pogatscher, P.J. Uggowitzer and J.F. Löffler, “High-strength low-alloy (HSLA) Mg-Zn-Ca alloys with excellent biodegradation performance”, JOM 66 4 (2014) 566– 572. D. Klaumünzer, S.B. Yi, D. Letzig, S.-H. Kim and J.J. Kim, “Entwicklung neuer Magnesiumbleche für die Automobilindustrie”, Konstruktion 9 (2018) IW4–IW6. S.-J. Kim, Y.-S. Lee and D. Kim, “Analysis of formability of Ca-added magnesium alloy sheets at low temperatures”, Materials Characterization 113 (2016) 152–159.

Thermally Activated Slip in Rare Earth Containing Mg–Mn–Ce Alloy, ME10, Compared with Traditional Mg–Al–Zn Alloy, AZ31 Vikaas Bajikar, Jishnu J. Bhattacharyya, Nathan Peterson, and Sean R. Agnew

Abstract

The thermally activated deformation of textured Mg alloys is evaluated using repeated stress relaxation tests analyzed with the assistance of elasto-viscoplastic self-consistent (EVPSC) polycrystal modeling. The data, presented in a Haasen plot, suggests that the superposition of at least two mechanisms controls the thermally activated glide of dislocations in both a rare-earth containing alloy, ME10, and the conventional alloy, AZ31: forest dislocation interactions and a mechanism with a lower activation volume (solute–dislocation interaction and/or cross-slip). Keywords








Strain rate sensitivity Activation volume plot Stress relaxation Solute




to changes in dislocation velocity alone can be interrogated (rather than a combination of velocity and dislocation recovery effects). A function commonly used to describe the decrease in stress over time for a single relaxation is the logarithmic
   model: Ds ¼ kT ln 1 þ t , where V is the apparent Va

activation volume and c is a time constant. Taking the time derivative of this equation yields the stress and strain rate during relaxation. In order to discern the true activation volume, repeated relaxation data can be used to calculate a correction factor X, such that V = Va/X. X is given by [1, 2],    V Ds a n expð kT Va Ds Þ1 , where n is the nth PkTn1 X1 ¼ 1  ln 1 Va

Haasen

Extended Abstract The repeated stress relaxation method [1] is employed to determine the activation volume, V, of textured, polycrystalline Mg alloys ME10 and AZ31, in both the hard-rolled (F) and annealed (O) tempers. During stress relaxation testing, the material is held at a constant total strain, and the elastic strain relaxes. This relaxation occurs because the sample continues to strain plastically, even though the load frame cross-head remains fixed. In repeated relaxation tests, the sample is reloaded multiple times following relaxation. This largely eliminates the effect machine compliance can have on the results. Furthermore, if the sample remains elastic during reloading, it can be assumed that the dislocation density at the end of one relaxation is the same before the start of the next relaxation [2]. Therefore, effects related V. Bajikar
J. J. Bhattacharyya
N. Peterson
S. R. Agnew (&) University of Virginia, Charlottesville, VA 22904-4745, USA e-mail: [email protected]

a

c

1

Dsj

exp

kT

1

relaxation in the series. In order to gain insight into the rate-controlling mechanisms operative in these alloys during plastic deformation, the so-called “Haasen plot,” V1 , versus post-yield flow stress, r  ry , was calculated (Fig. 1), for ME10 in both F and O tempers and for AZ31 in O temper. The Haasen plot will reveal a linear relationship between the two quantities if the strain rate sensitivity, m, is constant with straining and is directly proportional to the slope. This is also known as adherence to the Cottrell–Stokes relationship. If the only rate-controlling mechanism of dislocation motion is forest interactions, the flow stress r ¼ rd / 1l , where l is the mean dislocation spacing. It also follows from the definition of the activation volume, V ¼  @DG @s , that the activation volume Vd / l. Hence as r ! 0, V ! ∞, and intercept of the Haasen plot will be zero. If another mechanism is simultaneously operative, its strengthening contribution rs follows a linear superposition ðr ¼ rd þ rs Þ e.g. solute strengthening and it exhibits a lower activation volume (denoted here as Vs ) relative to that of the forest interactions; it can be shown [3] that the overall
 activation volume is given by: V1 ¼ V1s þ V1s rrds . The m values, derived from the Haasen plot slope for the ME10 in F temper and O temper, are 0.017 and 0.021,

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_4

21

22

V. Bajikar et al. 0.014

intercept equal to zero [3, 6]. Furthermore, studies of several alloy systems, e.g., AJ, AX, and AZ, have revealed similar values to those found presently, V * 50–100 b3 [7–9]. The current study seeks to determine what is the second mechanism responsible for the rate-sensitive behavior of Mg alloys, beyond forest interactions.

k/V (MPa/K)

0.012 0.010 0.008 0.006 0.004 ME10 AR ME10-O AZ31B-O

0.002 0.000

0

50

100

σ−σ Y (MPa)

Fig. 1 Haasen plot for ME10 in the as-rolled and annealed (O) tempers, i.e., inverse activation volume normalized by Boltzmann’s constant, k, and plotted as a function of the flow stress minus the 0.002 offset yield stress

respectively. The intercept values, however, are same and equal to (61 bcubed)−1, where b is the Burgers vector. Interestingly, the AZ31 alloy has a higher intercept value of (76b3)−1 and a lower m = 0.008. Therefore, the contribution of the second mechanism is greater in AZ31, as compared to ME10. The reasons for this are the subject of further investigation. However, it can already be surmised that AZ31 has a much higher solute content as compared to ME10. Prior work on pure single and polycrystal Mg [4, 5] yielded V * 102–104 b3, which is consistent with the fact that at very low stresses interrogated, the dislocation–forest interactions would lead to very high V, with Haasen plot

Acknowledgements The authors wish to thank the US National Science Foundation, Division of Materials Research, Metals and Metallic Nanostructures (NSF-DMR-MMN) program, Grant Number: 1810197, overseen by program manager Dr. Lynnette Madsen for their financial support.

References 1. D. Caillard and J.L. Martin: Thermally Activated Mechanisms in Crystal Plasticity, Pergamon Materials Series Vol 8, Elsevier, 2003. 2. P. Spätig, J. Bonneville, and J-L. Martin: Mater. Sci. Eng. A, 1993, vol. 167, pp. 73–79. 3. W. A. Curtin: Scr. Mater., 2010, vol. 63, pp. 917–20. 4. H Conrad, L Hays, G Schoeck, and H Wiedersich: Acta Metall., 1961, vol. 9, pp. 367–78. 5. D.H Sastry, Y.V.R.K Prasad, and K.I Vasu: Curr. Sci., 1970, pp. 97–100. 6. R. A. Mulford: Acta Metall., 1979, vol. 27, pp. 1115–24. 7. P Lukac and Zuzanka Trojanová: Key Enginnring Mater., 2011, vol. 465, pp. 101–4. 8. Zuzanka Trojanová, Kristián Máthis, Pavel Lukáč, Gergely Németh, and František Chmelík: Mater. Chem. Phys., 2011, vol. 130, pp. 1146–50. 9. Zuzanka Trojanová, Peter Palček, Pavel Lukáč, and Zdeněk Drozd: Magnes. Alloy. Solid Liq. States, 2014, pp. 3–48.

Part II Magnesium Technology 2019: Alloy Design and Casting

Bimodal Casting Process of Eco-Mg Series Alloys by Vertical High-Speed Press Machine Fabrizio D’Errico

Abstract

Ultimate advancements in non-flammable magnesium alloys (the so-called Eco-Mg alloys series, acronym of Environment Conscious Magnesium) have been recently achieved. Preliminary laboratory tests have been already confirmed safe window parameters (i.e. exposure time and maximum melting temperature) for processing Eco-Mg AZ91D-1.5CaO alloy in full liquid state. Further project challenge has been now completed; Eco-Mg samples have been realized in special novel pre-industrial vertical 4-column press machine, designed and constructed for processing Eco-Mg series alloys in air by two routes, by high-pressure die casting and by semisolid state injection. An external rotating stirrer finely controlled in temperature, and rotating velocity completes the semisolid process route, allowing Eco-Mg alloy remaining below flame-ignition temperature. Microstructure investigation on samples produced in both high-pressure die casting and semisolid state patterns has been performed. To date, such a bimodal pilot line is thought to introduce affordable industrial way e for producing near-net-shape Mg parts. Keywords














Eco-Magnesium alloys Life cycle assessment Recycling Lightweight design CO2 reduction Semisolid process

F. D’Errico (&) Department of Mechanical Engineering, Politecnico di Milano, Via La Masa 34, Milan, 20156, Italy e-mail: [email protected]

Drawbacks Preventing Magnesium to Become Key Material for Lightweight Strategies in Automotive Sector Driven in the past century by aerospace, historically magnesium alloy development has complied with transportation industry when had pursued weight-saving strategies. Despite weight reduction promises, safety and cost of Mg are still big issues due to its high flammability if treated in air and higher price than aluminum alloy, its main competitor in light weighting strategies in transportation sector. Mg cannot be processed in the same way as Al, since protective but expensive —as the necessary equipment—and extremely pollutant gases such as SF6 or SiO2 should be used to shelter melt Mg from oxygen. Additionally, even if Mg with very low viscosity is a key factor for many industry sectors (e.g. automotive, aerospace and also electronic goods), as it allows the production of very articulated shapes with reduced defects and very thin sections, highly skilled personnel are required in Mg foundries that have been limiting unlocking wider adoption in such a cost-sensitive mainstream automotive market. Notwithstanding Mg is one of most abundant elements in the earth’s crust, raw magnesium and consequently Mg master alloy prices are still less competitive than Al. Such an uneven competitiveness is due to shortage of Mg for Mg system-based alloy production. The availability of magnesium for producing Mg alloys is highly impacted by the demand for aluminum, titanium and steel because magnesium is largely used to make these metals. Primary magnesium metal is a reducing agent for the production of titanium, also being an aluminum alloying constituent and one element required in steelmaking production for desulfurization phase. This implies small percentage of the worldwide primary magnesium production is actually used for making Mg master alloys for structural applications, castings or wrought products. To complicate matters, almost eighty percent of the world’s magnesium supply is made in China by Pidgeon process at lower cost than Western facilities, causing in recent

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_5

25

26

decade many extractive plants to shut down thus subtracting to marketplace further Mg supplies. Despite several plant closures in the run-up to the 2008–2009 downturn, production in the USA, Russia and Israel has expanded. Producers in China still dominate magnesium production, but several projects were under development to increase North America primary magnesium metal capacity elsewhere [1]. The use of magnesium in automobile parts has been increasing although these are generally limited to expensive automobiles where the higher price of magnesium is a minor issue. By this way, the long-term outlook for Mg is expected to be positive and the world’s largest Mg producers see a return to robust demand in the coming years. As regard Europe, magnesium is today included in the list of the critical raw materials by the EC and considered fundamental to Europe’s economy. However, all such issues related to uncommon technical competences required for casting Mg, to high investment costs for magnesium die casting and to volatility of primary Mg price are constraining nowadays Mg casting part to a niche market in automotive sector. A forward-looking approach market-oriented to goal a sustainable competitive advantage in Mg cast part manufacturing would be possibility to have a non-SF6 (and other GHGs cover gases) die casting process, thus casting Mg with almost die casting press machine usually employed in aluminum foundry. Toward such strategy, high potentialities have been shown in recent year by CaO-modified magnesium system alloys, commercially known as Eco-Magnesium alloys. The Eco-Mg series was developed in 2000s’ decade by the Korean Institute of Industrial Technology (KITECH), and it is spreading rapidly as it does not require the protective use of highly environmental impacting gases as the presence of CaO provides a protective cover during the melting phase. As-cast part in magnesium can be manufactured in machine employed also for casting aluminum alloys, competitiveness of magnesium casting process costs would overcome existing drawbacks that still prevent transport market penetration on a wide scale. Moreover, as there is an urgent need to find new eco-compatible solutions to process Mg with a drastic reduction of the carbon footprint, reduction of total GHGs emitted over product cycle of Mg cast parts is also possible by elimination of protective gases during the cast part manufacturing (protective gases are still used for master alloy production).

The Semisolid Process Route Solution Further technical barriers that restrict use of Mg cast alloy to structural application in automotive sector are related to lower mechanical properties than equivalent cast part made

F. D’Errico

of aluminum alloy. This is physiologically due to lower mechanical properties of Mg cast alloys than usual aluminum cast alloys largely employed in vehicle part manufacturing. Comparing the common AZ91D alloy largely employed in magnesium die casting with the A356 aluminum alloy widely used in vehicle part manufacturing, the magnesium choice is not capable to target similar fatigue resistance and toughness achieved by aluminum alloy. In fact, in commercial affordable magnesium cast alloys alloyed with Al, the higher is the content of Al, the higher is YS and UTS, but the lower is the percentage elongation and toughness as consequence of increasing amount of a brittle network made of the intermetallic compound b (Mg17Al12) that forms at a-phase grain boundary. Furthermore, a physiologic low ductility and low toughness of Mg have to be taken into account because of particular crystal structure —the hexagonal close-packed structure—that reduces plastic deformation resources below 200 °C. At least, microstructure defects, typically matrix discontinuities provoked by casting, shall be considered as further sources of decay in above-mentioned mechanical properties. One strategy to reduce cast defects consists in: (a) reducing as possible shrinkage phenomenon that occurs in solidification phase of liquid metal inside the die and (b) interrupting/braking the intermetallic network of the brittle b phase. Both these tasks can be obtained by switching from full liquid process route to semisolid process route (SSM). Solidification is basically governed by nucleation and growth phenomena and usually driven by dendritic growth of grains that commonly results in shrinkage-related defects, such as porosities due to insufficient liquid feeding in those sections with latest solidification. By the early 1970s, it is known a material which is cooled in its semisolid state with stirring action shows a spheroidal microstructure and it also exhibits thixotropic behavior that consists in reducing by time its viscosity. As the material is not cool in die by 100% liquid state, there is less solidification shrinkage and minor occurrence of related defects. A family of various processing routes has been developed, but all start from production of spheroidal microstructure obtained by applying shearing force in metal when it is maintained in its semisolid phase; such a shearing force is necessary to brake dendritic structure and promoting solidified grains to grow with globular shape. After stirring, the material can be solidified into billet, a sort of precursor material, that can be reheated again in semisolid state, placed into the shot chamber of a die casting machine and forced into the die. This is the typical process route of the thixocasting process route. Otherwise, the precursor billet can be placed between closed dies, reheated in semisolid interval (usually lower temperature than thixocasting), thus pressed realizing the thixoforging process route.

Bimodal Casting Process of Eco-Mg Series Alloys …

High cost for acquiring precursor billets, however, precluded such SSM process routes to spread commercially in automotive sector part manufacturing. Alternatively, rheocasting the feedstock material is not solidified after stirring but poured in semisolid state directly into the die casting machine. If rheocasting is employed in industry for processing SSM aluminum alloys, it is not easily practicable at low cost for magnesium metal with a stirring unit furnace conventional die casting machine that works in air because of ignitable problems of conventional magnesium alloys. For this reason, in the past, by the 1990s, the thixomolding SSM process has been developed that use full solid Mg alloys provided in granule, directly fed in a hot chamber where semisolid metal is produced and agitated by mixing realized by a rotating screw, similarly to what the injection molding machine makes with thermoplastic polymeric materials. The presence of Ar insufflated in hot chamber prevents part of the magnesium liquid to react with oxygen and burn. Notwithstanding technology viability of thixomolding process, high investment cost and higher maintenance costs prevent such a process route to be affordable and interesting for automotive sector. The above-mentioned shortcomings due to difficulty to realize a cost-effective technology for industrial exploitation of SSM route have been carefully considered in research work here described. The result has been the set up of an experimental campaign to validate possibility to cast via SSM process route an Eco-Mg alloy series, a AZ91-1.5%CaO in a laboratory-scale vertical press machine without use of protective gas.

27

Fig. 1 Microstructure of as-cast supplied AZO9115 alloy

Experimental Part Material and Procedure A commercial Eco-Mg series alloy AZ91D with a nominal composition of 8.5% Al, 0.75% Zn, 0.3% Mn, Fe and Ni below 0.001%, and Mg as a balance modified with 1.5% CaO (hereinafter also AZO9115) provided by Korea Institute of Industrial Technology was employed as experimental material. The as-cast microstructure when supplied in ingot is shown in Fig. 1. It is composed of the coarse structure of a-Mg and the network of eutectic b-Mg17Al12 compound which mainly discontinuously distributes at the grain boundaries (see Fig. 1). An innovative injection press machine has been employed to develop a hot injection chamber die casting of metal. The prototype machinery is shown in Fig. 2. It is based on a 4-column vertical layout with the injection plunger positioned under the lower stationary plate of press machine. The injection unit is therefore constituted by: • a vertical chamber thermoregulated by mean of pressurized oil pumped by external thermal unit directly into

Fig. 2 Machine overviewing: 4-column vertical press machine equipped with semimold dies, visible in the middle; machine is covered over four faces with protective shield against hazardous liquid metal projections

complicated network of inner circumferential channels realized in the injection chamber; • a plunger thermoregulated by mean of pressurized water pumped by a further external thermal unit inside the plunger tip that is directly in contact with liquid metal. The nominal quantity of magnesium metal that can be introduced in the injection chamber is around 0.7 kg per cycle. Figure 2 shows some details of press machine. The concept of vertical architecture finally adopted for realizing the press machinery was based on the objective of reducing

28

F. D’Errico

as possible the injection plunger stroke (dimensioned at 100 mm as maximum) for targeting several benefits, such as: • reducing duration of the injection phase, thus decreasing total cycle time of the process with semisolid metal handling; • reducing the displacement velocity of plunger during injection in order to reduce turbulence; the maximum injection velocity set in the press machine is 135 mm/s, actually around 1 order of magnitude lower than conventional horizontal high-pressure die casting machine; • (thanks to the bottom-up vertical injection layout) reducing total quantity of air to evacuate during injection; this implies it is possible to reduce the maximum injection pressure required for air elimination, resulting in compact size of the machine and consequently low equipment cost. Table 1 shows gathered main machine dimensioning data, meanwhile Fig. 3 shows the shape geometry of part sample. Finally, the simplified diagram reported in Fig. 4 shows the three main phases of the adopted SSM process route. The production of semisolid material has been realized by stirring material by near-liquidus isothermal treatment conducted in an external unit. The external unit (refer to Fig. 5) was a small furnace with a graphite crucible with dosing function specifically designed and constructed for double function: (a) preparing semisolid metal by use of stirrer moved by brushless electric motor capable to maintain constant rotational velocity as set by operator; (b) maintaining stirring temperature for SSM preparation by precise control of temperature of metal operated by a submerged thermocouple. In order to check safety procedure, operators would have followed in laboratory to melt the AZO9115 alloy in no hazardous condition to melt and pour material investigated in the semisolid stirring furnace (see again Fig. 5); specific duration tests have been conducted in a laboratory chamber furnace crossing the near-liquid and full liquid phase in wide time interval. Particularly, 0.6 kg of AZO9115 alloy has been heated to 600 °C, and its temperature then fluctuated Table 1 Main machine press and cast part dimensioning data

Fig. 3 Definitive cast sample geometry obtained at the end of closed-loop CAE design optimization

changing to near-liquidus and liquidus conditions as reported in Table 2. No burning occurred during a prolonged duration test conducted with thermal cycling around temperatures near to the liquidus temperature for selected alloy (namely 595 °C). Results are shown in Table 2; they apparently show some incoherent results regarding with liquid-to-solid and reverse solid-to-liquid transitions. Semisolid state has detected during the prolonged duration test at 596.50 °C (refer to the time record 1:20:00 in Table 2), even though full liquid state has been also observed at 594.00 °C (refer to the record 0:31:00 in Table 2). However, this apparent inconsistency is due to thermal inertia phenomena as consequence of rapid thermal cycling around the liquidus temperature of the alloy.

Results and Discussion AZ91D system is the most used Mg alloy in foundry applications, and it has the widest solidification temperature interval among commercial Mg alloys. Addition of 1.5% CaO compound is one characteristic of the commercial Eco-Mg series alloys, and it is added for retarding oxidation of magnesium in molten state. The high risk for Mg alloys is that, once burning reaction starts, as it is endothermic, flame is not self-extinguishable, even in the case heat source is removed. The ignition of conventional Mg alloys occurs because of porous MgO surface oxide film at high temperature. MgO oxide film on surface of molten Mg cannot act as

Machine characteristics m/sec

Injection plunger velocity (two velocity stages available over plunger total stroke)

0.28

Plunger total stroke

100

Theoretical locking force

700

Projected complete shot area

135

cm2

Specific pressure on metal (max)

533

kg/cm2

Injection force

200

(+10%)

mm (+15%)

(+15%)

kN

kN

Bimodal Casting Process of Eco-Mg Series Alloys …

29

Fig. 4 SSM process route employed

for semisolid casting; accordingly, steps are depicted in Fig. 4. Two different SSM process conditions have been investigated varying stirring time at constant pouring temperature, at constant stirring temperature and rotational constant speed (see Table 3). Two SSM materials reported in Table 3 were therefore employed for casting two samples by use of the vertical short stroke pressure die casting machine. The injection was performed at fixed optimized parameters, specifically:

Fig. 5 Semisolid furnace working at constant temperature and constant stirring rate

a protective layer to prevent ignition of Mg, because of its porous nature that puts in contact directly the molten metal with oxygen in atmosphere [2]. CaO-added Mg alloy solidifies with reactive phase formation Mg2Ca, regardless of CaO contents and process condition [3]. During oxidation under an ambient atmosphere, Mg2Ca phase on the surface dissociates by oxygen to MgO and CaO. By this way, a dense oxide film consisting in porous CaO and MgO isolates metal bath from atmosphere [3]. However, it is reported by Lee and Kim [4] the 1.5 wt.% CaO added AZ91 Mg increased at 615 °C, thus about 300 °C compared to conventional AZ91D alloy [4]. The prolonged duration test above described (see again Table 2) was performed in order to define safe operating procedure

• injection rate, single stage at 135 mm/s for 100 mm total displacement; • 525 bar metal pressure obtained in mold die calculated in near-liquidus condition; • 245 °C heating temperature of mold die and hot chamber; • 135 °C temperature at the plunger tip. The two SSM samples are shown in Fig. 6. The microstructures obtained for the samples produced in the vertical press machine are reported in Fig. 7. Compared to as-cast microstructure shown in Fig. 1, the SSM sample shows the typical microstructures obtained at sub-liquidus temperatures made of primary solid particles and fine a-Mg dendrite and Mg17Al12 phase formed during final solidification after extraction of crucible from isothermal stirring unit, pouring in injection chamber, injected in mold die under pressure and finally completing solidification in mold die. Microstructures are characterized by small quantity of primary solid particles disperse in relatively uniform matrix composed of almost equiaxed a-Mg dendrite and network distributing Mg17Al12.

30 Table 2 Duration test for burning limit and safe holding time

Table 3 SSM process parameters of two samples

F. D’Errico Time h:mm:ss

T (°C)

Note

Burning incipience

0:00:00

600.00

Liq

No

0:04:00

583.00

Semisolid

No

0:11:00

600.00

Liq

No

0:20:00

602.00

Liq

No

0:21:00

604.00

Liq

No

0:24:00

606.00

Liq

No

0:26:00

604.80

Liq

No

0:28:00

601.50

Liq

No

0:31:00

594.00

Liq

No

0:38:00

581.70

Semisolid

No

0:43:00

580.60

Semisolid

No

1:20:00

596.50

Semisolid

No

Sample

T pouring (°C)

T stirring (°C)

Stirring time (s)

Rotational speed (rpm)

SSM_ 1_AZ91O

609.00

593.00

60

120

SSM_ 1_AZ91O

609.00

593.00

240

120

Fig. 6 Test samples obtained in: a HDPC mode and b SSM mode with process parameters detailed in Table 3

It is known [5] that the mechanism of non-dendritic structure formation in rheocasting process route is driven by both the mixing action and the shearing effect during stirring. The mixing action is thought to uniformly distribute temperature around solid nuclei, thus reducing the solute concentration gradient at the liquid–solid interface. As a result, it is usual a non-dendritic structure grows preferentially. On the other hand, shearing condition provoked by

Fig. 7 Microstructures obtained in central section of cylindrical body of cast sample for: a SSM process with stirring time 60 s and b SSM process with stirring time 240 s

stirring action makes partial dendrite arm brake and split to form further micro-particulates that work as nuclei for

Bimodal Casting Process of Eco-Mg Series Alloys …

homogeneous crystallization. However, some difference has been obtained in terms of the network of intermetallic compound for two conditions. According to previous experience of some authors, prolonged stirring time at low stirring rate velocity has reduced the extension of the intermetallic compound, as it has worked furthermore to increase the homogeneity of solute concentration in secondary nuclei formation during rapid solidification of the mass [6].

Conclusions A rheocasting process route has been performed with a AZ91D CaO-added magnesium alloy. The process route has been realized by preliminary preparation of the semisolid material employing an isothermal mechanical stirrer that worked at near-liquidus temperature. The semisolid injection has been successfully conducted in safe condition in air, thanks to very compact cycle time. By such results, this preliminary test campaign puts some promising premise to affordable SSM process route that matches faster cycle time typical of die casting process. Furthermore, possibility to process commercial Eco-Mg system alloys in air by compact

31

and low-cost press machine that is capable to process alternatively Mg and Al alloys is an interesting perspective for increasing occupancy of machine, which is one key economic aspect to consider in industrial manufacturing processes.

References 1. Lee Bray E, (2018), in: Magnesium Metal U.S. Geological Survey Yearly Bulletin: https://minerals.usgs.gov/minerals/pubs/commodity/ magnesium/mcs-2018-mgmet.pdf accessed: 8 September 2018. 2. Czerwinski, F. (2002), The oxidation behaviour of an AZ91D magnesium alloy at high temperatures, Acta Materialia, 50 (10): 2639–2654. 3. S.H. Ha, J.K. Lee, S.K. Kim, Effect of CaO on oxidation resistance and microstructure of pure. Mg, Materials Transactions. 49(5) (2008) 1081–1083. 4. J.K. Lee, S.K. Kim. Effect of CaO Addition on the Ignition Resistance of. Mg-Al Alloys. Materials Transactions, Vol. 52, pp: 1483–1488, 2011. 5. Flemings, M.C., (1991) Behavior of metal alloys in the semisolid state, Metallurgical Transactions B, 22 (3): 269–293. 6. Zhang, Y et al. (2008), Influence of Processing Parameters on Microstructure of Casting rolling Semi-solid AZ91D Magnesium Alloy, Solid State Phenomena (141–143): 535–538.

Investigation of the Evolution of the Microstructure in the Directionally Solidified Long-Period Stacking-Ordered (LPSO) Magnesium Alloy as a Function of the Temperature Daria Drozdenko, Kristián Máthis, Stefanus Harjo, Wu Gong, Kazuya Aizawa, and Michiaki Yamasaki

K. Máthis (&) Nuclear Physics Institute of the CAS, Řež 130, 250 68 Řež, Czech Republic e-mail: [email protected]

developed—Mg alloys with long-period stacking-ordered (LPSO) phase. This class of alloys exhibits better mechanical performance and promising high-temperature properties comparing to the commercial Mg alloys [1–4]. In general, besides the dislocation slip, “kinking” and twinning can control the plastic deformation of the Mg-LPSO alloys. Both are significantly dependent on materials parameters (e.g. size and orientation of the LPSO phase) and experimental conditions (loading direction, temperature, etc.). Twinning was found to be a common mechanism during deformation of a-Mg phase, but rarely observed in the LPSO phase [5]. Generally, conditions for kink and twinning activation in the LPSO phase are not clear yet and need further investigation. The research presented here deals with the characterization of the deformation behavior of directionally solidified Mg–Zn–Y alloy during uniaxial compression as a function of the testing temperature and the mutual orientation of the lamellar structure of LPSO phase and loading axis. The acoustic emission (AE) measurements can provide information about active deformation mechanisms. It was recently shown that deformation twinning, kinking and collective dislocation motion are strong sources of acoustic emission and, therefore, provide information about dynamic processes during plastic deformation [6]. Neutron diffraction (ND) measurements provide global information about microstructure changes. Thus, the combination of in situ AE and ND and the ex situ microscopy observations enables revealing the active deformation mechanism in Mg-LPSO materials at various experimental conditions.

D. Drozdenko
K. Máthis Faculty of Mathematics and Physics, Charles University, Ke Karlovu 5, 121 16, Prague 2, Czech Republic

Experimental

Abstract

The influence of the LPSO-phase orientation and the temperature on the deformation mechanisms of directionally solidified Mg–Zn–Y magnesium alloy has been investigated by neutron diffraction and acoustic emission (AE) technique. The results indicate that the kinking mechanism and activation of non-basal slip are significantly temperature and orientation dependent with respect to the loading axis. Keywords











Magnesium Long-period stacking-ordered structure Neutron diffraction Acoustic emission Non-basal slip

Introduction During last decades, Mg alloys have been attracting considerable interest in lightweight structural applications. However, the degradation of the mechanical properties at elevated temperatures significantly limits their wider expansion. Recently, a new generation of Mg alloys was D. Drozdenko
M. Yamasaki Magnesium Research Center, Kumamoto University, 2-39-1, Kurokami, Kumamoto, Japan

S. Harjo
W. Gong J-PARC Center, Japan Atomic Energy Agency, 2-4 Shirane Shirakata, Tokai-Mura, Naka-Gun, Ibaraki, 319-1195, Japan K. Aizawa Elements Strategy Initiative for Structural Materials, Kyoto University, Yoshida-Honmachi, Sakyo-ku, Kyoto, 606-8501, Japan

Directional solidification of Mg + 24 wt% Y + 12 wt% Zn alloy was conducting using Bridgman furnace (NEV-DS2, Nissin Giken, Japan) to receive billet with a diameter of 30 mm. Cylindrical samples with a gauge length of 16 mm and diameter of 8 mm were used for uniaxial compression

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_6

33

34

Fig. 1 a Initial microstructure of the specimen. The arrows indicate the loading directions with respect to the solidification direction (DS). b Initial texture of the material, as it was detected in axial detector (i.e. parallel with DS)

tests. In order to investigate orientation effect, the as-solidified material was cut from the ingot along the solidification direction (longitudinal direction—LD) and perpendicular to it (transversal direction—TD); see Fig. 1a. The neutron diffraction experiments were performed by the engineering neutron diffractometer—TAKUMI, at Japan Proton Accelerator Research Complex (J-PARC). The specimens were fixed horizontally to a deformation rig with its axial direction at +45° to the incident neutron beam, and two detector banks are used to collect the diffracted neutron patterns at +90 and −90° relative to the incident beam, respectively. Deformation tests were performed at a constant strain rate of 0.1 mm/min at 25 °C for both LD and TD orientations and additionally at 300 °C for the LD specimen. Concurrently, the AE activity was monitored by a MICRO-II AE system (Physical Acoustics Corporation). The system allows a continuous storage of the AE signals with a sampling frequency of 2 MHz. A preamplifier of 60 dB and two types of sensors, a piezoelectric PICO S/N 7549 and a S9215 high-temperature sensor for room and elevated temperature tests, respectively, were used. The microstructure was observed by a confocal optical microscope (OM; Lasertec C-130). The surfaces for OM observation were polished by the standard method with final etching in picric acid.

Results and Discussion The initial microstructure exhibits a lamellar structure, with an average thickness of lamellae of 250 lm (Fig. 1a). The lamellae are oriented mainly in solidification direction. The material exhibits a strong texture, with a distinct maximum of the f1120g peak in the axial detector (Fig. 1b), which means that the basal planes are parallel to the growth

D. Drozdenko et al.

direction. Similar texture of Mg–Zn–Y alloys was observed in [7]. The deformation curves exhibit a significant orientation and temperature dependence. At room temperature, the value of yield stress of the specimen deformed in LD exceeds that for sample in TD by almost four times (Fig. 2a, b). In the case of LD sample, there is a stress plateau after reaching the yield point followed by a secondary hardening stage. The AE response differs for the two sample orientations as well. In LD, there are high-amplitude bursts present and there is a local maximum of the signal near the yield point. Further maximum is present at the onset of the secondary hardening stage. Interestingly, there is virtually no AE during the elastic part, which is usually very common for the magnesium alloys [8]. In contrast, in TD, the AE response is significantly lower, without distinct maximum. In upper part of Fig. 2, we compared the diffraction data for three most intensive peaks: basal f0:0:0:18g and two pyramidal: 1:8g and f1:0: 1:10g Owing to the sample geometry, f1:0: Fig. 2a and c shows data for axial detector, whereas in Fig. 2b for radial one. It is obvious for both LD (Fig. 2a) and TD (Fig. 2b) that after reaching the yield point, the intensity of the basal peak increases, which indicates a significant crystal rotation towards f0:0:0:18g orientation. With increasing strain, in LD, the intensity of the basal peak suddenly drops, approximately at the same level as the second AE maximum appears. In TD, there is some decrease in the intensity, but significantly lower than that for the LD. The pyramidal planes behave differently for the two sample orientations. There is an obvious peak shift of f1:0:1:8g and 1:10g peaks after reaching the yield for LD, which f1:0: indicates plastic deformation on these planes. This effect is also present in TD, but it is less pronounced. Based on the above-listed experimental evidences, the following scenario of the plastic deformation of the directionally solidified LPSO magnesium alloys can be proposed. In LD, owing to the initial texture, the basal slip is difficult. Thus, high stress concentration is necessary for slip initiation (c.f. high yield stress) which starts most probably with prismatic slip. Therefore, there is a negligible AE during the elastic loading (in contrast to collective dislocation motion in basal plane in the case of typical Mg alloys [8]). The prismatic slip causes crystal rotation towards the f0:0:0:18g orientation, as it was shown in [9]. Around the yield point, the first-order f10 11gh 1 123i pyramidal slip becomes active, which causes a peak shift on f1:0: 1:8g and f1:0:1:10g planes. As it was shown by Matsumuto et al. [10], the pyramidal slip activates basal slip, which finally leads to kinking. Kinking requires collective movement of a large number of basal dislocations [7]; thus, it is followed by high emission of elastic waves. Consequently, a secondary peak is observed in the AE response. At the same time, owing to

Investigation of the Evolution of the Microstructure …

35

Fig. 2 Stress–strain curves and the corresponding strain evolution of the acoustic emission and neutron diffraction data for a LD at RT; b TD at RT; c LD at 300 °C

the kinking process, the basal planes come out from their Bragg position and the f0:0:0:18g decreases. In TD, the texture allows activation of the slip at lower stresses, which leads on hand to lower yield stress and to less intensive, continuous-like AE. The non-basal slip plays also here an important role, and further investigation of this issue is essential. The presence of the kinks in LD sample was confirmed by OM (Fig. 3a). In addition, twin-like objects were observed in sample deformed in TD (Fig. 3b). Our

preliminary EBSD results indicated that those fractions are not covered by known twin systems in Mg alloys. However, they have a common tendency to rotate the lattice towards the f0:0:0:18g orientation. At 300 °C in LD, there is a drop in the yield stress, but the degradation of the mechanical properties is smaller than for the commercially used magnesium alloys. The overall process of the deformation is similar to that at RT, but the diffraction data indicates that the kinking is less intensive. In Fig. 3c, the micrograph shows that the kinks formed at this

Fig. 3 Microstructure after deformation a LD at RT; b TD at RT; c LD at 300 °C. The twin-like objects for TD (see text) are indicated by arrow

36

temperature are narrower in comparison to those at RT (Fig. 3a). The dislocation slip is enhanced by thermal activation, which leads to the appearance of dislocation-originated AE signals already in the elastic regime.

Conclusions The deformation processes in the directionally solidified LPSO magnesium alloys were investigated as a function of the loading direction and temperature using acoustic emission and neutron diffraction. The following conclusions can be drawn: 1. At the beginning of the deformation, the samples exhibit a significant crystallographic rotation towards f0:0:0:18g by activation of prismatic slip. 2. Kinking is a dominant deformation mechanism in LD specimen, but its initiation takes place above 0.04% strain. 3. In TD, the activation of the basal dislocation slip is easier due to the texture. This leads to low yield stress. The kinking in this sample orientation is negligible. 4. At 300 °C in LD, the kinks are thinner and the dislocation slip is enhanced by thermal activation.

Acknowledgements This research was funded by Czech Science Foundation grant number GB14–36566 G. K.M. acknowledges the support of the Operational Programme Research, Development and Education, The Ministry of Education, Youth and Sports (OP RDE, MEYS) [CZ.02.1.01/0.0/0.0/16_013/0001794].

D. Drozdenko et al.

References 1. Inoue A, Kawamura Y, Matsushita M, Hayashi K, Koike J (2001) Novel hexagonal structure and ultrahigh strength of magnesium solid solution in the Mg–Zn–Y system. J. Mater. Res. 16:1894–1900 2. Kawamura Y, Hayashi K, Inoue A, Masumoto T (2001) Rapidly Solidified Powder Metallurgy Mg97Zn1Y2 Alloys with Excellent Tensile Yield Strength above 600 MPa. Mater. Trans 42:1172– 1176 3. Kawamura Y, Kasahara T, Izumi S, Yamasaki M (2006) Elevated temperature Mg97Y2Cu1 alloy with long period ordered structure. Scripta Mater. 55:453–456. 4. Yamasaki M, Hashimoto K, Hagihara K, Kawamura Y (2011) Effect of multimodal microstructure evolution on mechanical properties of Mg–Zn–Y extruded alloy, Acta Mater 59: 3646–3658. 5. Kishida K, Inoue A, Yokobayashi H, Inui H (2014) Deformation twinning in a Mg–Al–Gd ternary alloy containing precipitates with a long-period stacking-ordered (LPSO) structure. Scripta Mater. 89:25–28. 6. Garcés G, Máthis K. Medina J, Horváth K, Drozdenko D, Oñorbe E, Dobroň.P, Pérez P, Klaus M, Adeva P (2018) Combination of in-situ diffraction experiments and acoustic emission testing to understand the compression behavior of Mg-YZn alloys containing LPSO phase under different loading conditions, Int. J. Plasticity 106: 107–128. 7. Hagihara K, Yokotani N, Umakoshi Y (2010) Plastic deformation behavior of Mg12YZn with 18R long-period stacking ordered structure, Intermetallics 18:267–276. 8. Máthis K, Chmelík F, Janeček M, Hadzima B, Trojanová Z, Lukáč P (2006) Investigating deformation processes in AM60 magnesium alloy using the acoustic emission technique Acta Mater. 54:5361– 5366 9. Mayama T, Noda M, Chiba R, Kuroda M (2011) Crystal plasticity analysis of texture development in magnesium alloy during extrusion. Int. J. Plasticity 27:1916–1935. 10. Matsumoto R, Uranagase M, Miyazaki N (2013) Molecular dynamic analyses of deformation behavior of long-periodstacking-ordered structures. Mater. Trans. 54:686–692.

TEM Studies of In Situ Formation of MgO and Al4C3 During Thixomolding of AZ91 Magnesium Alloy Conducted in CO2 Ł. Rogal, L. Litynska-Dobrzynska, and Bogusław Baran

Abstract

MgO and Al4C3 compounds were produced in situ by reaction of CO2 with AZ91 alloy at semi-solid temperature range. Using modified magnesium injection molding, reactive carbon dioxide was introduced to hot zone of cylinder to conduct controlled oxidation reaction with partially melted Mg alloy. In result, nano-scale native MgO (30–50 nm) and small amount of Al4C3 carbide within eutectic mixture consisted of a(Mg) and bMg17Al12 were formed. Apart from eutectic obtained directly from liquid state, proeutectic magnesium solid solution a(Mg)p is crystallized. Homogeneously distributed a(Mg) globular grains (not melted during the process) with size 20–50 lm and volume 8–12% were also visible in the microstructure. Electron beam was aligned along [011] zone axis of MgO and [01-10] of aMg matrix. (11-1) planes of MgO were parallel to (0002) planes of the matrix. Orientation relationship was [011] MgO ║ [01-10] aMg and (11-1) MgO ║ (0002) aMg. Keywords

Thixomolding




MgO




Composites




In situ

using mechanical stirring [7] or ultrasonic waves [9] have been routine techniques to produce the MgMCs. An alternative route for the cost-effective way of MgMC production is an in situ synthesis developed in recent years [10]. Several ways to obtain magnesium matrix composites have been invented so far to form in situ the reinforcement phase, e.g. Mg2Si/Mg by mechanical milling the elemental Mg, Al, and Si powders [11, 12] or the reaction of Mg with Ti and C to obtain the TiC/Mg material [13] and the Mg–TiB2–TiB one created by the reaction of Mg with KBF4 and K2TiF6 [14]. We developed a new method based on the semi-solid metal (SSM) processing technology called thixomolding, which involves the in situ formation of nano-scale native MgO by the application of a mixture of reactive CO2 and Ar gases as the inert atmosphere to the partially melted AZ91 magnesium alloy. The detailed microstructural studies have been conducted, using scanning and transmission electron microscopes to identify the strengthening particles and to explain the mechanisms of their formation in the AZ91 composites.

Experimental Procedures Material Preparation and Microstructure Analysis

Introduction Magnesium alloys are attractive for the automobile and aircraft engine industries, because they have the lowest density of all technologically relevant alloys [1–6]. However, the relatively low ductility and strength are limiting their wide commercialization [5]. Magnesium matrix composites (MgMCs) provide the improvement of their mechanical properties [7]. The powder metallurgy [8] and the casting route in which the particles of SiC, TiC, and AlN are mixed Ł. Rogal (&)
L. Litynska-Dobrzynska
B. Baran Institute of Metallurgy and Materials Science, Polish Academy of Sciences, 25 Reymonta Str., 30-059 Krakow, Poland e-mail: [email protected]

In the present study, a mechanically chipped magnesium alloy with the nominal composition of 8.7% Al, 0.7% Zn, 0.2% Mn, 0.02% Si, Mg balance, which corresponded to the AZ91 alloy, was used as a matrix during the formation of in situ composites. In order to obtain composites, thixomolding technology was used, which also belongs to the net-shape forming technique. The process has been conducted according to the scheme presented in Fig. 1 (upper part). Mg chips were introduced to a hopper in a rebuilt Arburg prototype system tightly connected with the cylinder and preheated to 595 °C. The precise temperature distribution during the process is shown in Fig. 1 (bottom part). The hopper was tight closed and argon as inert gas and the CO2 reactive atmosphere

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_7

37

38

Ł. Rogal et al.

Fig. 1 Technology of Mg-based nano-composite production

were introduced. The machine with a clamp force of 120 tones equipped with rectangular shape steel mold covered with BN preheated to 150 °C was applied for injection molding. The observations were carried out using a Leica DM IRM metallographic microscope. The microstructure was also examined applying an FEI SEM XL30 scanning electron microscope (SEM) working in backscattered electron (BSE) or secondary electron (SE) modes. The phase analysis was carried out with a Philips PW 1710 diffractometer and Co Ka filtered radiation in the scan mode 20–100 2-Theta range at the anode voltage of 40 kV and current of 40 mA. The Vickers hardness was measured using a Zwick/ZHU 250 tester under the load of 5 kg in accordance with ASTM E92 standards. The compression test was performed following the PN-57/H-04320 standard using an INSTRON 3382 machine and samples 3 mm in diameter and 4.5 mm in height. The differential scanning calorimetry (DSC) was applied to measure the solidus–liquidus range and to determine the dependence of liquid phase amount in function of temperature using a Netzsch DSC404F1 apparatus. The samples, 3 mm in

diameter and 2 mm thick, were heated in an alumina crucible at the rate of 15 °C/min up to 700 °C in argon atmosphere.

Results and Discussion: Microstructure Analysis In the present study, the gas mixtures consisting of Ar, which was used to protect Mg against ignition, and CO2 to react with partially melted alloy (ratio: 10%–Ar and 90%–CO2) were applied so that to generate the composite microstructure in the AZ91 alloy processed using thixomolding. The obtained microstructure consisted of globular grains (not melted during the process) in the amount of about 12 vol.%, and the average grain size of 45 lm surrounded by fine proeutectic a(Mg)p irregular grains and the eutectic mixture consisting of a(Mg)e and b-Mg17Al12 (white areas in Fig. 2a). The X-ray analysis (Fig. 2b) confirmed, apart from a(Mg), the presence of cubic MgO (below 3 vol.%) and small amount of b-Mg17Al12 as well as Al4C3 (which formed by reaction of aluminum with carbon).

Fig. 2 AZ91 magnesium alloy after thixomolding in a mixture of Ar and CO2; a optical image, b X-ray diffraction pattern of the process conducted in Ar protective gas

TEM Studies of In Situ Formation of MgO and Al4C3 …

39

Fig. 3 TEM observations of AZ91 alloy thixomolded under Ar and CO2 atmosphere; a BF image of the eutectic area, b circular diffraction pattern

The transmission electron microscopy observations (Fig. 3a) from the area of structure obtained directly from the liquid, show the mixture of proeutectic a(Mg)p and eutectic phases with nanocrystalline particles with the size of 30–70 nm (dark contrasted areas). A detailed analysis of ring diameters in the circular diffraction pattern (Fig. 3b) allowed determining the lattice parameters of phases and assigning them as MgO. A high-resolution transmission electron microscopy images of the eutectic area, where MgO particles co-exist with a(Mg)e are presented in Fig. 4. The electron beam is aligned along [011] zone axis of MgO and [01-10] of aMg matrix. The (11-1) planes of MgO are parallel to (0002) planes of the matrix. Orientation relationship: [011] MgO ║ [01-10] aMg and (11-1) MgO ║ (0002) aMg. Wang et al. [15] studied the MgO obtained as a by-product of surface oxidation in pure Mg and AZ91 magnesium alloy (the liquid was dispersion mixed at 700 °C), identifying the cubic structure and truncated-octahedral morphology of the oxide particles, respectively. The authors explained that the various morphologies of MgO were connected with the interfacial energy

of certain crystal plane changes relative to other planes due to the adsorption or segregation of alloying elements at the interface (AZ91). That would affect the growth behavior of MgO crystals in the melt, and eventually, the planes with the lowest interfacial energy would be the faceted ones. The MgO identified in the present work had the cubic structure, which was different from that, what Wang reported, probably due to various conditions of the MgO formation (lower temperature of the process, different kind of reactive atmosphere, and shorter time for the reaction). Nevertheless, the semi-coherence of MgO with a(Mg)e was preserved, which had positive effect on mechanical properties. Additionally, the room temperature compression test was carried out to determine the deformation behavior of the thixomolded AZ91 alloy obtained in the mixture of Ar and CO2. The AZ91 processed under reactive mixture of Ar with CO2 revealed yield strength of 220 MPa at compression strength of 460 MPa and hardness 103 ± 2 HV. The increased properties were connected with the presence of nano-MgO particles and Al4C3 carbides which were homogenously distributed in the Mg matrix.

Ł. Rogal et al.

40 Fig. 4 High-resolution transmission electron microscopy image (HRTEM) of the area of MgO/a(Mg) interface taken of AZ91 composite. The inserted squares, which correspond to fast Fourier transform (FFT) from HRTEM images (marked in the micrograph), confirm the presence of MgO [011] with cubic structure and a(Mg) [01-10]

Conclusions (1) Microstructure of AZ91 nano-composites consisted of a(Mg) globular grains (20–50 lm) and volume of 8– 12% surrounded by a mixture of proeutectic a(Mg)p with irregular shape and average size of 4 lm. Fine eutectic mixture consisting of a(Mg)e + MgO + b-Mg17Al12 and Al4C3 carbides was observed around them. (2) High-resolution transmission electron microscopy images of eutectic area, where MgO particles and a(Mg)e coexisted, showed that (11-1) planes of MgO were parallel to (0002) planes of the matrix. Orientation relationship was [011] MgO ║ [01-10] aMg and (11-1) MgO ║ (0002) aMg.

Acknowledgements The authors gratefully acknowledge the financial support by the National Centre for Research and Development, Grant No.: LIDER/007/151/L-5/13/NCBR/2014.

References 1. Maruyama K, Suzuki M, Sato H. Creep strength of magnesium-based alloys. Metall. Mater. Trans. A. 2002;33:875– 882. 2. Alfredo Monteiro W, editor. Mg-Based Quasicrystals. New Features on Magnesium Alloys; 2012. p. 1–28. 3. Wei L, Doudou Y, Gaofeng Q, et al. Microstructure Evolution of Semisolid Mg-2Zn-0.5Y Alloy during Isothermal Heat Treatment. Rare Metal Materials and Engineering. 2016;45(8):1967–1972. 4. Wei LY, Dunlop GL. Crystal symmetry of the pseudo-ternary T-phases in Mg-Zn-rare earth alloys. Journal of Materials Science Letters. 1996;15:4–7.

TEM Studies of In Situ Formation of MgO and Al4C3 … 5. Fridlyander JN, Eskin DG, editors. Magnesium Alloys Containing Rare Earth Metals: Structure and Properties. Advances in Metallic Alloys. 2003, p. 34–68. 6. Avedesia MM, Baker H. ASM specialty hand book: magnesium and magnesium alloys. Ohio: ASM International;1999, p. 54. 7. Ye HZ, Liu XY, B. Luan B. In situ synthesis of AlN in Mg–Al alloys by liquid nitridation. Journal of Materials Processing Technology. 2005;166:79–85. 8. Mabuchi M, Kubota K, Higashi K. Tensile strength, ductility and fracture of magnesium-silicon alloys. Journal of Materials Science. 1996;31(6):1529–1535. 9. Joost WJ, Krajewski PE. Towards magnesium alloys for high-volume automotive applications. Scripta Materialia. 2017;128:107–112.

41 10. Azarniya A, Safavi MS, Sovizi S, et al. Metallurgical Challenges in Carbon Nanotube-Reinforced Metal Matrix Nanocomposites. Metals. 2017;7:384. 11. Aarstad K. Protective Films on Molten Magnesium [dissertation]. Norway: Norwegian University of Science and Technology; 2004. 12. Cao G, Kobliska J, Konishi H, et al. Tensile properties and microstructure of SiC nanoparticle-reinforced Mg-4Zn alloy fabricated by ultrasonic cavitation-based solidification processing. Metallurgical and Materials Transaction A. 2008;39:880–886. 13. Cao G, Kobliska J, Konishi H, et al. Mg–6Zn/1.5%SiC nanocomposites fabricated by ultrasonic cavitation-based solidification processing. Journal of Materials Science. 2008;43(16):5521–5526. 14. Tanaka K, Iijima S. Carbon Nanotubes and Graphene. Elsevier, 2014. 15. Wang Y, Fan Z, Zhou X, et al. Characterisation of magnesium oxide and its interface with a-Mg in Mg–Al-based alloys. Philosophical Magazine Letters. 2011;91(8):516–529.

FFF of Mg-Alloys for Biomedical Application M. Wolff, T. Mesterknecht, A. Bals, T. Ebel, and R. Willumeit-Römer

Abstract

Additive manufacturing is a very promising approach to patient-specific implants. In combination with degradability, individual tissue regeneration could be obtained. Like metal injection moulding (MIM), fused filament fabrication (FFF) of metal powders belongs to binder-based sintering technologies. However, FFF of metal powders does not require an expensive mould, but it offers individual prototyping of sophisticated shaped parts at low costs. FFF of metals is novel and processing of Mg powders is just at the start of development. In the present work, special Mg-alloy-powder-polymer blends were developed to enable manufacturing of flexible filaments and failure-free green parts. Consolidation to final metal parts took place using SF6-free powder metallurgical (PM) sintering technique. Test specimens and implant demonstrator parts were successfully produced. The specimens showed mechanical properties of up to 177 MPa UTS, 123 MPa yield strength and 2.8% elongation at fracture. Thus, the mechanical properties are equivalent to those of as-cast material. Based on these results, FFF appears to be a very promising approach to Mg implant production. Keywords

Metal injection molding




Magnesium




Sintering

M. Wolff
T. Mesterknecht
A. Bals
T. Ebel
R. Willumeit-Römer (&) Helmholtz-Zentrum Geesthacht, Centre for Materials and Coastal Research, Institute of Materials Research, Div. Metallic Biomaterials, Max-Planck Str. 1, 21502 Geesthacht, Germany e-mail: [email protected]

Introduction Additive manufacturing techniques, for example selective laser melting (SLM) and fused filament fabrication (FFF), become more and more interesting for model manufacturing, prototyping and low-cost production of tools and consumer applications in low batches, as the raw material can be transferred into the final part at comparably low manufacturing effort. In parallel, Mg and its alloys show a growing impact on consumer, lightweight [1–3] and biomedical applications as biodegradable implant materials [4–11]. Here, special biomedical Mg-based alloys demonstrate mechanical properties similar to those of human bone tissue [12–15]. The development of additive manufacturing routes for the processing of magnesium materials, this means using powder metallurgical techniques, is, thus, consequential. Sintering provides a perfect basis to achieve a homogeneous microstructure independent of the part’s geometry. Metal injection molding (MIM) for economic near net shape mass production and fused filament fabrication (FFF) for prototyping and individual parts is most promising. Both techniques enable manufacturing of complex-shaped components. For instance, a patient’s broken bone can be scanned by a laser tomography scanner and an optimal implant shaped by computer modelling software generating a 3D data file of the implant. In a next step, a 3D filament printer can generate the physical real implant using this data file. The screws for fixation would then be produced by MIM. Since the microstructure is the same in both cases, compatibility in terms of mechanic properties or degradation would not be an issue. The use of both MIM and FFF requires the production of an appropriate feedstock system made of Mg-based alloy powder and polymeric binder components. Hence, this feedstock can be used for the complete range from small batches to large numbers of parts (1 piece—1 mio. pcs. and beyond). However, sintering of Mg and its alloys is known

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_8

43

44

M. Wolff et al.

dedicated biomedical Mg-alloys, as there are Mg–Ca, Mg– Ca–Zn, Mg–Gd and Mg–Ag alloys.

Materials and Methods Powder, Feedstock and Green Part Printing

Fig. 1 Mg-alloy demonstrator parts and test specimens made by FFF of Mg-alloy at Helmholtz-Zentrum Geesthacht

as highly challenging due to its high oxygen affinity [16]. Nevertheless, recent work on sintering of Mg [19–21] and MIM of Mg [20–22] has shown that it is possible to overcome this major challenge. This study is focusing on deducing a new process chain from the knowledge of MIM processing of Mg-alloys to enable FFF of Mg metal powder-loaded polymeric filament. The study aims to identify a parameter set enabling 3D printing of first test specimen and demonstrator parts. For this, different test specimens and biomedical implant demonstrator parts as shown in Fig. 1 were produced and tested: • Right-hand side components: feedstock granules made of Mg-alloy powder and binder components. • Middle: filament for 3D printing, made of Mg-feedstock. • Middle part: dogbone tensile test specimen according to ISO 2740-B, green and sintered conditions. • Left-hand parts: bone plate demonstrator part in the as-sintered condition. • Lower parts: micro-tensile test specimen and small bone plate in the green and as-sintered condition. In this study, which is to the knowledge of the authors the first study on FFF of Mg-alloys worldwide, the technical Mg-alloy AZ81 was used due to its availability and good sintering performance. However, future studies will focus on

Fig. 2 Filament made of Mg-alloy and polymeric binder

For the feedstock, commercial spherical gas atomized Mg– 8Al–1Zn alloy powder, in the following referred to as AZ81, with a size smaller than 45 µm was used (SFM, Martigny, Switzerland). The feedstock consisted of the following polymeric binder components: polypropylene–copolymer– polyethylene, stearic acid, and thermoplastic elastomer (TPE). For the blending of the components at 160 °C, a planetary mixer (Thinky ARE-250 planetary mixer, Japan) applying approx. 500G acceleration was used. To avoid any additional oxygen uptake, all loose powder handling was carried out in a glovebox system (Unilab, MBraun, Germany) under protective argon atmosphere. For the feedstock granulation, a cutting mill (Wanner B08.10F, Germany) was used. The filament of 2.85 mm in diameter (see Fig. 2), necessary for 3D-printing, was produced using the single-screw extruder of an injection moulding machine (Arburg 320S, Germany). The fused filament fabrication of dogbone shape tensile test specimen green parts according to ISO 2740-B, micro-tensile test specimen and bone plate demonstrator parts as shown in Fig. 1 was performed using an Ultimaker3 3D printer at 185 °C mold temperature and 60–70 °C build plate temperature. The layer highness was set at 0.2 mm. Print speed was set to 5–8 mm/s to ensure the successful manufacturing of printed green parts. For variation of the print core nozzle diameter, the original Ultimaker print core (ULT18) was substituted by a hard core (3D Solex), which allows changing of the nozzle. The following nozzle diameters were used: 0.4, 0.6 and 0.8 mm. Figure 3 displays the technical setup of the print head.

Debinding and Sintering Solvent debinding of stearic acid and the organic TPE was done using hexane at 40–65 °C for 20–120 h (Lömi

FFF of Mg-Alloys for Biomedical Application

45

Fig. 4 Example of a printed part with large nozzle diameter leading to partial overflow of feedstock

Sintering Temperature and Time Fig. 3 Setup of a filament print head

EBA50/2006, Germany). Thermal debinding of the backbone polymer and consolidation of the 3D-printed parts through sintering were carried out in a combined debinding and sintering hot wall furnace (MUT Advanced Heating GmbH, Jena, Germany). Thermal debinding took place in protective argon atmosphere at low ambient pressure, using a flow of 1 L/min at 10–60 mbar. The sintering time was set to 4 h at 615 °C. According to the state of the art of sintering of magnesium, getter material inside the labyrinth-like crucible configuration was used [19]. For the getter material, irregular coarse pure magnesium powder grit (Sigma Aldrich, USA) was used.

Materials’ Characterization The microstructure was investigated by optical microscopy (Olympus PGM 3) and secondary electron microscopy SEM (Tscan). Micro-tensile tests of sintered FFF specimens were performed on a Zwick ZM005 materials’ testing machine.

Results and Discussion Nozzle Diameter The printing experiments revealed that changing the nozzle diameter involves the need of a recalibration of nozzle-build plate distance and layer thickness. A larger nozzle diameter enables feeding of more material and faster printing. However, too much feedstock may be squeezed into the part as shown in Fig. 4.

For MIM, as a result of a sintering optimization process, the chosen sintering time of 4 h at 615 °C sintering temperature leads to sufficient sintering performance [23]. This knowledge of MIM of Mg could directly be transferred to the sintering within the process of FFF, as shown in the following microstructure of a micro-tensile test specimen in Fig. 5. The microstructure demonstrates the typical liquid-phase sintered structure of the selected AZ81-alloy. It could be observed that 4 h sintering time is sufficient for the consolidation of the printed parts, analogous to the experience made in MIM. Some roundish pores are visible in the left-hand image. This is typical residual porosity for sintered MIM and FFF Mg-alloy parts. How does the fast sintering of the AZ81 material occur? On the one hand that it can be assumed that the Al-rich permanent liquid phases may be able to reduce the MgO-layer material in accordance with the relation of Gibbs free energy of oxide formation, if the powder particle surface is coated by magnesium oxide in the as-received condition On the other hand, Mg may be able to also reduce the oxides, if Al or Zn oxides coat the particle surfaces of the used powder. In order to answer this question in more detail, an adequate analysis of the particle surface, e.g. XPS, IR spectroscopy or µ-XRF, in combination with DSC and XRD analyses, has to be done in future work. Regarding the sintering performance of Mg–Al alloys, literature reveals that during magnesium oxide layer reduction, as shown in Eq. 1: MgO þ Al

! Mg þ Al2 O3

ð1Þ

is formed. This redox reaction takes place in dependency of element concentration, in both directions [25].

46

M. Wolff et al.

Fig. 5 Microstructure of sintered fused filament fabricated micro-tensile test specimen

Mechanical Properties The following diagram in Fig. 6 displays the micro-tensile test results of the first set of FFF micro-tensile test specimens, sintered at 615 °C for 4 h sintering time in comparison with their corresponding MIM parts. Comparing the tensile yield strengths, the FFF material achieved a slightly but not significantly better value of 123 ± 6 MPa than the corresponding MIM material (118 ± 2 MPa). The FFF parts attained an elastic modulus of 33 GPa which matches the typical range of human bone tissue. The left-hand side set of columns in the diagram

Fig. 6 Mechanical properties after tensile testing of 3D printed parts (FFF) in comparison with their MIM produced counterparts

displays the mechanical properties of the printed FFF parts in comparison with their MIM counterparts. The FFF material achieved a moderate UTS of 177 ± 9 MPa, which is quite sufficient for many consumer and biomedical implant applications. The corresponding MIM processed parts scored a UTS of 240 ± 6 MPa. For comparison, recent biodegradable polylactide acid-based implants have achieved a maximum of 120–140 MPa UTS. An elongation at fracture of 2.8 ± 0.7% of the FFF parts could be achieved in comparison to 5.1 ± 0.6% of the corresponding MIM parts. The lower ultimate tensile strength and ductility of the FFF material are probably caused by fabrication flaws like bigger pores and possible partly delamination, as shown exemplarily in Fig. 7. The delamination shown can be explained by residing TPE-binder residuals after solvent debinding. These TPE-binder residuals can form gas bubbles and cavities during thermal debinding. A reduction or avoidance of this effect may be realized by decreasing the heating rate during thermal debinding or by reducing or substituting the TPE-binder component, as shown exemplarily in Fig. 8. However, at the current stage of research, TPE is necessary for the failure-free printing of the filament. The corresponding MIM parts did not need TPE for the shape giving process. For the MIM parts, paraffin wax was used instead of TPE. For FFF, paraffin wax is not adequate due to embrittlement of the filament material.

FFF of Mg-Alloys for Biomedical Application

47

Fig. 7 Layer delamination within a sintered FFF micro-tensile test specimen

Fig. 8 Microstructure without delamination of a sintered FFF micro-tensile test specimen

Electron Microscopy of Fracture Surface In addition to the optical microscopy images in Figs. 7 and 8, the following SEM image in Fig. 9 reveals the effect of gas bubble formation and delamination within the printed component in more detail.

Printing and Sintering of First Bone Plate Demonstrator Parts In a further step, the sintering results described in the previous chapter were adapted to the sintering of bone plate demonstrator parts produced by FFF, as shown exemplarily in Figs. 1 and 10.

48

Fig. 9 SEM image of the fracture surface of a FFF micro-tensile test specimen

M. Wolff et al.

regime for the Mg alloy powder could be transferred from the MIM processing chain. The tensile test revealed an average UTS of 177 MPa at 2.8% elongation at fracture and 123 MPa yield strength. Corresponding parts produced by MIM showed significantly higher UTS and elongation at fracture. The lower mechanical strength and elongation at fracture of the FFF parts can be explained by macroscopic delamination effect within the microstructure of the printed parts. However, FFF of Mg-alloys is quite a new technique. The achievement of these first tensile test results and mechanical properties is sufficient for many consumer applications and, in many cases, even for the substitution of biodegradable polymer-based polylactide acid implants, which are currently in clinical usage. Further work will be done regarding optimal filament feedstock blending for failure-free printing of green parts and delamination-free sintering. Generally, MIM and FFF processing of Mg-alloy powders implement economic and SF6-free near net shape mass production of small-sized complex Mg-alloy parts. However, in this first study on FFF of Mg alloys, the technical Mg-alloy AZ81 was used due to its availability and good sintering performance. Future studies will focus on dedicated biomedical Mg-alloys, as there are Mg–Ca, Mg– Ca–Zn, Mg-Gd and Mg–Ag alloys, to enable additional biodegradation test procedures.

References

Fig. 10 First demonstrator parts made by FFF of Mg (left: green part; right: sintered part): typical bone plate

As shown in Fig. 10, the demonstrator parts exhibit significant shrinkage due to the sintering process and a silver shining surface, proving good sintering and pure sintering atmosphere.

Conclusions and Outlook This study displays the first worldwide published tensile test results of fused filament fabrication (FFF) Mg alloy parts. The investigation proved that Mg alloys like AZ81 can successfully be processed by additive manufacturing techniques like FFF and, subsequently, consolidated to a nearly dense metal part through sintering. The optimal sintering

1. H. E. Friedrich, B. L. Mordike, Springer, Berlin, Heidelberg, Germany, (2006). 2. K. U. Kainer, Wiley-VCH, Weinheim, Germany, (2010). 3. H. Dieringa, N. Hort, K. U. Kainer, Proceedings of LMT 2011, Trans Tech Publications Ltd, Material Science Forum, Vol. 690 (2011). 4. Z. Li, X. Gu, S. Lou, Y, Zheng. The development of binary Mg-Ca alloys for use asbiodegradable materials within bone, Biomaterials 29 (2008) 1329–1344. 5. M. P. Staiger, A. M. Pietak, J. Huadmai, G. Dias. Magnesium and its alloys as orthopaedic biomaterials- a review, Biomaterials 27 (2006), 1728–1734. 6. F. Witte, V. Kaese, H. Haferkamp, E. Switzer, A. Meyer-Lindenberg, C. J. Wirth, H. Windhagen, Biomaterials 26 (2005) 3557–3563. 7. F. Witte, J. Reifenrath, P. P. Müller, H. –A. Crostack, J. Nellesen, F. W. Bach, D. Bormann, M. Rudert, Materialwissenschaft und Werkstofftechnik 37 (2006) 504–508. 8. F. Witte, F. Feyerabend, P. Maier, J. Fischer, M. Störmer, C. Blawert, W. Dietzel, N. Hort, Biomaterials 28 (2007) 2163–2174. 9. F. Witte, H. Ulrich, M. Rudert, E. Willbold, Journal of Biomedical Materials Research 81A (2007) 748–756. 10. F. Witte, J. Fischer, J. Nellesen, H. A. Crostack, V. Kraese, A. Pisch, F. Beckmann, H. Windhagen, Biomaterials 27 (2006) 1013– 1018. 11. F. Witte, H. Ulrich, C. Palm, E. Willbold, Journal of Biomedical Materials Research 81A (2007) 757–765.

FFF of Mg-Alloys for Biomedical Application 12. G. Poumarat, P. Squire, Biomaterials 14 (1993) 337–349. 13. A. R. Cunha et al, International Journal of Hypertension, Hindawi Publishing Co. (2012) Art.-ID 754250. 14. C. Janning, E. Willbold, C. Vogt, J. Nellesen, A. Meyer-Lindenberg, H. Windbergen, F. Thorey, F. Witte. Magnesium hydroxide temporarily enhancing osteoblast activity and decreasing the osteoclast number in peri-implant bone remodelling, Acta Biomaterialica 6 (2010) 1861–68. 15. F. Witte, J. Fischer, J. Nellesen, H. A. Crostack, V.Kraese, A. Pisch, F. Beckmann, H. Windhagen. In vitro and in vivo corrosion measurements of magnesium alloys, Biomaterials 27 (2006) 1013– 1018. 16. M. Wolff, M. Dahms, T. Ebel. Sintering of Magnesium, Advanced Engineering Materials 12 (2010) 829–836. 17. M. Wolff, T. Guelck, T. Ebel. Sintering of Mg and Mg-Ca alloys for biomedical applications, Euro PM2009 Proceed. 2 (2009) 417– 422. 18. M. Wolff, C. Bischof, M. Dahms, T. Ebel, T. Klassen, 9th International Conference on Magnesium and their Applications, Vancouver, Canada, July 8–12, (2012) page 102.

49 19. M. Wolff, J. G. Schaper, M. Dahms, T. Ebel, K.U. Kainer and T. Klassen, Powder Metallurgy, Vol. 57 No. 5 (2014) 331–340. 20. M. Wolff, J. G. Schaper, M. R. Suckert, M. Dahms, F. Feyerabend, T. Ebel, R. Willumeit-Römer, T. Klassen, Metal Injection Molding - MIM of Magnesium and its Alloys, Metals, Vol. 6 No. 118 (2016), https://doi.org/10.3390/met 6050118. 21. M. Wolff, J. G. Schaper, M. R. Suckert, M. Dahms,, T. Ebel, R. Willumeit-Römer, T. Klassen, Magnesium powder injection molding (MIM) of orthopedic implants fpr biomedical application, JOM, Vol.68, No4. (2016) https://doi.org/10.1007/s11837–0161837-x 22. B. Wiese, The Effect of CaO on Magnesium and Magnesium Calcium Alloys, Dissertation, Clausthal University of Technology (2016). 23. M. Wolff, J. G. Schaper, M. Dahms, T. Ebel, R. Willumeit-Römer, T. Klassen (2018) Metal Injection Molding (MIM) of Mg-Alloys. In: TMS 2018 147th Annual Meeting & Exhibition Supplemental Proceedings. The Minerals, Metals & Materials Series. Springer, Cham, DOI:https://doi.org/10.1007/978–3-319-72526-0_22.

Effects of Gd/Y Ratio on the Microstructures and Mechanical Properties of Cast Mg–Gd– Y–Zr Alloys J. L. Li, D. Wu, R. S. Chen, and En-Hou Han

Abstract

Three Mg–xGd–yY–0.5Zr (x + y = 13, wt%) alloys were prepared by sand casting to investigate the effects of Gd/Y ratio on the microstructures and mechanical properties. The results show that Gd/Y ratio had little influence on the grain size and the phase constitution of the microstructures. However, the volume fraction of the second phase Mg24(Gd, Y)5 in as-cast state increased, while that of the cubic phase (Y, Gd)H2 in as-solutionized state almost unchanged with the decrease of Gd/Y ratio. The uniaxial tension tests of the three alloys show that for both as-solutionized and as-aged states, the yield strength was slightly increased but the ductility was apparently decreased with the decrease of Gd/Y ratio from 3.33 to 1.17. It was thought that Gd and Y atoms in Mg matrix have approximate solute strengthening effect but different adverse effect on ductility, which should be responsible for the mechanical properties difference for the three alloys. Keywords

Cast Mg–Gd–Y alloy Mechanical properties




Gd/Y ratio




Microstructure

J. L. Li
D. Wu (&)
R. S. Chen (&)
E.-H. Han The Group of Magnesium Alloys and Their Applications, Institute of Metal Research, Chinese Academy of Sciences, 62 Wencui Road, Shenyang 110016, China e-mail: [email protected] R. S. Chen e-mail: [email protected] J. L. Li School of Materials Science and Engineering, University of Science and Technology of China, 19 Jinzhai Road, Hefei 230026, China




Introduction Mg alloys have attracted much attention in recent years owing to their high specific strength, significant weight loss effect, and the resulting broad prospective applications in vehicles [1–3]. Among them, Mg alloys containing rare-earth elements (RE) have received considerable interest due to the high yield strength achieved through precipitation hardening and solution strengthening [4–6]. Especially, Gd has the high equilibrium solid solubility in Mg (23.49% in mass fraction (wt%) at 548 °C), which decreases exponentially with the decrease of temperature (3.82 wt% at 200 °C). Therefore, the Mg–Gdbased alloys are prone to form supersaturated solid solution, inducing excellent ageing strengthening by forming more precipitates. However, for Mg–Gd binary alloys, it was reported that only when the Gd contents are more than 10 wt % that the Mg–Gd binary alloys can exhibit significant ageing strengthening [7]. On consideration of cost, multicomponent Mg–Gd-based alloys were developed with addition of Zn, Nd, Y, Zr and other alloying elements [7, 8]; the most typical should be Mg–Gd–Y–Zr system alloys, which are found to exhibit outstanding mechanical properties and better creep resistance than conventional Al and WE alloys at both ambient temperature and elevated temperatures [9]. In the past few years, microstructure and mechanical properties of the Mg–Gd–Y–Zr system alloys have been widely reported [9–11]. Adjusting the content of each rare-earth (RE) element should be the primary method in a large number of studies, on the premise that the total addition of RE elements was controlled in a scale of certain extent. Nodooshan et al. [12] investigated the effects of different Gd contents (3–12 wt%) on the microstructure, age hardening response and mechanical properties of the Mg–xGd–3Y– 0.5Zr alloys. Wang et al. [13] investigated compositional dependence of age hardening response and tensile properties for Mg–10Gd–xY–0.4Zr (x = 1, 3, 5 wt%) alloys. The results show that both Gd and Y played a similar role of enhancing ageing hardening response and tensile properties [12, 13].

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_9

51

52

J. L. Li et al.

However, to our knowledge, research concerning the effects of Gd/Y values on the microstructures and mechanical properties is not widely developed. In this work, three Mg– Gd–Y–Zr alloys with same total RE content in mass fraction (wt%), but different Gd/Y values were prepared. The effects of Gd/Y values on the microstructures and mechanical properties of the sand cast Mg–Gd–Y–Zr alloys were investigated.

software. Vickers hardness testing was performed using 500 g load and a holding time of 10 s. No fewer than eight measurements were taken in each specimen. Tensile test specimens have gauge dimension of 10  3.5  2.5 mm. Tensile tests were performed by universal tensile testing machine at a strain rate of 1  10−3 s−1 and room temperature.

Experimental Procedures

Results and Discussion

Three alloys with nominal chemical compositions of Mg– 7Gd–6Y–0.5Zr (GW76), Mg–9Gd–4Y–0.5Zr (GW94) and Mg–10Gd–3Y–0.5Zr (GW103) (wt%) were produced by conventional sand–cast ingot metallurgy; the corresponding Gd/Y ratio is 7/6, 9/4 and 10/3, respectively. The chemical composition of the three alloys were determined to be Mg– 7.29Gd–6.10Y–0.34Zr (wt%), Mg–9.4Gd–3.76Y–0.41Zr (wt%) and Mg–10Gd–2.57Y–0.38Zr (wt%) by an inductively coupled plasma atomic emission spectrum apparatus. Solution treatment (T4) was carried out at 525 °C for 8 h [14], 510 °C for 6 h [15] and 525 °C for 12 h [11] for GW76, GW94 and GW103, respectively. Ageing after solution treatment was implemented at 200–250 °C. The microstructures of the specimens were observed using scanning electron microscopy (SEM) on a JEOL JSM– 5600 equipped with Energy Disperse Spectroscopy (EDS). The area fractions of second phases were estimated from five SEM images at magnification of 200% using Image J

Microstructures

Fig. 1 Cast microstructures of a GW76, b GW94, c GW103 and d their representative phase constituents

Figure 1 shows the microstructures of the as-cast alloys. As can be seen from Fig. 1a–c, the three alloys with different Gd/Y ratios have a similar cast microstructure that was mainly made up of equiaxed grains and discontinuous second phases around grain boundaries. Semi-quantitative EDS analysis indicated that the three Mg–Gd–Y–Zr alloys have the same phase compositions, i.e. a-Mg solid matrix, skeletal Mg24(Gd, Y)5 and cuboid Mg5(Gd, Y), as shown in Fig. 1d. However, it should be noted that although the three alloys have approximate phase compositions, the area fractions of the second phases of them are apparently different; the GW76 alloy has obviously more second phases. Figure 2 displays the microstructures of the investigated alloys in T4 state. It is apparent that after solution treatment, the eutectic phases in as-cast dissolved completely, only trace cubic phases with the size of 0.3–3 lm were observed

Effects of Gd/Y Ratio on the Microstructures …

53

Fig. 2 SEM images of the matrix of the a GW76–T4, b GW94–T4 and c GW103–T4

12 11 10 9 8 7 6 5 4 3 2 1 0 -1

Grain size {

as-cast as-solutionized

120

Mg24(Gd, Y)5 +Mg5(Gd,Y)

100

(as-cast ) 80 60 40 YH2 (as-solutionized)

7/6

Grain size, µm

Area fraction, %

in all the alloys. Our previous investigation has revealed that the cubic phase in T4 state is not the cuboid Mg5(Gd, Y) in as-cast samples but (Y, Gd)H2, which might be generated during solution treatment and insolvable in subsequent ageing treatment [15]. Figure 3 plots variations of area fractions of second phases and the grain size with Gd/Y (wt%) ratio in both as-cast and as-T4 states. It could be seen that the second phase in as-cast alloys decreases from 8.8 to 3.3% as the Gd/ Y value was increased from 7/6 to 10/3, keeping the total RE contents at 13 wt%. The variation of second phases in as-cast indicates that decreasing Gd/Y decreases the total solubility of Gd and Y at room temperature. According to reference [16], Y would likely segregate around the grain boundaries due to the smaller equilibrium solid solubility. Therefore, it is reasonable that the second phase would increase rapidly as Gd/Y value drops. However, the area fractions of the cubic phase were 0.63, 0.5 and 0.36% for the GW76, GW94 and GW103 alloys in as-T4 state, respectively. It seems that Gd/ Y ratio has little impact on the fraction of the cubic phase in as-T4 state, which kept in less than 1%. This is reasonable considering that H element in (Y, Gd)H2 might come from

20 9/4

10/3

0

Gd/Y (wt.%) ratio Fig. 3 Variations of area fractions of the second phases and the grain size with Gd/Y (wt%) ratio in both as–cast and as-T4 states

the environment or the melting process which were the same for the three alloys. These results also indicate that the formation of cubic RE hydride has a great dependence on the H element and cannot be decreased or eliminated by adjusting RE contents. Moreover, it also can be seen from Fig. 3 that the grain size of the three alloys exhibits small fluctuations within 50–68 lm and 85–68 lm for as–cast and as–T4 alloys, respectively.

Mechanical Properties The engineering stress–engineering strain curves of the three alloys in T4 state are shown in Fig. 4a, while the corresponding mechanical properties as a function of Gd/Y ratio values are displayed in Fig. 4b. As can be seen, Gd/Y ratio would not influence the yield strength (YS) of the as-T4 alloys, when keeping the alloying addition same. However, the elongation (EL) together with the ultimate tensile strength (UTS) exhibits increasing trend as Gd/Y ratio increases. Especially, the EL of GW76 alloy is only about 3.4%, while that of GW94 and GW103 alloys is 6.2 and 6.5%, respectively. It has been illustrated that the three Mg–Gd–Y alloys with different Gd/Y ratio (7/6–10/3) have similar microstructures in as-T4 state, consisting of a-Mg and trace (0.63–0.36%) cubic RE hydrides with similar grain sizes (85–68 lm). With neglecting the small difference due to cubic phase and grain size, the approximate yield strength of the as-T4 alloy indicated that Gd and Y have similar solid solution strengthening effect. Previously, it is widely accepted that the formation of cubic RE hydrides is neither expected nor preferred for Mg– RE alloys [17]. As hard particles, the cubic phases cause stress concentration and act as nucleation sites for the crack initiation, thus reducing the ductility of the Mg–RE alloy [12]. The slightly higher cubic phase might contribute to the lower ductility in GW76 alloy, but should not the main reason. Actually, the different contents of Gd and Y in the

54

J. L. Li et al.

Fig. 4 a Engineering stress– engineering strain curves and b the corresponding mechanical properties of the Mg–Gd–Y alloys (Gd + Y = 13 (wt%)) with different Gd/Y ratios

matrix should be responsible for the difference in ductility. Using the Cauchy pressures as an indicator of the bonding character, it was suggested that RE enhances the brittleness of Mg [18]. The tensile stress–strain curves for the Mg–Y and Mg–Gd alloys at room temperature investigated by Gao [19] showed that the elongation of Mg–Y is less than that of Mg–Gd alloy with approximate solute atomic contents. It seems that the adverse effect of Y on ductility is more serious than Gd. Consequently, it is reasonable to observe that the ductility of the investigated three alloys in as-T4 state increased with the increase of Gd/Y ratio. As an evidence, more smooth and coarse cleavage planes were observed in the fracture of GW76–T4 than GW103–T4, as can be seen from Fig. 5. Simultaneously, deep dimples at the sites of cubic phases were observed in the fracture surface of GW103–T4, while not in GW76–T4.

Fig. 5 Representative fracture morphology of a, b GW103 and c, d GW76 alloys in T4 states; b and d is the high magnification view of the rectangle region in (a) and (c), respectively

Figure 6 shows the ageing hardening curves of the three Mg–Gd–Y alloys (Gd + Y = 13 (wt%)) at 200 and 250 °C. As can be seen, the GW76 alloy has the highest peak-hardness compared with GW94 and GW103 at both 200 and 250 °C. For all the three alloys, the peak hardness was smaller and obtained in a much shorter time when the ageing temperature increased from 200 to 250 °C. Figure 7 displays the engineering stress–engineering strain curves and the corresponding mechanical properties of the three Mg– Gd–Y alloys after peak-ageing at 200 and 250 °C. As can be seen from Fig. 7a, at the early stage of the stress–strain curves were almost overlapped, corresponding to the similar yield strength values of the investigated alloys; the highest is 236 MPa, while the lowest is 219 MPa, obtained in GW76 and GW103 both peak-aged at 250 °C, respectively. Although the increase of strength with Gd/Y ratio decreasing

Vickers hardness, HV

Effects of Gd/Y Ratio on the Microstructures … 140 135 130 125 120 115 110 105 100 95 90 85 80 75 70 0.01

55

Conclusions

GW76 GW94 GW103

This work investigated the influence of Gd/Y ratio on the microstructure and mechanical properties of Mg–Gd–Y alloys by keeping the same total RE contents. The main conclusions are as follows: Solid lines Dash lines

0.1

1

10

200 250

100

1000

Aging time, h

Fig. 6 Ageing hardening curves of the three Mg–Gd–Y alloys (Gd + Y = 13 (wt%)) at 200 and 250 °C

was slight, the decrease of ductility was apparent. As can be seen, the elongation of GW76 alloy in T6 states is lower than 1%, while that of GW103–T6 is about 2.6%. It seems that both as-T4 and as-T6 alloys had similar relationship of mechanical properties and Gd/Y ratio. The result of the as-T6 alloys further supported that Gd and Y have similar strengthening effect but different adverse effect on ductility.

(1) The phase constituents of the investigated Mg–xGd– yY–0.5Zr alloys (x + y = 13 wt%) are not affected by Gd/Y ratio. They are a-Mg, Mg24(Gd, Y)5 and Mg5(Gd, Y) in the as-cast state, while a-Mg and trace cubic (Y, Gd)H2 in the as-solutionized state. (2) With Gd/Y ratio increasing, the volume fraction of second phases in the as-cast alloys decreases, while Gd/ Y ratio has little influence on the insoluble cubic phases of the alloys in as-solutionized state. (3) Mg–7Gd–6Y–0.5Zr alloy with the lowest Gd/Y ratio exhibited similar yield strength but much lower ductility compared with that of the Mg–9Gd–4Y–0.5Zr and Mg–10Gd–3Y–0.5Zr.

Acknowledgements This work was funded by the National Key Research and Development Program of China through Project No. 2016YFB0301104, the National Natural Science Foundation of China (NSFC) through Projects No. 51531002, No. 51301173, No. 51601193 and No. 51701218, the National Science and Technology Major Project of China through Project No. 2017ZX04014001, and the National Basic Research Program of China (973 Program) through Project No. 2013CB632202.

References

Fig. 7 a Stress–strain curves and b the mechanical properties of the Mg–Gd–Y alloys after different ageing treatments

1. B.L. Mordike, T. Ebert, Materials Science and Engineering: A, 302 (2001) 37–45. 2. B. Smola, I. Stulıková, F. von Buch, B.L. Mordike, Materials ́ Science and Engineering: A, 324 (2002) 113–117. 3. R.G. Li, J.F. Nie, G.J. Huang, Y.C. Xin, Q. Liu, Scripta Materialia, 64 (2011) 950–953. 4. M. Suzuki, H. Sato, K. Maruyama, H. Oikawa, Materials Science and Engineering: A, 252 (1998) 248–255. 5. G.W. Lorimer, P.J. Apps, H. Karimzadeh, J.F. King, Materials Science Forum, 419(2003) 279–284. 6. S.B. Li, W.B. Du, X.D. Wang, K. Liu, Z.H. Wang, Acta Metallurgica Sinica (English Letters), 54 (2018) 911–917. 7. M.E. Drits, Z.A. Sviderskaya, L.L. Rokhlin, N.I. Nikitina, Metal Science and Heat Treatment, 21 (1979) 887–889. 8. J.F. Nie, X. Gao, S.M. Zhu, Scripta Materialia, 53 (2005) 1049–1053. 9. I.A. Anyanwu, S. Kamado, Y. Kojima, Materials Transactions, 42 (2001) 1206–1211. 10. S.M. He, X.Q. Zeng, L.M. Peng, X. Gao, J.F. Nie, W.J. Ding, Journal of Alloys and Compounds, 427 (2007) 316–323. 11. V. Janik, D.D. Yin, Q.D. Wang, S.M. He, C.J. Chen, Z. Chen, C.J. Boehlert, Materials Science and Engineering: A, 528 (2011) 3105–3112.

56 12. H.R.J. Nodooshan, W. Liu, G. Wu, Y. Rao, C. Zhou, S. He, W. Ding, R. Mahmudi, Materials Science and Engineering: A, 615 (2014) 79–86. 13. J. Wang, J. Meng, D. Zhang, D. Tang, Materials Science and Engineering: A, 456 (2007) 78–84. 14. I.P.D.S.J.–F.N.M. Qian, Light Alloys: Metallurgy of the Light Metals, 2017. 15. J.-L. Li, N. Zhang, X.-X. Wang, D. Wu, R.-S. Chen, Acta Metallurgica Sinica (English Letters), 31 (2018) 189–198.

J. L. Li et al. 16. X. Liu, Z. Zhang, Q. Le, L. Bao, Journal of Magnesium and Alloys, 4 (2016) 214–219. 17. Y. Huang, L. Yang, S. You, W. Gan, K.U. Kainer, N. Hort, Journal of Magnesium and Alloys, 4 (2016) 173–180. 18. K. Chen, K. Boyle, Metallurgical and Materials Transactions A, 40(2009) 2751–2760. 19. Gao L (2010) Composition design, solid solution strengthening and precipitation strengthening mechanisms in high strength cast Mg–Gd–Y–Zr Alloys. Ph.D. thesis, Chinese Academy of Sciences.

Part III Magnesium Technology 2019: Thermomechanical Processing

Evolution of Heterogeneous Microstructure of Equal-Channel Angular Pressed Magnesium Qizhen Li

Abstract

This study investigated the distribution variations of the 0002 poles of magnesium after experienced a series of numbers of equal-channel angular pressing (ECAP) passes through either the Bc route or the C route. For the ECAP processing, the mold channel angle was 142° and the sample temperature was 200 °C. The results show that most 0002 poles are in the ring of about 60–80° away from the ED for all ECAPed samples, and the location of the maximum intensity oscillates in the 60–80° range. The first ECAP pass increased the nonuniformity of the distribution of 0002 poles, while the following ECAP passes resulted in more uniform distribution of 0002 poles in the 60–80° ring and reduced texture intensity through decreasing the maximum intensity and the intensity range of 0002 poles with the increase in the number of ECAP passes for both Bc and C routes. Keywords

Equal-channel angular pressing




Magnesium




Texture

Introduction There is a strong motivation to reduce the weight of transportation vehicles through the application of lightweight materials, and thus to improve fuel efficiency and lower waste gas emission. Magnesium is reported to be the lightest structural metals and can be an ideal candidate for the automobile applications [1–4]. To take the full advantage of the low density of magnesium, it is critical to improve the absolute strength. Among the various strategies such as adding alloying elements or reinforcing phases [5–18] and inputting heavy deformation [2, 19–23] for achieving Q. Li (&) School of Mechanical and Materials Engineering, Washington State University, Pullman, WA, USA e-mail: [email protected]

strength improvement, ECAP is one of the most popular method to severely deform the material and modify the microstructure to control the properties [24]. During ECAP, a sample can be ECAPed repeatedly without intermediate machining between the consecutive ECAP passes. ECAP was also extensively utilized to process magnesium and its alloys such as pure magnesium [21, 25], AZ91 [26], ZA62 [27], and ZK60 [28]. In general, the microstructure of processed samples is heterogeneous. Although most of the reported research provided the texture of the ECAPed samples, the analysis on the variation of intensity of texture is still lacking. This study aims to explore the change of the 0002 pole intensity due to the ECAP processing of pure magnesium. Two different ECAP routes (i.e. Bc and C) were used in the processing, and the number of ECAP passes were to be 1, 2, 4, and 8 for each ECAP route. The 0002 pole figures were obtained by the X-ray diffraction technique.

Experimental Commercially pure magnesium bars were cut into a cubic shape for ECAP processing. The geometry of these ECAP samples is 10 mm  10 mm  35 mm since the open channels of the ECAP mold have a square cross section of 10.5 mm  10.5 mm. The direction along the open channel is the extrusion direction (ED). Before the samples were pressed through the ECAP mold, they experienced heat treatment at 300 °C for 12 h. During an individual ECAP pass, the ECAP mold was heated to 50 °C and the sample was heated to 200 °C. For multiple-pass ECAP processing, both Bc route and C route were employed. For the Bc route, a sample was rotated by 90° with respect to the ED between two consecutive passes. For the C route, a sample was rotated by 180° with respect to the ED between two consecutive passes. Between any two continuous ECAP passes, a sample was heated at 200 °C for 10 min. This study investigated the samples experienced 1, 2, 4, and 8 ECAP

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_10

59

60

Q. Li

passes along the Bc and C routes, respectively. For each sample, the cross section perpendicular to the ED was prepared and observed through the X-ray diffraction (XRD) technique to investigate the texture evolution with the ECAP processing.

Results and Discussion The sample before being processed by ECAP (i.e. experienced zero ECAP pass) can be referred to as the unECAPed sample. The distribution of 0002 poles for this sample is reported in the polar coordinate system in Fig. 1a, and the ED is perpendicular to the polar coordinate system. The maximum intensity of 0002 poles is 282 as labeled in Fig. 1a and locates at about 60° away from the ED. The high intensity values fall in a ring that is 60 * 80° away for the ED. Figure 1b shows the geometry of an ECAP sample, and the ED is parallel to the 35 mm length direction. Figure 1c illustrates the orientation relation (60°) between the ECAP pressing force and the 0002 pole with the maximum intensity. The hexagonal close packing unit cell can rotate around the 0002 pole without changing this orientation relation. Figure 2 presents a schematic ECAP mold with a angle of 142° between the exit and entrance channels employed in this study. A sample was pressed through the channels under the pressing force. The green region shows the two-dimensional sample passing through the connection of the two channels. The sample experienced one ECAP pass for the Bc route is the same as that experienced the C route. The 0002 pole figure for this sample is provided in Fig. 3. The maximum intensity is 276 and about 75° away from the ED. Figure 4 reports the

(a)

(b)

Fig. 2 Sketch of ECAP mold with a channel angle of 142° between the two channels

0002 pole figures for the samples after being ECAPed for 2, 4, and 8 passes along the Bc route and the variation of maximum intensity with the increase in the number of ECAP passes. Figure 5 reports the 0002 pole figures for the samples after being ECAPed for 2, 4, and 8 passes along the C route and the variation of maximum intensity with the increase in the number of ECAP passes. For all ECAPed samples, it is clear that most 0002 poles are still in the ring of about 60–80° away

(c)

Fig. 1 a 0002 pole figure, b the geometry, and c the orientation relation between the ECAP pressing force and the 0002 pole with the maximum intensity for the samples after being heat treated at 300 °C for 12 h and before experiencing ECAP processing

Evolution of Heterogeneous Microstructure of Equal-Channel …

Fig. 3 0002 pole figure of the sample after one ECAP pass

from the ED. The maximum intensities are 70°, 75°, and 70° away from the ED for 2, 4, and 8 passes, respectively, for the Bc route. Similarly, the maximum intensities are 70°, 70°, and 80° away from the ED for 2, 4, and 8 passes, respectively, for the C route. Therefore, the location of the maximum intensity oscillates in the 60–80° range. Fig. 4 0002 pole figures of the samples after a 2, b 4, and c 8 ECAP passes along the Bc route. d The variation of maximum intensity for 0002 poles with the increase in the number of ECAP passes along the Bc route

61

Figure 6a provides the variations of maximum intensity with the increase in the number of ECAP passes for both Bc and C routes. After one ECAP pass, the maximum intensity decreased from 282 to 276. For the Bc route, this maximum intensity first increased to 281 after 2 passes and then decreased linearly to 271 after 4 passes and 253 after 8 passes. For the C route, this maximum intensity first decreased to 263 after 2 passes and then increased to 273 after 4 passes and decreased to 262 after 8 passes. In addition to the maximum intensity for each sample, the minimum intensity was collected and the intensity range was computed by subtracting the minimum intensity from the maximum intensity for an individual sample. Figure 6b shows the variations of the intensity range with the increase in the number of ECAP passes for both Bc and C routes. The unECAPed sample had the highest maximum intensity and some intermediate intensity range, which indicates that the 0002 poles were distributed in a narrow range and the sample had a strong texture. After the first ECAP pass, the maximum intensity decreased and the intensity range increased compared with the unECAPed sample. This phenomenon implies that some 0002 poles in the initial maximum intensity of the unECAPed sample rotated due to

(a)

(b)

(c)

(d)

62 Fig. 5 0002 pole figures of the samples after a 2, b 4, and c 8 ECAP passes along the C route. d The variation of maximum intensity for 0002 poles with the increase in the number of ECAP passes along the C route

Q. Li

(a)

(b)

(c)

(d)

Fig. 6 a The maximum intensity and b the intensity range of 0002 poles of the samples after 0, 1, 2, 4, and 8 ECAP passes along the Bc and C routes

twinning to broaden the intensity range and make the 0002 poles distributed more nonuniformly. After the first ECAP pass, the intensity range decreased with the increase in the number of ECAP passes for both Bc and C routes, which suggests that the 0002 poles distributed more and more uniformly in the space. The overall trend for the maximum intensity is also decreasing for both Bc and C routes, which indicates that the texture intensity was reduced.

Conclusions ECAP was employed to process the magnesium samples through two different routes (i.e. the Bc and C routes) for different number (i.e. 1, 2, 4, and 8) of passes, respectively. The used mold has a 142° channel angle, and the sample temperature is 200 °C. The 0002 pole figures were obtained for the

Evolution of Heterogeneous Microstructure of Equal-Channel …

unECAPed and all ECAPed samples using the XRD technique. The analysis of the pole figures provided the following knowledge. The unECAPed sample had the highest maximum intensity and some intermediate intensity range to possess a strong texture. The first ECAP pass reduced the maximum intensity and increased the intensity range compared with the unECAPed sample, because of twinning to rotate some 0002 poles away from the initial maximum intensity location and broaden the intensity range to make the 0002 poles distributed more nonuniformly. After the first ECAP pass, the decrease of the intensity range implies the more and more uniform distribution of 0002 poles with the increase in the number of ECAP passes for both Bc and C routes. Acknowledgements The author appreciates the financial support for this work from the Basic Energy Sciences Office at the US Department of Energy under Award no. DESC0016333.

References 1. M.M. Avedesian, H. Baker, ASM specialty handbook: magnesium and magnesium alloys, ASM international Materials Park, OH1999. 2. Q. Li, B. Tian, Mechanical properties and microstructure of pure polycrystalline magnesium rolled by different routes, Materials Letters 67(1) (2012) 81–83. 3. Q. Li, Dynamic mechanical response of magnesium single crystal under compression loading: Experiments, model, and simulations, Journal of Applied Physics 109(10) (2011) 103514. 4. Q. Li, Mechanical properties and microscopic deformation mechanism of polycrystalline magnesium under high-strain-rate compressive loadings, Materials Science and Engineering: A 540 (2012) 130–134. 5. M. Habibnejad-Korayem, R. Mahmudi, W. Poole, Enhanced properties of Mg-based nano-composites reinforced with Al 2 O 3 nano-particles, Materials Science and Engineering: A 519(1) (2009) 198–203. 6. H. Cay, H. Xu, Q. Li, Mechanical behavior of porous magnesium/alumina composites with high strength and low density, Materials Science and Engineering: A 574 (2013) 137–142. 7. Q. Li, Carbon nanotube reinforced porous magnesium composite: 3D nondestructive microstructure characterization using x-ray micro-computed tomography, Materials Letters 133 (2014) 83–86. 8. S. Hassan, M. Gupta, Effect of different types of nano-size oxide particulates on microstructural and mechanical properties of elemental Mg, Journal of Materials Science 41(8) (2006) 2229–2236. 9. Q. Li, Effect of porosity and carbon composition on pore microstructure of magnesium/carbon nanotube composite foams, Materials & Design 89 (2016) 978–987. 10. S.K. Thakur, G.T. Kwee, M. Gupta, Development and characterization of magnesium composites containing nano-sized silicon carbide and carbon nanotubes as hybrid reinforcements, Journal of Materials Science 42(24) (2007) 10040–10046. 11. N. Zou, Q. Li, Compressive mechanical property of porous magnesium composites reinforced by carbon nanotubes, Journal of Materials Science 51(11) (2016) 5232–5239.

63 12. H. Xu, N. Zou, Q. Li, Effect of Ball Milling Time on Microstructure and Hardness of Porous Magnesium/Carbon Nanofiber Composites, JOM (2017) 1–8. 13. H. Fukuda, K. Kondoh, J. Umeda, B. Fugetsu, Fabrication of magnesium based composites reinforced with carbon nanotubes having superior mechanical properties, Materials Chemistry and Physics 127(3) (2011) 451–458. 14. H. Xu, Q. Li, Effect of carbon nanofiber concentration on mechanical properties of porous magnesium composites: Experimental and theoretical analysis, Materials Science and Engineering: A 706 (2017) 249–255. 15. Q. Li, B. Tian, Compression behavior of magnesium/carbon nanotube composites, Journal of Materials Research 28(14) (2013) 1877–1884. 16. K. Kondoh, H. Fukuda, J. Umeda, H. Imai, B. Fugetsu, M. Endo, Microstructural and mechanical analysis of carbon nanotube reinforced magnesium alloy powder composites, Materials Science and Engineering: A 527(16) (2010) 4103–4108. 17. B. Tian, Z.G. Cheng, Y.X. Tong, L. Li, Y.F. Zheng, Q. Li, Effect of enhanced interfacial reaction on the microstructure, phase transformation and mechanical property of Ni–Mn–Ga particles/Mg composites, Materials & Design 82 (2015) 77–83. 18. B. Tian, F. Chen, Y.X. Tong, L. Li, Y.F. Zheng, Y. Liu, Q. Li, Phase transition of Ni–Mn–Ga alloy powders prepared by vibration ball milling, Journal of Alloys and Compounds 509 (13) (2011) 4563–4568. 19. G.S. Vinod Kumar, M. Mukherjee, F. Garcia-Moreno, J. Banhart, Reduced-Pressure Foaming of Aluminum Alloys, Metallurgical and Materials Transactions A 44(1) (2013) 419–426. 20. X. Fan, X. Chen, X. Liu, H. Zhang, Y. Li, Bubble Formation at a Submerged Orifice for Aluminum Foams Produced by Gas Injection Method, Metallurgical and Materials Transactions A 44 (2) (2013) 729–737. 21. X. Jiao, Q. Li, An observation about global microstructure of ECAPed magnesium, Emerging Materials Research 3(6) (2014) 261–264. 22. Z. Cao, Y. Yu, M. Li, H. Luo, Cell Structure Evolution of Aluminum Foams Under Reduced Pressure Foaming, Metallurgical and Materials Transactions A 47(9) (2016) 4378–4381. 23. Q. Li, X. Jiao, Exploration of equal channel angular pressing routes for efficiently achieving ultrafine microstructure in magnesium, Materials Science and Engineering: A 733 (2018) 179–189. 24. M. Furukawa, Y. Iwahashi, Z. Horita, M. Nemoto, T.G. Langdon, The shearing characteristics associated with equal-channel angular pressing, Materials Science and Engineering: A 257(2) (1998) 328–332. 25. F.S.J. Poggiali, C.L.P. Silva, P.H.R. Pereira, R.B. Figueiredo, P.R. Cetlin, Determination of mechanical anisotropy of magnesium processed by ECAP, Journal of Materials Research and Technology 3(4) (2014) 331–337. 26. J. Gubicza, K. Máthis, Z. Hegedűs, G. Ribárik, A.L. Tóth, Inhomogeneous evolution of microstructure in AZ91 Mg-alloy during high temperature equal-channel angular pressing, Journal of Alloys and Compounds 492(1) (2010) 166–172. 27. K. Yan, Y.-S. Sun, J. Bai, F. Xue, Microstructure and mechanical properties of ZA62 Mg alloy by equal-channel angular pressing, Materials Science and Engineering: A 528(3) (2011) 1149–1153. 28. Y. He, Q. Pan, Y. Qin, X. Liu, W. Li, Microstructure and mechanical properties of ultrafine grain ZK60 alloy processed by equal channel angular pressing, Journal of materials science 45(6) (2010) 1655–1662.

Novel Magnesium Alloy Processing via Shear-Assisted Processing and Extrusion (ShAPE) S. Mathaudhu, N. Overman, S. Whalen, M. Olszta, D. Catalini, K. Kruska, J. Darsell, V. Joshi, X. Jiang, A. Devaraj, and G. Grant

Abstract

Traditional magnesium alloy manufacturing has often relied on processing within a temperature range that results in liquid phase formation of some of the constituents. These methods are often limited by the equilibrium phase formation states available from the melt. In this work, we will present findings on a recently developed processing approach that enables complex, unique microstructural evolution (often to persistent metastable states) while remaining in the solid phase state. Specifically, the shear assisted processing and extrusion (ShAPE) method as applied to Mg alloys will S. Mathaudhu (&)
A. Devaraj University of California, Riverside, 3401 Watkins Dr, Riverside, USA e-mail: [email protected] A. Devaraj e-mail: [email protected] S. Mathaudhu
N. Overman
S. Whalen
M. Olszta
D. Catalini
K. Kruska
J. Darsell
V. Joshi
X. Jiang
G. Grant Pacific Northwest National Laboratory, 902 Battelle Blvd, Richland, USA e-mail: [email protected] S. Whalen e-mail: [email protected] M. Olszta e-mail: [email protected] D. Catalini e-mail: [email protected] K. Kruska e-mail: [email protected] J. Darsell e-mail: [email protected] V. Joshi e-mail: [email protected] X. Jiang e-mail: [email protected] G. Grant e-mail: [email protected]

be presented. Novel microstructural pathways, textural formation and mechanical properties will be discussed. These results point to the ability to design and engineering novel Mg materials with unprecedented properties and performance. Keywords

Magnesium




Extrusion




Shear




ShAPE




Friction

Introduction Extruded magnesium products have been implemented for a variety of applications due to their low density and availability. Most commonly, the billets begin as castings, which are heated to enable the uniform flow necessary for extruding the alloys into bulk products, such as rods and tubes. The result of the elevated temperatures is frequently the lack of ability to control the microstructural evolution (e.g. crystallographic texture, grain size, second phase particle distribution and morphology) in ways that maximize the desired outcome performance. Thus, novel approaches are necessary to circumvent the constraints of conventional extrusion technologies to allow a higher level of tailorability of properties. Solid phase processing (SPP) methods are promising to realize this vision. SPP approaches, akin to friction stir processing or welding, plastically deform the material in the solid state through the impingement of high shear at high strain rates. Three key conditions that enable new pathways to microstructural evolution are: (1) The plastic work of deforming the material generates the heat needed for material flow. In general, no external heating is required. (2) The high shear strains enable alternative diffusion and advection pathways for the movement of atoms, and

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_11

65

66

thus, novel microstructures and phases that are away from equilibrium may be realized. (3) Only a small volume of material is heated and deformed, therefore facilitating process intensification in a small region. The rapid heat-up and cool-down before and after may allow locking-in of curious microstructural states. The combination of these conditions represents a drastic departure from conventional extrusion processing methodologies and thus the opportunity for entirely new pathways to beneficial microstructural states with step increases in properties and performance.

Results There are many examples of processes that meet these conditions, such as friction stir process/welding, impact welding, and cold-spray, however, this paper will center on a new method termed shear-assisted processing and extrusion (ShAPETM) (Fig. 1). The ShAPE process drives a rotating bulk ingot or powders into a spiral-patterned die. At the interface, the material temperature increases to enable flow (and in the case of powder/particulates, densification) along the spirals into bulk products. The process intensification further minimizes overall energy consumption due to the low forces needed to only deform a small volume, and reduced external heating (if any) of the processed materials. The extant literature reveals a number of successful research implementations of this method. Joshi et al. [1] were able to demonstrate the ability of SPP conditions to refine grain structures and control particle phase dispersions and morphologies in commercial ZK60A-T5 and Mg–2 wt%Si (Mg– 2Si) alloy billets. Small tubes (nominal 7.45 mm outer diameter and 1.61 mm wall thickness) were produced. Whalen et al. [2] were able to demonstrate scalability of this method to produce a ZK60 alloy tube (50.8 mm outer diameter and 1.52 mm wall

Fig. 1 Schematic of the ShAPE process for the production of wire, rod or tube extrusions from bulk materials or particulates

S. Mathaudhu et al.

thickness) from ZK60A-T5 billets. The final grain size was very fine and was measured to 3.8 lm due to the rapid heating and cooling during the process. ShAPE has also been proven to effectively consolidate particulate materials into bulk products with similar features to those produced from bulk ingot. Overman et al. [3] densified AZ91E melt-spun flakes and studied the microstructural evolution in-depth. Their work showed the ability to control novel textures and attain uniform grain refinement and homogenization. As with the processing of the bulk materials [1, 2], lower extrusion forces were needed as compared to conventional billet extrusions. While these results point to promising avenues of attaining novel properties in Mg alloys, there yet remains room for improvements, many of which involve processing refinement and control, and tool materials, features and processing parameters, as shown by Darsell et al. [4]. Key conditions include RPM and torque (which controls heat generation, grain refinement and texture), ram speed, extrusion force, die face geometry (e.g. spiral flow enhancement) and boundary conditions (e.g. active cooling or atmospheric control).

Discussion These preliminary results on Mg alloys point to unprecedented control over microstructural and textural features that strongly affect the end properties and performance. Namely, they offer the opportunity for increasing the strength and ductility that have somewhat limited the implementation of Mg alloys in applications that use materials with high densities (e.g. Al and Fe alloys for transportation). Benefits such as reduced energy costs (due to lower forces and temperatures), single-step processing from ingot/particulate to rods/wires/tubes, and scalability powerfully point to acceleration applicability; however, some key scientific and technical opportunities exist: (1) The pathways to achieve desirable microstructural states (particularly pervasively metastable ones) are unknown and, given the complexity of the deformation state, will require targeted studies, such as beamline experiments or other marker studies that track microstructural evolution. (2) Thus far, usage of advanced computational tools at length and timescales needed to understand and control microstructural evolution have yet to be explored. Of key interests are the mechanisms for atomic transport under high shear, high strain rate events. (3) Opportunities exist for instrument and tool design, both from materials and engineering perspective. These, along with enhanced sensing and control algorithms, will hasten the likelihood of industrial adoption.

Novel Magnesium Alloy Processing via Shear-Assisted Processing…

67

(4) Because microstructure controls properties, including those beyond mechanical response, it will be essential to explore aspects of the properties, including for Mg alloys, corrosion and reactivity that will concurrently need to be improved to see further use in industry.

in part by the National Science Foundation under Grant no. 1463679. A portion of the research was performed using EMSL, a DOE Office of Science User Facility sponsored by the Office of Biological and Environmental Research.

References Summary Preliminary ShAPE research efforts have shown the possibility for disruptive control of Mg-based microstructure and the resulting properties. These are facilitated by the unique underlying deformation physics that solid phase processing (SPP) offers, namely high shear/high shear strain-induced heating, alternative kinetic and thermodynamic mass transport and process intensification. Scientific and technical opportunities exist to maximize the potential of this approach to for step change improvements in mechanical and functional properties, and given these break-through advancements, the ShAPE process is poised to serve as an energy-effecient and scalable approach for the manufacture of Mg alloys and other high-value advanced materials. Acknowledgements Partial financial support of this work was enabled through the MS3 (Materials Synthesis and Simulation across Scales) Initiative at the Pacific Northwest National Laboratory, a multi-program US Department of Energy Laboratory operated by Battelle under contract DE-AC05-76RL01830. The authors would also like to thank the US Department of Energy Vehicle Technologies Office and Office of Fuel Cell and Vehicle Technologies. S.N. Mathaudhu was supported

1. V.V. Joshi, S. Jana, D.S Li, H. Garmestani, E. Nyberg, C. Lavender, “High Shear Deformation to Produce High Strength and Energy Absorption in Mg Alloys”, in: M. Alderman, M.V. Manuel, N. Hort and N.R. Neelameggham (eds), Magnesium Technology 2014, The Minerals, Metals and Materials Series, Springer, Cham, 83–88, 2014. https://doi.org/10.1007/978-3-319-48231-6_19 2. S. Whalen, V. Joshi, N. Overman, D. Caldwell, C. Lavender, T. Szszek, “Scaled-up Fabrication of Thin-Walled ZK60 Tubing using Shear Assisted Processing and Extrusion (ShAPE)”, in: K. Solanki, D. Orlov, A. Singh and N.R. Neelameggham (eds), Magnesium Technology 2017, The Minerals, Metals and Materials Series, Springer, Cham, 315–321, 2017. https://doi.org/10.1007/978-3-31952392-7_45 3. N.R. Overman, S.A. Whalen, M.E. Bowden, M.J. Olszta, K. Kruska, T. Clark, E.L. Stevens, J.T. Darsell, V.V. Joshi, X. Jiang, K.F. Mattlin and S.N. Mathaudhu, “Homogenization and texture development in rapidly solidified AZ91E consolidated by Shear Assisted Processing and Extrusion (ShAPE)”, Materials Science and Engineering, A. Vol. 701, pp. 56–68, 2017. https://doi.org/10. 1016/j.msea.2017.06.062 4. J.T. Darsell, N.R. Overman, V.V. Joshi, S.A. Whalen and S.N. Mathaudhu, “Shear Assisted Processing and Extrusion (ShAPETM) of AZ91E Flake: A Study of Tooling and Features and Processing Effects, Journal of Materials Engineering and Performance, Vol 27 (8), pp. 4150–4161, 2018. https://doi.org/10.1007/s11665-0183509-1

Effects of the Extrusion Temperature on Microstructure, Texture Evolution and Mechanical Properties of Extruded Mg–2.49Nd–1.82Gd–0.19Zn–0.4Zr Alloy Lei Xiao, Guangyu Yang, Shifeng Luo, and Wanqi Jie

Abstract

Microstructure, texture evolution and mechanical properties of extruded Mg–2.49Nd–1.82Gd–0.19Zn–0.4Zr alloy were investigated at extrusion temperatures of 260 °C, 280 °C, 300 °C and 320 °C, with an extrusion ratio of 15 and RAM speed of 3 mm s−1, respectively. The results indicated that the coarse grains of homogenized billets were substantially refined after the extrusion process, which was caused by the refinement of dynamically recrystallization (DRX) and the pinning effect of precipitated Mg5Gd and Mg12(Nd, Gd) particles. The grain size decreased gradually when the extrusion temperature increased from 260 to 280 and 300°C, and then coarsened slightly once the extrusion temperature further increased to 320 °C. Moreover, the DRX process was promoted with the increasing extrusion temperature, and a completely DRX microstructure could be obtained when the extrusion temperature up to 300 °C. The room temperature tensile and compressive yield strength increased when the temperature increased from 260 to 300 °C and then decreased at 320 °C. All extruded alloys exhibited an extremely low tension–compression yield asymmetry, which was mainly attributed to the rare earth (RE) texture component as well as the fine microstructure developed during the extrusion process. Keywords










Magnesium alloy Extrusion temperature Texture analysis Tension–compression yield asymmetry

L. Xiao
G. Yang (&)
S. Luo
W. Jie State Key Laboratory of Solidification Processing, Northwestern Polytechnical University, No. 127 Youyi Western Road, Xi’an 710072, Shaanxi, People’s Republic of China e-mail: [email protected]

Introduction For their excellent specific strength [1], high formability [2], and high creep resistance [3], Mg–RE alloys are promising for further technical applications in aeronautics and automobile constructions. RE element addition on the microstructure transformation has been extensively studied by many literatures. Among them, it was reported by Ning et al. [4] that the addition of Nd in Mg exhibited a strong age-hardening response and an improvement in the tensile yield strength (TYS) after adequate heat treatments. Moreover, a trace addition of Zn [5] and Zr [6] elements into Mg– Nd-based alloys was also found to further enhance the mechanical properties. Recently, in order to further improve the heat resistance of Mg–Nd–Zn–Zr alloys, Gd was added due to the formation of high thermal stability secondary phases. These works suggested that developing Mg–Nd–Zn– Zr-based alloys with Gd addition would make a significance sense to the engineering applications in the future. In our previous research [7, 8], Mg–2.5Nd–1.5Gd– 0.2Zn–0.6Zr (all concentrations in wt%, unless otherwise stated) casting alloy was developed, which showed a comparative tensile strength and creep resistance, and it was very promising for further industrial application. However, Mg– 2.5Nd–1.5Gd–0.2Zn–0.6Zr casting alloy exhibited a poor ductility at room temperature, insufficient for the high load-bearing structural components. Fortunately, extrusion process was considered as an effective method to improve both the strength and the ductility of Mg alloys [9]. In this study, microstructure and mechanical properties of Mg–2.5Nd–1.5Gd–0.2Zn–0.6Zr alloy at different extrusion temperatures (260, 280, 300 and 320 °C) with an extrusion ratio of 15 and RAM speed of 3 mm s−1 are investigated. The correlation between the texture and tension–compression yield asymmetry is also analyzed, which may thereby provide a beneficial information to enlarge the industrial application of Mg–2.5Nd–1.5Gd–0.2Zn–0.6Zr alloy.

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_12

69

70

Experimental Procedures The experimental alloy with a nominal composition of Mg– 2.5Nd–1.5Gd–0.2Zn–0.6Zr was prepared from pure Mg, pure Zn, Mg–33Nd, Mg–28Gd and Mg–33Zr master alloys and melt in an electrical resistance furnace under a protective atmosphere (CO2 + SF6). After refining with C2Cl6 and holding at 780 °C for 20 min, the melt was poured into a steel mold which was preheated to 200 °C. The chemical composition of the casting billet (U130 mm  160 mm) was determined to be Mg–2.49Nd–1.82Gd–0.19Zn–0.4Zr by using an inductively coupled plasma analyzer (Perkin Elmer Plasma 400). The billets were homogenized at 515 °C for 18 h and then machined to cylindrical samples with 100 mm in diameter and 150 mm in height. These samples were then extruded into bars with an extrusion ratio of 15 and RAM speed of 3 mm/s at extrusion temperatures of 260 °C, 280 °C, 300 °C and 320 °C, respectively. The microstructure was analyzed with OLYMPUS-GX7 optical microscopy (OM) and JSM-6390A scanning electron microscope (SEM) equipped with energy dispersive spectrometer (EDS, Oxford Inca) operating at 20 kV. The details of preparation of the samples for the microstructure observation referred to the previous report [10]. The DRX fraction was obtained by measuring the microstructure on five micrographs of each extruded alloy; meanwhile, the average grain size was calculated five times by using the linear intercept method. Texture measurement was performed on the longitudinal section of extruded alloys by using a JEOL JSM-7000F SEM equipped with TSL electron backscattered diffraction (EBSD) operating at 15 kV. To ensure statistical rigor, more than 1000 grains were examined. Transmission

Fig. 1 a OM image, b SEM image of the as-cast experimental alloy, c EDS result of the position A, d EDS result of the position B

L. Xiao et al.

electron microscopy (TEM) observation was carried out using a Technai 30F. Thin foil specimens about 0.8 mm thick for TEM observations were cut from the transverse direction (TD), and they were ground to about 50 lm in a SiC sandpaper and prepared by ion beam milling. Then, they were examined in an acceleration voltage of 200 kV. Tensile specimens (a diameter of 6 mm and a gauge length of 25 mm) and cylindrical compressive specimens (15 mm in height and 10 mm in diameter) were both machined parallel to the extrusion direction (ED). Then, they were tested on Instron 4206 according to ASTM E8 with a rate of 1 mm/min.

Results and Discussion Figure 1a shows the optical microscope (OM) microstructure of the as-cast experimental alloy, which mainly comprised equiaxed a-Mg grains surrounded by discontinuous eutectic regions. Some gray needle-like precipitates and dotted particles are also observed within the a-Mg matrix. The corresponding SEM image is shown in Fig. 1b, and the EDS tested positions are marked by A and B, respectively. According to the EDS results in Fig. 1c, d, the eutectics along grain boundaries are determined to be Mg12(Nd, Gd, Zn). The dotted particle within the a-Mg matrix can be indexed as a-Zr, which was generally regarded as the nucleation site for a-Mg grains during solidification, leading to the grain refinement [11]. OM microstructure of the homogenized experimental alloy is shown in Fig. 2a, and a small amount of undissolved secondary phases can be observed along grain boundaries.

Effect of the Extrusion Temperature on Microstructure …

71

Fig. 2 a OM image, b SEM image, c TEM bright image of the long rod-like precipitates in (b), d EDS results of the long rod-like precipitates, e the corresponding SAED pattern of the homogenized experimental alloy

SEM image of the homogenized experimental alloy is shown in Fig. 2b, and the residual secondary phases marked by red arrows can also be identified as Mg12(Nd, Gd, Zn). In addition, as indicated by red dotted squares in Fig. 2b, the precipitates inside each grain are actually some long rod-like precipitates distributed in local regions. TEM bright image of the long rod-like precipitates is shown in Fig. 2c, and it can be seen the long rod-like precipitates are distributed within the a-Mg matrix with random orientations. These long rod-like precipitates are further identified to be Zn2Zr3 phase with tetragonal structure (a = 0.768, c = 0.699 nm, space group P42/mnm [12]) based on the corresponding EDS analysis results in Fig. 2d and the SAED pattern in Fig. 2e. The long rod-like precipitates could not be observed in the as-cast condition, and they may be formed by the diffusion reaction of Zr particles with the dissolved Zn element during the homogenization treatment, which has been proved to be true in previous works [13–15]. The optical micrographs of the extruded experimental alloys are shown in Fig. 3, which were taken from the mid-position of the longitudinal and transverse sections, respectively. It is quite clear that coarse grains in homogenized billets are substantially refined, suggesting that DRX took place during the extrusion processing, and thus the grain boundaries are found to be less resolved to the optical microscope. In addition, the secondary phases with dark contrast are oriented approximately parallel to the ED at all four extruded alloys, and the coarse elongated un-DRX grains disappear once the extrusion temperature increases to 300 °C. Figure 4a shows the SEM image of the experimental alloy extruded at 300 °C, and the broken secondary phases

(marked by white arrows) along the ED are confirmed as Mg12(Nd, Gd) phase using EDS analysis method. The local enlargement of the red dotted square region is further shown in Fig. 4b, and the image was obtained under the backscattering electron (BSE) mode to highlight the contrast between the matrix and precipitated particles. Numerous fine white particles can be clearly observed, and particles are too small to be analyzed by EDS. The corresponding TEM bright image is further shown in Fig. 4c, and fine precipitated particles are found to be randomly distributed along grain boundaries or within the matrix or in the triple grain

Fig. 3 Optical microstructures of the extruded experimental alloys

72

L. Xiao et al.

Fig. 4 Microstructures of the experimental alloy extruded at 300 °C: a SEM image, b BSE image, c TEM bright image, d, e the corresponding EDS analysis results and SAED pattern of white arrows in (c), f, g the corresponding EDS analysis results and SAED pattern of red arrows in (c)

boundary junctions. These fine precipitated particles can be divided into two groups according to their shapes: the cubic-shaped particles (marked by white arrows) and irregular-shaped particles (marked by red arrows). Based on the EDS analysis results in Fig. 4d and the corresponding SAED pattern in Fig. 4e, the cubic-shaped particle is indexed to be Mg5Gd (space group m, a = 2.2 nm) with face-centered cubic (FCC) structure [16]. The irregular-shaped particle can be confirmed as Mg12Nd with a tetragonal structure (space group I4/mmm, a = 1.031, c = 0.593 nm) from Fig. 4f, g. Figure 5 shows the variation trend of microstructure and mechanical properties of the extruded experimental alloys with the different extrusion temperatures. It can be seen from Fig. 5a that the increasing extrusion temperature is beneficial for the grain refinement when the extrusion temperature is below 300 °C. For instance, the extruded alloy with the extrusion temperature of 260 °C has the average grain size of 4.83 ± 0.41 lm and then decreases to 4.05 ± 0.26 lm and 2.31 ± 0.11 lm when the extrusion temperature increases to 280 °C and 300 °C, respectively. However, the grain coarsens slightly to 3.7 ± 0.20 lm when the extrusion temperature raises to 320 °C. In addition, the DRX fraction increases monotonically with the increasing extrusion temperature, which is in good agreement with the elimination of coarse elongated un-DRX grains as seen in Fig. 3, and a completely DRX microstructure can be obtained when the extrusion temperature increases to 300 and 320 °C.

Studies [17, 18] suggested that the increasing extrusion temperature leads to the coarse grain, and it seems contradictory that the grain size was gradually refined from 260 to 280 and 300 °C in the present case. Actually, high extrusion temperatures (above 300 °C) were often conducted in those previous studies, and a completely DRX microstructure could be obtained under such conditions. Furthermore, the extrusion temperature interval was much larger than that in the present study (20 °C). Therefore, the coarse elongated un-DRX grains did not exist, the replacement of coarse elongated un-DRX grains by the fine DRX grains would not occur in the previous studies, and increasing of the temperature would only result in coarse grain. On the contrary, in the present work, as the coarse elongated un-DRX grains were replaced by the fine DRX grains with the increasing DRX fraction from 260 to 300 °C, therefore, grain refinement could be observed, which has also been found in as-extruded AZ31 alloy by Tang et al. [19]. However, a complete DRX microstructure was obtained when the extrusion temperature raised to 300 °C, and then the coarse grain would occur when the temperature is further increased to 320 °C. Figure 5b, c shows the tensile and compressive engineering stress–strain curves, respectively. The corresponding tensile yield strength (TYS), compressive yield strength (CYS) and the tension–compression yield asymmetry ratio (CYS/TYS) are shown in Fig. 5d. All tensile curves show the characteristic of upper yield point following by yield point elongation of as much as 3% strain. The phenomenon

Effect of the Extrusion Temperature on Microstructure …

73

Fig. 5 Variation trend of microstructure and mechanical properties of the extruded alloys with extrusion temperatures: a the DRX fraction and average grain size, b tensile curves, c compressive curves, d the results of CYS, TYS and CYS/TYS

of the yield elongation has been confirmed as the pinning from solute atoms on the dislocations formed during the extrusion process [5]. In addition, high elongation to failure can be clearly observed from these curves. Compressive curves in Fig. 5c show a typical sigmoidal shape, and a yield phenomenon followed by rapid linear work hardening also can be clearly observed, which is considered as the feature of one kind of deformation mechanism predominant in the compressive yield behavior [20]. It can be concluded from Fig. 5d that both TYS and CYS increase with the increase of extrusion temperature from 260 to 280 and 300 °C, and then decrease once the extrusion temperature increases to 320 °C, which can be attributed to the evolution of the average grain size with the extrusion temperature. Moreover, all extruded alloys exhibited an extremely low tension–compression yield asymmetry, and the CYS/TYS increases slightly with the increasing extrusion temperature, in which value gets close to 1 gradually. The experimental alloy extruded at 300 °C is selected to be tested by EBSD, and the EBSD results are illustrated in

Fig. 6. As shown in Fig. 6a, the low-angle boundaries (2° * 15°) are depicted as red lines; however, they can hardly be observed in the IPF-Y orientation map. On the contrary, high-angle boundaries (15° * 100°) with black lines almost occupy the whole figure. In addition, the statistical result of grain boundary misorientation in Fig. 6b also proves it, suggesting a completely DRX microstructure with fine grains can be obtained after extruded at 300 °C. Furthermore, pole figures in Fig. 6c indicate that the orientation of basal plane deviates at an angle of 40° * 50° to the ED obviously. The corresponding inverse pole figure in Fig. 6d clearly reveals a typical non-fiber texture, in which grains do not lie on the arc between and parallel to the extrusion direction, but have an orientation with parallel to the extrusion direction, and this kind of texture component has been confirmed as RE texture by several previous studies [21, 22]. The Schmid factor of the basal slip inside each grain of the experimental alloy extruded at 300 °C is depicted in Fig. 7a, and the loading direction is parallel to the extrusion

74

L. Xiao et al.

Fig. 6 EBSD results of the alloy extruded at 300 °C: a the IPF-Y orientation map, b the statistical result of grain boundary misorientation, c pole figures, d inverse pole figure. Inverse pole figure refers to the extrusion direction

Fig. 7 Schmid factor results of the basal slip of the alloy extruded at 300 °C: a Schmid factor of each grain and b Schmid factor distribution

direction. It can be seen that large amount of grains have a high Schmid factor above 0.3. The corresponding distribution of the Schmid factor is shown in Fig. 7b, and it further proves that basal slip has a high Schmid factor, indicating the basal slip could activate easily under both tensile and compressive tests. In this case, tensile yield and compressive yield behaviors were both determined by the same deformation mechanism, which was quite different from that in the strong fiber texture-extruded Mg alloys [23]. Moreover, the Gd and Nd element addition would also result in the RE texture component in the other three extruded alloys. In addition, all the extruded alloys exhibited a comparative fine microstructure, which could also contribute to the low tension–compression yield asymmetry and high ductility [24]. Therefore, the extremely low tension–compression yield asymmetry with a high elongation to failure can be explained.

Conclusions In this study, Mg–2.49Nd–1.82Gd–0.19Zn–0.4Zr alloy was directly extruded at four different extrusion temperatures (260, 280, 300 and 320 °C) with an extrusion ratio of 15 and RAM speed of 3 mm s−1, respectively. Effects of the extrusion temperature on microstructure, texture evolution as well as the accompanying tension–compression yield asymmetry were investigated. The main conclusions were drawn: 1. The as-cast experimental alloy exhibited equiaxed a-Mg grains surrounded by discontinuous a-Mg + Mg12(Nd, Gd, Zn) eutectics and a-Zr particles within the matrix. After homogenized at 515 °C for 18 h, a small amount of residual Mg12(Nd, Gd, Zn) could be found along grain boundaries, and a new Zn2Zr3 phase was detected.

Effect of the Extrusion Temperature on Microstructure …

2. The coarse grains in homogenized billets were substantially refined after the extrusion process, which was caused by the refinement of DRX and the pinning effect of precipitate Mg5Gd and Mg12(Nd, Gd) particles. 3. The grain size gradually reduced when the extrusion temperature increased from 260 to 280 and 300 °C, and then coarsened slightly when the extrusion temperature raised to 320 °C. The DRX fraction was promoted with the increasing extrusion temperature, and a complete DRX microstructure could be obtained once the extrusion temperature up to 300 °C. 4. The RE tensile and compressive yield strength increased at first from 260 to 300 °C, and then decreased at 320 ° C. All extruded alloys showed an extremely low tension– compression yield asymmetry, which was mainly attributed to the RE texture component caused by RE element addition and the fine microstructure.

Acknowledgements This work was supported by the National Natural Science Foundation of China (Nos. 51771152, 51227001 and 51420105005) and the National Key Research and Development Program of China (Grant No. 2018YFB1106800).

References 1. Guan K, Yang Q, Bu FQ, Qiu X, Sun W, Zhang DP, Zheng T, Niu XD, Liu XJ, Meng J (2017) Microstructures and mechanical properties of a high-strength Mg-3.5Sm-0.6 Zn-0.5Zr alloy. Mater. Sci. Eng. A 703:97–107 2. Jung I, Sanjari M, Kim J, Yue S (2015) Role of RE in the deformation and recrystallization of Mg alloy and a new alloy design concept for Mg-RE alloys. Scr. Mater. 102:1–6 3. Li YX, Zhu GZ, Qiu D, Yin DD, Rong YH, Zhang MX (2016) The intrinsic effect of long period stacking ordered phases on mechanical properties in Mg-RE based alloys. J. Alloys Compd. 660:252–257 4. Ning ZL, Wang H, Liu HH, Cao FY, Wang ST, Sun JF (2010) Effects of Nd on microstructures and properties at the elevated temperature of a Mg-0.3Zn-0.32Zr alloy. Mater. Des. 31:4438– 4444 5. Ma L, Mishra RK, Balogh MP, Peng LM, Luo AA, Sachdev AK, Ding WJ (2012) Effect of Zn on the microstructure evolution of extruded Mg-3Nd (-Zn)-Zr (wt.%) alloys. Mater. Sci. Eng. A 543:12–21 6. Ning ZL, Liu HH, Cao FY, Wang ST, Sun J F, Qian M (2013) The effect of grain size on the tensile and creep properties of Mg-2.6Nd-0.35Zn-xZr alloys at 250 °C. Mater. Sci. Eng. A 560:163–169 7. Han WY, Yang GY, Xiao L, Li JH, Jie WQ (2017) Creep properties and creep microstructure evolution of Mg-2.49Nd-1.82Gd0.19Zn-0.4Zr alloy. Mater. Sci. Eng. A 684:90–100

75 8. Liu SJ, Yang GY, Luo SF, Jie WQ (2015) Microstructure evolution during heat treatment and mechanical properties of Mg-2.49Nd-1.82Gd-0.19Zn-0.4Zr cast alloy. Mater. Charact. 107:334–342 9. Wu SW, Oh-ishi K, Kamado S, Uchida F, Homma T, Hono K (2011) High-strength extruded Mg-Al-Ca-Mn alloy. Scr. Mater. 65 (3):269–272 10. Hantzsche K, Bohlen J, Wendt J, Kainer KU, Yi SB, Letzig D (2010) Effect of rare earth additions on microstructure and texture development of magnesium alloy sheets. Scr. Mater. 63(7):725–730 11. Song GL, Stjohn D (2002) The effect of zirconium grain refinement on the corrosion behavior of magnesium-rare earth alloy MEZ. J. Light. Met. 2:1–16 12. Gao X, Muddle BC, Nie JF (2009) Transmission electron microscopy of Zr-Zn precipitate rods in magnesium alloys containing Zr and Zn. Phil. Mag. Lett. 89(1):33–43 13. Gill LR, Lorimer GW, Lyon P (2007) The Effect of Zinc and Gadolinium on the precipitation sequence and quench sensitivity of four Mg-Nd-Gd alloys. Adv. Eng. Mater. 9(9):784–792 14. Li JH, Barrirero J, Sha G, Aboulfadl H, Mücklich F, Schumacher P (2016) Precipitation hardening of an Mg-5Zn-2Gd-0.4Zr (wt.%) alloy. Acta Mater. 108:207–218 15. Freeney TA, Mishra RS (2010) Effect of friction stir processing on microstructure and mechanical properties of a Cast-Magnesium-Rare earth alloy. Metall. Mater. Trans. A 41:73–84 16. Nie JF (2012) Precipitation and hardening in magnesium alloy. Metall. Mater. Trans. A 43(11):3891–3939 17. Zhang BP, Geng L, Huang LJ, Zhang XX, Dong C (2010) Enhanced mechanical properties in fine-grained Mg-1.0Zn-0.5Ca alloys prepared by extrusion at different temperatures. Scr. Mater. 63(10):1024–1027 18. Murai T, Matsuoka S, Miyamoto S, Oki Y (2003) Effects of extrusion conditions on microstructure and mechanical properties of AZ31B magnesium alloy extrusions. J. Mater. Process. Technol. 141(2):207–212 19. Tang WQ, Huang SY, Zhang SR, Li DY, Peng YH (2011) Influence of extrusion parameters on grain size and texture distributions of AZ31 alloy. J. Mater. Process. Technol. 211 (7):1203–1209 20. Park SH, Lee JH, Moon BG, You BS (2014) Tension-compression yield asymmetry in as-cast magnesium alloy. J. Alloys Compd. 617:277–280 21. Imandoust A, Barrett CD, Oppedal AL, Whittington WR, Paudel Y, Kadiri HE (2017) Nucleation and preferential growth mechanism of recrystallization texture in high purity binary magnesium-rare earth alloys. Acta Mater. 138(1):27–41 22. Wu XW, Jin L, Zhang ZY, Ding WJ, Dong J (2014) Grain growth and texture evolution during annealing in an indirect-extruded Mg-1Gd alloy. J. Alloys Compd. 585:111–119 23. Jiang Y, Chen YA, Gao GT (2016) Role of volume fraction of second phase particles, dislocation-twin and twin-twin interactions in the reduced tension-compression yield asymmetry. Mater. Des. 97:131–137 24. Dogan E, Karaman I, Ayoub G, Kridi G (2014) Reduction in tension–compression asymmetry via grain refinement and texture design in Mg-3Al-1Zn sheets. Mater. Sci. Eng. A 610:220–227

Influence of Thermomechanical Treatment on Tension-Compression Yield Asymmetry of Extruded Mg–Zn–Ca Alloy P. Dobroň, M. Hegedüs, J. Olejňák, D. Drozdenko, K. Horváth, and J. Bohlen

Abstract

Thermomechanical treatment consisting of pre-compression and isothermal aging at 150 °C for 16 h was applied to the extruded Mg–Zn–Ca (ZX10) alloy in order to reduce tension-compression yield asymmetry and improve mechanical properties via strengthening mechanism. With respect to the initial texture of the alloy, pre-compression leads to a formation of extension twins. A solute segregation and precipitation along twin boundaries is realized during a subsequent isothermal aging. After thermomechanical treatment, a solute solution and precipitation hardening contribute to the strengthening of the alloy. Active deformation mechanisms were monitored during compression or tension using the acoustic emission technique. Keywords





Magnesium Precipitation




Pre-compression Twinning Acoustic emission

P. Dobroň (&)
M. Hegedüs
J. Olejňák
D. Drozdenko
K. Horváth Department of Physics of Materials, Charles University, Ke Karlovu 5, 12116 Prague 2, Czech Republic e-mail: [email protected] D. Drozdenko Magnesium Research Center, Kumamoto University, 2-39-1 Kurokami, 860-8555 Kumamoto, Japan K. Horváth Nuclear Physics Institute, The Czech Academy of Sciences, Řež 130, 25068 Řež, Czech Republic J. Bohlen MagIC - Magnesium Innovation Centre, Helmholtz-Zentrum Geesthacht, Max-Planck-Straße 1, D21502 Geesthacht, Germany

Introduction Tension-compression yield strength asymmetry of extruded Mg alloys is a limiting factor for their use in technical applications [1]. A formation of extension twins during compression along the extrusion direction (ED) significantly reduces compressive yield strength (CYS) in comparison with tensile yield strength (TYS), where dislocation slip is dominant deformation mechanism due to strong basal texture. It is generally accepted that an influence of twinning on compressive deformation behavior can be reduced by texture weakening, grain size refinement or using thermomechanical treatment (TMT). Combination of pre-compression and isothermal aging lead to precipitation to newly formed twin boundaries and during subsequent compression, the twin boundaries should remain immobile. It was shown in [2] that proper TMT performed on the extruded ZX10 alloy increases CYS of the alloy. Tensile properties of pre-compressed Mg alloys are negatively influenced by detwinning [3, 4]. This process is characterized by a thickness reduction or disappearance of existing twin lamellae as stress required for moving the twin boundaries is lower than that for nucleation of new twins. The activity of deformation mechanisms during loading of Mg alloys can be monitored using the acoustic emission (AE) technique [5–9]. AE provides real-time information about the collective processes involved in plastic deformation of crystalline materials. It is based on sensitive detection of the transient elastic waves propagated within the material during deformation due to sudden localized structure changes. The present work is focused on examination the influence of TMT on tension-compression yield asymmetry of extruded ZX10 alloy. Due to a low content of Zn and Ca in the alloy, only Mg2Ca precipitates are formed as basal plates in microstructure during isothermal aging [2, 10–12]. The contribution of precipitates to the strengthening of the pre-compressed ZX10 alloy with respect to deformation loading is discussed.

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_13

77

78

P. Dobroň et al.

Experimental Procedure

Results

The ZX10 (Mg–1 wt%Zn–0.25 wt%Ca) alloy was prepared by a conventional casting process and prior extrusion the billet was solid-solution annealed for 20 h at 400 °C. Indirect extrusion (extrusion ratio of 1:25) was carried out at 400 °C with an extrusion speed of 0.5 mm/s to produce a round bar with a final diameter of 10 mm. For deformation tests, samples with a diameter of 5 mm and a gauge length of 12 mm (pre-compression—tension loading) and the Ø 9.5–14 mm samples (pre-compression— compression loading) were machined from the extruded bar along ED. Deformation tests were performed using a universal testing machine Instron 5882 at room temperature and at a constant strain rate of 10−3 s−1. The choice of pre-compression stress levels and isothermal aging conditions is based on our previous study [2]. Pre-compression of 118, 130, 140, and 155 MPa represents plastic strain of 1, 2, 3, and 4%, and applied isothermal aging at 150 °C for 16 h maximize strengthening effect at the yield strength. AE during deformation tests was monitored by the computer-controlled PCI-2 device (Physical Acoustic Corporation) using a piezoelectric transducer (Pico 30S—PAC) and a 2/4/6-preamplifier giving a gain of 60 dB. The sensor was attached to the specimen surface using a clamp. The threshold level of 0.1 V (the full scale of the device is 10 V) was used to parametrize the AE signal (AE count rate— count number per time unit [13]). Microstructure of the ZX10 alloy was investigated by a Zeiss Auriga scanning electron microscope (SEM) equipped with an EDAX electron backscatter diffraction (EBSD) camera. The surface of the samples were ground using SiC paper, polished by diamond pastes down to 0.25 lm particle size and finally polished by Ag ion milling (Leica EM RES102).

The microstructure of the extruded alloy is homogeneous with an average grain size of (11 ± 1) µm and exhibits relative weak basal texture (Fig. 1). Influence of precompression and isothermal aging on deformation behavior of samples subsequently deformed in compression or tension can be seen in Fig. 2a, b, respectively. All tensile curves exhibit S-shape, which become more pronounced with increasing pre-compression level. During compression, a yield plateau is observed on deformation curves for all tested samples and their following strengthening behavior is comparable for differently pre-strained samples. Thus, influence of pre-compression level on tensile or compressive strain of the thermomechanically treated samples was negligible. Higher plasticity was achieved in samples loaded in tension. Influence of TMT on yield strength is presented in Fig. 3. Compressive yield strength increases continuously with increasing pre-compression strain, whereas TYS for an initial state is higher than those for the thermomechanically treated samples and TYS has an increasing tendency only with a higher level of pre-compression. In the initial state, tension-compression yield strength asymmetry with a higher value of TYS, typical for Mg alloys, can be seen. A different scenario is observed for the thermomechanically treated samples, where CYS is higher than TYS for all samples and the asymmetry is more pronounced. Interestingly, there is no change in the asymmetry stress value (i.e. difference between CYS and TYS values) for all pre-compression levels. The AE activity was monitored during deformation of all samples, however due to similar tendencies in stress-strain and the AE count rate—strain behavior for thermomechanically treated samples, only data for the initial state (Fig. 4) and after pre-compression of 2% and isothermal aging at

Fig. 1 Initial microstructure and texture of the ZX10 alloy (X ED)

Influence of Thermomechanical Treatment on Tension-Compression … Fig. 2 Influence of pre-compression and isothermal aging at 150 °C for 16 h on deformation behavior in a compression and b tension

Fig. 3 Effect of the pre-compression level and a subsequent isothermal aging at 150 °C for 16 h on tension-compression yield strength asymmetry

Fig. 4 Stress and the AE count rate versus strain curves for tension (a) and compression (b) of the sample in the initial state

Fig. 5 Stress and the AE count rate versus strain curves for tension (a) and compression (b) of the thermomechanically treated sample (pre-compressed up to 2% and isothermally aged at 150 °C for 16 h)

79

80

150 °C for 16 h (Fig. 5) are presented as a representative one. For samples in the initial state, the AE count rate in the vicinity of yield point is much higher for deformation in compression with more pronounced following decrease in the AE activity in comparison to that during tensile loading. In the thermomechanically treated samples, the AE count rate during tension exhibits two maxima, which correlate with the yielding. The AE activity during compression is weak and only some higher AE count rate can be observed around the yield point.

Discussion The ZX10 alloy exhibits a fully recrystallized and homogeneous microstructure with a basal texture. During pre-compression, {10–12} extension twins are formed in the microstructure. More details about the evolution of twin volume fraction as a function of pre-compression strain can be found in [2]. Application of TMT led to an increase in CYS without any further changes in compressive deformation behavior. That demonstrates possibilities of improving the mechanical properties in compression using the proposed procedure, where higher dislocation density, twin volume fraction and a formation of Mg2Ca in the microstructure during heat treatment contribute to strengthening of the Mg alloy. A different situation arises for tensile properties. The S-shaped tensile curves clearly point out that the detwinning process contributes to plastic strain. The stress required for detwinning is lower than those for twin nucleation and therefore, during reverse tensile loading of already pre-compressed samples, detwinning plays a key role in deformation behavior. Similar results were observed during cyclic loading of Mg alloys [3]. The AE measurements confirm that during tensile test of samples in the initial state (as extruded), the dominant deformation mechanism is dislocation slip. Multiplication and collective dislocation movement results in a maximum of the AE count rate at the yield point and the following decline in the AE activity reflects increasing dislocation density, which reduces the mean free path of mobile dislocations. The very high AE at the yield point during compression is produced by a collective dislocation movement in basal planes and mainly by twin nucleation, which is an excellent source of AE [14]. The steep decrease in the AE count rate can be ascribe, beside high dislocation density, to newly formed twin boundaries, which also become as a barrier for collective dislocation movement. The second AE peak related to the inflection point on the deformation curve indicates on activation of another deformation mechanism and a competitive behavior between strengthening and softening of the alloy.

P. Dobroň et al.

Tensile loading of the thermomechanically treated sample was accompanied by a very interesting AE response at yielding. While twin nucleation produces detectable AE signals, twin growth or shrinkage is very slow processes compared to twin nucleation and therefore, it is below the resolution of the AE method [15, 16]. The first AE peak can be related to collective dislocation movement due to opposite loading direction with respect to previous pre-compression. Detwinning cannot contribute to the AE activity and the AE count rate is decreasing. However, during detwinning, basal planes are reoriented back in favorable position for slip and thus, a collective dislocation movement in basal planes causes the second AE peak. Higher dislocation density in the later stage of plastic deformation is responsible for a gradual decrease in the AE count rate. The very low AE activity during compression of the sample subjected to TMT confirms that higher dislocation density due to pre-compression together with a formation of Mg2Ca precipitates significantly reduced the mean free path for moving dislocation. With respect to our results, it is obvious that Mg2Ca precipitates created during heat treatment cannot effectively stop twin boundary movement and therefore, detwinning is responsible for a relative low TYS. This limitation is also reflected in a very high tension-compression asymmetry for samples subjected to TMT, where CYS can be increased using TMT, however TYS after applying TMT is lower than those in the initial state.

Conclusions The influence of thermomechanical treatment (TMT) consisting of pre-compression and isothermal aging at 150 °C for 16 h on tension-compression yield asymmetry of extruded ZX10 alloy was studied. It was shown that TMT can be used to increase compressive yield strength (CYS). However, tensile properties, particularly tensile yield strength (TYS) are negatively influenced by detwinning as a consequence of previous pre-compression. The Mg2Ca precipitates formed during isothermal aging in pre-deformed microstructure cannot effectively prevent twin boundary shrinkage during reverse tensile loading. Thus, tension-compression yield asymmetry for samples subjected to TMT is more pronounced in comparison with the initial state, however with higher CYS than TYS. Acknowledgements This work received support from the Czech Science Foundation under grant No. 17-21855S; the Grant Agency of the Charles University under grant Nr. 1262217; the Operational Programme Research, Development and Education, The Ministry of Education, Youth and Sports (OP RDE, MEYS) under the grant CZ.02.1.01/0.0/0.0/16_013/0001794.

Influence of Thermomechanical Treatment on Tension-Compression …

References 1. Bettles, C.J.; Gibson, M.A. Current wrought magnesium alloys: Strengths and weaknesses. JOM 2005, 57, 46–49. https://doi.org/ 10.1007/s11837-005-0095-0 2. Dobron, P.; Drozdenko, D.; Olejnak, J.; Hegedus, M.; Horvath, K.; Vesely, J.; Bohlen, J.; Letzig, D. Compressive yield stress improvement using thermomechanical treatment of extruded mg-zn-ca alloy. Materials Science and Engineering a-Structural Materials Properties Microstructure and Processing 2018, 730, 401–409. https://doi.org/10.1016/j.msea.2018.06.026 3. Drozdenko, D.; Bohlen, J.; Yi, S.; Minarik, P.; Chmelik, F.; Dobron, P. Investigating a twinning-detwinning process in wrought mg alloys by the acoustic emission technique. Acta Materialia 2016, 110, 103–113. https://doi.org/10.1016/j.actamat. 2016.03.013 4. Bohlen, J.; Dobron, P.; Nascimento, L.; Parfenenko, K.; Chmelik, F.; Letzig, D. The effect of reversed loading conditions on the mechanical behaviour of extruded magnesium alloy az31. Acta Physica Polonica A 2012, 122, 444–449. 5. Drozdenko, D.; Bohlen, J.; Chmelik, F.; Lukac, P.; Dobron, P. Acoustic emission study on the activity of slip and twin mechanisms during compression testing of magnesium single crystals. Materials Science and Engineering a-Structural Materials Properties Microstructure and Processing 2016, 650, 20–27. https://doi.org/10.1016/j.msea.2015.https://doi.org/10.033 6. Dobroň, P.; Balík, J.; Chmelík, F.; Illková, K.; Bohlen, J.; Letzig, D.; Lukáč, P. A study of mechanical anisotropy of mg-zn-rare earth alloy sheet. Journal of Alloys and Compounds 2014, 588, 628–632. https://doi.org/10.1016/j.jallcom.2013.11.142 7. Mathis, K.; Capek, J.; Clausen, B.; Krajnak, T.; Nagarajan, D. Investigation of the dependence of deformation mechanisms on solute content in polycrystalline mg-al magnesium alloys by neutron diffraction and acoustic emission. Journal of Alloys and Compounds 2015, 642, 185–191. https://doi.org/10.1016/j.jallcom. 2015.03.258

81

8. Horvath, K.; Drozdenko, D.; Mathis, K.; Bohlen, J.; Dobron, P. Deformation behavior and acoustic emission response on uniaxial compression of extruded rectangular profile of mg-zn-zr alloy. Journal of Alloys and Compounds 2016, 680, 623–632. https://doi.org/10.1016/j.jallcom.2016.03.310 9. Vinogradov, A.; Vasilev, E.; Linderov, M.; Merson, D. In situ observations of the kinetics of twinning–detwinning and dislocation slip in magnesium. Materials Science and Engineering a-Structural Materials Properties Microstructure and Processing 2016, 676, 351–360. http://dx.doi.org/10.1016/j.msea.2016.09.004 10. Oh, J.C.; Ohkubo, T.; Mukai, T.; Hono, K. Tem and 3dap characterization of an age-hardened mg–ca–zn alloy. Scripta Materialia 2005, 53, 675–679. https://doi.org/10.1016/j. scriptamat.2005.05.030 11. Oh-ishi, K.; Watanabe, R.; Mendis, C.L.; Hono, K. Age-hardening response of mg–0.3at.%ca alloys with different zn contents. Materials Science and Engineering: A 2009, 526, 177–184. https://doi.org/10.1016/j.msea.2009.07.027 12. Gao, X.; Zhu, S.M.; Muddle, B.C.; Nie, J.F. Precipitation-hardened mg–ca–zn alloys with superior creep resistance. Scripta Materialia 2005, 53, 1321–1326. https://doi. org/10.1016/j.scriptamat.2005.08.035 13. Standard test method for dynamic young’s modulus, shear modulus, and poisson’s ratio for advanced ceramics by impulse excitation of vibration. Standard Test Method for Dynamic Young’s Modulus, Shear Modulus, and Poisson’s Ratio for Advanced Ceramics by Impulse Excitation of Vibration 1994. 14. Heiple, C.R.; Carpenter, S.H. Acoustic emission produced by deformation of metals and alloys-a review: Part II. J. Acoust. Emission 1987, 6, 215–237. 15. Toronchuk, J.P. Acoustic emission during twinning of zinc single crystals. Materials Evaluation 1977, 35, 51–53. 16. Vinogradov, A.; Vasilev, E.; Seleznev, M.; Máthis, K.; Orlov, D.; Merson, D. On the limits of acoustic emission detectability for twinning. Materials Letters 2016, 183, 417–419. https://doi.org/ 10.1016/j.matlet.2016.07.063

Homogeneous Grain Refinement and Ductility Enhancement in AZ31B Magnesium Alloy Using Friction Stir Processing Vivek Patel, Wenya Li, Quan Wen, Yu Su, and Na Li

Abstract

Low ductility in magnesium (Mg) alloy hinders its application in material forming industries. Hence, it is indeed of the material processing technique to modify the microstructure of Mg alloy for increasing ductility. Present work aims to refine the grain size of thick AZ31B Mg alloy using friction stir processing (FSP) technique. A low heat input stationary shoulder tooling system is used for processing heat sensitive Mg alloy, exhibiting small temperature gradient across the thickness of processed material. This low heat input and small temperature gradient contributed to uniform grain refinement across the thickness due to dynamic recrystallization. Significant reduction in grain size achieved after FSP, which helped to enhance the elongation of FSP samples (top, middle, and bottom) in comparison with the unprocessed material. Furthermore, uniform and homogenous grain size and elongation obtained across the thickness of processed material, which was attributed to the stationary shoulder tooling system.





Keywords








AZ31B Ductility Friction stir processing Grain refinement Magnesium Stationary shoulder

V. Patel (&)
W. Li
Q. Wen
Y. Su
N. Li Shaanxi Key Laboratory of Friction Welding Technologies, School of Materials Science and Engineering, Northwestern Polytechnical University, Xi’an, 710072, Shaanxi, People’s Republic of China e-mail: [email protected]; [email protected] V. Patel Mechanical Engineering Department, School of Technology, Pandit Deendayal Petroleum University, Gandhinagar, 382007, Gujarat, India

Introduction Friction stir processing (FSP) was derived from friction stir welding (FSW), following same working principle of FSW. FSW was invented at The Welding Institute (TWI) of UK in 1991 as a solid-state joining technique, and it was initially applied to aluminum alloys [1]. FSP was developed by using the basic principles of FSW [2, 3]. Recently, FSP has been demonstrated as an effective and promising solid-state material processing technique to improve the mechanical and metallurgical properties of the base material (BM) [4, 5]. FSP has shown wide range of applications such as grain refinement [6, 7], superplasticity [8], surface composite manufacturing [9], modification of coating structure [10, 11]. FSP consists of a rotating tool, which plunges into the BM and advances in the direction of processing, as shown in Fig. 1a. When the rotating tool contacts the workpiece surface, sever plastic deformation of BM takes place and plasticized material flows around the tool probe. This deformation develops different zones such as nugget or stir zone (SZ), thermo-mechanically affected zone (TMAZ), and heat affected zone (HAZ) within processing region as displayed in Fig. 1b. Shoulder mainly contributes to the frictional deformation while probe to the plastic deformation. Moreover, shoulder is responsible to generate majority of the heat during processing due to its larger surface area [12]. Therefore, researchers have demonstrated additional cooing during FSP in order to reduce the process temperature and so that grain size in the SZ [13, 14]. Magnesium (Mg) alloys are gaining popularity as structural materials due to its lower weight to strength ratio in comparison with the aluminum alloys. However, magnesium alloys suffer from some processing issues due to its heat sensitivity. Recently, Mg alloys have been processed for grain refinement using FSP with additional cooling approach [15– 17]. To eliminate use of additional cooling, the tooling system could be modified in such a way that shoulder of tool remains nonrotating, like stationary shoulder friction stir welding

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_14

83

84

V. Patel et al.

Fig. 1 Schematic diagrams of: a FSP and b cross-sectional views of processed zones [5]

(SSFSW) tooling system. Use of stationary shoulder tool in welding has already demonstrated sound weld quality because of low heat input, smaller temperature gradient across the weld thickness, no or little flashes, and arc corrugations [18–20]. Such a low heat input and smaller temperature gradient characteristics are desired in FSP of Mg alloys to enhance the grain refinement by producing uniform fine-grained microstructure in SZ. Hence, stationary shoulder tooling system could be used in FSP as well to lower the heat input during processing of magnesium alloys without using additional cooling. We propose stationary shoulder friction stir processing (SSFSP) as a variant of FSP for microstructure refinement and properties enhancement. The present work is to study SSFSP of thick AZ31B magnesium alloy.

Materials and Methods FSP was applied on rolled plate of 6.35-mm-thick magnesium alloy. The plate was cut in dimension of 250  150 mm2, ensuring 250 mm side represents rolling direction, and FSP was conducted parallel to the rolling direction of plate. The chemical composition of BM is displayed in Table 1. Prior to the FSP, plates were ground and subsequently cleaned with acetone in order to obtain contamination free flat surface. Special fixture was used to clamp the plate in rigid manner to avoid deflection, since

Table 1 Chemical composition of AZ31B alloy

tool exerts large amount of axial downward as well as traverse force onto the BM during FSP. The customized stationary shoulder tooling package developed at the Shaanxi key laboratory of friction welding technologies, NWPU was used to install the stationary shoulder and rotating probe or pin. Stationary shoulder and probe assembly is displayed in Fig. 2. The probe tool was made of tool steel and subsequently followed heat treatment. The tool dimensions consist of stationary shoulder of 18 mm, threaded probe diameter of 6 mm root and 3.5 mm tip, and probe length of 6 mm. After series of experimental trails, FSP was conducted at the process parameters of 700 rpm probe rotational speed, 100 mm/min traverse speed, 2.5° tool tilt, and single pass. After FSP, the transverse cross section of the processing zone was cut to prepare specimen for optical microscopy. The cut specimens were mechanically ground, polished, and etched (4.2 g picric acid, 10 ml glacial acetic acid, 10 ml distilled water, 70 ml ethanol). Grain size in the SZ was measured using linear intercept method. Tensile specimens were cut using wire cut electrodischarge machining in longitudinal direction of the processing zone in such a way that gage dimensions (25 mm length, 1.5 mm thickness, 2 mm width) coincide with SZ only. Three tensile specimens were cut at fiddfrent locations in the SZ, i.e., top, middle, and bottom to evaluate the uniformity and homogeneity in tensile properties. The room temperature tensile testing was conducted at crosshead speed of 0.5 mm/min.

Al

Zn

Mn

Fe

Cu

Si

Mg

2.5 * 3.5

0.60 * 1.4

0.20 * 1.0

 0.003

 0.001

 0.008

Bal.

Homogeneous Grain Refinement and Ductility Enhancement … Fig. 2 Stationary shoulder tooling system

85

Rotating Probe

Stationary Shoulder

Results and Discussion Macrostructure of FSPed zone is shown in Fig. 3. As it can be seen that three distinct zones in the macrograph, such as SZ, TMAZ, and HAZ, are developed. Shape of the SZ clearly indicates that plastic deformation is produced by only probe rather than shoulder. Additionally, shape of the SZ is found symmetric on the advancing side (AS) as well as (RS), representing uniform and homogenous deformation within the SZ. Therefore, non-rotating action of the shoulder resulted into probe-dominated SZ throughout the thickness, which is not possible to obtain by conventional tooling system. Furthermore, probe-dominated SZ ensures the low heat input and smaller temperature gradient across the thickness during the process, which results in homogenous microstructure refinement. Despite of low heat input processing conditions, SSFSP produced defect-free SZ. The main reasons for the void related defects in the SZ are due to selection of inappropriate tool geometry, tool rotational speed, and traverse speed [21–23]. The SZ microstructure evolution is represented in Fig. 3c. Sever plastic deformation produced by tool probe resulted in fine equiaxed grains in the SZ due to dynamic recrystallization. The low heat input stationary tooling system could have restricted grain growth during recrystallization and thus reduced the grain size. The grain size in SZ is reduced significantly to 6.77 ± 0.72 from 25 ± 1.42 µm of the BM. It is worth to note that this grain refinement is achieved without using rapid cooling during the process, which attributes solely to the stationary shoulder tooling system. The other microstructural zones of TMAZ and HAZ evolved during process on both sides of AS and RS are shown as micrographs in Fig. 3a, c. TMAZ is developed adjacent to SZ due to heat generation by the rotating probe and sliding stationary shoulder. While HAZ

represents almost similar features of BM due to solid-state nature of the process. The grain refinement in the SZ influences the mechanical properties of the processed material. Fine-grained microstructure possesses more number of grain boundaries and thus it requires more energy to deform it. Tensile behavior of the FSP specimens along with as-received BM is shown in Fig. 4. The tensile specimens were tested for the three different locations in the SZ, i.e., top, middle, and bottom and almost similar tensile behavior reported as shown in Fig. 4, which confirms the uniform and homogenous grain refinement in SZ by stationary shoulder tooling system. FSP samples underwent strain hardening during the tensile deformation in comparison with that of BM. The FSPed SZ reported slight increment in ultimate tensile strength (UTS) to 201 MPa from the value of 181 MPa (BM). But, significant improvement in ductility is achieved for FSPed sample (21.67%) compared to that of BM (17.05%). This improvement in ductility could further be enhanced by adopting additional cooling during SSFSP. Moreover, Yuan et al. [24] studied the effect of texture on grin size and subsequent tensile properties in FSPed AZ31 Mg alloy. They found the presence of strong basal fiber texture in the processing direction of SZ, which favours the formation of extension twinning during tensile deformation for strain accommodation. At the beginning of tensile testing, the formation of extension of twinning can reduce the stress concentration and accommodate strain incompatibility arising from basal slip dislocation movement, which can accommodate further dislocations [25]. Therefore, twinning effect induced by basal slip contributes to the strain hardening and high ductility in the processed samples, which confirms the tensile curve behavior of the FSPed sample in present work. Overall summary of grain refinement and tensile properties is tabulated in Table 2.

86

V. Patel et al.

(a)

(b)

TMAZ

SZ

SZ

TMAZ

100 μm

100 μm

AS

RS SZ HAZ

(a) *

* (c)

TMAZ

* (b) 1 mm

(c)

50 μm

Fig. 3 Montage of the cross section of processing zone representing distinct microstructural zones

Engiineering stress (MPa)

250

200

Table 2 Comparative summary of the grain size and tensile properties

FSP top FSP middle FSP bottom BM

Alloy

150

Average grain size (µm)

UTS (MPa)

El (%)

BM

25 ± 1.42

181

17.05

FSP

6.77 ± 0.72

201

21.67

100

Summary

50

0 0.00

0.05

0.10

0.15

0.20

0.25

Engineering strain

Fig. 4 Room temperature tensile behavior of the FSPed SZ and BM

Stationary shoulder FSP has successfully refined the grain size of the thick Mg alloy without using additional cooling. Use of stationary shoulder exhibited probe-dominated SZ with uniform distribution of fine-grained size microstructure,

Homogeneous Grain Refinement and Ductility Enhancement …

resulted in homogenous enhanced ductility across the SZ. This enhancement in ductility is attributed to the low heat input tooling system and smaller temperature gradient across the SZ. Acknowledgements The authors would like to thank the financial support from the National Key Research and Development Program of China (2016YFB1100104). We also want to acknowledge the editorial committee, organizer (s) of Magnesium Technology 2019, and TMS for recognizing our research work.

References 1. Thomas W, Nicholas E, Needham J, Murch M, Templesmith P, Dawes C (1991) 2. Mishra RS, Mahoney M, McFadden S, Mara N, Mukherjee A (1999) High strain rate superplasticity in a friction stir processed 7075 Al alloy. Scripta Materialia 42 (2):163–168 3. Mishra RS, Mahoney MW (2001) Friction stir processing: a new grain refinement technique to achieve high strain rate superplasticity in commercial alloys. Materials Science Forum 357:507–514 4. Ma ZY, Feng AH, Chen DL, Shen J (2018) Recent Advances in Friction Stir Welding/Processing of Aluminum Alloys: Microstructural Evolution and Mechanical Properties. Critical Reviews in Solid State and Materials Sciences 43 (4):269–333. https://doi.org/10.1080/10408436.2017.1358145 5. Patel VV, Li WY, Vairis A, Badheka VJ (2018) Recent Development in Friction Stir Processing as a Solid-State Grain Refinement Technique: Microstructural Evolution and Property Enhancement. Critical Reviews in Solid State and Materials Sciences. In Press. 6. Patel VV, Badheka VJ, Kumar A (2016) Effect of Velocity Index on Grain Size of Friction Stir Processed Al-Zn-Mg-Cu Alloy. Procedia Technology 23:537–542. doi:http://dx.doi.org/10.1016/j. protcy.2016.03.060 7. Padhy GK, Wu CS, Gao S (2018) Friction stir based welding and processing technologies-processes, parameters, microstructures and applications: A review. Journal of Materials Science & Technology 34 (1):1–38. doi:https://doi.org/10.1016/j.jmst.2017. 11.029 8. Patel VV, Badheka V, Kumar A (2016) Friction Stir Processing as a Novel Technique to Achieve Superplasticity in Aluminum Alloys: Process Variables, Variants, and Applications. Metallography, Microstructure, and Analysis 5 (4):278–293 9. Rathee S, Maheshwari S, Siddiquee AN, Srivastava M (2018) A Review of Recent Progress in Solid State Fabrication of Composites and Functionally Graded Systems Via Friction Stir Processing. Critical Reviews in Solid State and Materials Sciences 43 (4):334– 366. https://doi.org/10.1080/10408436.2017.1358146 10. Yang K, Li WY, Niu PL, Yang XW, Xu YX (2018) Cold sprayed AA2024/Al2O3 metal matrix composites improved by friction stir processing: Microstructure characterization, mechanical performance and strengthening mechanisms. Journal of Alloys and Compounds 736:115–123. https://doi.org/10.1016/j.jallcom.2017.11.132 11. Yang K, Li W, Huang C, Yang X, Xu Y (2018) Optimization of cold-sprayed AA2024/Al2O3 metal matrix composites via friction stir processing: Effect of rotation speeds. Journal of Materials Science & Technology. doi:https://doi.org/10.1016/j.jmst.2018. 03.016

87 12. Patel VV, Badheka VJ, Kumar A (2017) Influence of Pin Profile on the Tool Plunge Stage in Friction Stir Processing of Al–Zn– Mg–Cu Alloy. Transactions of the Indian Institute of Metals 70 (4):1151–1158. https://doi.org/10.1007/s12666-016-0903-y 13. Patel VV, Badheka VJ, Patel U, Patel S, Patel S, Zala S, Badheka K (2017) Experimental Investigation on Hybrid Friction Stir Processing using compressed air in Aluminum 7075 alloy. Materials Today: Proceedings 4 (9):10025–10029. doi:https://doi. org/10.1016/j.matpr.2017.06.314 14. Patel VV, Badheka VJ, Zala SR, Patel SR, Patel UD, Patel SN Effects of Various Cooling Techniques on Grain Refinement of Aluminum 7075-T651 During Friction Stir Processing. In: ASME 2016 International Mechanical Engineering Congress and Exposition, 2016. American Society of Mechanical Engineers, pp V014T011A015-V014T011A015 15. Alavi Nia A, Omidvar H, Nourbakhsh SH (2014) Effects of an overlapping multi-pass friction stir process and rapid cooling on the mechanical properties and microstructure of AZ31 magnesium alloy. Materials & Design 58 (Supplement C):298–304. doi:https:// doi.org/10.1016/j.matdes.2014.01.069 16. Darras B, Kishta E (2013) Submerged friction stir processing of AZ31 Magnesium alloy. Materials & Design 47 (Supplement C):133–137. doi:https://doi.org/10.1016/j.matdes.2012.12.026 17. Chang CI, Du XH, Huang JC (2008) Producing nanograined microstructure in Mg–Al–Zn alloy by two-step friction stir processing. Scripta Materialia 59 (3):356–359. doi:https://doi.org/ 10.1016/j.scriptamat.2008.04.003 18. Barbini A, Carstensen J, dos Santos JF (2018) Influence of a non-rotating shoulder on heat generation, microstructure and mechanical properties of dissimilar AA2024/AA7050 FSW joints. Journal of Materials Science & Technology 34 (1):119–127. doi: https://doi.org/10.1016/j.jmst.2017.10.017 19. Ji S, Meng X, Liu J, Zhang L, Gao S (2014) Formation and mechanical properties of stationary shoulder friction stir welded 6005A-T6 aluminum alloy. Materials & Design (1980–2015) 62:113–117 20. Wen Q, Li WY, Wang WB, Wang FF, Gao YJ, Patel V (2018) Experimental and numerical investigations of bonding interface behavior in stationary shoulder friction stir lap welding. Journal of Materials Science & Technology. In Press. doi:https://doi.org/10. 1016/j.jmst.2018.09.028 21. Patel VV, Badheka V, Kumar A (2017) Effect of polygonal pin profiles on friction stir processed superplasticity of AA7075 alloy. Journal of Materials Processing Technology 240:68–76. doi:http:// dx.doi.org/10.1016/j.jmatprotec.2016.09.009 22. Patel VV, Badheka VJ, Kumar A (2016) Cavitation in Friction Stir Processing of Al-Zn-Mg-Cu Alloy. International Journal of Mechanical Engineering and Robotics Research 5 (4):317–321. https://doi.org/10.18178/ijmerr.5.4.317-321 23. Patel VV, Badheka V, Kumar A (2016) Influence of Friction Stir Processed Parameters on Superplasticity of Al-Zn-Mg-Cu Alloy. Materials and Manufacturing Processes 31 (12):1573–1582. https://doi.org/10.1080/10426914.2015.1103868 24. Yuan W, Mishra RS, Carlson B, Mishra R, Verma R, Kubic R (2011) Effect of texture on the mechanical behavior of ultrafine grained magnesium alloy. Scripta Materialia 64 (6):580–583 25. Yuan W, Mishra RS (2012) Grain size and texture effects on deformation behavior of AZ31 magnesium alloy. Materials Science and Engineering: A 558 (Supplement C):716–724. doi: https://doi.org/10.1016/j.msea.2012.08.080

Microstructure and Texture Evolution During Hot Compression of Cast and Extruded AZ80 Magnesium Alloy Paresh Prakash, Amir Hadadzadeh, Sugrib Kumar Shaha, Mark A. Whitney, Mary A. Wells, Hamid Jahed, and Bruce W. Williams

Abstract

Uniaxial compression tests were conducted on cast and extruded AZ80 alloys at 400 °C and a strain rate of 0.1 s−1 up to a true strain of 1.0. Microstructure and texture evolution during hot deformation was studied using optical microscopy, X-ray diffraction macrotexture analysis and electron backscatter diffraction. The results indicate that dynamic recrystallization (DRX) occurred in the samples during deformation for both cast and extruded starting materials and the DRX fraction was found to increase with deformation strain level. The DRX grain size for both cast and extruded materials was measured as *5 µm and was independent of the deformation strain. In both cast and extruded materials, hot deformation led to the development of a sharp basal texture along the compression direction, which was attributed to grain rotation occurring during deformation, and preservation of deformation texture by the DRXed grains. Keywords











Recrystallization texture Cast AZ80 Extruded AZ80 Magnesium DRX EBSD Microstructure evolution Texture evolution Flow stress

P. Prakash (&)
S. K. Shaha
M. A. Whitney
H. Jahed Department of Mechanical and Mechatronics Engineering, University of Waterloo, 200 University Ave W, Waterloo, ON N2L 3G1, Canada e-mail: [email protected] A. Hadadzadeh Marine Additive Manufacturing Centre of Excellence (MAMCE), University of New Brunswick, Fredericton, NB E3B 5A3, Canada M. A. Wells College of Engineering and Physical Sciences, University of Guelph, 50 Stone Road East, Guelph, ON N1G 2W1, Canada B. W. Williams CanmetMATERIALS, Natural Resources Canada, 183 Longwood Rd S, Hamilton, ON L8P 0A5, Canada





Introduction Stringent norms on automobile fuel efficiency in the last two decades have both pushed and motivated academicians and automobile manufacturers towards lighter weight automobiles. One popular route to reduce vehicle’s weight is using lighter materials to fabricate structural components. Magnesium, owing to its high-specific strength and stiffness, is one of the important contenders in this space [1, 2]. Currently, the use of magnesium alloys in automotive components is relatively low and mostly limited to cast components [1, 2]. However, a cast material is generally not considered suitable for use in high-performance structural applications, owing to the inferior mechanical properties it exhibits relative to wrought materials [3]. One of the main challenges in a more widespread adoption of magnesium alloys in automotive components is their low formability at low deformation temperatures [3, 4]. Forging of magnesium alloys can be successfully accomplished at high deformation temperatures, typically *225 °C and above [4], which help activate higher order slip systems in these alloys [5]. Forging at these temperatures can also activate dynamic recrystallization (DRX) in the material [6], which leads to the generation of new strain-free grains within the deformed material, thereby reducing the dislocation density inside the material and enhancing its forgeability [6, 7]. Hot working (or forging) and DRX is also known to break up and refine the cast microstructure and thereby significantly improving the mechanical properties [8–12]. It is of both academic and industrial interest to study and compare the forgeability of cast and wrought materials, particularly for the use in fatigue critical components; however, limited prior research has been reported in this area [13, 14]. The present research aims to contribute to this knowledge by studying the microstructure and texture evolution with flow stress for cast and extruded AZ80 alloy via laboratory-scale uniaxial compression tests. In the current study, the microstructure and texture at various strain levels during hot compression of

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_15

89

90

P. Prakash et al.

cast and extruded AZ80 alloys were studied by means of optical microstructure analysis, X-ray diffraction (XRD) macrotexture analysis and electron backscatter diffraction (EBSD) analysis.

Experimental Details Starting Material The starting materials for this study were obtained from Magnesium Elektron North America Inc. in the form of industrially produced cast billets and extruded rods of AZ80 alloy of diameters 300 mm and 63.5 mm, respectively. The nominal chemical composition of the starting alloys is provided in Table 1, and the microstructure and texture of the starting materials are presented in Fig. 1. The as-cast material showed a coarse dendritic microstructure with an initial grain size of 178.9 ± 67.0 µm and a random initial texture. The extruded material on the other hand had a much finer, equiaxed grain structure, with an average grain size of 21.2 ± 2.7 µm. Referring to Fig. 1d, the extruded material showed a typical extrusion texture displayed by magnesium alloys, with basal poles aligned perpendicularly to the extrusion axis [15]. Mg17Al12 is the main precipitate phase for both the materials and was found to be distributed mainly at prior grain boundaries in lamellar form. The chemical composition of the precipitates was confirmed based on phase analysis using a computational thermodynamics database, FactSage® and energy dispersive spectra (EDS) analysis, which for brevity is not presented.

constant true strain rate of 0.1 s−1. Friction at contact surfaces was kept low using a graphite-based lubricant. Samples were heated to the test temperature at 5 °C/s and held for 60 s at the test temperature prior to compression. Samples were deformed to various strain levels and immediately water quenched (within 1 s) to preserve the corresponding microstructure state. The flow stress data presented in Fig. 3 was directly obtained from the Gleeble® 3500 machine. No corrections for barrelling or temperature variation during the tests (±2 °C) were made on the flow data. The deformed samples were sectioned along the compression axis, cold mounted in epoxy and prepared using standard metallographic procedure, followed by etching using an acetic-picral based solution. The metallographic investigations were focused on the central region of the deformed samples. The grain size of the DRXed grains was measured using the linear intercept method. Macrotexture measurements, based on XRD, were carried out using a Bruker D8-Discover machine. A collimator of size 1 mm was used, and an area of *1.5  2.5 mm was scanned on each sample for macrotexture measurements. Further details of the machine setup and control parameters can be found elsewhere [13]. Macrotexture results are presented in terms of ð0002Þ pole figures. EBSD measurements were carried out using a field emission gun scanning electron microscope (FEG-SEM) (FEI Nova NanoSEM-650) equipped with an EDAX EBSD detector. Prior to scanning, mechanically polished samples were additionally chemically polished using a 10% Nital solution. Scanning was performed on an area of *500 µm  500 µm at a step size of 0.25 µm. The scanned data was processed and analysed using TSL OIM 8.0 software. Misorientations greater than 10° were used to determine the grain boundaries.

Experimental Procedure

Results and Discussion Cylindrical samples for compression tests with dimensions of U 10  15 mm were machined from the cast and extruded billets, with their longitudinal direction parallel to cast and extrusion directions, respectively. Isothermal uniaxial compression tests were conducted using a Gleeble® 3500 thermal-mechanical simulation testing system available at the University of Waterloo at a temperature of 400 °C and a

Table 1 Composition of the alloys used in the present study (weight %) Al

Zn

Mn

Mg

Cast billet

8

0.31

0.15

rest

Extruded rod

8.2

0.33

0.14

rest

Flow Stress Evolution The measured flow curves of the AZ80 during hot compression are shown in Fig. 2. Both cast and extruded materials showed an initial rapid work hardening, followed by a flow stress peak and subsequent softening to a steady-state flow stress level. DRX typically initiates at strains slightly less than the peak stress level in the material [16]. Referring to Fig. 2, the extruded material achieved the flow stress peak at a slightly lower strain than in the cast material (0.04 in case of the extruded material, vis-à-vis 0.10 in case of the cast material), indicating an earlier onset of DRX. The extruded material also showed a more rapid work softening post the peak

Microstructure and Texture Evolution During Hot Compression …

(a)

(b)

91

(c)

(d)

Fig. 1 Microstructure and macrotexture of starting raw materials. a and b cast material, c and d extruded material. Some Mg17Al12 lamellas in the microstructure are marked by arrows. RD: radial direction, CD: casting direction, ED: extrusion direction

Microstructure Evolution

Fig. 2 Flow stress curves of cast and extruded materials deformed at 400 °C, 0.1 s−1

stress, which can possibly be attributed to a faster DRX kinetics in case of the finer grained extruded material [6, 7, 14]. Alternatively, the rapid flow softening displayed by extruded material could also be related to texture related softening effects, whereby the starting orientations in the extruded material (basal poles oriented perpendicular to the compression direction, also marked as A in Fig. 6d) were favorable for easy prismatic slip and second-order pyramidal slip (based on Schmid factor analysis using the EBSD data of the deformed samples), while the intermediate orientations (marked as B in Fig. 6d) were favorable for easy basal slip. Since the cast material showed a rapid development of basal texture (Fig. 3d), which puts a majority of the grains in hard (to deform) orientation for any further compression, it showed higher flow stress values compared to the extruded material. As the macrotexture results in Fig. 4d–f show, the texture in the extruded material also gradually reoriented to this strong basal orientation. At the deformation strain of 1.0, both cast and extruded materials showed comparable texture values, which might be the reason why they displayed similar flow stress values at this strain level.

The microstructure and macrotexture evolutions with strain for cast and extruded materials are shown in Figs. 3 and 4, respectively. Both materials showed an occurrence of a significant amount of DRX even at low deformation strain of 0.15, which corresponded to a strain just after the peak flow stress in either materials (Fig. 2). Referring to Fig. 3a, the cast material showed clear evidence of deformation twinning, with DRX occurring within some of the twins. DRX was additionally found to occur along the grain boundaries (GBs) (Fig. 3a). The deformation twinning mode in the cast material was not verified in the present work. Extruded material showed no microstructural evidence of deformation twinning and DRX was found to occur only along the grain boundaries (Fig. 4a). An absence of twinning during deformation of the extruded material can be attributed to the grain size dependence of twinning, whereby twinning activity in a material during deformation is found to reduce, and eventually cease, with grain refinement [17]. In both cases, the extent of DRX (i.e. the DRX fraction) was found to increase with deformation strain, though neither were fully (i.e. 100%) recrystallized even at the final deformation strain of 1.0 (Figs. 3c and 4c, respectively). DRX grain size variation with deformation strain for cast and extruded materials is plotted in Fig. 5. Both cast and extruded materials showed an average DRX grain size *5 µm, and the DRX grain size was observed not to vary significantly with deformation strain. The relative insensitivity of DRX grain size to deformation strain is in line with previous findings in the literature on AZ series of magnesium alloys [6, 14, 18]. A similar value of DRX grain size for fine-grained extruded material and coarse-grained cast material indicates that DRX grain size does not depend on prior grain size for AZ80 alloy. Previously, Watanabe et al. [19] and Beer et al. [14] have reported finer DRX grain size with finer prior grain size of the starting material for AZ31 and AZ61 alloys, and since the present

92

Fig. 3 Microstructure (a), (b) and (c), and macrotexture (d), (e) and (f) evolution in cast material with deformation strain. Arrows mark various microstructural features of interest—A: DRXed grains at prior

P. Prakash et al.

grain boundaries, B: DRXed grains within twins, C: twin, D: Non-DRXed grains. Legend—CD: compression direction, ND: normal direction, TD: transverse direction

Fig. 4 Microstructure (a), (b) and (c), and macrotexture (d), (e) and (f) evolution in extruded material with deformation strain. Interpretation of arrows and legend is same as in Fig. 3

Microstructure and Texture Evolution During Hot Compression …

Fig. 5 Variation of DRX grain size of cast and extruded materials with deformation strain at 400 °C, 0.1 s−1

findings indicate otherwise, this aspect of material behaviour needs to be further investigated.

Texture Evolution Referring to Figs. 3d–f and 4d–f, both the materials showed the development of a strong basal texture (basal poles aligning parallel to the compression direction during compression) as the deformation proceeded, though the development of basal texture in the cast material was found to be more rapid than in the extruded material. The initial random texture of the starting cast material showed a strong basal texture already at a strain of 0.15 (Fig. 3d), and the texture intensities remained relatively stable at this level with further deformation (Fig. 3d–f). The texture reorientation of the extruded material, on the other hand, was found to be more gradual (Fig. 4d–f). The textures of the cast and extruded materials at the final deformation strain of 1.0 were comparable. In order to separate the contributions of texture of deformed and recrystallized grains to the bulk texture (macrotexture), EBSD was performed on samples deformed to a strain of 0.4 for both cast and extruded materials.

(a)

(b)

Fig. 6 Recrystallization and deformation texture of samples deformed to strain of 0.4 for a, b cast and c, d extruded material. A, B and C mark various grain orientations of interest—A: Grains oriented close to initial texture in the extruded material, and are suited for prismatic and

93

DRXed grains and deformed grains in the microstructure were identified based on a grain orientation spread (GOS) criteria [20]. A grain was considered to have undergone DRX if it had a GOS of  0.5° (based on GOS distribution of an undeformed extruded AZ80 sample, that was annealed at 420 °C for 24 h, and would be expected to be fully recrystallized), and a grain size of  10 µm (based on the grain size distribution of deformed samples with GOS  0.5°). A GOS of >2° was used to identify deformed grains. The pole figures of DRXed and deformed grains for cast-deformed and extruded-deformed samples are plotted in Fig. 6. Note that the pole figures are plotted as scatter plots and not as contour plots so that the individual orientations of the grains can be visualized more clearly. Referring to Fig. 6b, d, it is evident that the deforming grains in both the cast and the extruded materials were progressively reorienting to align their basal poles parallel to the compression axis. For the cast material, this texture reorientation was rapid and is attributed to a combined effect of deformation (and grain rotation) through slip and twinning, while in case of extruded material, the deformation and corresponding grain rotation is gradual and is attributed to slip only. With regard to texture and orientation of the DRXed grains, referring to Fig. 6a, c, it is evident that DRXed grains, for both cast-deformed and extruded-deformed samples, were oriented very similarly as the nearby deformed grains. This suggests that DRX preserved the deformation texture for both cast and extruded AZ80 materials, for the studied deformation condition, although a more detailed analysis to confirm this is necessary. It is noted that for either materials, the DRXed grains showed significantly more scatter compared to the corresponding orientations in the deformed material, which might be related to multiple small DRXed grains nucleating from individual large prior grain(s), which is supposed to have internal orientation gradients owing to the imposed deformation [20]. This aspect of DRX texture also needs to be further investigated.

(c)

(d)

second-order pyramidal slip, B: Grains oriented for easy basal slip, C: Grains oriented with basal pole (c-axis) closely aligned to the compression axis and are suited for second-order pyramidal slip

94

Conclusions Hot deformation behavior of cast and extruded AZ80 alloys was investigated via uniaxial compression tests at 400 °C and a strain rate of 0.1 s−1. The main findings of the study are summarized in the following: (1) Cast material showed occurrence of twinning during hot compression, while the extruded material did not. (2) Both cast and extruded materials showed a substantial occurrence of DRX during deformation. DRX occurred in the cast material within deformation twins and along the grain boundaries, while for the extruded material, DRX occurred only along the grain boundaries in the material. The DRX grain size (*5 µm) was found to be independent of the cast/extruded nature of the starting material (and correspondingly, the prior grain size) and of the deformation strain level. DRX resulted in a significant grain refinement for both the starting materials. (3) DRXed grains were found to be oriented similarly as the nearby deformed grains (i.e. DRXed grains preserved the deformation texture). (4) A basal texture developed for both cast and extruded materials with progressive deformation. Development of basal texture in the cast material was much more rapid compared to that in the extruded material. (5) At the final deformation strain of 1.0, the flow stress values for starting cast and extruded materials were found to be comparable.

Acknowledgements The authors would like to gratefully acknowledge the financial support of the Natural Sciences and Engineering Research Council of Canada (NSERC), Automotive Partnership Canada (APC) program under APCPJ 459269—13 Grant with contributions from Multimatic Technical Centre, Ford Motor Company, and Centerline Windsor. One of the authors (PP) would like to thank Mr. Massimo Di Ciano of University of Waterloo for help in carrying out some of the uniaxial compression tests and Dr. Jian Li and Ms. Renata Zavadil of CanmetMATERIALS for assistance with metallography and EBSD analysis.

References 1. A. A. Luo, “Magnesium: Current and potential automotive applications,” JOM, vol. 54, no. 2, pp. 42–48, 2002. 2. M. K. Kulekci, “Magnesium and its alloys applications in automotive industry,” The International Journal of Advanced Manufacturing Technology, vol. 39, no. 9, p. 851–865, 2008. 3. J. Hirsch and T. Al-Samman, “Superior light metals by texture engineering: Optimized aluminum and magnesium alloys for automotive applications,” Acta Materialia, vol. 61, no. 3, pp. 818– 843, 2013.

P. Prakash et al. 4. T. Al-Samman and G. Gottstein, “Room temperature formability of a magnesium AZ31 alloy: Examining the role of texture on the deformation mechanisms,” Materials Science and Engineering: A, vol. 488, no. 1–2, pp. 406-414, 2008. 5. A. Chapuis and J. H. Driver, “Temperature dependency of slip and twinning in plane strain compressed magnesium single crystals,” Acta Materialia, vol. 59, no. 5, pp. 1986–1994, 2011. 6. T. Al-Samman and G. Gottstein, “Dynamic recrystallization during high temperature deformation of magnesium,” Materials Science and Engineering, vol. 490, no. 1–2, pp. 411–420, 2008. 7. F. J. Humphreys and M. Hatherly, “Recrystallization of Single-Phase Alloys,” in Recrystallization and Related Annealing Phenomena, Elsevier, 2004, pp. 215–268. 8. Z. Wang, Y. Yang, B. Li, Y. Zhang and Z. Zhang, “Effect of hot-deformation on microstructure and mechanical properties of AZ80 magnesium alloy,” Materials Science and Engineering: A, vol. 582, pp. 36–40, 2013. 9. D. Toscano, S. K. Shaha, B. Behravesh, H. Jahed and B. Williams, “Effect of Forging on Microstructure, Texture, and Uniaxial Properties of Cast AZ31B Alloy,” Journal of Materials Engineering and Performance, vol. 26, no. 7, p. 3090–3103, 2017. 10. D. Toscano, S. K. Shaha, B. Behravesh, H. Jahed and B. Williams, “Effect of forging on the low cycle fatigue behavior of cast AZ31B alloy,” Materials Science and Engineering: A, vol. 706, pp. 342– 356, 2017. 11. S. M. H. Karparvarfard, S. K. Shaha, S. B. Behravesh, H. Jahed and B. W. Williams, “Microstructure, texture and mechanical behavior characterization of hot forged cast ZK60 magnesium alloy,” Journal of Materials Science & Technology, vol. 33, no. 9, 2017. 12. A. Gryguc, S. K. Shaha, S. B. Behravesh, H. Jahed, M. Wells, B. Williams and X. Su, “Monotonic and cyclic behaviour of cast and cast-forged AZ80 Mg,” International Journal of Fatigue, vol. 104, pp. 136–149, 2017. 13. A. Gryguc, S. B. Behravesh, S. K. Shaha, H. Jahed, M. Wells, B. Williams and X. Su, “Low-cycle fatigue characterization and texture induced ratcheting behaviour of forged AZ80 Mg alloys,” International Journal of Fatigue, vol. 116, pp. 429–438, 2018. 14. A. Beer and M. Barnett, “Microstructural Development during Hot Working of Mg-3Al-1Zn,” Metallurgical and Materials Transactions A, vol. 38, no. 8, p. 1856–1867, 2007. 15. M. Shahzad and L. Wagner, “Influence of extrusion parameters on microstructure and texture developments, and their effects on mechanical properties of the magnesium alloy AZ80,” Materials Science and Engineering: A, vol. 506, no. 1–2, pp. 141–147, 2009. 16. E. I. Poliak and J. J. Jonas, “Initiation of Dynamic Recrystallization in Constant Strain Rate Hot Deformation,” ISIJ International, vol. 43, no. 5, pp. 684–691, 2003. 17. M. R. Barnett, Z. Kehsavarz, A. G. Beer and D. Atwell, “Influence of grain size on the compressive deformation of wrought Mg–3Al– 1Zn,” Acta Materialia, vol. 52, no. 17, pp. 5093–5103, 2004. 18. X. Yang, H. Miura and T. Sakai, “Dynamic Evolution of New Grains in Magnesium Alloy AZ31 during Hot Deformation,” Materials Transactions, vol. 44, no. 1, pp. 197–203, 2003. 19. H. Watanabe, H. Tsutsui, T. Mukai, K. Ishikawa, Y. Okanda, M. Kohzu and K. Higashi, “Grain Size Control of Commercial Wrought Mg-Al-Zn Alloys Utilizing Dynamic Recrystallization,” Materials Transactions, vol. 42, no. 7, pp. 1200–1205, 2001. 20. N. Allain-Bonasso, F. Wagner, S. Berbenni and D. P. Field, “A study of the heterogeneity of plastic deformation in IF steel by EBSD,” Materials Science and Engineering: A, vol. 548, pp. 56– 63, 2012.

Experimental Investigation of Friction Coefficient of Magnesium Alloy Developed Through Friction Stir Processing with PKS Ash Powder Particles R. S. Fono-Tamo, Esther Titilayo Akinlabi, and Jen Tien-Chien

Abstract

Magnesium metal alloys have application in a variety of engineering field. The inclusion of a number of metal particles into pure magnesium to improve its properties has been on the rise. The method of inclusion has gone pass the conventional powder metallurgy or stir casting method. In the current study, friction stir processing (FSP) was used for the embedment of palm kernel shell (PKS) ash particle into a magnesium substrate. Microstructure analysis of the developed composite showed a well-distributed PKS ash particles into the magnesium metal matrix. The Vickers hardness test shows an improvement on the hardness of the developed surface composite, especially at the middle and end of the specimen with respective values of 62.65 and 63.27 when compared to that of the base metal. Friction test was done under various loading of 1 and 10 N at a constant speed and relative humidity of 70%. The results revealed mean coefficient of friction of 0.857 and 0.478 for 1 N and 10 N loads, respectively. Friction stir processing proves to be an adequate technique of improving the surface properties of magnesium alloy when using PKs ash powder as reinforcement.





Keywords








FSP Surface composites Magnesium alloy PKS ash Vickers hardness Friction coefficient

Introduction The use of magnesium in the military date back to 1886 in Germany where it was used in tracer bullets, flares and pyrotechnics. Gupta and Wong [1] alleged that magnesium R. S. Fono-Tamo (&)
E. T. Akinlabi
J. Tien-Chien Department of Mechanical Engineering Science, University of Johannesburg, Johannesburg, 2006, South Africa e-mail: [email protected]

was a choice material to produce fuselages, wheels and engine parts in some war vehicles during the second world war. Furthermore, other important uses of magnesium nanocomposites in the military comprise the manufacture of gear housing and engine blocks for their ground vehicle, helmet and ballistic protection as well as in body armor. Magnesium is described as the lightest structural metal and exhibits analogous melting temperature and yield strength to the already commonly used aluminium metal. Reinforcing magnesium with nanosized particles is an attractive way to improve the mechanical properties of the metals while its ductility remains unchanged. Magnesium composites are usually processed via liquid- or solid-based processing techniques [1]. The types of reinforcements used to synthesize magnesium nanocomposites include ceramic, metallic, intermetallics and recently industrial and agricultural wastes. In a bid to reduce the weight of vehicles and increase fuel economy while reducing pollution, the use of lightweight materials has been in increase in transportation technology. Magnesium-based products suffer of poor general and galvanic corrosion resistance, which in turn affect their maintenance and lifetime [2]. Abdullah et al. [3] assert that the US military industry currently has a keen interest in magnesium-based alloys. Abdullah and coworkers used disintegrated melt deposition technique to develop a AZ31B alloy in which Pb was included in certain quantities for ballistic application. The absorption energy was affected by the percentage of Pb added to AZ31B and 1% Pb was deemed to be the best percentage of Pb that can be included in AZ31B for better performance. Their choice of the AZ31B series from the magnesium alloy family was based on the fact that it is normally used in aerospace and automotive applications. Besides good room temperature strength, ductility, corrosion resistance, and excellent weldability, AZ31B also has high-energy absorption as far as ballistic impact is concerned. In this study, AZ61 sheet which has similar properties with AZ31B and even more that is, it is known to have high strength, is highly corrosion resistant and has excellent thermal stability of its crystalline

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_16

95

96

structure was adopted. Palm Kernel Shell (PKS) is recovered as waste product in palm oil producing plant. Whilst, large quantities of PKS are generated annually, only a small fraction is used for fuel and other applications, such as palliative for un-tarred road, producing activated carbon and as building material. The unused PKS are dumped around the processing mill, constituting environmental and economic liability to the processor [4]. Mechanical properties of magnesium alloys have been demonstrated to be greatly enhanced when using severe plastic deformation (SPD) techniques; equal channel angular pressing (ECAP) was used by Jahadi et al. [5], accumulative roll bonding (ARB) was used by Saito et al. [6], high-pressure torsion (HPT) was used by Kai et al. [7], multidirectional forging (MDF) was used by Chen et al. [8] and cyclic closed-die forging (CCDF) by Guo et al. [9]. The above-mentioned techniques are also very effective in refining the microstructure of magnesium alloys. In recent decade, friction stir processing (FSP) which is described as a solid-phase processing technique [10] has been added to the list. FSP as explained by Mishra and Ma [11] is majorly characterized by the fact that a rotating non-consumable tool that contains a pin at the end helps alter the microstructure of metallic sheets without melting the material and at the same time induces intense plastic deformation on the sheets. Furthermore, it is a technique broadly used in the production of a number of metallic surface composites. Being described as a class of materials in which the surface contains dispersed secondary phase while the core of the material remains the same, surface composites are known to be produced through centrifugal casting, chemical or physical vapor deposition, plasma spraying, and laser treatment [12– 14]. However, the drawback of these techniques is the melting of the substrate except for FSP in which the material does not melt. FSP has proven to be an effective technique to produce surfaces with improved corrosion resistance, surface hardness and wear resistance as found a number of studies in the literature [15–21]. Use of agricultural waste PKS ash in creating lightweight materials that cut down pollution and fuel consumption of vehicles is an appealing idea. PKS ash which is an organic derivative may substantially contribute in enhancing the corrosion resistance of the magnesium-based metal. In this study, FSP is used to produce magnesium surface composite with PKS ash as reinforcing powder. This is to develop a lightweight magnesium composite with enhanced surface hardness, and wear resistance which could be used to produce military accessories.

R. S. Fono-Tamo et al.

Materials and Methods In this study, the palm kernel shell (PKS) used was obtained from a local palm oil processing mill in Nigeria. A hammer mill and a ball mill were used to reduce the grains to finer size. The powder material was loaded into a graphite crucible then fired in electric resistance furnace at temperature of 1300 °C for an hour to obtain palm kernel shell ash [22]. Furthermore, mesh size under 50 lm were obtained via sieving the powder by means of a sieve shaker [23]. The FSP process followed here is as described by [24] whereby AZ61 Mg alloy surface plates were used as base metal. Chemical elements found in AZ61 Mg alloy adopted in the present study are in weight % of Al (6.17), Fe (0.005), Si (0.04), Mn (0.3), Ca (0.0012), Zn (0.75) and Mg (Bal). The plates were grooved and filled with PKS ash powder that was to be distributed as secondary phase by means of FSP. The groove was of V-shape type with depth of 3.5 mm and width of 1.9 mm. The tools used were made from H13 tool steel and hardened to 52 HRC. The friction stir process was conducted on five-Axis friction stir welding machine. The tool rotational and travel speed used were 1200 rpm and 20 mm/min, respectively. Perpendicular and lateral forces of 20 kN were applied during the process and for the material to withstand such force, the workpiece having dimensions of 650  265  25 mm3 was securely clamped. The tool used had a tilt angle of zero and three passes were executed to compress the powder across the material surface for all the experiments. Surface analysis of the developed magnesiumbased MMCs was conducted using optical microscope and the related micrographs obtained. Samples were prepared according to the normal process of grinding, polishing and etching. The micrographs were recorded at the beginning, middle and at the end of the plate after FSP. The micro-hardness was measured using a digital Vickers hardness (HV) tester HXD-1000TM/LCD at a load of 0.5 kN for 15 s and specimens were taken from the beginning, the middle and the end of the FSPed metal. Friction coefficient of the specimens was determined using a Anton Paar GmbH standard tribometer version 7.3.13 where a single ball runs in a single way mode at a linear speed of 4.97 cm/s in an ambient temperature of 25 °C. Two different loads of 1 N and 10 N were applied in other to appreciate the material behavior under diverse loading. The average running time for the experiment was about 15 min and a plot of the friction coefficient was generated with the values of the COF displayed. Focus was on the specimens with well-distributed PKS ash particles.

Experimental Investigation of Friction Coefficient …

Results and Discussion Microstructural visualization of the FSP processed surfaces is presented in Fig. 1. Micrographs were taken from the specimens at the beginning, middle and end. Substantial PKS ash distribution on the surface is noted. Figure 1a shows good distribution of the reinforcement material but sparingly distributed when compared to Fig. 1b, c which shows the reinforcement evenly distributed. It is noticeable from Fig. 1 that there is a good bonding relation between the matrix and the reinforcing powder. The ashes are black in color and are easily noticeable when compare to the surrounding environment with uniform color which can be described as the magnesium matrix. The result of the hardness analysis shows that there is an increase in the hardness of the processed magnesium

97

composite surface with the hardness at the middle and the end being relatively the same (62.65 and 63.28), respectively. The observed high hardness value at the end of the processed material can be explained by the much better distribution of the reinforcing agent as said earlier and as seen in the microstructure. This contributes in strengthening the metal composite in a distributed manner as the micro-indenter might have regularly touched areas strengthen by the reinforcement powder. In the case of the specimen acquired at the beginning, the micro-hardness indenter might have encountered more of the base metal. The assumption being that the PKS ash particle was not yet well distributed at this stage thus creating clusters of powder material in some places and pockets of base metal in other places (Fig. 2).

Fig. 1 Micrographs of the developed surface composite a beginning b middle and c end

Micro-Hardness

98

R. S. Fono-Tamo et al. 65 64 63 62 61 60 59 58 57 56 55 54

clouded the wear track creating plateaus which serve as barrier and inhibit the effect of the rolling ball on the contact surface thus reducing the friction effect reason why the low COF at 10 N.

63.27

62.65

59.36

Conclusion

Beginning

Middle

End

Fig. 2 Vickers Hardness of the surface composites

The acquired coefficient of friction emanates from the samples derived at the end of the experiment where even distribution through microstructure analysis and better hardness properties were observed. Obviously as seen in Fig. 3, the load has a considerable impact on the material as one can notice that the COF with 1 N load (0.857) is different to the COF with 10 N load (0.478). The graphs depart from each other right at the beginning of the experiment. High COF with 1 N load is an indication of considerable material loss during the friction process. Ordinarily, it is expected that the higher the load, the more pronounced the friction effect. This could be confirmed from the analysis of the wear track. But lower value of COF at 10 N load when compared to that of 1 N is a sign that some tribological phenomenon such as tribological influence of wear debris, debris agglomeration, transfer layers might be taking place on the wear track. Certainly in the case of 10 N, wear debris generated during the process might have

1.2

1N 10N

Coefficient of Friction

1.0

0.8

0.6

0.4

0.2

0.0 0

200

400

600

800

Time [s]

Fig. 3 Friction coefficient plots for the developed composites

1000

The current study has shown that palm kernel shell which is a waste product from agriculture can be used for engineering application. Friction stir processing is proven to be a viable technique to develop magnesium surface composites with palm kernel shell as reinforcement. Microstructure analysis showed evenly distributed ashes. Hardness of the base metal is improved with the inclusion of PKS ash. Coefficient of friction under different loading conditions is comparable to that of other surface composites already being used in various industries. Optimizing the use of palm kernel shell in engineering will increase its demand which may lead to it being a source of income to local populations where palm oil processing by-product is just a waste. Magnesium alloys are already being extensively used in the military and with the growing interested shown by the US military, and this study is providing a new magnesium alloy which is worth considering given it biodegradable nature. Acknowledgements Dr. Fono-Tamo and Prof. Tien-Chien Jen are thankful for the financial support from GES Fellowship of the University of Johannesburg, Johannesburg, South Africa.

References 1. Gupta M and Wong WLE (2015) Magnesium-based nanocomposites: Lightweight materials of the future. Materials Characterization 105: 30–46, http://dx.doi.org/10.1016/j.matchar.2015.04. 015 2. Mathaudhu SN, Nyberg EA (2014) Magnesium alloys in U.S. Military applications: Past, current and future solutions. In: Mathaudhu SN, Luo AA, Neelameggham NR, Nyberg EA and Sillekens WH (eds) Magnesium Technology 2014. The Minerals, Metals & Materials Society, Pittsburgh; John Wiley & Sons, Inc, p 71–76 3. Abdullah MF, Abdullah S, Omar MZ, Sajuri Z and Sohaimi RM (2015) Failure observation of the AZ31B magnesium alloy and the effect of lead addition content under ballistic impact. Advances in Mechanical Engineering 7(5): 1–13. https://doi.org/10.1177/ 1687814015585428 4. Fono-Tamo, R., “Agro-Waste Based Friction Material for Automotive Application,” SAE Technical Paper 2014- 01-0945, 2014, https://doi.org/10.4271/2014-01-0945 5. Jahadi R, Sedighi M, Jahed H (2014) ECAP effect on the micro-structure and mechanical properties of AM30 magnesium alloy. Materials Science and Engineering A 593:178–184. https:// doi.org/10.1016/j.msea.2013.11.042 6. Saito Y, Utsunomiya H, Tsuji N, Sakai T(1999) Novel ultra-high straining process for bulk materials—development of the

Experimental Investigation of Friction Coefficient …

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

accumulative roll-bonding (ARB) process. Acta Materialia 47(2): 579–583. https://doi.org/10.1016/S1359-6454(98)00365-6 Kai M, Horita Z, Langdon TG (2008) Developing grain refinement and superplasticity in a magnesium alloy processed by high-pressure torsion. Materials Science and Engineering A 488: 117–124. https://doi.org/10.1016/j.msea.2007.12.046 Chen Q, Shu D, Hu C, Zhao Z, Yuan B (2012) Grain refinement in an as-cast AZ61 magnesium alloy processed by multi-axial forging under the multitemperature processing procedure. Materials Science and Engineering A 541: 98–104. https://doi.org/10.1016/ j.msea.2012.02.009 Guo W, Wang Q, Ye B, Zhou H (2013) Enhanced microstructure homogeneity and mechanical properties of AZ31–Si composite by cyclic closed-die forging. Journal of Alloys and Compounds 552: 409–417. https://doi.org/10.1016/j.jallcom.2012.11.067 Ratna SB (2016) Different strategies of secondary phase incorporation into metallic sheets by friction stir processing in developing surface composites. International Journal of Mechanical and Materials Engineering 11:12. https://doi.org/10.1186/s40712-0160066-y Mishra, R.S., Ma, Z.Y., 2005, “Friction stir welding and processing”, Materials Science and Engineering R, Vol. 50, pp. 1–78. https://doi.org/10.1016/j.mser.2005.07.001 Ayers JD and Tucker TR (1980) Particulate-TiC-hardened steel surfaces by laser melt injection. Thin Solid Films 73(1): 201–207. https://doi.org/10.1016/0040-6090(80)90352-1 Chawla, Nikhilesh, Chawla, Krishan K (2013) Metal Matrix Composites. Springer New York. https://doi.org/10.1007/978-14614-9548-2 Kapranos P, Carney C, Pola A, Jolly M R (2014) Advanced Casting Methodologies: Investment Casting, Centrifugal Casting, Squeeze Casting, Metal Spinning, and Batch Casting. In: Hashmi, S (ed) Comprehensive Materials Processing. Elsevier Science, p 40–66 Faraji G, Dastani O and Akbari-Mousavi SAA (2011) Effect of process parameters on microstructure and micro-hardness of AZ91/Al2O3 surface composite produced by FSP. Journal of Materials Engineering and Performance 20: 1583–1590. DOI: https://doi.org/10.1007/s11665-010-9812-0 Devinder Y and Bauri R (2011) Processing, microstructure and mechanical properties of nickel particles embedded aluminum

99

17.

18.

19.

20.

21.

22.

23.

24.

matrix composite. Materials Science and Engineering A 528(3): 1326–1333. https://doi.org/10.1016/j.msea.2010.10.035 Soleymani S, Abdollah-zadeh A, Alidokht SA (2012) Microstructural and tribological properties of Al5083 based surface hybrid composite produced by friction stir processing. Wear 278–279: 41–47. https://doi.org/10.1016/j.wear.2012.01.009 Liu Q, Ke L, Liu F, Huang C and Xing L (2013) Microstructure and mechanical property of multi-walled carbon nanotubes reinforced aluminum matrix composites fabricated by friction stir processing. Materials & Design 45: 343–348. https://doi.org/10. 1016/j.matdes.2012.08.036 Rajiv Sharan Mishra, Partha Sarathi De, Nilesh Kumar (2014) Friction Stir Welding and Processing. Springer International Publishing Switzerland. DOI:https://doi.org/10.1007/978-3-31907043-8 Ratna Sunil B, Sampath Kumar TS, Chakkingal U, Nandakumar V and Doble M (2014a) Friction stir processing of magnesium– nanohydroxyapatite composites with controlled in vitro degradation behavior. Materials Science and Engineering C 39: 315–324. https://doi.org/10.1016/j.msec.2014.03.004 Ratna Sunil B, Sampath Kumar TS, Chakkingal U, Nandakumar V and Doble M (2014b) Nano-hydroxyapatite reinforced AZ31 magnesium alloy by friction stir processing: A solid state processing for biodegradable metal matrix composites. Journal of Materials Science: Materials in Medicine 25: 975–988. https://doi. org/10.1007/s10856-013-5127-7 Kong SH, Loh SK, Bachmann RT, Choob YM, Salimon J and Abdul Rahim S (2013) Production and Physico-Chemical Characterization of Biochar from Palm Kernel Shell. In AIP Conference Proceedings 1571: 749–752. https://doi.org/10.1063/1.4858744 K.K. Alaneme, P.A. Olubambi, A.S. Afolabi, M.O. Bodurin (2014): “Corrosion and Tribological Studies of Bamboo Leaf Ash and Alumina Reinforced Al-Mg-Si Alloy Matrix Hybrid Composites in Chloride Medium”. International Journal of Electrochemical Science, 9: 5663–5674 Sanusi KO and Akinlabi ET (2017) Fabrication of Friction Stir Processed 6082-T6 Aluminum Alloy With Reinforced Powder. International Mechanical Engineering Congress and Exposition, Advanced Manufacturing 2; Tampa, Florida, USA, November 3– 9, 2017, https://doi.org/10.1115/imece2017-71305

A Review and Case Study on Mechanical Properties and Microstructure Evolution in Magnesium–Steel Friction Stir Welding Suryakanta Sahu, Omkar Thorat, Raju Prasad Mahto, Surjya Kanta Pal, and Prakash Srirangam

Abstract

Weight minimization and global environmental policies on carbon content open a new research avenue towards materials and manufacturing processes in transport industries. Friction stir welding (FSW) process is a combination of frictional heating and stirring action where materials are joined in their solid state. In this study, a review has been made on the joining status of magnesium alloys to steel by using FSW. Present problems and future opportunities of magnesium to steel joining with the help of FSW are also stated. A case study has also been presented where the joint characteristics of AZ31B to AISI 304 sheets fabricated in lap configuration by FSW have been investigated by varying tool rotational speed (600, 1000, and 1800 rpm) and varying weld speed (40, 200, and 350 mm/min). A maximum weld joint efficiency of 79% of the AZ31B base alloy has been achieved with a parametric combination of 600 rpm and 350 mm/min. Keywords









Automotive industry Magnesium alloys Stainless steel Friction stir welding Dissimilar joints

S. Sahu (&) Advanced Technology Development Centre, Indian Institute of Technology Kharagpur, Kharagpur, 721302, India e-mail: [email protected] O. Thorat Department of Mechanical Engineering, Dr. Babasaheb Ambedkar Technological University, Raigad, 402103, India R. P. Mahto
S. K. Pal Department of Mechanical Engineering, Indian Institute of Technology Kharagpur, Kharagpur, 721302, India P. Srirangam Warwick Manufacturing Group (WMG), University of Warwick, Coventry, CV4 7AL, UK

Introduction In recent times, the demand for reduction in fuel consumption and also net weight of components has increased the interest of various transport industries towards the usage of lightweight materials. Aluminium (Al), magnesium (Mg) and plastics are used considerably in automotive industries to solve the weight-related problems [1–3]. Among these materials, Mg and its alloys are suitable for automotive applications due to their unique characteristics such as high specific stiffness, high specific strength, low density, ductility, and better damping characteristics [1, 4]. Table 1 lists the physical properties of Mg, Al, and steel [5], where it can be observed that Mg alloys can offer considerable weight savings as compared to Al and steel. Owing to the various characteristics of Mg alloys, many industries like automotive, aerospace and ship building are interested in using hybrid structures that result in desirable weight saving [6]. So, due to increasing usages of Mg alloys in industrial applications, the issues related to the joining of these materials need to be solved. Various conventional joining techniques like ultrasonic welding [7], resistance welding [8–10] laser welding [11, 12], hybrid laser–TIG welding [1, 13–17], diffusion-brazed [2] and laser penetration-brazed [18, 19] have been used for the development of Mg–steel joints. But, a reliable joint between Mg and steel is challenging by using conventional fusion welding process because of the significant difference in their melting point (about 900 °C) and immiscible characteristics in both solid and liquid states. The melting temperature of Mg is 649 °C and that of Fe is 1538 °C. The huge difference in melting points of Mg and Fe renders them difficult to join by using fusion welding methods. While the solubility of Fe in Mg is 0.00041 atomic weight %, the solubility of Mg in Fe is zero, as reported in Mg–Fe binary alloy diagram [20–26]. To produce a joint in between Mg and steel with appreciable joint strength with the help of fusion welding processes, different types of coated steels or zinc (Zn), nickel (Ni) and

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_17

101

102

S. Sahu et al.

Table 1 Mg, Al and steel physical properties [5]

Metals

Mg

Al

Fe

Melting point (oC)

650

660

1536

3

Density (at 20 °C) (g/cm )

1.74

2.70

7.86

Tensile strength (MPa)

240

320

350

Modulus of elasticity (106 MPa)

44.126

68.947

206.842

Thermal expansion coefficient (10−6/C)

25.5

23.6

11.7

Lattice structure

BCC

FCC

hcp

produce a sound joint by FSW, a rotating tool is utilized. This tool is inserted into the faying surface of the metal plates. The tool is then moved along the direction of the welding line when the shoulder is in proper contact with the top surface of the workpiece to avoid softening of the materials. Tool rotational speed (x), welding speed (v), tilt angle (a) and plunge depth (pd) are the major process parameters in FSW. The x produces both the heat and deformation simultaneously to produce a solid-state joint [35–37]. The schematic representation of the lap joint in FSW is depicted in Fig. 1. The research progress of Mg to various kinds of steel joints with FSW in lap configuration is reported in Table 2. The effect of pin length and surface conditions of steel on mechanical properties and microstructure were investigated for FSLW of AZ31 to Zn-coated steel [20, 21]. The Zn coating results in improved weldability of Mg alloy and steel. It has been observed that Mg alloys joints with Zn coating exhibit higher fracture load as compared to brushed finished steel. On the other hand, for brushed finished steel joints, the failure loads of the joint have been found to be increasing with increasing pin length. It has also been seen that in case of Zn-coated steel, with increasing pin length, the tensile strength did not improve. This is because of the formation of voids in the stir zone. The influence of v on mechanical properties and temperature near the weld was studied. The temperature during welding process was less when v was higher and vice versa. No intermetallic compounds were observed at a low v for the zinc-coated steel [26]. The effect of tool design and v on mechanical properties and

Fig. 1 Representation of lap joint

copper (Cu) interlayers (filler metals) in between Mg and steel interface were used for achieving excellent mechanical strength as compared to traditional Mg–steel joints. The formation of favorable interfacial structures or intermetallic compounds (IMCs) with the Mg alloys enhanced the joint strength during the fusion welding process [1, 7, 12–17]. But the use of these filler metals, as well as the coated steels, increases the manufacturing cost of end product. Friction stir welding (FSW) is being widely used because of its energy efficiency, environment friendliness and versatility [20, 27, 28]. FSW has the potential to join materials like magnesium, copper, copper alloys, lead, zinc, steel, and titanium with a thickness ranging from 1 to 50 mm [20 and 28–30]. FSW process is also capable of joining both similar and dissimilar combination of materials without melting, and thus, the problems related to fusion welding processes like distortions of sheets, hot cracking, porosity, oxidation, and slag inclusions, wide heat affected zone, high residual stress, low strength, partial melting zones and segregations of alloys, etc., are not present in this process [31–34]. To Table 2 Research progress of Mg to steel FSW in lap configuration Materials

Thickness

Joint efficiency or UTS or load

References

Mg

Steel

Mg

Steel

AZ31

Zn and brushed finished

1.6

0.8

2.3 and 0.6 kN

[20]

AZ31

Zn and brushed finished

1.6

0.8

–

[21]

AZ31

Low carbon steel Zn coated

1.6

0.8

3.4kN-65% of Zn-coated base metal

[22]

AZ31

Hot dipped galvanized high strength low alloy steel and electro-galvanized mild steel

2.33

0.8 and 1.5

80% for thinner sheet

[23]

AZ31

High strength low alloy steel and mild steel

2.33

0.8 and 1.5

–

[24]

AZ31

SUS 302

3

1

96.3 MPa

[25]

AZ31B (O-tempered)

Zn-coated DX54D

2

2

–

[26]

A Review and Case Study on Mechanical Properties …

103

microstructural evolution of lap joint of AZ31B and SUS302 steel was also investigated [25]. It was observed that at high v, less material is softened due to less heat generation which forms thin flashes. The flashes which are peeled off form macro-interlocked structure providing nail effect which enhances joining strength. Another characteristic feature such as saw tooth is observed due to the cutting and frictional effect. Mg may have been extruded into saw teeth by the stirring force that resulted in the formation of micro-interlocks at the weld interface. This interlocking helped in creating the zipper effect. The long flashes resulting in nail effect saw tooth because of zipper effect have been found to be responsible for enhancing the joining strength at higher values of v. The effect of x on joint strength and evolution of intermetallic compounds at joint interface of AZ31 and AZ61 alloys were investigated. It has been observed that with increasing Al content in Mg, joint strength also increases. The tensile test specimens fractured near the interface. In case of steel/AZ61 and AZ31, the intermetallic layer was found to be thicker at higher values of x which resulted in higher tensile strength, whereas the steel–pure magnesium joint has lower tensile strength. The aluminium depletion layer has been found to be significantly related to the tensile properties of the steel–Mg joint. From the literature, it can be concluded that the amount of work that has been carried out for Mg to steel joining in lap Table 3 AZ31B and AISI 304: chemical compositions (wt%)

Table 4 AZ31B and AISI 304: mechanical properties

Fig. 2 Schematic representation of a FSW setup and b FSW tool

configuration with the help of FSW is limited. Previous research work developed the joints with the help of coating on the steel sheet, or directly coated steels have been used. In the present work, the joints were developed with the help of Mg to conventional single-phase austenitic grade steel without any coating. Finally, the effects of x and v on various properties of the joints have been investigated keeping a and pd as 1o and 0.1 mm, respectively.

Materials and Methods The base materials chosen for the experiments in lap configuration were Mg (AZ31B) and austenitic stainless steel (AISI 304) sheets; the dimension of both was 100  80 mm2, and thicknesses were 1.5 and 1 mm, respectively. Tables 3 and 4 represent the chemical and physical properties of the two base alloys, respectively. The FSW joints were fabricated by using a NC linear FSW machine (Make: ETA India, WS004). The detailed welding arrangement and configuration are shown in Fig. 2a. The overlap width for the welding process was kept as 32 mm. A non-consumable tool made of tungsten carbide (WC-10%Co) with circular pin profile with shoulder diameter and the pin diameter of 16 mm and 5 mm, respectively, was used to form the joints that are shown in Fig. 2b.

Elements

Mn

Fe

Al

Si

Cu

Zn

Ni

Mg

AZ31B

0.44

0.0029

2.97

0.033

0.0076

0.82

0.00076

Bal

Elements

C

Cr

Fe

Mn

Ni

P

S

Si

AISI 304

0.073

19

68.64

1.6

9

0.036

0.03

0.7

Properties

Ultimate tensile strength (MPa)

Yield strength (MPa)

Elongation (%)

AZ31B

252

143

154.0

AISI 304

585

240

40.0

104

S. Sahu et al.

Table 5 Process parameters for the experimentation Experiment no.

x (rpm)

v (mm/min)

1

600

40

2

600

200

3

600

350

4

1000

40

5

1000

200

6

1000

350

7

1800

40

8

1800

200

9

1800

350

Pd (mm)

a (°)

0.1

1

Fig. 3 a Tensile testing and b sample dimension

The process parameters selected for the experimentations are mentioned in Table 5. The tensile specimens were tested on a 5 kN servo-hydraulic UTM (Instron, 3365). The crosshead speed was set as 1 mm/min. The tensile sample dimension is shown in Fig. 3b. Scanning electron microscopy was used for analysing the fractured surface of the tensile samples (Zeiss, Evo 60). To study the microstructural changes occurred during welding, a standard polishing method with 220–2000 grit emery papers followed by diamond polishing of the samples was done as per the ASTM E3-11 (2017) standard [38]. Microetching of the polished samples was done as per the ASTM E407-07 [39]. The microstructure of different zones was viewed and captured with the help of a stereo zoom microscope (Leica, S6D). The XRD analysis was performed to observe the different phases formed at the interface during the welding process. Microhardness of different zones of the weldment was measured using Vickers microhardness testing machine (Buehler, Micromet 5103) as per ASTM E384-16 [40].

Results and Discussion Figures 4 and 5 show the joint strength of FSW joints obtained with different x and v, respectively. All the samples were tested under the advancing side (AS) loading

Fig. 4 Tensile strength for different x

condition; i.e. AZ31B Mg alloy was gripped into the driven crosshead, as shown in Fig. 3a. The relationship between x and tensile strength is shown in Fig. 4. It is observed that with an increase in x, the joint strength decreases. The joint strength obtained with 40 mm/min welding condition was 158.1 MPa. Same nature of graphs is observed for v of 200 and 350 mm/min where other conditions such as pd and a were kept constant. The maximum value of joint strength obtained at 200 mm/min is 168.61 MPa and for 350 mm/min is 198.75 MPa which are shown in Fig. 4. The relationship between v and tensile strength is shown in

A Review and Case Study on Mechanical Properties …

Fig. 5 Tensile strength for different v

Fig. 5. The tensile strength of the joint increases continuously with the increase in v. For 600 rpm, the maximum strength of the joint obtained is 158.1 MPa, which is shown in Fig. 4. For 1000 rpm, the maximum tensile strength obtained is 167.59 MPa, and for 1800 rpm, it is found to be 156.54 MPa as shown in Fig. 5. At parametric combination of 600 rpm and 350 mm/min, a higher joint strength of 198.75 MPa (nearly 79% of the Mg base alloy) was obtained. The fracture of all the tensile specimens occurred in an area between the SZ and the AS. Further, it has been seen that the joint strength decreases with higher heat input and lower cooling rates. High heat input coarsens the microstructure that leads to ease

Fig. 6 Optical micrographs for different x and v

105

movement of dislocations. Eventually, weld strength was decreased. The optical micrograph observations for the different process parameters were represented schematically in Fig. 6. For all the process parameters, hooks were seen. It was also identified that for low v like 40 mm/min, the size of the hook was bigger. The size of the hook altered the joint strength in FSW. Formation of hook primarily reduced the thickness of the sheet in the SZ which can be considered as a large stress concentration for which failures or crack propagation are mostly found at that location. As the v increases to 350 mm/min, the size of the hook reduced as indicated in Fig. 6. It has also been observed that with increasing x at high v, less material is softened and less flash is trapped in the stir zone. While welding Mg plate to steel, the pin is entered through Mg and inserted into steel. A strong mechanical and metallurgical bonding in the joint was found because of the cutting effect. As the pin is penetrated into the steel sheet, hard steel surface layers were removed by pin, and peeled steel scraps fell into the stir zone which plays a vital role in the material transfer as shown in Fig. 7. Another typical feature is witnessed at the interface due to cutting and the frictional effect. This effect forms ‘saw tooth’-like structure in the SZ, as shown in Fig. 6f. Figure 8 shows the fracture morphology of AZ31B and AISI 304 joints. The fracture of joints reveals cleavage terrace and river pattern with small dimples which indicates

106

Fig. 7 SEM photographs showing steel scrap in SZ: a x = 600 rpm and v = 350 mm/min; b x = 1800 rpm and v = 350 mm/min

that the fracture behaviour of the joints is quite close to brittle cleavage fracture. At x = 600 rpm and v = 200 mm/min, and at x = 1000 rpm and v = 350 mm/min, deeper and denser dimples are found which possesses better ductility but lower strength. The microhardness observations were taken along the thickness direction from the Mg surface towards the steel surface. The observations were noted at three different

Fig. 8 Pictures showing fractured surface of Mg substrate

S. Sahu et al.

locations: AS, retreating side (RS) and SZ. Variation of microhardness with distance from the top Mg surface at a distance of 0.2 mm vertically is shown in Fig. 9. When hardness was measured from top Mg surface, it was observed that the hardness increases gradually up to the Mg and steel interface and then comes down when measured in the steel. As we move from the AS to RS, the hardness increases. The reason behind this can be the decreasing grain size. During welding, the area under the tool tip deforms plastically. The reason for Mg hardness increase was due to the higher rate of cooling. The higher cooling rate produces smaller grains for which hardness was high. The changes in grain structure with varying rotational speed are shown in Fig. 10. In the case of steel, high-temperature generation and re-crystallization near the interface increase the hardness, but as we move down, this effect reduces, so less hardness value is obtained. The grain sizes obtained are 8.09 µm, 7.65 µm and 8.69 µm for 600, 1000, and 1800 rpm, respectively, at 350 mm/min.

A Review and Case Study on Mechanical Properties …

107

Fig. 9 Microhardness along the thickness direction for different values of x and constant v (350 mm/min): a 600 rpm, b 1000 rpm, and c 1800 rpm

Fig. 10 HAZ grain structure at v = 350 mm/min: a x = 600 rpm, b x = 1000 rpm, and c x = 1800 rpm

Conclusions In this report, FSLW has been used successfully to join AZ31B and AISI 304 sheets. Effect of parameters, i.e. x and v, on mechanical properties of joint are studied. A joint strength of 198.75 MPa with respect to base AZ31B was obtained for lap welds joined by FSW. The microhardness across the joint indicates that the microhardness at the joint interface is higher than the base materials. ‘Hook’-like feature is observed at the interface of the joint which enhances the joint strength at high welding speed, and for low welding speed, it deteriorates the joint strength. Similarly, the ‘saw

tooth’ feature observed at the interface at high travel speed is the reason that promotes the zipper effect which ultimately increases the joint strength of the weld.

Future Scope Materials with a wide difference in melting point and with no solubility into each other are difficult to weld with the conventional welding process. Mg alloy and steel welded using laser and hybrid welding techniques showed defects. FSW process gives effective means to join dissimilar materials with minimized welding defects. By observing the feasibility of Mg

108

to steel joining in the present work, the research can be progressed in the direction of joining Mg with advanced steel families such as high strength steel (HSS) and advanced high strength steels (AHSS). The use of Mg with AHSS will provide the considerable weight loss and also high strength and stiffness to the automobile components. The Al to AHSS joints have already integrated into the automobile components, but there is also a scope to replace Al with Mg alloys.

References 1. Caiwang Tan, Liqun Li, Yanbin Chen, Wei Guo, “Laser-tungsten inert gas hybrid welding of dissimilar metals AZ31B Mg alloys to Zn coated steel,” Materials and Design 49 (2013) 766–773 2. Waled M. Elthalabawy, Tahir I. Khan, “Microstructural development of diffusion-brazed austenitic stainless steel to magnesium alloy using a nickel interlayer,” Materials Characterization 61 (2010) 703–712 3. Takehide Senuma, “Physical metallurgy of modern high strength steel sheets,” ISIJ International, Vol. 41 (2001), No. 6, pp. 520–532 4. Prakash Kumar Sahu, Sukhomay Pal, “Influence of metallic foil alloying by FSW process on mechanical properties and metallurgical characterization of AM20 Mg alloy,” Materials Science & Engineering A 684 (2017) 442–455 5. L. Liu (ed) (2010) Welding and Joining of Magnesium Alloys. Woodhead Publishing Limited, UK 6. V.K. Patel, S.D. Bhole and D.L. Chen, “Influence of ultrasonic spot welding on microstructure in a magnesium alloy,” Scripta Materialia 65 (2011) 911–914 7. V.K. Patel, S.D. Bhole, D.L. Chen, “Formation of zinc interlayer texture during dissimilar ultrasonic spot welding of magnesium and high strength low alloy steel,” Materials and Design 45 (2013) 236–240 8. W. Xua, D.L. Chena, L. Liub, H. Mori, Y. Zhoub, “Microstructure and mechanical properties of weld-bonded and resistance spot welded magnesium-to-steel dissimilar joints,” Materials Science and Engineering A 537 (2012) 11–24 9. L. Liu, L. Xiao, J.C. Feng, Y.H. Tian, S.Q. Zhou, and Y. Zhou, “The Mechanisms of Resistance Spot Welding of Magnesium to Steel,” Metallurgical and Materials Transactions A, Volume 41A, October 2010, 2651–2661 10. L. Liu, L. Xiao, D.L. Chen, J.C. Feng, S. Kim, Y. Zhou, “Microstructure and fatigue properties of Mg-to-steel dissimilar resistance spot welds,” Materials and Design 45 (2013) 336–342 11. G. Casalinoa, P. Guglielmi, V.D. Lorusso, M. Mortello, P. Peyre, D. Sorgente, “Laser offset welding of AZ31B magnesium alloy to 316 stainless steel,” Journal of Materials Processing Technology 242 (2017) 49–59 12. M. Wahba, S. Katayama, “Laser welding of AZ31B magnesium alloy to Zn-coated steel,” Materials and Design 35 (2012) 701–706 13. Xiaodong Qi, Gang Song, “Interfacial structure of the joints between magnesium alloy and mild steel with nickel as interlayer by hybrid laser-TIG welding,” Materials and Design 31 (2010) 605–609 14. Liming Liu, Xiaodong Qi, “Strengthening effect of nickel and copper interlayers on hybrid laser-TIG welded joints between magnesium alloy and mild steel,” Materials and Design 31 (2010) 3960–3963 15. Gang Song, Guangye An, Liming Liu, “Effect of gradient thermal distribution on butt joining of magnesium alloy to steel with Cu–Zn alloy interlayer by hybrid laser–tungsten inert gas welding,” Materials and Design 35 (2012) 323–329

S. Sahu et al. 16. Zhi Zeng, Xunbo Li, Yugang Miao, Gang Wu, Zijun Zhao, “Numerical and experiment analysis of residual stress on magnesium alloy and steel butt joint by hybrid laser-TIG welding,” Computational Materials Science 50 (2011) 1763–1769 17. L. M. Liu, X. Zhao, “Study on the weld joint of Mg alloy and steel by laser-GTA hybrid welding,” Materials Characterization 59 (2008) 1279–1284 18. Y. G. Miao, D. F. Han, J. Z. Yao & F. Li, “Microstructure and interface characteristics of laser penetration brazed magnesium alloy and steel,” Sci. Technol. Weld. Join., vol. 15, no. 2, pp. 97–103, 2010 19. C. W. Tan, Y. B. Chen, L. Q. Li & W. Guo, “Microstructure and properties of laser brazed magnesium to coated Steel,” Sci. Technol. Weld. Join., vol. 18, no. 6, pp. 466–472, 2013 20. Y. C. Chen and K. Nakata, “Effect of surface states of steel on microstructure and mechanical properties of lap joints of magnesium alloy and steel by friction stir welding,” Sci. Technol. Weld. Join., vol. 15, no. 4, pp. 293–298, 2010 21. Y. C. Chen and K. Nakata, “Effect of tool geometry on microstructure and mechanical properties of friction stir lap welded magnesium alloy and steel,” Mater. Des., vol. 30, no. 9, pp. 3913–3919, 2009 22. Y. C. Chen and K. Nakata, “Friction Stir Lap Welding of Magnesium Alloy and Zinc Coated Steel,” Mater. Trans., vol. 50, no. 11, pp. 2598–2603, 2009 23. S. Jana and Y. Hovanski, “Fatigue behaviour of magnesium to steel dissimilar friction stir lap joints,” Sci. Technol. Weld. Join., vol. 17, no. 2, pp. 141–145, 2012 24. S. Jana, Y. Hovanski, and G. J. Grant, “Friction stir lap welding of magnesium alloy to steel: A preliminary investigation,” Metall. Mater. Trans. A Phys. Metall. Mater. Sci., vol. 41, no. 12, pp. 3173–3182, 2010 25. H. Kasai, Y. Morisada, H. Fujii, “Dissimilar FSW of immiscible materials: Steel/magnesium,” Materials Science & Engineering A 624 (2015) 250–255 26. C. Schneider, T. Weinberger, J. Inoue, T. Koseki, and N. Enzinger, “Characterisation of interface of steel/magnesium FSW,” Sci. Technol. Weld. Join., vol. 16, no. 1, pp. 100–107, 2011 27. N. Afrin, D. L. Chen, X. Cao, and M. Jahazi, “Microstructure and tensile properties of friction stir welded AZ31B magnesium alloy,” Mater. Sci. Eng. A, vol. 472, no. 1–2, pp. 179–186, 2008 28. Z. Y. Ma, “Friction stir processing technology: A Review,” Metallurgical and Materials Transactions A, Volume 39A, March 2008, 642–658 29. A. Gerlich, P. Su, M. Yamamoto & T. H. North, “Material flow and intermixing during dissimilar friction stir welding,” Sci. Technol. Weld. Join., vol. 13, no. 3, pp. 254–264, 2008 30. P. Su, A. Gerlich, T.H. North, and G.J. Bendzsak, “Intermixing in dissimilar friction stir spot welds,” Metallurgical and Materials Transactions A, Volume 38A, March 2007, 584–595 31. Padmanaban, G. & Balasubramanian, V, “An experimental investigation on friction stir welding of AZ31B magnesium alloy,” Int J Adv Manuf Technol (2010) 49: 111. https://doi.org/10.1007/ s00170-009-2368-1 32. Y. U. Sirong, Chen Xianjun, Huang Zhiqiu, Liu Yaohui, “Microstructure and mechanical properties of friction stir welding of AZ31B magnesium alloy added with cerium,” Journal of Rare Earths, Vol. 28, No. 2, Apr. 2010, p. 316 33. B. Ratna Sunil, G. Pradeep Kumar Reddy, A.S.N. Mounika, P. Navya Sree, P. Rama Pinneswari, I. Ambica, R. Ajay Babu, P. Amarnadh, “Joining of AZ31 and AZ91 Mg alloy by friction stir welding,” Journal of Magnesium and Alloys 3 (2015) 330–334 34. Stephan W. Kallee Wayne M. Thomas E. Dave Nicholas, “Friction stir welding of lightweight materials,” TWI Ltd, Cambridge, United Kingdom, pp 175–190

A Review and Case Study on Mechanical Properties … 35. P. Cavaliere, P.P. De, “Marco Superplastic behaviour of friction stir processed AZ91 magnesium alloy produced by high pressure die cast,” Journal of Materials Processing Technology 184 (2007) 77–83 36. Yajie Li, Fengming Qin, Cuirong Liu and Zhisheng Wu, “A review: Effect of friction stir welding on microstructure and mechanical properties of magnesium alloys,” Metals 2017, 7(12), 524; https://doi.org/10.3390/met7120524 37. Y. Wei, J. Li, J. Xiong, F. Huang, and F. Zhang, “Microstructures and mechanical properties of magnesium alloy and stainless steel weld-joint made by friction stir lap welding,” Mater. Des., vol. 33, no. 1, pp. 111–114, 2012

109 38. ASTM E3-11(2017) Standard Guide for Preparation of Metallographic Specimens, ASTM International, West Conshohocken, PA, 2017, https://doi.org/10.1520/E0003-11R17 39. ASTM E407-07(2015)e1 Standard Practice for Microetching Metals and Alloys, ASTM International, West Conshohocken, PA, 2015, https://doi.org/10.1520/E0407-07R15E01 40. ASTM E384-17 Standard Test Method for Microindentation Hardness of Materials, ASTM International, West Conshohocken, PA, 2017, https://doi.org/10.1520/E0384-17

Effects of Sn on Microstructures and Mechanical Properties of As-Extruded Mg−6Al−1Ca−0.5Mn Magnesium Alloy Huajie Wu, Ruizhi Wu, Daqing Fang, Yuesheng Chai, and Chao Liang

Abstract

Effects of Sn on microstructures and mechanical properties of Mg–6Al–Ca–0.5Mn magnesium alloy were investigated. With Sn addition, the grain size of Mg–6Al–Ca– 0.5Mn–xSn alloy decreases and the volume fraction of Mg2Sn phase increases. The minimum average grain size of Mg–6Al–Ca–0.5Mn–xSn alloy is 15 lm. The distribution of the phase is more dispersive in the Mg matrix. Sn content plays a key role for the inhibition of b-Mg17Al12 phase and the promotion of Mg2Sn phase. However, excessive Sn addition results in the decline of strength and elongation. Tensile results show that the Mg–6Al–Ca–0.5Mn–3Sn alloy exhibits the best mechanical properties, and the ultimate tensile strength, yield strength and elongation of the alloy are 335.9 MPa, 261.1 MPa and 10.9%, respectively. The improved tensile properties are mainly related to grain refinement, solid solution strengthening of Sn and precipitation strengthening of Mg2Sn phase. Fractographic analysis demonstrates that quasi-cleavage fracture is the dominant mechanism of these alloys.




Keywords




Sn Mg–6Al–1Ca–0.5Mn magnesium alloy Microstructures Mechanical property




H. Wu (&)
R. Wu Key Laboratory of Superlight Materials & Surface Technology, Ministry of Education, Harbin Engineering University, Harbin 150001, People’s Republic of China e-mail: [email protected] H. Wu
D. Fang
Y. Chai
C. Liang College of Materials Science and Engineering, Taiyuan University of Science and Technology, Taiyuan 030024, China

Introduction Now magnesium alloys have received increasing attention due to its great promise in reducing vehicle weight. Among these alloys, those based on Mg–Al are the most promising for more development, especially in automobile industry because of an excellent combination of improved die castability, acceptable room temperature mechanical properties and reasonable cost. However, the limited strength and the poor creep-resistance of Mg–Al alloys have significantly astricted their more wide applications [1–6]. A method to improve the mechanical properties of Mg–Al alloys is alloying through grain refining, solid solution strengthening and secondary phase strengthening [7]. The suitable addition of minor element [8] for mass production is a simple and economical method to improve performance of the alloy. An appropriate amount of Mn added in the magnesium alloy has many advantages for its properties. First, Mn can eliminate Fe by forming Fe–Mn compounds, which has adverse impacts on the corrosion resistance of magnesium alloys [9–13]. Second, Mn can refine the microstructure of magnesium alloy [14, 15]. In addition, a-Mn interacting with dislocation can improve the creep resistance of magnesium alloys [14, 16]. It is well known that Ca also can refine the magnesium alloys [17]. At the same time, Jun et al. [18, 19] and Zhang et al. [20] found that Ca addition in Mg alloys can not only refine the matrix a-Mg grains but also change the form and distribution of the Mg–RE phases. Because of its cheap price and the high solid solubility in a-Mg matrix, Sn is commonly used as the additional element [21–23]. Sn can not only refine matrix grains but also improve the corrosion resistance and form a stable Mg2Sn precipitates during solidification [24–26]. As Mg2Sn phase has a markedly higher melting point (770 °C) than the Mg17Al12 phase (463 °C), the alloys with Sn addition have excellent mechanical properties at high temperature [27, 28]. The Mg2Sn phase with high melting point can be formed by adding Sn element to Mg–6Al–1Ca–0.5Mn alloy, which

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_18

111

112

H. Wu et al.

is very advantageous to the mechanical properties of high temperature wrought magnesium alloy. But the research of the effect of Sn element on magnesium alloy is very rare. Therefore, in this study, the effects of Sn addition on the microstructure and mechanical properties of the as-extruded Mg–6Al–Ca–0.5Mn alloy were investigated.

spectrometer (EDS). Phase components were identified by X-ray diffractometer using Cu Ka.

Results and Discussion Microstructure

Experimental The Mg–6Al–Ca–0.5Mn experimental alloys were prepared by adding pure Mg, pure Al, Mg–Ca, MnCl2 and pure Sn. The experimental alloys were melted in a crucible resistance furnace and protected by a flux addition. After being held at 740 °C for 20 min, the melts of the experimental alloys were respectively homogenized by mechanical stirring and then poured into a permanent mould with an ingot diameter of 100 mm which was coated and preheated to 200 °C in order to obtain a casting. The ingots were homogenized at 420 °C for 5 h and then under a controlled constant force by XJí500 horizontal extrusion machine. The homogenized ingots were hot extruded to bars of 10 mm in diameter at 360 °C with the extrusion ratio of 100. After extrusion, the alloys were cooled in the open air. Table 1 shows nominal compositions of the alloys (wt%). In order to study effects of Sn on microstructure and mechanical properties of Mg–6Al–1Ca–0.5Mn alloy extruded bars, tensile test, metallographic observation, SEM analysis, and XRD analysis were conducted for different specimens. The specimen was cut off from each extruded bar center section, and the TD-ND surface was selected as the study surface. Then they were polished with sandpaper and polished in a polishing machine and then corroded. Etchant: picric acid 1.5 g + glacial acetic acid 5 ml + distilled water 10 ml + ethanol 25 ml. Tensile tests were performed using tensile specimens with gauge length of 20 mm and gauge diameter of 4 mm. The tensile directions were parallel to the extrusion direction (ED). The tensile tests were performed on a CNMT5105 electronic universal testing machine in air at a speed of 2 mm/min at room temperature. Tensile properties under each condition were obtained as the average values of three tests. The microstructures of specimens were observed with OLYMPUS-BX60M laser metallographic microscope and HITACHI S-4800 scanning electron microscope equipped with an OXFORD-250 energy-dispersive X-ray

Table 1 Nominal compositions of Mg–6Al–Ca–0.5Mn–xSn alloys (wt%)

Alloy

Figure 1 shows microstructures of Mg–6Al–Ca–0.5Mn–xSn (x = 1, 3, 6) alloy. With the addition of Sn, the microstructure is refined further, and T2 Mg–6Al–Ca–0.5Mn– 2Sn alloy exhibits the most refined microstructure. There are two reasons to explain this result. On the one hand, if alloys extrusion temperature and deformation in the process of extrusion remain unchanged, the thinner the as-cast grain size is, the smaller the as-extruded grain size is after recrystallization. (The finer the original raw grain of the alloy is, the more the grain boundaries which are favorable areas of recrystallization nucleation are, so its nucleation rate is higher.) On the other hand, the higher Sn content is, the more Sn atoms in grain boundary are, which greatly reduce the grain boundary interfacial energy, reducing the driving force of the interface moving and making the grain boundary not be easy to move, so Sn atoms hinder the grains growth, getting finer grains. As a result, the higher the Sn content is, the smaller the alloy grain is, as shown in Fig. 1b. The average grain size of alloy with 1 wt%, 3wt%, 6wt% Sn is 19 lm, 15 lm, 20 lm, respectively. With higher Sn content (6%), grains of alloys began to coarsen and the second phase reunited, as shown in Fig. 1c. The XRD analysis results of the as-extruded Mg–6Al– Ca–0.5Mn alloys with different Sn content are shown in Fig. 2. It can be seen that the alloy consists of a-Mg, b-Mg17Al12 and Mg2Sn phase. It is also evident that the diffraction peak intensity of the Mg2Sn phase increases and that of the b-Mg17Al12 phase decreases with the increasing Sn content, which indicates that the addition of Sn promoted the precipitation of Mg2Sn phase. Figure 3 shows EDS analysis of Mg–6Al–Ca–0.5Mn–3Sn alloy. Based on the analysis of XRD results (Fig. 2) and energy spectrum analysis (Fig. 3), it can be concluded there is the presence of Mg2Sn phase. In addition, judging from electronegativity difference between elements, the larger the electronegativity difference is, the greater the bonding force between the elements is, so

Al

Mn

Ca

Sn

Mg

Mg–6Al–Ca–0.5Mn–1Sn

6.0

0.5

1.0

1.0

Bal.

Mg–6Al–Ca–0.5Mn–3Sn

6.0

0.5

1.0

3.0

Bal.

Mg–6Al–Ca–0.5Mn–6Sn

6.0

0.5

1.0

6.0

Bal.

Effects of Sn on Microstructures and Mechanical Properties …

113

Fig. 1 Optical Microstructures of Mg–6Al–Ca–0.5Mn magnesium alloy with different Sn contents: a 1% Sn b 3% Sn c 6% Sn

Fig. 2 XRD patterns of Mg–6Al–Ca–0.5Mn magnesium alloys with different Sn addition

it is easier to form metal compounds [29]. Because the electronegativity difference between Mg and Sn (0.65) is larger than that between Mg and Al (0.30) or between Al and Sn (0.35) (Table 2), Sn can preferentially react with Mg to form a particle phase. Based on energy spectrum analysis (Fig. 3), it also can confirm that the particle phase is Mg2Sn phase. Figure 4 shows the SEM images of the as-extruded Mg– 6Al–Ca–0.5Mn alloy with different Sn contents. As seen in Fig. 4, with the increase of Sn content, grains are refined, the volume fraction of Mg2Sn particles increases, and its distribution is more dispersive, as shown in Fig. 4a, b. However with further increasing Sn content, grains coarsened and Mg2Sn particles reunited, as shown in Fig. 4c. For high melting point (770 °C) of Mg2Sn phase, it first forms crystal nucleus in the solidification process. As the solidification continues, Mg2Sn can be used as nucleation core of a-Mg matrix phase, making the matrix grain refinement. In

114

H. Wu et al.

Fig. 3 EDS analysis of Mg–6Al–Ca–0.5Mn magnesium alloy with 3% Sn addition Table 2 Electronegativity of elements in alloys

Element

Mg

Al

Sn

Electronegativity

1.31

1.61

1.96

Fig. 4 SEM images of Mg–6Al–Ca–0.5Mn alloy with different Sn addition: a 1% Sn b 3% Sn and c 6% Sn

Effects of Sn on Microstructures and Mechanical Properties …

addition, due to the high melting point of Mg2Sn, so it has been in the form of a solid, which will be enriched in the forefront of a-Mg phase, resulting in composition supercooling and hindering a-Mg growth to make the grain refinement.

Mechanical Properties The mechanical properties of as-extruded Mg–6Al–Ca– 0.5Mn–xSn alloys at room temperature are shown in Fig. 5. It can be seen that ultimate tensile strength (UTS), yield strength (YS) and elongation (d) of alloy increase firstly and then decrease with the increase of Sn content. But an excess of Sn (>3%) made the strength and elongation of the alloy decrease. With the addition of 3% Sn, Mg–6Al–Ca– 0.5Mn–3Sn alloy exhibits the best mechanical properties, and the ultimate tensile strength, yield strength and elongation of the alloy are 343 MPa, 142 MPa and 21.6%, respectively. As seen in Fig. 5, the addition of Sn can improve the tensile strength of alloys. The increase of strength of the alloy depends not only on grain size, but also on many other factors such as shape, size and distribution of the second phase. According to Hall–Petch relationship [30], which reflects the general rule of grain size influence on yield strength, so reducing the grain size will increase the yield strength of material. As the average grain size of alloy with 1 wt%, 3wt%, 6wt% Sn are 19 lm, 15 lm, 20 lm, respectively, Mg–6Al–Ca–0.5Mn–3Sn alloy exhibits the maximum yield strength. Meanwhile, the decrease of the grain size means the increase of the grain boundary. The grain boundary can block the motion of dislocations, so elongation of alloys can also be improved. According to Mg–Sn binary phase diagram [31], solid solubility of Sn in Mg is very small ( LA156 > LA36. Keywords

Mg–Li alloy properties




Microstructure




Corrosion-resistance

Introduction Limitations of corrosion resistance are the main major obstacles to the industrial application of magnesium (Mg) alloys to date. Because of the low density of lithium (0.58 g/cm3) and its ability to form the body center cubic (bcc)-structured Mg (Li) solid solution phases [1], alloying with Li yields the lightest Mg–Li alloy [2, 3] with the significantly improved malleability, which is essential for structural applications, such as automotive, aerospace, and electronic industries [4–7]. Based on the Mg–Li binary phase diagram [8], alloying Mg with Li can also change the crystal structure of Mg. The Mg–Li alloy retains the hcp structure a phase when the Li content is lower than 5 weight percent (wt %). The hcp structure a phase changed to bcc structure b Y. Li
T. Li
Q. Wang
Y. Zou (&) School of Mechanical Engineering, Zhengzhou University, Zhengzhou, 450001, People’s Republic of China e-mail: [email protected]

phase, when the Li content is higher than 11 wt%. With the Li content between 5 and 11 wt%, Mg–Li alloy has a duplex structure mixed with a phase and bcc structure b phase. However, alloying Li in Mg alloys decreases the corrosion resistances, compared with pure Mg, because the standard electrode potential of Li (−3.045 V vs. SHE) is lower than that of Mg (−2.37 V vs. SHE), making the alloy extremely susceptible to galvanic corrosion. The effect of Al addition on corrosion resistance of Mg alloys is very complicated [9–13]. Song and Atrens [13, 14] et al. proposed that the Mg17Al12 phase in an AZ alloy has two different effects on the corrosion behavior: the Mg17Al12 phase can act either as a galvanic cathode to accelerate corrosion or as a corrosion barrier to hinder corrosion, depending on the amount and distribution of phases. According to Sahoo and Atkinson [15], the ternary bcc-structured b phase Mg–12.3 Li–1.5 Al alloy (in wt%) seems to have better corrosion performance than the binary Mg–12.2 Li alloy (in wt%) in an electrolyte solution of magnesium perchlorate. Lin et al. [16] also investigated the corrosion performance of the bcc-structured b phase Mg–Li–Al–Zn alloy (LAZ) with different Al compositions. The results showed that the LAZ alloy with a higher Al content exhibits a better corrosion resistance, owing to the distribution of numerous AlLi particles in the matrix of the alloy. However, how the Li contents affect the corrosion performance of the Mg–Li–Al alloy has not been fully understood. Here we systemically investigate the effects of Li content on microstructures and corrosion resistance of three Mg–Li– Al alloys with the a phase, (a + b) phases, and b phase, by changing the Li concentration.

Experimental Procedure Material Preparation Three ingots with the nominal chemical compositions of Mg–3Li–6Al (wt%) (LA36), Mg–9Li–6Al (LA96), and Mg–

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_20

127

128

15Li–6Al (LA156) alloys were prepared under argon atmosphere by induction-melting high-purity elemental metals (99.9 Mg, 99.9 Li, and 99.9 Al, wt%). The induction-melted alloy was poured into plates in a copper mold. The actual measured chemical composition of the as-cast alloys is Mg–3.8Li–6.1Al (wt%) (LA36), Mg–9.5Li– 6.1Al (LA96), and Mg–15.9Li–6.2Al (LA156), respectively, which was determined by the Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-AES, Thermo Elemental, IRIS Intrepid II XSP).

Microstructural Characterization The microstructure and element composition analysis of the alloys was carried out with optical microscope (OM), scanning electron microscope (SEM, JSM-6480A, Japan Electronics) and energy dispersive X-ray spectroscopy (EDS) (INC250, Japan Electronics), and transmission electron microscopy (TEM). To reveal the microstructure, the samples were metallographically polished, followed by etching in 4 volume percent (vol. %) nital solution. The TEM measurements were carried out on JEM-2100 high-resolution transmission electron microscope operated at 200 kV. The plane-view thin foils for TEM were prepared by mechanical grinding, followed by thinning via Gatan PIPS. Phase identification was investigated by X-ray diffraction on a Rigaku D/max-TTR-III diffractometer using Cu Ka.

Corrosion Testing The corrosion performance was studied by electrochemical impedance spectroscopy (EIS), potentiodynamic polarization curves, and hydrogen evolution tests in a 3.5 wt% NaCl solution. The electrochemical experiments were carried out on a workstation (CHI660D, China) based on the conventional three-electrode system. Prior to the measurements, the working electrode (tested sample) was immersed in the test solution for 15 min at the open-circuit potential (OCP), in order to stabilize the free corrosion potential. The EIS data were recorded at the free corrosion potential in the frequency range of 100 kHz to 10 MHz with a 5 mV amplitude perturbation (peak-to-zero), and, then, the Nyquist plots of the impedance spectrums were obtained. An analysis of the EIS data was performed by fitting the experimental data to an equivalent circuit using the ZSimpWin 3.10 software. After the EIS measurement, each sample was returned to OCP for 5 min, and, then, the potentiodynamic polarization curves were obtained at a scanning rate of 2 mV/s.

Y. Li et al.

Results and Discussion Microstructures The microstructures of the as-cast Mg–3Li–6Al (LA36), Mg–9Li–6Al (LA96), and Mg–15Li–6Al (LA156) are shown in Fig. 1 [17]. The LA36 alloy consists of the a phase matrix (hcp) and precipitates (Fig. 1a). Most of the precipitates are distributed along the grain boundaries, and a few particles uniformly disperse inside the grains. The average grain size is *40 lm. With the Li concentration increasing to 9 wt%, the LA96 alloy represents a typical duplex-phase (hcp + bcc) structure; the phase in the bright color is the hcp a phase and that in gray is the bcc b phase. The morphology of the a phase is an irregular block and long strip, as shown in Fig. 1c. Some small spherical particles are evenly distributed in the b phase as well as along the grain boundaries (Fig. 1d). For the LA156 alloy, obvious equiaxed grains with an average grain size of *165 lm are shown in Fig. 1e. The precipitate particles are gathered along the grain boundaries, and comparatively finer spherical particles as well as rod-like precipitates are evenly distributed in the grains (Fig. 1f). The XRD patterns of the as-cast LA36, LA96, and LA156 alloys are illustrated in Fig. 2 [17]. The XRD results show that the LA36 alloy consists of the hcp a phase and AlLi phase, the LA96 represents a typical duplex phase of the hcp a phase and bcc b phase as well as AlLi and MgLiAl2 phases, and the LA156 alloy is composed of the bcc b phase, AlLi, and MgLiAl2 phases. Clearly, with increasing the Li content, the crystal structure is transformed from the hcp to bcc structures, which is consistent with the Mg–Li phase diagram [8]. Moreover, with increasing the lithium content, the peak position of the all peaks in the XRD spectra is shifted to the high degree, suggesting a decrease of the lattice constants through the substitution of Mg atoms by Li atoms. As Li is added to Mg, the c value of the hcp crystal falls faster than does the a value, which leads to a decrease in the c/a ratio [17–19]. With increasing the Li content, the lattice constants further decreases. The SEM morphologies and the corresponding EDS results of the LA36, LA96, and LA156 alloys are shown in Fig. 3 and Table 1 [17]. The SEM images in higher magnification exhibit that the second phases are gathered at the grain boundaries of the LA36 alloy with an irregular herringbone structure (Fig. 3a). Some small spherical particles with an average size of approximately 1 lm are evenly distributed in the b phase as well as along the boundaries of the LA96 alloy (Fig. 3b). For the LA156 alloys, the precipitate particles are gathered along the grain boundaries; comparatively finer spherical particles with approximately

Effects of Li on Microstructures and Corrosion Behaviors …

129

Fig. 1 Optical micrographs of the LA36 (a, b), LA96 (c, d), and LA156 (e, f) alloys [17]

Fig. 2 XRD patterns of the LA36, LA96, and LA156 alloys [17]

below 1 lm as well as short rod-like precipitates with a width of *0.5 lm and a length of 2–5 lm are evenly distributed in the grains (Fig. 3c). The corresponding EDS results show that the compounds of the three alloys are composed of Mg and Al elements, as the Li element is too light to be detected by EDS. Combining results of the previous studies and XRD analysis, the second phase in the LA36 alloy (“A” in Fig. 3a) is determined as AlLi [20]. Based on the similar content of Mg and Al elements, it can be inferred that the relatively large particles in LA96 (“B” in Fig. 3b) and LA156 (“D” in Fig. 3c) are AlLi, and the small size particles in LA96 (“C” in Fig. 3b) and short rod-like precipitates in LA156 (“E” in Fig. 3c) are MgLiAl2. The TEM morphologies and corresponding selected-area electron diffraction (SAED) patterns of the LA36, LA96, and LA156 alloys are shown in Fig. 4 [17], in order to further confirm the second precipitated phases. The precipitates

130

Y. Li et al.

Fig. 3 SEM morphologies of the LA36 (a), LA96 (b), and LA156 (c) alloys. The “A, B, C, D, and E” marked in figures are the locations where the EDS measurements were performed [17]

Table 1 EDS results of the LA36, LA96, and LA156 alloys. The “A, B, C, D, and E” marked in Fig. 3 are the locations where the EDS measurements were performed [17]

Number

Element wt%

at.%

Mg

Al

Mg

Al

A

86.6

13.4

87.8

12.2

B

87.2

12.8

88.3

11.7

C

97.6

2.4

97.8

2.2

D

86.4

13.6

87.6

12.4

E

94.3

5.7

94.9

5.1

Note The Li element was not involved in the table because the Li element is too light to be detected by EDS

show a capsule-like morphology with a length of 200– 500 nm and a diameter of 40–100 nm in LA36 alloy (Fig. 4a). The corresponding SAED patterns consist of spots from not only the a phase but also the AlLi phase (Fig. 4a). For the LA96 alloy, the precipitates show a shot-bar morphology with a length of 500–1000 nm and a diameter of 35–90 nm. The corresponding SAED patterns verify the existence of both the AlLi phase and MgLiAl2 phase (Fig. 4b). For the LA156 alloy, the second phase exhibits a cystiform morphology instead of capsule or bar morphology. The corresponding SAED patterns consist of the MgLiAl2 phase and AlLi phases (Fig. 4c). There are plenty of fine particles in the vesica as revealed by dark-field TEM image in Fig. 4d. The results further confirm that the XRD and SEM analyses of the precipitate phases in the LA36 alloy are the AlLi phase, and the precipitate phases in the LA96 and LA156 alloys are the AlLi phase and MgLiAl2 phase.

Corrosion Properties The corrosion performance of the cast LA36, LA96, and LA156 samples, which are studied by EIS and potentiodynamic polarization curves, are shown in Fig. 5. The Nyquist plots of the samples, which are obtained from the EIS, are exhibited in Fig. 5a. All the samples exhibit a high-frequency capacitive loop and a low-frequency inductance loop. The capacitive loop in the high-frequency region represents the double-layer capacitance and charge-transfer resistance at the electrode/electrolyte interface. The inductance loop in the low-frequency domain is mainly attributed to the corrosion nucleation at the initiation stage of the localized corrosion [21]. The EIS data were fitted by the ZSimpWin 3.10 software using the given equivalent circuit RS(QRct(LR1)) (Inset in Fig. 5a), and the fitting parameters are summarized in Table 2. In the equivalent circuit, RS and

Effects of Li on Microstructures and Corrosion Behaviors …

131

Fig. 4 TEM images of the LA36 (a), LA96 (b), and LA156 (c, d) alloys and the corresponding SAED insets, respectively [17]

Rct represent the solution resistance and charge-transfer resistance. The resistance, R1, and inductance, L, are used to specify the inductive behavior. The constant phase element (CPE, designated as Q) is used instead of the ideal double-layer capacitance (Cdl) to account for the non-ideal behavior of the double-layer due to the surface inhomogeneity, roughness, and adsorption effects [22]. The polarization resistance, Rp, is calculated as Rp = R1 + Rct. With the Li content increasing from 3 to 9 and 15 wt%, the polarization resistance, Rp, is increased from 127.7 to *277 X cm2, indicating the enhanced corrosion-resistance properties with the increasing Li content in the ternary Mg–Li–Al alloy. The calculated electrochemical corrosion data determined from the potentiodynamic polarization curves are summarized in Table 3. The data include the corrosion current density (icorr), corrosion potential (Ecorr), and the cathodic Tafel slope (bc), which can be obtained via linear region fitting. The detailed methods can be found in Reference [23]. With the Li content increasing from 3 to 9 and to 15 wt%, the corrosion potential (Ecorr) decreases from −1.41 to −1.52 and to −1.54 V, while the corrosion current density (icorr) decreases from 6.5  10−4 to 1.9  10−4 A cm−2 and then increases to 3.1  10−4 A cm−2 (Fig. 5b), implying the duplex-phase LA96 alloy has a better corrosion resistance than the other two single-phase alloys. To determine a longer-term corrosion-resistance performance, hydrogen evolution tests were also conducted. The hydrogen evolution of the LA36, LA96, and LA156 alloys

during immersion in a 3.5 wt% NaCl solution for 24 h is demonstrated in Fig. 6. The specimens exhibit an increase in the hydrogen evolution rate with increasing the immersion time. Obviously, the LA36 alloy exhibits a higher hydrogen evolution rate after 2 h immersion, and the rate becomes faster with increasing the immersion time. As a consequence, the hydrogen evolution volume of the three samples can be ranked as LA36 > LA156 > LA96, confirming that the duplex LA96 alloy has better corrosion-resistance performance than the other two single-phase alloys. Corrosion performance is also highly dependent on the microstructures. According to the previous studies on the Mg–Al–Zn (AZ) series alloys [31, 35, 36], either the Mg17Al12 acts as a galvanic cathode to accelerate corrosion or a corrosion barrier to hinder corrosion, depending on the amount and distribution of the phase. The galvanic corrosion acceleration is dependent on the anode (matrix)/cathode (Mg17Al12) area ratio, whereas the Mg17Al12 phase can act as a barrier against corrosion propagation, if it is finely and evenly distributed. In the present work, the microstructures of both the matrix and second phase vary with increasing the Li content, which makes the situation more complicated than the AZ series alloys that mainly consist of a matrix of a grain with the Mg17Al12 phase along the grain boundaries. The main second phase in the LA36, LA96, and LA156 alloys is the AlLi phase, which acts as an anode while the a or b phase acts as a cathode in the Mg–Li system, since the electrochemical potential of an AlLi phase [−1.960 V vs.

132

Y. Li et al.

Fig. 5 Nyquist plots (a), potentiodynamic polarization curves (b) of the LA36, LA96, and LA156 alloys

Table 2 EIS parameters for the LA36, LA96, and LA156 samples in a 3.5 wt% NaCl solution

Parameter

LA36

2

Rs (X cm ) −2

CPE (S cm

28.7 s)

2

LA156

28.4 −5

n

n

Table 3 Electrochemical parameters of the LA36, LA96, and LA156 samples obtained from the potentiodynamic polarization curves

LA96

25.7 −5

2.8  10

8.5  10

3.6  10−5

0.9

0.8

0.7

Rct (X cm )

94.8

211.8

205.2

L (H cm2)

40.8

646.7

537.3

R1 (X cm2)

32.9

65.8

71.0

Rp (X cm2)

127.7

277.6

276.2

Sample

Ecorr (V)

bc (V decade−1)

icorr (A cm−2)

LA36

−1.41

−0.21

6.5  10−4

LA96

−1.52

−0.24

1.9  10−4

LA156

−1.54

−0.23

3.1  10−4

Fig. 6 Hydrogen evolution of the LA36, LA96, and LA156 alloys during immersion in a 3.5 wt% NaCl solution for 24 h

standard hydrogen electrode (SHE)] is much more negative than that of a bcc-structured Mg matrix (−1.258 V vs. SHE) [24].

The corrosion mechanisms of the LA36, LA96, and LA156 alloys are summarized in Fig. 7, according to the corrosion characteristics of the various microstructures. The large AlLi precipitates are mainly gathered at the a phase grain boundaries like a circuit in the LA36 alloy. It is easy to form the galvanic corrosion of the anode (AlLi) and cathode (a phase) and to promote the corrosion. The small AlLi and MgLiAl2 particles are finely and evenly distributed in the b phase in both LA96 and LA156 alloys, building up a certain degree of continuity in the barrier; that is, the small particles can mainly act as a corrosion barrier to hinder corrosion and improve the corrosion resistance, as compared to the large AlLi precipitates in the LA36 alloy. In the duplex Mg–Li alloy, the b phase is more active than the a phase, and the localized corrosion propagates toward the direction of the b phase [21]. In the present work, the LA96 alloy presents a typical duplex structure, no Al-contained precipitates form in the a phase while large amounts of Al-contained precipitates (AlLi and MgLiAl2) form in the b phase, resulting aluminum-poor areas in the b phase, as compared to the a phase [12], which makes the b phase further less corrosion resistant than the a phase. Moreover, the fine precipitates

Effects of Li on Microstructures and Corrosion Behaviors …

133

Fig. 7 Corrosion mechanisms schematic of the LA36, LA96, and LA156 alloys

form an almost continuous barrier along the a/b boundaries, which can, to some extent, stop the development of corrosion [13]. Consequently, the LA96 alloy shows a better corrosion resistance than the LA156 alloy. As a result, the corrosion resistance of the three samples can be ranked as LA96 > LA156 > LA36.

Conclusions The microstructure and corrosion behaviors of three Mg–Li– Al alloys have been systematically investigated by changing the Li concentrations. The corrosion behaviors are related to compositions, phase contents, and microstructures. The corrosion performance of the cast LA36, LA96, and LA156 samples can be ranked as LA96 > LA156 > LA36. The different corrosion performance of the alloys can be related to the various microstructures: in the LA36 alloy, the large AlLi phase mainly appears at the a phase grain boundaries like a circuit and promotes the galvanic corrosion; in both LA96 and LA156 alloys, the AlLi particles are small and evenly distributed in the b phase as well as along the grain boundaries, which can act as a corrosion barrier to hinder corrosion. The LA96 alloy presents a typical duplex structure, no Al-contained precipitates form in the a phase while large amounts of Al-contained precipitates (AlLi and MgLiAl2) form in the b phase, resulting aluminum-poor areas in the b phase, as compared to the a phase, which makes the b phase further less corrosion resistant than the a phase. Moreover, the fine precipitates form an almost continuous barrier along the a/b boundaries, which can, to some extent, stop the development of corrosion. Consequently, the LA96 alloy shows a better corrosion resistance than the LA156 alloy. Acknowledgements This work was supported by the National Science Foundation of China (51705470, 51801185), Key Research Project of the Higher Education Institutions of Henan Province, Henan Provincial Department of Education, China (18A460032, 19A460007), and Special Research and Promotion Project of Henan Province, China (182102210009).

References 1. Z.K. Qu, L.B. Wu, R.Z. Wu, J.H. Zhang, M.L. Zhang, B. Liu, Microstructures and tensile properties of hot extruded Mg–5Li– 3Al–2Zn–xRE(Rare Earths) alloys, Mater. Des. 54 (2014) 792–795. 2. D.K. Xu, T.T. Zu, M. Yin, Y.B. Xu, E.H. Han, Mechanical properties of the icosahedral phase reinforced duplex Mg–Li alloy both at room and elevated temperatures, J. Alloy. Compd. 582 (2014) 161–166. 3. R.Z. Wu, Y.D. Yan, G.X. Wang, L.E. Murr, W. Han, Z.W. Zhang, M.L. Zhang, Recent progress in magnesium-lithium alloys, Int. Mater. Rev. 60 (2015) 65–100. 4. I. Shin, E.A. Carter, First-principles simulations of plasticity in body-centered-cubic magnesium-lithium alloys, Acta Mater. 64 (2014) 198–207. 5. W.A. Counts, M. Friak, D. Raabe, J. Neugebauer, Using ab initio calculations in designing bcc Mg–Li alloys for ultra-lightweight applications, Acta Mater. 57 (2009) 69–76. 6. V. Kumar, Govind, R. Shekhar, R. Balasubramaniam, K. Balani, Microstructure evolution and texture development in thermomechanically processed Mg–Li–Al based alloys, Mater. Sci. Eng. A 547 (2012) 38–50. 7. T.L. Zhu, J.F. Sun, C.L. Cui, R.Z. Wu, S. Betsofen, Z. Leng, J.H. Zhang, M.L. Zhang, Influence of Y and Nd on microstructure, texture and anisotropy of Mg–5Li–1A1 alloy, Mater. Sci. Eng. A 600 (2014) 1–7. 8. T.B. Massalski (Ed.), Binary Alloy Phase Diagrams, ASM, Metals Park, OH, 1986. 9. M.C. Zhao, M. Liu, G. Song, A. Atrens, Influence of the b-phase morphology on the corrosion of the Mg alloy AZ91, Corros. Sci. 50 (2008) 1939–1953. 10. R. Ambat, N.N. Aung, W. Zhou, Evaluation of microstructural effects on corrosion behaviour of AZ91D magnesium alloy, Corros. Sci. 42 (2000) 1433–1455. 11. T. Zhang, Y. Li, F. Wang, Roles of b phase in the corrosion process of AZ91D magnesium alloy, Corros. Sci. 48 (2006) 1249–1264. 12. G.L. Song, A.L. Bowles, D.H. StJohn, Corrosion resistance of aged die cast magnesium alloy AZ91D, Mater. Sci. Eng. A 366 (2004) 74–86. 13. G.L. Song, A. Atrens, X. Wu, B. Zhang, Corrosion behaviour of AZ21, AZ501 and AZ91 in sodium chloride, Corros. Sci. 40 (1998) 1769–1791. 14. G.L. Song, A. Atrens, M. Dargusch, Influence of microstructure on the corrosion of diecast AZ91D, Corros. Sci. 41 (1998) 249–273. 15. M. Sahoo, J. Atkinson, Magnesium-lithium-alloys—constitution and fabrication for use in batteries, J. Mater. Sci. 17 (1982) 3564–3574.

134 16. M.C. Lin, C.Y. Tsai, J.Y. Uan, Electrochemical behaviour and corrosion performance of Mg–Li–Al-Zn anodes with high Al composition, Corros. Sci. 51 (2009) 2463–2472. 17. Yun Zou, Lehao Zhang, Yang Li, Hongtao Wang, Jiabin Liu, Peter K. Liaw, Hongbin Bei, Zhongwu Zhang, Improvement of mechanical behaviors of a superlight Mg–Li base alloy by duplex phases and fine precipitates, Journal of Alloys and Compounds 735 (2018) 2625–2633. 18. R.S. Busk, Lattice parameters of magnesium alloys, Trans. Aime 188 (1950) 1460–1464. 19. L.W.F. Mackenzie, M. Pekguleryuz, The influences of alloying additions and processing parameters on the rolling microstructures and textures of magnesium alloys, Mater. Sci. Eng. A 480 (2008) 189–197. 20. B. Jiang, C.H. Zhang, T. Wang, Z.K. Qu, R.Z. Wu, M.L. Zhang, Creep behaviors of Mg–5Li–3Al–(0, 1) Ca alloys, Mater. Des. 34 (2012) 863–866.

Y. Li et al. 21. Y. Song, D. Shan, R. Chen, E.H. Han, Corrosion characterization of Mg–8Li alloy in NaCl solution, Corros. Sci. 51 (2009) 1087– 1094. 22. H.P. Liu, N. Li, S.F. Bi, D.Y. Li, Z.L. Zou, Effect of organic additives on the corrosion resistance properties of electroless nickel deposits, Thin Solid Films 516 (2008) 1883–1889. 23. Y. Zou, Z.W. Zhang, S.Y. Liu, D. Chen, G.X. Wang, Y.Y. Wang, M.L. Zhang, Y.H. Chen, Ultrasonic-Assisted Electroless Ni-P Plating on Dual Phase Mg–Li Alloy, J. Electrochem. Soc. 162 (2015) C64-C70. 24. H. Haferkamp, F.W. Bach, P. Bohling, P. Juchmann. Production, processing and properties of lithium-containing Mg-alloys, in: G. W. Lorimer (Ed.), Proceedings of the Third International Magnesium Conference, Institute of Materials, London, UK, 1997, pp. 177–192.

Galvanically Graded Interface: A Computational Model for Mitigating Galvanic Corrosion Between Magnesium and Mild Steel Kurt A. Spies, Vilayanur V. Viswanathan, Ayoub Soulami, Yuri Hovanski, and Vineet V. Joshi

Abstract

The impact of a graded metallic spacer on the galvanic corrosion between magnesium and mild steel is investigated in this work using a COMSOL model based on a validated numerical model. A graded spacer of 4 mm thickness decreases the peak galvanic corrosion by 50% over a system without a spacer, and 37% over a system with an aluminum spacer. The impact of the electrochemical properties of spacer materials is also investigated. Keywords

Multimaterial Corrosion




Joining




Magnesium




Alloy



Steel

Introduction No system in its entirety can be developed out of a single material. Design engineers are increasingly faced with the need to join dissimilar materials, as they are seeking creative new structures or parts with tailor-engineered properties. Widely different physical characteristics (melting point mismatch, coefficient of thermal expansion, etc.) and chemical incompatibility (presence of galvanic couple, chemical reaction at the interface) make dissimilar material joining a challenging task [1–4]. The applications of these materials systems are needed in all the emerging fields such as power electronics, fuel cells, automobiles, and aerospace. As an example, magnesium alloys, owing to their high specific strength, are widely being investigated for a range of structural applications such as in automotive, aerospace, and electronics packaging [5–8]. However, the application of magnesium alloys gets limited in multi-material systems K. A. Spies
V. V. Viswanathan
A. Soulami
Y. Hovanski
V. V. Joshi (&) Pacific Northwest National Laboratory, Richland, WA 99354, USA e-mail: [email protected]

because of their poor corrosion resistance [9, 10]. Magnesium being the most anodic of the structural materials [1, 11] is susceptible to galvanic corrosion and corrodes severely when interfaced with popular structural alloys such as aluminum and steel [12–15]. Conventional techniques of joining such as mechanical fastening [16], adhesive bonding [17], welding [18, 19], etc. are being developed to overcome this problem. However, a sacrifice in strength is usually accompanied with it. Additionally, coatings of zinc, cadmium, and Teflon/other polymers are also used on magnesium alloys at the interface of the two mating surfaces, which also comes at a cost penalty [16]. Apart from being susceptible to galvanic corrosion, magnesium alloys are also the least wear resistant [20]. The aforementioned coatings being applied on Mg alloys are also accompanied by maintaining a clean zone that is free of abrasives, fluxes, and oils which are commonly used to clean, weld, and lubricate during component manufacturing, since magnesium is susceptible to other forms of corrosion and wear in their presence [21]. Despite the plethora of literature on magnesium joining, all the techniques compromise on some aspects making it difficult or pricey to use magnesium alloys in mainstream products. In such systems where galvanic corrosion is a dominant concern, there are a variety of methods for mitigating the corrosion. These methods can be classified by mechanism: electrically insulating the interface of the two metals to prevent electron flow [10, 22, 23], the use of a spacer to separate the two metals to increase the resistance to ionic diffusion in the electrolyte [22, 24, 25], preventing/reducing ion transfer into the electrolyte by coating one or both metals [26–28], and the use of a sacrificial anode to preferentially corrode [29, 30]. Structural and material compatibility constraints of automobiles limit the practicality of plastic insulating spacers [31]. While coatings such as paint can dramatically reduce galvanic corrosion, scratches, and defects in the coatings can lead to high rates of corrosion due to the small surface area of exposed material [32]. Such a large difference in surface area generated by the use of coatings can exacerbate galvanic corrosion [33]. The salient feature of this work is the creation

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_21

135

136

of a graded galvanic interface using spacers that will improve the corrosion resistance. To explore galvanic corrosion mitigation by a spacer, a range of experimental systems are evaluated using a COMSOL finite element model based on a numerical model developed by Deshpande [25, 34]. There has been a significant amount of work [34–36] developing and optimizing numerical and computational models of galvanic corrosion. Models based on solving coupled partial differential equations for the distribution of potential and dissolved chemical species in the electrolyte have been experimentally validated [25, 27, 29, 34–40]. Numerical models have the advantage of speed over experimentally testing each system variable. They can be useful to frame experiments around important system characteristics and to generally explore scientific phenomena. The focus of this model is galvanic corrosion, because it is significant for magnesium incorporation into multi-metal systems. All general corrosion reactions between the individual metal and the electrolyte are ignored. Deshpande’s [25, 34] work with aluminum spacers has shown a positive impact on mitigating galvanic corrosion. However, materials with larger exchange current densities, different equilibrium potentials, and different charge transfer coefficients offer an opportunity to improve on the effectiveness of the spacer. Cross et al. explored galvanically graded interfaces as protection from corrosion and showed a positive impact from the graded system [27, 35]. Certain combinations of material properties should improve the corrosion mitigation of the spacer by altering the electrolyte potential due to surface oxidation of the spacer metal. The goal of this work was to use COMSOL modeling to study the impact of different spacer materials on the galvanic corrosion between mild steel and magnesium.

Model Development The objective of this study is to harness available corrosion numerical models to investigate the mitigation ability of spacer materials.

Fig. 1 Schematic view of the galvanic joint model geometry

K. A. Spies et al.

Geometry For this investigation, the modeled system is equal sized sections of mild steel and magnesium separated by various spacer materials. The basic geometry of the system is designed to be the simplest way to study the impact of spacers on the galvanic corrosion between magnesium and mild steel. The magnesium and mild steel are set to lengths of 10 cm, and the spacer length, also referred to as thickness, is set to 0–50 mm. On top of the metal joint is a NaCl electrolyte solution with a depth, in the y direction, of 10 mm. The electrolyte was chosen due to its relevance to corrosion in the automobile industry. The dimensions of the system were chosen to approximate a semi-infinite system. The electrolyte potential at the far ends of the magnesium and mild steel show no impact from the galvanic couple. The system is inherently 2D because there are no variations in the z-coordinate dimension. However, to aid in visualization, the system was modeled as a 3D system with a z-dimension of 10 mm, Fig. 1.

Equations The standard electrode potential in a half-cell is defined as the equilibrium reaction potential for an electrode in an electrolyte compared to a standard potential. In a galvanic system, the electrode with the more negative standard electrode is defined as the anode and the more positive electrode is defined as the cathode. For real systems, the potential deviates from equilibrium based on the overpotential of the system, η. This overpotential is a combination of resistances in the electrolyte to the transport of charged species, electrical resistances to the flow of electrons, and resistances from reaction kinetics. The sign of the potential is defined relative to the frame of reference. For example, at the anode the electrode loses ions and the electrolyte gains ions, so the electrolyte potential, /, is defined as the inverse of the electrode potential, E, Eq. 1. EðxÞ ¼ /ðxÞ

ð1Þ

Galvanically Graded Interface: A Computational Model …

137

The framework of the corrosion model is based on the numerical model developed and experimentally validated by Deshpande [25]. The electrolyte in the system is assumed to be well-mixed, incompressible, and electro-neutral. This reduces the governing equations of the electrolyte potential to the Laplace Equation, 2. r2 / ¼ 0

ð2Þ

All of the electrolyte boundaries away from the metal electrodes are modeled with the insulation boundary condition, 3 [25]. r/ ¼ 0

ð3Þ

Ohm’s Law is the boundary condition at the metal/electrolyte interface, 4 [41]. Where j is the electrical conductivity of the system, and i(/) is the net cathodic and anodic current density [42]. ið/Þ ¼ j
r/

ð4Þ

The current density function is a combination of a several electrochemical mechanisms. First, there are a range of surface reactions that can occur including: metal oxidation (5), hydrogen gas formation (6), water formation (7), and the creation of hydroxide (8). M ðsÞ ! M n þ þ ne

ð5Þ

2H þ þ 2e ! H2 ðg)

ð6Þ

Fig. 2 Nonlinear polarization behavior of steel, magnesium, aluminum, and magnesium alloys in 0.1 M NaCl at 25 °C from literature [43–45]. The electrode potential is plotted relative to a saturated calomel reference electrode

O2 ðg) þ 4H þ þ 4e ! H2 O(l)

ð7Þ

O2 ðg) þ 2H2 O(l) þ 4e ! 4OH

ð8Þ

The metal oxidation is the anodic reaction, while the cathodic reaction is based on the local electrolyte pH and dissolved gas composition. To mimic real systems, our model uses experimentally measured polarization curve data to simulate the surface reaction rates on the electrode materials, Fig. 2. Materials with low exchange current densities, such as aluminum, have lower rates of reaction at their equilibrium potential than materials with high exchange current densities, such as Mg24Y5. For materials without available experimental polarization curve data, their electrochemical kinetics is modeled by the Butler–Volmer equation, 9 [41].      aa F ac F ðgÞ  exp ð gÞ ð9Þ i ¼ i0 exp RT RT where F is the Faraday constant, R is the universal gas constant, T is the reaction temperature, η is the surface overpotential, a is the charge transfer coefficient, and i0 is the exchange current density [41]. The surface overpotential is the difference between the electrode potential and the equilibrium potential. The equations and boundary conditions are solved by finite element modeling to determine a steady-state potential distribution in the electrolyte. A corrosion current density can be determined from the system potential by the polarization curve data for magnesium and the spacer materials.

138

K. A. Spies et al.

Using the corrosion current density, a corrosion rate can be calculated using Faraday’s law, Eq. 10. CR ¼ i

M zFq

ð10Þ

where q is the density of the corroding metal, M is its molecular mass, and z is the number of electrons per metal ion.

Parameters To ensure realistic and meaningful results from the corrosion model, relevant parameters are chosen from literature and shown in Table 1. The electrolyte is a neutral pH 0.1 M NaCl (0.58 wt% NaCl). The salt concentration is about half that of the similar Deshpande galvanic corrosion model [25], but was chosen due to its relevance to atmospheric conditions and wide use in commercial magnesium alloy literature. Systems with lower electrolyte conductivities have a more acute corrosion rate position distribution along the anode electrolyte interface, making these systems more responsive to the incorporation of a spacer [31]. The conductivity of the electrolyte is calculated by interpolation from published NaCl concentration and conductivity relationship data [46]. Equilibrium potential data for the relevant materials are taken from literature tests of single material polarization curves in 0.1 M NaCl electrolyte [43–45]. Due to the availability of a wide range of electrochemical polarization data at 25 °C, the model is also performed at that temperature [43–45]. For the scenario with a gradated spacer, the electrochemical properties were calculated by using a linear distribution of materials between the most promising single

Table 1 Parameter values used in the model

material spacer and magnesium. For the range of potentials that were important for the spacer, only the exchange current density was important so the equilibrium potential and charge transfer coefficient were fixed.

Results and Discussion To explore the impact of spacer material characteristics that influence their corrosion mitigation properties, three types of scenarios were investigated. First, the impact of spacer thickness was evaluated in a scenario with an electrochemically inactive spacer. Second specific material properties of the spacers were studied in a single spacer material scenario. Finally, spacers with a gradient of materials were explored.

Electrochemically Inactive Spacer Spacer materials with very low exchange current densities for their polarization behavior, such as aluminum, have very slow reaction kinetics for both anodic and cathodic reactions in the electrolyte, Fig. 2. This minimal reaction rate means the concentration of ions in the electrolyte, and the electrons in the electrode are not directly impacted by the spacer itself. This simplifies the corrosion mitigation of spacer purely to an increased transport resistance in the electrolyte as the spacer thickness increases. This simplification makes electrochemically inactive spacers a good first scenario. Compared to a system with no spacer, the addition of a spacer drives the electrolyte potential at the magnesium anode in a positive direction, Fig. 4. This potential shift is beneficial because it decreases the rate of corrosion on the

Property

Variable

Value

References

Electrolyte conductivity

r

1.276 S/m

[44]

Equilibrium potential, mild steel

E0S

−0.631 V

[43–45]

Equilibrium potential, Mg2Al3

E0Mg2Al3

−1.152 V

[43–45]

Equilibrium potential, Mg17Al12

E0Mg17Al12

−1.350 V

[43–45]

Equilibrium potential, Al

E0Al

−1.363 V

[43–45]

Equilibrium potential, Mg2Si

E0Mg2Si

−1.563 V

[43–45]

Equilibrium potential, Mg–ZK60

E0ZK60

−1.586 V

[43–45]

Equilibrium potential, Mg24Y5

E0Mg25Y5

−1.607 V

[43–45]

Equilibrium potential, Mg

E0Mg

−1.642 V

[43–45]

Equilibrium potential, Mg–WE54

E0WE54

−1.673 V

[43–45]

Spacer thickness

xS

1–50 mm

[43–45]

Equilibrium potential, model

E0M

−1.607 V

[43–45]

Charge transfer coefficient, model

aM

0.04

NA

Exchange current density, model

i0M

0.0065–0.0135 A/cm2

NA

Galvanically Graded Interface: A Computational Model …

magnesium. Since the spacer is not contributing or removing ions or electrons from the system, the potential shift can only come from an increase in the overpotential due to increased ion transport resistance through the electrolyte as the steel and the magnesium are separated by the spacer thickness. The experimental data published by Deshpande [25] and their numerical model show a linear decrease in peak corrosion current for a system with magnesium and mild steel as an aluminum spacer increases in thickness from 1 to 5 mm. However, our work shows that applying their numerical model to larger spacer thicknesses leads to a decreasing effectiveness of the spacer as the thickness increases past 5 mm, Fig. 4. This finding is corroborated by Jia et al. who experimentally observed a nonlinear relationship between the thickness of a spacer and the peak corrosion current [22–24].

Single Electrochemically Active Spacer While the spacer can have a significant impact on the peak corrosion current by increasing the electrolyte resistance from increased path-length, spacers with sufficient exchange current densities can add an additional beneficial or deleterious impact to the corrosion rate. The polarization behavior of potential spacer materials can be very diverse, Fig. 2, but only the behavior in the voltage window that the spacer traverses is important. For a system with a mild steel cathode and a magnesium anode, the spacer experiences electrolyte potentials between about 1.20 and 1.50 VSCE (e.g., Fig. 3).

Fig. 3 Current density and electrolyte potential for a baseline system and for a system with a 10-mm electrochemically inactive spacer plotted as a function of position. The electrolyte potential is plotted relative to a SCE reference electrode. The position of the mild steel in the baseline system is plotted offset to overlay the mild steel in the spacer system

139

The value of the anodic terminus of the electrolyte potential (*1.50 VSCE) fundamentally controls the rate of magnesium corrosion because the magnesium rate at this interface is almost always the location of peak corrosion. A change from a 5-mm electrochemically inactive spacer to a 10-mm spacer shifts the electrolyte potential from 1.464 to 1.482 which equates to a 12% decrease in the peak magnesium corrosion rate. To explore how the electrochemical properties of the spacer impacts its ability to reduce corrosion six magnesium alloys and aluminum were modeled over a range of spacer thicknesses, 1–50 mm, Fig. 4. The corrosion mitigation was quantified by the peak current density in the respective systems, which corresponds to the corrosion rate by Faraday’s law. Materials that have cathodic behavior in the spacer potential range (e.g., Mg2Al3 & Mg17Al22), Fig. 2, lead to an increase in corrosion rate compared to an electrochemically inactive spacer. Cathodic materials consume electrons that are generated by the corroding magnesium increasing the driving for magnesium corrosion, Fig. 4. The cathodic behavior, shown as a negative current density, of Mg2Al3 and Mg17Al22 can be seen in Fig. 5. Materials that have anodic behavior in the spacer potential range, such as Mg2Si, Mg–ZK60, Mg24Y5, and Mg–WE54, are promising as spacer materials because they show a decline in the peak corrosion current density. The decline is due to the spacer metal experiencing oxidization which generates electrons and decreases the corrosion driving force for magnesium corrosion. However, if the oxidation rate is too high, the peak corrosion location shifts from the magnesium to the

140 9.0E-03 Electrochemically Inactive Mg2Al3

8.0E-03

Mg17Al12 Al

7.0E-03 Peak Current Density, A/cm2

Fig. 4 Peak corrosion current density for 7 different spacer materials systems as a function of spacer thickness. The electrochemically inactive spacer line (—) illustrates a system that has no anodic or cathodic influences on the electrolyte potential

K. A. Spies et al.

Mg2Si Mg-ZK60

6.0E-03

Mg24Y5 Mg-WE54

5.0E-03 4.0E-03 3.0E-03 2.0E-03 1.0E-03 0.0E+00 0

10

spacer. This corrosion location shift can be seen in Fig. 4 where the corrosion rate plateaus at a specific thickness for each material. The spacer thickness that causes the peak corrosion rate to plateau occurs at smaller spacer thickness for more anodic spacer materials. The anodic behavior of these materials for a scenario with a 10-mm spacer is shown in Fig. 5. With this spacer thickness, the peak corrosion current density for Mg–WE54 is located at the magnesium Fig. 5 Current density for seven spacer materials with a 10-mm thickness plotted as a function of position

20 Spacer Size, mm

30

40

50

interface while the peak current density for the more anodic Mg24Y5 occurs at the spacer interface with the steel, Fig. 5. The electrochemical behavior for various 10-mm spacer materials and the mild steel and magnesium as a function of position is shown in Fig. 5. A negative behavior represents reduction occurring while a positive value represents oxidation. There are two main material properties that impact the effectiveness of the spacer materials, (1) their exchange

Galvanically Graded Interface: A Computational Model … Table 2 Corrosion mitigation as a function of spacer material and thickness

141

Spacer material

Spacer thickness (mm)

% Decrease in peak corrosion rate from baseline

Inactive

50

65.0

Mg2Si

50

72.3

Inactive

10

35.1

Mg2Si

10

36.2

Mg–WE54

10

58.2

Inactive

5

26.2

Mg–WE54

5

38.1

Mg24Y5

5

47.6

Inactive

1

10.8

Mg24Y5

1

21.8

current density and (2) if they are anodic or cathodic in the spacer potential range. Materials with low exchange current densities (Mg2Al3, Mg17Al12, Al, and Mg2Si) show very similar impact on the current density on the magnesium anode. This is because the rate of oxidation or reduction reactions for these materials is not significant compared to the magnesium and mild steel. Whether these materials are anodes or cathodes in the spacer potential range is mostly irrelevant. Three materials that were modeled had sufficient electrochemical kinetics to make a noticeable impact on the corrosion behavior of the system, Mg–ZK60, Mg–WE54, and Mg24Y5. For these materials, their equilibrium potential which controlled the electrochemical behavior as a spacer was directly related to how effectively they reduced the peak corrosion rate. The scenario with a Mg24Y5 spacer showed the largest corrosion decline on the magnesium because of the high anodic polarization behavior of Mg24Y5, Fig. 2. However, the anodic rate is so large on the Mg24Y5 material that it became the location of the peak corrosion rate instead of the magnesium in the 10-mm scenario. In the 10-mm scenario, the Mg–WE54 spacer was the most effective in reducing the peak corrosion. The effectiveness of the Mg–WE54 spacer can be traced to the fact that at 10 mm the peak corrosion rate on the spacer and the magnesium is nearly identical. This dualism effectively leads to the highest average corrosion rate for a given peak corrosion rate. This is an optimal scenario because it will lead to the lowest possible peak corrosion rate for a given spacer thickness. A breakdown of the relationship between spacers’ thickness and spacer material on the peak corrosion rate of the system is shown in Table 2.

Graded Spacer As noted earlier for single spacer material systems, when the peak corrosion on the magnesium and the spacer is about the same, the system has the lowest possible peak corrosion rate.

However, single material spacers are limited by the fact that the maximum current density peaks can only occur at only one point on the spacer. Spacers with multiple materials offer the opportunity to optimize each spacer material so that it has the same peak corrosion current density as the other spacer materials. By effectively holding the corrosion rate constant across the spacer at the peak system corrosion rate, the rate of corrosion is spread out from a localized weak point to the whole surface dramatically improving the effectiveness of the spacer at corrosion mitigation. A graded spacer system offers many possible design variables for an optimal design but it is unclear what is the most important. The authors understand that the use of a series of intermetallics can be detrimental to the joint strength. The intermetallics used in this model are to represent the impact of alloying metals to shift their Tafel plots in a similar way to recent work done by Gusieva et al. [30]. One way that a graded material spacer could be effectively generated is to friction stir two alloys together [11]. Starting with the Butler–Volmer equation as a guide (Eq. 8), there are three independent variables: (1) exchange current density, (2) equilibrium potential, and (3) charge transfer coefficient. Based on the single spacer findings, a sufficient exchange current density is needed to allow the spacer to electrochemically interact with the electrolyte. In addition, an appropriate equilibrium potential that sets the material kinetics to be anodic in the potential range of the spacer is important. However, the impact of charge transfer coefficient is not known. The charge transfer coefficient is a representation of the impact on potential on the kinetics of the material both as an anode and as a cathode. A small charge transfer coefficient indicates a small relationship between current density and over potential. While a large charge transfer coefficient (a*1) indicates a large dependency of the current density on the over potential. For the corrosion system, the impact of the charge transfer coefficient is on the slope of decay of corrosion current density from its peak value as the potential of the spacer shifts more

142

K. A. Spies et al.

Fig. 6 Nonlinear polarization behavior of steel, magnesium, and modeled gradient spacer materials in 0.1 M NaCl at 25 °C. Plotted is the 8 gradient spacer material scenario. The shaded area represents the approximate potential range of the spacer. The electrode potential is plotted relative to a saturated calomel reference electrode

anodic with position towards the magnesium. To model a realistic graded transition across the spacer, a spacer was assumed to be made up of Mg24Y5 and magnesium. The Mg24Y5 was chosen because it has the highest anodic reaction kinetics of the materials tested. It was chosen to be mixed with magnesium because magnesium has an exchange current density that is more anodic and it would

Fig. 7 Current density for spacer scenarios with multiple layers compared to single material spacers plotted as a function of position

simulate a friction stirred transition between a spacer alloy and the magnesium. To generate polarization curves for assumed materials in the gradient, the Butler–Volmer equation, 8, was used, Table 1. A plot of the polarization performance for an 8 layer of graded spacer is shown in Fig. 6. The potential range of the spacer, shaded grey, is the only important potential range for this scenario.

Galvanically Graded Interface: A Computational Model … Table 3 Corrosion mitigation of gradient spacer systems

Spacer size (mm)

143 1

2

3

4

5

% Decrease in peak corrosion rate from baseline system Inactive

10.8

16.5

20.5

23.7

26.2

Mg24Y5

21.8

32.4

39.5

48.3

47.6

2-Layer

25.5

36.9

43.4

47.1

50.4

4-Layer

27.1

38.9

45.9

48.8

50.6

8-Layer

27.9

39.7

47.0

49.8

50.7

16-Layer

28.0

39.8

47.6

50.0

50.8

% Decrease from inactive spacer Mg24Y5

12.3

19.0

23.9

34.9

29.0

16-Layer

19.0

27.9

34.1

37.1

32.8

The graded system was tested in four different scenarios of 2, 4, 8, and 16 discrete material layers. The impact of increasing the number of layers in the graded system had a decreasing impact on the corrosion mitigation, Fig. 7. This is because the benefit of more layers is related to the deviation of the current density from an ideal horizontal line of peak current density across the spacer. Similar to the single material results, for a given graded spacer material system, an optimal thickness occurs where the peak corrosion rate decrease is the highest. Any thicker and the peak corrosion location shifts to the spacer and its mitigation plateaus, Table 3. The decreasing impact of the number of spacers can be seen in Table 3. For this assumed gradient between Mg24Y5 and magnesium, the performance plateaued around a 4-mm total spacer for the 16-layer scenario. The Mg24Y5 intermetallic reaches its optimum around the same spacer thickness. Compared to an electrochemically inactive spacer such as aluminum, the graded 16-layer system shows a dramatic 37.1% decrease in peak corrosion rate for a 4-mm spacer. This improvement is more dramatic at larger spacers up to 4 mm, because there is more spacer surface area that can oxidize and generate electrons to protect the magnesium.

Conclusions (1) Non-galvanic spacers decrease the rate of corrosion in a non-linear relationship with spacer size. (2) Only spacer materials with equilibrium potentials near the anode equilibrium potential (Mg) have a favorable impact on the corrosion mitigation of the spacer. If the spacer potential is too cathodic, the effectiveness of the spacer is reduced relative to an electrochemically inactive spacer. (3) Spacers with less than 10–7 A cm−2 exchange current density had little or no impact on the electrolyte

potential, due to their low reaction kinetics. The optimum spacer materials had exchange current densities that allowed for anodic polarization current densities in a similar range to the anode and cathode materials. (4) A graded system shows significant improvement over a single material spacer. However, the impact of the number of layers in the graded system drops off quickly. (5) For a graded material between Mg24Y5 and magnesium as a spacer, a 4-mm spacer was found to be the optimal size for a system of mild steel and magnesium.

References 1. M. G. Fontana, Corrosion Engineering. McGraw Hill, 2005. 2. P. Groche, S. Wohletz, M. Brenneis, C. Pabst, and F. Resch, “Joining by forming–A review on joint mechanisms, applications and future trends,” J. Mater. Process. Technol., vol. 214, pp. 1972–1994, 2014. 3. K. I. Mori, N. Bay, L. Fratini, F. Micari, and A. E. Tekkaya, “Joining by plastic deformation,” CIRP Ann. - Manuf. Technol., vol. 62, pp. 673–694, 2013. 4. Z. Sun and R. Karppi, “The application of electron beam welding for the joining of dissimilar metals: an overview,” J. Mater. Process. Technol., vol. 59, pp. 257–267, 1996. 5. E. F. Emley, Principles of Magnesium Technology. Oxford: Pergamon Press Ltd., 1966. 6. Q. Ge, D. Dellasega, A. G. Demir, and M. Vedani, “The processing of ultrafine-grained Mg tubes for biodegradable stents,” Acta Biomater., vol. 9, no. 10, pp. 8604–8610, 2013. 7. J. Hirsch and T. Al-Samman, “Supreior light metals by texture engineering: Optimized aluminum and magnesium alloys for automotive applications,” Acta Materialia, vol. 61, pp. 818–843, 2013. 8. W. Joost, “Reducgin Vehicle Weight and Improving U.S. Energy Efficiency Using Integrated Computational Materials Engineering,” JOM, vol. 64, pp. 1032–1038, 2012. 9. A. Atrens, Z. Shi, and G. L. Song, “Numerical modelling of galvanic corrosion of magnesium (Mg) alloys,” in Corrosion of Magnesium Alloys, G. I. Song, Ed. 2011, pp. 455–483.

144 10. G. Song, B. Johannesson, S. Hapugoda, D. St Jonn, and D. StJohn, “Galvanic Corrosion of Magnesium Alloy AZ91D in contact with an aluminium alloy, steel, and zinc,” Corros. Sci., vol. 46, no. 4, pp. 955–977, 2004. 11. L. Liu, “Introduction to the welding and joining of magnesium,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 3–8. 12. L. Liu, “Welding and Joining of Magnesium Alloys to Aluminum Alloys,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 38–63. 13. L. Liu, “The Joining of Magnesium Alloy to Steel,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 63–79. 14. L. Liu, “Corrosion and Protection of Magnesium Alloy Welds,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 79–94. 15. T. Watanabe, “Brazing and Soldering of Magnesium Alloys,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 97–121. 16. M. Heger and M. Horstmann, “Mechanical Joining of Magnesium Alloys,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 122–148. 17. L. Liu, “Adhesive Bonding of Magnesium Alloys,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 149–159. 18. G. Song, “Metal Inert Gas Welding of Magnesium Alloys,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 178–196. 19. G. Song, “Hybrid Laser-Arc Welding of Magnesium Alloys,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 229–253. 20. B. L. Mordlike and T. Ebert, “Magnesium: Properties - Applications - Potential,” Mater. Sci. Eng., vol. 302, pp. 37–45, 2001. 21. L. Liu, “Preparation for welding of magnesium alloys,” in Welding and Joining of Magnesium Alloys, L. Liu, Ed. Woodhead Publishing, 2010, pp. 16–22. 22. J. X. Jia, G. Song, and A. Atrens, “Influence of geometry on galvanic corrosion of AZ91D coupled to steel,” Corros. Sci., vol. 48, no. 8, pp. 2133–2153, 2006. 23. G. Song, A. Atrens, X. Wu, and B. Zhang, “Corrosion behaviour of AZ21, AZ501 and AZ91 in sodium chloride,” Corros. Sci., vol. 40, no. 10, pp. 1769–1791, Oct. 1998. 24. J. X. Jia, G. Song, and A. Atrens, “Experimental measurement and computer simulation of galvanic corrosion of magnesium coupled to steel,” Adv. Eng. Mater., vol. 9, no. 1–2, pp. 65–74, 2007. 25. K. B. Deshpande, “Effect of aluminium spacer on galvanic corrosion between magnesium and mild steel using numerical model and SVET experiments,” Corros. Sci., vol. 62, pp. 184–191, 2012. 26. F. Mansfeld and E. P. Parry, “Galvanic corrosion of bare and coated Al alloys coupled to stainless steel 304 or Ti-6Al-4 V,” Corros. Sci., vol. 13, no. 8, pp. 605–621, Jan. 1973. 27. S. R. Cross, R. Woollam, S. Shademan, and C. A. Schuh, “Computational design and optimization of multilayered and functionally graded corrosion coatings,” Corros. Sci., vol. 77, pp. 297–307, 2013. 28. T. Amorim, C. Allély, and J. Caire, “Modelling coating lifetime: first practical application for coating design,” 2008. 29. F. Thébault, B. Vuillemin, R. Oltra, K. Ogle, and C. Allely, “Investigation of self-healing mechanism on galvanized steels cut

K. A. Spies et al.

30.

31. 32.

33. 34.

35.

36.

37.

38.

39.

40.

41. 42.

43.

44.

45.

46.

edges by coupling SVET and numerical modeling,” Electrochim. Acta, vol. 53, no. 16, pp. 5226–5234, 2008. K. Gusieva, C. H. J. Davies, J. R. Scully, and N. Birbilis, “Corrosion of magnesium alloys: the role of alloying,” Int. Mater. Rev., vol. 60, no. 3, pp. 169–194, 2015. “Bimetallic Corrosion: Guides to Good Practice in Corrosion Control,” Teddington, Middlesex UK, 2000. H. S. Isaacs, A. J. Aldykiewicz, D. Thierry, and T. C. Simpson, “Measurements of corrosion at defects in painted zinc and zinc alloy coated steels using current density mapping,” Corrosion, vol. 52, pp. 163–168, 1996. F. Mansfeld, “Area Relationship in Galvanic Corrosion,” Corrosion, vol. 27, no. 10, pp. 436–442, 1971. K. B. Deshpande, “Validated numerical modelling of galvanic corrosion for couples: Magnesium alloy (AE44)-mild steel and AE44-aluminium alloy (AA6063) in brine solution,” Corros. Sci., vol. 52, no. 10, pp. 3514–3522, 2010. S. R. Cross, S. Gollapudi, and C. A. Schuh, “Validated numerical modeling of galvanic corrosion of zinc and aluminum coatings,” Corros. Sci., vol. 88, pp. 226–233, 2014. F. Thébault, B. Vuillemin, R. Oltra, C. Allely, and K. Ogle, “Reliability of numerical models for simulating galvanic corrosion processes,” Electrochim. Acta, vol. 82, pp. 349–355, 2012. C. R. Crowe and R. G. Kasper, “Ionic Current Densities in the Nearfield of a Corroding Iron-Copper Galvanic Couple,” J. Electrochem. Soc., vol. 133, no. 5, pp. 879–887, 1986. K. B. Deshpande, “Experimental investigation of galvanic corrosion: Comparison between SVET and immersion techniques,” Corros. Sci., vol. 52, no. 9, pp. 2819–2826, 2010. P. Doig and P. E. J. Flewitt, “A Finite Difference Numerical Analysis of Galvanic Corrosion for Semi-Infinite Linear Coplanar Electrodes,” J. Electrochem. Soc., vol. 126, pp. 2057–2063, 1979. S. Palani, T. Hack, J. Deconinck, and H. Lohner, “Validation of predictive model for galvanic corrosion under thin electrolyte layers: An application to aluminium 2024-CFRP material combination,” Corros. Sci., vol. 78, pp. 89–100, 2014. J. Newman and K. E. Thomas-Alyea, Electrochemical Systems, 3rd ed. Hoboken, NJ: John Wiley & Sons, 2004. A. Stenta, S. Basco, A. Smith, C. B. Clemons, D. Golovaty, K. L. Kreider, J. Wilder, G. W. Young, and R. S. Lillard, “One-dimensional approach to modeling damage evolution in galvanic corrosion,” Corros. Sci., vol. 88, pp. 36–48, 2014. R. Shimamura, A. Sugimoto, T. Fujiwara, and O. Seri, “Polarization behavior and its analysis of 1050 aluminum in solution containing chloride ions,” J. Japan Inst. Light Met., vol. 61, no. 7, pp. 303–309, 2011. A. D. Südholz, N. T. Kirkland, R. G. Buchheit, and N. Birbilis, “Electrochemical Properties of Intermetallic Phases and Common Impurity Elements in Magnesium Alloys,” Electrochem. Solid-State Lett., vol. 14, no. 2, p. C5, 2011. N. L. Sukiman, X. Zhou, N. Birbilis, A. E. Hughes, J. M. C. Mol, S. J. Garcia, and G. E. Thompson, “Durability and Corrosion of Aluminium and Its Alloys: Overview, Property Space, Techniques and Developments,” in Aluminium Alloys - New Trends in Fabrication and Applications, Z. Ahmad, Ed. 2012, pp. 47–97. G. R. Corporation, “Conductivity Ordering Guide.” [Online]. Available: https://www.grc.com/dev/ces/tns/conductivity_v_ concentration.pdf. [Accessed: 01-Jan-2015].

Iron Content in Relationship with Alloying Elements and Corrosion Behaviour of Mg3Al Alloys Ha Ngoc Nguyen, Jong Il Kim, Young Min Kim, and Bong Sun You

Abstract

In the present study, the effect of various alloying elements on controlling of iron content in magnesium alloys was investigated. A number of alloying elements including manganese, calcium, and yttrium were added separately and simultaneously to the melt then the iron content was determined. The effect of elements on iron removing was evaluated and compared. The results were interesting since it has changed the idea of considering manganese as the best to remove iron from the melt. Other elements would be a better candidate to improve the corrosion resistance of magnesium alloys, especially yttrium due to its double effect on iron removing and forming of a protective film to protect the matrix. This study would contribute effectively to a design of high-performance magnesium alloys in the future. Keywords

Magnesium alloy




Corrosion




Iron remove

Introduction Magnesium alloys have been replacing other structural materials as steels and aluminum in aerospace, auto and electronic device industries due to their lightest weight. However, the expansion of using magnesium alloys has been restricted by their disadvantage of low corrosion resistance [1]. Therefore, a ton of studies on the development of H. N. Nguyen (&)
Y. M. Kim
B. S. You Korea University of Science and Technology, Daejeon, Republic of Korea e-mail: [email protected] J. I. Kim Chungnam National University, Daejeon, Republic of Korea Y. M. Kim
B. S. You Korea Institute of Materials Science, Changwon, Republic of Korea

corrosion behaviour of magnesium alloys has been reported. Most of the studies focused on improving the corrosion resistance of magnesium alloys by alloying and explain the mechanism of the corrosion process [1–3]. On the other hand, a number of researches have been published about the effect of impurities on the corrosion behaviour of magnesium alloys [4, 5]. The worst impurity is iron which can seriously deteriorate the corrosion resistance of magnesium alloys even with a small amount. According to the ASTM standard specification for unalloyed magnesium ingot B92/B92 M-11, the iron content can vary in the range of 0.005–0.04 wt% depend on each type of the commercial ingot. In fact, this range of iron content can cause very different corrosion behaviour. This was proved by our experiment as shown in Fig. 1. Two different commercial pure magnesium ingots, type I and type II, which contain different iron level as indicated in Table 1, were used to compare the corrosion behaviour of pure magnesium and Mg3Al alloys. The results show that the iron content of around 310 ppm can cause much higher corrosion rate than that of 20 ppm. Liu [4, 5] found that the corrosion rates of magnesium alloys were typically below 10 mm/y; however, if the iron impurity level is more than 40 ppm, the corrosion rates can vary from 20 mm/y to greater than 200 mm/y. Their experimental results revealed that there is a critical Fe-to-Mn weight ratio, above which the corrosion rate of magnesium alloy sharply increases as the Fe/Mn weight ratio increase, Fig. 2. Although none of critical Fe/Mn weight ratio above which the corrosion rate sharply increases was given in their study, this critical value was reported in others to vary in the range from 0.010 to 0.032 depending on the alloy type [6–9]. In order to avoid this unexpected effect of iron, manganese has been used widely for the purpose of removing of iron from the melt due to the formation of intermetallic between aluminum, manganese and iron which is heavy and settle down on the bottom of crucible during melting [10]. Manganese was used in almost magnesium aluminum-based commercial alloys. Although it was widely used for

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_22

145

146

H. N. Nguyen et al.

Fig. 1 Corrosion behaviour of pure Mg and Mg3Al alloys a, c type I, b, d type II Fig. 2 The corrosion rates against Fe/Mn weight ratio [4]

magnesium alloys production, the use of manganese has some disadvantages. First of all, the use of manganese chloride as the additive of manganese caused serious problems to the health of worker as well corrosion of tools for casting. Furthermore, the formation of aluminum–manganese intermetallic phases which have much higher electrochemical potential than the matrix can cause galvanic corrosion [11]. From the other point of view, a question was given, that is whether the manganese is the best element to remove iron. Besides the use of manganese to remove iron from the melt, the use of alloying elements has been reported as a good method to improve the corrosion resistance of magnesium alloys [11, 12]. Yttrium was reported to improve remarkably the corrosion resistance for magnesium alloys [11–14]. However, the mechanism of improving corrosion resistance by using manganese and yttrium or other elements is different. Recently, the alloys that contain calcium and yttrium were reported to have excellent properties such as high mechanical performance, good corrosion resistance and nonflammable property [11, 15]. The good corrosion behaviour of these alloys was assigned to the formation of passive films which can protect the matrix. There was no report on another way to explain the good corrosion behaviour of magnesium alloys by addition of these elements. Therefore, there are two different approaches to improve corrosion properties of magnesium alloys, by removing of iron content and by forming of passive protective film.

Table 1 Chemical composition (wt%) of pure magnesium by ICP

In this study, we considered the effect of alloying elements involve yttrium and calcium, on removing of iron content, and compare their effect to manganese in order to understand fully the role of these elements on the improvement of corrosion resistance of magnesium alloys.

Experimental The commercial pure magnesium (type II) that contains a higher content of iron (340 ppm) was used for casting of a number of alloys involve Mg3Al, Mg3Al0.1Mn, Mg3Al0.3Mn, Mg3Al0.1Ca, Mg3Al0.3Ca, Mg3Al0.1Y, Mg3Al0.3Y, and Mg3Al0.1Mn0.1Y. The resistant furnace was used for melting. Graphite crucibles coated by boron nitride were used to avoid the contamination during casting. Mg–20Ca and Mg–30Y master alloys were used to add calcium and yttrium to the melt at 780 °C. Commercial manganese chloride was used to add manganese at 750 °C. The melt was manually stirred for good distribution of elements, and hold 90 min at 700 °C for stabilizing before casting into the stainless steel billet mould (diameter 90 mm, high 240 mm) which was preheated at 200 °C. The samples were sectioned at 20 mm above the bottom of the billet to check iron content by optical emission spectrometry equipment.

Pure Mg

Mn

Fe

Al

Si

Ca

Type I

0.0610

0.0019

0.0160

0.0098

0.0010

Type II

0.0023

0.0310

0.0030

0.0021

0.0010

Iron Content in Relationship with Alloying Elements …

147

Fig. 3 Supersaturated iron particles were found in the pure magnesium type II

Results and Discussion Since the impurity of iron in pure magnesium (type II) is higher than its solid solubility, the supersaturated iron particles were found, Fig. 3. The chemical composition of the alloys in this study is listed in Table 2. Aluminum alloying is well known to decrease the iron tolerance limits; therefore, the iron content in the Mg3Al alloy is lower than that of pure magnesium. In order to predict the trend of results, the Factsage 6.4 software was used to draw the phase diagram and calculate the solubility of iron in the melt of all alloys in this study, Fig. 2. In the binary alloy, the solubility of iron at 700 °C was determined to be about 122 ppm (Fig. 4a). The addition of 0.3 wt% of manganese reduced this value to 74 ppm (Fig. 4b). On the other hand, the alloys that contain calcium and yttrium were predicted that have not a remarkable change of the solubility of iron in the melt at this temperature, 120 ppm (Fig. 4c) and 125 ppm (Fig. 4d), respectively. These calculated data have explained for the use of 0.3 wt% of manganese in almost commercial magnesium-aluminum based alloys in the industry, and it seems that there is no reason for use of calcium or yttrium for iron removing purpose. Despite this, the real experimental data presented an interesting discovery.

Table 2 Chemical composition (wt%) of alloys

The measured iron content in the binary alloy was 167 ppm instead of 122 ppm as calculated. The difference between calculated data and measured data can be assigned to the difference of real condition to the ideal condition that was used for calculating. Similarly, the measured iron content in the ternary alloys that contain manganese was higher than that of the predicted value, Fig. 5. The addition of 0.1 wt% of manganese reduced about 30% of iron content in the melt (from 167 to 119 ppm). The higher manganese content of 0.3 wt% further decreased the iron content up to 50% (to 82 ppm). These data have confirmed the effectiveness of manganese as iron remover in magnesium–aluminum alloys. In the case of calcium, it was similar to the predicted value, the alloys that contain 0.1 wt% Ca and 0.3 wt% Ca showed the same iron content, Fig. 6, which changes a little bit compared to Mg3Al alloy (138 ppm vs. 167 ppm). The lower the content of iron in Mg3Al0.1Ca and Mg3Al0.3Ca alloys comparing to binary alloy could be explained by longer casting process due to the addition of calcium. This means that calcium has none remarkable effect on iron removing. An interesting data were found with the addition of yttrium, Fig. 7. In contrast to the predicted value, the addition of 0.1 wt% as well 0.3 wt% Y dramatically pull out the iron content to 47 ppm which means the reduction of 72% of iron content compares to the binary alloy. This is even much more effective than using 0.3 wt% Mn (50% reduction of iron). These impressive data indicated the double effect of yttrium on improving the corrosion behaviour of the Mg3Al alloy. Yttrium has not only contributed to the formation of the passive film to protect the matrix but also helped to remove unexpected iron from the melt. The addition of 0.3 wt% Y did not further remove iron since this was extremely diluting the content of iron in the melt. Since both manganese and yttrium presented good effect on iron removing, they were added to the melt together in Mg3Al0.1Mn0.1Y alloy to consider their combined effect. As shown in Fig. 8, the iron content was quite similar to Mg3Al0.1Y alloy (39 ppm vs. 47 ppm). This suggested that by addition of yttrium (0.1 wt% or more), the usage of

Alloys

Al

Mn

Ca

Y

Mg

Mg3Al

2.698

0.0030

0.0001

0.0030

Bal.

Mg3Al0.1Mn

2.752

0.0498

0.0001

0.0030

Bal.

Mg3Al0.3Mn

2.683

0.2429

0.0001

0.0030

Bal.

Mg3Al0.1Ca

2.777

0.0032

0.0431

0.0030

Bal.

Mg3Al0.3Ca

2.772

0.0034

0.2450

0.0030

Bal.

Mg3Al0.1Y

2.678

0.0030

0.0009

0.0930

Bal.

Mg3Al0.3Y

2.666

0.0031

0.0026

0.3030

Bal.

Mg3Al0.1Mn0.1Y

2.737

0.0857

0.0002

0.1160

Bal.

148

H. N. Nguyen et al.

Fig. 4 Phase diagrams by FactSage 6.4 of alloys: a Mg3Al–xFe, b Mg3Al0.3Mn–xFe, c Mg3Al0.3Ca–xFe, d Mg3Al0.3Y–xFe

manganese is not necessary any more. As we mentioned above, the addition of manganese can cause problems for the health as well formation of intermetallic phases which have higher electrochemical potential than the matrix. Therefore, Fig. 5 Effect of Manganese addition on Fe content (wt%) in Mg3Al alloys

the use of yttrium seems to be the best approach to improve corrosion property of magnesium alloys. This study would be helpful for alloy design and process design of the magnesium industry in the future.

Iron Content in Relationship with Alloying Elements … Fig. 6 Effect of Calcium addition on Fe content (wt%) in Mg3Al alloys

Fig. 7 Effect of Yttrium addition on Fe content (wt%) in Mg3Al alloys

Fig. 8 Effect of Manganese and Yttrium added separately and together on Fe content (wt%) in Mg3Al alloys

Fig. 9 Comparison the effect of Mn, Ca and Y on iron removing in Mg3Al alloys

149

150

Conclusion The relationship between iron content and alloying elements was investigated in this study. The results are summarized in Fig. 9. There were differences between calculated and experimental data. As predicted, manganese has a good ability to remove iron from the melt, and the addition of calcium has no effect on iron removing. On the other hand, yttrium showed an impressive effect on iron removing. It is even much better than the use of manganese. The use of 0.1 wt% Y seems to be optimum to remove iron from the melt of Mg3Al alloy. This suggested the double effect of yttrium on the improvement of corrosion resistance of magnesium alloy. It has not only contributed to the formation of the passive film to protect the matrix but also helped to remove unexpected iron content effectively from the melt. Further study should be done in order to clarify the mechanism of iron removing by yttrium as well as confirm the corrosion behaviour of alloys. On the other hand, this study has given a hint of considering the optimization of alloying with calcium and yttrium in order to improve other properties such as non-flammability which is essential for auto and aerospace application. Acknowledgements This work was supported by the National Research Council of Science & Technology (NST) grant by the Korea government (MSIP) (No. CRC-15-06-KIGAM).

References 1. M. Esmaily. Fundamentals and advances in magnesium alloy corrosion. Progress in Materials Science 89: 92–193

H. N. Nguyen et al. 2. Guang Ling Song. Corrosion prevention of magnesium alloys. Woodhead publishing limited, 2013 3. Qu Liu. Enhanced corrosion resistance of AZ91 magnesium alloy through refinement and homogenization of surface microstructure by friction stir processing. Corrosion science 138: 284–296 4. Ming Liu. Impurity control and corrosion resistance of magnesium-aluminum alloy. Corrosion science 77: 143–150 5. Ming Liu. Calculated phase diagrams and the corrosion of die-cast Mg–Al alloys. Corrosion Science 51 (2009): 602–619 6. K.N. Riechek. Controlling the Salt Water Corrosion Performance of Magnesium AZ91 alloy. Society of Automotive Engineers Technical paper 850417, Inc., Warrendale, Pennsylvania, USA, 1985 7. J.E. Hillis. High Purity Magnesium AM60 alloy: the critical contaminant limits and the salt water corrosion performance. Society of Automotive Engineers Technical paper 860288, Inc., Warrendale, Pennsylvania, USA, 1986 8. J.E. Hillis. Composition and performance of an Improved Magnesium AS41 alloy. Society of Automotive Engineers Technical paper 890205, Inc., Warrendale, Pennsylvania, USA, 1989 9. W.E. Mercer. The critical contaminant limits and salt water corrosion performance of magnesium AE42 alloy. Society of Automotive Engineers Technical paper 920073, Inc., Warrendale, Pennsylvania, USA, 1992 10. ASM specialty handbook: magnesium and magnesium alloys. ASM Int Mater Park OH 1999 11. B. Mingo. Corrosion of Mg-9Al alloy with minor alloying elements (Mn, Nd, Ca, Y and Sn). Materials and Design 130: 48–58 12. Li Li. Effect of yttrium on corrosion behavior of extruded AZ61 Mg alloy. Journal of Magnesium and alloys 4, 2016: 44–51 13. Meisam Nouri. Beneficial effects of yttrium on the performance of Mg–3%Al alloy during wear, corrosion and corrosive wear. Tribology International 67, 2013: 154–163 14. Qiang Wang. Evaluating the improvement of corrosion residual strength by adding 1.0 wt.% yttrium into an AZ91D magnesium alloy. Materials Characterization 61, 2010: 674–682 15. Fan Jianfeng. Effect of yttrium, calcium and zirconium on ignition-proof principle and mechanical properties of magnesium alloys. Journal of Rare Earth 30, 2012: 74

Microstructures, Corrosion and Mechanical Properties of Mg–Si Alloys as Biodegradable Implant Materials Weidan Wang, Ming Gao, Yuanding Huang, Lili Tan, Ke Yang, and Norbert Hort

Abstract

Magnesium alloys attracted more and more attentions as biodegradable implant materials because of their properties similar to cortical bone. From the perspective of element biosafety and dietetics, the ideal alloying elements suitable for biodegradable applications should be those essential to or naturally presented in the human body. This study presents a novel aluminum-free magnesium alloy system with Si selected as a major alloying element, due to its superior biocompatibility in biological environment, especially in bone regeneration and repairment. Mg–Si binary alloys with different Si contents were prepared in a permanent mould via gravity casting and direct-chill casting. The microstructures, corrosion properties and mechanical properties were inves- tigated as a function of alloy composition. Keywords





Magnesium alloys Permanent mould casting Microstructure Corrosion properties




W. Wang
M. Gao
L. Tan (&)
K. Yang Institute of Metal Research, Shenyang, 110016, China e-mail: [email protected] W. Wang
Y. Huang
N. Hort (&) Magnesium Innovation Centre (MagIC), Helmholtz-Zentrum Geesthacht, 21502 Geesthacht, Germany e-mail: [email protected] W. Wang University of Chinese Academy of Sciences, Beijing, 100049, China M. Gao School of Materials Science and Engineering, University of Science and Technology of China, Shenyang, 110016, China

Introduction As important lightweight metallic structural materials, magnesium alloys have received increasing attentions in the last two decades because of their great potential for use in automotive and aerospace industries [1, 2]. In addition, as a new generation of biodegradable implants, Mg-based alloys are promising candidates for orthopedic clinical applications, owing to their biodegradability, biocompatibility and mechanical properties similar to that of natural bone [3]. However, most alloys currently investigated for biomedical applications were adopted from commercial Mg alloys, which were originally designed for engineering application [4, 5]. For example, AZ91, WE43, LAE442 contains aluminium (Al) and rare earth elements (REE). Al has controversially been implicated as a factor in Alzheimer’s disease. It could also inhibit the phosphorylation process, ATP synthesis and subsequently reduce the intracellular energy reserve. Epidemiological data are reinforced by indications that aluminium exposure can result in excess inflammatory activity within the brain [6, 7]. On the other hand, REE (rare earth elements) have been arising concerned since they are not naturally present in human body. Prospects studies were suggested, to evaluate long-term effects of REE exposures [8]. From this point of view, the alloying elements in magnesium matrix should be selected without harmful effects, to guarantee the biosafety of implant materials. Silicon (Si) was regarded as an essential mineral in the human body. It was reported that Si contributes to the growth of bone and connective tissue and helps to build the immune system [9, 10]. Meanwhile, Si is an essential trace element for bone development and is reported to only locate in the active areas of young bone. Its ions are proved to be involved in the calcification process of young bones; therefore, its existence has a beneficial effect on the osteogenesis process [11]. Review results from Luigi et al. also showed a positive relationship between dietary Si intake and bone regeneration [12].

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_23

151

152

Apart from the biological effects of Si ions, the addition of Si to Mg alloys also causes an increased fluidity of the molten metal and helps with the grain refinement. What’s more, very small amount of Si addition in the alloy can result in a pronounced increase in ductility, while the tensile strength remaining almost unaffected [13]. With the increasing content of Si, the hardness and tensile strength are improved due to the presence of Mg2Si precipitates. Mg–Si alloys have shown high potential as high-temperature structural material since Mg2Si reinforcement exhibits a high melting temperature of 1085 °C, high hardness of 4.5  109 N/m2, low density of 1.99  103 kg/m3 and a low thermal expansion coefficient of 7.5  10−6/K [14, 15]. This particle is very stable, thus can restrain grain boundary sliding at elevated temperature. However, the solubility of Si in magnesium is quite limited, which is only about 0.006 wt %. Therefore, as-cast Mg–Si alloys always exhibit a low ductility, due to the presence of coarse block Mg2Si particles or the brittle Chinese script Mg2Si phase. Microstructural refinement is an effective way to improve mechanical properties as well as corrosion properties. Many studies demonstrated the morphology of Mg2Si can be modified into finer shape by alloying with a third element. Zhang et al. [9] reported the effects of Zn and Ca additions to Mg–0.6Si alloys. Both element additions refined the grain size and modified the morphology of Mg2Si, as a result, the bio-corrosion resistance and mechanical properties were enhanced significantly. Jiang et al. [16] studied the effect of Y addition to Mg–Si alloys, indicating the minimum amount of Y is 0.8 wt% for Mg–5 wt% Si alloy, aiming to refine both the primary Mg2Si and Chinese script Mg2Si. However, when the Y content was higher than 1.2 wt%, the primary Mg2Si became coarse again. According to Guo’s investigation [17], the modification by Bi element also functioned. When Bi content exceeded 0.8 wt%, the over-modification of primary Mg2Si occurred. Other studies have also reported that the addition of Ba, Sb, La, Sr, P elements can change the morphology of Mg2Si [18–21]. Up to now, very few researches reported the influence of casting process on microstructural refinement of the Mg–Si based alloys. The casting process plays an important role in affecting the distribution of the alloying elements prior to the heat treatment step, and thus can influence the morphology, size and amount of second phase particles. Interests in solidification phenomenon were progressively increased during the past decades since microstructures and material phases were modified possibly by advanced processing techniques [22]. Permanent mold casting is a wellestablished manufacturing route for casting magnesium alloys. Recently direct-chill casting was reported [23, 24], which used a faster cooling rate compared with the previous techniques including sand casting or gravity casting. In the present study, Mg–Si binary alloys with different Si

W. Wang et al.

additions were prepared in a permanent mould via gravity casting and direct-chill casting at a laboratory scale. The microstructure, corrosion properties and mechanical properties were focused. The effects of cooling rate and alloy compositions on them were discussed.

Experimental Permanent mould casting was used to prepare all the investigated alloys. The melting and gravity casting process were carried out in an electric resistance furnace with a protection gas mixture of SF6 and Ar. The raw materials were continuously heated to above 750 °C and kept for 10 min to complete the dissolution of Si, then stirred for 5 min, and finally poured into a permanent mould preheated to 300 °C. The direct-chill casting method was also used to prepare the alloys as described previously [25]. First, pure Mg ingots and small quantities with few grams-weight Si chips were melted in a resistance furnace at 750 °C, under a constant flux of an Ar–SF6 mixture. The oxides on the surface of melt were cleaned out with a boron nitride-coated stainless steel paddle, after stirring. The permanent mould was preheated in a tubular furnace and was introduced again after pouring, with the same temperature of 750 °C. The mould and molten alloy were hold for 5 min, and then, a yaxis movement inductor was applied to achieve a uniform pulling velocity of 12 cm/min for the mould into a water bath below the tubular furnace. For each type of measurement, specimens were cut from the same position in cylinder ingots of different alloys. The contents of Si were determined by spark emission spectrometer (Spectrolab M, Spektro, Germany). Samples for microstructural observations were prepared by grinding with SiC emery papers up to 2500 grit and then polished with a lubricant containing 1-lm diamond particles and 0.05-lm colloidal silica (OPS). SEM observations were performed using a scanning electron microscope (SEM, TESCAN VEGA3-SB, Brno, Czech Republic) equipped with energydispersive X-ray spectroscopy (EDS) at an acceleration voltage of 15 kV. In order to identify the Mg2Si phases in the alloys, X-ray diffractions (XRD) were performed with Cu Ka radiation. The Vickers hardness measurements (HV5) were carried out in order to investigate the influence of Si content on the mechanical properties, using a standard microhardness tester (EMCOTEST M1C010 universal hardness tester) with a load of 5 kg and a dwell time of 10 s. An average of 10 measurements was evaluated for each alloy. The corrosion properties were obtained by electrochemical test carried on an Gamry electrochemical workstation. A standard three-electrode system was employed for all electrochemical measurements in simulated body fluid (Hank’s solution). In the potentiodynamic polarization test, the change in the

Microstructures, Corrosion and Mechanical Properties …

153

open-circuit potential (OCP) was first monitored for 5 min, to check the stability of the potential and to make sure air bubbles do not disturb the measurements.

Results and Discussion Microstructure The actual chemical compositions of all the investigated alloys are listed in Table 1. The contents of Si in the alloys are more or less deviated from the nominal chemical compositions, due to the burn off and extra addition when taking into account of mass loss during casting. Two different alloy groups are distinguished based on the casting method. The alloys names are shortened () and numbered with G representing gravity casting and DC referring to direct-chill casting, respectively. The impurity contents presented in Table 1 show that higher level of Fe existed in G group (320, 513 and 797 ppm), compared with *150 ppm for DC group alloys. While the content of Ni, Cu and Mn for gravity casting is less than that for DC casting. According to the solidification parameters of previous DC casting [25, 26], the specially optimized laboratory-scale installation was developed to ensure the “clean” ingots with homogeneous distribution of alloying elements, no macrosegregation, free of porosity and inclusions. In addition, the solid–liquid interface had a y-axis movement simultaneously as the mould pulled down and contacted with the water bath. The thermal gradient pushed iron impurities to the later-solidified part of the ingots, which was discarded before chemical analysis. Figure 1 shows the optical microstructures of Mg–Si alloys by gravity casting and direct-chill casting. A clear difference in the size, morphology and distribution of the Mg2Si particles can be observed. Mg–Si binary alloy is a typical eutectic system, according to the binary phase diagram, the eutectic point is at the 1.34 wt% Si (1.13 at.%) [27]. Thus, for the as-cast Mg–Si hypoeutectic alloys (G-1#, DC-1#, DC-2#, DC-3#), the microstructure mainly consists of primary Mg dendrite and eutectic structure which is made up of a-Mg and b-Mg2Si phase. It can be observed that

Table 1 Actual chemical compositions (wt%) of the investigated alloys

a-Mg formed by gravity casting is much coarser than that by DC casting. The significant reduction was mainly due to the higher cooling rate for DC casting. Thus, there was no sufficient time for primary Mg phase to grow. Meanwhile, rod-like eutectics were formed in the interdendritic regions. The morphology of eutectic phase is dependent on the entropy of melting and the amount of phases [28]. When the two phases have low fusion entropies, the rod-shaped eutectic structures appear. In addition, if the amount of phase is no more than *30%, the eutectic structure shows rod-like [23]. In the present Mg–Si alloys, at the end of liquid–solid phase transition, the interdendritic regions have a higher content of Si, leading to less amount of a-Mg and thus formation of rod-like eutectic structures. As for the hypereutectic alloys (G-2#, G-3#, DC-4#), the coarse polygonal primary Mg2Si, with blue colour in the etched micrographs, appeared only in the gravity casting. Formation of the pseudo-eutectic in DC-4# could be a result of the non-equilibrium solidification process caused by the high cooling rate of water quenching. Normally, the higher difference of melting points between two solid phases in the metallic–nonmetallic system, the more eutectic coupled zone deviated o high melting point phase. While for G-2# and G-3# alloys, an Mg-rich zone can be seen around the primary Mg2Si phases. This is due to the formation of primary crystallized Mg2Si phase. The concentration of Si drops dramatically in its surrounding area. Figure 2 displays the representative back-scattered electron (BSE) micrographs of Mg–Si alloys by gravity casting and DC casting. As described above, the as-cast hypoeutectic alloys of DC-2# and G-1# exhibited a microstructure consisted of the primary a-Mg phase and the net-shaped eutectic phases at the interdendritic regions. The mapping result of DC-4# was presented with colourful image, since the volume fraction of eutectic phases increases with increasing the content of Si. In this case, the microstructure was difficult to discern using back-scattered electron. As indicated in the mapping result, the dark dendritic structure was a-Mg phase and the coloured field was Si-rich area. The EDS results of the corresponding areas in BSE micrographs are shown in Table 2. The higher concentration

Alloys

Si

Fe

Ni

Cu

Mn

Mg

Name as

Mg–0.6Si

0.66

0.0032

0.00091

0.00116

0.00153

Bal.

G-1#

Mg–1.2Si

1.54

0.00513

0.00128

0.00106

0.00187

Bal.

G-2#

Mg–1.6Si

1.81

0.00797

0.00078

0.00136

0.00209

Bal.

G-3#

Mg–0.3Si

0.20

0.0011

0.00122

0.00152

0.00191

Bal.

DC-1#

Mg–0.6Si

0.35

0.00156

0.00156

0.00161

0.00213

Bal.

DC-2#

Mg–1.2Si

1.02

0.00122

0.00114

0.00155

0.00193

Bal.

DC-3#

Mg–1.6Si

1.50

0.00159

0.00118

0.00159

0.00205

Bal.

DC-4#

154

W. Wang et al.

Fig. 1 Optical microstructure of Mg–Si alloys by gravity casting and DC casting

Fig. 2 Representative back-scattered electron (BSE) images of Mg–Si alloys by gravity casting and DC casting, enlarged area for DC group were shown inside

Microstructures, Corrosion and Mechanical Properties … Table 2 EDS results of the corresponding areas in Fig. 2

155

Elt. (at.%)

Point a

Point b

Point c

Point d

Point e

Mg Ka

99.22

98.80

99.87

66.47

66.66

Si Ka

0.78

1.20

0.13

33.53

33.34

of Si in Mg matrix for DC casting suggested its supersaturation, which was attributed to the fast cooling rate of DC casting. The supersaturation phenomenon was also reported by Jiang et al. [23] in WE43 magnesium alloy fabricated by DC casting. EDS analysis of the light-grey coarse blocks in G-2# and G-3# verified the compound was Mg2Si, with a ratio of at.% Mg: at.% Si to be 2. The above obtained results were well consistent with the XRD results, as shown in Fig. 3. All alloys are composed of two phases, a-Mg and intermetallic compound Mg2Si with face-centred cubic crystal structure. For both casting group, the amount of Mg2Si particles increased with increasing the content of Si. What’s more, a higher intensity of Mg2Si phase peak was observed in gravity casting XRD patterns, owing to their higher amount in hypoeutectic phases compared with DC casting.

Fig. 3 XRD pattern identifications of Mg–Si alloys

Mechanical and Corrosion Properties Figure 4 displays the Vickers hardness of Mg–Si binary alloys as a function of Si content and casting method. In both casting condition, the hardness increased obviously with increasing Si content. According to the Mg–Si binary phase diagram, there is no solubility of Si in Mg and more Mg2Si phase can be formed owing to the increased addition of Si; thus, the improvement in hardness was mainly attributed to the second phase strengthening. Moreover, the hardness of DC casting exhibited higher value, owing to the more homogenous distribution of eutectic phases and finer a-Mg dendrites. Potentiodynamic polarization curves are displayed in Fig. 5. As known, the cathodic polarization current reflects the severity of hydrogen evolution reaction while the anodic branch represents the dissolution of magnesium alloys. DC casting alloys presented a lower cathodic current density compared with the gravity casting alloys, implying enhanced corrosion resistance was obtained by increasing the solidification cooling rate. In addition, the lower corrosion current density was found with increased Si content. For magnesium alloy, the a-Mg matrix is usually anodic to the second phase or compound; thus, these particles are normally inert in the corrosion reaction and can act as two different roles, depending on their distribution and structure. On the one hand, eutectic phases can play the role of cathode to a-Mg

Fig. 4 Vickers hardness as a function of Si content by gravity casting and DC casting

156

Fig. 5 Potentiodynamic polarization curves of Mg–Si alloys in Hank’s solution

and cause an accelerated corrosion rate. While on the other hand, they can be a corrosion barrier to retard the corrosion when they have sufficient amount leading to the formation of a network structure. As a result, the overall corrosion properties are controlled by the dominated process.

Conclusions The influence of Si additions on microstructure, mechanical and corrosion properties was investigated. Both gravity casting and direct-chill casting were performed when preparing the binary alloys in a permanent mould. A different microstructure with finer a-Mg dendrites and more homogenous eutectic phase was obtained for DC casting under a higher cooling rate. Moreover, the mechanical and corrosion behaviours of DC casting alloys were improved compared to gravity casting. Acknowledgements The first author would like to acknowledge the Chinese Scholarship Council (CSC) for a scholarship. The research was supported by Key program of China on biomedical materials research and tissue and organ replacement (No. 2016YFC1101804, 2016YFC1100604) and funding from Institute of Metal Research, Chinese Academy of Sciences (No. 2015-ZD01). Sincerest gratitude goes to the colleagues from Magnesium Innovation Centre (HZG) for their technical supports.

References 1. J. Li, G. Sha, T. Wang, W. Jie, S. Ringer, Precipitation microstructure and age-hardening response of an Mg–Gd–Nd– Zn–Zr alloy, Materials Science and Engineering: A 534 (2012) 1–6.

W. Wang et al. 2. G. Riontino, M. Massazza, D. Lussana, P. Mengucci, G. Barucca, R. Ferragut, A novel thermal treatment on a Mg–4.2 Y–2.3 Nd–0.6 Zr (WE43) alloy, Materials Science and Engineering: A 494(1–2) (2008) 445–448. 3. D. Zhao, F. Witte, F. Lu, J. Wang, J. Li, L. Qin, Current status on clinical applications of magnesium-based orthopaedic implants: A review from clinical translational perspective, Biomaterials 112 (2017) 287–302. 4. F. Witte, N. Hort, C. Vogt, S. Cohen, K.U. Kainer, R. Willumeit, F. Feyerabend, Degradable biomaterials based on magnesium corrosion, Current opinion in solid state and materials science 12 (5–6) (2008) 63–72. 5. W. Wang, J. Han, X. Yang, M. Li, P. Wan, L. Tan, Y. Zhang, K. Yang, Novel biocompatible magnesium alloys design with nutrient alloying elements Si, Ca and Sr: Structure and properties characterization, Materials Science and Engineering: B 214 (2016) 26–36. 6. S.C. Bondy, Low levels of aluminum can lead to behavioral and morphological changes associated with Alzheimer’s disease and age-related neurodegeneration, Neurotoxicology 52 (2016) 222–229. 7. Z. Wang, X. Wei, J. Yang, J. Suo, J. Chen, X. Liu, X. Zhao, Chronic exposure to aluminum and risk of Alzheimer’s disease: A meta-analysis, Neuroscience letters 610 (2016) 200–206. 8. G. Pagano, F. Aliberti, M. Guida, R. Oral, A. Siciliano, M. Trifuoggi, F. Tommasi, Rare earth elements in human and animal health: state of art and research priorities, Environmental research 142 (2015) 215–220. 9. E. Zhang, L. Yang, J. Xu, H. Chen, Microstructure, mechanical properties and bio-corrosion properties of Mg–Si (–Ca, Zn) alloy for biomedical application, Acta biomaterialia 6(5) (2010) 1756– 1762. 10. A. Gil-Santos, I. Marco, N. Moelans, N. Hort, O. Van der Biest, Microstructure and degradation performance of biodegradable Mg-Si-Sr implant alloys, Materials Science and Engineering: C 71 (2017) 25–34. 11. E.M. Carlisle, Silicon as an essential trace element in ammal nutrition, Silicon biochemistry 703 (2008) 123. 12. L.F. Rodella, V. Bonazza, M. Labanca, C. Lonati, R. Rezzani, A review of the effects of dietary silicon intake on bone homeostasis and regeneration, The journal of nutrition, health & aging 18(9) (2014) 820–826. 13. Y. Guangyin, L. Manping, D. Wenjiang, A. Inoue, Microstructure and mechanical properties of Mg–Zn–Si-based alloys, Materials Science and Engineering: A 357(1–2) (2003) 314–320. 14. J. Zhang, Z. Fan, Y. Wang, B. Zhou, Microstructural development of Al–15wt.% Mg2Si in situ composite with mischmetal addition, Materials Science and Engineering: A 281(1–2) (2000) 104–112. 15. G.F.S. Beer, E. Schmid Proc. Conf. Magnesium Alloys and their Applications. (1992), pp. 317–324. 16. Q. Jiang, H. Wang, Y. Wang, B. Ma, J. Wang, Modification of Mg2Si in Mg–Si alloys with yttrium, Materials Science and Engineering: A 392(1–2) (2005) 130–135. 17. E. Guo, B. Ma, L. Wang, Modification of Mg2Si morphology in Mg–Si alloys with Bi, Journal of Materials Processing Technology 206(1–3) (2008) 161–166. 18. A. Srinivasan, S. Ningshen, U.K. Mudali, U. Pillai, B. Pai, Influence of Si and Sb additions on the corrosion behavior of AZ91 magnesium alloy, Intermetallics 15(12) (2007) 1511–1517. 19. M. Baoxia, W. Liping, G. Erjun, Modification Effect of Lanthanum on Primary Phase Mg2Si in Mg-Si Alloys, J.Chin. Rare Earth Soc. 26(1) (2008) 87. 20. A. Gil-Santos, N. Moelans, N. Hort, O. Van der Biest, Identification and description of intermetallic compounds in Mg–Si–Sr cast and heat-treated alloys, Journal of Alloys and Compounds 669 (2016) 123–133.

Microstructures, Corrosion and Mechanical Properties … 21. J.J. Kim, D.H. Kim, K. Shin, N.J. Kim, Modification of Mg2Si morphology in squeeze cast Mg–Al–Zn–Si alloys by Ca or P addition, Scripta Materialia 41(3) (1999) 333–340. 22. J. Zhang, Z. Fan, Y. Wang, B. Zhou, Effect of cooling rate on the microstructure of hypereutectic Al–Mg2Si alloys, Journal of materials science letters 19(20) (2000) 1825–1828. 23. H.S. Jiang, M.Y. Zheng, X.G. Qiao, K. Wu, Q.Y. Peng, S.H. Yang, Y.H. Yuan, J.H. Luo, Microstructure and mechanical properties of WE43 magnesium alloy fabricated by direct-chill casting, Materials Science and Engineering: A 684 (2017) 158–164. 24. M. Jahedi, B.A. McWilliams, F.R. Kellogg, I.J. Beyerlein, M. Knezevic, Rate and temperature dependent deformation behavior of as-cast WE43 magnesium-rare earth alloy manufactured by direct-chill casting, Materials Science and Engineering: A 712 (2018) 50–64.

157 25. F.R. Elsayed, N. Hort, M.A. Salgado Ordorica, K.U. Kainer, Magnesium permanent mold castings optimization, Materials Science Forum, Trans Tech Publ, 2011, pp. 65–68. 26. M. Salgado-Ordorica, W. Punessen, S. Yi, J. Bohlen, K.-U. Kainer, N. Hort, Macrostructure evolution in directionally solidified Mg-RE alloys, Magnesium Technology 2011, Springer2011, pp. 113–118. 27. T.B. Massalski, H. Okamoto, P. Subramanian, L. Kacprzak, Binary Alloy Phase Diagrams; ASM International Materials Park; 1990. 28. P. Mengucci, G. Barucca, G. Riontino, D. Lussana, M. Massazza, R. Ferragut, E.H. Aly, Structure evolution of a WE43 Mg alloy submitted to different thermal treatments, Materials Science and Engineering: A 479(1–2) (2008) 37–44.

The Influence of Temperature and Medium on Corrosion Response of ZE41 and EZ33 M. AbdelGawad, A. U. Chaudhry, and B. Mansoor





Abstract

Keywords

Mg-based implants offer a promising alternative to commonly used permanent implants due to their biodegradability that eliminates the need for a follow-up surgery, along with the associated medical and economic risks. Several of the commercial Mg alloys for various applications including potential implant applications contain rare earth elements that are known to improve mechanical strength and corrosion resistance. However, it remains a significant challenge to better understand in vitro corrosion behavior of Mg–RE alloys and predict in vivo behavior, which is useful for biomedical applications, since in vitro corrosion rates tend to be significantly higher than those reported in vivo. In this work, we study the mechanical and corrosion behavior of two Mg–RE alloys, ZE41 and EZ33, at physiologically relevant temperature of 37 °C in 3.5 wt% NaCl and Hank’s solution. Tensile and compression tests were used to evaluate mechanical properties while electrochemical techniques were used to investigate the corrosion response. Both alloys demonstrated improved corrosion resistance in Hank’s solution which was attributed to the formation of a more protective surface film. In addition, the increased RE concentration positively impacted the corrosion behavior of EZ33 compared to ZE41 in both mediums.

Mg alloys Corrosion behavior EIS

M. AbdelGawad
A. U. Chaudhry
B. Mansoor (&) Mechanical Engineering Program, Texas A&M University at Qatar, Doha, Qatar e-mail: [email protected] M. AbdelGawad
B. Mansoor Mechanical Engineering Department, Texas A&M University, College Station, TX, USA B. Mansoor Materials Science and Engineering Program, Texas A&M University, College Station, TX, USA




Rare earths




Mechanical

Introduction The corrosion behavior of magnesium (Mg) and it alloys has been extensively studied due to the alloys recent popularity in different applications [1–7]. They are of great interest in the aerospace and automotive industries because of their low density and desired strength-to-weight ratio which can reduce emissions and save fuel and energy. They have also attracted attention as superior substitutes to metallic and polymeric biodegradable implants due to the closeness of their mechanical properties to those of the human cortical bone, biocompatibility with the human body and biodegradation which eliminates the need for an implant removal surgery [8–10]. However, research has indicated that the fast corrosion response of Mg alloys prevents their use to their full potential. Most Mg alloys are characterized as two-phase alloys due to the limited solubility of the alloying elements in Mg. These secondary phases have been reported to have a dual effect either by enhancing corrosion resistance and acting as a barrier or causing a galvanic effect with Mg matrix and accelerating corrosion [1, 3, 9, 11]. In common environments such as 3.5 wt% NaCl solution, Mg alloys develop a partially protective film made up of magnesium hydroxide which can provide some protection over a wide range of pH, as indicated by the Mg Pourbaix diagram [12]. Typically, corrosion occurs in the form of localized corrosion and initiates at the breaks within the developed surface film resulting in the production of Mg2+ ions and hydrogen gas [1]. For biomedical applications, in vitro corrosion behavior is characterized based on immersion at 37 °C in biological mediums, such as Hank’s solution, which have a chemical composition close to human blood plasma. Unfortunately, corrosion rates in vitro are higher than those reported in vivo

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_24

159

160

M. AbdelGawad et al.

[13], and therefore, it has been a challenge to better understand the corrosion mechanism of Mg alloys in physiological conditions. Currently, the most commonly used biodegradable metallic implants are Mg alloys that contain rare earths. Mg– RE alloys were primarily developed for the automotive and aerospace industries due to their high mechanical strength and creep resistance [14, 15]. ZE41 is a commercially available Mg–RE alloy whose corrosion characteristics have been studied previously in 3.5 wt% NaCl and Hank’s solution. EZ33 is another Mg–RE alloy that is available in the market but its corrosion behavior has not been reported as widely as ZE41 and has three times the RE concentration of ZE41. In our previous work, we investigated the effect of microstructure on the corrosion behavior of ZE41 and EZ33 in 3.5 wt% NaCl at room temperature [16]. We found that the increased volume fraction of REs in EZ33 resulted in the formation of a more stable protective film which improved corrosion resistance. Several studies on the corrosion of ZE41 in Hank’s solution are available in the literature; however, most of them compared its corrosion mechanism to aluminum-containing Mg alloys [17–20]. Therefore, our preliminary results on the effect of temperature and solution on the corrosion characteristics of ZE41 and EZ33 are presented in this paper. Both alloys were immersed in Hank’s solution and 3.5 wt% NaCl at 37 °C and compared to our previous results at room temperature. The influence of the increased RE content in EZ33 is studied to determine the impact of rare earths on corrosion behavior. Electrochemical techniques such as Open Circuit Potential (OCP), Electrochemical Impedance Spectroscopy (EIS) and Potentiodynamic Polarization (PD) were used to evaluate the corrosion response of both alloys in the different environments.

Experimental Methods Materials Table 1 presents the chemical composition of Mg alloys ZE41 and EZ33 that were cut from as-cast plates in T5 Table 1 Chemical compositions of ZE41A-T5 and EZ33A-T5

Element

condition. These alloys were chosen due to their commercial availability and the similarity in the main rare earth elements in them.

Microstructure Specimens were mounted in cold setting resin with an exposed surface of about 1 cm2 and grounded successively up to 1200 grit SiC paper. The samples were then polished using 3-µm diamond suspensions, washed with ethanol and dried with air. The microstructures were examined in the un-etched condition by scanning electron microscopy (SEM) equipped with backscatter electron (BSE) imaging.

Mechanical Properties To determine the effect of temperature on the mechanical properties of ZE41 and EZ33, tensile and compression tests were performed at ambient temperature and 37 °C. Tests were conducted at a strain rate of 10−3/s using electromechanical MTS Insight 30kN machine equipped with a three-zone furnace and an extensometer. Small size dogbane specimens with a gauge length of 11 mm and 2.85 mm 3 mm cross-sectional area were used for tension tests while samples of 3 mm diameter and 5.5 mm in length were used for compression testing.

Electrochemical Tests A typical three-electrode setup with Ag/AgCl as a reference electrode, a graphite rod as a working electrode and a specimen with a 1 cm2 exposed area as a working electrode was used. The electrochemical tests were carried out in a Gamry MultiportTM cell in which the volume of the electrolyte was kept at 800 ml. An insulated wire was attached to the back of the specimen and encapsulated in cold resin to provide an electrical connection. The exposed front surface of the sample was prepared using the same metallurgical Composition, wt% ZE41A-T5

EZ33A-T5

Zinc

4.16

2.66

Zirconium

0.71

1.32

Cerium

0.41

1.16

Lanthanum

0.12

0.47

Neodymium

0.11

0.26

Manganese

0.01

0.01

Magnesium

Remainder

Remainder

The Influence of Temperature and Medium on Corrosion Response …

procedure as was done to conduct microstructural analysis. The open circuit potential (OCP) was obtained for each alloy at the 3 different conditions: 3.5 wt% NaCl at room temperature, 3.5 wt% NaCl at 37 °C and Hank’s solution at 37 ° C and monitored throughout the total time of the experiment. Tests evaluating the effect of temperature using 3.5 wt% NaCl at ambient temperature and at 37 °C where monitored for 320 min while the experiments conducted to investigate the effect of electrolyte, i.e., using Hank’s solution and NaCl at 37 °C continued for 570 min. The main reason for prolonging the immersion time in the second experiment was because of the drastic difference in response that was observed which indicated a difference in surface film formation, as will be shown in the results section. Electrochemical Impedance Spectroscopy (EIS) was measured to gain insight on how temperature and electrolyte affected the corrosion mechanism by applying an AC amplitude of 10 mV and a range of frequencies between 105 and 10−2 Hz. The Gamry Echem Analyst software was used to fit the data from EIS and obtain the equivalent circuit.

161

identified the secondary phases using EDX analysis as mainly Mg7Zn3RE in ZE41 [21, 22]. Siebert-Timmer et al. also identified the secondary phases in EZ33 as MgZnxREy intermetallics [23]; however, various stoichiometric ratios have been reported for the prevalent secondary phases in the literature [15]. Wei et al. studied the solidification behaviors and phase constituents of different alloys in Mg–Zn–RE system and found the composition of the secondary phase, or T-phase, to vary depending on the alloy composition [24]. Therefore, as the wt% of RE increased and wt% of Zn decreased, the T-phase had a higher RE content and hence it is presumed that the secondary phase in EZ33 would contain more REs than ZE41. In our study, we observed that the larger wt% of REs in EZ33 resulted in a larger volume fraction of intermetallic phases; *7.0% compared to *5.0% found in ZE41, and the b-phase (which appears in the SEM as white regions) formed a more continuous network in EZ33 in contrast to the intermittent nature found in ZE41. In addition, the SEM images revealed particles within the grains surrounded by a halo-shaped zone of a lighter shade. Neil et al. identified these particles as Zr-rich intermetallics [25, 26].

Results and Discussion Microstructure Analysis

Mechanical Properties

Figure 1 shows SEM micrographs of ZE41 and EZ33 in T5 condition. Both alloys are made up of equiaxed a-Mg grains, with a grain size of *40 µm, and a secondary phase distributed along the grain boundaries as well as within the grains. Due to their difference in RE concentration, the composition and volume fraction of the intergranular intermetallic phase varied between both alloys. Zhao et al.

The tensile and compressive properties of ZE41 and EZ33 at ambient temperature and at 37 °C are shown in Fig. 2. Overall, ZE41 exhibited higher ultimate and yield strength during tension and compression at both temperatures compared to EZ33. As mentioned previously, RE-containing Mg alloys are widely used in transport industries because of their ability to considerably improve the mechanical

Fig. 1 Microstructure of ZE41A (left) and EZ33A (right) in T5 condition

162

M. AbdelGawad et al. 250 350

225

300

200

Stress (MPa)

Stress (MPa)

175 150 125 100 75

EZ33 (RT)

25

EZ33 (37°C) 0.02

0.04

0.06

0.08

150

ZE41 (RT) EZ33 (RT) ZE41 (37°C) EZ33 (37°C)

50

ZE41 (37°C)

0 0.00

200

100

ZE41 (RT)

50

250

0 0.00

0.10

0.05

0.10

0.15

0.20

0.25

0.30

Strain

Strain

Fig. 2 Stress–strain curves of ZE41 and EZ33 at room temperature (RT) and at 37 °C under tensile loading (left) compressive loading (right)

properties of Mg [9, 14]. The formation of secondary phases in such alloys acts as deterrents to dislocations and hence increasing their strength. Although EZ33 had a higher volume fraction of REs than ZE41, it had lower mechanical properties. This is due to the limited solubility of REs in Mg, in comparison with Zn. The solubility limit of Zn is about 2–6 times more than Ce, La and Nd; therefore, its strengthening effect is more dominant and results in ZE41 having superior mechanical properties due to the higher concentration of Zn. In EZ33, tensile properties were consistent between the two temperatures; however, a notable difference was found during the compression test. Although the stress–strain curves of both EZ33

specimens were analogous initially, the specimen at 37 °C seemed to have fractured in a brittle manner which could imply the presence of casting defects. In examining the effect of temperature, it can be observed that during both tension and compression tests, the difference in the response was not significant for both alloys. In tensile tests, there was a slight increase observed in stress levels for both alloys at 37 °C which is rather interesting; however, overall it was statically inconsequential and was not investigated in this study. Since the temperature difference between ambient and 37 °C is relatively minor, the behaviors of both alloys were deemed expected and largely in agreement with the literature [15, 27].

(a)

(b) -1550

-1600 -1650 -1700 -1750 -1800

ZE41 (NaCl RT) EZ33 (NaCl RT) ZE41 (NaCl 37°C) EZ33 (NaCl 37°C)

-1850 0

50

100

150

200

Time, min

250

300

Potential,mV (vs Ag/AgCl)

Potential,mV (vs Ag/AgCl)

-1550

-1600 -1650 -1700 -1750 -1800

ZE41 (Hanks 37°C) EZ33 (Hanks 37°C) ZE41 (NaCl 37°C) EZ33 (NaCl 37°C)

-1850 0

80

160

240

320

400

480

560

Time, min

Fig. 3 Open circuit potential (OCP) of ZE41 and EZ33 in: a 3.5 wt% NaCl at room temperature (RT) and 37 °C after 320 min of immersion and b in Hank’s solution and 3.5 wt% NaCl at 37 °C after 570 min of immersion

The Influence of Temperature and Medium on Corrosion Response …

Open Circuit Potential Open circuit potential (OCP) is usually used as a method for understanding the reactivity of the metal surface. Figure 3 presents the evolution of OCP of ZE41 and EZ33 at the different experimental conditions. In Fig. 3a, the effect of temperature can be observed throughout the total immersion time in 3.5 wt% NaCl while in Fig. 3b, the effect of changing the electrolyte is shown. In Fig. 3a, during the first few minutes of immersion, the OCP of ZE41 at room temperature seemed to have reached its maximum and then started to decrease as immersion time increased, which shows that the passive layer formed is unstable and is starting to saturate. When the temperature increased, the initial OCP values recorded were about 50 mV more negative proving that the surface layer was not as protective as when corrosion occurred at room temperature. Overall, the trend in the OCP of ZE41 did not experience much variability with respect to time for both temperatures. The variability is observed more when the electrolyte is changed from 3.5 wt% NaCl to Hank’s solution (Fig. 3b). In Hank’s solution, the OCP values for ZE41 ranged from around −1850 mV up to −1600 mV where a significant increase was shown at the start of immersion. Although this OCP value is the most negative value observed, the sharp increase is an indication of the increased stability of the film being formed, which did not exist during immersion in 3.5 wt% NaCl. However, the integrity of the film was not maintained since the OCP then reached a plateau at around 320 min before starting to decrease. Looking at both Fig. 3(a, b), EZ33 presented similar corrosion behaviors which seems independent of temperature and solution. However, throughout all operating

(a)

conditions, it is evident that EZ33 is nobler than ZE41 indicating the formation of a more stable surface film. The effect on the corrosion behavior due to the increase in temperature was not as substantial as the effect due to the change in corrosion medium. Mg alloys corrode aggressively in NaCl solutions due to the presence of chloride ions that are known to attack and break down the developed surface films [1]. The chloride ion concentration [Cl]− in Hank’s solution is about 0.14 M while in 3.5 wt% NaCl it is 0.6 M; therefore, it is expected to see a difference in the corrosion behavior of the alloys when immersed in both solutions.

EIS in 3.5% NaCl at Different Temperatures Electrochemical Impedance Spectroscopy (EIS) is a technique that uses AC polarization to investigate the processes occurring at the metal surface [28]. Figure 4 presents the EIS Nyquist plots for ZE41 and EZ33 immersed in 3.5 wt% NaCl at room temperature (RT) and 37 °C for 10 and 320 min. The plots were characterized by defined capacitive loops at high and medium frequencies and an inductive loop at low frequencies. As the immersion time increased, both alloys exhibited decreased corrosion resistance at both temperatures which confirmed similar findings in the OCP plots (Fig. 3), where a surface film was formed but was not stable and degraded. Although the effect of temperature was not quite evident in the OCP results, the Nyquist spectra show that the corrosion resistance of both alloys was higher at RT than at 37 °C and that was apparent by the higher impedance and larger capacitive loop. Zainal Abidin et al., calculated the corrosion rates of different Mg alloys, including ZE41, through hydrogen evolution and also

(b)

200

150

ZE41 (NaCl RT) EZ33 (NaCl RT) ZE41 (NaCl 37°C) EZ33 (NaCl 37°C)

150

100

Zimag, Ω.cm2

100

Zimag, Ω.cm2

163

50 0 -50

50

0

-50

-100 -150

ZE41 (NaCl RT) EZ33 (NaCl RT) ZE41 (NaCl 37°C) EZ33 (NaCl 37°C)

-100

-200 0

100

200

300

Zreal, Ω.cm2

400

500

600

0

50

100

150

200

250

300

350

Zreal, Ω.cm2

Fig. 4 Nyquist plots of ZE41 and EZ33 in 3.5 wt% NaCl at room temperature (RT) and 37 °C after a 10 min and b 320 min of immersion

164

M. AbdelGawad et al.

Fig. 5 Equivalent circuit used for the fitting of EIS data for 3.5 wt% NaCl at 37 °C

determined that increasing the temperature increased the corrosion rate due to increase in microgalvanic corrosion [18]. During the first 10 min of immersion, the corrosion resistance of the ZE41 at RT was the highest compared to EZ33 at RT and to both alloys at 37 °C. However, by the end of the immersion, the corrosion resistance of EZ33 was higher at both temperatures (Fig. 4b). The improvement in EZ33’s corrosion behavior was also observed during OCP with the slow, steady increase in contrast to the decreased or plateau behavior witnessed in ZE41. A similar result was also seen in potentiodynamic polarization (PD) results. The EIS spectra for ZE41 and EZ33 immersed in 3.5 wt% NaCl 37 °C were fitted using an equivalent electrical circuit shown in Fig. 5. The spectra for ZE41 and EZ33 immersed in 3.5 wt% NaCl at RT were previously fitted in [16]. The proposed circuit model was based on representing the characteristics of the metal surface by the higher frequency capacitive loop, the properties of the corrosion product layer by the capacitive loop in the medium frequencies and the inductance loop in the low frequencies attributed to the adsorption. In this equivalent circuit, Rs represented the resistance of the solution/electrolyte, R1 and CPE1 are in

(a)

Fig. 7 Equivalent circuit used for the fitting of EIS data for Hank’s solution at 37 °C

parallel and represented the resistance and capacitance of the corrosion product layer, and R2 and C2 are also in parallel and represented the charge transfer resistance and double layer capacitance, respectively. L was added to model the inductive behavior found at the low frequency range. Constant phase elements (CPE) are usually used instead of capacitances to describe a nonideal behavior due to different surface factors such geometry, roughness and inhomogeneity [29]. Echem Analyst was used to fit the EIS data using nonlinear least squares within an error of 10% and the fitted spectra were superimposed on the data plots in Fig. 4.

EIS at 37 °C in Different Electrolytes Figure 6 presents the Nyquist plots for ZE41 and EZ33 immersed in the two different solutions: 3.5 wt% NaCl and Hank’s solution at 37 °C. The significant difference in the spectra of the alloys when immersed in Hank’s solution versus when immersed in NaCl is an indication of the difference in reactivity between the metal’s surface and the electrolyte. Both alloys revealed improved corrosion resistance across the entire tested period when immersed in

(b)

500

500 400 400 300

200

Zimag, Ω.cm2

Zimag, Ω.cm2

300

100 0 -100 -200

-400 0

200

400

50

200

25 0 -25

100

-50 0

-300 800

1000

Zreal, Ω.cm2

1200

1400

1600

80

120

160

ZE41 (Hanks 37°C) EZ33 (Hanks 37°C) ZE41 (NaCl 37°C) EZ33 (NaCl 37°C)

-200

600

40

0 -100

ZE41 (Hanks 37°C) EZ33 (Hanks 37°C) ZE41 (NaCl 37°C) EZ33 (NaCl 37°C)

-300

75

0

200

400

600

800

1000

1200

1400

Zreal, Ω.cm2

Fig. 6 Nyquist plots of ZE41 and EZ33 in Hank’s solution and 3.5 wt% NaCl at 37 °C after a 10 min and b 570 min of immersion

1600

165

0.0

0.0

-0.5

-0.5

-1.0

-1.0

E, Volt vs Ag/AgCl

E, Volt vs Ag/AgCl

The Influence of Temperature and Medium on Corrosion Response …

-1.5 -2.0 -2.5 -3.0 -3.5 -4.0 1E-8

ZE41 (NaCl RT) EZ33 (NaCl RT) ZE41 (NaCl 37°C) EZ33 (NaCl 37°C) 1E-7

1E-6

1E-5

-1.5 -2.0 -2.5 -3.0 -3.5

1E-4

0.001

Current Density, A/cm

0.01

0.1

2

-4.0 1E-8

ZE41 (Hanks 37°C) EZ33 (Hanks 37°C) ZE41 (NaCl 37°C) EZ33 (NaCl 37°C) 1E-7

1E-6

1E-5

1E-4

0.001

Current Density, A/cm

0.01

0.1

2

Fig. 8 Potentiodynamic curves of ZE41 and EZ33 in: a 3.5 wt% NaCl at room temperature (RT) and 37 °C and b in Hank’s solution and 3.5 wt% NaCl at 37 °C

Hank’s solution which is evident with the higher impedance levels and broader inductive loops that imply the formation of a thicker corrosion layer. Zainal Abidin et al. showed that the corrosion rates for the RE Mg alloys immersed in Hank’s solution were lower than when immersed in NaCl at room temperature [17]. The improvement of corrosion resistance was ascribed to the thick layer of corrosion products that were more protective when compared to the products formed during NaCl immersion. In addition, the lower concentration of [Cl]− in Hank’s solution helped to maintain the integrity of the developed film, allowing it to thicken as immersion time increased. At the start of immersion (Fig. 6a), ZE41 performed better than EZ33 which was contrary to what was observed in 3.5 wt% NaCl. On the other hand, after 570 min (Fig. 6b), EZ33 presented much higher impedance values and hence lower corrosion rates when compared to ZE41. The change in the size of the capacitive and inductive loops indicated that the surface film on EZ33 is thickening and decreasing the diffusion across the layer resulting in improved surface protection [30]. An equivalent circuit (Fig. 7) was also developed to model the EIS data, and the fitted results were imposed on the data in Fig. 6. The circuit components are similar to the circuit developed for both alloys when immersed in 3.5 wt% NaCl at 37 °C, however, with the presence of RL which represented the resistance of absorbed species in the corrosion product layer.

Potentiodynamic Polarization Curves The potentiodynamic (PD) results for ZE41 and EZ33 at different experimental conditions are shown in Fig. 8.

Regardless of the temperature and corrosion medium, EZ33 still showed a more positive value than ZE41 which agreed with the OCP results. This was also in consistence with the ennobling effect of REs as presented by Birbilis et al. [31] where the corrosion of binary Mg–RE alloys were studied and compared to pure Mg. It was observed that as the volume fraction of REs increased, there was an increase in the cathodic reactions and a decrease in the anodic kinetics, which in turn enhanced the Ecorr values. A similar response was also seen in the PD results above. In Fig. 8a, the Ecorr values for the alloys at 37 °C were slightly negative than at RT, which is an indication of an increased corrosion rate. In Fig. 8b, EZ33 immersed in Hank’s solution at 37 °C had the most positive potential which agreed with Fig. 6b and proved the formation of the most stable protective film throughout all the experiments. Although the Ecorr for ZE41 in Hanks at 37 °C was more negative than in NaCl at 37 °C, the Icorr seems to be larger causing lower corrosion resistance, which is also consistent with Fig. 6b. Comparing Fig. 8a, b, the differences in the Ecorr values are greater when the solution was changed which is a similar trend to that observed in the OCP Fig. 3a, b. Our preliminary results on mechanical and corrosion response of Mg–RE alloys indicate that although mechanical responses are largely similar, the differences in film formation and more aggressive attack at physiologically important temperature of 37 °C warrants further investigation. We will continue our study by employing electron microscopy and XRD to examine the precise nature of surface films relevant to the two corrosive mediums to ascertain their break down mechanisms.

166

Conclusions The following are the conclusions of this work: 1. Tensile and compression tests showed that increasing the temperature to 37 °C did not have a significant difference on the mechanical properties of ZE41 and EZ33. As expected, ZE41 exhibited higher ultimate and yield strength values during tension and compression compared to EZ33. The limited solubility of the REs in comparison with Zn caused EZ33 to have lower mechanical properties since it contains a larger volume fraction of REs. 2. EZ33 presented a more positive OCP than ZE41 which was independent of the temperature and solution. This was attributed to the presence of REs that caused the secondary phases to have a more positive potential than the a-Mg matrix and therefore a lower corrosion rate. The same trend in potentials was seen in PD results. 3. Both alloys demonstrated lower corrosion resistance at 37 °C throughout the total immersion time. Improved corrosion resistance was observed when ZE41 and EZ33 were immersed in Hank’s solution, in contrast to immersion in 3.5 wt% NaCl. The presence of a larger concentration of chloride ions in NaCl is believed to attack the surface film forming on the alloys causing it to be unstable. Also, the higher impedance levels in the Nyquist plots show that the surface film formed in Hank’s solution is more stable than in NaCl. 4. Generally, ZE41 formed a stronger passive film than EZ33 at the start of immersion in both solutions. However, as time progressed, the film disintegrated as shown by the decreased impedance levels in the EIS spectra and reported values less than EZ33.

Acknowledgements This research was performed with support from the Qatar Foundation under the National Priorities Research Program grant# NPRP 8-856-2-364. The authors acknowledge this financial support with gratitude.

References 1. Song G, Atrens A. (2003) Understanding Magnesium Corrosion— A Framework for Improved Alloy Performance. Adv Eng Mater 5:837–858. https://doi.org/10.1002/adem.200310405 2. Song GL, Atrens A (1999) Corrosion Mechanisms of Magnesium Alloys. Adv Eng Mater 1:11–33. https://doi.org/10.1002/(sici) 1527-2648(199909)1:1%3c11::aid-adem11%3e3.0.co;2-n 3. Song G (2005) Recent progress in corrosion and protection of magnesium alloys. Adv Eng Mater 7:563–586. https://doi.org/10. 1002/adem.200500013 4. Atrens A, Song GL, Cao F, et al (2013) Advances in Mg corrosion and research suggestions. J Magnes Alloy 1:177–200. https://doi. org/10.1016/j.jma.2013.09.003

M. AbdelGawad et al. 5. King AD, Birbilis N, Scully JR (2014) Accurate electrochemical measurement of magnesium corrosion rates: A combined impedance, mass-loss and hydrogen collection study. Electrochim Acta 121:394–406. https://doi.org/10.1016/j.electacta. 2013.12.124 6. Gusieva K, Davies CHJ, Scully JR, Birbilis N (2015) Corrosion of magnesium alloys: the role of alloying. Int Mater Rev 60:169–194. https://doi.org/10.1179/1743280414y.0000000046 7. Esmaily M, Svensson JE, Fajardo S, et al (2017) Fundamentals and advances in magnesium alloy corrosion. Prog Mater Sci 89:92–193. https://doi.org/10.1016/j.pmatsci.2017.04.011 8. Zhao D, Witte F, Lu F, et al (2017) Current status on clinical applications of magnesium-based orthopaedic implants: A review from clinical translational perspective. Biomaterials 112:287–302. https://doi.org/10.1016/j.biomaterials.2016.10.017 9. Chen Y, Xu Z, Smith C, Sankar J (2014) Recent advances on the development of magnesium alloys for biodegradable implants. Acta Biomater 10:4561–4573. https://doi.org/10.1016/j.actbio. 2014.07.005 10. Agarwal S, Curtin J, Duffy B, Jaiswal S (2016) Biodegradable magnesium alloys for orthopaedic applications: A review on corrosion, biocompatibility and surface modifications. Mater Sci Eng C 68:948–963. https://doi.org/10.1016/j.msec.2016.06.020 11. Ding Y, Wen C, Hodgson P, Li Y (2014) Effects of alloying elements on the corrosion behavior and biocompatibility of biodegradable magnesium alloys: a review. J Mater Chem B 2:1912–1933. https://doi.org/10.1039/c3tb21746a 12. Pourbaix M (1974) Atlas of electrochemical equilibria in aqueous solutions. 2nd Eng. ed. by Marcel Pourbaix; translated from the French by James A. Franklin (except sections I, III 5, and III 6, which were originally written in English). Houston, Tex.: National Association of Corrosion Engineers, 1974 13. Sanchez AHM, Luthringer BJC, Feyerabend F, Willumeit R (2015) Mg and Mg alloys: How comparable are in vitro and in vivo corrosion rates? A review. Acta Biomater 13:16–31. https://doi.org/10.1016/j.actbio.2014.11.048 14. Tekumalla S, Seetharaman S, Almajid A, et al (2015) Mechanical Properties of Magnesium-Rare Earth Alloy Systems: A Review. Metals (Basel) 5:1–39. https://doi.org/10.3390/met5010001 15. Sediako D, Bichler L, van Hanegam M, Shook S (2013) Compressive Creep Properties of Wrought High Temperature Magnesium Alloys in Axial and Transverse Orientation—A Neutron Diffraction Study. Magnes Technol 2013 3–8. https:// doi.org/10.1002/9781118663004 16. AbdelGawad M, Mansoor B, Chaudhry AU (2018) Corrosion Characteristics of Two Rare Earth Containing Magnesium Alloys BT—Magnesium Technology 2018. In: Orlov D, Joshi V, Solanki KN, Neelameggham NR (eds). Springer International Publishing, Cham, pp 43–53 17. Zainal Abidin NI, Martin D, Atrens A (2011) Corrosion of high purity Mg, AZ91, ZE41 and Mg2Zn0.2Mn in Hank’s solution at room temperature. Corros Sci 53:862–872. https://doi.org/10.1016/ j.corsci.2010.10.008 18. Zainal Abidin NI, Atrens AD, Martin D, Atrens A (2011) Corrosion of high purity Mg, Mg2Zn0.2Mn, ZE41 and AZ91 in Hank’s solution at 37 °C. Corros Sci 53:3542–3556. https://doi. org/10.1016/j.corsci.2011.06.030 19. Taltavull C, Shi Z, Torres B, et al (2014) Influence of the chloride ion concentration on the corrosion of high-purity Mg, ZE41 and AZ91 in buffered Hank’s solution. J Mater Sci Mater Med 25:329– 345. https://doi.org/10.1007/s10856-013-5087-y 20. Johnston S, Shi Z, Atrens A (2015) The influence of pH on the corrosion rate of high-purity Mg, AZ91 and ZE41 in bicarbonate buffered Hanks’ solution. Corros Sci 101:182–192. https://doi.org/ 10.1016/j.corsci.2015.09.018

The Influence of Temperature and Medium on Corrosion Response … 21. Zhao MC, Liu M, Song GL, Atrens A (2008) Influence of pH and chloride ion concentration on the corrosion of Mg alloy ZE41. Corros Sci 50:3168–3178. https://doi.org/10.1016/j.corsci.2008. 08.023 22. Zhao MC, Liu M, Song GL, Atrens A (2008) Influence of microstructure on corrosion of As-cast ZE41. Adv Eng Mater 10:104–111. https://doi.org/10.1002/adem.200700246 23. Siebert-Timmer A, Fletcher M, Bichler L, Sediako D (2013) Creep performance of wrought AX30 and EZ33 magnesium alloys. Can Metall Q 52:430–438. https://doi.org/10.1179/1879139513y. 0000000069 24. Wei LY, Dunlop GL, Westengen H (1997) Solidification behaviour and phase constituents of cast Mg-Zn-misch metal alloys. J Mater Sci 32:3335–3340. https://doi.org/10.1023/a: 1018695927717 25. Neil WC, Forsyth M, Howlett PC, et al (2011) Corrosion of heat treated magnesium alloy ZE41. Corros Sci 53:3299–3308. https:// doi.org/10.1016/j.corsci.2011.06.005 26. Neil WC, Forsyth M, Howlett PC, et al (2009) Corrosion of magnesium alloy ZE41—The role of microstructural features.

167

27.

28.

29.

30.

31.

Corros Sci 51:387–394. https://doi.org/10.1016/j.corsci.2008.11. 005 Trojanová Z, Lukáč P (2005) Compressive deformation behaviour of magnesium alloys. J Mater Process Technol 162–163:416–421. https://doi.org/10.1016/j.jmatprotec.2005.02.024 Kirkland NT, Birbilis N, Staiger MP (2012) Assessing the corrosion of biodegradable magnesium implants: a critical review of current methodologies and their limitations. Acta Biomater 8:925–936. https://doi.org/10.1016/j.actbio.2011.11.014 Wang J, Jang Y, Wan G, et al (2016) Flow-induced corrosion of absorbable magnesium alloy: in-situ and real-time electrochemical study. Corros Sci 104:277–289. https://doi.org/10.1016/j.corsci. 2015.12.020 Zakiyuddin A, Lee K (2015) Effect of a small addition of zinc and manganese to Mg-Ca based alloys on degradation behavior in physiological media. J Alloys Compd 629:274–283. https://doi. org/10.1016/j.jallcom.2014.12.181 Birbilis N, Easton MA, Sudholz AD, et al (2009) On the corrosion of binary magnesium-rare earth alloys. Corros Sci 51:683–689. https://doi.org/10.1016/j.corsci.2008.12.012

Alloy Design Strategies of the Native Anti-corrosion Magnesium Alloy Tao Chen, Yuan Yuan, Jiajia Wu, Tingting Liu, Xianhua Chen, Aitao Tang, and Fusheng Pan

Abstract

Application of Mg alloy is limited because of its poor corrosion resistance. Due to low standard electrode potential of Mg, severe galvanic corrosion can happen if other alloyed elements form high electrode potential precipitate in Mg alloy. Moreover, natively formed oxide film on the surface of pure Mg is not compact and cannot hinder further oxidation of inner substrate. In this work, alloy design strategies are proposed to improve the native anti-corrosion property of Mg alloys. The first is to purify the Mg-melt by forming high-density precipitates in the settling process to increase the efficiency of the settling process. The second is to enclose extra impurities in harmless compounds to avoid the severe galvanic corrosion. The third is to control the composites of oxides formed on the surface by alloying defined elements to obtain passivate, close packing oxides film. Keywords

Mg alloy




Anti-corrosion




Alloy design

applications of Mg alloys are limited due to their low ductility and native poor anti-corrosion property [5]. The poor anti-corrosion property of Mg is attributed to two major reasons [6, 7]. One reason is the naturally formed magnesium oxide film on its surface is not protective for further oxidation of inner substrate. The Pilling-Bedworth ratio (P-B ratio) of pure magnesium is 0.81. The natively formed oxide film on Mg alloy is not compact and is not able to hinder further oxidation [7]. Another reason is Mg has the lowest hydrogen evolution potential among common metals and severe galvanic corrosion often happens with magnesium alloy when other elements were alloyed in the alloy. Meanwhile, solubility of many metals (e.g., Fe, Ni, Cr, Cu [8–13]) in magnesium is very low (in the order of magnitude of ppm of weight fraction). Therefore, the impurities in Mg alloy are easy to precipitate and result in galvanic corrosion as formed galvanic couples with Mg substrate phases. In the current work, three alloy design strategies of magnesium alloy are suggested to enhance the native corrosion resistance of Mg alloy.

Alloy Design Strategies Introduction Magnesium is the lightest metal material among common structural metals and consider being energy-efficient. Therefore, magnesium alloys have attracted extensive attentions in transportation field [1–4]. However, abroad

T. Chen
Y. Yuan (&)
J. Wu
T. Liu
X. Chen A. Tang
F. Pan College of Materials Science and Engineering, Chongqing University, Chongqing 400000, China e-mail: [email protected] T. Chen
Y. Yuan
J. Wu
T. Liu
X. Chen
A. Tang
F. Pan National Engineering Research Center for Magnesium Alloys, Chongqing University, Chongqing 400000, China

Purification of Mg-melt by Controlling Precipitates in Settling Process Fe, Ni, Cu, and Cr have small solubility in the primary a(Mg) phase thus can precipitate as Mg–X compounds during its solidification process and cause severe galvanic corrosion. Many publications and patents were devoted to decrease the content of these impurities [13–25]. Among these impurities, Fe has attracted more attentions [18, 23, 26] as it is easy to be picked up during the synthesis process. Among many purification methods, one approach is to let the precipitates settle down to the bottom. As illustrated in our previous work [27], the required fully settling time of precipitates is much longer than the usually applied settling process as half an hour by industry. Clearly, efficiency and

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_25

169

170

T. Chen et al.

Table 1 The terminal settling velocity of precipitates in Mg-melt Alloy

T (° C)

Sol. of Fe in L (ppm in w.f.)

Precipitated particles

Comp. of prec. (at.%)

Density of pre.

m (m/s)

Mg

660

180

Bcc(Fe)

Fe

7.86

4.53e−5

This work

AZ91 without Mn

760

62

Bcc(Fe, Al)

Al50Fe50

5.61

3.6e−5

[27]

730

36

Al2Fe

Al67Fe33

4.08

2.1e−5

660

8

Al5Fe2

Al5Fe2

3.96

1.7e−5

AZ91 + 0.1 wt% Mn

730

35

662

9

Bcc(Fe, Al)

Al50Fe50

5.61

3.4e−5

Al2Fe

Al67Fe33

4.08

1.8e−5

AZ91 + 0.3 wt% Mn

700 660

15

Al8Mn5

Al54Fe16Mn30

4.00

2.2e−5

8

Al8Mn5

Al54Fe16Mn30

Mg + Zr

700

2

Fe2Zr

Fe66Zr34

References

[27] This work

2.0e−5 7.65

4.9e−5

This work

Sol. of Fe in L (ppm in w.f.): Calculated solubility of Fe in Liquid (ppm in weight fraction) (The calculation is based on PanMg database) Comp. of Prec.: The composition of the precipitates (experimental measured results) m: The terminal settling velocity of the precipitates calculated using Stokes equation

effectiveness of the settling process is actually the determining factor of the final purity of the alloy. According to Stokes Eq. (1), terminal velocity of different precipitates during settling process were calculated. m¼

2gr(DqÞ 9g

ð1Þ

where Dq denotes the difference of the densities of the particles and of the fluid, η denotes the viscosity of the fluid, g is the gravitational acceleration (9.8 m/s2), and r is the radius of the spherical participation particles, which, according to our experimental measured results, is around 2 µm. Viscosity with temperature dependence of AZ91D magnesium alloy reported by Abaturov [28] is employed as a general data for Mg-melt in this calculation. As shown in Table 1 and Fig. 1, the addition of Mn does not change much on the solubility of Fe in the Mg-melt, but the precipitations on the defined temperature were changed. The terminal settling velocity relates to the density of the precipitate. With different precipitates, the terminal settling velocity can be quite different. Accordingly, the effectiveness of the settling process can be quite different in the limited settling period. As shown in Table 1, with 0.1 wt% Mn addition, the Bcc (Fe) with Al solution can be stabilized up to 730 °C. At 730 °C, the terminal settling velocity of Bcc(Fe, Al) is 3.4e−5 m/s, 1.6 times of that of Al2Fe. However, with more Mn addition, like 0.3 wt% Mn addition, the precipitation at 700 °C is changed to Al8Mn5 with Fe dissolved in Mn sublattice, which have low settling velocity again. It is consistent with the observation that too much Mn addition will deteriorate the anti-corrosion property of Magnesium alloy [29]. Calculated isopleth of AZ91 at 700 °C based on PanMg database is shown in Fig. 1.

According to Fig. 1, the suggested Mn addition is between 0.14 wt% and 0.27 wt% for AZ91 alloy. According to our calculation, terminal settling velocity of Fe2Zr is even bigger if Zr is added in Mg melt. However, Zr will react Al and form Al3Zr instead of Fe2Zr with Al presence. Hence, the Zr additive is not recommended for the purification of Mg–Al–X series alloy. Hence, by adding defined elements to control the type of the precipitates is an effective way to improve the efficiency of the settling process and, accordingly, improve the final purity level of the Mg alloy.

Fig. 1 Isothermal section of AZ91 with variation of Mn and Fe at 700 °C [27]

Alloy Design Strategy of the Native Anti-corrosion Magnesium Alloy

The Increasing of the Tolerance Limit of the Impurities by Controlling the Formed Precipitates in the Final Alloy Another cost-efficient way to decrease the galvanic corrosion is to increase the tolerance limit of the impurities in Mg alloy. It is observed the tolerance limit of Fe has a fixed ratio with the content of Mn in Mg alloy [8–11, 13]. The mechanism of Mn addition to increase the tolerance limit of Fe in Mg–Al–X alloy is illustrated in our previous work [27]. For Mg–Al–X alloy, addition of Mn can modify the precipitates in the final alloy and, therefore, change the distribution of Fe in different phases. When most of the extra Fe atoms are dissolved in compounds, which have less effect on the galvanic corrosion, no severe galvanic corrosion will happen to the Mg alloy. As reported in [27], the preferred formed phases in the alloys were Al–Mn compounds with Fe dissolved in the Mn sublattice place. Al–Fe compounds or other compounds where Fe as a main component should be avoid. An isothermal section of AM60 with variation of Fe and Mn at 500 °C, which is just below the solidify temperature, is shown in Fig. 2. An approximately phase boundary line as y = 0.022 * (x − 0.0070) is observed in Fig. 2. When Fe content in final alloy does not exceed the boundary line, there will be no Al–Fe compound formed. Accordingly, no severe galvanic corrosion can happen. This value 0.022 is quite close to the experimental obtained limiting ratio of Fe/Mn as 0.021 for AM60 alloy [8]. Considering possible experimental errors for such small value, the agreement between current calculation and experimental results is excellent. Therefore, this kind of calculation can guide search of other additional elements or conditions.

Fig. 2 Isothermal section of AM60 with variation of Fe and Mn at 500 °C [27]

171

It is suggested, by enclosing Fe or other impurities in some defined compounds when the impurities are not the main component, harm effect on the galvanic corrosion coming from these impurities can be reduced or even avoided. Hence, by controlling the final formed precipitates with different elements addition in Mg-alloy can control the galvanic corrosion of Mg-alloy.

Controlling Natively Formed Oxides Film on Surface of the Alloy When contacting with air, an oxide film will instantly form on the surface of Mg alloys. The PBR is 0.81 for Mg, less than 1. For general studies, it is suggested to add another element, which have a higher value of PBR, to obtain compact protective oxides film. In this case, the formed combination of this oxide and MgO could have a PBR value above 1. However, in our experimental observation [30], Mg–Ca alloy shows better anti-oxidation property than that of Mg–Gd alloy, where the P-B value of Ca is even small than that of Mg. A structure of difference size closing packing oxides (DS-CP) film is suggested in our work [30]. As shown in Fig. 3, only with small amount Ca addition (3 wt%) in the alloy, the fraction of finally formed CaO in the oxide film is up to 80% [30]. In addition, it is shown the fraction of CaO is increasing when it is closing to the inner substrate, which means Ca can diffuse fast to the surface and be oxidized. According to the formation energy of oxide, both Gd and Ca are more prone to be oxidized than Mg at high temperature. Hence, the oxidations and enrichments of Gd and Ca in the surface were observed. With both Ca and Gd added to

Fig. 3 Distribution of oxides in the surface of Mg–3.5Gd–0.5Ca (wt%) alloy by X-ray photoelectron spectroscopy (XPS) depth profiling [30]

172

Mg alloy, a phenomena of preferential oxidation of Ca and a competing diffusion kinetics between Ca ang Gd was observed [30]. With Ca addition to Mg–Gd alloy, Gd2O3 and enrichment of Gd at the surface were disappeared. Instead, the formation of CaO and enrichment of Ca at the surface was observed, as shown in Fig. 3. A mechanism of selective oxidation resulting in competing diffusion kinetics is suggested in our work. Based on this selective oxidation and competing diffusion kinetics mechanism, the anti-corrosion property of current commercial alloys can be improved by adding small amount defined diffuse fast elements.

Conclusion Three alloy design strategies for native anti-corrosion Mg alloy were proposed: (1) Purifying of the Mg-melt by forming high-density precipitates in the settling process; (2) Enclosing impurities in harmless compounds by controlling the formed precipitates in the final alloy; (3) Controlling the composites of oxides formed on the surface by alloying defined elements to obtain passivate, close packing oxides film. Based on the above three strategies, new designed Mg alloy with better corrosion resistance property can be expected. According to our experiments and calculation, it is observed that Mn has the both purification effect and enclosing effect for the impurities [27] and Mg–Ca alloy is passive at the high temperature [30]. According to above three strategies, a broad selective of anti-corrosion Mg–X (−Y) alloy and as well, the improvement of the current commercial alloy by alloying design can be performed. Acknowledgements The authors thank the fund from the National Key Research and Development Program of China with No. 2016YFB0301100.

References 1. Abbott TB (2015) Magnesium: Industrial and Research Developments over the last 15 years. Corrosion 71:120–127. https://doi. org/10.5006/1474 2. Mordike BL, Ebert T (2001) Magnesium Properties - applications potential. Mater Sci Eng A 302:37–45. https://doi.org/10.1016/ s0921-5093(00)01351-4 3. Wang TXJJ, Xu DK, Wu RZ, et al (2017) What is going on in magnesium alloys? J Mater Sci Technol 12–14. https://doi.org/10. 1016/j.jmst.2017.07.019 4. Joost WJ, Krajewski PE (2017) Towards magnesium alloys for high-volume automotive applications. Scr Mater 128:107–112. https://doi.org/10.1016/j.scriptamat.2016.07.035

T. Chen et al. 5. Brady MP, Joost WJ, Warren CD (2017) Insights from a recent meeting: Current status and future directions in magnesium corrosion research. Corrosion 73:452–462. https://doi.org/10. 5006/2255 6. Song G, Atrens A (2000) Corrosion Mechanisms of Magnesium Alloys. Adv Eng Mater 11–33. https://doi.org/10.1002/(sici)15272648(199909)1:1%3c11::aid-adem11%3e3.0.co;2-n 7. Esmaily M, Svensson JE, Fajardo S, et al (2017) Fundamentals and advances in magnesium alloy corrosion. Prog Mater Sci 89:92–193. https://doi.org/10.1016/j.pmatsci.2017.04.011 8. Hillis JE, Reichek KN (1986) 860288 High purity magnesium AM60 alloy: the critical contaminant limits and the salt water corrosion performance. SAE Tech Pap Ser 9. Hfllis JE, Shook S (1989) 890205 Composition and performance of an improved magnesium AS41 alloy. SAE Tech Pap Ser 10. Mercer WE, Hillis JE (1992) 920073 The Critical Contaminant Limits and Salt Water Corrosion Performance of Magnesium AE42 Alloy. SAE Tech Pap Ser. https://doi.org/10.4271/920073 11. Reichek KN, Clark KJ, Hillis JE (1985) 850417 Controlling the Salt Water Corrosion Performance of Magnesium AZ91 Alloy. SAE Tech Pap Ser. https://doi.org/10.4271/850417 12. Hillis J. (1983) 830523 The effects of heavy metal contamination on magnesium corrosion performance. SAE Tech Pap Ser 13. Liu M, Uggowitzer PJ, Schmutz P, Atrens A (2008) Calculated phase diagrams, iron tolerance limits, and corrosion of Mg-Al alloys. Jom 60:39–44. https://doi.org/10.1007/s11837-008-0164-2 14. Qiao Z, Shi Z, Hort N, et al (2012) Corrosion behaviour of a nominally high purity Mg ingot produced by permanent mould direct chill casting. Corros Sci 61:185–207. https://doi.org/10. 1016/j.corsci.2012.04.030 15. Chen X, Pan F, Mao J (2012) CN 102672148 A, Chinese Patent 16. Chen X, Yan T, Pan F, Mao J (2015) CN 104593612 A Chinese Patent 17. Pan F, Mao J, Chen X, et al (2015) CN 104630516 A Chinese Patent 18. Prasad A, Uggowitzer PJ, Shi Z, Atrens A (2012) Production of high purity magnesium alloys by melt purification with Zr. Adv Eng Mater 14:477–490. https://doi.org/10.1002/adem.201200054 19. Parthiban GT, Palaniswamy N, Sivan V (2009) Effect of manganese addition on anode characteristics of electrolytic magnesium. Anti-corrosion Methods Mater 56:79–83. https://doi. org/10.1108/00035590910940069 20. Matsubara H, Ichige Y, Fujita K, et al (2013) Effect of impurity Fe on corrosion behavior of AM50 and AM60 magnesium alloys. Corros Sci 66:203–210. https://doi.org/10.1016/j.corsci.2012.09. 021 21. Birbilis N, Williams G, Gusieva K, et al (2013) Poisoning the corrosion of magnesium. Electrochem commun 34:295–298. https://doi.org/10.1016/j.elecom.2013.07.021 22. Liu M, Song GL (2013) Impurity control and corrosion resistance of magnesium-aluminum alloy. Corros Sci 77:143–150. https:// doi.org/10.1016/j.corsci.2013.07.037 23. Pan F, Chen X, Yan T, et al (2016) A novel approach to melt purification of magnesium alloys. J Magnes Alloy 4:8–14. https:// doi.org/10.1016/j.jma.2016.02.003 24. Wu GH, Gao HT, Ding WJ, Zhu YP (2005) Study on mechanism of iron reduction in magnesium alloy melt. J Mater Sci 40:6175– 6180. https://doi.org/10.1007/s10853-005-3161-7 25. Chen X, Pan F, Mao J (2011) CN 102296184 A Chinese Patent 26. Gao H, Wu G, Ding W, et al (2004) Study on Fe reduction in AZ91 melt by B2O3. Mater Sci Eng A 368:311–317. https://doi. org/10.1016/j.msea.2003.11.017 27. Yuan Yuan, Jiajia Wu, Tao Chen, Tingting Liu, Dajian Li, Xianhua Chen, Aitao Tang, Fusheng Pan (2018) The CALPHAD investigation of the Mn effect on the melt purification and Fe

Alloy Design Strategy of the Native Anti-corrosion Magnesium Alloy tolerance limit in AZ and AM series of Magnesium alloy, submitted 28. Abaturov IS, Popel PS, Brodova IG, et al (2008) Exploration of the viscosity temperature dependences and microstructure of magnesium-based commercial alloy AZ91D with small additions of calcium. J Phys Conf Ser 98:6–10. https://doi.org/10.1088/ 1742-6596/98/6/062023

173 29. Simanjuntak S, Cavanaugh MK, Gandel DS, et al (2015) The Influence of Iron, Manganese, and Zirconium on the Corrosion of Magnesium : An Artificial Neural Network Approach. corrosion 71:199–208 30. Jiajia Wu, Xiaowen Yu, Dajian Li, Yuan Yuan, Bin Jiang, Fusheng Pan (2018) The study of high temperature oxidation behavior of Mg-Gd and Mg-Gd-Ca Alloys, submitted

Corrosion Bending Fatigue of RESOLOY® and WE43 Magnesium Alloy Wires Petra Maier, Adam Griebel, Matthias Jahn, Maximilian Bechly, Roman Menze, Benjamin Bittner, and Jeremy Schaffer

Abstract

RESOLOY®, a magnesium resorbable alloy based on Mg–Dy is the focus of this study. Corrosion bending fatigue behavior of RESOLOY wires was investigated with WE43 serving as a reference. Since these wires are developed for absorbable implant applications like stents, clips and anastomotic nails, circulating Ringer solution of 37 °C was used to simulate body conditions. The alloys were first extruded and finally cold-drawn to a wire diameter of 500 µm. Both alloys show very fine grains. The microstructure of WE43 was found homogeneous and equiaxed. RESOLOY shows recrystallized but non-equiaxed grains. RESOLOY is slightly harder than WE43. Both alloys were subjected to strain-controlled fatigue and corrosion fatigue in a sequence which mimicked stent crimping, expansion, and in-vessel cycling. Fatigue life was strongly influenced by corrosion. Fatigue data for RESOLOY highlight the need for further wire processing optimization work that is currently underway. Keywords








Magnesium RESOLOY WE43 Bending fatigue Corrosion fatigue




Corrosion




P. Maier (&)
M. Jahn
M. Bechly University of Applied Sciences Stralsund, Zur Schwedenschanze 15, D-18435 Stralsund, Germany e-mail: [email protected] A. Griebel
J. Schaffer Germany Fort Wayne Metals Research Products Corp., Fort Wayne, IN, USA R. Menze
B. Bittner MeKo Laserstrahl-Materialbearbeitungen e.K., Sarstedt, Germany

Introduction and Motivation As interest in absorbable metals continues to grow, there remains a need for both development and characterization of high-performance alloys. Magnesium alloys show significant promise and are already used clinically [1]. WE43 (Mg-4Y-3RE-0.4Zr), where RE is a mixture of rare earth elements containing primarily Nd, is a high-strength alloy with an acceptable biological response [2–4]. The intermetallic particles observed in this alloy are fully characterized [5–8]. Intermetallic particles Mg12Nd, Mg24Y5 and Mg14Y4Nd are observed and the mechanical properties of the alloy depend strongly on the volume fraction and spatial distribution of these intermetallics. Whereas WE43 was developed as a creep resistant alloy for aerospace applications, RESOLOY® was developed specifically for absorbable implants by MEKO Laser Materials Processing and the Helmholtz Center Geestacht in Germany and internationally patented [9]. RESOLOY is an alloy based on Mg–Dy with excellent strength and ductility. The corrosion rate depends on many factors, like grain size, thermomechanical history, and surface quality, but is relatively low [10]. Fort Wayne Metals, USA, having a strong expertise in the field of wire drawing for medical applications [11] provides RESOLOY in wire form with tensile strengths exceeding 500 MPa in the cold-drawn state. Additional annealing reduces strength, but increases ductility substantially. The high elasticity of RESOLOY provides excellent fatigue life in air [12]. In a study by Li et al. [13], a rapid ageing response of a Mg–Dy– Gd–Nd alloy has been found. In situ solidification experiments on Mg20Dy and Mg20Dy alloyed with Zr [14] determined Mg24Dy5 as secondary phases. In a study by Steinacker et al. [15], heat treatment on an extruded RESOLOY shows increased amount of LPSO phase with increasing heat-treatment time. Short-time heat treatment in an extruded bar reveals a negative effect on the degradation behavior of the alloy, whereas longer heat-treatment time leads to a similar degradation performance compared to the

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_26

175

176

as-extruded alloy. The heat treatment induced a finer distribution of intermetallics, which improved mechanical properties and acted as a barrier to the corrosion process. Biomedical applications require a moderate, homogenous corrosion rate to avoid strong hydrogen evolution and pitting corrosion. The corrosion morphology is often described by the pitting factor resulting from deepest corrosion pits, the area percentage of the corroded surface as well as the shape and size of corrosion pits, if present [16–18]. Corrosion pits, especially when deep and narrow, act as notches and can cause a strong reduction in the cross-sectional area and result in increased stress intensity under mechanical loading and should be avoided. Understanding the interactive effects of fatigue and corrosion is critical for absorbable implants which are designed to corrode while in the body. While some work has been reported in this area [19, 20], more investigations in clinically relevant materials, product forms, and loading schemes are needed.

Experimental Two magnesium alloys, WE43 and RESOLOY, have been first extruded and then drawn into the wires with a diameter of 0.5 mm. Cold-drawn wire samples were then annealed while in a fixture to “shape set” the samples to the initial apex test geometry (Fig. 1). Samples for metallographic investigations were prepared according to Kree et al. [21]. Vickers hardness was tested with a ZHU2.5 by Zwick with approximately ten indents under a load of 1 kg (9.807 N) and the average value is reported. Tensile tests are done with wires of a length of 127 mm and at a speed of speed 25.4 mm/min. Immersion tests, to evaluate the amount and form of corrosion, were performed using ten samples each, each wire had a length of 2 cm. The wires were exposed to 500 ml Ringer–Acetate solution at 37 ± 1 °C for 7 days.

Fig. 1 Deformation sequence of the samples during corrosion bending fatigue

P. Maier et al.

Corrosion bending fatigue tests were done according to the sequence shown in Fig. 1. Undeformed sample (distance between the epoxy clamps of the wire: 25 mm—wire was shape-set to a triangular apex shape) was loaded into the tester and compressed to distance x (−15 mm) to simulate the crimping of the stent. Afterwards, the sample was expanded to distance y (−7 mm) to simulate the balloon expansion of the stent. Before the cyclic test started, the sample was moved to distance z (−9 mm) to simulate the stent recoil after balloon deflation. The speed of these movements was 0.3 mm/s and every step followed the next immediately, apart from two samples each, where crimping lasted 24 h to see if the material relaxes during static loading. Cyclic loading to simulate the cyclic fatigue during stent life was here applied with an amplitude of 0.5, 1.0 and 2.0 mm at a stress of R = −1 and a frequency of 1 Hz. Up to five wires were tested in both air and Ringer-Acetate solution at 37 ± 1 °C. Samples were removed from the test setup very soon after either the limit in numbers of cycles (250,000) was reached or fracture took place. Scanning electron microscopy was performed using an EVO40 by Carl Zeiss to image fracture surfaces.

Results and Discussion Microstructure and Mechanical Properties: Both wires show a recrystallized, fine-grained microstructure, see Figs. 2 and 3. However, the microstructure of RESOLOY still shows clearly the wire drawing direction (horizontal), whereas the WE43 wires show homogeneous globular grains. The elongated grains in wire drawing direction in RESOLOY may be due to the blocking of grain growth in the transverse orientation by elongated Dy-rich regions. The

Fig. 2 Microstructure cross-section)

of

the

RESOLOY

wire

(longitudinal

Corrosion Bending Fatigue of RESOLOY® and WE43 Magnesium …

177

Fig. 5 RESOLOY wire after immersion of 7 days

Fig. 3 Microstructure of the WE43 wire (longitudinal cross-section)

Fig. 6 WE43 wire after immersion of 7 days

Fig. 7 Macrograph of corrosion attack of RESOLOY wire after immersion of 7 days (unetched)

Fig. 4 Stress–strain curve of the RESOLOY and the WE43 wire

hardness of the RESOLOY wire (78.9HV0.2 ± 4.5) was higher than that of the WE43 wire (68.8HV0.2 ± 2.34), which correlates to the higher strength of the RESOLOY wire (see Fig. 4). This may be attributable to increased solutions strengthening and a higher volume fraction of second phases in the RESOLOY. Figure 2 presents the secondary phases, which have been also observed in [15]. Figure 4 also shows increased ductility of the RESOLOY wire. Immersion: Figs. 5 and 6 show photographs of the wires after immersion over 7 days. The WE43 wire shows more local corrosion and an overall stronger corrosion attack, as indicated by the higher amount of white corrosion product Mg(OH)2 in Fig. 6. The RESOLOY wire corroded more uniformly, and interestingly lost its straightness during the immersion test (Fig. 5), which could be a sign of relieving residual stress. The morphology of the corrosion attack is shown in Figs. 7 and 8. In both cases, corrosion pits overlap and cause a rather homogenous corrosion; there are (apart from a very few cases) no deep and narrow or large undercutting pits. RESOLOY corrodes after 7 days up to 25 µm penetration

Fig. 8 Macrograph of corrosion attack of WE43 wire after immersion of 7 days (unetched)

Fig. 9 Wohler curve of RESOLOY wire: samples test in Ringer and air—two samples were crimped over 24 h

with small pits. The surface appears rather rough. WE43 shows corrosion proceeding as deep as 100 µm from the initial surface. The corrosion pits here are rather wide and shallow, which will not lead to a stress intensity increase when additional mechanical loading is applied. Corrosion Bending Fatigue: Figures. 9 and 10 show the Wohler curves of RESOLOY and WE43 wires; the number of cycles to fracture is given in dependence of strain amplitudes (0.5 and 1.0 mm, only RESOLOY was tested at

178

Fig. 10 Wohler curve of WE43 wire: samples test in Ringer and air— two samples were crimped over 24 h

2.0 mm). The bar chart in Fig. 11 shows the values (mean) of the cycles to fracture of the RESOLOY and WE43 wires in Ringer and in air at the two different amplitudes 0.5 and 1.0 mm. Samples with a triangle symbol in Fig. 8 and 9 have been tested in air, where both wires did not fail at a strain amplitude of 0.5 mm. RESOLOY wire in air at an amplitude of 1.0 mm performed much better than WE43 samples, which failed before reaching 8000 cycles compared to approximately 140,000 cycles for RESOLOY. The higher strength of RESOLOY (hardness of RESOLOY wires is 15% higher than of WE43 wires) may explain this increase in fatigue strength. In the presence of the Ringer corrosion media, the WE43 wires outperformed RESOLOY in the bending fatigue tests (Fig. 11). Even though the corrosion rate of RESOLOY during immersion was less than WE43, the corrosion surface of wires used in this study looks rather rough (perhaps due to the relatively inhomogeneous microstructure seen in Figs. 2 and 3). In this corrosion fatigue test, reducing the strain amplitude from 1 to 0.5 improves the fatigue life significantly in both alloys. WE43 performs better than RESOLOY at both strain amplitudes: At an amplitude of 1 mm WE43 reaches a mean value of 4350 cycles, three times higher than RESOLOY, and at 0.5 mm amplitude all but one WE43 wire sample reached the machine limit of 250,000 cycles while RESOLOY wires reached a mean value of only 65,000 cycles. It is important to note that in an actual stent

Fig. 11 Bar charts of number of cycles to fracture of RESOLOY and WE43 wires in Ringer and air at two different amplitudes: 0.5 and 1 mm

P. Maier et al.

application, the strain amplitude will likely be much smaller, and both wire materials will presumably perform much better. Samples with a diamond symbol in Figs. 9 and 10 have been crimped over 24 h and then expanded. Crimping over 24 h did not result in appreciable relaxation; the resulting stress has not changed and the number of cycles to fracture is not noticeably different. Figure 12 and 13 show representative wires, which failed under corrosion bending fatigue at 1 mm amplitude. The samples were removed from the setup and dried in air, but kept in their epoxy clamps. The RESOLOY sample (Fig. 12) reached 2710 cycles and the WE43 sample reached 2270 cycles at the same strain amplitude. Of note is the location of fracture of the two samples; the fracture occurred further from the apex in the RESOLOY. The yielding and work-hardening at the apex during crimping and expanding prior to the cyclic loading seems to have shifted the peak stress during fatigue to the point where work-hardening is reduced and residual stresses are less favorable. Figure 14 shows the fracture surface of a RESOLOY wire after a cycle number of 850 at a strain amplitude of

Fig. 12 Representing RESOLOY wire after corrosion bending fatigue test: off-center fracture

Fig. 13 Representing WE43 wire after corrosion bending fatigue test: off-center fracture

Corrosion Bending Fatigue of RESOLOY® and WE43 Magnesium …

179

Fig. 16 Macrograph of RESOLOY wire after corrosion bending fatigue test to cycle number of 900 cycles (strain amplitude 1 mm)

Fig. 14 SEM fracture surface of RESOLOY wire after corrosion bending fatigue test to a number of 850 cycles (strain amplitude of 1 mm)

1 mm. Assuming the thicker corrosion product (white regions) is the region of crack propagation and the grey area (which still has a corrosion layer from exposure to the media after fracture) is the area of the final fast fracture, the critical length for final catastrophic rupture is rather small. Figure 15 shows a higher magnification of the fracture surface in Fig. 14, revealing a more ductile fracture by features like dimples and plastic deformation. Cross-sectional macro and micrographs of tested wires (both fractured and unfractured) reveal the form of corrosion during corrosion bending fatigue. Figure 16 shows a broken

Fig. 15 SEM fracture surface of RESOLOY wire: higher magnification of Fig. 12

RESOLOY wire (900 cycles at 1 mm amplitude). The wire surface has transitioned from its bright metallic finish to a more dull matte surface, but does not show a high amount of local corrosion products, apart from the two areas (marked with the black arrow). One of these two areas with high local corrosion is where the sample ultimately failed. In Fig. 17, a WE43 wire (250,000 cycle runout at 0.5 mm amplitude) showed a large amount of corrosion products, but these did not lead to failure. The corrosion layer of RESOLOY seems to become porous and prone to the development of deep pits or cracks during corrosion fatigue (Fig. 18). Interestingly, only the intrados side of the apex is attacked by pitting. This would indicate that the displacement scheme outlined in Fig. 1 may result in tensile stresses, which drive pit growth, on this side. Figure 18 also reveals that these pits act as notches and crack initiation points—see propagated cracks starting from corrosion pits; the second longest is almost 240 µm long (Fig. 19). The fatigue crack propagation area is

Fig. 17 Macrograph of WE43 wire after corrosion bending fatigue test to cycle limit of 250,000 cycles (strain amplitude 0.5 mm)

180

P. Maier et al.

Fig. 18 Macrograph of RESOLOY wire after corrosion bending fatigue test to cycle number of 900 cycles (strain amplitude 1 mm)

Fig. 20 Macrograph of WE43 wire after corrosion bending fatigue test to cycle limit of 250,000 cycles (strain amplitude 0.5 mm)

Fig. 19 Micrograph of RESOLOY wire after corrosion bending fatigue test to cycle number of 900 cycles (strain amplitude 1 mm)

Fig. 21 Micrograph of WE43 wire after corrosion bending fatigue test to cycle limit of 250,000 cycles (strain amplitude 0.5 mm)

difficult to determine from the cross-sectional macrograph in Fig. 18. The macrograph of a WE43 wire after corrosion bending fatigue test, cycled to the limit of 250,000 cycles at a strain amplitude of 0.5 mm, is seen in Fig. 20. As seen in the RESOLOY wire, the corrosion pits, here much wider and shallower, are mostly at the inner side. The outside also shows corrosion, but more uniform. Figure 21 shows that none of these corrosion pits at the inside act as crack initiation points. The WE43 wire exhibits a similar corrosion morphology to the immersion test and is comparatively resistant to corrosion fatigue, perhaps due to the homogenously recrystallized microstructure. WE43 markedly outperformed RESOLOY in corrosion fatigue at 0.5 mm amplitude, but only marginally so at

1 mm amplitude. However, WE43 did not perform very well in air at an amplitude of 1 mm, and the addition of corrosion will only reduce performance. The strain-hardening behavior of both alloys at the displacements used in this study is worthy of further investigation. Because of the lower corrosion rate of RESOLOY in the immersion tests and its better fatigue behavior in air, the corrosion fatigue resistance may improve at lower strain amplitudes. In total, after all it can be seen that the combination of fatigue loading and corrosion in this study is acting worse on RESOLOY than on WE43, which leads to the need of improving the process route of wire drawing for RESOLOY. It also should be mentioned that Ringer solution is known to be slightly more aggressive compared to, for example, HBSS, DMEM, PBS or SBF [22].

Corrosion Bending Fatigue of RESOLOY® and WE43 Magnesium …

Summary In this study RESOLOY, a magnesium resorbable alloy based on Mg–Dy, and the commercial Mg-alloy WE43 were investigated in the form of cold-drawn wires. The microstructure, hardness, corrosion in immersion and corrosion bending fatigue behavior were investigated and data were compared to each other. Since these wires are developed for applications as biodegradable implants, like stents, clips and anastomotic nails, Ringer solution at a temperature of 37 °C was used. The alloys were first extruded and finally cold-drawn to a wire diameter of 500 µm. Both alloys show very fine recrystallized microstructure. The grains in WE43 were found more homogeneous and spherical. RESOLOY shows non-equiaxed grains. The microhardness of RESOLOY is slightly higher than in WE43. Immersion tests reveal relatively uniform corrosion in both alloys, with a higher corrosion rate in WE43. Neither wires exhibit deep and undercutting pits. The corroded surface of RESOLOY shows smaller, sharper pits, while the pits in WE43 are very wide. Both wires were compared by the number of cycles to failure during strain-controlled corrosion bending fatigue at different amplitudes. These amplitudes are much higher than a stent would see in real application, but should be useful in the screening of materials. In air, RESOLOY outperformed WE43 in fatigue. However, introduction of corrosion to the fatigue tests reduces lifetime in both alloys, but more so in RESOLOY, resulting in WE43 outperforming RESOLOY in corrosion fatigue. The tendency of RESOLOY to pit and crack in corrosion fatigue may be due to the relatively inhomogenous microstructure. The thermomechanical processing used to produce the test samples will likely need optimized to achieve a more uniform microstructure (work which is currently underway). Furthermore, the use of a higher number of samples will help to reduce the uncertainty of the results. In future work, it is anticipated that RESOLOY with improved microstructure and tested at smaller strain amplitudes will result in better fatigue life.

181 Acknowledgements The authors acknowledge the support from Hartmut Habeck (UASS) for the corrosion experiments.

References 1. Biotronik AG, www.biotronik.com/de-de/products/coronary/magmaris. 2. Y.F. Zheng, X.N. Gu and F. Witte, Mat. Sci. Eng. R, 77, 1 (2014). 3. L. Choudhary, R.K. Singh Raman, J. Hofstetter and P.J. Uggowitzer, Mater. Sci. Eng. C 42, 629–636 (2014). 4. Magnesium Elektron UK, data sheet 467. 5. D. Tolnai, C.L. Mendis, A. Stark, G. Szakács, B. Wiese, K.U. Kainer and N. Hort, Mater. Lett., 102–103, 62 (2013). 6. B. Smola, I. Stulı́ková, F. von Buch and B.L. Mordike, Mat. Sci. Eng. A, 324, 113 (2002). 7. L.L. Rokhlin, T.V. Dobatkina, N.I. Nikitina and I.E. Tarytina, Met. Sci. Heat Treat., 52, 588 (2011). 8. Y.H. Kang, D. Wu, R.S. Chen and E.H. Han, J. Magnesium Alloys 2, 109 (2014). 9. US patents: 9,566,367 B2 & 9,522,219 B2, EU patents: 2744531 & 2744532. 10. Meko Laser Materials Processing, www.meko.de/RESOLOY. 11. A.J. Griebel, J.E. Schaffer, T.M. Hopkins, A. Alghalayini, T. Mkorombindo, K.O. Ojo, Z. Xu, K.J. Little and S.K. Pixley, J Biomed. Mater. Research—Part B Applied Biomaterials, 106, 5, 1987–1997 (2018). 12. Fort Wayne Metals, www.fwmetals.com/RESOLOY-a-magnesiumalloy-for-absorbable-devices. 13. D. Li, J. Dong, X. Zeng, C. Lu and W. Ding, J Alloys and Compounds 439, 1–2, 254-257 (2007). 14. D. Tolnai, P. Staron, A. Staeck, H. Eckerlebe, N. Schel, M. Müller, J. Gröbner and N. Hort, Magnesium Technology, 17–21 (2016). 15. A Steinacker, CL Mendis, M Mohedano, F Feyerabend, M Stekker, P Maier, KU Kainer, N Hort, Cells and Materials 32, 6, 22 (2016). 16. P. Maier, G. Szakács, M. Wala and N. Hort, Magnesium Technology, 393–398 (2015). 17. P. Maier, J. Gonzalez, R. Peters, F. Feyerabend, T. Ebel and N. Hort, European Cells and Materials 32, 6, 22 (2016). 18. P. Maier, F. Zimmermann, M. Rinne, G. Szakács, N. Hort and C. Vogt, Corrosion and Materials 69, 2, 178–190 (2018). 19. A.J. Griebel and J.E Schaffer, Magnesium Technology, 303–307 (2015). 20. L. Choudhary and R.K. Singh Raman, Acta Biomater. 8, 916–923 (2012). 21. V. Kree, J. Bohlen, D. Letzig and K.U. Kainer, Practical Metallography 41, 5, 233–246 (2004). 22. P. Maier, L. Gentsch and N. Hort, Magnesium Technology, 429– 437 (2017).

Sacrificial Cathodic Protection of Mg Alloy AZ31B by an Mg–5Sn Surface Alloy C. F. Glover, T. W. Cain, and J. R. Scully

Abstract

A solid solution Mg–5Sn alloy is evaluated as a sacrificial anode for the cathodic protection of AZ31B. Uncoupled Mg–5Sn is shown to have superior barrier properties and reduced cathodic kinetics relative to AZ31B. The performance as a sacrificial anode was studied in situ with global and local measurements of galvanic coupling between the Mg–5Sn alloy and AZ31B when immersed in 0.6-M aqueous NaCl solution. The scanning vibrating electrode technique (SVET) was utilized to map the local current density distributions across the interface of the galvanic couple. Undesirable polarity reversal was observed during the initial 10 h of immersion, after which protection was offered. A self-corrosion rate of 52% was observed. Keywords

Magnesium




Cathodic protection




AZ31B




SVET

Introduction The high specific strength of Mg alloys makes them an attractive option for use in lightweight transport applications [1, 2]. The high susceptibility of Mg to corrosion is now well documented [3–7]. Water reduction (2H2O + 2e− ! H2 + 2OH−) is the primary cathodic reaction [8], so there is no requirement for oxygen and Mg is high reactive in aqueous environments [9]. Whilst alloying Mg has so far been the most common route to improved corrosion resistance [4, 10–15], developments on the protection of Mg via a sacrificial coating are limited. The aim of the current chapter is to determine the viability of an Mg–5Sn alloy for C. F. Glover (&)
T. W. Cain
J. R. Scully Center for Electrochemical Science and Engineering, Department of Materials Science and Engineering, University of Virginia, Charlottesville, VA 22904, USA e-mail: [email protected]

the sacrificial cathodic protection of AZ31B in chloride-containing, aqueous solutions. A viable sacrificial anode will dissolve preferentially and prevent detrimental polarity reversal by maintaining a galvanic couple potential that is consistently lower than that of the lowest open-circuit potential (OCP) of the metal in need of protection [9, 16]. There are a limited number of alloying elements that can produce a negative open-circuit potential with respect to an Mg alloy substrate, making it difficult to find an alloy that satisfies this requirement [9]. Sn has received significant attention [17–21] as an alternative alloying element to highly toxic As [12, 22, 23] in an attempt to improve the corrosion performance of Mg. It has been demonstrated that Mg–Sn alloys form a more tenacious passive film compared to that of pure Mg in chloride-containing aqueous environments [18, 19, 24–26]. A Sn-enriched surface layer that accumulates at the interface between the passive film and the underlying alloy has been reported [19, 25]. This Sn-enriched layer may provide sites of sluggish HER kinetics and reduce the fraction of sites available for fast HER on the matrix [19] in a similar manner to that reported in recent studies for both As [12, 22, 23] and Ge [27] alloying additions. Furthermore, Liu et al. found that anodically enhanced cathodic kinetics do not occur on pure Sn [17], and recent studies report low hydrogen evolution reaction (HER) exchange current density on Mg–Sn alloys [10, 17] suggesting that the detrimental effects of the negative difference effect (NDE) are limited [28], thus reducing self-corrosion which is an important factor in the longevity of a sacrificial anode. The performance of an Mg–5Sn alloy as sacrificial anodes for the protection of AZ31B is presented herein. Zero-resistance ammeter (ZRA) experiments and the in situ scanning vibrating electrode technique (SVET) were employed to evaluate net galvanic current densities and spatially localized anodic and cathodic current density across the couple interface, respectively. Previous studies have used SVET extensively to examine the corrosion mechanism of Mg under immersion conditions [29–35] and to study

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_27

183

184

C. F. Glover et al.

galvanic couples [34, 36]. Measurement of galvanic couple potentials was also performed. The work presented in this chapter aims to provide a comprehensive picture of the protection efficiency of a Mg–5Sn alloy, advancing the understanding of sacrificial anode systems for the protection of Mg alloys.

resistance associated with the anodic reaction (R3). The capacitor labelled C1 represents the charge separation at the sample surface of the oxide/hydroxide, and the double-layer capacitance is represented by C2. The relaxation of the coverage of adsorbed intermediates in actively corroding areas of the Mg surface is represented by the inductor (L). Rp was calculated via the following equation with values taken from the circuit:

Experimental Methods

1 1 1 ¼ þ Rp R 1 þ R 2 R3

Materials and Alloy Preparation High purity (HP) Mg was obtained from Alfa Aesar, Mg alloy AZ31B-H24 from Magnesium Elektron, and custom Mg–Sn alloys were prepared in-house. Table 1 gives the compositions of each alloy used in the investigated. Mg–Sn alloys were prepared with pellets of HP Mg, and 99.99% pure Sn flakes containing 0.85 wt% Y and dislocations in magnesium. Scripta Mater., 127:68–71, 2017. 22. Z. Wu and W.A. Curtin. The origins of high hardening and low ductility in magnesium. Nature, 526:62–67, 2015. 23. J.F. Stohr and J.P. Portier. Etude en microscopie electronique du glissement pyramidal {1122} dans le magnesium. Phil. Mag., 25:1313–1329, 1972. 24. T. Obara, H. Yoshinga, and S. Morozumi. {1122} slip system in Magnesium. Acta Metall., 21:845–853, 1973. 25. K.Y. Xie, Z. Alam, A. Caffee, and K.J. Hemker. Pyramidal I slip in c-axis compressed Mg single crystals. Scripta Mater., 112:75–78, 2016.

Part VII Magnesium Technology 2019: Poster Session

Forging of Mg–3Sn–2Ca–0.4Al Alloy Assisted by Its Processing Map and Validation Through Analytical Modeling K. P. Rao, K. Suresh, Y. V. R. K. Prasad, C. Dharmendra, and N. Hort




Abstract

Keywords

The processing map for hot working of cast Mg–3Sn– 2Ca–0.4Al (TXA320) alloy has been validated using forging experiments to form a rib–web (cup) shaped component. Finite-element (FE) simulation has also been conducted to obtain the strain and strain rate variations in the components as well as the load–stroke curves. TXA320 has been successfully forged under optimum processing conditions (450 °C at press speeds of 1 and 0.1 mm s−1) predicted by the processing map, where dynamic recrystallization (DRX) occurs. The microstructure obtained on these components revealed fully DRX grains and the average grain size has increased with increase in temperature. The load–stroke curves predicted by FE simulation of the forging process correlate well with experimental curves, although the simulated curves are slightly lower. Forging done in the flow instability regime of the processing map resulted in specimen fracture and the microstructure exhibited cracks at flow localization bands.

Mg alloy Processing map simulation

K. P. Rao (&) Department of Mechanical and Biomedical Engineering, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong e-mail: [email protected] K. Suresh Department of Physics, Bharathiar University, Coimbatore, India e-mail: [email protected] Y. V. R. K. Prasad Processingmaps.Com, Bengaluru, India e-mail: [email protected] C. Dharmendra Department of Mechanical Engineering, University of New Brunswick, Fredericton, Canada e-mail: [email protected] N. Hort Magnesium Innovation Centre, Helmholtz-Zentrum Geesthacht, Max-Planck-Str. 1, 21502 Geesthacht, Germany e-mail: [email protected]




Forging




Process

Introduction Magnesium alloys are used for structural components in automotive and aerospace industry due to their low density. The forming of these alloys at room temperature is challenging, because of its limited slip systems which lead to poor ductility [1]. Also, their strength and corrosion properties need improvements to make them attractive as structural materials. Commercial Mg alloys are based on systems such as Mg–Al–Zn, Mg–Al–Mn and Mg–Al–RE, but they have limited structural applications due to its inadequate corrosion and creep resistance [2, 3]. The recent Mg–Sn–Ca (TX) series of alloy are promising in overcoming some of these problems. Mg–3Sn–2Ca (TX32) alloy is found to possess optimum combination of creep strength (due to Ca addition) and corrosion resistance (due to Sn addition) [4], and the cast TX32 alloy also shows good hot workability behavior in a wide range of temperature and strain rates [5]. To further strengthen this TX32 alloy, Al addition has been attempted since Al causes solid solution strengthening of Mg [6]. Since Al addition is detrimental to the corrosion resistance, only minor additions are preferred which is the basis for the development of the alloy Mg–3Sn–2Ca–0.4Al (TXA320). This alloy also exhibits good hot workability domains in its processing map [7]. The aim of the present investigation is to validate the interpretations of various domains and regimes of the processing map developed earlier for the TXA320 alloy. The approach consists of using physical modeling technique involving laboratory-scale forging of a rib–web (cup) shape combined with analytical modeling involving finite-element (FE) simulation method. Forging tests have been conducted under temperature and

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_46

313

314

K. P. Rao et al.

strain rate conditions not only in the workability domains but also in the flow instability regime of the processing map.

Experimental Details Cylindrical billets of TXA320 alloy were prepared by melting elemental metals in the proportion (wt%) of 3Sn, 2Ca, 0.4Al and balance Mg under a protective cover of Ar + 3% SF6 gas and casting in a permanent mold. The detailed method of preparation was given earlier [7]. The suitable slices of the billets were used to prepare the cylindrical specimens with 12.5 mm diameter and 14 mm height with a 45° chamfer at the one end. Details of the specimen and die geometry along with the forging load train are given in an earlier paper [8]. The forging tests were conducted in a computer-controlled servo-hydraulic machine at temperatures of 300, 350, 400, 450, and 500 °C and at speeds of 0.01, 0.1, 1 and 10 mm s−1 to produce rib–web (cup) shape components. A thermocouple was inserted into the specimen near the chamfer side to measure the temperature rise during the forging process. The stroke was limited to about 11 mm and the corresponding load–stroke curves were recorded. The forging process was simulated using the FEM model embedded in the software program DEFORM, which was based on a code developed by Kobayashi et al. [9] considering rigid viscoplastic behavior. The principles and equations involved in this model are given in an earlier paper [8]. The program predicts the local values of stress, strain and strain rate as well as the load–stroke curves for the forging process. For metallographic investigations, the deformed forged components were sectioned in the center parallel to the forging axis. The cut surface was mounted, polished, and etched with an aqueous solution mainly containing 3-g picric acid. The microstructures were examined using optical microscope.

Results and Discussion The initial microstructure of the starting billet (unforged) of as-cast TXA320 alloy is shown in Fig. 1. The average grain diameter is about 250 lm. The microstructure exhibits two intermetallic phases: (Mg, Al)2Ca typically existing at the grain boundaries and Ca2SnMg particles appearing in the matrix. The existence of these two phases has been predicted by a thermodynamic model [10] and confirmed by detailed energy dispersive spectroscopy (EDS) analysis as reported in earlier publication [6].

Fig. 1 Microstructure of as-cast Mg–3Sn–2Ca–0.4Al alloy a optical and b SEM images

The processing map developed on TXA320 alloy at strain of 0.6 is shown in Fig. 2, which exhibits two domains of dynamic recrystallization (DRX) in the temperature and strain rate ranges of (1) 300–360 °C and 0.0003–0.001 s−1 (2) 400–500 °C and 0.005–0.7 s−1. At about 360 °C and 0.001 s−1, a changeover has occurred between the domains. Hot working of the alloy is best done under peak efficiency conditions in Domain 2, i.e. 450 °C and 1 or 0.1 mm s−1, where DRX occurs. Also, in view of higher temperatures, the forming loads will be lower. Conditions in Domain 1 may be selected for finishing operations since DRX occurring in this domain will result in fine-grained microstructures in the product. However, the workability in this domain will be lower than in Domain 2. The map also shows a large regime of flow instability at temperature range of 300–450 °C at higher strain rates and this regime will have to be avoided in

Forging of Mg–3Sn–2Ca–0.4Al Alloy Assisted by Its Processing …

315

Fig. 2 Processing map of cast Mg–3Sn–2Ca–0.4Al alloy at a strain of 0.6. Validation tests are done at conditions marked by and . Specimens at have fractured

Forging speed, mm s

-1

10

1

0.1

0.01

300

350

400

450

500

Temperature, oC Fig. 3 Geometry of TXA320 alloy forged specimens—top view

processing since it results in flow localization leading to highly inhomogeneous microstructure. The top view photographs of TXA320 components forged at different temperature and speeds are shown in Fig. 3.

It may be noticed from Fig. 3 that forgings done under conditions corresponding to Domain 2 (400–500 °C and 0.01–1 mm s−1) exhibited regular shaped sound components while those forged in the instability regime have fractured. FE simulation was conducted under isothermal conditions; i.e., the die and workpiece both have the same temperature. The process simulations were conducted at the range of 300– 500 °C and at speed range of 0.01–10 mm s−1 until the stroke reaches 11 mm in steps of 0.1 mm. As a typical example, the effective strain distribution obtained at the end of the stroke in the component forged at 450 °C and 1 mm s−1 is shown in Fig. 4. It has been observed from the simulation, the minimum and maximum effective strain range is about 0.14–3.97. The material flow patterns and the strain distributions obtained from the simulations conducted for the forging conditions of 400 °C and 500 °C at strain rates of 0.1 mm s−1 and 1 mm s−1 under Domain 2 of the processing map are largely similar. Typical load–stroke curves obtained from the simulations are shown in Fig. 5 for the forging conditions of 400 °C, 450 °C, and 500 °C at speeds of 0.1 and 1 mm s−1, along with

316

K. P. Rao et al.

Fig. 4 Strain contours obtained in FE simulation at the end of stroke (11 mm) obtained at temperature and speed of 450 °C and 1 mm s−1, respectively

Fig. 5 Comparison of load–stroke curves obtained at a 400 °C, b 450 °C (Domain 2), and c 500 °C at the speeds of 0.1 mms−1 and 1 mms−1 from the FE simulation and forging experiments

those obtained experimentally. The agreement between these two has been reasonable in all the cases although the simulated curves predict slightly lower values for the load. The discrepancy may be attributed to the contribution of some redundant work involved in forming this shape. The curves exhibit three stages typical of semi-close die forging process: (i) load increase until plastic flow initialization, (ii) material flow until the cup formation is complete and (iii) direct material compression in the bottom of the cup, resulting in a steep increase in the load with stroke. The optical microstructures obtained between the rib and web region of the components forged at 400, 450 and 500 °C at a speed of 1 mm s−1 are shown in Fig. 6a–c, which confirm the occurrence of DRX in Domain 2. The DRX microstructure is uniform at 450 °C and 1 mm s−1 (optimum forging conditions) while it is incomplete at 400 °C with grain growth occurring at 500 °C. The average grain diameters are about 22, 28, and 42 lm at forging temperatures of 400, 450 and 500 °C, respectively. Forging done at 350 °C and 10 mm s−1 falls in the instability regime of the processing map (Fig. 2). All the specimens forged in the instability regime exhibited shear fracture (Fig. 3) and the microstructure exhibited shear cracks associated with flow localization (Fig. 7a). The microstructure of the specimen forged just at the border of Domain 1 and flow instability regime (300 °C and 0.01 mm s−1) is shown in Fig. 7b, which exhibits fine grains (about 10 lm) confirming that Domain 1 also represents DRX. However, processing under conditions in this domain will have to be done by imposing limited strain followed by annealing as is generally practiced for finishing operations. Very large strains, if imposed in a single step, will lead to fracture of the specimen (Fig. 3).

Forging of Mg–3Sn–2Ca–0.4Al Alloy Assisted by Its Processing …

317

Fig. 7 Microstructures of TXA320 alloy forged at a 350 °C and 10 mm s−1 in the instability regime indicates shear cracks (shown by arrows), b 300 °C and 0.01 mm s−1 (border region of Domain 1)

Conclusions The hot forging behavior of Mg–3Sn–2Ca–0.4Al (TXA320) alloy was studied in the temperatures of 300–500 °C and at speeds of 0.01–10 mm s−1 with a view to validate the predictions of the processing map. The following conclusions are drawn from this study.

Fig. 6 Microstructures of TXA320 alloy forged at a 400 °C and 1 mm s−1, b 450 °C and 1 mms−1, c 500 °C and 1 mm s−1 (Domain 2). The forged axis is vertical

(1) Rib–web (cup-shaped) component was successfully forged from TXA320 alloy under optimum processing conditions predicted by the processing map, namely 450 °C at press speeds of 1 and 0.1 mm s−1, where DRX occurs.

318

(2) The load–stroke curves predicted by FE simulation of the forging process correlate well with experimental curves, although the simulated curves are slightly lower. (3) The microstructure obtained on the components forged in the higher-temperature domain revealed fully DRX grains and the average grain size has increased with increase in temperature. (4) Forging done in the flow instability regime of the processing map resulted in cracking of the specimen and the microstructure revealed cracks associated with flow localization, in accordance with the predictions of the processing map.

Acknowledgements This work was fully supported by a research grant (project #115108) from the Research Grants Council of the Hong Kong Special Administrative Region, China.

References 1. Kaiser F, Letzig D, Bohlen J, Styczynski A, Hartig Ch, Kainer KU (2003) Aniosotropic properties of magnesium sheet AZ31. Mater. Sci. Forum 419–422:315–320

K. P. Rao et al. 2. Luo AA (2004) Recent magnesium alloy development for elevated temperature applications. Int. Mater. Rev. 49(1):13–30 3. Pekguleryuz M, Celikin M (2010) Creep resistance in magnesium alloys. Int. Mater. Rev. 55(4):197–217 4. Hort N, Huang YD, Abu Leil T, Rao KP, Kainer KU (2011) Properties and processing of magnesium-tin-calcium alloys Kovove Mater. 49(3):163–177 5. Dharmendra C, Rao KP, Prasad YVRK, Hort N, Kainer KU (2013) Hot workability analysis with processing map and texture characteristics of as-cast TX32 magnesium alloy. J. Mater. Sci. 48 (15):5236–5246 6. Suresh K, Rao KP, Prasad YVRK, Hort N, Kainer KU (2013) Microstructure and mechanical properties of as-cast Mg–Sn–Ca alloys and effect of alloying elements. Trans. Nonferr. Met. Soc. China 23(12):3604–3610 7. Dharmendra C, Rao KP, Prasad YVRK, Hort N, Kainer KU (2012) Hot working mechanisms and texture development in Mg-3Sn-2Ca-0.4Al alloy. Mater. Che. Phy. 136(2–3):1081–1091 8. Rao KP, Prasad YVRK, Suresh K (2011) Materials modeling and simulation of isothermal forging of rolled AZ31B magnesium alloy: Anisotropy of flow. Mater. Des. 32(5):2545–2553 9. Kobayashi S, Oh S-I, Altan T (1989) Metal forming and the finite-element method. Oxford University Press, New York 10. Grobner J, Schmid-Fetzer R (2013) Key issues in a thermodynamic Mg alloy database. Metall. Mater. Trans. A 44(7):2918– 2934

Development of Manufacturing Processes for Magnesium Sheet A. Javaid and F. Czerwinski

Abstract

Although casting is still the dominant manufacturing process of magnesium components, to open new application opportunities a quality sheet as a semi-finished product is required. At present, applications for sheet have been limited due to the poor cold formability of magnesium combined with the perceived high cost of the rolled products. In this report, research activities oriented towards development of advanced magnesium sheet technology at CanmetMATERIALS are described. They cover twin-roll casting and solid-state rolling trials both conducted using the pilot-scale equipment. Results are presented for commercial alloys and experimental compositions with additions of rare earths, in-house designed and manufactured. The present challenges are outlined, including those caused by high affinity of magnesium to oxygen and their effect on hardware performance and quality of produced strips and sheets. Keywords

Magnesium sheet earth elements




Twin-roll casting




Rolling




Rare

Introduction Magnesium alloys, offering substantial weight reduction, are seen as attractive structural materials for many applications, including transportation vehicles. Due to unique solidification behaviour of high fluidity and low susceptibility to hydrogen porosity, casting is at present the process of choice for magnesium components, representing about 98% of its overall structural applications [1]. To expand magnesium use and manufacture wrought components with tight A. Javaid (&)
F. Czerwinski CanmetMATERIALS, Natural Resources Canada, 183 Longwood Rd South, Hamilton, ON L8P 0A5, Canada e-mail: [email protected]

dimensional tolerances, surface quality and optimal mechanical properties as that achieved for steel or aluminum, the technology of sheet metal forming is needed [2]. A sheet is a convenient raw material for downstream production of net-shape parts using a number of techniques, including laser or water jet cutting, stamping, bending, perforating, punching, curling, or roll forming. Although there are examples of successful sheet production, some properties of magnesium like mechanical anisotropy, relatively low absolute strength and poor formability at room temperature still hinder the large-scale manufacturing expansion. The low plastic flow properties of magnesium at room temperature are caused by its crystal anisotropy and hexagonal close packed structure offering limited number of active slip systems, primarily just involving the basal planes [1]. The critical resolved shear stress for basal plane slip in magnesium single crystal is about two orders of magnitude lower than that for non-basal plane slip involving prismatic or pyramidal planes so the distribution of the former plays an important role in determining its formability. As a result, deformation resistance along direction parallel to the basal plane is small. In contrast, deformation resistance in direction parallel to the prismatic plane is very large. Since during rolling deformation, very strong basal texture develops in magnesium sheet, modifying and lowering a contribution of the basal component is seen as the effective route to formability enhancement. To overcome fundamental obstacles with magnesium rolling, there are efforts focused on development of novel alloys with improved formability and development of sheet manufacturing technology. First, a better understanding of microstructure and texture development is necessary to design new wrought magnesium alloys. Then, formability modification may also be achieved through selecting deformation conditions as is the case of superplastic forming, equal channel angular rolling, cross-roll rolling, asymmetric,

© Her Majesty the Queen in Right of Canada, as represented by the Minister of Natural Resources 2018 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_47

319

320

or shear rolling. The program of NSERC Magnesium Network (MagNET) devoted to magnesium revealed a very different texture in the commercial alloy ZEK100 containing a small amount of rare earth element neodymium as compared to the strong basal texture of conventional alloys such as the AZ31 grade. Using the ZEK100 alloy, automotive parts were successfully produced at industrially acceptable forming rates and temperatures. In this report, research activities oriented towards development of advanced magnesium sheet technology at CanmetMATERIALS, involving twin-roll casting and solid-state rolling, are described.

Sheet Development via Twin-roll Casting Twin-roll continuous casting (TRC) is seen as the technique reducing problems caused by magnesium formability and allowing manufacturing sheet or strip directly from a molten state. The technique combines casting and hot rolling into a single step operation, thus decreasing the cycle time, energy, pollutant emissions, and final cost compared to traditional sheet production using a direct chill ingot casting. Moreover, high solidification rates, typical for TRC, have positive effects on the strip microstructure and properties resulting in fine and homogeneous grains with reduced chemical segregation, increased solid solubility and enhanced precipitate nucleation within the matrix. In contrast to ferrous and non-ferrous metal industries where for decades TRC found wide acceptance in producing cast strips of different thicknesses and widths, its application for magnesium strip production has proven difficult. As a cause, high affinity of magnesium to oxygen, its low specific heat and large

A. Javaid and F. Czerwinski

freezing ranges, which can cause formation of casting defects such as segregation, hot cracking, and oxide inclusions, are often cited. An optimization of TRC process parameters in a combination with alloy design led to a microstructure showing promising mechanical properties like high formability accompanied by acceptable strength.

TRC Facility A pilot-scale TRC facility, used for experiments with manufacturing of magnesium strip with its basic geometry, is shown in Fig. 1. The equipment can produce 250-mm-wide magnesium strip using 355-mm-diameter water-cooled casting rolls that are 610 mm wide. The system is instrumented with optical sensors and thermocouples for monitoring and temperature control in the furnace, headbox, nozzle inlet/outlet, roll surface, cooling water, and sheet exit. The TRC is also equipped with sensors for roll separating force, liquid alloy level in the head box, and rolling speed. The TRC has a fully integrated Human Machine Interface system that collects all the process information/data during strip casting.

Melt Temperature As a source of molten magnesium alloy, the casting line includes a 400-kg melting furnace and automated melt delivery system. The molten alloy temperature in the furnace is kept at 705 °C. Then, alloy is continuously transferred to the headbox with a temperature kept at 700 °C. The system

Fig. 1 The pilot-scale TRC facility at CanmetMATERIALS with 355-mm-diameter rolls and furnace of molten metal having 400 kg capacity

Development of Manufacturing Processes for Magnesium Sheet

is used to feed the nozzle, which supply alloy directly to twin rolls. Increasing melt temperature may prevent premature freezing of the alloy within the nozzle, which causes its breakage and process interruption. At the same time, however, higher temperature increases reaction of magnesium with oxygen and formation of oxide entrapments hindering flow. The temperature values were selected based on extensive experiments.

321

Since experiments with large pilot equipment are costly, modeling is used to provide quantitative basis for optimizing process variables and make trials more effective. An example of modeling temperature distribution within alloy injected from the nozzle towards the rolls is shown in Fig. 2b. Air pockets trapped in nozzle that travel along sides up to the nozzle-box interface persists throughout simulation. When casting speed was reduced, the solidification front moved upstream [3].

Rolls Feeding Challenges There are still unsolved challenges with hardware performance in harsh environment of highly reactive liquid magnesium. An example may be the ceramic nozzle and rollers, distributing liquid metal at the start of the process and adversely affecting the strip quality. To gain a better understanding of the impact of the nozzle design and operating parameters, a three-dimensional model of the TRC was constructed and multi-phase computations were carried out [3]. Key insights, to date, are that there is significant recirculation of the flow in the nozzle which could lead to insufficient melt supply to the gap between the rolls and contribute to the observed defects of the magnesium strip. This finding should help optimizing the nozzle design. The delivery of molten metal to casting rolls through ceramic nozzle/tip is shown in schematic of Fig. 2a.

Role of Magnesium Reactivity in Sheet Production via TRC Magnesium exhibits high reactivity with oxygen at increased temperatures, especially that corresponding to semisolid or liquid states. When exposed to an oxidizing atmosphere, it reacts rapidly leading to catastrophic oxidation, ignition, and fire. Solid-state oxidation kinetics of magnesium alloys depend on temperature and time. For Mg–9%Al alloy and air atmosphere, protective oxidation is valid at approximately 200 °C for several hours, while above 430 °C after 1 h, there is a catastrophic reaction [4]. At initial stage, the oxide has a form of nodules and with growing frequency of nodules nucleation it loses integrity. As opposed to solid-state reaction where oxide remains permanently on the metal surface, oxide formed initially on a surface of the liquid alloy is subsequently entrained into the alloy volume. The entrained

Fig. 2 Distribution of molten magnesium alloy to rolls during TRC: a delivery of molten metal to casting rolls through ceramic nozzle (called also tip) with a width of 250 mm; b modeling result of the temperature within the solidified alloy near the nozzle tip

322

A. Javaid and F. Czerwinski

oxides may reside inside a component as nonmetallic inclusions or serve during solidification as nucleation sites of pores and other casting defects, thus reducing final properties, especially the fatigue life.

Melt Protection in Furnace To prevent magnesium oxidation, particularly at temperatures of liquid states, two types of atmospheres are used in practice: inert and active. This is in addition to fluxes and vacuum [5]. A concept of melt protection by inert atmospheres relies on complete replacement of the oxidizing atmosphere with an inert gas. The most frequently used for this purpose is argon, and injection molding is an example of its full-scale industrial implementation. In general, however, a use of inert atmospheres for melt protection is limited due to the high vapor pressure of magnesium, which at its melting temperature of 650 °C reaches 360 MPa and is about 8 orders of magnitude higher than aluminum [1]. An inert gas does not prevent evaporation and when a free space is available above the molten alloy surface, evaporation leads to a condensation of the highly reactive gas that is prone to explosion upon contact

Fig. 3 General view of AZ31 strip with thickness of 5 mm, manufactured via TRC: a quality strip surface; b strip surface damages by magnesium oxidation; c high magnification view of surface oxidation, SEM; d MgO inclusions embedded on the 5-mm-thick strip surface, SEM

with air. As a result, an active protection with SF6 gas mixture was used in the TRC furnace.

Oxidation in Liquid State During Roll Feeding When molten alloy inside furnace is protected by a gas mixture, during its flow into rolls, ways of protection are limited. There are attempts to blow a protective gas around the nozzle to create a protective blanket and displace oxygen. Although there is a short time of possible melt contact with oxygen, some reaction takes place forming harmful oxides.

Oxidation of Hot Strip in Solid State There is an extended exposure to atmosphere of semisolid and solid alloy at temperature reaching solidus. As a result, strip which should have a clean surface (Fig. 3a) reacts with oxygen creating black oxide defects (Fig. 3b). If strip does not solidify completely before its free contact with oxygen, local liquid pools of magnesium alloy will be subjected to ignition and burning. In this case, holes throughout the strip thickness may be formed, as shown in Fig. 3c, d.

(a)

(b)

(c)

(d)

100 um

300 um

Development of Manufacturing Processes for Magnesium Sheet

323

Alloys Suitable Specifically for TRC

Challenge of Reactivity with Oxygen

Advantage of Rapid Solidification

The evident challenge of TRC is preventing reaction of magnesium with oxygen. Since ways of applying a protective atmosphere outside of the melting furnace are limited, alloys used should itself have a resistance against oxidation. By modifying alloy composition, the integrity and structure of oxide layers is changed to form a dense and protective surface scale after exposures to high temperature. There is a group of elements, which substantially reduce the alloy oxidation rate. It covers rare earths and other elements with high affinity to oxygen referred together as reactive elements such as Y, Ce, Hf, Gd, La, Zr, Ti, or Ca. The challenge here is that alloying additions preventing oxidation should not deteriorate the alloy castability, formability and mechanical properties.

The near-rapid solidification experienced by alloys during TRC requires the design of new alloys that take advantage of the increased amount of solute in the a-Mg solid solution, thus resulting in enhanced nucleation of precipitates within the matrix and their dense distribution. The features of microstructure modifications, related to rapid solidification, positively affect the strip formability at room temperature. At the same time, however, strip is subjected to solid-state deformation and also requires formability. There is a limited information in the literature since solidification characteristics are primarily studied for cast alloys and for wrought alloys, as-cast features do not attract such detailed attention. In our research, the solidification characteristics of wrought magnesium alloys containing minor additions of rare earth metal of neodymium were evaluated. The composition of alloys tested along with their measured liquidus temperatures is shown in Table 1. An example of melting and solidification characteristics of Mg–Nd alloys is shown in Fig. 4. Thermal analysis, performed during solidification, confirmed two major reactions: the formation of a-Mg dendrites followed by the eutectic transformation. There was a slight difference in both the liquidus and solidus temperature between alloys tested, aligned with growing Nd content. The cooling rate during solidification affected the microstructure refinement and a volume fraction of intermetallic precipitates. Increasing the cooling rate from 30 to 110 °C/s resulted in a reduction in dendrite arm spacing in Zr-free alloys from roughly 45 lm to 25 lm. For all alloys, an increase in the solidification rate was accompanied by a noticeable reduction in the number of intermetallic precipitates.

Sheet Production Through Solid-State Rolling

Another part of development of magnesium sheet focusses on solid-state rolling. The sheet metal rolling process consists of passing metal stock through one or more pairs of rolls to generate a flat product with a specific and uniform thickness. In addition to modifying hardware to change deformation characteristics during rolling, there are also efforts to assess the role of all subtle processing parameters. CanmetMATERIALS rolling facilities allow researching rolling conditions for a variety of advanced material including magnesium alloys. The pilot-scale rolling mill is a single stand reverse mill that can be configured in 2 high for hot rolling and 4 high for cold rolling under tension (Fig. 5a). It is driven by two 300-hp motors applying a maximum load of 500 tons. For low scale experiments, the laboratory is equipped with 50 tons Stanant reverse rolling mill with preheated rolls (Fig. 5b). The major portion of our research, involving experiments Table 1 Examples of alloys tested containing rare earth element of with large alloy volume, were focused on commercial neodymium and their liquidus temperature determined through thermal magnesium alloy ZEK100 (Mg–1.2Zn–0.35Zr–0.17Nd, in analysis wt%). As major processing parameters, the rolling temperAlloys Liquidus, °C ature and the effect of post-rolling heat treatment on sheet Mg–1Zn–0.5Nd 647 properties were examined. As-cast plates with thickness of Mg–1Zn–1Nd 646 25 mm were rolled to the final sheet thickness of 1.5– Mg–2Zn–1Nd 640 1.7 mm at temperatures up to 450 °C with a thickness reduction of 10–15% per pass. It was found that the tensile Mg–2Zn–2Nd 637 and compressive properties of the hot rolled sheet exceeded Mg–4Zn–1Nd 635 substantially that for the as-cast state with values strongly Mg–4Zn–2Nd 634 affected by the rolling temperature. For example, an increase ZEK100 648 in the rolling temperature from 350 to 450 °C caused

324

A. Javaid and F. Czerwinski

Fig. 4 Temperature versus time (a) and first derivatives of heating/cooling curves (b) of Mg-Zn-Nd alloys tested by UMSA. In heating/cooling plots, critical points are marked as: liquidus—c, d solidus—a, f and eutectics—b, e

reduction in tensile strength from 257 to 228 MPa accompanied by a reduction in tensile yield stress from 237 to 185 MPa (Fig. 6). At the same time, the alloy elongation increased from 17 to 21%. For comparison, after casting tensile strength was of the order of 175 MPA, yield stress below 65 MPa and elongation reached 13%. The rolling temperature, being an essential process parameter used to control the magnesium formability, had effect on microstructural refinement. For example, increasing temperature in the range from 350 to 450 °C caused grain

coarsening with the highest growth seen for the temperature range of 400–450 °C. The correlation between the alloy grain size after rolling and the sheet properties was established where a reduction in grain size was accompanied by an increase in both the tensile/compressive strength and yield stress. At the same time, a reduction in alloy grain size was accompanied by a reduction in sheet elongation. Based on Hall–Petch relationship and an average grain size, the correlation developed may be used as the prediction model for properties of the hot rolled magnesium sheet.

Development of Manufacturing Processes for Magnesium Sheet

325

Fig. 5 Pilot-scale rolling mills at CanmetMaterials: a 500 tons; b 50 tons

Fig. 6 Tensile properties of ZEK100 sheet rolled at different temperatures. (R-rolling direction, T-transverse direction)

Alloy Development for Improved Formability In addition to development of the modern rolling process, there is a quest for novel magnesium alloys with improved formability. In this search, an experimental work and modeling based on thermodynamic calculations are used to provide a deeper understanding of the complex chemistry of magnesium alloys that will assist in the design of new alloys for sheet production. Although there is a universal objective of improved formability for the successful candidate, alloys tailored specifically for TRC should also take advantage of unique solidification characteristics of this technology. In our research to find optimal alloying elements, the primary interest was focused on rare earths. In this respect,

neodymium was selected with contents earlier shown in Table 1. A combination of calculations using the FACTsage software and examinations using a number of experimental techniques was explored to determine the solidification characteristics of wrought magnesium alloys containing rare earth metal of neodymium [6]. Under equilibrium solidification conditions, the FACTsage software predicted for ZEK100, Mg–1Zn–0.5Nd, and Mg–1Zn–1Nd alloys the same phases of a-Mg, Mg12Zn13 and Nd5Mg41. During nonequilibrium solidification, an additional phase of Mg51Zn10 was predicted by the FACTsage software to form, mainly at the expense of Mg12Zn13. FACTsage calculations of phases present under nonequilibrium (Scheil) solidification conditions in Mg–2Zn–1Nd alloy are shown in Fig. 7.

326

A. Javaid and F. Czerwinski

Magnesium alloys with small additions of rare earth element neodymium show to be promising material for sheet production through TRC and solid-state rolling. Increasing Nd content up to 2 wt% in Mg–Zn–Nd alloys caused increasing MgxNdy content while the content of Mg12Zn13 remained unchanged. The rolled sheet from this alloy family exhibited tensile strength up to 260 MPa, yield stress up to 240 MPa, elongation up to 20% and very low anisotropy of properties. As measured by compressive tests, the anisotropy between rolling and transverse direction was in the range of 0–5% for the rolling temperature range up to 450 °C. Acknowledgements The authors thank the Innovative Casting team at CanmetMATERIALS for support during equipment modernizing and experimental work. Fig. 7 FactSage calculations of phases present under nonequilibrium (Scheil) solidification conditions in Mg–2Zn–1Nd

References Summary Metal rolling is one of the most important manufacturing processes in the modern industry, and therefore, its expansion to magnesium alloys is of paramount interest. Although magnesium has poor room temperature formability due to its crystallographic structure, there are efforts to overcome this barrier in order to produce magnesium sheet at the industrial scale. CanmetMATERIALS with its state-of-the-art pilot-scale casting, twin-roll casting and rolling facilities is well suited to contribute to the progress in development of low cost magnesium strip and sheet. To achieve this, we focus on both the development of the modern process of twin-roll casting/rolling and new magnesium alloys with improved formability at room temperature.

1. F. Czerwinski, Magnesium Injection Molding, New York: Springer, 2008. 2. K. Kainer, R. Hoppe, J. Bohlen, G. Kurz, S. Yi and D. Letzig, “Challenges and solutions in the development of magnesium sheet for Sustainable vehicle concepts,” Materials Science Forum, Vols. 828–829, pp. 15–22, 2015. 3. A. Javaid, J. Hanke, C. Simha and M. Kozdras, “Twin Roll Casting of Magnesium Strip at Canmet Materials—Modeling and Experiments,” in Magnesium Technology 2015. Springer, Cham, TMS, pp. 461–464, 2015. 4. F. Czerwinski, “The oxidation behavior of an AZ91D magnesium alloy at high temperatures,” Acta Materialia, vol. 50, pp. 2639– 2654, 2002. 5. F. Czerwinski, “The reactive element effect on high temperature oxidation of the magnesium,” International Materials Review, vol. 60, pp. 264–296, 2015. 6. A. Javaid, F. Czerwinski, R. Zavadil, M. Aniolek and A. Hadadzadeh, “Solidification characteristics of wrought magnesium alloys containing rare earth metals,” in Magnesium Technology 2014, TMS, pp. 197–202, 2014.

Part VIII TMS-DGM Symposium on Lightweight Metals: Magnesium

Incorporating an ICME Approach into Die-Cast Magnesium Alloy Component Design J. P. Weiler

Abstract

The work presented here summarizes a methodology to incorporate an integrated computational materials engineering (ICME) approach into die-cast magnesium alloy component design. The framework of this approach was developed through process structure–property relationships developed through both Meridian’s internal and published studies.




Keywords





ICME Mechanical properties Process-structure-property predictions

Die-casting

The use of magnesium alloy die-castings for structural automotive applications is currently increasing, while these components, such as closure inners and instrument panel beams, are being designed with more geometrical complexity at decreasing nominal cast thicknesses. An accurate representation of the material behavior in computer-aided engineering (CAE) simulations is an important factor in the design process of these castings. Meridian has previously published [1] a work describing the development of a robust material model for die-cast magnesium alloy AM60B. The work presented here, also published in the Journal of The Minerals, Metals & Materials Society (TMS) [2], demonstrates robustness validation of the conservative failure criteria through correlated bench component testing. At the time of this work, the authors acknowledge that further investigations are required to advance the understanding of die-cast magnesium properties and, more specifically, relationships between microstructure and these properties. The work presented here summarizes the development and validation of an approach that incorporates variability and robustness into CAE performance simulations of die-cast magnesium alloy castings through use of casting process simulations and J. P. Weiler (&) Meridian Lightweight Technologies, Strathroy, Ontario, Canada e-mail: [email protected]

relationships developed between process, structure and properties by Meridian Lightweight Technologies. Meridian’s integrated computational materials engineering (ICME) approach was developed through several studies focusing on the magnesium die-casting process and casting simulations, the microstructure of die-cast magnesium material, the resulting mechanical properties, and the relationships between each. Several principal process–structure– property relationships were identified to ultimately predict local mechanical properties. The variability in local mechanical properties observed in different casting and throughout different regions in a large, complex die-cast component due to variations in local filling and solidification conditions are accounted for by these relationships. An approach was formulated to incorporate these process– structure–property relationships into the die-cast component product design process. The ICME approach summarized here uses proprietary results from casting simulations of the filling and solidification processes of a die-cast component as inputs into a Meridian-developed open-source-coded algorithm. This algorithm creates an input file for CAE simulations that automatically calculates and maps local mechanical properties into a mesh. In this manner, each element in a CAE simulation of a complex die-cast component can possess a unique material response based upon the simulation local processing history. Meridian has performed extensive coupon and component correlation to validate the ICME methodology described. One case study example is demonstrated here. A CAE input file with mapped local mechanical properties was created for an instrument panel beam cast component utilizing the casting filling and solidification simulations and the algorithm developed here. A CAE simulation was performed subjected the instrument panel beam to a tensile loading mode, simulating the experimental bench-testing performed on die-cast components. The results of the simulation using this methodology demonstrate that the failure location and average force–displacement response correlate with the experimental results in all tested components. Further, the

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_48

329

330

failure displacement prediction using this methodology in CAE is a statistically conservative performance prediction compared to the experimental data; normally distributed exhibiting a 99% reliability with 99% confidence interval. The results summarized in this case study provide an example of the accuracy and robustness of this ICME methodology. In conclusion, it is expected that this method can help improve the product design process of magnesium alloy die-castings to further optimize material usage.

J. P. Weiler

References 1. Regis Alain, Tony Lawson, Pat Katool, Gerry Wang, John Jekl, Richard Berkmortel, Len Miller, Jorild Svalestuen, and Hakon Westengen, SAE Technical Paper: 2004–01-0131, (2004), https:// doi.org/10.4271/2004-01-0131 2. J.P. Weiler, JOM (2018) 70:2338, https://doi.org/10.1007/s11837018-2985-y

Influences of SiC Particle Additions on the Grain Refinement of Mg–Zn Alloys Yuanding Huang, Jiang Gu, Sihang You, Karl Ulrich Kainer, and Norbert Hort

Abstract

A homogeneous microstructure of as-cast magnesium alloys is desired to improve their mechanical properties when achieving lightweighting. Recently, it was demonstrated that the addition of SiC refines both Mg–Al and Mg–Zn alloys. The present work investigates the effect of SiC particle additions on the grain refinement of Mg–Zn alloys, including their addition amount, particle size, addition temperature and holding time. The microstructures were characterized using XRD, SEM and EDS. It was found that the addition of SiC particles refines the grains of Mg–Zn alloys. With increasing their amount and reducing the addition temperature and holding time, the grain size decreases. The optimal SiC particle size for nucleation of alpha-Mg was found to be around 2 µm. The responsible refinement mechanism is attributed to the formation of (Mn, Si)-enriched intermetallics by the interactions between SiC and impurity Mn in alloys. Keywords

Magnesium–Zinc alloys Grain refinement




Microstructure




Solidification

Introduction Due to its HCP crystal structure, magnesium has a poor room temperature formability. Previous investigations indicate that alloying with rare earths elements and microstructural refinement can improve its formability effectively. Microstructural refinement not only increases its ductility but also its yield strength. Adding foreign particles into the melt prior to casting named as inoculation treatment is a popular approach to refine the as-cast microstructure. It Y. Huang (&)
J. Gu
S. You
K. U. Kainer
N. Hort MagIC-Magnesium Innovation Centre, Helmholtz-Zentrum Geesthacht, Max-Planck-Str. 1, 21502 Geesthacht, Germany e-mail: [email protected]

promotes the heterogeneous nucleation rate depending on their nucleation potency, size and distribution [1, 2]. Mg alloys can be generally classified into two broad groups: Al-free and Al-bearing. Most of the newly developed grain refiners such as Al2Y [3], AlN [4], Al2Ca [5], CaO [6, 7] for Al-bearing Mg alloys are not as efficient as the commercially available ones, such as Zr in Al-free Mg alloys. However, there are still several limitations for Zr addition in Al-free Mg alloys, such as Zr master alloys are expensive and difficult to be added into the melt [8]. In addition, Zr is only available for non-Al/Mn/Si/Fe-containing Mg alloys [9]. Recently, it was found that SiC, which shows potential in commercial applications in Al-bearing Mg alloys due to its environmental-friendly, low-cost and relatively good grain refining effect [10, 11], cannot only refine the as-cast Al-bearing Mg alloys but also the as-cast Al-free Mg alloys, such as Mg–Zn binary alloys [12, 13]. Huang et al. reported that the main mechanism responsible for grain refinement of Mg–Al alloys caused by SiC inoculation is due to the formation of Al2MgC2 by the reaction of SiC with Al and Mg [11]. This Al2MgC2 has a very close crystal structure to that of magnesium. However, the role mechanisms of SiC particles responsible for the grain refinement of Mg–Zn alloys still remain unclear. Recent investigations indicated that the possible grain refinement mechanisms should be related to the formation of Mn-containing particles [14]. After the addition of SiC in Mg–Zn alloys, these particles were often detected inside the grains, which were formed by the interaction of SiC with those impurities such as Mn and Fe. The results demonstrate Mn plays an important role in the grain refinement of Mg–Zn alloys inoculated by SiC particles. The present work further investigated the effects of SiC particle additions on the grain refinement of Mg–Zn alloys, including their addition amount, particle size, addition temperature and holding time. These parameters were optimized based on microstructural observations. The related mechanisms were discussed.

© The Minerals, Metals & Materials Society 2019 V. Joshi et al. (eds.), Magnesium Technology 2019, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-030-05789-3_49

331

332

Y. Huang et al.

Experimental Procedures

(EDS) was also used to observe the microstructures of selected samples.

Mg–3Zn alloys inoculated with different contents of SiC were prepared using commercial purity Mg or high purity Mg and Zn (Table 1). SiC particles (supplied by Alfa Aesar GmbH & Co KG, Germany) with different sizes (an average size of 0.1–10 µm) were used as refiner. Mg and Zn ingots were melted at 700 °C in an electrical resistance furnace using a mild steel crucible under a protective gas mixture of high pure Ar + 0.2% SF6. The melt was manually stirred for 2 min, and then its surface was skimmed. SiC particles preheated to 500 °C under Ar atmosphere were added into the melt directly. After that the melt was stirred vigorously at 100 rpm for 5 min to ensure a good dispersion of SiC particles. Before casting the melt was held at 700 °C for 15 min. Each ingot was cast by pouring the melt into a steel mold preheated to 200 °C with a diameter of 70 mm at the bottom and 80 mm at the top and a height of 250 mm. In order to investigate the effects of SiC amount, SiC particle size, its addition temperature and holding time on microstructure, various alloys with different SiC particle sizes and amount and under different addition temperatures and holding time were prepared (Table 2). Metallographic samples were transversally sectioned from the same position of 20 mm from the bottom of the castings. They were prepared according to a standard procedure. The samples for optical observations were etched with a solution of 8 g picric acid, 5 ml acetic acid, 10 ml distilled water and 100 ml ethanol. The average grain size was measured by the linear intercept method from the micrographs taken using polarized light in an optical microscope. A Zeiss Ultra 55 (Carl Zeiss GmbH, Oberkochen, Germany) scanning electron microscope (SEM) equipped with energy dispersive spectroscopy Table 1 Measured chemical compositions of source materials (wt%)

Table 2 Detailed preparation parameters used for Mg–3Zn alloys

Results Addition Amount Figure 1 shows the changes in average grain size of as-cast Mg–3Zn alloy inoculated with different contents of SiC. It clearly shows a slight decrease of average grain size from 330 ± 12 to 285 ± 10 lm with the increase of SiC content from 0 to 0.2%. Then, a sharp decrease of average grain size from 285 ± 10 to 180 ± 9 lm was obtained with increasing SiC content from 0.2–0.3%. After that, the average grain size remains relatively stable with the increase of SiC content, even up to 10% SiC. Figure 2 shows the optical micrographs of as-cast Mg– 3Zn–0.2SiC, Mg–3Zn–0.5SiC and Mg–3Zn–10SiC alloy. Homogeneous equiaxed grain morphologies were obtained in both Mg–3Zn–0.5SiC and Mg–3Zn–10SiC alloys. Compared with Mg–3Zn–0.2SiC (Fig. 2a) and Mg–3Zn–0.5SiC (Fig. 2b) alloy, SiC clusters that marked by black arrows in Fig. 2c were observed in Mg–3Zn–10SiC alloy, indicating that SiC was not well inoculated in Mg–Zn alloy with high content of SiC. Conversely, SiC can be added into the melt well and distributed uniformly within the matrix when its content is less than 0.5%. Figure 3 is a typical BSE image of as-cast Mg–3Zn– 10SiC alloy and its corresponding EDS mapping analysis. Alloying element Zn segregates at the dendritic and grain boundaries. Round-shaped Mg7Zn3 particles were also detected inside Mg grains. EDS mapping clearly shows that SiC clusters aggregated at the grain boundaries. This result

Alloys

Zn

Al

Zr

Mn

Fe

Cu

Ca

Ni

Mg

Mg

0.003

0.006