Magnesium alloys and their applications 3527302824, 9783527302826, 3883552550, 352729936X
In the recent years there has been a dramatic increase in research activity and also applications of magnesium alloys. T
395 40 47MB
English Pages 816 Year 2000
Cover Page......Page 1
Further Titles of Interest:......Page 3
Title: Magnesium Alloys and their Applications......Page 4
ISBN 3527302824......Page 5
Preface......Page 6
Alloy Development......Page 7
Texture and Microstructure......Page 9
Magnesium Matrix Composites......Page 10
Mechanical Development......Page 11
Corrosion and Surface Treatment......Page 13
Processing......Page 15
Creep Behaviour......Page 17
Subject Index......Page 18
Alloy Development......Page 19
2 Die Casting......Page 21
3 Thixomolding®......Page 23
4 Critical Success Factors for Growth......Page 24
5 Observations and Conclusions......Page 25
6 References......Page 26
2 Database Development......Page 27
3 Alloy Selection in the Mg-Mn-Sc-Gd System......Page 29
6 References......Page 31
1 Introduction......Page 32
1 High Performance Alloys......Page 33
2 Commercial Alloys......Page 35
3 Practical Foundry Considerations......Page 38
5 References......Page 39
1 Introduction......Page 41
2 Method of Investigation......Page 42
3 Material Properties......Page 43
5 Acknowledgements......Page 45
6 Literature......Page 46
1 Introduction......Page 47
2 Preparation of Test Material and Corrosion Testing......Page 48
3 Mechanical Properties and Castability......Page 49
5 Conclusion......Page 50
6 References......Page 51
4 Results and Discussion......Page 53
5 Conclusion......Page 57
7 References......Page 58
2 Introduction......Page 59
4 Modified Wrought Alloys Containing Zinc......Page 60
5 New Wrought Alloys Containing Lithium......Page 61
8 References......Page 63
2 Experimental Procedures......Page 65
3 Results and Discussion......Page 66
4 Conclusions......Page 69
5 References......Page 70
1 Introduction......Page 71
3 Results and Discussion......Page 72
4 Conclusions......Page 75
5 References......Page 76
2 Experimental Procedure......Page 77
3 Results and Discussion......Page 78
5 References......Page 82
1 Introduction......Page 83
2 Materials and Experimental Procedure......Page 84
3 Results......Page 85
4 Discussion......Page 88
6 References:......Page 89
2 Introduction......Page 91
3 Experimental Details......Page 92
4 Results......Page 93
5 Discussion......Page 94
6 Conclusion......Page 95
7 References......Page 96
2 Thermodynamic Calculation of the Ternary System Al-Mg-Sc......Page 97
3 Thermodynamic Calculation of the Ternary System Al-Mg-Gd......Page 99
6 References......Page 101
2 Experimental Investigation......Page 102
3 Thermodynamic Modeling......Page 104
6 References......Page 105
2 Experimental Investigation......Page 106
3 Thermodynamic Modeling......Page 108
6 References......Page 109
2 Results and Discussion......Page 110
3 Conclusions......Page 114
5 References......Page 115
2 Thermodynamic Calculations......Page 116
3 Experimental......Page 118
5 Conclusion......Page 119
7 References......Page 120
2 Experimental Procedure......Page 121
3 Results and Discussion......Page 122
4 Conclusion......Page 126
6 References......Page 127
2 Experimental Procedure......Page 128
3 Results and Discussion......Page 129
6 References......Page 133
Texture and Microstructure......Page 135
1 Introduction......Page 137
2.1 Mechanical Test Results......Page 138
4 References......Page 142
3 Sample Preparation......Page 143
4 Experimental Details of Sample Characterization......Page 144
5 Results......Page 145
7 References......Page 148
2 Experimental Procedure......Page 149
3 Results and Discussion......Page 150
5 References......Page 154
2 Experimental and Results......Page 155
4 Discussion......Page 158
6 References......Page 159
2 Experimental Methods......Page 161
3 Results and Discussion......Page 162
5 Acknowledgments......Page 165
6 References......Page 166
2 Experimental Procedure......Page 167
3 Results and Discussion......Page 168
4 Conclusions......Page 171
6 References......Page 172
2 Introduction......Page 173
4 Simulation Models......Page 174
5 Results and Discussion......Page 176
7 References......Page 178
3 Materials and Experimental Methods......Page 179
4 Results and Discussion......Page 180
6 Acknowledgements......Page 184
7 References......Page 185
Magnesium Applications in Aerospace and Electronic Industries......Page 186
Joining......Page 191
2 Introduction......Page 193
3 State of the Art Friction Stir Welding Tools......Page 194
4 Probe and Shoulder Design of Whorl™ Friction Stir Welding Tools......Page 195
5 Workpiece Materials......Page 197
6 Weld Quality......Page 198
7 Microstructure of Friction Stir Welds in Magnesium......Page 200
8 Friction Stir Welded Aluminium Tanks in the Aerospace Industry......Page 202
9 Friction Stir Welded Aluminium Panels in the Shipbuilding Industry......Page 203
11 Friction Stir Welded Hollow Aluminium Panels in the Rolling Stock Industry......Page 205
12 Commercially Available Powerstir™ and SuperStir™ Machines......Page 206
14 References......Page 207
3 Basic Observation......Page 209
4 TheMechanism of Pore Formation......Page 211
5 The influence of the Welding Parameters......Page 212
6 Reduced Porosity Compared to Die-casted Aluminium......Page 213
7 Influence of the Plate Thickness......Page 214
8 Other Pore ReductionMethods......Page 215
10 Acknowledgment......Page 216
11 References......Page 217
3 Inert-gas Metal-arc Welding (MIG)......Page 218
4 Metallographic Investigations and Weld Profile......Page 220
5 Strength of MIG-welded Magnesium Alloys......Page 221
7 References......Page 222
Magnesium Matrix Composites......Page 223
2 Inserts......Page 225
3 Embedding of Inserts......Page 227
4 Modeling......Page 228
6 Conclusions......Page 230
8 References......Page 231
2 Introduction......Page 233
3 Experimental......Page 234
4 Results and Discussion......Page 235
6 Acknowledgements......Page 237
7 References......Page 238
2 Introduction......Page 239
3 Experimental Procedures......Page 240
4 Results and Discussion......Page 241
5 Conclusions......Page 245
6 Reference......Page 246
2 Experimental Procedure......Page 247
3 Experimental Results and Discussion......Page 248
6 References......Page 251
1 Introduction......Page 252
2 Experimental Materials and Procedures......Page 253
3 Results and Discussion......Page 254
6 References......Page 256
1 Introduction......Page 258
3 Chemical Reactions Between Preform and Magnesium......Page 259
4 Structure Investigations......Page 260
6 Creep-Behaviour of Magnesium Matrix Composites......Page 261
8 Literature......Page 262
1 Introduction......Page 264
3 Results and Discussion......Page 265
6 References......Page 268
1 Introduction......Page 270
3 Investigation Results......Page 271
5 References......Page 274
2 Experimental Procedure......Page 275
3 Result and Discussions......Page 276
5 References......Page 280
2 Materials......Page 281
3 Dynamic Compression Tests......Page 282
4 Fractography......Page 284
6 References......Page 285
Mechanical Development......Page 287
2 Electrochemical Behavior......Page 289
4 Corrosion Protection......Page 290
5 References......Page 291
2 Experimental Method......Page 292
3 Results......Page 293
4 Discussion......Page 295
5 Conclusions......Page 296
6 References......Page 297
2 Experimental Procedure......Page 298
3 Results and Discussion......Page 299
5 Acknowledgments......Page 301
6 References......Page 302
2 Experimental Procedure......Page 303
3 Results and Discussion......Page 304
5 References......Page 308
2 Feedstock Material......Page 309
3 Heating Operation......Page 310
5 Testing Operation......Page 311
6 Results......Page 312
8 References......Page 313
2 Determination of Material Properties......Page 314
3 Verification Tests......Page 317
4 Summary......Page 320
5 References......Page 321
2 Materials and Mechanical Properties......Page 322
3 Fatigue Test Results and Comparison with other Materials......Page 324
4 Conclusions......Page 328
5 Acknowledgement......Page 329
2 Introduction......Page 330
4 Results and Discussion......Page 331
6 References......Page 335
2 Precipitate Phases in Alloy WE54......Page 336
3 Precipitates in Alloy AZ91......Page 338
4 References......Page 341
2 Material Behaviour at High Strain Rates and Temperatures......Page 342
3 Constitutive Material Law for Dynamic Loading......Page 343
4 Influence of Strain Rate and Temperature on Ductility of MgAl8Zn......Page 345
6 References......Page 347
3 Experimental Details......Page 348
4 Results and Discussion......Page 350
7 References......Page 353
2 Introduction......Page 354
3 Experimental Procedure......Page 355
4 Results and Discussion......Page 356
5 Conclusions......Page 358
6 References......Page 359
2 Superplasticity......Page 360
3 Experimental Procedure......Page 361
4 Results......Page 363
6 References......Page 365
2 Experimental......Page 366
3 Results and Discussion......Page 367
5 Acknowledgements......Page 370
6 References......Page 371
2 Experimental Procedure......Page 372
3 Results and Discussion......Page 373
5 References......Page 376
1 Introduction......Page 377
2 Characterization of the Deformation and Damage Behavior......Page 378
3 Component Tests and Simulations......Page 381
5 References......Page 382
1 Extended Abstract......Page 383
2 USCAR Program......Page 384
3 EUCAR Chassis Program......Page 385
Application......Page 389
2 Basic Knowledge......Page 391
3 Drilling......Page 392
4 Compression of Dry Chips......Page 395
5 Summary / Outlook......Page 396
6 Literature......Page 397
2 Introduction......Page 398
3 Preliminary Design, Material Selection and FEM Analyses......Page 399
4 Prototype Casting Process Development......Page 400
5 Prototype Static, Impact and Fatigue Tests......Page 402
6 Summary......Page 403
7 References......Page 404
2 The Starting Point......Page 405
5 Selection of Development Partners......Page 406
6 Selected Development Priorities......Page 407
8 Outlook......Page 408
2. Wheel Engineering......Page 409
3. Alloy Selection......Page 410
4. Manufacturing......Page 411
5. Conclusion......Page 413
Acknowledgement......Page 414
3 Cost and Weight Saving......Page 415
6 References......Page 419
Cast Magnesium Alloys For Wide Application......Page 420
2 Cutting Experiments......Page 424
3 Burnishing Experiments......Page 427
5 Acknowledgement......Page 428
6 References......Page 429
1 Introduction......Page 430
2 Mechanisms of Tool Wear......Page 431
3 ToolWear......Page 432
4 Surface Topography and Subsurface Zone......Page 434
5 References......Page 435
3 Effects on Properties......Page 436
5 References......Page 440
Corrosion and Surface Treatment......Page 441
2 Experimental Results......Page 443
3 Discussion......Page 447
6 References......Page 448
2 Experimental......Page 450
3 Results......Page 451
5 Conclusion......Page 455
6 References......Page 456
1 Introduction......Page 457
3 Results and Discussion......Page 458
4 Conclusions......Page 461
6 References......Page 462
2 Experimental Procedures......Page 463
3 Results and Discussion......Page 464
5 References......Page 468
2 Introduction......Page 469
3 Molten Handling......Page 470
4 Casting Knock-out and Cleaning......Page 472
5 Casting Storage/Temporary Protection......Page 473
7 References......Page 474
2 Introduction......Page 475
3 Experimental Procedure......Page 476
4 Results and Discussion......Page 477
5 Summary......Page 479
6 References......Page 480
Introduction......Page 481
Experimental......Page 482
Results and Discussion......Page 483
References......Page 486
2 Experimental......Page 487
3 Adhesive Bonding Tests......Page 488
4 Results and Discussion......Page 489
5 Conclusions......Page 491
7 References......Page 492
3 Material......Page 493
4 Development of Alternative Conversions......Page 494
5 Characterisation......Page 495
8 Bibliography......Page 501
2 Introduction......Page 502
3 The R&D ANOMAG Pilot Plant......Page 504
5 Corrosion and Paint Adhesion Test......Page 506
6 Costs......Page 508
8 Bibliography......Page 509
1 Introduction......Page 510
2 Paint Finish for Automotive Bodies......Page 511
3 Experimental Work......Page 512
4 Conclusion......Page 515
5 References......Page 516
1 Introduction......Page 517
3 Results and Discussion......Page 519
4 Conclusions......Page 522
5 References......Page 523
2 Experimental......Page 524
3 Results and Discussion......Page 527
6 References......Page 531
1 Introduction......Page 532
2 Experimental procedures......Page 533
3 Results and Discussion......Page 534
5 References......Page 536
2 Experimentals......Page 537
3 Results......Page 538
5 Conclusions......Page 541
6 References......Page 542
Introduction......Page 543
Experimental......Page 544
Results and Discussion......Page 545
References......Page 547
Processing......Page 549
2 The New Rheocasting (NRC) Process......Page 551
3 Microstructural Considerations and Mechanical Properties......Page 554
6 References......Page 556
2 Design Considerations......Page 558
3 Input Data......Page 560
4 Summary......Page 564
5 References......Page 565
2 Simulation in the Design Process of a Magnesium Wheel......Page 566
4 Simulation of the Start Design......Page 567
5 Optimization of Casting Design and Casting Process......Page 569
6 Summary......Page 570
2 Introduction......Page 571
3 A new Device for Magnesium Low Pressure Sand Casting......Page 572
4 Development Results......Page 573
5 Thin-walled Parts......Page 574
7 Conclusion......Page 575
2 Introduction......Page 576
3 Determination of Material Parameters......Page 578
4 Discussion......Page 579
6 References......Page 580
1 Introduction......Page 582
2 Continuous Casting Technology......Page 583
4 Conclusion......Page 587
6 References......Page 588
2 Experimental Procedures......Page 589
3 Results and Discussion......Page 590
6 References......Page 593
1 Introduction......Page 595
2 Injection Cycle......Page 596
3 Semi–Solid Magnesium Influences the Processing Conditions......Page 597
4 Molds......Page 598
5 MoldedParts......Page 599
6 Summary......Page 600
8 References......Page 601
2 Introduction......Page 602
4 Results......Page 603
5 Conclusions......Page 606
7 References......Page 607
2 Introduction......Page 608
4 Fundamentals and Results......Page 609
5 Applications......Page 612
7 References......Page 613
2 Production of the Investigated Alloys......Page 614
3 Results and Discussion......Page 615
4 Conclusions......Page 618
5 Literature......Page 619
2 Wrought Magnesium Alloys......Page 620
3 Magnesium Section Technology......Page 621
4 Magnesium Sheet Technology......Page 623
5 Summary and Outlook......Page 625
6 Literature......Page 626
2 The Principles of Hydrostatic Extrusion......Page 627
4 Mechanical Proerties and Micro Structure......Page 629
5 Feasibility of Hydrostatic Extrusion......Page 631
2 Properties of the Tested Sheet Materials......Page 633
4 Deep Drawing Tests of Cylindrical Cups at Room Temperature......Page 635
5 Drawing and Press Bending of Structural Components......Page 637
7 References......Page 638
2 Material – Original Microstructure – Experimental......Page 639
3 Hot Forming Behavior of Mg Material AZ 61......Page 641
4 Cold Forming Behavior of Mg Material AZ 61......Page 642
5 Results of Rolling Tests......Page 643
7 References......Page 645
2 Literature Review......Page 646
3 Research Methodology......Page 648
4 Results......Page 649
5 Discussion......Page 651
7 References......Page 652
2 Experimental Procedure......Page 653
3 Results and Discussion......Page 654
4 References......Page 656
1 Introduction......Page 657
2 Compression Test Simulation under Dynamic Loading......Page 658
3 Conclusions......Page 661
5 References......Page 662
2 Experimental......Page 663
3 Results and Discussion......Page 664
6 References......Page 667
2 Experimental Procedures......Page 669
3 Results and Discussion......Page 670
4 Conclusions......Page 673
6 References......Page 674
Physical Properties......Page 675
2 Experimental Procedure......Page 677
3 Experimental Results and Discussion......Page 678
4 Conclusions......Page 681
6 References......Page 682
3 Experimental Part......Page 683
4 Results and Discussion......Page 684
6 Acknowledgements......Page 687
7 References......Page 688
2 Introduction......Page 689
3 Experimental Procedure......Page 690
4 Results and Discussion......Page 691
6 References......Page 694
2 Introduction......Page 696
3 Experimental Procedure......Page 697
4 Results and Discussion......Page 698
5 Conclusions......Page 700
7 References......Page 701
Creep Behaviour......Page 703
2 Results......Page 705
3 Discussion of the Tension/Compression Asymmetry......Page 707
4 Acknowledgments......Page 709
5 References......Page 710
2 Experimental Procedure......Page 711
3 Microstructure......Page 712
4 Mechanical Properties......Page 713
5 Discussion......Page 715
7 References......Page 716
2 Experimental Procedures......Page 717
3 Results and Discussion......Page 718
6 References......Page 722
2 Experimental Procedure......Page 723
3 Results......Page 724
4 Discussion......Page 726
5 Conclusions......Page 727
6 References......Page 728
3 Experimental......Page 729
4 Results and Discussion......Page 730
6 References......Page 733
3 Experimental......Page 734
4 Experimental Results......Page 735
5 Discussion......Page 737
7 Acknowledgements......Page 738
8 References......Page 739
Recycling, Melting, Environmental......Page 741
1 General......Page 743
2 Casting methods......Page 744
4 Pre-heating and Charging Machine......Page 745
5 Melting and Holding Furnaces......Page 746
6 Furnace Design......Page 748
8 Protective Gas Mixing System (Figure 11)......Page 750
10 Practical Casting Operation......Page 752
11 Remelting Plants......Page 753
12 Summary......Page 756
2 Experimental......Page 757
3 Results and Discussion......Page 758
4 Conclusions......Page 762
5 References......Page 763
Experimental procedure......Page 764
Results......Page 765
Discussion......Page 767
Summary and Outlook......Page 768
References......Page 769
1 Introduction......Page 770
2 Melting of Dross and Sludge with Flux......Page 771
3 Utilization of the Oxide/Salt Fraction by Leaching......Page 774
5 Acknowledgement......Page 777
6 References......Page 778
2 Sulfur Hexafluoride and Sulfur Dioxide......Page 779
4 ReactionMechanisms......Page 780
5 Distribution and Control of the SO2 Gas Mixture......Page 781
6 Health, Environment and Safety (HES) Aspects......Page 782
8 Conclusions......Page 783
9 References......Page 784
2 Impurities inMagnesium......Page 785
3 Sources of Impurities......Page 786
4 Effects of Impurity Elements and Inclusions......Page 788
5 Recovery of Magnesium from Scrap......Page 791
7 References......Page 796
Simulation......Page 799
2 Applied Methods for Determination of the Solidification Curves of Magnesium AZ-Alloys......Page 801
3 Results......Page 802
4 Discussion......Page 804
5 Conclusions......Page 805
7 References......Page 806
Author Index (with page links, to the first page of the respecting article)......Page 807
Subject Index (with page links, to the first page of the respecting article)......Page 811
Recommend Papers
![Magnesium: Proceedings of the 6th International Conference Magnesium Alloys and their Applications [1st edition]
9783527309757, 3527309756](https://ebin.pub/img/200x200/magnesium-proceedings-of-the-6th-international-conference-magnesium-alloys-and-their-applications-1st-edition-9783527309757-3527309756.jpg)
![Magnesium, Magnesium Alloys, and Magnesium Composites [1 ed.]
0470494174, 9780470494172](https://ebin.pub/img/200x200/magnesium-magnesium-alloys-and-magnesium-composites-1nbsped-0470494174-9780470494172.jpg)
![Alloys and Their Industrial Applications [4 ed.]
5519380236, 9785519380232](https://ebin.pub/img/200x200/alloys-and-their-industrial-applications-4nbsped-5519380236-9785519380232.jpg)
![Corrosion Resistance of Aluminum and Magnesium Alloys: Understanding, Performance, and Testing [1 ed.]
047171576X, 9780471715764](https://ebin.pub/img/200x200/corrosion-resistance-of-aluminum-and-magnesium-alloys-understanding-performance-and-testing-1nbsped-047171576x-9780471715764.jpg)
File loading please wait...
Citation preview
Magnesium Alloys and their Applications Edited by K. U. Kainer
Magnesium Alloys and their Applications. Edited by K. U. Kainer. ©WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
Further Titles of Interest: P. J. Winkler et al. (Eds.) Materials for Transportation Technology ISBN 3-527-30124-0 T. W. Clyne et al. (Eds.) Metal Matrix Composites and Metallic Foams ISBN 3-527-30126-7 B. L. Mordike, K. U. Kainer (Eds.) Magnesium Alloys and their Applications ISBN 3-527-29936-X H. Kunze et al. (Eds.) Competitive Advantages by Near-Net-Shape Manufacturing ISBN 3-527-29924-6
Magnesium Alloys and their Applications Edited by K. U. Kainer
Deutsche Gesellschaft für Materialkunde e.V.
Weinheim · New York · Chichester Brisbane · Singapore · Toronto
Prof. Dr.-Ing. K. U. Kainer GKSS-Forschungszentrum Geesthacht GmbH Institut für Werkstofforschung Max-Planck- Straße D-21502 Geesthacht International Congress „Magnesium Alloys and their Applications“ held from 26-28 September 2000 in Munich Organizer: DGM • Deutsche Gesellschaft für Materialkunde e.V.
This book was carefully procuced. Nevertheless, authors, editors and publisher do not warrant the information contained therein to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.
Library of Congress Card No. applied for. A catalogue record for this book is available from the British Library. Deutsche Bibliothek Cataloging-in-Publication Data: A catalogue record for this publication is available from Die Deutsche Bibliothek ISBN 3-527-30282-4 © WILEY-VCH Verlag Gmbh, D-69469 Weinheim (Federal Republic of Germany), 2000 Printed on acid-free and chlorine-free paper. All rights reserved (including those of translation in other languages). No part of this book may be reproduced in any form – by photoprinting, microfilm, or any other means – nor transmitted or translated into machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law. Composition: WGV Verlagsdienstleistungen GmbH, Weinheim Printing: betz-druck, Darmstadt Bookbinding: Buchbinderei Osswald, Neustadt/Wstr. Printed in the Federal Republic of Germany
Preface The success of the International Conference on „Magnesium Alloys and their Applications“ in April 1998 in Wolfsburg demonstrated the renewed interest in innovative applications of magnesium materials. This demand has resulted in an increased research and development activity in companies and research institutes in order to achieve an improved property profile and a better choice of alloy Systems. First results have been presented at the last Conference. Due to Stagnation in development in the 80’s and 90’s there is much ground to be covered to fulfill the demands. This is a critical period for the successful application of magnesium based materials. It is particularly important to promote exchange of information and discussion in which development trends and application potential in different fields like the automotive industry and communication technology are discussed in an interdisciplinary framework. This year’s Conference, to be held from September 26-28 in Munich, will cover sessions on the following topics: alloy development, mechanical properties, physical properties, texture and microstructure, creep behavior, corrosion and surface treatment, processing, joining, magnesium matrix composites, Simulation, application, recycling, melting, environmental aspects. The book in hand presents all the Conference topics and the papers are grouped according to the Conference program. The International Conference Magnesium 2000 will be held in parallel to the „Materials Week“ and the International Trade Fair „Materialica“. This exhibition for innovative materials, processes, and applications highlights the topic magnesium within a Special sector. A joint presentation of research centers, institutes, and universities and a forum will be organized.
K. U. Kainer
Contents
Alloy Development Global Overview on Demand and Applications for Magnesium Alloys R. L. Edgar, Hydro Magnesium, Brussels (B) ............................................................................3 Focused Development of Magnesium-Alloys Using Computational Thermochemistry as a Powerful Tool R. Schmid-Fetzer, J. Gröbner, Institute of Metallurgy, Technical University of Clausthal (D); D. Kevorkov, Institute of Metallurgy, Technical University of Clausthal (D)............................9 Development of Practical High Temperature Magnesium Casting Alloys J. F. King, Magnesium Elektron, Manchester (GB) .................................................................14 Development of a Low-cost, Temperature- and Creep-resistant Magnesium Die-casting Alloy F. von Buch, S. Schumann, Volkswagen AG Wolfsburg (D); E. Aghion, B. Bronfin, Dead Sea Magnesium, Beer-Sheva (IL); B. L. Mordike, Department of Materials Engineering and Technology, Technical University Clausthal (D); M. Bamberger, Technion, Haifa (IL); D. Eliezer, Ben-Gurion University of the Negev, Beer-Sheva (IL)...........................................23 Creep Resistant Mg Alloy Development K. Pettersen, H. Westengen, Norsk Hydro Research Centre, Porsgrunn (N); J. I. Skar, M. Videm, Norsk Hydro Research Centre, Porsgrunn (N); L.-Y. Wei, Luleå University of Technology (S) .........................................................................29 Development of High Temperature Creep Resistant Alloys B. L. Mordike, Institute of Material Science and Engineering, Technical University Clausthal (D); F. von Buch, now: VW Forschung Wolfsburg (D)....................................................................35 New Magnesium Wrought Alloys C. Jaschik, H. Haferkamp, M. Niemeyer, Institut of Materials Science, University of Hanover (D) ..............................................................................................................................41 Effects of Alloying Elements of the Creep Resistance of Thixomolded Mg-Al-Ca-X (X=Si, Zn, Mm, Ba, Sr) Alloys T. Tsukeda, R. Uchida, K. Saito, The Japan Steel Works, Ltd. (J); M. Suzuki, J. Koike, K. Maruyama, H. Kubo, Tohoku University, Tohoku (J).........................47 Precipitation Processes in Magnesium-Heavy Rare Earth Alloys during Ageing at 300°C. P.J. Apps, G. W. Lorimer, Manchester Materials Science Centre, Manchester (GB); H. Karimzadeh, J.F. King, Magnesium Elektron, Manchester (GB)........................................53
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
VIII Grain Refinement of Magnesium Casting Alloys by Boron Addition N. Nishino, H. Kawahara, Y. Shimizu, H. Iwahori, Toyota Central R & D Labs., Inc., Aichi (J).....................................................................................................................................59 Studies on Cadmium and Silver Trace Element Modified AZ91C Magnesium Alloy R. Shabadi, E. S. Dwarakadasa, UGC Center for Advanced Studies, Indian Institute of Science, Bangalore (IND); R. Ambat, University of Birmingham (GB); K. L. Bhat, Department of Metallurgical and Materials Engineering, KREC, Surathkal. (IND); V. Gopalakrishnan, Foundry and Forge Division, Hindusthan Aeronautics Limited., Bangalore (IND) .......................................................................................................................65 Electrophysical Properties of Mg-Pb Based Liquid Alloys and Their Application. Y. Plevachuk, Institute of Applied Physics, Ivan Franko National University, Lviv (UA) .......73 Mg-Al-(Sc, Gd) Alloy Design Using Computational Thermochemistry J. Gröbner, D. Kevorkov, R. Schmid-Fetzer, Institute of Metallurgy, Technical University of Clausthal (D); A. Pisch, Laboratoire de Thermodynamique et Physico-Chimie Métallurgiques Institut National Polytechnique de Grenoble, Saint Martin d'Hères (F)..............................................79 Mg-alloy Database Construction: Investigation of Al-Li-Si Phase Equilibria D. Kevorkov, J. Gröbner, R. Schmid-Fetzer, Institute of Metallurgy, Technical University of Clausthal (D) ............................................................................................................................84 Mg- alloy Database Construction: Investigation of Al-Ca Phase Equilibria D. Kevorkov, R. Schmid-Fetzer, Institute of Metallurgy, Technical University of Clausthal (D) ........................................................................................................................88 Aging Response of Mg-Rare Earth Alloys with Low Scandium Content B. Bohumil Smola, Institute of Materials Engineering, Czech Technical University, Prague (CZ); I. Ivana Stulíková, J. Jitka Pelcová, Faculty of Mathematics and Physics, Charles University, Prague (CZ); F. Frank von Buch, B. L. Barry L. Mordike, Institute of Material Science and Engineering, Technical University Clausthal (D) ..........................................................................................92 Phase Equilibria, Microstructure and Properties of Novel Mg-Mn- Y Alloys A. Pisch, C. Antion, C. Tassin, F. Baillet, Laboratoire de Thermodynamique et PhysicoChimie Métallurgiques Institut National Polytechnique de Grenoble, Saint Martin d'Hères (F); J. Joachim Gröbner, R. Schmid-Fetzer, Institute of Metallurgy, Technical University of Clausthal (D) ............................................................................................................................98 Microstructure and Protium Absorbing/desorbing Characteristics of Mg-Ni-Mn Alloys H. Hayato Okumura, T. Tohru Tabata, A. Akihiro Matsui, S. Shigeharu Kamado, Y Kojima, Department of Mechanical Engineering, Nagaoka University of Technology, Nagaoka (J) ........................................................................................................103
IX Heat and Corrosion Resistance of Mg-Zn-Al-Ca Alloys I. A. Anyanwu, T. Honda, Nagaoka University of Technology, Nagaoka (J); S. Kamado, Y. Kojima, Department of Mechanical Engineering, Nagaoka University of Technology, Nagaoka (J); S. Takeda, T. Ishida, Magnesium Manufacturing Division, Ahresty Corporation, Tokyo (J) .................................................................................................................................110 Texture and Microstructure Texture Analysis as a Tool for Wrought Magnesium Alloy Development S. R. Agnew, M. H. Yoo, J. A. Horton, Oak Ridge National Laboratory, Oak Ridge, TN (USA)..............................................................................................................119 Microstructure, Texture and Residual Microstrains in MgAl8Zn Deformed at Very High Strain Rates A. Pyzalla, Hahn-Meitner-Institut, Berlin (D); M. Brodmann, Lehr- und Forschungsgebiet Werkstoffkunde, RWTH Aachen (D); P. L. Lee, D. Haeffner, Advanced Photon Source, Argonne National Lab., Argonne, IL (USA) ..................................................................................................................125 Zn Incorporation within the Intermetallic Mg12(LaxCe1-x) Lattice in Elektron MEZ C. J. Bettles, C. J. Rossouw, K. Venkatesan, CSIRO Manufacturing Science and Technology, Victoria (D) ........................................................................................................131 Orientation Relationship in Mg-Mg17Al12 Eutectic S. Guldberg, Hydro Magnesium, Porsgrunn (N); N. Ryum, Department of Materials Technology and Electrochemistry, NTNU, Trondheim (N).........................................................................................................................137 Influence of Texture on Deformation Behaviour of Magnesium Alloy AZ31 R. Gehrmann, M. M. Frommert, G. Gottstein, Institut für Metallkunde und Metallphysik, RWTH Aachen (D) ..................................................................................................................143 Ageing Behaviour and Microstructure of a Mg-9Al-3Zn Alloy. K. Venkatesan, C. J. Bettles, CSIRO Manufacturing Science and Technology, Victoria (D).............................................................................................................................149 Magnesium SiC Reinforced Composites - Texture and Residual Strain Investigation by Simulation and Experiments H.-G. Brokmeier, E. M Jansen, IWWTU Clausthal and GKSS-Forschungszentrum Geesthacht (D); P. Spalthoff, J. A. Signorelli, P. A. Turner, Instituto de Física Rosario (RA); R. E. Bolmaro, IGDL-Univ. Göttingen (D) and FLNP-JINR Dubna-Russia (RUS) ..............155 Effect of Thermomechanical Treatments on the Microstructure of AZ91 Alloy N. V. Ravi Kumar, J. J. Blandin, M. Suéry, Institut National Polytechnique de Grenoble (INPG), Génie Physique et Mécanique des Matériaux (GPM2), Saint-Martin d'Hères (F) ..161
X Magnesium Applications in Aerospace and Electronic Industries B. Landkof, Technion, Israel Institute of Technology.............................................................168 Joining Friction Stir Welding of Lightweight Materials S. W. Kallee, W. M. Thomas, E. D. Nicholas, TWI Ltd, Cambridge (GB)..............................175 Strategies to Reduce Porosity in Electron Beam Welds of Magnesium Die-Casting Alloys C. Vogelei, D. von Dobeneck, pro-beam KGaA, Planegg/München (D); I. Decker, H. Wohlfahrt, Institut für Schweißtechnik, Technische Universität Braunschweig (D) ...................................................................................................................191 Influences on the Static and Dynamic Strength of MIG-welded Magnesium Alloys M. Rethmeier, S. Wiesner, H. Wohlfahrt, Institut für Schweißtechnik, Technische Universität Braunschweig (D) ................................................................................................200 Magnesium Matrix Composites Design Rules for Selective Reinforcement of Mg-Castings by MMC Inserts H. P. Degischer, Institute of Materials Science and Testing, Vienna University of Technology, Vienna (A); F. G. Rammerstorfer, Institute of Lightweight Structures and Aerospace Engineering, Vienna University of Technology, Vienna (A); O. Beffort, Swiss Federal Laboratories for Material Testing and Research, Thun (CH) ......207 The Influence of Ca-Additions on the Mechanical Properties of T300-C-Fibre/Mg(Al) Metal Matrix Composites O. Beffort, Swiss Federal Laboratories for Material Testing and Research, Thun (CH); C. Hausmann, VonRoll Druckguss AG, St.Gallen (CH).........................................................215 Fabrication and Properties of Cast and Extruded SiCw/AZ91Mg Composites J. Kaneko, M. Sugamata, J. Kim, M. Masataka Kon, Nihon University Izumi-cho, Narashino, Chiba (J) ..............................................................................................................221 Thermal Fatgue of Magnesium Matrix Composites F. Chmelík, P. Lukác, Department of Metal Physics, Charles University, Praha (CZ); S. Kúdela, Institute of Materials and Mechanical Engineering, Slovak Academy of Sciences, Bratislava (SK); J. Kiehn, B. L. Mordike, Department of Materials Engineering and Technology, Technical University Clausthal (D); K.-U. Kainer, Institute for Materials Research, GKSS Investigation Centre, Geesthacht (D) ........................................................................................................................229 Microstructure and Creep Behavior of SiC Particulate Reinforced QE 22 Composite M. Svoboda, M. Pahutova, J. Brezina, V. Sklenicka, Institute of Physics of Materials, Academy of Sciences of the Czech Republic, Brno (CZ); F. Moll, Institute of Materials Engineering and Technology, Technical University of Clausthal (D) ..........................................................................................................................234
XI Formation of Mg2Si by Infiltration of C-Fibres-Si-Hybrid-Preforms K. U. Kainer, H. Dieringa, Inst. für Werkstoffforschung GKSS-Forschungszentrum Geesthacht (D); P. Schulz, J. Reiter, Leichtmetall Kompetenzzentrum Ranshofen (A) ....................................240 Flow Mechanisms in Creep of an AZ 91 Magnesium-based Composite V. Sklenicka, M. Pahutova, K. Kucharova, M. Svoboda, Institute of Physics of Materials, Academy of Sciences of the Czech Republic, Brno (CZ); T. G. Langdon, Departments of Materials Science and Mechanical Engineering, University of Southern California, Los Angeles, CA (USA) ...................................................246 Possibilities of the Heat Treatment of MagnesiumMatrix Composites Reinforced with SiC Particles K. N. Braszczynska, Technical University of Czestochowa (PL)............................................252 Fabrication and Microstructure Analysis of Al18B4O33 Ceramics Reinforced Magnesium Alloy Matrix Composites G. Sasaki, S. Hara, M. Yoshida, J. Pan, H. Fukinaga, Dept of Mech. Eng., Hiroshima Univiversity, Higashi-Hiroshima (J); N. Fuyama, T. Fujii, Western Hiroshima Pref. Industrial Res. Inst., Kure (J).......................257 Dynamic Behaviour of Magnesium Matrix Composites in Elevated Temperatur E. Lach, T. U. Benzler, K. U. Kainer, GKSS Research Centre – Geesthacht (D); A. Bohmann, M. Scharf, ISL, French-German Research Institute of Saint-Louis (F)............263 Mechanical Development The Use of Magnesium in Cars – Aspects of Corrosion M. Brettmann, B. Reinhold, H. Schnattinger, Audi AG, Ingolstadt (D) .................................271 Mechanical Properties of Extruded Magnesium Alloys B. Closset, Timminco SA, Geneva (CH) .................................................................................274 The Grain Size Dependence of Strength in the Extruded AZ91 Mg Alloy M. Mabuchi, Y. Yamada, K. Shimojima, C. E. Wen, Y. Chino, M. Nakamura, T Asahina, Materials Processing Department, National Industrial Research Institute of Nagoya (J); H. Iwasaki, Himeji Institute of Technology, Shosha, Himeji, Hyogo (J); T. Aizawa, Department of Metallurgy, The University of Tokyo, (J); K. Higashi, Department of Metallurgy and Materials Science, Osaka Prefecture University, Gakuen-cho, Sakai, Osaka (J)..............................................................................280 Formability and Strain Rate Sensitivity of a Mg-8.5Li-1Zn Alloy Sheet H. Takuda, Kyoto University, Kyoto (J); S. Kikuchi, The University of Shiga Prefecture, Hikone (J); K. Kubota, Mitsui Mining & Smelting Co. Ltd., Ageo (J).......................................................285 Mechanical Properties of MagnesiumAlloys Processed by Semi-Solid Casting W. Wagener, D. Hartmann, EFU Gesellschaft für Ur-/Umformtechnik mbH, Simmerath (D); F. Lehnert, BMW Group München (D); K. Scholz, GeorgFischer Automotive, Schaffhausen (CH) .....................................................291
XII Determination of Material Properties and Numerical Simulation to Predict the Mechanical Performance of Die Casted Components M. Wuth, E. Lieven, PETRI AG, Aschaffenburg (D); W. Böhme, Fraunhofer-Institut für Werkstoffmechanik (IWM), Freiburg (D) .......................296 Fatigue Design with Cast Magnesium Alloys C. M. Sonsino, K. Dieterich, L. Wenk, Fraunhofer-Institute for Structural Durability (LBF), Darmstadt (D); A. Till, German Foundrymen's Association, Düsseldorf (D)..................................................304 Isothermal Fatigue of Magnesium Wrought Alloy AZ31 U. Noster, I. Altenberger, B. Scholtes, Institute of Materials Technology, University Gh Kassel (D) ...............................................................................................................................312 Characterisation of Precipitate Phases in WE54 and AZ91 Alloys J. F. Nie, Department of Materials Engineering, Monash University, Clayton (AUS); X. L. Xiao, C. P. Luo, Department of Mechano-Electronic Engineering, South China University of Technology, Guangzhou (VRC) ........................................................................318 Compression Test on MagnesiumAlloy MgAl8Zn at High Strain Rates and Temperatures E. El-Magd, M. Abouridouane, Department of Materials Science, RWTH Aachen (D) ........324 Wear Behaviour of Laser Surface Treated Magnesium Alloys U. Kutschera, DLR Stuttgart (D); R. Galun, IWW, TU Clausthal (D)..........................................................................................330 Mechanical Behavior and Residual Stresses in AZ31 Wrought Magnesium Alloy Subjected to Four Point Bending J. P. Nobre, M. Kornmeier, A. Dias, Department of Mechanical Engineering, University of Coimbra (P); U. Noster, J. Gibmeier, I. Altenberger, B. Scholtes, Institut of Materials Technology, University Gh Kassel (D)........................................................................................................336 Superplasticity of Magnesium-Based Alloys U. Draugelates, A. Schram, C.-C. Kedenburg, Institute of Welding and Machining, (ISAF), Clausthal (D)..............................................................................................................342 Cyclic Deformation Behavior of the Cast Magnesium Alloy AZ91 H. W. Höppel, G. Eisenmeier, B. Holzwarth, H. Mughrabi, Institut für Werkstoffwissenschaften, Universität Erlangen-Nürnberg, Erlangen (D) .............................348 Deformation Twinning of AZ31 Alloy in Quasistatitic and Dynamic Compression Tests E. Lach, A. Bohmann, M. Scharf, ISL, French-German Research Institute of Saint-Louis (F); U. Kainer, GKSS Research Centre – Geesthacht (D).............................................................354 Deformation and Fracture Behavior of Magnesium Structural Components A. Ockewitz, C. Schendera, D.-Z. Sun, Fraunhofer-Institut für Werkstoffmechanik (IWM), Freiburg (D); B. Grosser, A. Hamann, Volkswagen AG Wolfsburg (D) .......................................................359
XIII Internationalization of Magnesium Research Through USCAR in North America and EUCAR in Europe G. Cole, Ford Motor Company Research Laboratories, Dearborn, MI (USA)......................365 Application High-speed-drilling in AZ91 D without lubricoolants F. Tikal, M. Schmier, Universität Gesamthochschule Kassel (D); C. Vollmer, Volkswagen AG Wolfsburg (D) ...........................................................................373 Design, Optimization and Reliability of Magnesium Safety Vehicle Parts Y. Tzabari, Israel Institute of Metals, Technion, Israel Institute of Technology, Haifa (IL); I. Reich, Y. Bahalul, Ortal Diecasting LTD. Kibbutz Neve-Ur (IL) .......................................380 Development and Production of a Die-cast Magnesium Convertible Soft-top Cover P. Geist, F. Lehnert, BMW Group München (D); U. Kwasny, EDAG AG Munich (D) ........................................................................................387 Magnesium Motorcycle Wheels for Racing Applications K. J. Schemme, Otto Fuchs Metallwerke, Meinerzhagen (D).................................................391 Weight and Cost Saving with Magnesium Die Castings A. Mertz, Honsel GmbH & Co KG, Meschede (D).................................................................397 Cast Magnesium Alloys For Wide Application P. G. Detkov, I. Yu. Mukhina, A. D. Zhirnov, Solicamsk Magnesium Works, All-Russian Institute of Aviation Materials (RUS) .....................................................................................402 Improving the Characteristics of MagnesiumWorkpieces by Burnishing Operations H. K. Tönshoff, T. Friemuth, J. Winkler, C. Podolsky, Institute for Production Engineering and Machine Tools, University of Hannover (D) ...................................................................406 Machining of Light-Metal Matrix Composites K. Weinert, M. Lange, M. Schroer, Institut für Spanende Fertigung, University of Dortmund (D)..........................................................................................................................412 Nitriding of Pressure Die Casting Dies and Tool Elements H. R. Schmauser, Drei-S-Werk, Schwabach (D) ....................................................................418 Corrosion and Surface Treatment Corrosion Behaviour of the Microstructural Constituents of AZ Alloys G. Song, A. Atrens, D. StJohn, L. Zheng, Department of Mining, Minerals and Materials Engineering, The University of Queensland, Brisbane (AUS) ...............................................425 Corrosion Properties of Die Cast AM Alloys M. Videm, J. I. Skar, P. Bakke, Norsk Hydro Research Centre, Porsgrunn (N) ....................432 Improved Corrosion and Oxidation Resistance of AMand AZ Alloys by Ca and RE Additions H. Alves, U. Köster, Dept. Chem. Eng., University of Dortmund (D) ....................................439
XIV Corrosion Behavior of thin Wall MagnesiumProducts Molded by Thixomolding I. Nakatsugawa, H. Takayasu, K. Saito, The Japan Steel Works, Ltd. (J)..............................445 Effect of Foundry Processing on the Corrosion Performance of High Purity MagnesiumSand Casting Alloys H. Karimzadeh, Magnesium Elektron, Manchester (GB).......................................................451 Corrosion of the Magnesium Alloy AZ91 and its Influence on Fatigue Properties C. Müller, R. Koch, Institut für Materialwissenschaft, Technische Universität Darmstadt (D); G. H. Deinzer, Adam Opel AG Internationales Technisches Entwicklungszentrum, Rüsselsheim (D) ......................................................................................................................457 Effect of MEchanical Surface Treatment and Environment on Fatigue of Wrought Magnesium Alloys M. Hilpert, L. Wagner, Chair of Physical Metallurgy and Materials Technology, Technical University of Brandenburg at Cottbus (D) .............................................................................463 Phosphate-Permanganate: A Chrome Free Alternative for Magnesium Pre-treatment J. I. Skar, D. Albright, Norsk Hydro ASA, Porsgrunn (N)......................................................469 Alternatives to Cr(VI) Conversion Coatings for Magnesium Alloys I. Azkarate, P. Cano, A. Del Barrio, M. Insausti, P. Santa Coloma, Fundación INASMET, San Sebastián (P)....................................................................................................................475 Surface Treatments for Large Automotive Magnesium Components G. Guerci, C. Mus, TEKSID S.p.A., (I); K. Stewart, Meridian GTC (CND) ..........................................................................................484 Coating Systemfor Magnesium Diecastings in Class A Surface Quality R. Gadow, D. Scherer, Institute for Manufacturing Technologies of Ceramic Components and Composites, University of Stuttgart (D); F. J. Gammel, DaimlerChrysler AG, München (D); F. Lehnert, BMW Group München (D); J. I Skar, Norsk Hydro ASA, Porsgrunn (N)...........................................................................492 Corrosion Fatigue and Corrosion Creep of Magnesium Alloys A. Eliezer, E. M. Gutman, E. Abramov, Y. Unigovski, E. Aghion, Dept. of Materials Engineering, Ben-Gurion University of the Negev, Beer-Sheva (IL) .....................................499 Corrosion Resistance of Die Casting AZ91D Magnesium Alloys in the Atmosphere H. Umehara, National Institute of Materials and Chemical Research, Tsukuba, Ibaraki (J); M. Takaya, Chiba Institute of Technology, Narashino, Chiba (J), ; T. Itoh, Japan Weathering Test Center, Choshi, Chiba (J) ....................................................506 Fabrication of Pure Magnesium Films on Magnesium Alloys by Vapor Deposition Technique H. Tsubakino, A. Yamamoto, S. Fukumoto, Himeji Institute of Technology, Himeji (J); A. Watanabe, Graduate student of Himeji Institute of Technology (J)...................................514
XV Corrosion Creep of Magnesium and Die-Cast Magnesium Alloys E. M. Gutman, Y. Unigovski, A. Eliezer, E. Abramov, Dept. of Materials Engineering, Ben-Gurion University of the Negev, Beer-Sheva (IL) ...........................................................519 Response of Light Alloys to Mechanical Surface Treatments: Comparison of Magnesium and Aluminum Alloys M. Hilpert, L. Wagner, Chair of Physical Metallurgy and Materials Technology, Technical University of Brandenburg at Cottbus (D).............................................................525 Processing The Fundamentals of the New Rheocasting – Process for Magnesium Alloys H. Kaufmann, Leichtmetall Kompetenzzentrum Ranshofen (A); P. J. Uggowitzer, Swiss Federal Institute of Technology, ETH Zürich (CH).........................533 High Performance Die Castings - Utilizing Magnesium’s Properties T. Aune, Norsk Hydro ASA, Porsgrunn (N); H. Westengen, D. Albright, Norsk Hydro ASA, Porsgrunn (N)..............................................540 Optimized Development for Magnesium Castings and Casting Processes; Increase in Value by Applying a Closed Process Chain for the Development of Automotive Magnesium Castings G. Hartmann, A. Egner-Walter, MAGMA Gießereitechnologie GmbH, Aachen (D) ............548 Thin-walled Mg Structural Parts by a Low-pressure Sand Casting Process F. J. Edler, G. Lagrené, Honsel Fonderie Messier, Arudy (F); R. Siepe, Honsel GmbH & Co KG, Meschede (D)..................................................................553 Quality Index Charts for Mg-based Casting Alloys C. H. Cáceres, Department of Mining, Minerals and Materials Engineering, The University of Queensland, Brisbane (AUS) .....................................................................558 Magnesium Adapted Continuous Casting Technology U. Holzkamp, H. Haferkamp, M. Niemeyer, Institut of Materials Science, University of Hanover (D) ............................................................................................................................564 Microstructure Evolution and Mechanical Properties of AZ91 Mg Foams C. Wen, Y. Yamada, K. Shimojima, M. Mabuchi, M. Nakamura, T. Asahina, National Industrial Research Institute of Nagoya (J); T. Aizawa, The University of Tokyo, Tokyo (J); K. Higashi, Osaka Prefecture University, Osaka (J)..............................................................571 Semi Solid Injection Molding of MagnesiumAlloys A. Dworog, Gerhard-Mercator-Universität Duisburg (D); M. Kothen, D. Hartmann, EFU Gesellschaft für Ur-/Umformtechnik mbH, Simmerath (D); K. Kuhmann, C. Boehnke, Hengst Filterwerke GmbH & Co. KG, Münster (D) ....................577
XVI Characterization of Melt Spun Mg-Ca-Zn Alloys P. M. Jardim, I. G. Solórzano, Department of Material Science and Metallurgy, PUC-Rio, Rio de Janeiro (BR); J. B. Vander Sande, Department of Material Science and Engineering – MIT, Cambridge, MA (USA); B. S. You, W. W. Park, Korea Institute of Machinery & Materials, Changwon, Kyungnam (ROK)....................................................................................................................584 Properties and Perspectives of Magnesium Rolled Products J. Enss, T. Everetz, T. Reier, P. Juchmann, Salzgitter AG Steel and Technology, Salzgitter (D)...........................................................................................................................590 Microstructure, Mechanical Properties And Deformation Behaviour of Extruded Magnesium Alloys K. U. Kainer, GKSS Research Centre – Geesthacht (D); E. Doege, S. Janssen, Institute for Metal Forming and Metal Forming Machine Tools, University of Hanover (D); T. Ebert, Department of Materials Science and Engineering, Technical University of Clausthal (D) ..........................................................................................................................596 Properties and Processing of Magnesium Wrought Products for Automotive Applications W. Sebastian, K. Dröder, S. Schumann, Volkswagen AG Wolfsburg (D)...............................602 Hydrostatic Extrusion of Magnesium K. Savage, J. F. King, Magnesium Elektron, Manchester (GB); A. van Kooij, HME, Holland (NL) ..........................................................................................609 Deep Drawing of Magnesium Sheet Metal at Room Temperature H.-W. Wagener, Metal Forming Laboratory, University of Kassel (D); F. Lehnert, Bayerische Motoren Werke AG, München (D) ....................................................615 Hot and Cold Forming Behaviour of Magnesium Alloys AZ31 and AZ61 L. Chabbi, W. Lehnert, R. Kawalla, Institute of Metal Forming, Freiberg University of Mining and Technology, Freiberg (D); F. Lehnert, BMW Group München (D)...................................................................................621 Suppression of Mold-Metal Reactions during Investment Casting M. H. Idris, A. Ourdjini, E. Hamzah, Universiti Teknologi Malaysia, Johor (MAL); A. Clegg, Loughborough University, Loughborough (GB) ....................................................628 Microstructure and Mechanical Properties of Melt-quenched Mg-Gd-( Ni, Cu or Zn) Alloys K. Matsuzaki, M. Takahashi, T. Sano, Mechanical Engineering Laboratory, Tsukuba (J)..............................................................................................................................635 Simulation of Open-cell Magnesium Foams under Dynamic Loading K. Shimojima, M. Mabuchi, Y. Yamada, C. E. Wen, Y. Chino, M. Nakamura, T. Asahina, Materials Processing Department, National Industrial Research Institute of Nagoya (J); T. Aizawa, Department of Metallurgy, The University of Tokyo, (J); K. Higashi, Department of Metallurgy and Materials Science, Osaka Prefecture University, Gakuen-cho, Sakai, Osaka (J)..............................................................................639
XVII
Processing of Cellular Magnesium Alloy Y. Yamada, C. Wen, K. Shimojima, M. Mabuchi, M. Nakamura, T. Asahina, National Industrial Research Institute of Nagoya (J); T. Aizawa, The University of Tokyo, Tokyo (J); K. Higashi, Osaka Prefecture University, Osaka (J)..............................................................645 Semi-solid Forming of New Mg-Zn-Al-Ca Alloys S. Kamado, N. Ikeya, R. Suhardi Rudi, T. Araki, Y. Kojima, Department of Mechanical Engineering, Nagaoka University of Technology, Nagaoka (J) .............................................651 Physical Properties Damping in Magnesium and Magnesium Alloys W. Riehemann, Institut für Werkstoffkunde und Werkstofftechnik der TU Clausthal (D)......659 Thermal Diffusivity and Thermal Conductivity of Mg Alloys A. Rudajevová, P. Lukác, Department of Metal Physics, Charles University, Praha (CZ)..............................................................................................................................665 Experimental Determination of Interfacial Heat Transfer Coefficient for AZ91E Castings S. Bergeron, M. Lepage, J. Renaud, INTERMAG Technologies Inc., Sainte-Foy (CND); D. Dubé, V. Lambert, Dept. of Mining and Metallurgy, Laval University, Sainte-Foy (CND) ...................................................................................................................671 Stress Relaxations in a Magnesium Alloy and Composite Z. Trojanová, P. Lukác, Department of Metal Physics, Charles University, Praha (CZ); W. Riehemann, Department of Materials Engineering and Technology, Technical University Clausthal (D); F. von, Volkswagen AG Wolfsburg (D) ..................................................................................678 Creep Behaviour Tensile and Compressive Creep Behavior of Magnesium Die Casting Alloys Containing Aluminum S. R. Agnew, S. Viswanathan, E. A. Payzant, Q. Han, K. C. Liu, E. A. Kenik, Oak Ridge National Laboratory, Oak Ridge, Tn (USA) ...........................................................................687 Creep of Mg-Zn-Al-Alloys M. Vogel, O. Kraft, E. Arzt, Max-Planck-Institut für Metallforschung, Stuttgart (D); E. D. Reese, R. Rauh, DaimlerChrysler AG, München (D)....................................................693 Creep Behavior and Deformation Substructures of Thixomolded Mg-Al-Ca Alloys M. Suzuki, J.-I. Koike, K. Maruyama, H. Kubo, Tohoku University, Tohoku (J); T. Tsukeda, K. Saito, The Japan Steel Works, Ltd. (J)............................................................699 Creep and Bolt-Load Retention of Sand Cast Elektron MEZ C. J. Bettles, CSIRO Manufacturing Science and Technology, Victoria (D); M. S. Dargusch, The University of Queensland (AUS)...........................................................705
XVIII The Microstructure and Creep of an Extruded Mg-Y-Nd Alloy R. Azari-Khosroshahi, Semnan University, Semnan (IR) .......................................................711 Microstructure and Creep Properties of Die-cast Mg-Al-base Alloys AZ91and AS21 P. Zhang, R. Agamennone, W. Blum, Institut für Werkstoffwissenschaften, Universität Erlangen-Nürnberg, Erlangen (D); B. von Grossmann, H.-G. Haldenwanger, Audi AG, Ingolstadt (D) ......................................716 Recycling, Melting, Environmental MagnesiumMelting / Casting and Remelting in Foundries H. W. Dörsam .........................................................................................................................725 The Impact of Metal Cleanliness on Mechanical Properties of Die Cast Magnesium Alloy AM50 P. Bakke, K. Pettersen, S. Guldberg, S. Sannes, Norsk Hydro Research Centre, Porsgrunn (N) .........................................................................................................................739 Remelting and Cleaning of Magnesium Scrap U. Galovsky, M. Kühlein, Leichtmetall Kompetenzzentrum Ranshofen (A) ...........................746 Utilization of Residues fromFluxless Remelting of Compact Magnesium Scrap A. Ditze, C. Scharf, Institut für Metallurgie, Technische Universität Clausthal (D)..............752 Use of SO2 as Protection Gas in Magnesium Diecasting W. Schubert, LM LeichtMetall-Systemtechnik GmbH, Fellbach (D); H. Gjestland, Hydro Magnesium, Porsgrunn (N)...................................................................761 Impurities in Magnesium and Magnesium Based Alloys and their Removal H. Singh Tathgar, T.A. Engh, Norwegian University of Science and Technology, Trondheim (N); P. Bakke, Norsk Hydro Research Centre, Porsgrunn (N) ......................................................767 Simulation An Approach to Determine Solidification Curves of Commercial Magnesium Alloys D. Mirkovic, J. Gröbner, R. Schmid-Fetzer, Institute of Metallurgy, Technical University of Clausthal (D).....................................................................................................783
Author Index .........................................................................................................................789 Subject Index.........................................................................................................................793
Alloy Development
Global Overview on Demand and Applications for Magnesium Alloys Robert L. Edgar Hydro Magnesium, Brussels
1
Introduction
Magnesium and its alloys are becoming widely recognized as playing an increasingly important role in automotive and electronic consumer products. This paper will identify where the growth is coming from and where it could be expected to come from in the future. As well, it will highlight some of the critical success factors for proliferating and broadening the number of applications for magnesium alloys in the future. Magnesium alloys are used in a variety of applications based on the following processes: 1) Liquid and semi-solid molding processes: - high and low pressure die casting - gravity casting - Thixomolding® - squeeze casting 2) Wrought forming processes: - extrusion - rolling of sheet and plate - forging Using statistics from the International Magnesium Association (IMA), it can be deduced that more than 90% of the shipments of alloy magnesium are consumed in the manufacture of high pressure die castings (Figure 1). Over the next 10 years, high pressure die casting is expected to increase in importance relative to other processes. This paper will therefore deal mainly with the situation relating to high pressure die casting (referred to as "die casting" from hereon), but will also discuss the status of the other processes.
2
Die Casting
Figure 2 illustrates that until recently most of the growth of demand for die castings has taken place in North America, averaging 26% per year since 1991. However, Europe is rebounding after many years of decline following curtailment of production of the Volkswagen Beetle. This time the development is more wide-spread, driven by ever tightening legislation dictated by environmental concerns, particularly with respect to automobiles. Figure 2 also projects industry growth to 2005. North America and Europe are by far the largest contributors over this period. In fact, Europe is projected to be a larger consumer of magnesium die castings by 2001/2002. Developments in Japan, Korea and elsewhere are slow in coming, but in the author's opinion have the potential to develop faster than shown. Projected growth for the Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
4 period 2000 – 2003 is based analysis of an internal database built up by Hydro Magnesium and containing more than 1000 different automotive die cast applications.
120000 100000 000 tons
80000 60000 40000 20000 0 1990
1992
1994
1996
1998
Year High pressure die
All other processes
Figure 1: Demand for alloy magnesium
Automotive die castings make up 85% to 90%, of the total annual weight of die castings produced, and are expected to retain this share for the foreseeable future. Table 1 lists many of the automotive castings currently in use in motor vehicles as well as some potential new applications presently under development. One can see that in principle, magnesium alloy die castings can be found throughout a motor vehicle. In practice each vehicle contains a limited range of applications, and consequently there is considerable scope for proliferation. Based on the growth forecast in figure 2 and annual production of 50 million vehicles, magnesium content per vehicle is expected to rise from the present 3 kg to about 6 kg in 2005. This is still a modest quantity, but as table 2 shows, there are several notable examples of vehicles with contents well above the present average, ranging up to 26 kg. Recent growth has been fueled by the AM alloy series. These alloys possess excellent energy absorbing properties, permitting use in safety related applications such as steering wheels, instrument panels and beams, seat structures, brackets and inner door panels. While automotive castings dominate in terms of total weight of magnesium consumed, certain non-automotive applications are being made in increasing numbers with considerable added value. Most require high quality surface finish and are usually painted. These applications are generally very thin walled and relatively small in size, and are most commonly found in laptop computers as well as in hand held devices such as cell phones and various types of cameras. Here, magnesium brings a unique combination of advantages versus plastic, namely 100% recyclability, good heat dissipation, electromagnetic and RF shielding and compactness arising from thin walls supported excellent physical properties. It is worth pointing out that such products are being produced as well by Thixomolding (see below).
5
200 180
'000 tons castings
160 140 Other
120
Japan & Korea
100
Europe
80
North America
60 40 20 0 91
93
95
97
99
01
03
05
Year Figure 2: Demand for magnesium die castings
3
Thixomolding®
Thixomolding is the high-speed injection molding of semi-solid thixotropic alloys. Mechanical shearing of the semi-solid metal generates a thixotropic structure that allows these materials to be molded utilizing a process similar to plastic injection molding, eliminating a number of environmental issues associated with die casting. Unlike die casting, the process does not require separate melting and transfer systems for handling of molten metals. One potential disadvantage is the cost of granulated alloy feedstock relative to the alloy ingot used in the die casting process. Thixomolding is attracting particular interest from plastic injection molders who see similarities between Thixomolding and their own process. Traditional die casters have not yet embraced Thixomolding as an alternative process. Properties for well made Thixomolded products are generally similar to those of well made die castings, although claims are being made for lower porosity and thinner walls in Thixomolded products. Certain injection molding are now being replaced by Thixomolded magnesium products and, as well, by die castings. There are now 150 Thixomolding machines installed world wide , mostly in Japan and Taiwan, but with a growing number in North America. Activity in Europe is still in an early stage of development.
6 Table 1: Applications of die castings in motor vehicles Interior Exterior Seat components Engine hood Instrument panels & beams Roof panels* Knee bolsters Rear deck lid* Steering column components Road wheels* Steering wheels Radiator support* Pedal brackets Grill opening support* Air bag retainers Drivetrain Pedal brackets Manual transmission housing Radio heat sink/frame 4WD transfer case Radio HVAC covers Automatic transmission housing* Sunroof components Engine Mirror brackets Cylinder head covers Headlight retainers Intake manifolds Inner door frames Accessory drive brackets Support pillars* Electrical connectors Engine block* Oil Pan* Starter-alternator housing*
4
Critical Success Factors for Growth
The prospects for growth of the use of die castings, especially for automotive applications are particularly attractive. However, the actual outcome is not a forgone conclusion. Cole(1) has thoroughly reviewed the critical elements for success to remain competitive and support growth. One of these is the cost Metal cost is being improved by implementing operational refinements, increasing plant scale and introducing new technology. The trend of declining real prices in the 1990’s can be expected to continue, as long as new greenfield plants or brownfield expansions with improved performance can be brought on stream. Growth will be required to justify such investments, with the current outlook being about 5% per annum. Table 2: Vehicle platforms with high magnesium content Vehicle platform kg per vehicle GM Full Sized Vans - Savana & Express up to 26,3 kg Daimler Benz SL 17,0 to 20,3 kg GM Minivans - Safari & Astro up to 16,7 kg Ford F-150 Truck 14,9 kg VW Passat, Audi A4 & A6 13,6 to 14,5 kg Porsche Boxster Roadster 9,9 kg Buick Park Avenue 9,5 kg Alfa Romeo 156 9,3 kg Daimler Benz SLK Roadster 7,7 kg Chrysler Minivans 5.8 kg
7 Recycling costs have the potential to decline more quickly than the cost for primary metal. In a die casting industry expected to grow at 10% to 15% annually, recycling volumes will grow likewise, with economies of scale driving down costs. Furthermore, die casters are expected to increasingly reach a scale of operation allowing cost effective in-house recycling of class 1 returns, eliminating the freight expense for recycling externally in the same way aluminium die casters do today. Possibly the most dramatic cost improvement potential rests with component design. General Motors have successfully introduced magnesium instrument panel substrate beams into their North American cars and trucks. One early unit weighed as much as 12.4 kg, but an updated version will be 14% lighter with the same functionality. More recent versions for other GM vehicles weigh about 6 kg, or about half the weight of the first generation, with enhanced functionality. Volkswagen’s magnesium manual transmission housing introduced in 1996, weighed 13.7 kg. A more recent magnesium version is more than 30% lighter with much higher performance requirements in terms of allowable torque and power. In both examples significantly less magnesium is purchased, and less is recycled. Die casters are constantly improving their general operations and bring costs down. In an industry growing at 10% to 15% per year, increasing scale of operation is a prominent feature helping to average down fixed costs. The true cost of a product over its life cycle is receiving increasing attention(2). Industry participants should be sensitive to the positive and negative factors describing impact magnesium’s contribution. Implementation of magnesium into motor vehicles can have a net positive environmental effect compared to the materials and components it replaces. In this regard, die casters and producers alike should be taking steps to eliminate SF6 from their operations. This gas has a strong undesirable green house effect, 23,900 times that of an equivalent quantity by weight of carbon dioxide, and its elimination will allow users to focus on the positive contributions from magnesium die castings.
5
Observations and Conclusions
1. The die casting industry is the driver for growth of demand for magnesium alloys. Annual growth is expected to average 10% to 15% over the next six years, in a general magnesium market growing at about 5%. 2. The collective demand for magnesium alloys other than for die casting is expected to exhibit very modest if any growth over the next several years. One potential exception is Thixomolding, which is attracting considerable interest as an alternative for converting plastic injection moldings used in portable devices to magnesium. 3. The automobile industry is and will continue to be the most important consumer of alloy magnesium. 4. Growth of metal demand will create opportunities for expansion of metal production and consequent reduction of cost for some producers. 5. The high rate of growth of individual die casters will create important scale advantages, leading to lower cost and an increasingly sophisticated approach to casting design. 6. Component design shows considerable potential for cost reduction. 7. To sustain growth, the industry will need to focus much more of its collective resources to make magnesium alloys more usable for automotive designers
8
6
References
[1] Gerald Cole; Proceedings of the 56th Annual Meeting of the International Magnesium Association, June 1999 [2] D.L. Albright and J.O. Haagensen; Proceedings of the 54th Annual Meeting of the International Magnesium Association, June 1997
Focused Development of Magnesium-Alloys Using Computational Thermochemistry as a Powerful Tool Rainer Schmid-Fetzer, Joachim Gröbner, Dmytro Kevorkov Technical University of Clausthal, Institute of Metallurgy, Germany
1
Introduction
Computational thermochemistry enables not only the calculation of multicomponent phase diagrams. Furthermore the tracking of individual alloys during heat treatment or solidification by calculation of phase distributions and phase compositions is possible. These are the basic data to understand and control behavior of any novel or modified Mg-alloy. In traditional alloy development long-term experimental investigations with numerous samples of different alloy compositions are performed. Once three- or multicomponent alloys are considered, the selection criteria for alloying elements and their compositions become diffuse in a traditional approach. Computational thermochemistry provides a clear guideline for such selections and helps to avoid long-term experiments with less promising alloys. Thus it is a powerful tool to cut down on cost and time during development of Mg-alloys.
2
Database Development
Development of thermodynamic data for multicomponent alloys requires a combination of experiments and computational thermochemistry with data from alloy application (Figure 1). Since numerous binary and ternary subsystems have to be treated before multicomponent alloys can be calculated, this development becomes a long-term project. In our group at the TU Clausthal a thermodynamic database for several promising alloying elements like Al, Li, Si, Mn, Ca, Sc, Y, Zr and Rare Earth elements, is currently under construction. Long term project: Multicomponent Magnesium Alloy Database Applications
Experiments
Thermodynamic database
Figure 1: Corner stones of magnesium alloy database
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
10 To create the thermodynamic phase descriptions the Calphad method is used. The principle of this method is shown in Figure 2.
Models
Software
Stable & Metastable Phase Equilibria
Process Simulation
Data Key Experiments
Direct Application
Figure 2: Schematic approach of database development and applications
Each phase is described in a suitable model with parameters optimized by available experimental data. This data set is used to calculate stable and metastable equilibria for process simulation or direct applications. The calculations are improved continuously by new experimental data coming from application and were checked by key experiments to improve their reliability. An example for Process simulation is the coupling with solidification models. An example for a Direct Application is given in chapter 3. In Mg-systems the experimental literature data were very sparse in both, phase diagram data and thermodynamic properties. Therefore several experimental methods are being applied in our group to produce a sufficient experimental database (Figure 3).
Experimental Methods Sample Preparation Levitation Melting Arc Melting Electron Beam Melting Reaction Sintering
Sample Analysis Thermal Analysis (DTA,STA) X-ray Diffractometry (XRD, 25-1500°C) Electron Microscopy (SEM/EDX/WDX) Metallography, Optical Microscopy
Figure 3: Experimental methods
Our work focuses on ternary and quaternary systems for improvement of creep resistance [1], thermal stability (Mg-Mn-Sc-RE, Mg-Al-Ca-RE) and density reduction (Mg-Li-Al-Si, Mg-Al-Ca-Si). The thermodynamic and technical application of the ternary Al-Mg-Sc is already discussed intensively in literature [2, 3]. Here the quaternary system Mg-Mn-Sc-Gd is shown as example for alloy selection using computional thermochemistry.
11
3
Alloy Selection in the Mg-Mn-Sc-Gd System
Investigations started with binary Mg-Sc alloys. Scandium was chosen for hardening by ageing because of its large solubility in Mg and the retrograde solubility at lower temperatures. Binary MgSc precipitates were formed very slowly during ageing and improved the mechanical properties only slightly because of their incoherent interface. Therefore, Mn was added as second alloying element. The precipitation of Mn2Sc was predicted by thermodynamic calculations. Mn2Sc precipitations form coherently and were found useful for improving creep resistance, hardness and strength of Mg alloys. New MgSc15Mn1 or MgSc6Mn1 alloys show about 100 times better creep-resistance than best commercial WE43 alloy at 350°C and 30 MPa [1]. Further improvement of the properties and cost reduction of high-price Sc metal initiated a search for additional alloying elements. Gd, Y and Zr were considered for this purpose to achieve a large quantity of suitable precipitations to improve mechanical properties using a minimum of expensive alloy element addition. These element combinations Mg-Mn(Sc, Gd, Y, Zr) form a variety of quaternary systems and within those there is a huge amount of possibilities to select alloy compositions. Therefore phase diagram calculations were performed to give hints for selecting promising candidates. Here we report only for the MgMn-Gd-Sc as an short example.
L (liquid)
L+ Mn2Sc
Temperature [°C]
L + (Mg) (Mg) (Mg) + Mn12 Gd
(Mg) + Mn2Sc
(Mg) + Mn23Sc6
(Mg) + Mn 12Gd (Mg) + Mn2Sc + + GdMg5 GdMg5 + Mn23Sc6 + Mn23Sc6 5.0 Gd 1.0 Mn 94.0 Mg 0.0 Sc
(Mg) + Mn 2Sc + GdMg5
wt.% Sc
(Mg) + Mn2Sc + GdMg5 + MgSc 5.0 Gd 1.0 Mn 93.0 Mg 1.0 Sc
Figure 4: Phase diagram section with constant 1 wt.% Mn, 5 wt.% Gd from 0- 1 wt.% Sc
Several vertical sections in the ranges of 0-1.5 wt.% Mn, 0-10 wt.% Sc and 0-10 wt.% Gd were studied. The calculated quaternary phase equilibria are presented in two-dimensional sections with constant Mn and Gd content. Figure 4 shows a T-x section with constant 1 wt.% Mn, 5 wt.% Gd from 0- 1 wt.% Sc. A large one-phase field of (Mg) and several different solid phases stable at lower temperatures can be seen. At this point the question arise how to identify promising alloy candidates from all these calculated diagrams. What phase diagram features are related to what alloy processing steps?
12 What is needed is a “priority list” of beneficial phase diagram features, derived from the relevant alloy processing steps. The most important points are given in Table 1. Table 1: Beneficial phase equilibrium features and their relevance for (Mg)-alloy processing Phase diagram feature Relevance for alloy processing Liquidus temperature (or ), whether the primary slip plane is basal (e.g. cadmium, magnesium, or zinc) or prismatic (e.g. titanium or zirconium). Even if both slip planes operate, there is still no way to accommodate strains along the c-axis. Deformation twinning can help2,3; however, twinning modes are unidirectional. Because there is only one major twinning mode in magnesium, the { 10 1 2 } tension twin, the crux of the problem is c-axis compression. If a grain is forced to deform in this fashion, a large plastic mismatch with neighboring grains will develop and fracture will ensue. Stohr and Poirier4 and Obara et al.5 identified this problem and performed careful experiments of c-axis aligned single crystal compression. Optical microscopy revealed slip occurring on { 1122 } planes in a 1/3< 1 1 23 > direction, also known as . Burgers vector analysis by transmission electron microscopy (TEM) confirmed the slip plane and direction. However, the critical resolved shear stress (CRSS) of this slip system was shown to be 40-80 times that of the basal slip system at room temperature. In addition to focusing on c-axis compression, we also draw special attention to crystallographic texture. When materials are plastically deformed and the accommodation occurs by crystallographic mechanisms, the crystallites tend to rotate to preferred orientations. This texture is a “fingerprint” of the imposed deformation geometry and the deformation modes that sustained the plasticity. The objective of this study is to use deformation texture data to probe the effects of solid solution alloying on the balance between the deformation modes. The results should help to guide efforts develop wrought magnesium alloys with better forming characteristics. There are two principles involved: (i) Alloying can change the lattice parameters (i.e. c/a ratio) and electronic structure, resulting in a change in the relative activities of the different deformation modes. (ii) By activating different deformation modes, the crystallographic texture that results from deformation will be altered.
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
120
2
Experimental Procedures
Magnesium solid solution alloys with lithium (1, 3 and 5 wt%) and yttrium (1 and 3 wt%) were cast into a ø25 x 75 mm copper mold inside a glove box with a protective cover gas (Ar + 1.5% SF6 + 0.5%O2). A resistive furnace was used to melt the alloys in mild steel crucibles at 750-850°C. The as-cast microstructures varied from large grained and dendritic, to fine grained and equiaxed (~10 µm). In order to compare alloys with similar microstructures, thermo-mechanical treatment schedules were developed to obtain grain sizes of 25-100 µm (the details are reported elsewhere6.) A servo-hydraulic load frame (MTS 810) was used for compression testing and deforming the alloys for deformation texture comparisons. Mechanical behavior of the alloys was determined using compression samples (ø6.4 x h9.6 mm). Samples, ø25 x 3.5 mm, were compressed to a true strain of ~25% for texture comparisons. Additionally, a 12.7 mm channel die was used for deforming samples in plane strain. This lower symmetry deformation, which simulates rolling, made subsequent texture comparisons clearer. Each of the tests were conducted at a constant displacement rate, with an initial strain rate of 5x10-3s-1. The crystallographic texture was measured with x-ray diffraction using the reflection method. Pole figures were obtained from ( 10 1 0 ), ( 0002 ), and ( 10 1 1 ) Bragg peaks. Complete orientation distributions were calculated using the WIMV method of the popLA code7. Complete recalculated ( 0002 ) and ( 10 1 0 ) pole figures are presented. Before the pole figures were collected, standard θ -2θ scans were collected to obtain the positions of the Bragg peaks. These data revealed lattice parameter changes induced by alloying. In short, both Li and Y additions decrease the c/a ratio from that of pure Mg (1.624), to minimum c/a ratios of 1.608 and 1.619 for 5 wt% Li and 3 wt % Y, respectively. Few elements have this effect on magnesium’s lattice. The approach outlined above for determining deformation mode activity was verified using TEM dislocation analysis. Mg-5wt% Li compression samples were with a significant population of grains having their c-axis close to the compression axis. These polycrystalline samples, were compressed to plastic strains, εp = 1 and 3%, to introduce a low density of dislocations suitable for TEM. Thin sections containing the compression axis were cut from the compressed samples using electro-discharge machining. Final thinning was performed using a double-jet electropolisher (Struers Tenupol) with a solution of 450 ml ethanol, 50 ml perchloric acid, and 25 ml butyl cellosolve cooled to temperatures less than -30°C.
2.1 Mechanical Test Results The compression test results are shown in Figure 1. All samples failed catastrophically along a shear-type crack. The alloys’ ductilities were dramatically improved over the pure magnesium, and the ultimate strengths were greater as well. The lithium containing alloys continued to harden up to failure, indicating a resistance to plastic instability that is critical for material deformed under less forgiving conditions (e.g. tension).
121
250
Stress (MPa)
200 150 pure Mg 1 wt% Y 3 wt% Y 3 wt% Li 5 wt% Li
100 50 0 0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Strain Figure 1: True stress vs. true strain curves obtained from compression tests of the alloys
Hauser et al.8 noted a “knee” in their tensile flow curves in the lithium containing alloys not present in any of the other alloys they tested. They speculated that the rapid decrease in the hardening rate coincided with the general activation of a non-basal slip mode (prism slip.) General plasticity through non-basal slip had previously only been observed in magnesium deformed at high temperatures. Secondary deformation modes may indeed enable these alloys to accommodate significantly larger strains prior to failure, however, the current results focus on another mode, slip. 2.2
Experimental and Simulated Texture Results
All the samples show an overall tendency for a basal compression texture. However, as reported earlier6, the peak texture component is off basal. The primary deformation modes, basal slip and tension twinning, generate a basal texture and another mode is necessary to develop the final off-basal texture. The plane strain compression textures clearly show a difference between the alloys (Fig. 2). Although there is a tendency to align basal poles with the “normal direction” (ND) there is a spreading towards the “rolling direction” (RD) that increases from pure Mg to the Y alloys to the Li alloys. In fact, the basal poles have clearly rotated away from ND ~20° towards RD for 3 wt% Li. Methods of deformation texture simulation are summarized elsewhere (Kocks et al.9). Recent advances, such as techniques to explicitly incorporate the effects of strong single crystal plastic anisotropy and deformation twinning10, were essential in this study. For the current case, the strength of the different deformation modes were normalized to CRSSbasal = 1. The effects of prism slip, pyramidal slip, and tensile twinning were investigated. A range of CRSS values from 1 to 24 were explored for each mode, with guidance provided by single crystal CRSS values in the literature. It was found that prism slip played little role in the texture evolution of the current alloys. Basal slip and tension twinning dominated the straining with a strength ratio of 1:2. The strength of the slip mode varied significantly between the different alloys. Figure 3 shows the best matches for the experimental data, with CRSS = 6, 5 and 3 for pure Mg, 1 wt% Y and 3 wt% Li, respectively. A strain rate sensitivity of 0.05 was assumed for each deformation mode. The
122 increased activity in the alloys helps to explain their improved ductilities since c-axis compression can be accommodated.
Figure 2: Plane strain compression textures (εp ~ 25%) as shown by complete recalculated ( 0002 ) and ( 10 1 0 ) pole figures for pure Mg, 1 wt% Y, and 3 wt% Li. (Equal area projections with “rolling direction” right and “transverse direction” down.)
Figure 3: Simulated plane strain compression textures (εp ~ 25%) as shown by complete ( 0002 ) and ( 10 1 0 ) pole figures for CRSS = 6, 5 and 3. (Discrete textures were smoothed over 10°)
It has long been known that Mg and its alloys exhibit ND-RD plane splitting of the peak intensity in the basal pole figure after plane strain deformation such as rolling11. Wonsiewicz and Backofen12 suggested that it might be due to a secondary twinning within the primary tension twins. The current solution based on activity has not, to the authors’ knowledge, been proposed before.
123 2.3
TEM Dislocation Analysis
The details of this analysis will be presented in a paper in preparation, only the highlights will be presented here. The g · b = 0 and g · b × u = 0 invisibility criteria and the case of g · b = 2 contrast doubling were used to identify the different dislocations. All of the samples had large numbers of dislocations. In addition, there were dislocations parallel to the basal plane with strong contrast for g = [0002]* (* indicates “reciprocal lattice vector” after Thomas and Goringe13), indicating a c component to their Burgers vector (Fig 4a.) They are visible for g = [ 1120 ]*, [ 10 1 1 ]*, and [ 10 1 1 ]* and invisible for g = [ 10 1 0 ]*, [ 1 101 ]* and [ 01 1 1 ]*.These dislocations were identified as b = 1/3[ 12 13 ].
Figure 4: TEM images show dissociated edge dislocation that appears faulted under some diffraction conditions (i.e. g = [ 10 1 1 ]*). Note dislocation invisible for g = [0002]* condition
For some of the diffraction conditions, these dislocations appeared faulted (Fig. 4b). The extinction distance for Mg is long, therefore, fault contrast is difficult to observe. The change in apparent fault width while tilting along a ( 10 1 1 ) band is consistent with faulting on the ( 121 2 ) slip plane. Two possible dissociation reactions are as follows: 1 /3[ 12 13 ] → 1/6[ 2203 ] + SF(1212) + 1/6[ 0223 ] (1) /3[ 12 13 ] → 1/3η[ 12 13 ] +
1
SF(121 2 ) +
/3(1-η)[ 12 13 ]
1
(2)
Stohr and Poirier4 concluded from their TEM analyses that the dissociation of edge dislocations is consistent with Eq. (1), which was origianlly proposed on the basis of a hardsphere model14. There is a problem with their analysis because they suggest that the dissociation occurs on the basal plane, while this reaction would actually occur on the slip plane. The co-linear splitting of Eq. (2) was obtained by atomistic simulation studies in hcp metals15,16. Recent work on the generalized stacking faults17 confirms the earlier result of a symmetric dissociation with η== 1/2. In the current situation, the partials exhibit the same invisibilities (and visiblilities) as the original 1/3[ 12 13 ] would, with both partials simultaneously in or out of contrast for all of the diffraction conditions under investigation. The Shockley type 1/6< 2023 > partials should exhibit some contrast asymmetry in the g = < 10 1 1 >* and < 112 0 >* diffraction conditions. This is the first report of experimental evidence for the reaction of Eq. (2).
124
3
Acknowledgments
Research supported by the LDRD Fund at the ORNL, operated by UT-Battelle, LLC, for the U.S. Department of Energy under contract DE-AC05-00OR22725. SRA is currently a Eugene P. Wigner Fellow at the laboratory. The authors also thank C. A. Carmichael, E. D. Specht, and J. W. Jones for assistance with alloy preparation, texture measurement, and TEM sample preparation, respectively.
4
References
[1] F. Hehmann, in Magnesium Alloys and Their Applications, eds. B. L. Mordike and F. Hehmann, Verlag, Germany, 1992, p. 21-28. [2] U. F. Kocks and D. G. Westlake, Trans. Metall. Soc. AIME, 239, 1967, 1107-1109. [3] M. H. Yoo, Metall. Trans. A, 12A, 1981, 409-418. [4] J. F. Stohr and J. P. Poirier, Phil. Mag., 25, 1972, 1313. [5] T. Obara, H. Yoshinga and S. Morozumi, Acta Metall., 21, 1973, 845. [6] J. S. Kallend, U. F. Kocks, A. D. Rollet and H.-R. Wenk, Mat. Sci. Eng., A132, 1991, 111. [7] S. R. Agnew and M. H. Yoo, Magnesium Technology 2000, Eds. H. I. Kaplan, J. Hryn, and B. Clow, TMS, Warrendale, PA, 2000, p. 331-340. [8] F. E. Hauser, P. R. Landon, and J. E. Dorn, Trans. ASM, 48, 1956, 986. [9] U.F. Kocks, C.N. Tomé and H.-G. Wenk, Texture and Anisotropy, Cambridge, 1998. [10] R. A. Lebensohn and C. N. Tomé, Acta Metall., 41, 1993, 2611-2624. [11] C. S. Roberts, Magnesium and Its Alloys, New York, John Wiley, 1960, p. 180-193. [12] B. C. Wonsiewicz and W. A. Backofen, Trans. AIME, 239, 1967, 1422-1431. [13] G. Thomas and M. J. Goringe, Transmission Electron Microscopy of Materials, John Wiley, New York, 1979, p. 340-347 [14] F. C. Frank and J. F. Nicholas, Phil. Mag. 44, 1953, 1213. [15] Y. Minonishi, S. Ishioka, M. Koiwa, and S. Morozumi, Phil. Mag. A, 45, 1982, 835. [16] M. H. Liang and D. J. Bacon, Phil. Mag. A 53, 1986, 181. [17] J. R. Morris, J. Scharff, K. M. Ho, D. E. Turner, Y. Y. Ye, and M. H. Yoo, Phil. Mag. A, 76, 1997, 1065.
Microstructure, Texture and Residual Microstrains in MgAl8Zn Deformed at Very High Strain Rates A.Pyzalla [1], M.Brodmann [2], P.L.Lee [3] and D.Haeffner [3] [1] Abt. Werkstoffe, Hahn-Meitner-Institut, Glienicker Straße 100, D - 14109 Berlin, Germany [2] Lehr- und Forschungsgebiet Werkstoffkunde, RWTH Aachen, Augustinerbach 4, D-52062 Aachen, Germany [3] Advanced Photon Source, Argonne National Lab., Argonne, IL 60439-4856, USA
1
Abstract
The influence of the degree of natural strain and the strain rate on the microstructure, the texture and the residual microstrains after compressive deformation of MgAl8Zn is studied using X-rays and high energy synchrotron radiation.
2
Introduction
Cold forming processes besides establishing the desired shape induce microstructural changes, texture and residual stresses into the workpiece. These affect the formability of semi-finished material as well as the mechanical properties and the corrosion resistance of the formed product. While the formation of residual macrostresses in cold and hot formed products have been the subject of a number of experimental and FEM studies the formation, the distribution and the value of residual microstresses have not been studied as intensively. The residual microstrains and microstresses evolve as a consequence of texture formation, since the differences in the mechanical properties of the crystal orientations lead to microstrains and microstresses between adjacent grains. These residual microstrains amd microstresses are small with respect to their range but not with respect to their magnitude, thus they often have a strong impact on the fatigue and the corrosion resistance of a heavily deformed material. The aim of the experiments described here is the determination of the influence of the deformation ratio and the deformation velocity on the texture and the residual microstains in the magnesium alloy MgAl8Zn, also commonly referred to by AZ80.
3
Sample Preparation
Dynamic compression tests with strain rates in the range of 2400 s-1 ≤ ε& ≤ 3400 s-1 were carried out on the samples at the Lehr- und Forschungsgebiet Werkstoffkunde of the RWTH Aachen using a Split-Hopkinson-Pressure-Bar /1/. The arrangement consists of two bars, the input and output bar, with a cylindrical specimen between their faces. The specimens with a ratio between the initial height and diameter H0 / D0 = 1 were machined with their load axis Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
126 parallel to the extrusion direction of the initial material AZ80A-F. A pneumatically accelerated projectile hits the free end of the input bar and induces an elastic mechanical wave running through the bar. At the other end of the input bar, a part of the stress wave is transmitted through the specimen as a plastic wave, because of the smaller cross section of specimen, and then introduced into the output bar again as elastic wave. Reflections take place at the free end of the output bar and at the specimen. In order to limit the deformation, an additional ring with a specific height is placed around the specimen acting as a stopper. The deformation was stopped at a nominal global strain of 5 %.
4
Experimental Details of Sample Characterization
Only diffraction methods allow accessing the residual microstrains in different oriented crystallite groups, usually by analysing a multitude of reflections. In order to avoid surface effects the experimental determination of residual microstrains which are caused by texture formation has to be performed in the bulk of the material. For achieving a sufficient penetration depth either neutrons or high energy synchrotron radiation have to be used. A rapid and simultaneous access of the texture and the residual microstrains is possible by a recently developed method using white high energy synchrotron radiation, where the whole diffraction spectrum is available at the same time /2/. The measurements were carried out at the beamline 1IDB of the Advanced Photon Source at Argonne National Lab., Argonne, USA using white high energy synchrotron radiation. By choosing a diffraction angle 2θ = constant = 7°, inserting an 100 µm slit in the incoming beam and inserting an 100µm slit in the diffracted beam, the gauge volume is a parallelepiped with a length of 900 µm and a width of 110 µm, approximately. The gauge volume was thus small enough to be kept completely within the sample. A typical spectrum obtained in the direction parallel to the sample axis, χ = 90° (in the direction of the compressive load), respectively in transversal direction, χ = 0°, is given in fig.1. 90000
10-11
Intensity [counts/300s]
80000
χ = 0°, transversal direction χ = 45° χ = 90°, longitudinal direction
70000 60000
11-20
50000
10-12
40000
0002
30000 20000 10000
10-13
10-10
0
35
40
45
50
55
60
65
70
Energy [keV] Figure 1: Spectrum, sample after 5% compressive deformation ,deformation rate 3349/s
127 From the energy spectrum, after identifying the reflection hkl, the lattice spacing dϕψhkl can be calculated according to Bragg’s law, which can be written as a function of the energy Eϕψhkl: 1 1 hc hkl (1) = ⋅ = const. ⋅ d ϕψ hkl 2 sin θ hkl Eϕψ
hkl Eϕψ
hkl denote the Miller’s indices, θhkl is the Bragg angle, h is Planck’s constant and c is the velocity of light. The strain εϕψ the sample orientation ϕ, ψ is evaluated from the shift of the lattice spacing : hkl = ε ϕψ
hkl E0 d ϕψ − 1 = hkl −1 d0 Eϕψ
(2)
d0 denotes the lattice parameter of the stress-free material and E0 is the corresponding energy value. Strain measurements in different directions of the samples allow for the determination of the strain tensor ε and the εϕψ - sin2ψ - curves which are a measure for the microstrains caused by the texture evolution. The stress tensor σ can be calculated from the strain tensor ε using Hooke’s law taking into account the Diffraction Elastic Constants (DEC).
5
Results
The microstructure of an MgAl8Zn sample after 5% deformation at a deformation rate of dε/dt = 3349 1/s is shown in fig. 2. The microstructure shows unequal grain sizes which originate from the hot extrusion during the manufacturing of the samples. Within several grains precipitates are visible. But, the amount of these precipitates is not high enough for significant contributions to the diffraction spectra obtained using X-rays and high energy synchrotron radiation. In the micrographs corresponding to the longitudinal cut of the samples the shortening of the grains due to the compressive deformation of the samples is visible.
Figure 2: Microstructure in longitudinal and transversal direction, 5% strain, strain rate 3349/s
After the extrusion process, before the compressive deformation, a - fiber texture is present in the samples. This fiber texture, which is typical for compression of hexagonal materials at ambient temperature, increases in strength with increasing compressive deformation of the samples. X-ray pole figure analyses as well as high energy synchrotron radiation data reveal that the texture strength depends not only on the degree of natural strain imposed on the samples but,
128 also on the deformation rate. Quantitative analyses of X-ray data show that the intensity of the - fiber decreases with increasing deformation speed (fig. 3).
Intensity of 0001 fiber component [mrd]
11,0 10,5 10,0 9,5 9,0 8,5
5% - deformation
8,0 2400
2600
2800
3000
3200
3400
deformation rate d ε/dt [1/s] Figure 3: Intensity of the - fibre for different deformation rates
intensity [counts/300s]
In agreement with X-ray data the energy spectra obtained in the high energy synchrotron diffraction experiment reveal a decrease in the intensity of the 0002 reflection with increasing speed of the compressive deformation (fig.4). 3,0x10
4
2,5x10
4
2,0x10
4
1,5x10
4
1,0x10
4
5,0x10
3
0002
deformation rate 3349/s 2863/s 2427/s
10-11 10-12 11-20 10-13
0,0 35
40
45
50
55
60
65
70
energy [keV] Figure 4: High energy synchrotron radiation spectra, samples after 5% compressive strain and different strain rates, χ = 90°
The intensity of different reflections for various inclination angles ψ and the corresponding strains of the lattice planes for the sample with 5% strain and the highest strain rate available are given in fig. 5 and fig. 6. Especially high residual microstrain values (fig.5) are visible in the longitudinal and the transversal direction of the sample as well as near the texture poles (fig.4). These microstrains reach magnitudes of up to 1x10-3, they are particularly high for the basal 0002 lattice plane and the 10-11 plane whose axes is situated in an angle of 60° to the 0002 axes. Similar high residual microstrain values have been reported in /3,4/ after tensile deformation of magnesium rods.
intensity [counts/300s]
129 2,5x10
6
2,0x10
6
1,5x10
6
1,0x10
6
5,0x10
5
reflections 10-12, 0002,
11-20 10-11
0,0 0,0
0,2
0,4
2
sin ψ
0,6
0,8
1,0
Figure 5: Intensity versus inclination angle ψ, strain 5%, strain rate 3349/s
6x10
-4
4x10
-4
2x10
-4
error margin -4 +- 1x10
strain
0 -2x10
-4
-4x10
-4
-6x10
-4
-8x10
-4
-1x10
-3
0,0
lattice planes 10-12, 0002,
0,2
0,4
2
0,6
11-20, 10-11
0,8
1,0
sin ψ Figure 6: Residual microstrain versus sin2ψ - curves, strain 5%, strain rate 3349/s
The strain versus sin2ψ curves are nonlinear for all lattice planes investigated here. For residual stress analyses on magnesium components this implies that residual strain values determined in the poles and in the load and transversal axis of strongly plastic deformed samples have to be checked carefully with respect to residual microstrains, e.g. by checking the linearity of the strain versus sin2ψ curve. Samples deformed with high strain rates show a tendency towards lower residual microstrains and less non-linear microstrain versus sin2ψ – curves (fig. 7). This presumably can be attributed to their weaker texture (fig. 3) compared to slowly deformed samples.
130 -4
4x10
-4
strain
2x10
0 -4
-2x10
-4
error margin +- 1x10
-4
-4x10
-4
-6x10
0002, degree of natural strain 5%, app. deformation rate: 2862 1/s, 3349 1/s
-4
-8x10
-3
-1x10
0,0
0,2
0,4
2
sin ψ
0,6
Figure 7: Residual microstrain versus sin2ψ – curve, deformation rates
6
0,8
1,0
0002 lattice plane after 5% deformation, different
Acknowledgements
Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Energy Research, Contract No. W-31-109-Eng-38. We gratefully acknowledge the assistance of Jörg Wegener, HMI Berlin, with the experiments.
7
References
[1] E. El-Magd, C. Treppmann., M. Brodmann, Werkstoffprüfung, DGM, Bad Nauheim, 1998, 193 - 202 [2] W. Reimers, A. Pyzalla; M. Broda, G. Brusch, D. Dantz, K.-D. Liss, T. Schmackers, T. Tschentscher‚ J. Mat. Sci. Letters 1999, 19, 581 – 583 [3] J.W.L. Pang, T.M. Holden, P.A. Turner, T.E. Mason, Acta mater. 1999, 47, 373-383 [4] P.A. Turner, C.N. Tomé, Acta mater. 1994, 42, 4143-4153
Zn Incorporation within the Intermetallic Mg12(LaxCe1-x) Lattice in Elektron MEZ C.J.Bettles*, C.J.Rossouw and K.Venkatesan* CSIRO Manufacturing Science and Technology, Private Bag 33, Clayton South MDC, Victoria, Australia 3169. * Members of CAST CRC
1
Introduction
MEL (Magnesium Elektron) have developed a new magnesium alloy containing a rare earth misch metal, Zn and Mn (designated Elektron MEZ). This alloy has superior high temperature creep properties compared to AE42, and shows potential for both HPDC and sand casting applications. The sand cast microstructure is comprised of equiaxed dendrites of a Mg solid solution separated by an intermetallic interdendritic phase. In this work systematic electron diffraction and analytical electron microscopy studies are used to identify the crystal structure and composition of this interdendritic phase. Variations in emission rates of characteristic X-rays, resulting from high energy electrons systematically scanned in angle of incidence near a zone axis orientation, are used to form two-dimensional channelling patterns. When this experimental data is used in conjunction with calculated channelling patterns for the specific crystal structure, characteristic features associated with individual sublattice or interstitial sites can be identified. This is particularly useful in enabling the specific sublattice site for minority atomic species to be determined. In this paper, channelling patterns are used to identify the particular type of sublattice site occupied by the minority Zn atoms within the intermetallic interdendritic phase.
2
Experimental Procedure
Thin foils of alloy Elektron MEZ, in the sand cast condition, were prepared for electron microscopy by electropolishing with a magnesium perchlorate-based electrolyte at 90v and 55°C. The preliminary crystal structure determination was carried out using selected area diffraction patterns (SADP) on a JEOL 2000EX electron microscope equipped with a double tilt holder at an accelerating voltage of 200keV. X-ray incoherent channelling patterns (ICPs) were recorded using a Philips CM30 electron microscope with a collimated and focussed 200 keV electron probe, 0.4 µm in diameter and with total convergence angle of either 1.8 mrad (30 µm condenser aperture) or 3 mrad (50 µm aperture). The theoretical basis for the use of ICP contrast and atom location by channelling enhanced microanalysis (ALCHEMI) to determine specific atomic lattice site occupancy of minority atomic species introduced into a lattice has been well documented [1]. The crystalline phase under consideration is tilted to align the electron beam close to a low index zone axis, and the total energy dispersive x-ray (EDX) spectrum is recorded. Variations in characteristic X-ray emissions as a function of beam orientation are thus obtained. With the Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
132 overall crystallography of the phase being known, these channelling patterns can be compared with calculated patterns for different sublattice or interstitial site-occupancies. Correlation between experiment and theory enables the distribution of the introduced minority alloying atoms over sublattice sites to be directly determined.
3
Results and Discussion
Figure 1(a) shows a segment of the pole figure map for the intermetallic interdendritic phase, as derived from SADPs. This established that the unit cell is tetragonal with lattice parameters of a = 10.3 Å and c/a = 0.586. Analysis of the EDX spectra from thinner areas yielded the composition (in atomic percent) as approximately 92 % Mg, 3.3 % La and 4.4 % Ce, with a small amount (~ 1 %) Zn and even smaller amounts (less than 0.1 %) Cu. Thus, the phase has an overall stoichiometry close to Mg12RE, where RE is La or Ce, and a composition ratio for the rare earths (La:Ce) of about 43:57. It conforms with the general structural type Mn12Tn (space group I4/mmm, number 139) [2,3] where the rare earth (RE) is uniquely distributed over one particular sublattice site, denoted by the Wyckoff letter a, and located at the origin and body centre of the tetragonal unit cell. Mg is uniformly distributed over three separate sets of sublattice sites within this cell, denoted by Wyckoff letters f, i and j (Figure 1(b)). In this paper these are referred to equivalently as M(k), with index k = 1, 2 and 3 respectively.
Figure 1: (a) Zone axis pole figure map from the intermetallic phase and (b) projected 107) are calculated. The values for SSC magnesium are between 50 – 70 N/mm². The values for the AM70 material (fine and globular grain structure) are about 10% higher than those for the AZ91 material, which, however, could be rather due to scattering of the data than to a metallurgical background. Comparing earlier fatigue strength testing with the KFP component, the S-N curves for the semi-solid produced specimen are between those for AZ91 specimens from die casting and AlSi7Mg specimens.
295
Stress amplitude [MPa]
1000
AM70 semi solid
AlSi7Mg-T6 perm. mould cast
100
AZ91 se mi solid
AZ91 die casting
10 1,00E+03
1,00E+04
1,00E+05
1,00E+06
1,00E+07
1,00E+08
Number of cycles to failure
Figure 6: S-N curves for dynamic testing of AZ91, AM70, and AlSi7Mg
7
Summary
The process of semi-solid casting combines the advantages of forging with the advantages of casting by producing near-net-shape components with thin and thick- walled cross sections with low porosity and shrinkage. The mechanical properties are investigated by analyzing components made of semi-solid casting of the alloys AZ91 and AM70. Standard tensile test specimen are used for the investigation of UTS or elongation, for example. A specially designed KFP-component is dynamically tested for fatigue values (S-N curves). Significant influences of the heat treatment processes are observed on the mechanical properties of tensile test specimens, which display the most interesting potential of the analyzed material. The fatigue results from the KFP components are satisfactory, which is based on the higher S-N curves in comparison with magnesium die castings. These results encourage to enlarge the dynamic testing of magnesium (especially wrought or hard-castable alloys) to provide material data for FEM calculations of future automotive applications.
8
References
[1] G. Chiarmetta, Thixoforming of automobile components, Proceedings 4th Intl. Conf. Semi Solid Processing of Alloys and Composites, Sheffield (England), p. 204-207, 1997 [2] J.Langemann, J.-P. Gabathuler, H. Huber, Serienproduktion von Thixoforming-Sicherheitsteilen für die Automobilindustrie, Symposium Thermprocess, GIFA, Düsseldorf (Germany), 1999 [3] W. Wagener, D. Hartmann, Feedstock Material for Semi-Solid Casting of Magnesium, proceeding 6th Intl. Conf. Semi Solid Processing of Alloys and Composites, Turin (Italy), Sep. 2000 [4] N.N., DIN 50125, German Industry Standard, Zugproben Form B, Deutsches Institut für Normung e.V., Berlin, p.3 [5] N.N., Datenblatt Magnesium-Druckgusslegierungen, April 1997, Hydro Magnesium Marketing GmbH, Bottrop (D) [6] N.N., TALMAC-Project BE95-1244, Properties of Thixformed parts, 1998
Determination of Material Properties and Numerical Simulation to Predict the Mechanical Performance of Die Casted Components Moritz Wuth, Elke Lieven PETRI AG, Aschaffenburg;
Wolfgang Böhme Fraunhofer-Institut für Werkstoffmechanik (IWM), Freiburg
1
Introduction
An important feature of the modern industrial development process is the use of calculations and simulations at an early stage of the design process. A precise description of the materials behavior is the necessary prerequisite for a correct mechanical finite element analysis (FEanalysis). This is not too difficult for the materials usually used with explicit linear-elastic performance like Steel and Aluminum. In the case of Magnesium alloys, which today are materials of great interest in automotive applications, it has been found that this is difficult because of usual preparation of material data sets and software restrictions. Preliminary calculations show significant differences between the numerical results and the performance of the components in reality. Steering wheels have to be designed to bear bending loads besides a variety of other mechanical service and safety related loads (torsion, fatigue, crash absorption). The most important carrier of a steering wheel’s mechanical behavior is the so called “frame” which today is almost entirely diecasted from Magnesium or Aluminum alloys. For this reason, accurate measurements of material properties have been performed for the MgAl6Mn (AM60)- as well as for the AlMg5Si2Mn diecasting alloy. These two materials are of interest for steering wheel production. The named alloys allow for sufficient ductility and strength to meet the complete range of service conditions in the as casted state. Besides, these alloys can be processed efficiently in the die casting process. The material database has been detected to have a significant impact on the reliability of the results: Not all the material’s properties can be taken from the same source, usually engineering stress strain curves are available which have to be transposed to true stress strain curves. Furthermore there are indications, that the usual type of material behavior, performing linear-elastically up to a proof stress, do not apply for Magnesium alloys [1, 2].
2
Determination of Material Properties
2.1
Principle of Measurement
Two different techniques to determine true stress strain curves have been used on a statistically relevant number of casted specimens: Strain gage instrumentation of the Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
297 specimens to determine true stress, elongation and transverse contraction up to 2 % as the most precise technique available. Optical determination of elongation and transverse contraction up to failure of specimens to enable the measurement of elongation and transverse contraction uninfluenced by non-linear strain gage performance. The specimens have been casted by Petri AG into a round testbar shown in fig. 1 in the cold chamber diecasting process.
Figure 1: Test specimen for true stress-strain measurements
2.2
Results
Two basic findings are the result of the measurements: The first finding is, that there is little, if any, linear-elastic performance in MgAl6Mn in comparison to AlMg5Si2Mn, as shown in fig. 2. This means furthermore, that the 0.2 % proof stress is not the essential value for calculation of mechanical performance as it has nothing to do with the true yield point. The non-elastic performance of a material can be pointed out much better if one plots the slope of the stress strain curve for a Magnesium alloy as a function of true stress, in other words the load dependent evolution of the modulus. This curve must have a slope of 0 if an elastic modulus exists. In case of MgAl6Mn we do not find a constant value up to a proof stress. Only in the region below 20 MPa a constant performance according Hooke’s law may exist, see fig. 3. The measurements do not allow a definite evaluation with regard to the value of 20 MPa due to small preload at the beginning of the test. For the Aluminum alloy, this constant value exists up to about 140 MPa, which is also according to the stress strain curve (s. fig. 2). 300
σ (True Stress) [MPa]
200
0,002
100
Specimen 212, AlMg5Si2Mn Specimen 113, MgAl6Mn MgAl2Mn Elastic modulus = 70 GPa Elastic modulus = 45 GPa 0 0,000
0,005
0,010
0,015
0,020
ε (True Strain) [1]
Figure 2: Comparison of initial part of stress strain curves for MgAl6Mn, MgAl2Mn, and AlMg5Si2Mn
298 80
60
dσ / dε [GPa]
Specimen 212, AlMg5Si2Mn Specimen 111, MgAl6Mn 40
20
0 0
50
100
150
200
250
σ (True Stress) [MPa]
Figure 3: Load dependent evolution of dσ/dε, comparison MgAl6Mn / AlMg5Si2Mn
The second finding of relevance is the fact, that a constant Poisson’s ratio “ν” does not exist for MgAl6Mn as is demonstrated in fig. 4. 0,45
ν (Poisson Ratio) [1]
0,40
0,35
0,30
Specimen 213, AlMg5Si2Mn
Specimen 112, MgAl6Mn 0,25 0
50
100
150
200
250
σ (True Stress) [MPa]
Figure 4: Poisson’s ratio as a function of stress, comparison of MgAl6Mn and AlMg5Si2Mn
Fig. 4 shows, that Poisson’s ratio is not constant especially up to the usually determined 0.2 % yield point for an Magnesium AM-type alloy. This behavior is well in agreement with the plastic, non-linear behavior. Whether there might be constant behavior below 20 MPa for MgAl6Mn cannot be answered due to the applied preloading. The Aluminum alloy nevertheless shows an obvious constant Poisson’s ratio up to more than 200 MPa, which is significantly more than the determined proof stress. With regard to the prediction of the performance of components by FE-analyses, deviations in the determination of plastic deformation after a certain loading can thus be explained. The differences are significant, if parts are exposed to stresses in the region of the usually determined 0.2 % proof stress (fig. 2). The latter defines a state, which occurs locally under
299 usual service conditions of technical components. The designer basically has to design for those stresses.
3
Verification Tests
3.1
Model, Testing Conditions
To give prove of the significance of a correct material database and a proper FE-model, experiments on bending a flat beam and a steering wheel are conducted and finally a comparison between FE-analyses and real, mechanical performance is drawn. The beam is a model of little complexity and allows an analytical calculation of the elastic deformation expressed by elastic deformation distance fmax, el according to equation (1) [3]. fmax, el = F * l³ / ( 3 * Iy * E) (1) The flat beam has a geometry of b = 8 mm, t =3 mm and a bending length of l = 51.5 mm (fig. 5). l F
w (x) fmax, el
bending line x
Figure 5: Test beam, bending test installation, sketch defining the variables used for analytic calculation
The same conditions have been defined for bending experiments, simulation and analytical calculation. The load is imposed onto the beam along a line perpendicular to the specimen’s longitudinal axis and all degrees of freedom are fixed at one end. The frame is mounted by its hub while being exposed to a bending load perpendicular to the ring plane in 12:00 o’ clock position. The meshes of beam and frame consist of hexaeder elements. The fineness of the meshes is optimized by preceding calculations. FE-analyses are processed with MSC Nastran, version 70.5, non-linearly, and with elasto-plastic material definition. All parts are tested in both alloys in diecasted condition. Results are determined for the loaded and the unloaded state of the specimen.. 3.2
Results
Bending of the beam under different loads results in deformation distances fmax, el which can quite exactly be determined by FE-simulation for all materials, if the material database correctly takes into account proof stress and elastic modulus (table 1). Tables 2 and 3 summarize the comparison of the results from experiments, simulations and analytic calculations of the bending of the beam for both materials. The deviation of measured
300 values from simulated as well as calculated values for the maximum stresses proof indirectly, that the Magnesium alloy does not behave linearly already at very low stresses. With the Aluminum alloy a deviation does not occur below the defined yield point, which in this case is 7 times higher than for the Magnesium alloy. Because of the distinct linear-elastic behavior of the Aluminum alloy, calculated results show better conformity with the measured values. Table 1: Mechanical properties of mentioned materials as defined in Petri FE-database Mechanical Property MgAl6Mn MgAl2Mn AlMg5Si2Mn (AM60) (AM20) Elastic modulus [GPa] 41.0 35.6 69.6 a) Proof stress [MPa] 20 15 142 Poisson ratio 0.310 ----0.325 a)
Value has a tendency towards 0
Table 2: Comparison of bending results from FE-analyses, real experiment and analytic calculation of MgAl6Mn-testbeam Load Stress fmax ,el fmax , el fmax , el fmax ,total fmax , total Fmax [N] σxxmax (measurement) (simulation) (analytic) (measurement) (simulation) [MPa] [mm] [mm] [mm] [mm] [mm] 3.9 16.7 0.185 0.231 0.23 0.19 0.231 5.9 25.3 0.305 0.348 0.353 0.39 0.352 9.8 42.1 0.505 0.585 0.586 0.60 0.590 19.5 83.7 1.300 0.857 1.160 1.55 1.237 Table 3: Comparison of bending results from FE-analyses, real experiment and analytic calculation of AlMg5Si2Mn-testbeam Load Stress fmax ,el fmax , el fmax , el fmax ,total fmax , total Fmax [N] σxxmax (measurement) (simulation) (analytic) (measurement) (simulation) [MPa] [mm] [mm] [mm] [mm] [mm] 9.80 42.06 0.38 0.353 0.357 0.38 0.353 19.50 83.69 0.77 0.702 0.709 0.87 0.702 49.80 213.73 2.00 1.794 1.815 2.20 1.985 Table 4: Comparison of bending results from FE-analyses and real experiment of frame, load 500 N fmax, total fmax, plastic fmax total (FEfmax plastic (FE(measurement) (measurement) simulation) simulation) [mm] [mm] [mm] [mm] Mean Value 27.8 3.3 29.2 (22.2) 5.6 (0.65) Standard Deviation 0.5 0.2 --------Values in brackets: Database from literature, table 5
The bending of the frame can be predicted only by the help of a FE-analysis because of its complex geometry. The results are summarized in table 4. The values in brackets have been determined with a material database which supposes a linear-elastic behavior up to 0.2 % proof stress and the commonly given elastic modulus as given in table 5.
301 Measured values and deformation in loaded state are of good conformity with simulation. The deformation in unloaded state could not be determined with similar precision. The difference between FE-results and experimental measurements may be due to differences between the material model used within the FE-software and real material performance: The model is not able to take into account a stress dependent Poisson’s ratio. Moreover the conditions for simulations have not been adopted to the special material behavior under unloading conditions. An unknown amount of preload within the frame should be of minor relevance. Table 5: Mechanical properties. Literature Mechanical property MgAl6Mn MgAl2Mn (AM60) (AM20) Elastic modulus [GPa] 45 [4] 45 [4] 0.2 Yield Strength [MPa] 130 [4] 90 [4] Poisson ratio [1] 0.33 [1] ------
AlMg5Si2Mn 70-73 [5, 6] 145 [7] 0.30 –0.34 [5]
Figure 6 shows on the left side the frame during measurements and on the right hand side the result of the FE-Simulation.
Figure 6: Frame bending test. Comparison experiment (left) with FE-result (right)
Frames casted from Al-alloy have not been available for the bending test. Therefore a comparison is not possible. If material databases do not take into account the true material’s performance, especially plastic deformation will not be calculated correctly. The results in form of values of deformation will be too small because the linear-elastic approach does not take into account the existing plastic deformation, which is displayed in fig. 7. The results will assume a design of sufficient firmness whereas the component really is too weak. This is detrimental if a designer designs a safety part. Calculations of deformation will be increasingly wrong the higher the definition of elastic modulus and the proof stress. The mistake will increase with the load.
302 200
σ (True Stress) [MPa]
150
εel
εpl
100
50
Specimen 113, MgAl6Mn elastic modulus 45 GPa true modulus 42,337 GPa 0 0,000
0,002
0,004
0,006
0,008
0,010
ε (True Strain) [1]
Figure 7: Impact of elastic modulus on prediction of true deformation of a component
4
Summary
Precise determination of the material behavior as well as correct material laws for FEsimulation are a necessary prerequisite for the prediction of a component’s mechanical behavior especially for components made from Magnesium alloys. To determine the material behavior of a non-linear performing material it is necessary to measure precisely the very beginning of the true stress strain curve. This is very important with regard to the elastic modulus and the true proof stress. The 0.2 % proof stress must not be used to substitute the true yield point for a non-linear performing material. Poisson’s ratio and transverse contraction are for a wide range between true yield stress and commonly applied 0.2 % proof stress not constant but stress dependent for the MgAl6Mn alloy whereas Poisson’s ratio is constant even above true yield point for the AlMg5Si2Mn alloy. The fact, that the stress strain curve for MgAl2Mn (AM20) shows a similar shape as for MgAl6Mn (AM60) indicates, that similar conclusions have to be drawn for all MgAl-type alloys (AM-alloys). It is necessary to enable the software to calculate on a basis of a completely non-linear material model, to allow a definition of a stress dependent Poisson’s ratio, and to distinguish between the state of loading and unloading. The commonly used material models are suitable for materials which perform linear-elastic as Aluminum and Steel do, because the mathematical error stays within tolerable limits.
303
5
References
[1] W. Buchmann in Magnesium und seine Legierungen, (Ed. A. Beck), Verl. Springer, 1939, Chapter Festigkeitseigenschaften 3, p. 152 ff. [2] K. Pettersen, S. Fairchild, Stress Relaxation in Bolted Joints of Die Cast Magnesium Components, SAE 970326, 1997, p. 2. [3] G. Rumpel, H.D. Sondershausen, (Ed.: W. Beitz, K.-H. Küttner), Dubbel - Taschenbuch für den Maschinenbau, 17th ed., Springer, Berlin, 1990, C 21 [4] Hydro Magnesium, Data Sheet Die Cast Magnesium Alloys, 1996. [5] H. Kuchling, Nachschlagebücher für Grundlagenfächer Physik, 14th ed.., VEB Fachbuchverlag, Leipzig, 1978, p. 511. [6] Aluminium Taschenbuch, Chapter 12.3, p. 770. [7] H. Koch, U. Hielscher, H. Sternau, A. Franke, Magsimal 59, an AlMgMnSi-Type Squeezecasting Alloy Designed for Temper F, Aluminium Rheinfelden GmbH, Germany, 1996, p. 4
Fatigue Design with Cast Magnesium Alloys CM. Sonsinol), K. DieterichO, L. Wenk2), A. Till 1) 1) Fraunhofer-Institute for Structural Durability (LBF), Darmstadt/Germany 2) German Foundrymen's Association, Diisseldorf/Germany
1 Introduction The light weight of magnesium alloys make them attractive specially for automotive applications which require ecological contributions by saving of resources, i.e. material and energy. However, for a broader use of these materials, a good knowledge of fatigue design relevant properties like S-N curves, their scatters, influence of mean-stresses and notches on fatigue strength as well as the influence of thickness and surface (casting skin) is necessary. The following paper presents the present data generated for the die cast materials AZ 91 HP (GD-MgA19Znl), AM 50 HP (GD-MgA15Mn) and AM 20 HP (GD-MgA12Mn) and compares them with wrought steels, cast nodular iron and cast aluminium alloys. This investigation enlarges the present data basis for fatigue applications [1,2].
2 Materials and Mechanical Properties The materials AZ 91 HP, AM 50 AP and AM 20 HP were die cast. Table 1 gives the chemical compositions and Table 2 the conventional mechanical properties. Table 1: Chemical compositions of the investigated alloys Chemical comp. in %
Zn
Al
Mn
Cu
Si
Fe
Ni
Pb
Be
Mg
9.0000 0.8100 0.1500 0.0064 0.0310 0.0010 0.0006 0.0050 0.0003
rest
AM 50 HP 4.7000 0.0000 0.3100 0.0060 0.0030 0.0010 105 cycles extremely shallow, k = 45, which means a drop of endurable stress amplitude of 5% from one decade to the next one. The results are covered by the scatter of the endurable stress amplitudes between the probability of survivals Ps = 10 and 90% TCT = 1 : [aa(10%)/aa(90%)] = 1 : 1.30. This is a value which is found also for other cast materials. The curves for pulsating loading (R=0) lie in a lower level compared with fully reversed loading (R=-l) due to the mean-stress sensitivity. The endurable stresses of the notched specimens, presented by nominal stress amplitudes a an =F a / Anett0, are lower than the endurable values of the unnotched ones due to the existing notch sensitivity. Also the S-N curves for the alloys AM 50 HP and AM 20 HP exhebit same characteristics found for AZ 91 AP. The crack initiation of these materials started always from present microshrinkages, Fig. 5. G-MgA19Znl (AZ 91 HP)
G- MgAl5Mn (AM 50 HP)
G-MgAI2Mn (AM 20 HP)
Crack initiation in the area of Crack initiation in the area of Crack initiation in the area of a shrinkage cavity a shrinkage cavity a shrinkage cavity
Figure 5. Crack initiation sites of the investigated alloys The endurable stress amplitudes of these materials for N = 105 and 5 • 106 cycles with P s = 50% are compiled in Table 3. From these values the mean-stress sensitivity M = [a a (R=-l)/a a (R=0)]-l
(1)
Kf=aan(Kt=1.0)/aan(Kt=2.5),
(2)
and the fatigue notch factor
which corresponds to the notch sensitivity, was calculated, Table 4.
Table 3: Endurable stress amplitudes Stress concentration factor Kt -1 AZ91HP GD-MgA19Znl
die casting
1.0
2.5
84/77
61/56
52/48
38/35
70/64
49/45
51/46
36/33
61/56
43/39
43/39
32/29
*)
0 *)
-1 AM 50 HP GD-MgA15Mn
die casting
*)
0 *)
-1 AM 20 HP GD-MgA12Mn
die casting
*)
0 •)
*) Stress ratio R^ = (JU/CT0 Table 4: Mean-stress and notch sensitivity Alloy
Casting process
Mean-stres:5 sensitivity M = [a a (R=-l) /a a (R=0)]-l
Fatigue-stress concentration factor Kf
K t =1.0
Kt = 2.5
R = -l
R=0
AZ 91 HP
die casting
0.61
0.61
1.38
1.37
AM 50 HP
die casting
0.37
0.36
1.43
1.42
AM 20 HP
die casting
0.42
0.34
1.42
1.34
These values are compared in Figs. 6 and 7 with other metallic materials [4]. It can be concluded that the increase of ultimate tensile strength also increases the mean-stress sensitivity. The values are on a comparable level with cast aluminium alloys, Fig. 6. Fig. 7 shows that the decrease of fatigue strength by increasing the stress concentration is for the investigated alloys much lower than aluminium, cast nodular iron or wrought steels. Despite the axial loading, the notch sensitivity of magnesium is lower than for cast-nodular iron or wrought steels under bending. This means that under bending the notch sensitivity of magnesium alloys will be even lower and for fatigue design more advantageous.
309
. Cast steel
I
Steel
c ro Scatter band of round specimens fc under constant amplitude bending— Scatter band of flat specimens under constant amplitude axial loading
1500 2000 MPa Tensile strenght S.
1000
Figure 6. Mean-stress sensitivity of different metallic materials
LOw^v
G-Mg (axial)
pSfN. A
I
X^^^^^.'-TX.
AZ 91 HP AM 50 HP/AM 20 HP / /
G-AI (axial)
Cast iron ^^C* s^> •
"a CD N
~ro
E o
0.1-
Ps= 50 % N = 2-106 R = -1 1
1.0
= ==—-f
^ ^ S i ^ S ; ^ ^ ^ - ' ' " - " " ^--155555^
0)
1
1 —
2.0
i
3.0
•
i
i
4.0
*sis=^=sj
1
"
5.0
Stress concentration factor Kt Figure 7. Decrease of fatigue strength by increased stress concentration
6.0
310 As magnesium alloys are an alternative for substituting aluminium, also a comparison of endurable stresses of both alloys is necessary. In Fig. 8 the investigated alloys are compared for the present porosity degree of P=0 [5] with the presently mostly used aged cast aluminium alloy G-AlSi7Mg0.6 T6 [6]. Fig. 8 reveals comparable endurable high-cycle fatigue strength amplitudes. The fact of comparable fatigue strengths explain why magnesium can be a good alternative to aluminium, provided that corrosion and creep effects can be circumvirated by good design and manufacturing.
1.0 1.0 2.5 G-AISi7Mg0.6 T6
1.0 1.0 2.5 G-MgAI9Zn1 (AZ91HP)
1.0 1.0 2.5 G-MgAI5Mn (AM50HP)
1.0 1.0 2.5 G-MgAI2Mn (AM20HP)
Figure 8. Comparison of fatigue strengths of cast aluminium and magnesium alloys
4 Conclusions Presently used cast magnesium alloys reveal sufficient high fatigue strength for a substitution of cast aluminium alloys and for taking advantage from the lower density. However, for time and environment dependent effects more knowledge and feed-backs from field applications are necessary. With regard to fatigue design with cast magnesium alloys same fatigue design principles as for other metallic materials can be applied but particular properties of magnesium alloys like low Young's modulus, tendencies to creep, mean-stress, temperature and corrosion sensitivity and continuous fatigue strength decrease in the high-cycle fatigue area must be considered and compensated by a magnesiumadjusted design.
311
5 Acknowledgement The presented results were obtained within a research project funded by the Community of Industrial Research Associations (AiF), Koln, (Funding No. 11 726) and supported by the Working Group „Light Metal Casting" of the Association of German Casters (VDG), Dusseldorf, chaired by R. Woltmann. The authors acknowledge the mentioned supporters of this investigation.
6 References [ 1] [2]
[3]
[4]
[ 5]
[6]
CM. Sonsino, H. Kaufmann, R. Keiper, Light Turbocharger Compressor Wheels from Aluminium and Magnesium Investment Casting, SAE Paper 990371 (1999) M. Witt, K. Potter, H. Zenner, K. Sponheim, P. Heuler, Fatigue Strength of Cast Aluminium and Magnesium Chassis Parts, Proceedings of the Second Israeli International Conference on Magnesium Science & Technology, February 2000, Dead See, Isreal, 263-275 CM. Sonsino, Methods to Determine Relevant Material Properties for the Fatigue Design of Powder Metallurgy Parts, Powder Metallurgy International 16 (1984) No. 1,34-38; 16 (1984) No. 2, 73-77 CM. Sonsino, V. Grubisic, Requirements for Operational Fatigue Strength of High Quality Cast Components, Materialwissenschaft und Werkstofftechnik 27 (1996) No. 8, 373-390 ASTM Designation El55, Standard Reference Radiographs for Inspection of Aluminium and Magnesium Castings, American Society for Testing and Materials, Philadelphia (1979) CM. Sonsino, K. Dieterich, Einflufi der Porositat auf das Schwingfestigkeitsverhalten von Aluminium-GuBwerkstoffen, Fraunhofer-Institute for Structural Durability (LBF) Darmstadt, Report No. FB-188 (1990)
Isothermal Fatigue of Magnesium Wrought Alloy AZ31 Ulf Noster, Igor Altenberger and Berthold Scholtes Institute of Materials Technology, University Gh Kassel, Germany
1
Abstract
AZ31 in the as wrought condition shows a distinct anisotropy of deformation and strength, if tension and compression loading is compared. Cyclic deformation behavior in the temperature range beetwen 20°C and 300°C was investigated in stress controlled tension-compression tests. With increasing temperature fatigue strength decreases and the damage process shifts from fatigue-controlled to a more creep-controlled mechanism. The asymmetrical deformation behavior leads to cyclic creep for symmetrical cyclic loading. The macroscopically measured deformation was correlated with microstructural observations, in particular, with the extent of deformation twinning. Fatigue strength of AZ31 can be enhanced by mechanical surface treatments, e.g. by shot peening or deep rolling. As a consequence of the induced residual macro and microstress, plastic strain amplitude, which characterizes the fatigue damage process, decreases. Deep rolled AZ31 exhibits also an increase of fatigue strength at higher loading temperatures in comparison to the untreated state, but because of the instability of the induced residual stress and strain hardening, the enhancement is much smaller compared to loading at lower temperatures.
2
Introduction
Magnesium is the lightest structural metal with a relatively high strength-to-density ratio and a good corrosion resistance due to alloying development. This advantage makes magnesium alloys very attractive for light-weight constructions, e.g. in the automotive and aerospace industry. At this time, most of the components are produced as castings like gear and clutch housings, computer components and video-camera housings. On the other hand wrought and extruded materials states, which find an increasing demand in industrie (1), often have the advantages of lower costs and better mechanical properties than cast components (2). Of special interest is the use of magnesium alloys for structural applications under dynamic loads at higher service temperatures. The possibility to enhance the fatigue strength at room temperature of magnesium wrought alloys by mechanical surface treatments, like shot peening and deep rolling, has been demonstrated by different authors (see e.g. 3-6). In this paper the cyclic deformation behavior of magnesium wrought alloy AZ31 in the temperature range beetwen 20°C and 300°C is described. To analyse the possibility of lifetime enhancement by mechanical surface treatment even at higher service temperatures first results of deep rolled materials states are also included.
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
313
3
Materials and Experimental Details
Cylindrical specimens with a diameter of 7 mm and a gauge length of 15 mm were taken in rolling direction out of a 15 mm thick sheet of wrought magnesium alloy AZ31 (3 wt.-% Al, 0.88 wt.-% Zn, 0.33 wt.-% Mn, balance Mg). Due to the manufacturing process small differences of the mechanical properties of individual plates occured. No further heat treatment was applied. The investigated material has a highly inhomogeneous grain size and a distinct rolling texture with basal planes of the hexagonal cells parallel to the surface of the sheet. To avoid an influence of the machining operations the material was electrolytically polished at the gauge length up to a depth of 100 µm. For deep rolling processes, a hydraulic rolling device (Ecoroll HG 6-9) with a 6.6 mm ball (pressure 100 bar) was applied. The quasistatic tension and compression tests were carried out using a mechanical testing machine at a deformation velocity of approx. dε/dt = 10-3 s-1. Tension-compression fatigue tests were performed with a servohydraulical testing device under stress control without mean stress (R = -1) and, if not otherwise specified, with a frequency of 5 Hz. Strain was measured using capacitative extensometers. For experiments at higher temperatures up to 300°C specimens were heated with two controlled resistance heating elements. Residual stresses were measured on a ψ-diffraction unit at (112)-planes with CrKα radiation. For stress evaluation the sin2ψmethod with X-ray elastic constant 1/2 s2 = 2.93x10-5 mm2/N was used. Residual stress depth profiles were determined by successive material removal achieved by electropolishing.
4
Results and Discussion
For magnesium wrought alloys an asymmetrical deformation behavior at room temperature is known if tension and compression tests are compared (2,7). After compressive loading a higher density of deformation twins was observed than in tension for the same applied stresses (6,10). AZ31 investigated in this paper shows a tensile 0.2%-proof stress of 160 MPa compared to a compressive 0.2%-proof stress of 92 MPa. As one can see in Fig. 1 hysteresis loops at the beginning of fatigue life are asymmetrical with pronounced plastic yielding during the compression phase. This behavior leads to cyclic creep with negative mean strains for symmetrical cyclic push-pull loading, and hence, to a reduction of the length of the specimens, especially at the beginning of fatigue tests. In Fig. 2 for room temperature tests, plastic strain amplitudes and mean strains are shown as a function of number of cycles. The material exhibits cyclic hardening and the amount of mean strain increases with higher stress amplitudes. With increasing temperature quasistatic tension tests show a significant decrease of strength and an increase of elongation to fracture. In Fig. 3 cyclic deformation curves for two different stress amplitudes at test temperatures up to 300°C are given. For the higher stress amplitude cyclic hardening at smaller test temperatures occurs, but with increasing temperature cyclic softening becomes more and more important. At a lower stress amplitude of σa = 75 MPa, cyclic softening is more pronounced and only for T ≤ 150°C, nearly constant plastic strain amplitudes are observed.
314
Figure 1: Stress-strain-curves during symmetrical push-pull loading of AZ31 (T = 20°C)
Figure 2: Cyclic deformation curves of AZ31 for symmetrical push-pull loading with different stress amplitudes (T = 20°C)
In Fig. 4 plastic strain amplitudes at half the number of cycles to failure vs. number of cycles to failure (Nf) are drawn (Manson-Coffin Plot). It can be seen that up to a test temperature of approximately 200°C all data can be described by a single straight line. For higher temperatures, the slope of the line decreases and for a given plastic strain amplitude, a smaller fatigue lifetime occurs. In addition, at higher test temperatures, a pronounced influence of the test frequency on the number of cycles to failure is observed. At 250°C and σa = 50 MPa lifetime was approx. 60 times higher at a test frequency of 10Hz compared with 0.01Hz. Metallographic observation of crack formation and cracked surfaces indicate an increased influence of creep processes at higher temperatures and lower frequencies. For T ≥ 250°C, asymmetry of hysteresis loops at the beginning of the tests disappears as well as the pronounced formation of deformation twins during compressive plastic loading (2,8-10).
315
Figure 3: Cyclic deformation curves of AZ31 for symmetrical push-pull loading with different stress amplitudes at temperatures in the range beetween 20°C and 300°C
Figure 4: Manson-Coffin plot of AZ31 for symmetrical push-pull loading with different stress amplitudes at temperatures in the range beetween 20°C and 300°C
In (3-6) the possibility to enhance cyclic strength at room temperature by mechanical surface treatments is shown. In the case of AZ31 the plastic strain amplitudes of shot peened or deep rolled materials states are much lower compared with polished ones. For the surface conditions investigated a cyclic hardening behavior was observed. Compressive residual stresses and strain hardening induced during surface treatment were identified as responsible for the alterations of the cyclic deformation behavior (5,6). In Fig. 5 plastic strain amplitudes of polished and deep rolled materials states for fatigue tests at higher temperatures are compared. For the stress amplitude investigated polished specimens, at 100°C exhibit cyclic hardening, but with increasing temperature, for higher number of cycles, cyclic softening becomes more and more of importance. For deep rolled
316 specimens, cyclic softening is more pronounced. Consequently, although plastic strain amplitudes at the beginning of the test for all temperatures are much smaller for deep rolled compared with polished specimens, this difference decreases with increasing number of cycles. As a consequence, lifetime enhancement by deep rolling is not very pronounced for fatigue at higher temperatures.
Figure 5: Cyclic deformation curves of polished and deep rolled materials states for symmetrical push-pull loading at different temperatures (σa = 75 MPa)
Figure 6: Depth profiles of residual stress and FWHM of deep rolled materials states as a function of fatigue parameters (σa = 75 MPa)
One explanation for the only small lifetime enhancement at higher temperatures is given in Fig. 6. Compared with the starting condition after deep rolling, residual stresses were considerable reduced after applying Nf/2 load cycles at higher temperatures. The same is valid
317 for full width at half maximum of x-ray interference lines (FWHM), which indicates the strain hardening state of the material. During room temperature fatigue these values are much more stable (5).
5
Acknowledgement
Sincere thanks is expressed to the Deutsche Forschungsgemeinschaft for financial support.
6
References
[1] B. M. Closset, J.-F. Perey, Proc. Magnesium Alloys and their Applications, ed. B. L. Mordike, K. U. Kainer, Verlag Werkstoff-Informationsgesellschaft, Frankfurt 1998, 195200. [2] Materials Science and Technology, Structure and Properties of Nonferrous Alloys, Vol. 8, ed. K. H. Matucha, VCH, Weinheim 1995. [3] D. Deiseroth, W. Zinn, B. Scholtes, Proc. Magnesium Alloys and their Applications, ed. B. L. Mordike, K. U. Kainer, Verlag Werkstoff-Informationsgesellschaft, Frankfurt 1998, 409-414. [4] T. Dörr, M. Hilpert, P. Beckmerhagen, A. Kiefer, L. Wagner, Proc. 7th Int. Conf. on Shot Peening (ICSP 7), Poland, ed. A. Nakonieczny, Warsaw 1999, 153-160. [5] Altenberger, U. Martin, B. Scholtes, H. Oettel, Proc. 7th Int. Conf. on Shot Peening (ICSP 7), Poland, ed. A. Nakonieczny, Warsaw 1999, 79-87. [6] Altenberger, S. Jägg, B. Scholtes, Proc. Werkstoffprüfung 1998, Bad Nauheim 1998, DVM, 55-65. [7] M. M. Avedesian, H. Baker, ASM Specialty Handbook, Magnesium and Magnesium Alloys, Materials Information Society, 1999. [8] Beck, Magnesium und seine Legierungen, Springer, Berlin 1939, [9] J. Polmear, Light Alloys, Metallurgy of Light Metals, Edward Arnold, London 1981. [10] R. Gehrmann, G. Gottstein, Proc. 12th Int. Conf. Text. Mater, Canada, Montreal 1999.
Characterisation of Precipitate Phases in WE54 and AZ91 Alloys J.F. Nie1, X.L. Xiao1,2 and C.P. Luo2 1 Department 2 Department
1
of Materials Engineering, Monash University, Clayton, Australia of Mechano-Electronic Engineering, South China University of Technology, Guangzhou, China
Introduction
Lightweight magnesium alloys have attracted increasing interest in recent years for potential applications in the aerospace, aircraft and automotive industries. One important group of magnesium alloys, including alloys designated WE54, WE43, AZ91, ZC63, and QE22 [1], are those strengthened essentially by precipitation hardening. Although the maximum hardness achievable is substantially lower than that in precipitation-hardened aluminium alloys [1], the hardening response in these alloys is such as to have practical significance. However, attempts [2,3] to improve the hardness and strength of these alloys, and to develop new higher strength magnesium alloys, have to date been restricted by a lack of understanding of the structure, orientation, morphology, composition, and distribution of the strengthening precipitate phases, and the factors that control these microstructural parameters. It is the purpose of this paper to provide an overview of recent results of characterisation of the structure, morphology and orientation of fine-scale precipitates of intermediate and equilibrium phases in magnesium alloys WE54 and AZ91 using transmission electron microscopy and electron microdiffraction.
2
Precipitate Phases in Alloy WE54
The maximum strength of magnesium alloy WE54 (Mg-5wt%Y-2wt%Nd-2wt%Heavy Rare Earth-0.4wt%Zr) is achieved by conventional age-hardening treatments, which typically involve a solution treatment of 8h at 525°C, a hot water quench and a subsequent ageing treatment of 16h at 250°C [4]. The aged microstructure at maximum hardness has been reported [4,5] to contain metastable β' and equilibrium β phases as dispersed precipitates, and both of these phases have been described as forming as plates on {1 100 }α planes of the magnesium matrix (α) phase. The β' phase has been determined [5,6] to have a base-centred orthorhombic Bravais lattice, with lattice parameters a = 0.640 nm, b = 2.223 nm, c = 0.521 nm, based on selected area electron diffraction (SAED) patterns. The orientation relationship reported between β' and the matrix phase has the form (l00)β' // (1 210 )α, [001]β' // [000l]α. The proposed structure and orientation of the β' phase are similar to those of the β' formed in binary Mg-Y alloys [5]. The equilibrium phase β has been reported to have a a face-centred cubic (f.c.c.) Bravais lattice, with lattice parameter a = 2.223 nm [6]. The orientation relationship reported between β and the matrix phase is such that ( 112 )β // (1 100 )α, [110]β // [000l]α, and is identical in form to that observed between β' and the matrix phase in binary
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
319 Mg-Nd alloys [6]. Although Bravais lattices have been proposed for the β' and β phases, full crystal symmetries remain to be established.
Figure 1: Transmission electron micrographs showing [0001]α images of (a) globular β' and plate-shaped β1 precipitates, and (b) a β1 plate that has partially transformed to β phase in samples of WE54 aged for 48h at 250°C
The structure, morphology and orientation of strengthening precipitate phases in peak-aged samples of alloy WE54 have recently been characterised [7] in detail using TEM and electron microdiffraction. Figure 1(a) shows a bright field image, recorded parallel to [0001]α, typical of the microstructure in samples aged to maximum hardness (48h at 250°C). The microstructure contained a dispersion of irregular, globular particles, in contact with plateshaped precipitates. Electron microdiffraction patterns obtained from globular precipitates [7] could be indexed consistently according to the based-centred orthorhombic structure of β' phase. The orientation relationship implied by the superimposed precipitate and matrix patterns was also consistent with that reported previously, i.e. (l00)β' // (1210 )α, [001]β' // [000l]α. The observed morphology of the β' phase was, however, clearly in contrast to previous suggestions [4-6] that β' formed as plates parallel to {1 100 }α. The projected images of the plate-shaped precipitates had the form of parallelopipeds, with the extended broad surfaces parallel to {1 100 }α planes. Electron microdiffraction patterns recorded from these=plates suggested that most of them had an f.c.c. structure, with a ~ 0.74 nm, and an orientation relationship that was of the form ( 112 )β1 // (1 100 )α, [110]β1=// [0001]α [7]. This phase, designated β1, is a major intermediate precipitate phase in samples aged to maximum hardness at 250°C, but has not been reported previously in WE alloys. Although it proved difficult to obtain full crystal symmetry from electron microdiffraction patterns, the proposed structure is similar to that of phases of the from Mg3X (X = Nd, La, Ce, Pr, Dy, and Sm). All such phases have an f.c.c. structure (space group Fm3m ) and a lattice parameter in the range 0.74±0.01 nm, and all form exclusively as plates parallel to {1 100 }α [7]. Inspection of {1 100 }α plates of the β1 phase in samples aged 48h at 250°C revealed that some of them had transformed in situ to the equilibrium phase β, Fig. 1(b). Analysis of the pattern symmetries and absent reflections in convergent beam electron diffraction patterns recorded from β plates suggested that the β phase had a F43m space group [7]. The
320 orientation relationship was such that ( 112 )β //(1 100 )α, [110]β=// [0001]α. This orientation relationship was similar to that observed between β1 and the matrix phase.
3
Precipitates in Alloy AZ91
Binary Mg-Al alloys can contain up to 12 at.% Al in solid solution at about 430°C, but only about 1 at.% at 100°C, and there would appear potential to establish a large volume fraction of strengthening precipitates by appropriate thermal processing. However, during isothermal ageing in the temperature range 100-300°C, the precipitation process appears to involve solely the formation of relatively coarse particles of the equilibrium phase β (Mg17Al12). The β phase has a body-centred cubic structure (space group I43m , a ~ 1.06 nm), and has been reported [8] to form as plates on (0001)α. Based on SAED patterns, the orientation relationship between β and the matrix phase has been reported [9,10] to be such that (011)β // (0001)α, [1 11]β // [ 21 1 0 ]α, i.e. the Burgers orientation relationship that has been observed frequently in h.c.p./b.c.c. transformations. Several other orientation relationships have also been reported [9,11], but none of them are supported by electron diffraction patterns recorded from individual precipitates and the surrounding matrix phase.
Figure 2: Transmission electron micrographs showing (a) < 1120 >α, and (b) [0001]α images of (0001)α laths in samples of AZ91 aged for 8h at 200°C, and (c) electron microdiffraction pattern recorded from the precipitate in (b)
321 Systematic examination of a wide range of precipitates in samples of AZ91 (Mg-9wt%Al1wt%Zn-0.3wt%Mn) alloy, which were solution treated 24h at 425°C, cold water quenched, then aged at 200°C, indicated that most precipitates had a faceted lath morphology, with the broad surface parallel to (0001)α. Figure 2 shows projected images of the (0001)α laths in a sample aged 8h at 200°C, and an electron microdiffraction pattern recorded from a single precipitate lath. The microdiffraction pattern could be indexed according to the b.c.c. structure of β phase. Examination of a range of the precipitate laths revealed that they had a near Burgers orientation relationship with respect to the matrix phase, but one which departed significantly from the rational Burgers relationship [12]. Figure 2(b) shows an [0001]α projected image of an (0001)α precipitate lath. Since the precipitate lath was thin and embedded in the matrix phase, several sets of Moiré fringes, which were the product of double diffraction involving a precipitate reflection and an adjacent matrix reflection, were also visible within the precipitate. In this orientation, the precipitate laths were observed to have two faceted interfaces, which were defined by two sets of Moiré fringes, parallel to irrational planes in both the precipitate and matrix phases. Analysis of the image of the lath, Fig. 2(b), and the corresponding microdiffraction pattern, Fig. 2(c), indicated that the normal to the major precipitate/matrix interface (I) of the precipitate lath was parallel to the diffraction vector connecting the (1 100 )α and ( 033 )β reflections, while the normal to the minor precipitate/matrix interface (II) was parallel to the diffraction vector connecting the (10 10 )α and ( 4 11)β reflections. Examination of a wide range of such (0001)α laths revealed that the major precipitate/matrix interface was invariably parallel to the Moiré fringes defined by (1 100 )α and ( 033 )β reflections, and that the minor precipitate/matrix interface was always parallel to the Moiré fringes defined by (10 10 )α and ( 4 11)β reflections. Since ( 033 )β and ( 4 11)β are the most closely-packed planes in the β lattice, and {1 100 }α are the second most closely-packed planes in the α lattice, the resultant Moiré will be expected to define a low energy precipitate/matrix interface that contains a high density of atoms [13]. In addition to the (0001)α laths, a minor fraction of particles having the form of faceted [0001]α rods were also observed in samples aged 8h at 200°C, Fig. 3. Electron microdiffraction patterns recorded from these [0001]α rods could again be indexed consistently according to β phase, but the orientation relationship was of the form (1 10 )β // (1 100 )α, [111]β // [0001]α, which is identical to that reported by Crawley and Miliken [9]. Viewed along their long axis, i.e. in the [0001]α direction, the [0001]α rods had a hexagonal cross-section, with the bounding facets parallel to {1 100 }α=// { 330 }β. A few particles of β phase were occasionally observed to develop a rod shape with their long axes apparently inclined with respect to the [0001]α direction, Fig. 4. Electron microdiffraction patterns recorded from these rods implied an orientation relationship that was of the form (1 10 )β ~// (1 100 )α, [11 5 ]β ~// [0001]α.
322
Figure 3: Transmission electron micrographs showing (a) < 1120 >α, and (b) [0001]α images of [0001]α rods in samples of AZ91 aged for 8h at 200°C, and (c) electron microdiffraction pattern recorded from the rod in (b)
Figure 4: Transmission electron micrographs showing (a) < 1100 >α, and (b) [0001]α images of precipitate rods in samples of AZ91 aged for 16h at 200°C, and (c) electron microdiffraction pattern recorded from the rod in (b)
323
4
References
[1] I.J. Polmear, Light Alloys, 3rd ed., Arnold, London, 1995. [2] C.J. Bettles, P. Humble and J.F. Nie, in Proc. 3rd Int. Magnesium Conf. (ed.: G.W. Lorimer), The Institute of Materials, London, 1997, 403-417. [3] J.F.Nie and B.C. Muddle, in Proc. Magnesium Alloys and Their Applications (eds: B.L. Mordike and K.U. Kainer), Werkstoff-Informationsgesellschaft, Frankfurt, 1998, 229234. [4] R.A. Khosrhoshahi, R. Pilkington, G.W. Lorimer, P. Lyon and H. Karimzadeh, in Proc. 3rd Int. Magnesium Conf. (ed.: G.W. Lorimer), The Institute of Materials, London, 1997, 241-256. [5] G.W. Lorimer, in Proc. London Conf. on Magnesium Technology (ed.: G.W. Lorimer), The Institute of Metals, London, 1987, 47-53. [6] H. Karimzadeh, Ph.D. Thesis, University of Manchester, U.K. 1985. [7] J.F. Nie and B.C. Muddle, Acta Mater., 2000, 48, 1691-1703. [8] J.B. Clark, Acta Metall., 1968, 16, 141-152. [9] A.F. Crawley and K.S. Miliken, Acta Metall., 1974, 22, 557-562. [10] D. Duly, W.Z. Zhang and M. Audier, Phil. Mag. A, 1995, 71, 187-204. [11] J. Gjönnes and T. Östmoe, Z. Metall., 1970, 61, 604-606. [12] J.F. Nie, X.L. Xiao, C.P. Luo and B.C. Muddle, Micron, submitted, 2000. [13] P.H. Pumphrey, Scripta Mater., 1972, 6, 107-114.
Compression Test on Magnesium Alloy MgAl8Zn at High Strain Rates and Temperatures E. El-Magd and M. Abouridouane Department of Materials Science, RWTH Aachen
1
Introduction
Under quasi-static loading strain hardening causes an increase of force and acts stabilising on the deformation process. In case of dynamic loading, the deformation process is characterised by the propagation of mechanical waves and their reflections [1]. Furthermore, additional influences on the ductility of the material have to be taken into consideration. With increasing deformation rate, the strain rate sensitivity increases leading to a higher local value of flow stress and stabilises the deformation. On the other hand, the adiabatic character of the deformation process reduces the flow stress and promotes instability. Mass inertia in the lateral direction arise in connection with radial acceleration. This causes the initiation of either lateral tensile or lateral compressive stresses depending on the time function of the global strain of the specimen [2].
2
Material Behaviour at High Strain Rates and Temperatures
Quasi-static and dynamic compression tests were carried out on the high-strength wrought magnesium alloy AZ80 A-F (MgAl8Zn; nominal composition: 8.3 % Al, 0.6 % Zn, 0.2 % Mn, balance: Mg) at strain rates in the range of 0,001 s-1 ≤ ε ≤ 10.000 s-1 and temperatures between 20 °≤ T ≤ 300 °C. The investigated cylindrical specimens with a relation between the initial height and diameter H0 / D0 = 1 were machined with their load axis parallel to the extrusion direction of the initial material, which was available in two conditions: The as-delivered condition with a mean grain size of dG = 40 µm and dG = 120 µm, and an additionally homogenised condition (400 °C / 4 h, according to [3]) with dG = 40 µm. The experiments at lower strain rates ε ≤ 100 s-1 were carried out using a servo-hydraulic universal test system and the dynamic tests with strain rates ε > 1000 s-1 using a Split-HopkinsonBar arrangement. The experimentally determined flow curves of quasi-static ( ε = 0, 001 s-1) and dynamic compression tests for the as-delivered condition, dG = 120 µm are shown in Fig. 1. The stress of the quasi-static flow curves increases continuously with increasing strain due to strain hardening. With higher strain rates ε ≥ 1.000 s-1, the flow curves show on the one hand an increase of flow stress for lower strains and on the other hand a decrease of stress beyond a stress maximum due to the adiabatic character of the deformation process resulting in a temperature increase of the specimen for higher strains. At room temperature (T = RT), the 0.2% proof stress lies between 200 MPa and 250 MPa according to strain rate. The true compressive strain at fracture increases from 0.1 to 0.18 Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
325 when the true strain rate increases from 0,001 s-1 to 9.218 s-1. The higher the temperature the higher is the degree of influence of strain rate on strain at fracture in the dynamic compression tests. At temperatures of 250 °C and 300 °C, no fracture was observed in quasi-static compression tests even after reaching a true strain of 1,6. 500 T = RT
400
T = 100 °C mean true strain rate
mean true strain rate
-1
-1
9218 s -1 8260 s -1 7776 s -1 7321 s -1 6344 s -1 6340 s -1 4327 s -1 4326 s -1 2353 s -1 2348 s -1 0.001 s
300 200 100
MgAl8Zn Compression
True compression stress , MPa
0
8438 s -1 8194 s -1 7456 s -1 7356 s -1 6285 s -1 6188 s -1 4558 s -1 4527 s -1 2445 s -1 2388 s -1 0.001 s
MgAl8Zn Compression
500 T = 150 °C
400
T = 200 °C mean true strain rate
300
mean true strain rate
-1
8398 s -1 8318 s -1 7228 s -1 7179 s -1 6067 s -1 5876 s -1 4602 s -1 4449 s -1 4341 s -1 0.001 s
200 100
MgAl8Zn Compression
0 500
-1
8665 s -1 8626 s -1 7369 s -1 7309 s -1 6803 s -1 6736 s -1 4865 s -1 4815 s -1 0.001 s
MgAl8Zn Compression
mean true strain rate
mean true strain rate
-1
8796 s -1 8546 s -1 7825 s -1 7396 s -1 7248 s -1 6243 s -1 6133 s -1 5813 s -1 4608 s -1 4478 s -1 0.001 s
T = 250 °C
400 300 200 100 0
MgAl8Zn Compression
0
0.1
-1
7988 s -1 7787 s -1 7081 s -1 6481 s -1 6331 s -1 4632 s -1 4632 s -1 4612 s -1 2404 s -1 2174 s -1 0.001 s
T = 300°C
MgAl8Zn Compression
0.2
0.3
0.4
0.5 0
0.1
0.2
0.3
0.4
0.5
True plastic strain
Figure 1: Flow curves of MgAl8Zn at different mean true strain rates and temperatures
3
Constitutive Material Law for Dynamic Loading
In case of dynamic loading with strain rates higher than 2000 s-1, the influence of strain rate can be represented by a linear relation σ = σ h + ηε according to the damping mechanism, as it was confirmed e.g. in [4]. The parameter σh signifies the stress extrapolated from the range of high strain rates down to ε = 0 s-1; η is the damping parameter. The influence of a temperature increase on the flow stress can be considered by a multiplicative function [5], so that the material behaviour can be described by [6]
326 é
σ = é K(B + ε) n + ηε ù exp ê −β ë
û
êë
( T − T0 ) ù , Tm
(1)
ú úû
where B, K and n are material constants. Tm, T0 and T are the absolute melting point, room temperature and actual temperature respectively. β is a material constant which can be set to 3 for several materials [6]. Under dynamic loading, the deformation process can be considered to be approximately adiabatic. Assuming that the major part of deformation energy is transferred to heat during a dynamic deformation process and that the remaining part is consumed by an internal energy increase e.g. due to dislocation multiplication as well as mass acceleration, the temperature increase can be determined from ρ c dT = κ σ dε . In this equation κ ≈ 0.9 represents the fraction of energy transferred to heat; ρ and c are the density and the specific heat capacity of the material. With the relation between flow stress and temperature according to eq. (1), the increase of temperature can be determined as a function of strain by integration. The stressstrain relation under consideration of the two contrary influences of strain rate sensitivity and adiabatic character of the deformation results from the substitution of temperature in eq. (1): K ( B + ε ) + ηε n
σad =
1 + a ò éK(B + ε) n + ηε ù dε ë
,
(2)
û
with a = κβ /(Tm ρ c) . True compression stress , MPa
600 MgAl8Z n Compression T = RT
500 400 300
me an true strai n rate
200
K = 730 MPa B=0 n = 0.267 η = 0.011 MPa s -1 a = 0.00166 MPa
100 0
9218 8260 7776 6344 4327 4326
0
0.05
0.10
s- 1 -1 s- 1 s s- 1 -1 s s- 1
0.15
True plastic strain Figure 2: Comparison between computations following eq. (2) (continuous curves) and compression tests at room temperature with different mean true strain rates (markers) for MgAl8Zn
Fig. 2 shows a comparison between computational results following eq. (2) (continuous curves) and experimental results from compression tests (markers) for MgAl8Zn at different true mean strain rates. The needed material parameters used in the computations are listed in Fig. 2. A procedure for its systematic determination is shown in [7]. The deviation of the experimental data from the flow function at high strains indicates successive damage in connection with the starting fracture process.
327
4
Influence of Strain Rate and Temperature on Ductility of MgAl8Zn
Examples for the fracture mode of specimens after compression tests at room temperature are shown in Fig. 3. Both of the quasi-statically ( ε = 0,001 s-1) and the dynamically ( ε = 4.000 s-1) tested specimen show a low-ductility shear fracture with an inclination of nearly 45 °. While in the quasi-static case |∆H / H0|f reaches values of 0,09, the deformation at fracture increases up to a value of 0.16 when the strain rate reaches a value of ε = 4.000 s-1 (Fig. 3, diagrams).
Quasi-static
Dynamic 500 STRESS , MPa
STRESS , MPa
500 400 300
ε = 0.001s −1
200 100
0
0 .0 5
0.10
0.15
PLASTIC STRAIN
0
0.20
400 300
ε = 4000s − 1
200 100
0
0 .0 5
1 mm
0
0
0
0.10
0.15
0.20
PLASTIC STRAIN
1 mm
0.1 mm
0.2 mm
Figure 3: Damage and fracture modes within specimens of magnesium alloy MgAl8Zn, deformed at room temperature under quasi-static and dynamic conditions
The nominal deformation at fracture (-∆H/H0) in compression tests as a function of the nominal mean strain rate is represented in Fig. 4 for different temperatures. At room temperature the dependence of the deformation at fracture |∆H / H0|f on the strain rate indicates that with increasing strain rate the deformation at fracture increases. Under dynamic compression load, beyond a transition range of strain rates up to ε > 1000 s-1 in which the deformation at fracture shows a weak dependence on strain rate, the MgAl8Zn shows a much higher ductility than under quasi-static loading resulting from the stabilising effect of an increased strain rate sensitivity. With increasing temperature, the dependence of |∆H / H0|f on
328 temperature under dynamic compression load increases and higher values of |∆H / H0|f are reached. 40 d G = 40 µm as-delivered d G = 40 µm homogenized d = 120 µm as-delivered
d G = 120 µm as -delivered
G
30
T = 100 °C 20
T = 20 °C
10 MgAl8Zn Compression
Deformation at fracture (-∆ H/H0)f , %
MgAl8Zn Compre ssion
0 40 d G = 120 µm as-delivered
d G = 120 µm as-deliv ered
30
T = 200 °C
T = 150 °C 20 10
MgAl8Zn Co mp ression
MgAl8Zn Compression
0 40 d G = 120 µm as-delivered
dG = 120 µm as-deliv ered
30 20
T = 250 °C
T = 300 °C
No fracture up to -∆H/H = 80 %
No fracture up to -∆H/H = 80 %
0
0
10 MgAl8Zn Compression
0 -3 10
10
MgAl8Zn Compression -1
10
1
10
3
10
-3
10
-1
10
1
10
3
-1
Mean strain rate, s
Figure 4: Deformation at fracture as a function of the mean strain rate of MgAl8Zn
In case of quasi-static loading with ε = 0,001 s-1 the deformation at fracture increases continuously with increasing temperature, and at temperatures of 200 °C and higher no fracture could be detected even up to values of |∆H / H0| = 80 %. At a strain rate of ε = 0,01 s-1 |∆H / H0|f starts to increase at a temperature of 150 °C. For the strain rates of ε = 0,01 s-1 and -1 ε = 0,1 s the absence of fractures was observed at temperatures of 250 °C and higher.
329
5
Summary
Quasi-static and dynamic compression tests were carried out on magnesium alloy MgAl8Zn in order to investigate its flow behaviour and ductility in dependence on strain rate and temperature. Under dynamic loading conditions, the strain rate sensitivity of the material increases with increasing strain rate and stabilises the deformation process while the increase in temperature caused by the adiabatic character of the deformation process reduces the stress and promotes instability. A constitutive material model was introduced which allows the description of the adiabatic flow curves under consideration of these two contrary effects. Compression tests with MgAl8Zn showed, that at room temperature the deformation at fracture |∆H / H0|f increases with increasing strain rate, so that in dynamic compression tests with strain rates ε > 1.000 s-1 |∆H / H0|f was 1,6 times higher than under quasi-static loading conditions ( ε ≤ 0,1 s-1). At high strain rates, an increase of temperature leads to an increase of the dependence of deformation at fracture on strain rate resulting in higher values of |∆H / H0|f. In the range of quasi-static loading (0,001 s-1 ≤ ε ≤ 0,1 s-1) elevating temperature results in a corresponding increase of |∆H / H0|f. With increasing strain rate, the starting point of the increased values for |∆H / H0|f is shifted to higher temperatures. However, for temperatures of 250°C and higher, no fracture could be observed in quasi-static compression tests even up to deformations of |∆H / H0| = 80%.
6 [1] [2] [3] [4] [5] [6] [7]
References H. Lippmann, Springer Verlag, Berlin, 1981, 210-229 E. El-Magd, C. Treppmann, H. Weisshaupt, J. Phys IV, suppl. III, 1997, 7, 511-516 H. J. Bargel, Springer Verlag, Berlin, 1999, 702-707 K. Sakino, J. Shiori, J. Phys. IV, suppl. III, 1991, 1, C3-35 N. J. Petch, Phil. Mag. 1958, 8, 1089-1097 E. El-Magd, J. Phys. IV, suppl. J. Phys. III 1994, 4, 149-170 E. El-Magd, H. Scholles, H. Weisshaupt, Mat.Wiss. u. Werkstoff., 1996, 2, 408-413
Wear Behaviour of Laser Surface Treated Magnesium Alloys U. Kutschera DLR Stuttgart
R. Galun IWW, TU Clausthal
1
Abstract
Laser surface cladding and laser surface alloying are two methods to improve the surface wear resistance of magnesium alloys. AZ91, WE54 and cp-Mg were used as substrate materials. As additional materials elements forming hard and stable intermediate phases like Al, Si, Ni and Cu as well as SiC were chosen to produce surface layers with an improved wear resistance. The wear behaviour was investigated by pin on ring (POR) and pin on disc (POD) tests under different atmospheric and lubricating conditions. The results of these investigations will be presented in this paper.
2
Introduction
Since the beginning of the last decade the interest on light weight magnesium alloys increased especially at the automotive industry. A necessary reduction of fuel consumption led to a renaissance of magnesium as construction material [1]. Laser surface alloying and laser surface cladding are two possible treatments to modify the surface of a material. In this work the aim of both methods is to improve the wear resistance of the surface by addition of alloying elements, which form intermediate phases with magnesium or with other elements to produce fine dispersed hard particles in the layers.
3
Experimental Details
The investigations on laser surface cladding are running at the DLR-Stuttgart. For this process a solid state laser (Nd:YAG) with a max. power of 3 kW is used. The cladding material is added in form of powder. In a single step process the powder is transported by a powder feeder to the nozzle. At the tip of the nozzle a shielding gas is inserted, focussing the powder beam and protecting it as well as the melt pool from atmospheric oxidation. The focal point of the powder beam is set to be on the surface plane of the substrate [2]. Fig. 1 shows schematically the equipment and the process used for laser surface cladding. The process of laser surface alloying is similar to laser cladding. These experiments are running in Clausthal using a 5 kW CO2-laser. To investigate the wear resistance of the laser treated layers, two different wear tests are used. The testing equipment of the pin on disc test (POD, Fig. 2a) is placed in a vacuum Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
331 chamber to vary the environmental conditions from atmospheric air to space-like vacuum. These tests are also useful to understand generally the effect of oxygen on the occurring wear mechanisms. The sample with fine grinded surface finish is fixed in a rotating holder. As test partner a hardened steel sphere is used, fixed in a holder to avoid rolling. The steel sphere is pressed on the sample surface with an applied load of 1 N. This load results in a Hertzian pressure which is below the critical value for the magnesium substrate. The test runs for 1000 m.
laser optic powder nozzle
shielding gas nozzle
Figure 1: Laser surface cladding with powder
load
load
steel ball
steel pin sample
Figure 2a: Pin on disc test (POD)
Figure 2b: Pin on disc test (POD)
Table 1: Process parameters of the pin on disc test (POD) and pin on ring test (POR) POD POR load 1N 27 N revolutions/distance 26540 / 1000 m 5300 / 1000 m track diameter 12 mm 60 mm velocity 6 m/min 11 m/min test partner hardened steel sphere, d = 6 mm hardened steel pin, r = 12 mm lubricant ---/motor oil 5W40 The influence of lubricants on the wear process is investigated with the pin on ring test (POR, Fig. 2b). The rings, cladded and alloyed at the periphery and with a fine grinded
332 surface finish, are fixed on a rotating axis. The test partner, a hardened steel pin, is pressed on the ring with a load of 27 N. This is also below the critical Hertzian pressure of the magnesium substrate. The sample rotates for a distance of 1000 m. Optionally an oil sump can fixed at the equipment to lubricate the sample. The parameters of both wear tests are summarised in table 1. During both tests the friction coefficient as a function of time is recorded. To get information about the volume loss caused by wear the profile of the wear tracks was measured by a profilometer. The microstructure of the wear tracks is characterised by SEM to get some information about potential wear mechanisms. For the laser surface treatment casted Mg of commercial purity (cpMg,) casted AZ91 and WE54 are used as substrate materials. Table 2 gives an overview of the cladding and alloying materials used in the form of powder mixtures or alloy powders (AlSi30 only). The composition of the tested surface layers is listed in table 3. Table 2: Layer materials investigated during POR / POD tests Cladding materials Alloying materials WE54 + AlSi40 Al+Cu WE54 + AlSi40+SiC Al+Ni AlSi30 Al+Si Table 3a: Surface cladded layers; approximated content (wt%) of elements Name Mg Al Si (except SiC) SiC rare earth AlSi30 -70 30 --0SiC 77 10 4,5 -7,6 5SiC 73 9,5 4,3 5 7,2 10SiC 69 9 4 10 6,8 15SiC 65 8,5 3,8 15 6,5 Table 3b: Surface alloyed materials; approximated content (at%) of elements (EDX) Name Mg Al Cu Ni Si WE(Al+Cu) 79 10 11 --Mg(Al+Ni) 47 38 -- 16 -AZ(Al+Ni) 44 44 -- 12 -WE(Al+Si) 32 43 -- -- 25
4
Results and Discussion
4.1
Pin on disc test
In the test series in air an increase of wear resistance compared to the untreated substrates takes place for all laser treated materials. In Fig. 3 the volume loss of the tested surfaces is displayed. As example: AlSi30 cladded on AZ91 shows a decrease of the volume loss of nearly 38 % and cladded on WE54 a decrease of nearly 57 %. In air the dominant wear mechanism primary is adhesion, followed by an oxidation of the wear particles, resulting in hard and abrasive oxide particles (Fig. 4a). The best results in air can be achieved by layers laser alloyed with Al+Ni. They show no adhesion wear but the
333 formation of a good bonded oxide layer (Fig. 4c). The steel spheres generally show small wear marks due to abrasion and adhesion to the tested materials. Only the spheres tested on SiC-dispersed layers show strong abrasion caused by the hard particles. The results in vacuum are remarkable. For nearly all tested materials the wear attack is significantly reduced. For the laser claddings with AlSi30 no measurable volume loss occurred (Fig. 4b). The general reason is the missing oxidation of the adhesion particles,which remain soft and only adhesion and redeposition but no abrasion occurred. The only exception in the test series are the layers alloyed with Al+Ni. Here the good bonded oxide layer is not formed and so these layers show a similar volume loss by adhesion wear like most of the other layers. Pin on Disc
volume loss [mm³]
3 2,456
2,5 2
1,854
1,5
1,16
1,16
1,07
1 0,5
0,073
0,03 0,109 0,043 0,006 0,069
0
0,97
0
A Z9 1 su A bs Z9 tra 1 te su ai bs r tra W te E5 va 4 c. su W b str E5 at 4 ea su bs ir W tra E5 te va 4+ c. (A W L E5 +C 4+ u) (A air L+ Cu M )v g+ ac (A . l+ M N g+ i) (A ai r l+ N i ) A va Z+ c. A lS A i 3 Z+ 0 ai A r lS i3 0 W va E+ c. A lS W i 3 E+ 0 air A W lS i3 E+ 0 (W W va E+ c. E5 (W 4+ A E5 l) 4+ ai A r lS i6 0) ai r
0
Figure 3: Volume loss of different materials tested on the POD wear test
a) AlSi30 / air b) AlSi30 / vacuum Figure 4: Wear tracks of the POD test (SEM)
c) Al+Ni / air
Also the samples behave different for longer testing cycles in air. After several hundred revolutions the friction coefficient increases from a constant low level, whereas it remains on the same level during the whole test in vacuum (Fig. 5). The reason for this phenomenon might be the formation of oxidised particles which increase friction and wear attack.
334 POD (AlSi30-air)
POD (AlSi30-vacuum) 1,2 Friction coefficient
Friction coefficient
1,2 1 0,8 0,6
1 0,8 0,6 0,4
0,4 0
200
400
600 800 Distance [m]
1000
0
1200
200
400
600 800 Distance [m]
1000
1200
Figure 5: Friction coefficient during the POD wear test of AlSi30 in air and under vacuum conditions
4.2
Pin on ring test
The results of the non-lubricated POR tests are comparable with the POD tests in air. Except the Al+Cu alloyed materials, all laser surface treated materials show an improvement of wear resistance (Fig. 6). Pin on Ring 30
volume loss [mm³]
25,4 25 20
16,2
10
14,9
14,2
15
6,5
14,5
12,5 9,8
8,5
7,8 7,5
5
0
0
0
0
0
0
0
AZ A 91 Z9 s u b 1 su str W bstr ate E5 at e E5 4 su oil bs 4 su tra bs tr te W ate o E W +( i l E+ Al ( A +S l+ i) AZ Si) AZ +(A oil +( l+ Ni A l W +N ) E i) W +(A oil E+ l (A +Cu W l+C ) u E W 54 ) oi E5 +A l 4+ lS Al i30 Si 30 oi l 0S i 0S C iC oi l 5S i 5S C iC oi 15 l S 15 iC Si C oi l
0
Figure 6: Volume loss of different materials tested at the POR wear test
Due to the lower magnesium content the improvement is more significant for laser alloyed materials than for cladded layers. Under lubricated conditions with 5W40 motor oil no volume loss occurs at the laser treated rings, even at elevated temperatures up to 140 °C, whereas the untreated substrates show again a volume loss by the wear attack. Only a smoothing of the surface can be detected at the alloyed and cladded layers and in the case of the SiC-dispersed layers absolutely no wear marks are found. The steel pins generally show minimal wear attack except at the pins of the SiC-dispersed materials. Their strong volume loss is caused by abrasion based on the SiC particles. In addition to the wear attack for several laser cladded and alloyed layers also a chemical reaction between the oil and the wear particles occurs, visible by a dulling of the oil. The reaction mainly occurs for these layers with the highest magnesium content, the magnesium seems to be a reactive partner. Further investigations have to be made to find out the type of reaction and evaluate its influence on the lubricating properties of the oil.
335
5
Conclusions
The surface treatments of laser cladding and laser alloying are found to be promising methods to improve the wear resistance of magnesium base alloys. Except of the alloying combination Al+Cu all layer materials are able to improve the wear resistance. Laser alloying with other element combinations leads to better results than laser cladding with magnesium containing powders. By a reinforcement with SiC-particles the wear resistance of these layers can be increased significantly, but here the wear attack at the pins is very strong. The comparison of the results of the POD tests in air and vacuum as environmental conditions shows that a good bonded tribooxide layer can improve the wear resistance (laser alloying with Al+Ni). But if the bonding is not good enough particles will break out from the surface and increase the wear attack due to an additional abrasion (laser cladding with AlSi30). Further investigations have to be focussed on the laser claddings containing SiC or other hard particles and on some different laser alloyed layers. But a compromise has to be found between the wear attack of the laser treated layers and of the tested material, which should be a real component of a possible application. Also the investigations of the chemical reaction between magnesium containing materials and lubricants like motor oil have to be intensified.
6
Acknowledgements
The investigations on wear resistance of the laser treated layers were carried out at the Aerospace Materials Technology Testhouse (AMTT), Seibersdorf, Austria and was supported by the European Commision under Contract No. FRBFMGECT980141 at the program „Training and Mobility of Researchers (TMR)“.
7
References
[1] H. Friedrich, S. Schumann, Proc. 2nd Israeli Int. Conf. on Magnesium Sci & Technol., 22.-24. Feb 2000, Dead Sea Israel, 9-18 [2] Arnold, J. et al., J. Thermal Spray Technol., 8 (2), June 1999, 243-248 [3] Galun, R. et al., Metall, 53, 1999, 672-675
Mechanical Behavior and Residual Stresses in AZ31 Wrought Magnesium Alloy Subjected to Four Point Bending J.P. Nobre/U. Noster*/J. Gibmeier*/I. Altenberger*/M. Kornmeier/A. Dias/B. Scholtes* Department of Mechanical Engineering, University of Coimbra, Portugal *Institut of Materials Technology, University Gh Kassel, Germany
1
Abstract
Several specimens of AZ31 magnesium alloy were taken in different directions from a rolled plate and subjected to tensile, compressive and four-point bending tests. Tensile and compressive tests allowed to characterize the mechanical properties of the material in the rolling and cross rolling directions. The bending tests were carried out in four different specimen types to take the effect of the anisotropy induced by the rolling procedure itself into account. Six strain gages per specimen allowed to observe the evolution of the strain with the bending moment during the tests. All bending specimens were bent until a total compressive deformation of 2.5%. It was observed that the corresponding tensile strain was significant lower and slightly different for each specimen type. In addition, a characteristic non-uniform distribution of deformation twinning was observed. The induced residual stresses after bending were characterized by X-ray diffraction (XRD) and incremental hole-drilling (IHD). Due to the different mechanical behavior in tension and compression, an asymmetric residual stress distribution after bending could be observed. The neutral axis was, in all cases, shifted towards the tensile side. This observation agrees with the strain measurements during bending tests.
2
Introduction
The relatively high strength-to-weight ratio is the most attractive characteristic of magnesium alloys. Consequently, their use plays meanwhile an important role in structural applications in, e.g., automotive and aerospace industries. The characterization of the mechanical behavior of these materials is, therefore, of great interest. Typical applications of AZ31 wrought magnesium alloys (∼ 3% Al, ∼ 1% Zn) are forgings and extruded bars, rods, shapes, structural sections, sheets and plates with good formability and moderate strength, high resistance to corrosion (AZ31B) and good weldability [1]. Magnesium has a hexagonal close-packed crystalline structure (hcp). The limited number of slip systems in hcp crystals is the reason for their extreme orientation dependence of, e.g., the mechanical properties and relatively low ductility [2]. For that reason, twinning has an important role on the mechanism of plastic deformation of magnesium alloys. The effect and importance of textures on plastic deformation of pure polycrystalline magnesium were reported in [3]. However, extrusion and rolling procedures seem to have less influence on the mechanical properties of the AZ31 magnesium alloys, in different directions relatively to the rolling direction, than in the case of pure magnesium [4]. Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
337 In this work the mechanical behavior of an AZ31 wrought magnesium alloy subjected to monotonic tensile, compressive and four point bending was analyzed at room temperature. In addition, the residual stress distribution after bending was determined.
3
Experimental Procedure
Four different specimen types (A, B, C and D) were cut from a rolled plate of AZ31 magnesium alloy, as shown in figure 1a. This way, the effect of the rolling procedure on the mechanical behavior of the AZ31 alloy can be studied. Specimens taken out of the plate like bending specimens A and C, respectively, were used in tensile tests, while specimen types A, B, C and D were used in bending tests. Before bending, the specimens were subjected to electrolytic polishing to remove the surface layer affected by the machining procedure (about 300 µm on each side). ROLLING DIRECTION B
D
A
Strain gages
C
10 x 1 20 (A,C); 15 (B,D) z
160 Tensile side
AZ31 magnesium alloy plate F/2
a)
6 5
4 3
50
F/2
F/2
a
b
50
y 2
50
F/2
b)
Figure 1: a) Specimen types used in four point bending tests. A, B, C and D – tensile surfaces of the bent specimens. b) Geometry of the specimens, strain gage positions and loading conditions
The geometry of the specimens and the loading conditions during the bending tests are shown in figure 1b. Six strain gages were bonded on the material’s surface to observe the evolution of the strain with the bending moment in the tensile and compressive surfaces and along the lateral sides of the specimens. In all cases, the tests were carried out until an imposed total compressive strain of 2.5%. X-ray diffraction (XRD) and incremental hole-drilling (IHD) techniques were used to characterize the residual stresses after the bending test. For the IHD technique, high speed drilling equipment (milling guide RS-200* with air turbine) and a strain gage rosette (CEA-06-062UM-120*, *Measurements Group, Inc.) were used. In depth increments of 0.04 mm, strain relief was measured to about 1 mm below the surface. The typical hole diameter was about 1.8 mm. Two residual stress calculation procedures were applied, the power series method [5] and the differential method [6]. XRD residual stress analysis was performed on the tensile, compressive and lateral surfaces of the specimens. Lattice deformations of the Mg {213} planes were determined on a conventional ψ-diffractometer for 11 ψ-angles between -60° and +60° using CuKα radiation. Residual stresses were calculated for the plain stress conditions using X-ray elastic constants ½ s2 = 2.93×10-5 MPa-1 and s1 = -6.59×10-6 MPa-1. Two collimators of different diameters were used: 2 mm and 3.2 mm.
338
4
Results and Discussion
Figure 2a shows the results of the tensile and compressive tests. It is obvious that the material presents a completely different plastic behavior in tension and compression.
Engineering Stress [MPa]
400
300
200
compression (CRD) compression (RD) tension (CRD) tension (RD)
100
0 0
a)
2
4 6 8 10 Engineering Strain [%]
12
b)
Figure 2: a) Results of the monotonic tension and compression tests of the AZ31 alloy. RD – rolling direction; CRD – cross to rolling direction. b) Microstructure of the material after a comparable stress level in tension (125 MPa) and compression (-110 MPa)
The compressive yield strength reaches only around 50% of the tensile yield strength. These differences are typical for wrought magnesium alloys, while cast magnesium alloys show nearly identical tensile and compressive yield strength [7]. Specimens cut in the rolling direction show slightly smaller values (Rp0.2 = 92 MPa (compression); Rp0.2 = 160 MPa (tension)) compared to those cut cross to the rolling direction (Rp0.2 = 99 MPa (compression); Rp0.2 = 195 MPa (tension)). On the other hand, tensile strength was about 278 MPa in cross rolling direction and 265 MPa in rolling direction. These observations agree with those reported by Closset et al. [4]. In the optical micrographs of figure 2b, a different twinning distribution can be seen when the material is subjected to the almost same stress level in tension and compression. While no twinning occurs for the tensile case (above), a great density of deformation twins can be observed for the compressive case (below). Deformation twinning occurs more easily when the material is in compression and, therefore, orientation changes resulting from twinning may place new slip systems in a favorable orientation with respect to the stress axis, allowing additional slip to take place [2]. Figure 3 shows the apparent stress versus the total strain measured during the four point bending tests. The apparent stress was calculated using the assumptions of the simple beam theory for pure elastic bending [8], i.e.:
339 σx = −
Mzy Iz
(1)
where Mz is the bending moment (constant between points a and b – see figure 1b), y is the distance between the neutral axis and the corresponding point where the deformation is determined (corresponding to the middle point of each strain gage area) and Iz is the area moment of inertia – see fig. 1b. 300
300
Stress [MPa]
Stress [MPa] 200
200
100
100
Total strain [%]
Total Strain [%] 0
3
-2
-1
0 0
1
2
3
-3
-2
-1
0
1
2
3
Strain gage 1
-100
Strain gage 2
-100
Type B (RD-90º)
Strain gage 3
-200
Strain gage 4
Type A (RD)
-200
Type C (CRD)
Strain gage 5 Type D (CRD-90º)
Strain gage 6
-300
-300
a) b) Figure 3: Apparent stress versus total strain during bending tests. Tests were carried out until an imposed total compressive strain of 2.5%. a) Results of the six strain gages obtained in the type C specimen. b) Comparison of the results obtained in the tensile and compressive sides for all specimen types. See also figure 1.
Figure 3a shows, as an example, the results obtained in the specimens of type C (cross to rolling direction). The apparent stress was plotted against the total strain measured by the six strain gages, bonded as shown in figure 1b. It can be observed that, for a total compressive strain of 2.5%, the corresponding tensile strain was significantly smaller (less than 1.5%). Regarding the results of the tension and compression tests, plastic flow should begin in the compressive surface and grow into the interior. Only after a certain layer has been plastically deformed in the compressive side, the plastic flow should begin in the tensile surface. As a result of this asymmetric behavior, the neutral axis is shifted towards the tensile side. This shift can be verified by the inversion of the strain signal (tension→compression) of the strain gage number 4 (see also figure 1b). This behavior was observed for all specimen types and bending tests. Furthermore, the non-linear unloading behavior seems to indicate a possibly strong Bauschinger effect. For an accurate analysis of the elastic-plastic bending, factors such as the strain-hardening behavior, the shift of the neutral axis from the centroidal axis and the effect of transverse stresses must be considered [8]. Therefore, in the plastic region, the stress given by equation (1) is no longer valid but still allows a comparison between all specimens types. In figure 3b is shown a comparison between the results of all specimen types, considering only the strain measured in the tensile and compressive surfaces (strain gages 1 and 2 – see figure 1b). All specimen types present the same asymmetric tensile and compressive behavior. No substantial differences could be found between all specimen types, which indicates a small influence of the rolling procedure on the mechanical properties of the AZ31 in different directions, only influencing the plastic anisotropy in tension and compression. The slight differences in the tensile strain response could be related with the possible minor differences on the
340 homogeneity of the grain size depending on the rolling direction. However, those differences would be considerably higher for pure magnesium [3-4].
-60
Residual Stress [MPa] -20 0 20 40
-60
-40
Residual Stress [MPa] -20 0 20 40
60
0 -1
Tensile side
-3 -4 -5 -6 -7 -8 -9 -10
60
-2
NA centroidal axis
y [mm]
y [mm]
0 -1 -2
-40
Tensile side
-3
NA
-4 -5
centroidal axis
-6 -7
Compressive side
-8
Compressive side
-9 Longitudinal Residual Stress - 3.2 mm collimator Longitudinal Residual Stress - 2.0 mm collimator Longitudinal Residual Stress at surface - 3.2 mm collimator Longitudinal Residual Stress at surface - 3.2 mm collimator IHD - Differential method IHD - Power series method
-10
Figure 4: Residual stresses obtained by XRD and IHD in two rolling direction specimens (type A) (left) and one cross to rolling direction specimen (type C) (right)
The main characteristic of the bending tests is the inhomogeneous (non-uniform) nature of the deformation, since the strain and stress at a given point depend on its location relatively to the neutral axis of the cross-sectional area of the specimen. When the bending moment is removed, the layers that are subjected to more deformation prevent those less deformed to recover their initial length and, therefore, macro residual stresses will be present in the specimen. Figure 4 shows the longitudinal residual stresses obtained by X-ray diffraction and incremental hole-drilling in the specimens type A (rolling direction) and C (cross to rolling direction) after bending. Similar results were obtained for the other specimen types. The interrupted lines are only an approximate extrapolation from the measured points obtained by XRD. Nevertheless, these results clearly show an asymmetric residual stress distribution thus confirming the asymmetric results of the strain measurements during the bending tests and the shift of the neutral axis (NA).
5
Conclusions
All specimen types of a AZ31 wrought magnesium alloy subjected to four point bending presented a clearly asymmetric behavior in the tensile and compressive regions. The obtained longitudinal residual stress distribution confirmed this asymmetric behavior, according to the strain measurements during bending. The neutral axis was always clearly shifted towards the tensile side. The rolling procedure itself does not seem to have great influence on the plastic deformation behavior of these alloys in different directions, as the hcp structure presented by the magnesium could imply. However, a different twinning distribution was found when the
341 specimens were subjected to the same stress level in tension and compression, at room temperature.
6
References
[1] S. Housh, B. Mikucki, A. Stevenson, ASM Metals Handbook 9th ed. 1991 Vol. 2, p. 480. [2] G. Dieter, Mechanical Metallurgy, 2nd ed., McGraw-Hill, Tokyo, 1976, p. 118, 135. [3] R. Gehrmann, G. Gottstein in Proc. 12th Int. Conf. Text. Mater, Montreal, Canada, 8-19 August, 1999 [4] B. Closset, J. Perey, C. Bonjour, P. Moos in Proc. of Magnesium Alloys and their Applications (Ed.: B. Mordike, K. Kainer), Verlag Werkstoff-Informationsgesellschaft, Frankfurt, Germany, 1998, 195-200. [5] G. S. Schajer, Journal of Eng. Mat. and Tech. 1988, 110 (4), 338-349. [6] T. Schwarz, H., Kockelmann, Meßtechnische Briefe (HBM) 1993, 29 (2), 33-38. [7] S. Housh, B. Mikucki, A. Stevenson, ASM Metals Handbook 9th ed. 1991 Vol. 2, p. 456. [8] P. Dadras, ASM Metals Handbook 9th ed. 1991 Vol. 8, p. 118.
Superplasticity of Magnesium-based Alloys Ulrich Draugelates, Antonia Schram and Claus-Christian Kedenburg Institute of Welding and Machining, (ISAF), Clausthal, Germany
1
Introduction
The increasing demand for high strength light weight products which can easily be produced at low cost with respect to a reduction of energy consumption and material saving has caused a growing interest in superplastic forming in the fields of the automobile and aviation industry. The suitability of this production process to manufacture thin-walled structural components for light weight applications is said to be the reason for the growing interest in the technology of superplastic forming. Magnesium alloys fulfill the requirements of modern light weight construction materials with properties such as low density, good castability and a large recycling potential. The limited reforming properties of magnesium alloys can be associated with its hexagonal lattice structure which causes problems during processing. The utilization of the superplastic deformation behavior of magnesium materials can give remedy to this limitation because this way the productivity of the mechanical processing can be increased considerably in comparison to conventional forming processes. Superplastic forming of magnesium is a perfect and cost saving process for near-net-shape forming as an alternative for complicated machining or joining processes. As of now, there is only limited knowledge about the superplastic behavior of magnesium alloys which add a further decisive contribution to weight reduction of components due to the fact that their density is 30% less than that of aluminum alloys. The exploitation of the superplastic behavior is especially reasonable for magnesium alloys because of their restricted cold-working properties.
2
Superplasticity
Superplastic materials are polycrystalline solids which have the ability to undergo large uniform strains prior to failure without constriction and practically no strain-hardening, that exceed the usual limits of „normal-plastic“ materials of 10 to 40 % by some 100 to more than 1000%. Another characteristic of superplastic behavior of materials is the strong dependence of yield stress in relation to the elongation speed. This dependence is described in the StrainRate-Sensitivity-Parameter m (m-value) that is part of the valid flow equation for structural superplasticity (1) with the work hardening exponent n and a material constant k:
σ = k ⋅ε n ⋅ε m (1) Superplastic forming requires temperatures as high as or above 0.6 Tm which thermally activate recovery- and recristallization- processes so that the work hardening exponent n
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
343 drops to 0 and consequently εn adopts the value 1. Thus equation (1) can be simplified to (2) which is valid for structural superplasticity [1]. σ = k ⋅ε m (2) The yield stress σ, the force F and the strain rate ε describe the time dependent reduction of the specimen diameter A0 related to the initial diameter A0. As for areas of microscopic constriction (diameter A) equation (2) is also applicable. Equation (3) describes the relation of the time dependent variation of the specimen diameter as a function of the m-value and the diameter ratio. 1− m æ = çç A0 è
A
A0 ö m ÷ A ÷ø
(3)
As far as superplastic materials are concerned m=0.3 is a characteristic m-value, therefore a linear relation exists between the strain rate ratio and the cross section ratio. The formation of a local reduction of the cross section during the process of superplastic deformation leads to an increase of the local strain rate according to (3) and considering the valid yielding equation (2) an increase of the yield stress in the zone of constriction can be perceived. Due to the increase of the local yield stress, the superplastic deformation shifts to areas where a lower force counteracts the deformation velocity [2]. The extremely high strains which occur in materials with superplastic properties can be related to softening as well as hardening processes. A high m-value is, however, only a prerequisite for high strains but not necessarily a guarantee. Because of recristallisation, diffusion processes, and grain defects such as cavities, the m-value may be reduced during the deformation process which in turn may lead to an early material failure caused by local constrictions. As of now many plausible hypotheses to explain the origin of superplasticity have been developed but none of them is capable to exactly describing either mechanical or microstructural processes which take place during superplastic deformation [3, 4, 5]. The main mechanism of superplastic deformation is grain boundary sliding [3, 6]. Although other processes in addition to grain boundary sliding are responsible for the superplastic material deformation, grain boundary sliding accounts for 60% to 80% of the total deformation depending on the material. During the deformation process, stress peaks occur at grain boundary intersections and at irregularities of grain boundaries which are overcome by accommodating deformation processes (diffusion and dislocation processes) in the case of an ideal superplastic deformation process.
3
Experimental Procedure
Since the investigated ZRE1 magnesium alloy in the cast condition showed a very coarse grain structure with an average grain size of d > 250µm (see Fig. 3a) the following process was applied to refine the microstructure. From the cast ingots of a conventional ZRE1 magnesium alloy extruding billets were manufactured with a diameter of 75mm. The billets were heat treated and cooled down to room temperature in air. After a preliminary heating of 350°C for one hour the heat treated billets were extruded in a 400t hot axial plodder. The billets were reduced by a 2-step nozzle to a final diameter of 14mm, which corresponds to an
344 extrusion ratio of 1:26. The press velocities were between 3.0 and 5.2m/min. The examined tensile specimens were cut from the extruded sections parallel to the extruding direction. 3.1
Determination of superplastic properties
The determination of the m-value is affected by the determination of the incremental slope of the logσ = log ε- flow curve (fig. 1b). The slope of the inclining curve results in the m-value and is hereby a precondition for the quantification of the superplastic behavior of the tested materials. The flow curve is constructed by the determination of (σ; ε )- value pairs (fig. 1b), that are determined by velocity change tests [1, 5, 6, 7].
Figure 1: Determination of the m-value
At first, the actual flow stress σn under consideration of volumetric constancy derived from the actual force Fn of the initial length l0 and the actual elongation ln as well as the initial cross section S0 is calculated (4). σn =
Fn (l0 + l n ) l0 ⋅ S 0
(4)
The actual elongation rate ε n is calculated with the aid of crosshead speed vn (5). εn =
vn l0 ⋅ l n
(5)
In contrast to [5], the calculation of the m-value is not effected by (σ; ε)-value pairs that are characterized by identical elongation but by (σ; ε )-values (6) that had been stored when reaching a maximum flow stress, [8]. σn σ è n +1 æ
m =
ln çç
εn
ö ÷ ÷ ø
ö ÷ ÷ è ε n +1 ø æ
ln çç
(6)
Supposing that the slope between two value pairs is constant, the value will be calculated directly as the differential quotient of these value pairs. Many investigations [1, 2, 3] reveal a large dependency of the maximum elongation on the strain rate during the superplastic deformation. For this reason the maximum elongation to fracture of the modified ZRE1 magnesium alloy was determined by tensile tests with constant strain rates at temperatures from 290°C to 530°C. For that kind of tensile tests the change of the cross head speed can be described by equation (7). As opposed to normal tests with a constant cross head speed it is possible to obtain a very short testing time while still achieving
345 large elongations. During the tensile tests with a constant strain rate the cross head speed is constantly changed by a PC control system. (7) v (t ) = l0 ⋅ ln(1 + ε ) ⋅ (1 + ε )t
4
Results
4.1
Microstructure
The magnesium alloy ZRE1 was chosen for the examination because of its large content of zinc and the resulting relatively fine microstructure. Optical and scanning electron microscopic investigations of the magnesium base alloy ZRE1 in the as cast condition show precipitations at the grain boundaries with an average grain size of d > 150µm, fig. 2a. The precipitations at the grain boundaries are dissolved by a two step heat treatment of the cast material resulting in a very homogeneous but particularly rough microstructure with a grain size of d > 300µm, fig. 2b. Fig. 2c shows the microstructure after the extrusion process. The average grain size of the resulting extruded bar is reduced to d = 7.6 µm, which represents an excellent prerequisite for good superplastic forming characteristics. The investigations of the microstructure after the superplastic deformation showed that the grain size of the tensile specimen in the deformed area was even more reduced by recrystallisation processes during the constant strain rate tests. Thus it can be assumed that the superplastic characteristics of deformed tensile specimens can not only be preserved but even improved by further grain refinement.
Figure 2: Overview of the crystalline structure of ZRE1 (a: as cast, b: cast heat treated, c: heat treated and extruded )
4.2
m-value ZRE1
The determination of strain rate dependencies (m-value) of the modified ZRE1 magnesium alloy by velocity change tests showed a maximum m-value of 0.86 at a forming temperature of 480°C. In general the modified ZRE1 alloy reveals an m-value > 0.3 within a large range of parameters and thus superplastic characteristics. Even at a relatively high constant strain rate of 5×10-3s-1 and at a forming temperature of 460°C the m-value of the modified alloy is as high as 0.35, fig. 3.
346 1 300°C 340°C 380°C 420°C 460°C 480°C
0,9 0,8
m-value
0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 1,0E-04
1,0E-03
log
1,0E-02
-1
strain rate [s ] Figure 3: m-values of the optimized ZRE1 magnesium alloy at different forming temperatures and strain rates
4.3
Elongation to fracture of the optimized ZRE1 magnesium alloy
The maximum elongation to failure achieved for the modified ZRE1 magnesium alloy are shown in fig. 4. The tensile tests carried out at constant strain rates revealed a minimum elongation of 200% at temperatures above 380°C within the entire examined strain rate rage. A maximum elongation of 1050% was achieved at a strain rate of 1.6×10-4 s-1 and a forming temperature of 500°C, fig. 5. 1200 1000
290°C
380°C
410°C
440°C
470°C
500°C
elongation [%]
530°C
800 600 400 200 0 1,0E-04
1,0E-03
1,0E-02 -1
strain rate [s ] Figure 4: Elongation to fracture of the optimized ZRE1 magnesium alloy at constant strain rates and different forming temperatures
347
5
Conclusions
The investigations showed that it is possible to generate superplastic characteristics in a ZRE1 magnesium alloys by modification of the microstructure. For this purpose a procedure was developed, which makes it possible to significantly improve the superplastic behavior of the ZRE1 magnesium cast alloy by a refinement of the microstructure to a grain size d < 10µ m. The results corroborate the theory that the grain size has the main influence on the superplastic deformation characteristics. By means of an appropriate heat treatment in combination with a extrusion process it was made possible for a conventional ZRE1 magnesium alloy to achieve an elongation to fracture of 1050% during a tensile test with a constant strain rate of 1.6×10-4 s-1 and a forming temperature of 500°C, fig. 5.
Figure 5: Tensile specimen ZRE1; A: unbiased, B: elongation ε = 1050% at a constant strain rate of 1.6×10-4 s-1 and a forming temperature of 500°C
6
References
[1] J. Pilling, and N. Ridley, Superplasticity in Crystalline Solids (Houghton, Michigan: Metallurgical Engineering, Michigan Technological University, 1988), 12–32. [2] K. Mukherjee, “ Deformation mechanisms in superplasticity”, Ann. Rev. Mater. Sci., 9 (1979), 191-217. [3] Arieli, and A. K. Mukherjee, “A Model for the Rate-Controlling Mechanism in Superplasticity”, Materials Science and Engineering, 45 (1980), 61-70. [4] B.P. Kashyap, A. Arieli and K. Mukherjee, “Microstructural aspects of superplasticity", Journal of Materials Science, 2 (1985), 2661-2686. [5] W.-A. Backoven, I.-R. Turner and H. Avery, “Superplastic in an Al-Zn Alloy”, ASM Trans., 57 (1964), 980-990. [6] Arieli, and A. Rosen, “Measurements of the Strain-Rate-Sensitivity Coefficient in Superplastic Ti-6Al-4V Alloy”, Scripta Metallurgica, 10 (1976), 471-475. [7] J. Hedworth and M.J. Stowell, “The Measurement of Strain-Rate Sensitivity in Superplastic Alloys,” Journal of Material Science, 6 (1971), 1061-1069. [8] U. Giegel, “Superplastische ultrahoch borlegierte Stähle,” (Ph.D. thesis, Technical University of Clausthal, 1992).
Cyclic Deformation Behavior of the Cast Magnesium Alloy AZ91 H. W. Höppel, G. Eisenmeier, B. Holzwarth and H. Mughrabi Institut für Werkstoffwissenschaften, Lehrstuhl I, Universität Erlangen-Nürnberg, Martensstr. 5, D-91058 Erlangen, Fed. Rep. of Germany
1
Introduction
In the last few years, due to their low specific weight, magnesium alloys were "re-discovered" for light-weight constructions of components in the automotive industry. For example, they are currently used for steering wheels, cross-beams, gear boxes, door structures and oil sumps [1]. One of the most common magnesium die-casting alloys in use is the so-called AZ91 alloy (ASTM-notation [2]), with about 9 wt.-% Al, 1 wt.-% Zn and 0.5 wt.-% Mn. In engineering applications, cyclic load spectra must be considered in many cases and, therefore, the fatigue behavior and, in particular, damage mechanisms during cyclic deformation are of common interest for this alloy. This report deals with the cyclic deformation behavior of the alloy AZ91 at room temperature (20 °C) and at 130 °C and is complemented by a detailed characterization of the microstructure before and after fatigue testing.
2
Experimental
The AZ91HP alloy (HP: high purity, i. e. low content of iron, copper and nickel), provided by the AUDI AG, Ingolstadt, Germany, was manufactured by using a so-called vacuum diecasting system. The evacuation of the die leads to a lower content of porosity due to gaseous inclusions in comparison to the conventional casting process, as reported in [3]. Unfortunately, shrinking porosity (cavities), due to the high volume leap during solidification cannot be avoided by using this technique. The obtained microstructure of the cast alloy, which is shown in fig. 1a, consists of Al-poor dendrites with an eutectic Al-rich solid solution and the intermetallic phase Mg17Al12 (βphase) in between, as reported by Kopp [4]. In addition, contraction cavities are observed in the interdendritic regions, as shown in fig. 1b, with an overall volume fraction of fv = 1.2 %. a)
β-Phase (Mg17Al12)
b)
Primary solidified Al-poor solid solution Secondary solidified Al-rich solid solution Contraction cavity 50 µm
50 µm
Figure 1: a) Microstructure of AZ91, etched cross section, LM, b) contraction cavities, SEM Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
349 The cyclic deformation tests were performed at both room temperature (20 °C) and 130 °C using a servohydraulic testing system (MTS 810). The specimens had a cylindrical gauge length. All experiments were done in dried air under total strain controlled conditions and with a constant total strain rate of 2×10-2 s-1 at a total strain-ratio of R(εt) = -1. The total strain amplitudes ∆εt/2 (∆εt: total strain range) varied between 1.4×10-3 to 2×10-2. Engineering strains and loads were recorded for each test using a computer data acquisition system. The fatigue failure criterion was defined as a 20 % drop in maximum tensile load or specimen fracture. Common metallographical methods and equipment were used for the microstructural characterization of the alloy, i. e. standard preparation techniques and light and scanning electron microscopy (LM, SEM). Due to the reaction between water and Mg to magnesium hydroxide, contact of water with the magnesium alloy was avoided in all cases. Therefore grinding as well as polishing were carried out by using a mixture of 1 part glycerin and 3 parts ethyl alcohol, which was modified by adding Al2O3-particles (particle size: 1 µ m) in the case of polishing. This preparation technique was also used to minimize the influence of surface roughness on fatigue life.
3
Results and Discussion
Fig. 2 shows the fatigue life data (total strain amplitude ∆εt/2 vs. 2Nf, Nf: number of cycles to failure) of the investigated alloy AZ91 at 20 °C and at the elevated temperature of 130 °C. Additionally, the elastic strain amplitudes ∆εel/2 and the plastic strain amplitudes ∆εpl/2 are plotted for both temperatures. As reported by Eisenmeier et al. [5] the fatigue life data for both temperatures is very well described in terms of both the Basquin and the Manson-Coffin laws. All data for the Basquin and Manson-Coffin laws were taken from the half-life hysteresis loops. The obtained results are in good agreement with Goodenberger and Stephens [6], taking into account that they have used sand-cast AZ91E in the T6 condition. Concerning the cyclic hardening behavior and first results of two step fatigue tests, the reader is referred to refs. [7, 8]. In order to reveal the main damage mechanisms during fatigue detailed metallographic investigations were carried out. Crack initiation predominantly occurs at contraction cavities sited at or beneath the surface. An example for the latter case is shown in fig. 3a. Similar observations of crack initiation for die-casting AZ91 were reported by Mayer et al. [9, 10] and for cast aluminum alloys by Ting and Lawrence [11] and Couper et al. [12]. As reported by Mughrabi [13] based on FEM-calculations performed by Borbély [14], it is evident that crack initiation will take place preferentially at cavities sited at the surface. Nevertheless, subsurface crack initiation is to be considered as a "very probable and plausible" mechanism for crack initiation due to local plastic strain concentrations. Hence, due to the high number of contraction cavities, corresponding to a mean area density of pores of 853 mm-2 and to a volume fraction of 1.2 %, crack initiation at cavities must be considered as a probable event as it is also indicated by SEM investigations. Turning now to crack propagation, see fig. 3b, the cracks run along the interdendritic (especially along the incoherent β-phase) as well as along the transdendritic areas. In addition, there is a strong tendency for a crack to propagate preferentially along those interdendritic areas which show a higher amount of local porosity [8].
350
∆ε/2
∆ε/2
1 ∆εt/2
T = 20 °C
10
-1
10
-2
10
-3
Basquin:
10
-4
∆εel/2 = 0.014 (2N f )
10
-5
∆εel/2 ∆εpl/2
-0.143
M anson-Coffin: ∆εpl/2 = 0.043 (2N f )
-0.465
T = 130 °C
10
-1
10
-2
10
-3
Basquin:
10
-4
∆εel/2 = 0.013 (2N f )
10
-5
10
-6
-0.156
M anson-Coffin: ∆εpl/2 = 0.061 (2N f )
1
10
1
10
2
10
3
10
4
10
5
-0.444
10
6
10
7
2Νf
Figure 2: Total, elastic and plastic strain amplitudes ∆εt/2, ∆εel/2, ∆εpl/2 vs. number of reversals to failure 2Nf , R(εt) = -1, at room temperature (20 °C)and at 130 °C
Figure 3: SEM micrographs of: a) a cross section showing an example of crack initiation at contraction cavities near the surface, b) an etched cross section showing an example for the interdendritic (especially along the βphase) and transdendritic propagation of a crack
Using a surface replica technique and unloading tests the crack propagation behavior could be determined more exactly. Hence, the fatigue tests at 20 °C were interrupted at certain numbers of cycles to produce replicas of the surface taking series of cracks drawn from replicas, as reported in more detail in [5]. It comes out clearly that crack propagation occurs mainly by the coalescence of smaller cracks. Additionally, it is found that the fraction of fatigue life spent in the crack initiation stage increases with increasing fatigue life [8]. Moreover, the performed unloading tests (elastic reductions) within closed cycles reflect the
351 changes of stiffness within one closed cycle and during the fatigue life. Fig. 4 shows an example of elastic reductions within a closed cycle. The sudden change of the stiffness can be explained with the opening and closing of the main crack in the material, whereas the steady change reflects the non-linear elastic behavior characterized by the dependence of the Young’s modulus on the stress [15]. Obviously the cracks open at small tensile stresses and close at small compressive stresses. 42
100
41 50
E D /GPa
σ /MPa
40 0
∆E D
39
-50 38 from compression to tensile stress from tensile to compression stress -100 0.00
0.02
0.04 εt - σ/E 0 /%
0.06
37 -100
-75
-50
-25
0
25
50
75
100
σ /MPa
a) b) Figure 4: a) Elastic reductions within a loading cycle after N/Nf = 80 %, ∆εt/2 = 2.25×10-3 at 20 °C. (εt – σ/E0: plastic strain corresponding to the linear Hooke’s law, where E0 is the Young’s modulus for the undamaged specimen at σ = 0). The unusual shape of the hysteresis loop reflects the non-linear elastic stress-strain response and the stiffness changes due to the damage; b) stiffness values (differential Young´s modulus ED) determined by the elastic reductions shown in a) in dependence of σ during a hysteresis loop.
With the difference ∆ED between the stiffness values of the specimen with opened and closed cracks one can derive a damage parameter ∆ED/E0 (E0: Young’s modulus of the undamaged specimen at σ = 0), which corresponds approximately to the relative loss of loadbearing cross sectional area. Fig. 5 demonstrates that ∆ED/E0 is approximately proportional to the crack length of the main crack (which leads to the failure of the sample) up to approximately 80 % of number of cycles to failure, whereas afterwards ∆ED/E0 deviates to higher values. This behavior can be explained by the fact that the main crack propagates at first mainly along the surface but later also into the bulk of the material. Comparing the results from the unloading tests with the other data of the damage parameter ∆ED/E0, which have been automatically calculated from the stiffness values at the tensile reversal points of the hysteresis loops [5, 13, 16, 17], both values for the damage parameter are in good agreement.
352 14
16 crack length from replicas ∆E D /E 0 by unloading tests
10
∆E D /E 0 from stiffnesses at
12
tensile peak stresses
8
∆εt /2 = 5×10
6
8
-3
T = 20 °C 4
∆E D /E 0 /%
crack length /mm
12
4
2 0
0
400
800 N
1200
0 1600
Figure 5: Comparison of the damage parameter ∆ED/E0, determined both by elastic reductions within closed cycles and from the stiffness values at the tensile unloading points, with the crack length of the main crack determined by replicas
4
Conclusions
The cyclic deformation and fatigue behavior of the vacuum die-casting alloy AZ91HP was investigated under total strain controlled conditions at 20 °C and 130 °C. The fatigue life data are in good correspondence with the laws of Manson-Coffin and Basquin. For crack initiation, the cavities at or beneath the surface play an important role in terms of "crack starters". Crack propagation can occur interdendritically as well as transdendritically and depends on the local microstructure, in particular on the local porosity. Surface replicas, elastic reductions and loss of stiffness determined at the tensile reversal points of the hysteresis loops provide a very detailed picture of evolution of crack length and damage during fatigue. It is concluded that the cracks close at small compressive stresses and open at small tensile stresses. Values of the damage parameter ∆ED/E0, obtained from the losses of stiffness at the tensile reversal points of the hysteresis loops and from the elastic reductions within closed cycles, are in good agreement and explain quite well the crack propagation at the surface measured by the replica technique.
5
Acknowledgements
The authors express their thanks to the Deutsche Forschungsgemeinschaft (DFG) for their financial support and to the AUDI AG, Ingolstadt, for the free supply of the samples.
353
6
References
[1] S. Schumann, F. Friedrich in Proc. of Magnesium Alloys and Their Applications (Eds.: B. L. Mordike, K. U. Kainer), MATINFO, Frankfurt, 1998, pp. 3. [2] J. Polmear, Light Alloys, Arnold, London, 3. Ed., 1995. [3] Müller-Weingarten, Vacuum Diecasting System, brochure. [4] J. Kopp, Doctorate Thesis, University of Erlangen-Nürnberg, 1996. [5] G. Eisenmeier, B. Holzwarth, H. W. Höppel, H. Mughrabi in Proc. of ICSMA-12: 12th Int. Conf. on Strength of Materials, Asilomar, USA, 2000, submitted. [6] D. L. Goodenberger, R. I. Stephens, J. Engng. Mater. Tech. 1993, 115, 391. [7] G. Eisenmeier, M. Ottmüller, H. W. Höppel, H. Mugrabi in Fatigue ’99, Proc. of the 7th Int. Fatigue Congress (Eds.: X. R. Wu, Z. G. Wang), Higher Education Press, Beijing, China, EMAS Ltd. West Midlands, UK, 1999, Vol. 1, pp. 253. [8] G. Eisenmeier, H. Mughrabi, B. Holzwarth, H. W. Höppel, H. Z. Ding in DVMBerichtsband des DFG-Schwerpunktprogramms "Lebensdauervorhersage" (Ed.: Deutscher Verband für Materialforschung und -prüfung), DVM Verlag, Berlin, 2000, pp. 153. [9] H. Mayer, H. Lipowsky, M. Papakyriacou, R. Rösch, A. Stich, B. Zettl, S. StanzlTschegg in Fatigue ’99: Proc. of the 7th Int. Fatigue Congress (Eds.: X. R. Wu, Z. G. Wang), Higher Education Press, Beijing, China, EMAS Ltd. West Midlands, UK, 1999, Vol. 3, pp. 2059. [10] H. R. Mayer, H. Lipowsky, M. Papakyriacou, R. Rösch, A. Stich, B. Zettl, S. StanzlTschegg, Fatigue Fract. Engng. Mater. Struct. 1999, 22, 591. [11] J. C. Ting,. F. V. Lawrence, Fatigue Fract. Engng. Mater. Struct., 1993, 16, 631. [12] M. J. Couper, A. E. Neeson, J. R. Griffiths, Fatigue Fract. Engng. Mater. Struct. 1990, 13, 213. [13] H. Mughrabi in Proc. of 13th European Conf. on Fracture ECF-13, San Sebastian, Spain, Elsevier Science, 2000, in press. [14] Borbély, private communication. [15] Sommer, H.–J. Christ, H. Mughrabi, Acta metall. mater. 1991, 39, 1177. [16] M. Kemnitzer, Diploma Thesis, University of Erlangen-Nürnberg, 2000. [17] H. Biermann, M. Kemnitzer, O. Hartmann in Proc. of ICSMA-12: 12th Int. Conf. on Strength of Materials, Asilomar, USA, 2000, submitted.
Deformation Twinning of AZ31 Alloy in Quasistatitic and Dynamic Compression Tests Lach E., Kainer K. U.*, Bohmann A. und Scharf M. ISL, French-German Research Institute of Saint-Louis, F-68301 SAINT-LOUIS *GKSS, Research Center, Institute of Materials Research, D-21502 GEESTHACHT
1
Introduction
In general, magnesium alloys show limited deformability at ambient temperatures, which is related to their microstructure and crystallographic structure. About the deformation behavior of conventional magnesium alloys under quasistatic condition with low strain rates are many information available. But there is a lack of information on the deformation behavior at high strain rates. In this paper dynamic compression tests (split-Hopkinson-pressure-bar) and subsequent metallographic study were performed to investigate the magnesium alloy AZ31 with respect to deformation twinning. The alloy was extruded at 350 °C and 400 °C, respectively, and subsequent heat-treated at 400°C for 16h. The extrusion and treatment parameter are listed in table 1. The specimen had been compressed dynamically at ambient temperature up to fracture. Some of the dynamic compression tests had been performed with a stopper ring in order to limit the deformation. This enables to study the twin formation at the onset of plastic deformation. A dynamic compression test at –196 °C was performed also. In this way the development of the twinned substructure was investigated by metallography. The mechanical behavior under quasistatic and dynamic compression loading of specimens 7 to 10 (table 1) is reported in reference [1]. Table 1: Change of the grain size in different treated AZ31 alloys ters of the production and treatment Specimen Extrusion temExtrusion ratio Extrusion rate No perature [°C] ln(A/A0) [m/min] 7 400 3.06 4.1 8 400 3.06 4.1 9 350 2.83 6.9 10 350 2.83 6.9
2
depending on the parameCondition as extruded 400°C/16h as extruded 400°C/16h
Grain size [µm] 21.2 25.6 33.9
Experimental Procedure
The tensile and quasi-static compression tests were carried out in an Instron universal testing machine at the strain rates indicated in table 2. The fracture work was determined using an instrumented Charpy instrument. Dynamic compression tests have been performed on a classical split-Hopkinson-pressure-bar [2-3].
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
355 Table 2: Ductility of different treated AZ31 alloys tested under various test conditions depending on the parameters of the production and treatment as shown in table 1 Charpy tests [J] Specimen Grain size Elongation to fracture Compression strain unnotched notched No [µm] in tensile tests [%] [%] dε/dt = 3 x 10-4s-1 7 21.2 14.95 12.8 61.5 12 8 25.6 16.3 16.65 63.5 8 10 33.9 18.3 17.5 68 31.9
3
Results and Discussion
Figure 1 shows the true stress-true strain-curves of the specimen from table 1 tested under dynamic compression condition. There is no different in the maximum stress for the different treated material tested at ambient temperatures. Nevertheless differences in ductility occur. The alloy tested in, as extruded condition, which shows smaller grain sizes (No. 7, table 2), possesses the lowest ductility. With increasing grain size the ductility increases. At low temperatures (No 10, -193°C) the materials shows an hardening effect, which leads to the reduction of the deformability. The reason is the limited deformation by dislocation and twinning. The microstructures in figures 2 – 6 are in relation with the different deformation behavior of the tested materials and confirm the results of the dynamic compression tests. In one example the microstructure of quasi-static tested specimen is presented to show the differences in twinning formation compared with dynamic tested specimen. Figure 2 shows a micrograph of a quasistatic compressed specimen (number 10 in table 1). The compression flow curve is shown in reference [1]. A relatively homogeneously distributed grain size is the result of the heat treatment. 600 MPa
-193 °C
TRUE STRESS
500 400 300 200
strain rate: 2000 - 1000 s -1
100
7
8
9
0 0,00
0,05
0,10
10 0,15
0,20 0,25 0,30 TRUE STRAIN
0,35
0,40
0,45
Figure 1: Dynamic compression tests with thermo-mechanically treated AZ31 performed at room temperature and one at –193 °C. The numbers indicate specimen numbers of table 1 and 2
356
Figure 2: Specimen 10, cast, extruded and heat-treated, quasistatic compression test up to failure, 500x
Figure 3: Specimen 10, cast, extruded and heat-treated, dynamic compression test up to failure, 500x
A high density of twins characterizes the micrograph. They are responsible for the higher compressive strain to failure of a quasistatic compression test (figure 1). Figure 3 shows a micrograph of a specimen of the same specimen number but it was subjected to a dynamic compression test. It shows clearly that the density of twins is lower. The dynamic compression stress is about 100 MPa higher as the quasistatic compression stress, but the strain to failure under quasistatic condition is higher [1]. It can be concluded that twinning is increasing the ductility of Mg-alloys. The alloys in the as-extruded condition have a very inhomogeneously distributed grain size as shown in figure 4. This specimen was dynamically compressed. The deformation was stopped before failure by a stopper ring. It shows that the twin density in the coarse grains is higher. In reference [1] was shown that coarse grained material fracture only after the formation of twins. The competitive mechanism of dislocation and twinning leads to transcrystalline fracture in the coarse grains and leads to crack propagation in fine grains. In figure 5 the micrograph of specimen 10 is shown, which was also dynamically compressed by using a stopper ring. The grain size is homogeneously distributed compared to the specimen in figure 4. A high density of twins characterizes the microstructure.
357 The twin density is higher than in the specimen of figure 3, which was dynamically compressed until fracture. It seems that twinning governs the onset of plastic deformation. Figure 6 shows the micrograph of the specimen, which was dynamically compressed at –196 °C. A lower number of twinning is visible.
Figure 4: Specimen 9, cast and extruded, dynamic compression test with stopper ring, 6 mm to 5.8 mm, 1000x
Figure 5: Specimen 10, cast, extruded and heat treated, dynamic compression test with stopper ring, 6 mm to 5.8 mm, 500x
358
Figure 6: Specimen 10, cast, extruded and heat treaded, dynamic compression test at –196 °C, 500x
4
Conclusions
It was shown that the deformation behavior under dynamic compression condition of the wrought alloy AZ31 is influenced by the microstructure, which is determined by the thermomechanical treatment. Materials with lower grain sizes show lower amount of deformation by twinning. The result is a lower ductility. In further investigation the influence of the texture on the deformation behavior and hence on the twinning formation has to be verified.
5 [1] [2] [3] [4] [5]
References E. Lach, K. U. Kainer, A. Bohmann, M. Scharf, Deformation Behaviour of AZ Alloys at High Strain Rates, ISL-Report PU 319/2000 B. Hopkinson, Proc. R. Soc. A, 74 (1905), p 498 E. Lach, Aufbau und Darstellung einer split-Hopkinson-pressure-bar-Anlage, ISL-Report, RT 509/89 (1989)
Deformation and Fracture Behavior of Magnesium Structural Components Andrea Ockewitz, Christoph Schendera, Dong-Zhi Sun Fraunhofer-Institut für Werkstoffmechanik, Freiburg
Bernd Grosser, Andreas Hamann Volkswagen AG, Fahrzeugforschung, Wolfsburg, Germany
1
Introduction
The door of a model of VW Polo consists of an aluminum exterior part and a magnesium interior part which is a die cast component made from AM 50. The compression tests performed at VW on the magnesium door interior show that damage concentrates on the window frame and sill regions. To analyze the deformation and fracture behavior of these regions in more detail, segments were cut from the magnesium door interior and were tested at VW under quasistatic three-point bending and compression, respectively (Figure 1).
Region for bending tests
Region for compression tests
Figure 1: Magnesium inner part of the VW Polo door
It is well known that damage development depends on loading situations, especially on the stress triaxiality σm/σe which is defined as the ratio of the mean stress σm to the effective stress σe. Since the stress triaxiality in a component under crash loading changes from one position to another, a description of the influence of the stress triaxiality on damage evolution is required to predict the component behavior based on tests of small specimens. In this work two damage concepts i.e. the forming limit diagram and the micromechanical Gurson damage model [1, 2] were applied and compared by testing and modeling different specimens and components. Both concepts take into account the influence of the stress triaxiality on fracture processes. The forming limit diagram is considered as a phenomenological concept and does not account for the deformation history, while the Gurson damage model is based on the
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
360 micromechanical description of fracture processes and describes the influence of the deformation history.
2
Characterization of the Deformation and Damage Behavior
The material properties for the deformation and fracture were characterized by tension tests on round and flat tensile specimens as well as by punch tests. To achieve a variation of the stress triaxiality the flat tension specimens were notched with three notch radii 2, 4 and ∝ mm; the curvature radius of the stamp for the punch tests was varied from 5 to 10 to 20 mm. Marking circles with a diameter of 2 mm were applied to the surfaces of the flat tensile and punch specimens to measure the local strain components. The left side of the forming limit diagram was evaluated from the tension tests and the right side from the punch tests. The damage parameters for the Gurson model were determined by simulating the tension test on a flat specimen with notch radius 4 mm. 2.1
Tension tests and simulations
The narrowest cross-section of all flat specimens was 2x10 mm2 and the diameter of the round tension specimens was 2 mm (Figure 2). The round specimens showed a more pronounced scatter of the fracture strain in comparison with the large flat tension specimens [3]. This scatter of the fracture strain may be caused by imperfections like porosity in the cast material. The influence of the imperfections for the small round specimens is larger than for the large flat specimens. The measured load vs. displacement curves of the smooth and notched flat specimens are plotted in Figure 3. All tension tests were simulated by using the Gurson model in the finite element program PAM-CRASH for shell elements. At first, the damage parameters of the Gurson model were determined by fitting the calculated displacement at fracture of a flat specimen with notch radius ρ=4 mm to the corresponding experimental value . Then, these damage parameters were used to simulate the other notched specimens. Figure 3 shows that the influence of the stress triaxiality on the fracture behavior of the magnesium alloy is well predicted by the damage Gurson model.
Figure 2: Tensile flat specimens with different notch radii (ρ=2, 4, ∞ mm) and round mini-tensile specimen after fracture
361 5
Load [kN]
4
3
2 Tension tests =2, Exp. FE =4, Exp. FE = , Exp. FE
1
0 0
1
2
3
4
5
Elongation l [mm] Figure 3: Tensile flat specimens with different notch radii (ρ=2, 4, ∞ mm) and round mini-tensile specimen after fracture
2.2
Punch tests and simulations
Punch tests were performed on round sheet specimens with a diameter of 55 mm and a thickness of 2 mm. The first appearance of damage in the punch specimens depends on the curvature radius r of the stamp (Figure 4). While at r=10 and 5 mm the damage occurs in the center of punch region, at r=20 mm the crack forms in the bending region around the sharp edge of the die which fixes the punch specimen together with a sheet holder with a load of 10 kN. The damage parameters determined from the tension tests on the specimens with the notch radius ρ=4 mm were used to simulate the different punch tests. The damage positions for the different stamp radii were well predicted by the Gurson damage model ( Figure 5). The calculated load vs. stamp displacement curves agree with the measured curves in a satisfactory way (Figure 6).
Figure 4: Specimens after punch tests for 3 stamp radii
Figure 5: Deformed mesh (one quarter of the punch specimens) with damage distribution
362 10
Experiment FI05T Experiment FI06T Simulation
9 8
Load [kN]
7 6 5 4 3 2 1 0 0
1
2
3
4
5
Stamp displacement [mm] Figure 6: Measured and calculated load vs. stamp displacement curves for stamp radius r=10 mm
2.3
Forming limit diagram
To determine the forming limit diagram the marking circles close to the cracks of all specimens were measured after the tests and two principal strains were evaluated from these data. Figure 7 shows the results from the tensile specimens with the filled symbols and the results from the punch tests with the open symbols. Since not all measured marking circles were loaded to the forming limit due to different distances from the crack, only the maximum values were used to determine the forming limit diagram. The critical curve shown in Figure 7 with the solid line was later used to simulate the component tests.
Figure 7: Forming limit diagram of the cast magnesium alloy AM 50 determined from tension and punch tests
363
3
Component Tests and Simulations
The quasistatic three point bending tests on the segments of the window frames (Figure 8) and compression tests on the segments from the sill region (Figure 11) were performed at VW. To verify the applicability of the Gurson model and the forming limit diagram for crash simulations, the bending tests were simulated with both concepts. Since the damage development calculated in the models depends on the element sizes and generally, much coarser elements have to be used for component simulations than for specimen simulations, the damage parameters and the forming limit diagram represented before can not directly be used for the component simulations. They were calibrated for the element edge length (about 3 mm) in the component models by simulating a punch test with a fine (0.5 mm) and a coarse (3 mm) mesh. Figure 9 and Figure 10 show that the location of damage in the bending component and the corresponding global response were well predicted by the Gurson model. The deviations between the measured and calculated loads may be explained by several simplifications taken in the numerical model, such as not exact wall thickness. The damage position predicted with the forming limit diagram does not agree with the experimental observation as well as that predicted by the Gurson model, it is only a rough correspondence.
Load [kN]
Figure 8: Component test under three point Figure 9: Component simulation with the damage bending model (Gurson) 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Three point bending Experiment Simulation
0
5
10
15
20
25
30
35
40
Stamp displacement [mm] Figure 10: Measured and calculated load vs. displacement curves of the component test under three point bending
The compression tests on the magnesium segments from the sill region of the VW Polo door were simulated with the Gurson damage model as well. The predicted fold form in the
364 mid-rib is not identical to that in the experiment. This may be attributed to the simplified modeling of the variation of the wall thickness by shell elements, e.g. without modeling of ejector pins. The damage positions are well predicted by the Gurson model with the calibrated parameters.
Figure 11: Component test under compression
4
Figure 12: Component simulation with the damage model (Gurson)
Conclusions
The deformation and fracture behavior of the magnesium cast alloy AM 50 was characterized by tension tests with round and flat tension specimens as well as by punch tests. The influence of the stress triaxiality on the damage development was systematically investigated. Both the forming limit diagram and the Gurson damage model were applied to simulate the fracture behavior of the material. To validate the applicability of both damage concepts to crash simulations, component tests under three point bending and compression were performed and simulated with the finite element program PAM-CRASH. The global and local responses of the components predicted by the Gurson model agree with the measurements very well. The important advantages of the Gurson damage model are that it describes the influence of the deformation history and the experimental expenditure for the determination of the damage parameters is much smaller than for the forming limit diagram.
5
References
[1] A.L. Gurson, J. Engng. Mater. Technology, 1977, 99, 2-15. [2] D.-Z. Sun, F. Andrieux, J. Christlein, Progress in Mechanical Behaviour of Materials, (Eds. F. Ellyin and James W. Provan), 1999 Vol. III,1104-1109. [3] Ockewitz, Ch. Schendera, D.-Z. Sun, B. Grosser, A. Hamann, to be published at the Material Week 2000 München.
Internationalization of Magnesium Research Through USCAR in North America and EUCAR in Europe Gerald Cole Ford Motor Company Research Laboratories, Dearborn, MI, USA
1
Extended Abstract
Both European and American automotive manufacturers have committed to reducing fuel consumption and CO2 emissions over the next five years. To achieve these goals, they are performing significant amounts of R&D to develop alternate fuel sources, improve powertrain efficiencies, reduce vehicle drag and decrease vehicle mass. When all the options are examined, mass reduction turns out to be one of the most cost-effective ways to reduce fuel use. And since magnesium has the lowest density of all commercial metals, automotive engineers are increasingly turning to using it in vehicle construction. Lightweight magnesium components also improve driveability, handling and safety, add customer value using the casting process to provide new shapes, and reduce noise by virtue of magnesium's excellent damping properties. In addition, magnesium castings can simplify assembly by replacing aluminium and steel fabrications with large, individual, complex components. It is for these reasons that the magnesium components in North America have grown by over 18% per year from 1.2 kg/vehicle in 1993 to almost 4.0 kg/vehicle in 2000. If this growth rate could continue for the next 20 years, there would be the same 120 kg of magnesium castings as those produced in aluminium. But based on a lack of infrastructure and using the current "business as usual" approach, this will not happen. Compared to the aluminium and plastics industries, which sponsor $ hundreds of millions for research at industrial facilities and universities around the world, the magnesium industry is very small and spends only a small fraction of this amount. It is for this reason that auto companies in North America and Europe have decided that since they need lighter weight vehicles, they will have to help the magnesium industry promote the material more aggressively. They will personally have to manage changing the historic supplier-customer relationship, as demonstrated by the aluminium and plastics industries where the suppliers have performed most of the component-and-process R&D that has led to the significant growth in the automotive usage of their materials. Without a concerted effort by the auto industry, magnesium components will not be developed and we will not obtain all of the potential fuel reductions that mass reduction with magnesium could provide. In order to stimulate growth in the use of magnesium components, the two large precompetitive automotive organizations in NA and Europe, the United States Council for Automotive Research (USCAR) and EUCAR, are sponsoring R&D programs in their respective communities to address issues that are impeding the large-scale application of magnesium in chassis components. They are also beginning to work together to forge an alliance that will lead to a global pre-competitive knowledge base that can be used by the global auto industry. This document describes the programs that the two communities are proposing to address the problems that beset large-scale commercialisation of magnesium. Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
366 Each has developed a chassis program to promote more magnesium use, and each has proposed sharing the data both to reduce redundant R&D and to share pre-competitive knowledge on a global scale.
2
USCAR Program
In 1994, the United States Department of Energy (DOE) and USCAR created the Partnership for a Next Generation of Vehicles (PNGV) with the goal of developing a 5-6 passenger “Supercar” which could attain 80 mpg (3L/100 km). In early 1995 a materials engineering team within USCAR (the U.S. Automotive Materials Partnership, or USAMP) created a cast light metals project that supported the PNGV goals of reducing the mass of structural/chassis components (primarily aluminum castings) by 50%; this was completed in April 2000. Since aluminium was essentially a mature technology it was decided that the next program should focus on increasing the applications for lightweight magnesium. But it was recognized also that there were many open issues that prevented automotive engineers from specifying magnesium components: there was a need for comprehensive design guidelines and a reliable mechanical property database, improved casting processes; improved corrosion resistance, especially at bimetallic junctions; failure (fail-safe) mode analysis and NDE, and improved joining/fastening technologies. The new magnesium project that is being proposed to USCAR will focus on chassis components such as front cradles and rear engine supports that require good ductility, good fatigue and impact resistance, and high to moderate strength,. Project learning’s and simulation models are expected to be applicable to other components in the chassis (such as control arms, steering knuckles, road wheels), body (door and lift gate inner, A/B pillars), and interior (front cross-car beams, seats). 2.1
Deliverables
This will be a 4-year project. The annual cost will be $1.2 M for the DOE, and $1.2M for industry where each of the Big 3 will provide $167,000 of in-kind resources and each of the 30 or so commercial partner will be expensed also in-kind at $25,000. The project will be coordinated by a contractor, who was involved with the initial USAMP project. We will use the management and personnel of the NA Die Casting Association and the International Magnesium Association for industry direction, as well as input from the Society of Automotive Engineers (SAE). towards the end of the project. The program is divided into two phases: 2.1.1 Phase I • •
•
Develop baseline cost models for production-intent front and rear cradles. Develop casting processes and gate/runner designs to produce 6 sigma quality, low cost, high integrity, structural magnesium castings that are cost competitive with aluminium, using for example high pressure die casting, squeeze casting and vacuum low pressure die casting. The objective is to produce as-cast ductile components that have significantly reduced porosity, and improved fatigue properties throughout all casting cross-sections and at all locations in the component. Investigate solutions to bimetallic corrosion of common magnesium alloys associated with fastening magnesium components to aluminum, cast irons and steels.
367 • •
Develop a suitable low cost corrosion-protection system to ensure durability of chassis components against chipping and environmental corrosion. Develop robust, low-cost fastening methodology that reduces the tendency for.magnesium creep.
2.1.2 Phase II • • • • •
•
•
3
Incorporate chassis component properties into the existing magnesium properties database. Link cast magnesium microstructure to fracture, and to monotonic/cyclic mechanical properties. Develop math-based models that simulate key cast microstructure features using boundary conditions of mold temperature, metal temperature, mold fill, distance of melt travel and solidification.. Simulate corrosion performance of various cast microstructures using boundary conditions of composition and process parameters. Continue the partnership with the Country's National Laboratories: Sandia National Laboratory (to predict monotonic and cyclic mechanical properties), Lawrence Livermore National Laboratory (to develop radioscopic inspection standards and quality indicators for thin wall magnesium castings), and Oak Ridge National Laboratory (to develop microstructure-based process models). Support casting process development activities by developing process measurement techniques of mold temperature and pressure, and product characterization methods to analyze casting discontinuities such as quantitative x-ray or other non-destructive evaluation methods. Produce a prototype cradle, which will meet automotive durability, corrosion, strength, crashworthiness and NVH requirements.
EUCAR Chassis Program
The EC along with individual countries such as Germany has had much more experience in joint programs related to improving magnesium technology that has the U.S. Germany has been the leading supporter with such programs as MADICA (Magnesium Die Casting), INMAK (Innovative Magnesium for Wrought Alloy Development), and SFB 390 (a large 15year project covering many R&D subjects). EUCAR has sponsored programs under the heading of MAGDOOR (Magnesium Door Inner), TALMAC (Thixocasting of Al and Mg) and MAGPLAT (Magnesium plating and Corrosion Protection). The European Automobile Manufacturing Association (ACEA) has committed to reduce the CO2-emission for new cars by 25% through the year 2008, compared to 1995. Three new projects have been proposed by EUCAR to address this problem: a joining project (which has been approved), a high temperature alloys project (which has been resubmitted after an initial rejection) and a chassis project, that has followed a similar history. The objective of the Chassis EUCAR project is the subject of this report. Its goal is to develop suitable manufacturing technologies for larger magnesium components in chassis and exterior applications where high strength, high ductility and good corrosion performance are required. EUCAR considers that this should integrate the best world-wide experts, and should be global in its conceptualization. As such
368 the project justifies the new R&D investments that could allow substantial breakthroughs and expand the whole global demand for magnesium technology. The goals of the EUCAR project are very similar to those described above for the USCAR project and are outlined below: • Develop design guidelines for low mass designed magnesium chassis structures for different manufacturing processes and various design/joining methods. Generate a thermo-physical and mechanical database for high strength, ductile magnesium alloys produced with different manufacturing technologies. Measure the durability of optimized magnesium components under load tests as a function of component design, material, and manufacturing process.. • Technically and economically evaluate all magnesium processes (via cost/benefit analyses) for thick magnesium castings, such as road wheels. • Design and manufacture two major components that demonstrate that viability of magnesium components in chassis applications; e.g. wheel, and front cradle, (the latter to be performed in N.A. with USCAR) and a pedal bracket. • Solve the corrosion issues associated with road wheels. The participants agree that corrosion and bimetallic corrosion are the most important phenomena that limit the large-scale application of magnesium components on vehicles. The deliverables of the corrosion research are divided in 6 subtasks: fundamental studies on atmospheric corrosion; bimetallic corrosion; crevice corrosion; corrosion fatigue; surface treatment of magnesium alloys; complete wheel corrosion tests and complete vehicle testing. The corrosion products will studied by XRD, ion chromatography, SEM-EDX and XPS. Appropriate conversion coatings and paints will be examined and analysed using IRspectroscopy and electrochemistry to understand the effects of coating parameters and their influence on adhesion, surface structure and quality. A major goal of the corrosion part of the program is to improve the reliability of accelerated corrosion tests on magnesium alloys, perform fundamental studies on magnesium's corrosion properties to understand magnesium's limitations and be able to outline how to design an accelerated corrosion test. Field exposed test specimens will be used as references in all tests. At the end of the project a complete car test will be run to verify the corrosion performance of the optimized wheel. A finishing system free from chromium will be developed/evaluated along with design against galvanic corrosion as a result of contact with steel and/or aluminum at bushings, wheel bolts, rim to disk and wheel to break disk. Guidelines will be published as to the designs that should be avoided and preferred from a corrosion chemistry point of view. In addition the project will: • Develop improved/new alloys that optimize mechanical properties while improving corrosion resistance, and with higher strength and ductility properties for structural applications. • Optimize casting processes to either optimize current magnesium alloy functions or produce new alloys that have better properties in a road wheel situation. • Select and evaluate fasteners including available protection systems
369 •
Apply information for assembling a 2-part wheel using information obtained from the concurrent EUCAR program on joining • Develop quality assurance measures (x-ray, ultrasound) to validate magnesium wheel quality • Develop finite element procedures for optimizing magnesium component designs A magnesium wheel is being studied because it includes all the features necessary for a safety structural component. And if the wheel could be manufactured at only a small premium over the cost of an aluminum wheel, there could be a very large market for the product. The best compromise to develop a cost effective wheel is to produce it in two parts and then join together the hub (the inner part of the wheel) and the rim (the outer part). The hub will be cast (by high pressure die casting, thixocasting, squeeze casting and/or low pressure die casting); whereas the rim could be produced either from a magnesium thin wall casting or fabricated from aluminum wrought stock via a rolling operation. One industrial wheel manufacturer within the consortium is willing to develop a new technology for manufacturing the complete wheel by a roll forming operation. Casting will be carried out in parallel using numerical process simulation; such as Magmasoft and Simulor software to validate the numerical model. Material characterization will be carried out to evaluate material data to be implemented into the codes for process simulation. By means of both numerical and experimental analysis it will be possible to optimize the process parameters for the different technologies involved. In addition, the possibility to employ hot isostatic pressing after casting will be investigated to avoid possible problems with porosity due to bad filling during casting as well as to increase toughness; it is the intention here to use an innovative liquid hipping process developed by Teksid. Parts for wheel assembly will be analyzed by non-destructive analysis,by microstructural evaluation and by mechanical testing. The project consortium consists of 12 partners from 4 European nationalities, 6 automotive companies, 3 R&D organisations and 2 institutional partners. The selection of the partners and the links to parallel projects the United States underline the systematic approach that is fundamental for the success of the project. The project will be coordinated by a contract research organisation.
Application
High-speed-drilling in AZ91 D without Lubricoolants Franz Tikal Universität Gesamthochschule Kassel
Marcus Schmier (Sp) Universität Gesamthochschule Kassel
Christian Vollmer Volkswagen AG Wolfsburg
1
Introduction
Aim of an industry-connection-project of the IPL at the university Gesamthochschule Kassel was the changeover of the cutting procedures of machining magnesium-alloys on drytreatment and the simultaneous increase of the cutting-parameters. The development of geometrical optimized tools with inner air-supply and particular coatings to the diminution of the adhesion-tendency enables the economic dry-treatment of complex magnesium-workpieces. Specifically when drilling, feeding speeds of 12 m/min and cutting-ways of more than 1 km in dry cutting (without minimal lubrification) could be reached through the development of new tool-geometries.
2
Basic Knowledge
Magnesium and magnesium-alloys are known because of their excellent cuttingcharacteristics. The cutting treatment of magnesium offers crucial advantages, for example low specific cut-strengths and with it inferior performance-demand, short chip-pieces through the producing of shear-chips, inferior tool-wear, high surface quality, high cutting speeds and higher feeding speeds. With the comparison of the specific cutting-forces, it becomes clear that the necessary relative performance-demand for machining magnesium is very low in relation to the treatment of other metals: • magnesium-alloys 1,0 • aluminum-alloys 1,9 • cast iron 4,0 • steel (Rm = 600 N/mm²s) 6,5 • titanium-alloys 7,8 With the election of the concept of lubrification for the magnesium-treatment, different solutions with non water-mixable and water-mixed cool-lubricants become favourite. The advantages of the different concepts are shortly summarized here: removal of chips. keeping clean the machines. decrease of the tool-wear. avoidance of spark- and dust-formation. Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
374 lubrication of ways. better heat removal. Heeding are however also the dangers of the wet cutting. Through high cutting-speeds and enclosed machines, deflagration-danger exists with the oil-treatment. With the machining under application of water-mixed emulsion, hydrogen originates in the workroom and in the chip-conveyor. Hydrogen has a low ignition point, is fleeting however also easily. If a fire occurs with the magnesium-treatment with emulsion, the burning magnesium reacts intensively with water. But not only these security-misgivings but also health, ecological and economic aspects are very important. Another problem appears with the recycling of the chips. The very simply recyclable dry chips - in case of application of lubricants - have to be cleaned first and dried afterwards, frequently that happens through centrifuge. If one wants to briquette the chips to the diminution of the fire-danger between formation and recycling, the expenses grow also here with the share of the lubricant. The consequence of these disadvantages is the total dry-treatment, i.e. the consistent relinquishment on cool-lubricants. The dry-treatment cannot simply be reached by turning off the lubricant-influx however. It is to be taken a multiplicity of factors into account, that mutually influence itself. Particularly the used workpiece-material, the workpiece-geometry, the used tools, the toolmachine and the machining process have to be mentioned here. Only through an optimal interplay of these factors, the functions of the cool-lubricant - transporting and flushing, lubrication, cooling and temper as well as preserve – can be replaced. /1/, /2/, /3/, /4/, /5/, /6/, /7/
-
3
Drilling
The twist drill developed to the dry machining of magnesium distinguishes itself through its versatility. With this tool in the practice common feeding speeds of 1,2 to 3,6 m/min (∅ 6,8 mm, n = 6.000 rpm, f = 0,2 to 0,6 mm/r) can be realized with and without lubricant as well as high cutting-speeds of 15.000 rpm with f = 1,2 mm/r and inner compressed air-influx. 3.1
Drilling with vf = 2,4 m/min
The feeding speed of 2,4 m/mins (6.000 rpm with f = 0,4 mm/r) was first chosen for the attempts to the dry-treatment, because this corresponded to the usual cutting parameters of the wet machining at present. If one uses tools for wet machining for dry cutting without geometrical changes, adhesive wear occurs after a few hundred holes, which leads to tool-destruction as well as missing dimensional stability (fig. 1). In comparison to the standard tool fig. 2 shows one specifically for the dry cutting of magnesium developed drill after more than 60.000 holes. With a drilling depth of 3 x D, one can get a tool-life of 1,2 km that could already be safeguarded statistically in repeating attempts. No one of the further developed tools had reached the end of tool-life after 60.000 holes. There is no adhesive wear at the tool. The exits of the cool-canals and the relief faces are clean. The flute is coated from a thin silver-shiny film, that doesn't hinder the chip transportation however. Also the normally problematic transition of the land to the body clearance shows no adhesions of the material. The left part of fig. 2 clarifies the still sharp lip without
375 outbreaks and without wear at the outer corners. For the control, also the measurements of the strength Fz in tool-longitudinal-axis are used besides the quality of the holes and the toolwear. If the workpiece-temperature in machining is held low by a suitable tool-geometry, the depositions are avoided also in the dry-cut. This can be controlled by in-processmeasurements of the strength (fig. 3).
Figure 1: Standard-tool after 300 holes (dry)
Figure 2: Carbide-drill after 60.000 holes with vf = 2,4 m/min (dry) 600
Max. Strength [N]
500 400
64.414 holes
300 200
average strength flow
100 0 0
10000
20000
30000
40000
Number of Holes [-]
Figure 3: Strength-measurement
50000
60000
70000
376 The strengths in z-direction working with the treatment on the tool settled down after short increase between 400 and 550 N. The average roughness height Rz fluctuated after more than 60.000 holes about the value of 4,6 µm, the diameter-deviation amounted to 11 to 13 µm and is assigned into IT-class 7. Meanwhile, one drill has reached a number of 100.000 holes which corresponds to a tool-life of 2 km. The edges of the tool are still sharp and the depositions increased unimportantly. The drill is still usable. 3.2
High-Speed-Drillings with up to vf = 12 m/min
For more loading-attempts the speed was increased until 12.000 rpm and the feeding speed until 1,0 mm/r. Also here, all attempts were enforced exclusively with inner air-supply at a pressure of 5 bar. For the new requests further changes of the geometry were implemented. The grinding was varied (four lands, normal), the included point angle (140°, 150°, 160°) and the coating (uncoated, CCplusC, TiN). Also here the tool which was developed for the lower speeds was the best. More than 10.000 holes could be produced without problems. The tools showed easy depositions in the area of the chisel-edge and in the flute. On all other surfaces, no abrasion of the gold-shiny TiN-coating was to be determined, also the edges showed no chipping. The high load of 12 m/min feeding speed and over 10.000 holes doesn't reflect itself in the tool-wear. An end of the tool-life could not be determined until now. 3.3
Ultra-High-Speed-Drillings with vf = 18 m/min
With an increase of the feeding speed on vf = 18 m/mins (n = 15.000 rpm, f = 1,2 mm/r) increases the wear at the chisel-edge and in the flute of the tool in comparison to the lower cutting parameters. A tool-life of more than 10.000 holes could however nevertheless be realized with several tools, before adhesive wear at the cool-canal-exits led to the demolition of the attempts (fig. 4).
Figure 4: Tool after 10.000 holes with vf = 18 m/min
To high feeding speeds bring no more time-savings with a depth of 3 x D however, because of the dynamic characteristics of the tool-machine. A sketch of the drilling process (fig. 5) should help to show the real available feeds.
377 Werkstückoberkante Bremsweg 8 mm
Werkzeug
Werkstück
k Bohrtiefe
Z
- 38 mm
Sicherheitsabstand -20 mm
Figure 5: Braking distance with vf = 18 m/min
With a drilling depth of 3 x D, the machine already has a braking distance of 8 mm with a feeding speed of 18 m/min , which is clarified by the recording of the acceleration- and braking-ramp of the Heckert CWK 400 tool machine by using above described parameters. With the forward movement, the machine accelerates on the maximum feeding speed of 200 mm/s (≈ 12 m/min), reaches it after 8 mm and brakes 8 mm before reaching the reversion-point again as well. /8/
4
Compression of Dry Chips
The compression of chips directly after the formation has, especialy at the magnesiumtreatment, several advantages: diminution of the fire-danger. simplification of the handling. low transportation-volume. diminution of additional impurities. avoidance of the declaration as dangerous goods. no suspend of the chips when melting down. Prerequisite to take optimal advantage of this potential is the dry machining including a dry chip transportation. With the wet cutting, moist chips must be centrifuged before the compression; that causes considerable expenses and rescues additional dangers in itself. Furthermore, the water-share after compression still amounts to 1 to 2 %. If moist chips are compressed, the compression machine and their surroundings will become dirty and the briquettes considerably already start to rot internally after a short time. By the transaction of the examinations, a compression machine with a power of 2,2 kilowatts, a weight of 1.000 kg, a maximum pressure of 500 bar and a diameter of the compression cylinder of 60 mm was used. The power of compression of the machine amounted to 120 kg/h. Aim was to find out the machine-parameters so that is reached a high solidity and density of the briquettes. The variable machine-parameters were: pressure of supercharger: 60 to 200 bar compression pressure: 300 to 500 bar maximum filling time: 0,1 to 12 s
378 After optimally adjusted parameters, cylindrical briquettes with a height of 80 mm and a density of 1,4 to 1,5 g/cm³ could be pressed. The transportability (baffle-solidity) was determined with help of dropping attempts from a height of 1,75 m. Fig. 6 shows the results.
100
100
96,88
90
93,65
1st impact
mass [%]
70 60 50 40
2nd impact
80
30 20 10 0 PD = 500 bar
Figure 6: Investigation of the pellet-solidity through dropping-attempts (pressure PD = 500 bar)
The illustration shows the average rest-size of the compressed cylinders in weight-percent after the first and the second drop from 1,75 m. The crumble away run with each impact 3% of the original mass approximately. The pressure of the machine amounted to 500 bar, which with a changed pressure of 2.170 bar corresponds to by a cylinder-diameter of 60 mm. These pellets have a weight about 400 g and a volume of about 10 l of chips is needed to compress one which means a big and dangerous quantity of chips. The further attempts will be made with pellets of a weight of about 50 to 100 g to prevent the dangers which result of big chip quantities. /8/
5
Summary / Outlook
An allround-tool for the diameter of 6,8 mm and a drilling depth of 3 x D could be developed, that is able to reach a tool-life of more than 10.000 holes in an feeding speed area of f = 0,2 to 1,2 mm/r and a cutting speed between 6.000 and 15.000 rpm without remarkable wear. Feeding speeds of 1,2 to 18 m/min are possible. The quality of the holes corresponds to those of the wet machining. The tool-machine-manufacturers are now demanded to the practical transposition of the total dry-treatment of complex magnesium workpieces. Intelligent constructions in the machine to the avoidance of chip conglomeration and to the patronage of the chip transportation, including chip compression equipment are necessary beside more, active and passive safety measures to the realization of the whole concept.
379
6
Literature
[1] BGR 204. Umgang mit Magnesium. Berufsgenossenschaftliche Regel für Sicherheit und Gesundheit bei der Arbeit. Köln 1999: Carl Heymanns. [2] Feinauer, A. u. a.. Trockene Komplettbearbeitung komplexer Aluminium-bauteile. VDIZ. Special Werkzeuge April 1999, S. 24-28. [3] Fleischer, R.. Feinbohren und HSC-Fräsen von Magnesium-Legierungen. Zwischenbericht 1999, Projekt MaDiCa. [4] Schnier, M. u. a.. Magnesium HSC-bearbeiten. Werkstatt und Betrieb 131, 1999, S. 140144. [5] Schwaiger, H.. Druckgießen von Magnesium mit spanender Bearbeitung und Sicherheitsvorkehrungen. VDI-Berichte Nr. 1276, 1996, S. 305-316. [6] Serwe, G.. Magnesiumgussteile im Volkswagen. VDI-Berichte Nr. 58. 1962, S. 71-75 [7] Tönshoff, H. K. u. a.. Trockenbearbeitung von Aluminium- und Magnesiumlegierungen, 1997. IDR 31 Nr. 4, S. 357-364. [8] Vollmer, Chr. u. a.. Trockenbohren in Magnesium. maschinen anlagen verfahren (mav) 7-8/99, S. 44-46
Design, Optimization and Reliability of Magnesium Safety Vehicle Parts Yehuda Tzabari Israel Institute of Metals, Technion, Israel Institute of Technology, Haifa
Israel Reich, Yoel Bahalul Ortal Diecasting LTD. Kibbutz Neve-Ur
1
Abstract
This paper describes the development of three interior safety parts, airbag housing, steering wheel and seat component from preliminary design to commercial mass production. For all the parts the development process includes loading characterization material selection and testing, static and dynamic nonlinear finite element analyses, prototype diecasting production, x-ray examination and static, impact and fatigue acceptance tests for reliable parts. The design optimization is then considered to reduce weight as well as price production. The optimization results are discussed as a function of diecasting parameters, mechanical properties of specimens that were taken from different critical part zones, x-rays and fatigue tests. Development and design recommendations are presented based on the above experience and study. Keywords: Design, Magnesium alloys, Interior safety parts, Optimization, FEM Analysis.
2
Introduction
The applications of Magnesium alloys have been extended in the last decade to various vehicle parts, and a similar growth has been expected for future interior vehicle components [1]. The above is based on improvements in diecasting processes [2], mechanical properties and means of detecting casting defects. The interior safety parts of the vehicle are exposed to random dynamic loads created by the vehicle’s driving conditions and the impact loads generated in the course of accidents. In conjunction, the requirements will include reliable resistance to the above loads, with an adequate energy absorption potential. This paper sums up the development of three interior safety parts which include - airbag housing, steering wheel and seat component, from initial design to commercial mass production. In the development process, load characteristics, material selection, finite element analyses, prototype casting, detection of casting defects and final acceptance tests were made, as specified in previous papers [3], [4]. In the above mentioned development process the following differences were measured between: the calculated and the measured stresses, the theoretical properties of the materials and those actually measured on the cast products, as well as the impact of the casting defects on the fatigue tests. In the aforesaid paper these differences and impacts are discussed. Finally, a cautious process of optimization is dictated, especially in relation to fatigue loads. Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
381
3
Preliminary Design, Material Selection and FEM Analyses
In the first design stages the application of the load characteristics and the mechanical properties of the selected materials enabled an initial calculation and finite element analyses for geometric definition of the parts. The designed interior safety components are as shown in figures 1-3.
Figure 1: Airbag housing
Figure 2: Steering Wheel
Figure 3: Seat beam
The load characteristics in these parts are divided as following: 1. Operation loads - Dynamic random loads created by the manner of the vehicle driving which expose the part to fatigue. 2. Impact loads Developed in the course of an accident and dictate energy absorption requirements from the part during plastic deformations. 3. Static loads Express the assembling and removal forces of the part or extreme loads from (1) or (2) evaluated for the design requirements.
Figure 4: Seat beam FE model and stress distribution
Normally, in the design processes, as well as in the aforesaid case, most stress analyses were performed while using static loads which signify defined dynamic states and the dynamic analyses for the completion of the part behavior. In both cases, the loads are not identical to the real loads, and the stress analyses serve as a tool for a relative improvement of the design and characterization of the critical areas as first stage in the process of optimization. Table 1 represents the material selection and some of the mechanical properties as declared by the manufacturer for various parts. Figure 4 demonstrates the model and stress distribution of the seat beam as well as indicates the critical zones of high stresses.
382 Table 1: Material selection and some of the mechanical properties at room temperature Part Material Ultimate tensile Tensile yield Elongation selected strength [Mpa] Strength [Mpa] [%] Air bag housing AM50 210 125 10 Steering wheel AM50 210 125 10 Seat beam AM60 225 130 8
4
Prototype Casting Process Development
Developing the casting processes for the prototypes of the above parts included the diecasting die and the various casting parameters. In the development the MAGNA software was applied to study the pressure effects, flow of material, temperature, material entrance and cooling duct. Table 2 summarizes the casting methods and a number of iterations to obtain casting at a sufficient quality. Table 2: The casting machine, number of cavities and casting improvement iterations Part Casting machine No. of No. of cavities iterations Airbag housing Frech DAW 500 ton – hot chamber 3 8 Steering wheel Frech DAW 500 ton – hot chamber and 1 5 vacuum Seat beam Buhler, SCD-84,840 ton - cold chamber 1 3 For the purpose of defining sufficient quality the cast parts were tested, as following: Micro-structure tests, density measurement, X-ray tests, hardness measurement tests, mechanical properties tests and static loading tests for each part. From our experience, the most effective tests were mechanical properties and part loading tests. For density, hardness and micro-structure, no correlation was found with the casting quality, beyond the basic level. However, significant differences were obtained in the mechanical properties in different regions of the part, affected by the casting parameters. Table 3 demonstrates the mechanical properties in the casting process improvement stages of the seat beam - in the top and bottom part of the beam. Usually in x-ray tests extensive use is made for quality control of the part, but one should remember the limitation of size and orientation of defects that are detectable. In addition, not all the detected defects are critical, For example, in developing the steering wheel, the x-ray tests indicate defects at the ejector pin points of the steering wheel as described in Figure 5. The possibility that these defects will develop especially during fatigue crack growth and cause failure, depends on the nature of loading and the location of these points. In order to assess the effect of the defects, analyses of axial and radial loading were performed, and the stresses in the ejector points area were calculated using finite element method. The results showed that the stresses in their areas are significantly lower than the maximum stresses in the critical points in the steering wheel, which leads to the inference that the failure in the fatigue will also develop in the structural critical areas and not in the ejector points. Figure 6 describes the location, and Table 4 describes the maximum stresses in the various ejector points.
383 Table 3: Test results of mechanical properties taken from locations. Specimen Location unlimited Tensile yield identification tensile strength strength [Mpa] [Mpa] Prototype-1 Top 147 114 S-6R4 Bottom 123 101 Prototype-2 Top 168 109 S-62L1 Bottom 157 105 Prototype-3 Top 193 106 S-16L Bottom 155 97
Figure 5: A typical description of the defects in the ejector pin points
Figure 6: Location of the ejector pin points in the steering wheel
AM60 seat beam different Elastic modulos [Gpa] 44 26 34 34 38 35
Elongation [%] 1.4 0.9 3.3 2.8 6.1 4
384 Table 4: Stress calculation in the area of the ejector pin points described in Figure 6 Max stress [MPa] at various axial wheel loads Point No. 400 [N] 400 [N] 800 [N] 800 [N] 12 o’clock 3 o’clock 12 o’clock 3 o’clock W1 20 30 30 10 W2 10 25 50 50 W3 20 25 20 20 W4 30 30 30 50 W5 50 45 10 20 W6 60 45 30 50 W7 50 25 50 50 W8 20 20 20 20 W9 40 20 50 20 R11 20 30 20 20 R12 20 30 20 20 R13 30 20 20 20 R14 50 50 10 20 R15 50 50 10 20 R21 10 10 20 10 R22 10 20 20 10 R23 20 20 40 40 R24 20 20 10 10 Therefore the x-ray tests, especially the mechanical properties test and the part loading tests constitute an additional important stage in the optimization process of the final prototype.
5
Prototype Static, Impact and Fatigue Tests
The physical tests that complete the acceptance tests of the final prototype development should express the types of the existing and the required loads, whereas the permissible stresses and deformations dictate the limit of the optimization process. In these components, while the airbag housing was mainly tested under the impact load, the steering wheel and the seat beam were required to meet the static forces, the impact loads and various fatigue tests. Figure 7 describes the results of the fatigue tests that were performed on the first and final prototypes of the steering wheel, and demonstrates significant improvements in the results. Table 5 specifies the permissible stresses and deformations, which were used in the process of optimization of the various parts and the results on the prototype weight. In all the discussed cases, the first prototype failed during the tests. The results also demonstrate that the prototypes of the airbag housing and the steering wheel after optimization have a larger weight by about 10% than the initial prototype. The main reason is due to the fact that the initial design adopted a light weight and low stiffness components. However, this slight addition in weight enabled the part to meet the demands and improve the performances, including fatigue.
385 450 400 Steering whe
350
failure curv
First steering
]300 N [, E D U TI250 LP M A200 EC R FO
wheel
The curve
Final steerin
after
wheel
initial crack
150 100 50 0 1.E+03
1.E+04
1.E+05
CYCLES
1.E+06
1.E+0
Figure 7: Fatigue test results for the first and optimized steering wheel
Table 5: Optimized prototypes compared to the first design for the different parts Max. permissible stress [Mpa]/ deformation [%] Optimized Part First measured or calculated prototype prototype weight [kg] Static test Impact test Fatigue weight [kg] Airbag 0.120 Plastic 0.135 housing deformation 2% Steering 0.450 80 MPa Plastic 60 MPa 0.480 wheel deformation 4% Seat beam 0.750 80 MPa Plastic 50 MPa 0.690 deformation 3%
6
Summary
Design and optimization for obtaining a reliable product of Magnesium alloy AM60-AM50 which was used in the framework of developing an airbag housing, a steering wheel and a seat beam, requires a process that includes: Design, selection of materials and finite element analyses. At this stage one should take into consideration that the obtained stresses and deformations are not accurate due to the estimated loads, and the variation in the mechanical properties of the cast part. development of a casting process to arrive at a part of better quality. An iterative process which is assisted in its first stages by micro-structure , density and xray tests, but for further improvement, tests of the mechanical properties of samples taken from the part (if possible) and load tests for the whole part should be performed. Static, dynamic and fatigue load tests.
386 Testing the components by using loads which can express the existing and required loads from safety standards and limitation of permissible stresses and deformations. This process was able to demonstrate reliable airbag housing, steering wheel and seat beam components that met all the requirements. In general, the process can increase the weight of the part, in case the first prototype is under-designed, yet, in conjunction, it will improve the performance of the part accordingly. -
7
References
[1] Friedric H., Schumann S., The Second Age of Magnesium - Research Strategies to Bring the Automotive Industry’s Vision to Reality. Magnesium 2000, Proceedings of the second Israeli International Conference on Magnesium Science & Technology, Dead Sea, Israel 2000, pp 9-18. [2] Fink R., Beck W., Optimization of the Magnesium Die Casting Process. Magnesium 2000. Proceedings of the second Israeli International Conference on Magnesium Science & Technology, Dead Sea, Israel 2000, pp 112-119. [3] Tzabari Y., et al., Design and Modification of Magnesium Airbag Housing. Magnesium 97, Proceedings of the First Israeli international Conference on Magnesium Science & Technology, Dead Sea, Israel 1997, pp. 373-379. [4] Tzabari Y., et al., Light Weight and Low Stiffness Magnesium Steering Wheel Design. Magnesium 2000, Second Israeli international Conference on Magnesium Science & Technology, Dead Sea, Israel 2000, pp. 35-42.
Development and Production of a Die-cast Magnesium Convertible Soft-top Cover Peter Geist, Frank Lehnert BMW Group Munich (Germany)
Uwe Kwasny EDAG AG Munich (Germany)
1
Introduction
Weight reduction has become more and more important in vehicle engineering in the last few years. Vehicles have become heavier rather than lighter due to increased comfort and convenience requirements and safety measures. Ambitious goals such as cars consuming only three litres of fuel per 100 kilometres will force us to design cars in future that have the same weight as today’s models or less. Apart from intelligent engineering solutions, lightweight construction using alternative material concepts will become increasingly important. Thanks to its low density (1.70 –1.80 g/cu. m.), magnesium is particularly important here.
2
The Starting Point
A soft-top compartment cover was to be developed for a new convertible (Figure 1) with the main objectives of lightweight construction and the accumulation of expertise for manufacturing and processing alternative materials in car body engineering with the accompanying preconditions of top-quality design and compliance with cost targets.
Figure 1: New BMW 3 series Convertible
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
388
3
Choice of Concept
During the development of the soft-top compartment cover (Figure 2), several versions using potentially suitable materials (different plastics, aluminium, magnesium etc.) were evaluated during the concept stage in order to achieve optimum weight reduction compared with covers built in the past. The choice of materials was influenced by the requirement profile of the part and the pre-defined cost targets. As a result of these investigations, the following concept was chosen: a single-shell die-cast magnesium cover (material AM50 HP), with laminated exterior surface to achieve a highquality appearance and feel and to avoid mirror effects in the rear window. The surface in the rear area painted in the vehicle’s body colour is an SMC trim with integrated radio aerial (Figure 2).
Figure 2: Folding top compartment cover
4
Project Timetable
The project timetable was as follows: • Project start as a pre-development stage with the evaluation of different concepts • Commissioning the Mössner company after selection of development partners • Final definition of the magnesium cover for series-production development • Introduction on the new 3 Series Convertible.
5
Selection of Development Partners
Suitable development partners were chosen by a benchmarking procedure. It was decided that the cover would be developed jointly with the Mössner die-casting company. It was also decided that the die-casting company would also be responsible as system supplier for the entire die-casting, coating and lamination work. The OZF company would apply the CDC and powder-coated paint finishes. The Parat company would carry out cover lamination in one of the three optional colours in a vacuum lamination plant. All partners involved would collaborate closely in a “Simultaneous Engineering Team” during the development phase.
389
6
Selected Development Priorities
When developing the folding top compartment cover, demanding challenges were established by the defined goal that an exterior car body part with a diagonal dimension of about 1.5 metres was to be manufactured as a die-cast part. Important development elements were intensive FE calculations to reach rigidity targets, cast-specific part design (gating method) for cost optimisation and corrosion investigations concerning the surface lining and joining technique. The Mössner company has developed a new concept for mechanical processing (drilling, milling, grinding). 6.1
Casting technique
Due to the size of the part, extensive investigations were necessary in cooperation with the casting company, in particular with regard to the gating method and mould filling. By gating optimisation, cost optimisation of the cast part could be achieved by reducing shot weight to approx. 9.2 kg for a part weight of approx. 4.6 kg. As experience with other die-cast parts indicates, the ratio of the shot to the part weight is usually between 3:1 and 4:1. Another problem was meeting the tolerance requirements of car body engineering, which made several stages of process optimisation necessary. Among other things, an automated measuring and adjustment process was implemented at the Mössner company in order to meeting the tolerance definitions, for example, in order to compensate for convexity errors on the flap during the manufacturing process. The process chain for a new drilling, milling and grinding processing concept for the unmachined cover was set up jointly with a machine tool manufacturer at the Mössner company. 6.2
Rigidity calculation
In view of the counterforce exerted by the soft-top when closing, a manually operated soft-top compartment cover requires sufficient torsional rigidity. The impression of quality when closing the cover manually also depends on cover rigidity. Extensive use of FE calculations for a single-shell construction enabled the required rigidity goals to be reached (Figure 3). Using the FE calculations an optimised rib pattern (diagonal ribs) from a rigidity point of view was defined with the aid of geometrical framework preconditions and co-ordinated with the casting company for technical implementation. Sufficient bending strength had also to be achieved, to allow for misuse of the cover as a seat. Moreover, the cover was not to represent any risk in the event of a rear-end crash. The calculated results have been confirmed by component testing. 6.3
Corrosion resistance
The basis for a corrosion-resistant magnesium cover construction is the use of AM50 alloy in “high purity” quality. Extensive corrosion tests were carried out in the BMW laboratories to determine the optimal corrosion-resistant surface coating and at the same time to evaluate the known problem of contact corrosion with steel bolts in magnesium. On the basis of these examinations, suitably coated steel bolts with self-tapping threads were screwed into the magnesium cover for attaching the hinges and locks. The cover surface coating process consists of the following steps: pre-treatment/chromatising/CDC coating /powder coating.
390
Figure 3: Calculation
7
Properties Achieved
The weight reduction of the magnesium cover ready for installation is more than 3 kilograms compared with a sheet steel cover of the same rigidity. It was also possible to realise highquality design features using the die-casting technique. The special cover styling with a design edge requires a vacuum lamination process in which the magnesium flap serves as the “upper tool”; this would not have been possible using a steel sheet design. Due to the long lifespan of the die-cast tools, tool costs for this part are less expensive than for the steel solution. The process chain for a die-cast part is simpler than for a conventional multi-section sheet steel cover. Within the framework of function integration, it was also possible to integrate the radio diversity signal aerial into the SMC trim, which is in the vehicle’s body colour.
8
Outlook
During this first application of die-cast magnesium technology in the car body division of the BMW Group, extensive know-how was accumulated throughout the process chain by the close integration of technology, tests and the production plant. The potential for future applications of magnesium can thus be precisely evaluated. From our point of view, the following problems arise for future applications: • A suitable outer-skin quality cannot be achieved on magnesium die-castings at the moment. • A specific crash calculation for magnesium is not yet state-of-the-art. • Further alloy optimisation is required for good corrosion properties allied to satisfactory mechanical characteristics. • Further convergence and coordination of car body engineering and casting processes are required. • Further enhancement of the competence of the tool manufacturers for outer skin elements is necessary.
Magnesium Motorcycle Wheels For Racing Applications Knut J. Schemme, OTTO FUCHS Metallwerke, Meinerzhagen, Germany
1. Introduction Weight saving efforts receive highest priority by the automotive and transportation industry to increase fuel economy and/or pay load. These efforts are much higher if racing applications are considered, especially since 1/100 second can decide about the position on the starting grid. Every kg saved will increase the performance in regard to acceleration and high speed providing an advantage against the competitors. A regular racing motorcycle used for the 2000 world championship of 250cc series weighs approx. 92 - 95 kg. The weight target, i.e. the minimum weight is 162 kg including rider and fuel. With a 105 hp engine a maximum speed of up to 280 km/h can be achieved. Weight saving can be realised most successfully if both design and material related engineering work can be combined. Thus, a variety of materials is used for different parts and components of racing motorcycles depending on certain technical requirements. Besides aluminium (frame, subframe, forks), titanium (screws, small parts) and carbon fibre compounds (soft trim, brake discs, wheels) magnesium alloys are preferred for applications like gear boxes and wheels. The main portion of wheels (approx. 70%) used for the 2000 world championship are made of magnesium sandcast alloy AZ91. Further 20% are carbon fibre compound wheels exclusively fitted on Aprilia motorcycles. Only 10% are forged magnesium wheels, however, the demand is increasing due to high weight saving potential compared to the sand cast version.
2. Wheel Engineering Low weight is the main requirement for wheels to be used for racing applications. To meet this requirement first of all an appropriate material has to be selected providing excellent mechanical properties regarding static strength, fatigue strength and ductility combined with low density. Secondly, the wheel styling has to be optimised basing on the current load which can be done by performing FEA studies to localise the high and low stress areas. Basing on the FEA result the wheel engineering can be completed by determining the final spoke dimension. Magnesium alloys are preferred materials to fulfil all these requirements, especially due to high strength to density ratio. In this context the need to reduce material thickness for weight saving reasons can be best achieved by using wrought magnesium alloys. These alloys processed by forging and flow forming provide superior mechanical properties, which are almost constant over the wheel cross section (see Table I). In contrast, the properties of cast alloys are depending on differing local Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCHVerlag GmbH, Weinheim ISBN: 3-527-30282-4
392 solidification conditions resulting in high strength/ductility in the flange and low strength/ductility in spoke and hub sections. Also, the possible formation of shrinkage cavities has to be considered in dimensioning. Consequently, cast wheels will have thicker spoke cross sections to compensate such casting defects affecting a weight increase. Table L Tensile properties of some cast and wrought magnesium alloys (numbers takenfromofficial specifications) Alloy designation AZ 91-sandcastT6 AZ 80 - forged T6 ZK 30 - forged T6 ZK 60 - forged T6
Composition Mg A19 Znl Mg A18 Zn Mg Zn3 Zr Mg Zn6 Zr
Rmax [MPa]
240 - 300 280 - 320 290-310 300 - 330
Rp0,2 [MPa]
A5 [%]
150 200 200 220 -
2-7 6-10 7-10 7-10
190 230 230 250
3. Alloy Selection The plasticity of metallic materials depends on the number of available slip planes. Due to its hep lattice structure magnesium has only one slip plane - the (0001) basal plane - resulting in a limited formability at ambient temperature. Thus, additional (pyramidal) slip planes have to be activated to obtain the plasticity required for technical processes. This can be achieved with increased temperature (above 220°C). However, with rising temperature the tensile properties decrease and the directionality increases. Due to the hep lattice structure magnesium alloys show a characteristic deformation behaviour, i.e. the deformation process favours an orientation of the inherent slip plane (0001) parallel to the flow direction of the material resulting in anisotropic tensile properties [1]. In fact, the material flow in upsetting direction is primarily lateral to the applied load, so that tensile proof stress in metal flow direction is higher than in upsetting direction. Also of great importance is a fine grained microstructure, because the lower the grain size the higher the achievable deformation speed. Consequently, pressing temperature, grain size and strain rate are very important processing parameters for hot forming of magnesium alloys allowing just slight variations for proper processing [2]. As a result the magnesium forging process is much more complex in comparison to conventional sand- or die-casting. The commonly used alloys for forged parts and components are based on the alloy systems Mg-Al-Zn and Mg-Zn-Zr (see Table I). The latter is the preferable one providing the required fine grained microstructure even in as-cast condition. This is due to its chemical composition in combination with special melting and casting techniques using the semi-continuos direct chill method. Thus, the alloy ZK60 has been chosen for manufacturing of the subject motorcycle wheels.
4. Manufacturing For manufacturing the two following process steps are applied: forging the wheel disc blank and flow-forming the rim. Due to limited plasticity of magnesium alloys hydraulic forging presses are used especially if the billets are pressed in as-cast condition. The lower speed compared to hammer presses favours crystal recovery during the deformation process improving the formability. Further advantage is a greater control over the extent and temperature of deformation, and especially the ability to dwell for a period of time at maximum pressure to improve die filling. The forging equipment used for motorcycle wheel production consists of two hydraulic presses with different pressing powers. Cast billet sections ( 0 290mm) of ZK60 alloy are heated up to an appropriate forging temperature which should be in a range of 250 - 400°C. The process starts with a single forging operation providing the forged wheel blank (press 1). Subsequently follows the deburring operation as the second step (press 2, see Fig. 1).
Figure 1. Heating oven and forging line (schematic) The flow-forming technique is used to form the rim from the flange of the forged wheel disc. Prior to flow-forming the flange has to be heated up to an appropriate temperature (> 225°C). The operation itself runs in three steps: splitting up the flange, flow forming the rim and calibrating the rim contour. The flow forming device consists of mandrel and tailstock plate to be used for locking the wheel disc. Three forming rollers are positioned approximately in 120° orientation: roller 1 splits up the flange, rollers 2 and 3 flow form the rim including calibration of the final rim contour. After flow forming the wheels can be either stabilised at ambient temperature (Tl), artificially aged (T5) or fully heat treated (T6) depending on the required materials properties.
394
Figure 2. Flow forming operation (schematic) The flow formed wheel raw parts are machined by turning the face sides and milling the spokes. The subsequent steps are turning of the rim and drilling of the valve hole. Finally, the wheels have to be inspected, whereas coating and painting represent the final step of the manufacturing process. The wheel sizes produced vary between 3,5" x 17" (front wheel) and 6,25" x 17" (rear wheel).
Figure 3. Forged racing motorcycle wheel made of magnesium alloy ZK60. Wheel weight: 2,25 kg (Courtesy: PVM wheels and brakes, Mannheim, Germany)
395
5. Conclusion A new magnesium racing motorcycle wheel could be successfully launched by using the forging and flow-forming technique in combination with the Mg-alloy ZK60 (see Fig. 3). Due to superior mechanical properties of this wrought alloy a weight saving of almost 1 kg could be achieved in comparison to the magnesium sand-cast version weighing approx. 3,2 kg (wheel size 3,5x17). In this special application the anisotropy of tensile properties represents a remarkable benefit. The applied loads occurring in service are parallel to the metal flow direction, so that higher tensile proof stress in metal flow direction could be fully used by reducing the spoke cross section resulting in weight saving. Although full milling of the spokes generally is a time consuming process, the excellent machining properties of Mg-alloys (high cutting speed) help to limit the cycle times and cost respectively. On the other hand the dimension of the forged and flowformed raw part provides the chance to realise almost every possible spoke geometry without the need to modify the forging tooling. This is advantageous especially for racing applications since the volumes are very low. The wheels are currently used by the Chesterfield Yamaha Tech3 team for the 2000 world championship of 250cc series (see Fig. 4). The official supplier is PVM wheels and brakes, Mannheim, Germany responsible for engineering, full machining, painting and distribution of the racing wheels. The two drivers of the Chesterfield Yamaha TechS team are presently (June 2000) on 1st (S. Nakano) and 2nd ranking (0. Jacque) of the current 2000 world championship.
Figure 4. Chesterfield Yamaha TechS 250cc racing motorcycle. Rider: Shinya Nakano (Courtesy: PVM wheels and brakes, Mannheim, Germany)
396
6. References [1]
[2]
A. Beck, Magnesium und seine Legierungen Springer Verlag, Berlin 1939 J. Becker, G. Fischer, K. Schemme; Light Weight Construction Using Extruded and Forged Semi-Finished Products Made of Magnesium Alloys In: B.L. Mordike, K.U. Kainer (eds), Magnesium Alloys and Their Applications, Werkstoff-Informationsges. mbH, Frankfurt/Main 1998, 15-28
Acknowledgement The support by Jens P. Bdgel (General Manager PVM wheels and brakes) who gave some background information about motorcycle racing is gratefully acknowledged.
Weight and Cost Saving with Magnesium Die Castings Andreas Mertz Honsel GmbH & Co KG, Meschede
1
Introduction
The application of magnesium in the automotive industry is growing steadily. During the last 5 years, the annual growth rate of Mg-die casting alloy tonnage is progressively beyond 25%. The main advantages of magnesium alloys, low weight and superior casting properties offer a great potential with respect to weight saving, integration and simplification of structural components. The replacement of complex sheet metal components by thin walled Mg-die castings allows more economical production and assembly.
2
The Magnesium Pressure Die Casting Process
In the high pressure die casting process, a steel cavity is filled in an short time with liquid metal, allowing the combination of near-net-shape production, thin walls and high solidification rates. Near-net-shape magnesium components with a weight from several grams up to 25 kg, a maximum projected area of about 1 m² and minimum wall thickness of about 1 – 3 mm can be produced. After metering the liquid alloy into the shot sleeve, the plunger pushes the melt into the cavity in usually 20 ms to 100 ms. The ingate velocity of the melt is of the order of 20 m/s to 100 m/s. To support the feeding process, the melt solidifies under a pressure in the range from 20 MPa to 100 MPa. After solidification, the die is opened and the casting is ejected. Especially when manufacturing structural components, the cavity may be evacuated just before filling starts. Cycle time depends on the size of the casting and the pressure die casting machine and is about 20s to 120s.[1] Because of the high reactivity of Mg-melt, measures to avoid oxidation have to be applied.
3
Cost and Weight Saving
Modern light weight design is determined by adapting the structure to the expected working load. Besides manufacturing cost and the pure mass reduction further aspects concerning adapted manufacturing, assembly-, dismantling- and recycling processes, as defined by the "total cost of ownership" have to be considered. Additionally, energy consumption and pollution, respectively, both for the manufacturing process and during the life time of the product have to be minimized. Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
398 Cost and weight optimized solutions can only be realized by adapting component design to the specific material properties. For example component stiffness can be increased easily by an order of magnitude by applying suitable design methods, whereas increasing Young's modulus from magnesium (E = 45.000 MPa) via aluminum (E= 70.000 MPa) to steel (E = 210.000 MPa) only leads to a comparably moderate improvement. [5] Using the pressure die casting process, stiffeners, e.g. ribs or indentations, can be "mounted" without any additional working cycle. In this connection, virtual product development in combination with appropriate rapid prototyping processes vitally support the design process. The current trend in the automotive industry shows two main application ranges for Mgpressure die castings, the "powertrain" and the "body structure". A typical magnesium application in the "power train" is the gear box shown in Figure 1.
Figure 1: Magnesium gear box
Figure 2: Manufacturing cost Al- / Mg-die casting
399 Here, substituting the common aluminum alloy by magnesium is possible without increasing wall thickness. Appropriate strength and stiffness can be achieved by local stiffeners. Thus, almost the entire weight reduction caused by the density difference of 30% can be exploited. Comparing metal cost of AZ91 magnesium alloy with the commonly used AlSi9Cu3 aluminium alloy, AZ 91 is about 60% more expensive, relating to volume. However, because of more economic processing, and increased tool lifetime, in this example, production cost of the magnesium gear box is "only" about 20% more compared with the aluminum version. An other calculation of a gear box cover (Figure 2) shows an itemized list regarding the individual cost pro rata. It can be seen that the increased metal cost are at least partially compensated by lower processing and tooling expenses. In this specific case, the magnesium version is slightly less expensive compared with the aluminum casting. The other big application range for magnesium die castings are structural parts of the body structure. (Figure 3)
Figure 3: Magnesium structural castings large magnesium structural castings großflächige Mg-Strukturteile empfohlene minimal min. Wanddicke recommended wall thickness
min. wallWanddicke thickness [mm]
3,5 3 2,5 2 1,5 1 0,5 0 0
200
400
600
800
1000
flow Fließweg distance [mm]
Figure 4: Minimum recommended wall thickness for structural Mg-castings
400 The minimum recommended wall thickness mainly depends on the flow distance of the melt. [1]. Compared with steel sheet components much more complex shapes can be produced in one operation cycle. Thus, the "rear-bulk head" made of AM60 can be produced in quite a short manufacturing chain ready to be installed. (Figure 5) pressure die casting trimming sampled X-ray inspection finishing final inspection Figure 5: Manufacturing process of a structural Mg-casting
To calculate the total cost of a given number of steel or magnesium components, both investment and production have to be considered. (Figure 6). Tooling cost for deep-drawing and joining processes are much higher compared to pressure die casting. However, especially, when producing high volume products, this fact is compensated by reduced material cost and higher productivity. Usually, the break even point is of the order of 600.000 to 1.000.000 units. In this context it should be mentioned, that automotive industry usually pays about 1,5 to 2 € per saved kg of weight. In individual cases, one kg of weight saving is worth up to 5 to 8 €. [6] T otal P roduction Cost Car Body Com ponent 100000 90000 80000 tota l cost [k ]
€
70000 60000
c asting
50000
s teel s heet
40000 30000 20000 10000 0 0
200
400
600
800
1000
qua ntity [1000]
Figure 6: Total production cost, casting versus steel sheet
That means, over and above the weight reduction, Mg-die castings can represent an economic alternative to steel design, especially, in small and medium volume production.
401
5
Summary
With the background of rising fuel costs demand for light weight components will increase steadily. With decreasing metal cost, for the future, magnesium die casting technology has as a good basis to manufacture highly integrated light weight components with superior technological, ecological and economical properties.
6
References
[1] Mertz, D. Brungs, Opportunities to Combine Mg-Extrusions with Die Castings, IMA 1999 Annual World Magnesium Conference, Rome, 1999 [2] Mertz, K. Weiß, R. Vomhof, R. Heller, Near-Net-Shape Magnesium Structural Components, 7. Mg-Abnehmerseminar und Automotive Seminar, Aalen, 1999 [3] M. Siedersleben, Vakuum-Druckguß von Magnesiumlegierungen für hochbelastete Bauteile, Aluminium Nr. 3, 1997 [4] Verbundprojekt MADICA: Sichere Produktionsprozesse für die Magnesiumver- und bearbeitung, Produktion 2000 - 02PV13157 [5] W. Hufenbach, Lösungsstrategien beim Einsatz von Metallen in Leichtbaustrukturen, Leichtbau mit metallischen Werkstoffen, VDI-Gesellschaft Werkstofftechnik, 2000 [6] Leichte Teile dürfen teurer sein, Automobilindustrie 4/2000
Cast Magnesium Alloys For Wide Application P.G. Detkov, I.Yu. Mukhina, A.D. Zhirnov Solicamsk Magnesium Works, All-Russian Institute of Aviation Materials
Nowadays there is a growing interest to the magnesium-based materials, which due to their specific properties are applied in aerospace, automobile instrument-making, chemical industries, etc. A wide range of cast magnesium alloys on the base of the following systems, which have various compositions and properties, has been developed in Russia: • Mg-Al-Zn – ML-5 alloy, alloys for injection casting, • corrosion resistant alloys ML5hp (high purity), ML20-1; • Mg-Zn-Zr – high strength alloys; • Mg- Rare Earth Metal –Zr - heat resistant alloys. Development of new engineering, increase of radius of action, necessity of increase weight saving of products have required development of new light alloys, having more high heat resistant, strength and corrosion resistant properties at operating temperatures from minus 253°C up to plus 350°C with saving of minimal weight of working parts. ML5, ML23 alloys, intended for manufacturing of parts of devices of the radioelectronic equipment, engine parts, cases, pumps, etc., and also passenger seats for planes and automobiles, are used for die casting. ML23 alloy of Mg-Al-Si system surpasses AZ91, AZ88, ML5 alloys in technological properties and low cycle fatigue LCF = 26000 (as against 10500 for ML5). In die casting the alloy has hot crack formation by 2 times less in comparison with that of ML5 alloy (AZ91). The alloy composition does not contain toxic and scarce additions, differs by a fine grain. The use of the alloy allows to lower rejection of casting on cracks and strips and to ensure weight saving of a construction by 30% in comparison with that of Al alloys, • σb20° = 185-250 MPa, σ0,2 = 130-170 MPa, δ = 1,0-4,0% • σb150° =140-170 MPa, δ =2,5-12,5% Alloys of Mg-Al-Zn system: ML5, ML5hp etc. These alloys work at the temperature up to 150°C , at higher temperature they are rather strongly softened. The alloys of Mg-Zn system with Zn, Nd, La, In,Y additions meet new requirements. The presence of Zr in Mg alloys provides significant reduction in size of grain; binding and removing of harmful impurities of Fe, Si, Ni, H from the melt; achievement high homogeneous mechanical properties; high leakless of the castings. VIAM has developed high strength alloys with Zr – ML8, ML12, ML22; heat resistant alloys ML9, ML10, ML11, BML14, BML17. In casting these alloys Zr is introduced from Mg-Zr alloying composition, being developed specially by Solicamsk Magnesium Works in conjunction with VIAM. Manufacture of alloying composition is carried out by Works and is complex process of reduction of Zr from K fluoro- zirconate. Quality and reliability of parts, made from Mg alloys essentially depends on quality of Mg-Zr alloying composition and extent of introduction of Zr from it in alloys. The effect of alloying composition components: Zr, Hf, Fe, Si, O, H and Cl on extent of introduction of Zr is investigated in detail. Is established, that Zr content in alloying composition in quantity from 10% to 20% and Cl content in quantity from 0,5% to 2,5% does not effect practically on extent of introduction of Zr in alloys. Decrease of Cl content to 0,5-1,0% results in decrease of flux inclusions and both Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
403 increase of purity of alloying composition and alloys. Hf in quantity 0,05-2,0% being present in alloying composition as the analogue of Zr does not deteriorate the introduction of Zr. Hf has a refining effect also, grouping harmful impurities, which ingredients can be introduced from dirty raw material or a mixture, and forming refractory compositions with them. In developing various Mg-Zr alloying composition radiographic analysis of various sets of alloying composition of L2 type. The presence of Mg and Zr phases with lattice constant aMg =3,203 kX, cMg = 5,140kX, aα-Zr= 3,220 kX, cα-Zr=5,140 kX, and also ZrO2, ZrNH etc. is shown. In case of use of poorly purified raw material, Zr phase is not pure α-Zr and is a solid solution of introduction of oxygen in α-Zr lattice with larger lattice constant than at α-Zr. In this case Zr phase is not isomorphous to Mg phase and can not be its modifier. High hydrogen content in alloying composition and the melt and also Fe, Si, Al, Mg imparities promotes Zr liquation on the bottom of intermetallic composition and decrease of modifying effect (table 1). Table 1a: Effect of oxygen content in Mg-Zr alloying composition on extent of introduction of Zr in ML12 alloy hemical composition of alloying Oxygen Zr content Accordance composition of L2 grade, [%] content in in ML12 to GOST alloying alloy, [%] composition L2, [%] Zr Hf Cl Si Fe no 1 19,0 0,98 0,60 0,003 0,082 1,56 0,50 2 16,0 0,99 0,92 0,047 0,091 0,97 0,49 3 14,5 1,6 0,56 0,030 0,054 0,61 0,70 yes 4 15,7 traces 0,70 0,013 0,054 0,58 0,70
№ С №
Table 1b: Effect of hydrogen content in Mg-Zr alloying composition on extent of introduction of Zr in ML12 alloy Zr content Accordance hemical composition of alloying Hydrogen in ML12 to GOST composition of L2 grade, [%] content in alloy, [%] alloying composition L2, [cm3/g] Zr Hf Cl Si Fe 1 15,0 0,9 1,60 0,0025 0,062 403,2 0,49 no 2 14,8 1,0 0,70 0,003 0,071 117,6 0,53 3 14,9 1,0 0,60 0,010 0,070 93,0 0,70 yes 4 14,5 0,6 0,60 0,008 0,051 80,0 0,72
№ С №
The carried out all-round researches of Mg-Zr alloying compositions have allowed to develop technological process of manufacture of high –quality alloying composition, having fine-dispersed inclusions of undissolved elementary Zr, which can be partially dissolved in Mg-base solid solution; alloying composition can contain not more than 20% of fine dispersed intermetallic phases. Quality of alloying composition provides chemical composition and mechanical properties of high strength Mg alloys (table 2,3). High strength ML8 and ML22 alloys, being developed
404 on the base of Mg-Zn-Zr-Cd system compete with Al alloys; ML22 alloy surpasses domestic alloys and also ZK61 alloy in strength with preservation of high plasticity. It surpasses widely used ML5 and ML12 alloys in strength by 35%. Table 2: Mechanical properties of Mg-Zr alloys ML8, ML10, ML12, ML15, ML22 Grade of Quantity of Zr, Zr content Typical mechanical Specific strength, alloy, H/T being in alloy, properties [σB/p, km] introduced, [%] [%] σB, MPa δ,% ML8T6 2,0 0,75 270-280 4,0 15 ML10T6 1,3 0,51 230-260 3,0 12,9 ML12T1 1,5 0,75 230-260 5,0 12,9 ML15T1 1,5 0,70 210-230 5,0 11,0 ML22T6 2,0 0,75 310-330 5,0 17 Table 3: Mechanical properties of high strength ML22, ML8, ML12 at 20°C and 150°C Alloy 20°C 150°C σB, MPa σ0,2 MPa δ,% σB, MPa σ100, MPa ML22T6 310-330 220-230 5 210 95 ML8T6 270 190 5 180 85 ML12T1 230 130 5 160 80 ML15T1 220 150 5 145 105 ZK61 250-270 130-170 5 USA High strength ML9, ML10, ML11, ML19, BML14 alloys on the base of Mg-Nd-Zr system with In, Y, mish-metal and etc. (table 4), intended for long-term operation at the temperatures of 250-350°C and short-term operation at 400°C are of great interest. The alloys have good combination of mechanical properties at room and high temperature (table 5), good foundry and corrosion properties. Heat resistant ML19 and BML14 alloys have little tendency to formation of microporosity and provide high leakness of castings. The cast parts of complex configuration , made of these alloys, are characterized by high stability of sizes and are used in aggregate-building industry, successfully competing with known alloys WE43 and WE54. BML17 alloy surpasses WE54 alloy in strength, working temperature of BML17 alloy is higher by 50°C than that of WE54. New alloy BML10 surpasses widely used ML10 alloy in long-term strength by 45%. σ100 250ML10 = 70 MPa, σ100 250BML17 = 130 MPa. The alloys can be produced from virgin metals, however it is more preferable to use ingots and pig alloys, e.g. MA8Z 8Zhp, ZR1 3 etc., previously cast at Solicamsk Magnesium Works. Specific process with adding special salts and oxides into the mixture are developed, which provide producing high purity alloys of the specified composition by liquation refining . The increase of a level of reliability and operation characteristics of Mg alloys is achieved by optimization of existing foundry technology and development of new methods of melt processing, ensuring homogeneous chemical and phase compositions, fine grain structure of
МА
М
Н
405 cast structure, lack of liquation phenomena , increase of purity on metal impurities and nonmetal inclusions. Table 4: Chemical composition of cast magnesium alloys Alloy grade ML8 ML9 ML10 ML11 ML12 ML15 ML19 BML14
BML17
Mg balance -«-«-«-«-«-«-«-
Zn Cd 5,5-6,6 0,2-0,8 0,1-0,7 0,2-0,7 4,0-5,0 4,0-5,0 0,1-0,6 -
Zr Nd In Y La 0,7-1,0 0,4-1,0 1,9-2,6 0,2-0,8 0,4-1,0 2,2-2,8 0,4-1,1 Sum of Rare earth metals 2,5-4,0 0,6-1,1 0,7-1,1 0,6-1,2 0,4-1,1 1,6-2,3 1,4-2,2 In Mg-Nd-Y-Zr system
-«-
Table 5: Properties of Heat resistant Mg alloys Alloy H/T Working Mechanical properties, Mpa temperature, long-run, 0°C 20°C 250°C 300°C σB σ0,2 σB σ0,2 σB σ0,2 ML10T6 250 230 140 170 125 135 110 ML19T6 300 220 120 200 115 150 60 BML14T6 325 270 165 260 130 190 70 1 BML17T6 350 300 200 275 145 200 185 1 WE54T6 300 270 185 210 160 England AL9T5 300 340 230 170 115 120 65
350°C σB σ0,2 110 25 150 35 165
135
-
-
75
35
To make the most best use of alloy properties resources-economy melting and casting process without use of fluxes in protective gas atmosphere was developed. The process allows to increase the quality of casting , improve work conditions, decrease pollution of an environment. Losses of expensive alloying elements –Nd, Y,and Mg- are essentially reduced.
Improving the Characteristics of Magnesium Workpieces by Burnishing Operations H.K. Tönshoff, T. Friemuth, J. Winkler, C. Podolsky University of Hannover, Institute for Production Engineering and Machine Tools
1
Introduction
In the automotive industry a need to reduce fuel consumption is generally accepted. One way discussed is to reduce vehicle weight or - at least - to compensate additional masses caused by increased demands for comfort and safety. Lightweight metallic materials like aluminium and magnesium as well as synthetics are expected to replace steel and glass for many parts in the near future. Due to the hexagonal crystal lattice, magnesium parts are produced by casting processes, especially by die-casting. However, a final machining of functional elements is often required. In turning experiments carried out at the Institute for Production Engineering and Machine Tools the chip formation and surface generation mechanisms that are influenced by adhesive effects between workpiece material, cutting tool material and coating, respectively, were analysed. As machining was performed dry, attention was also paid to the problems related with the high reactivity of magnesium leading to the danger of chip ignition. Especially at small feed rates and small depths of cut , the chip-temperature might exceed the materials melting point of approximately 600° C [1]. Burnishing operations after the cutting process can be used to improve the characteristics of the machined surface. The effect of burnishing operations on surface formation and damage as well as on residual stresses in dependence on the burnishing force is presented. Burnishing velocity and burnishing feed rate showed only little influence on the machining result.
2
Cutting Experiments
Sand cast bars with a diameter d = 150 mm and a length of l = 320 mm were machined. The alloy used was AZ91 HP with approx. 9% Al, 0.7% Zn and 0.2% Mn. All tests including burnishing experiments were carried out on a CNC inclined-bed lathe Gildemeister MD10S with a main power P = 50 kW and a maximum number of revolutions of n = 10,000 min-1. Surface roughness Rz and Ra were measured with a contact stylus instrument Hommel T1000 with a tip radius of 5 µm and a tip angle of 90°. To detect the effect of different cutting tool materials and coatings on the surface formation scanning electron microscopes (SEM) served. Uncoated carbide tools and tools with polycrystalline diamond (DP) tips as well as TiN and DP-coated carbide tools CCMW 120408 and CCMT 120408 were used in a tool holder SCMCN 3225 P12. Resulting angles at the cutting edge (tip radius 0.8 mm) were flank α = 7°, rake γ = 0° and tool angle κ = 50°. The rake for the CCMT-geometry is γ = 5°. Tools with solid DP tips that are soldered to a carbide body were only available with a rake of γ = 0°.
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
407 The influence of different cutting tool materials and coatings, respectively, on adhesive effects is shown in Fig.1. Cutting conditions were cutting speed vc = 900 m/min, depth of cut ap = 1.5 mm, feed rate f = 0.4 mm and a cutting length lc = 750 m. Flank build-up can be observed if uncoated and TiN-coated carbides are used. The variation of the rake does not show a significant influence. If DP-tipped tools are used, adhesive effects can not generally be avoided as workpiece material gets into contact with the carbide body. DP-coated tools show a superior behaviour. No adhesion of magnesium on the flank occurs. However, workpiece material is found on the rake of all tools. Prerequisites for the formation of flank build-up in machining AZ91 HP are • a certain affinity between cutting tool material and workpiece material, • the existence of a sufficient temperature in the tool-workpiece contact and • a soft material component (Mg17Al12-phase on grain boundaries) in which hard particles (Mn-Al reactive products) are embedded [2,3].
Figure 1: Influence of cutting tool materials and coatings on adhesive effects
Adhesive effects between cutting tool material and workpiece material do not only have a negative influence on machining forces, but also lead to an inferior surface quality. Fig.2 shows SEM photographs of machined surfaces after a cutting length lc = 10 m. Grooves caused by the tool feed can be observed. If cemented carbides are used at a cutting speed of vc = 900 m/min additional grooves are caused by the tool material´s grains. At a cutting speed vc = 2100 m/min magnesium particles are torn out of and welded on to the workpiece surface forming flank build-up. Subsequently the microstructure of the flank build-up, in contrast to chips and subsurface of the machined element, shows strong plastic deformation [3]. According to fig.2, photographs of surfaces machined with TiN and DP-coated cutting tools are shown in fig.3. Adhesive effects can be observed when machining with TiN-coated tools even at a cutting speed of vc = 900 m/min. Molten workpiece material can be found at vc = 2100 m/min. DP-coatings are an adequate mean to suppress adhesion, but tracks caused by the pyramidal structure of the coating can be observed within the workpiece surface.
408
Figure 2: Influence of the cutting tool material on the machined surface
Figure 3: Influence of the tool coating on the machined surface
409
3
Burnishing Experiments
An Ecoroll EG14 burnishing tool was used to carry out burnishing tests on the named AZ91 HP bars. All specimens had been prepared by identical turning operations before burnishing. Fig.4 shows the influence of the burnishing force on surface quality, hardness and the residual stresses in the subsurface. Burnishing speed and feed rate had only minor effect on the experimental results. The average roughness Ra can be reduced to approx. 15% of the initial state after turning (Ra = 1.5 µm). However, if the burnishing force is chosen too high (e.g. Fr > 3 kN), the surface is damaged resulting in increasing Ra values. In spite, the gain in surface hardness is most significant for highest burnishing forces (108 HV10 for Fr = 5 kN compared to 68 HV10 after turning). Residual stresses parallel and perpendicular to the burnishing direction for both burnishing forces Fr = 1 kN and Fr = 5 kN are compared in the right section of fig.4. It can be shown that • compressive residual stresses can be induced in the workpiece subsurface by burnishing operations, • the maximum of residual stresses moves towards the workpiece centre for higher burnishing forces, • the maximum value of residual stresses is independent of the burnishing force and • for the burnishing force Fr = 5 kN tensile residual stresses can be detected in the workpiece surface. Tensile stresses in the workpiece surface can cause damages to the workpiece and may decrease its working life [4]. Surface damage caused by high mechanical loads (Fr = 5 kN) are shown in fig.5. Whereas surface and subsurface appear smooth and undamaged for a burnishing force of Fr = 1 kN, intercrystalline and transcrystalline cracks as well as a plastic deterioration can be observed in SEM photograph of the surface and in photographs of the ground section of the subsurface for Fr = 5 kN. However, no grooves caused by the feed of the cutting tool in the previous machining process can be detected (compare to fig.2 and fig.3) for both burnishing forces.
Figure 4: Influence of the burnishing force Fr on surface roughness, hardness and residual stresses
410
Figure 5: Influence of the burnishing force Fr on damage in surface and subsurface
4
Conclusion
To observe the interactions between the workpiece material AZ91 and tool materials and coatings, respectively, turning experiments have been carried out. When machining magnesium dry adhesion between cutting tool and workpiece can lead to flank build-up at cutting speeds of vc = 900 m/min and more if uncoated or TiN-coated carbides are used. Also the danger of chip ignition exists in dry machining if the materials melting point of approx. 600°C is exceeded which is especially significant for small depths of cut und small feed rates. Tools with DP insert or CVD diamond coating can be used to reduce friction and adhesion in the tool-workpiece contact resulting in a superior workpiece surface quality even at high cutting speeds of vc = 2400 m/min. Diamond coatings can also be applied to tools with complex geometries. Furthermore, burnishing operations are a useful mean to improve surface quality, surface hardness and to induce compressive stresses in the subsurface if adequate burnishing conditions have been chosen. For AZ91 HP a burnishing force of Fr = 1 kN gives good results whereas Fr = 5 kN leads to serious damage in the surface and subsurface, respectively.
5
Acknowledgement
The work described in this paper has been undertaken with support of the German Research Council (DFG) in a Special Research Programme on magnesium technology (SFB 390).
411
6
References
[1] H.K. Tönshoff, B. Karpuschewski, J. Winkler, Trockenbearbeitung von Aluminium- und Magnesiumlegierungen, Industrie Diamanten Rundschau, 4/1997, 357-364 [2] N. Tomac, K. TØnnessen, Formation of Flank Build-up in Cutting Magnesium Alloys, Annals of the CIRP 1/1991, 79-82 [3] Winkler, J.: Herstellung rotationssymmetrischer Funktionsflächen aus Magnesiumwerkstoffen durch Drehen und Festwalzen, Dr.-Ing. Diss., Hannover, 2000 [4] Broszeit, E., Steindorf H.: Mechanische Oberflächenbehandlung, DGM Informationsgesellschaft-Verlag, 1996
Machining of Light-metal Matrix Composites Klaus Weinert, Matthias Lange, Michael Schroer Institut für Spanende Fertigung, University of Dortmund, Germany
1
Introduction
The need for light-metal matrix composites as a material for lightweight constructions has arisen, due to the efforts by the automobile industry to achieve lower fuel consumption. Metal matrix composites (MMC) consist of a light-metal matrix and reinforcing particles or fibres (Figures 1 and 2).
Figure 1: Metal matrix composites
Figure 2: Microstructure of metal matrix composites
Reinforced light metals with short fibres or particles allow us to adapt more exactly the workpiece material properties to later requirements. There is an increasing trend in the automobile industry to use these materials for various parts (Figure 3). The material properties of metal matrix composites are:
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
413 •
increased apparent limit of elasticity, stiffness, ultimate tensile strength and fatigue strength, • improved creep stability at higher temperatures, • improved thermal fatigue resistance, • improved material damping, • increased wear resistance, • decreased thermal expansion. But at the same time the machinability of MMC is influenced in a negative way by the hard and abrasive ceramic phases.
Figure 3: Applications of metal matrix composites
2
Mechanisms of Tool Wear
To understand the specific problems encountered, some basic correlations are presented, which could be observed during the machining of MMC [1]. The main problem in machining MMC is high tool wear which under certain circumstances, leads to an uneconomical production process or renders the process impossible. The extensive tool wear is caused by the very hard and abrasive reinforcements (Figure 4). This may be explained by looking at the tribological system (Figure 5). The main reason for wear is the direct contact between the particles or fibres and the cutting edge, which causes both a mechanical and a thermal load at the cutting edge. The dominant wear mechanism is abrasion, which is generated by impacts at the cutting edge and by the sliding motion of the particles relative to the rake and clearance face [2]. Additionally, a thermal load stresses the cutting edge. This thermal load results from hot spots which are generated by microcontacts between the cutting edge and the reinforcements. Despite the relatively low process temperature this thermal load is limited by the melting temperature. Different wear mechanisms are responsible for the abrasive tool wear. These are known as microploughing, microfatigue, microcutting and microcracking [3]. Another important aspect regarding tool wear is the microstructural composition of the cutting material. Commonly used cutting materials like cemented carbide are also composites consisting of carbides as hardphase (WC) and cobalt as binder. Due to this, depending on the
414 ratio of the properties of both the cutting material and the reinforcements (e.g. grain size to particle size), one of the above-mentioned different wear mechanisms predominates.
Figure 4: Hardness of cutting tool materials and reinforcements
Figure 5: Tribological system
3
Tool Wear
In order to demonstrate the way in which different wear mechanisms cause tool wear the following scanning microscope photographs show worn cemented carbide cutting edges used for drilling of fibre and particle reinforced magnesium alloys (Figure 6). In both cases a very regularly formed flank wear is typical [4]. When drilling δ-Al2O3 short fibre reinforced magnesium, the dominant wear mechanisms are microcracking and fatigue. Because of the
415 low hardness of short δ-Al2O3-fibres of approx. 800 HV it is not possible for the short fibre to abrade the cemented carbide by microcutting. The hard phases of the cutting tool are abraded due to microcracking and fatigue at the cutting edge. These particles, which are separated from the substrate, contribute to the amount of wear because they slide over the rake and clearance face. They act like a polishing compound because of their small grain size of approx. 1-3 µm and cause a very smooth topography of the worn surface. Some larger cracks on the cutting edge could also be observed. The tool wear which results from drilling MMC containing harder reinforcements is different. The topography of the worn surface is characterised by grooves. These grooves are mainly oriented parallel to the cutting direction. Since the hardness of cemented carbide is much lower than the hardness of SiC (2400 HV), these particles abrade the cemented carbide cutting tool by microcutting, which is shown by the grooves at the worn clearance face. This wear mechanism leads to extensive tool wear. The only possibility of limiting tool wear is to use harder cutting tool materials, e.g. PCD. But considering the high price of PCD drill tools, alternatives, such as drills with coatings based on TiAlN or diamond, have to be found.
Matrix material : AZ 91 Reinforcement : 20 vol.% δ-Al2 O 3 -short fibres
AZ 91 20 vol.% SiC-particles
Tool: Cemented carbide K10 Cutting speed: v = 75 m/min Feed rate: f = 0.1 mm/rev
Drill hole length: Drill length: Coolant:
l = 20 mm L = 400 mm Dry machining
Figure 6. Worn cutting edges
In order to examine the efficiency of these different cutting materials, drill tests were carried out. The results of these tests are depicted in Figure 7. It shows the development of wear of the coated and uncoated tools as a function of the drilling length when machining magnesium containing different reinforcements. Short δ-Al2O3-fibres cause only a slight deterioration of the tool depending on the hardness of the reinforcements. The cemented carbide tool shows the most wear, because of its low hardness compared to the reinforcements. The best results were obtained when using TiAlN-coated tools. Surprisingly diamond-coated tools are unfavourable. The diamond-coated tool apparently shows more wear if tool wear is predominantly caused by microcracking and fatigue, which occurs when the δ-Al2O3 fibres hit the cutting edge. In this case the TiAlN-coating shows the best results. When drilling the ZC 63 magnesium alloy with 12 vol.% SiC the wear of the uncoated and
416 the TiAlN coated drills increases progressively because of the high hardness of the SiC and causes an early end of the tests. In this case the diamond-coating shows the best results. After a drilling length of Lf = 400 mm no tool wear could be measured. The suitability of diamond coatings for the machining of MMC could also be observed in the drill tests on MMC containing hybrid reinforcements. Therefore it could be summarised that when tool wear can be traced back to microcutting, diamond-coatings are favourable. 0.4
Material:
mm
ZC 63 + 12 vol.% SiC AZ 91 + 20 vol.% Al2 O3
0.3
Cutting speed: v c= 75 m/min Feed rate: f = 0.05 mm/rev Drill hole length: l = 20 mm Cemented Carbide TiAlN-Coating Diamond-Coating
0.2
0.1
Flank wear VB
Flank wear VB
0.4
Material: AZ 91 Reinf.: 5 vol.% δ-Al 2O3 + 15 vol.% SiC mm Cutting speed: v c= 100 m/min Feed rate: f = 0.25 mm/rev Drill hole length: l = 20 mm 0.3 Cemented Carbide TiAlN-Coating Diamond-Coating 0.2
0.1
0 0
100
200
300
mm 400
500
Drilling length L f
0
0
100
200 300 Drilling length L f
mm 400
500
Figure 7: Tool wear when drilling magnesium MMCs
4
Surface Topography and Subsurface Zone
Besides tool wear, another important factor is the surface integrity of MMC. Damage of the reinforcements, caused by the machining process, can lead to a decrease in the properties of the components. The cutting and accompanying deformation mechanism affect the surface quality in a multitude of ways. Examples of this are damage in the form of microcracks, pores, microstructural transformation, plastic deformation, residual stresses and changes in the hardness. In order to analyse the influence of turning, milling and drilling on the subsurface zone of composite materials, SEM and optical microscope analyses of the machined surfaces and of special cross sections were carried out. Figure 8 shows a SEM photograph of the subsurface zone of a magnesium alloy with short fibre reinforcement. A cemented carbide drill was used at a cutting speed of vc = 75 mm/min and a feed rate of f = 0.1 mm. Fibres which lie up to approximately 50 µ m below the surface are fractured. Figure 9 shows SEM photographs of the subsurface zone of a magnesium alloy with particle reinforcement. Cemented carbide and PCD drills were used at a cutting speed of vc = 75 mm/min and a feed rate of f = 0.1 mm. In the case of the subsurface machined with cemented carbide, the SiC-particles which lie up to approximately 40 µ m below the surface are fractured as a result of plastic deformations. On the other hand the SiC-particles of subsurface zone machined with the PCD drill are almost cleanly cut and nearly undamaged because of the very sharp cutting edge of the PCD tool.
417
Figure 8: Surface integrity of a drilled fibre reinforced magnesium alloy
Figure 9: Surface integrity of a drilled particle reinforced magnesium alloy
5
References
[1] Weinert, K.: Biermann, D.; Liedschulte, M.: Machining of Reinforced Aluminium and Magnesium. Proc. ICCM-11, Australia, July 14-18, 1997, Vol. III [2] Weinert, K.: A Consideration of Tool Wear Mechanism when Machining Metal Matrix Composites. Annals of CIRP Vol. 42/1/1993 [3] Biermann, D.: Untersuchungen zum Drehen von Aluminiummatrix-Verbundwerkstoffen. Fortschrittberichte VDI, Reihe 2, Nr.: 338, 1995 [4] Weinert, K.: Liedschulte, M; Schroer, M.: Machining of Magnesium Alloys And Magnesium Matrix Composites. Magnesium 97, Israel, November 10-12, 1997
Nitriding of Pressure Die Casting Dies and Tool Elements Harald R. Schmauser Drei-S-Werk, Schwabach
1
Surface Treatment by Nitriding Reduces Tooling Costs
A part from case-hardening, nitriding has become the most frequently used diffusion process in the heat treatment of ferrous materials. Treatment in salt baths has persisted and has even expanded because convection ensures that very uniform heat transfer conditions are obtained and therefore reaction with the carbon-releasing and nitrogen-releasing substances is rapid and uniform. Furthermore, because salt baths have higher density compared with the gas atmosphere, they generally have higher carbon or nitrogen potential and variation of their chemical composition and therefore of their action is very slow [1]. The special surface treatment of the ejector pins and similar tool elements is advantageously suitable in diecasting dies. The tools undergo a bath nitriding-treatment. In this way, pins obtain a spring hardened core with progressive reinforced zone running to the outside finally having a surface hardness of approx. 950 HV-0,3. Therefore the high degree of abrasion- and wear resistance influences the service life in a very positive way.
2
Salt-bath Nitriding
At the nitriding temperature of 570°C, active nitrogen and carbon diffuse into the surface of the workpiece. After a short treatment time there is a continuous boundary layer of ironnitrogen-carbon compounds on the surface, and this is called the ‘compound layer’. On the other hand, in bath nitriding the production of a compound layer that is as homogenous as possible is important for good nitriding. The thickness of compound layer attainable in a given time depends to a very large extent on the treatment temperature and the material used. The properties of the diffusion zone are also determined to a decisive extent by the material used, its structural constitution and the treatment times.
3
Effects on Properties
The compound layer mainly affects wear resistance, sliding and anti-frictional behaviour, polish ability and corrosion resistance. In addition to the increase of fatigue strength, which is of less interest for tools, the increase of hardness resulting from nitrogen enrichment in the diffusion layer leads to an increase of compressive strength. The harder diffusion layer supports the compound layer, as well as giving improved mould rigidity [2]. Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
419 3.1
Wear resistance
Plain steels are generally bath-nitrided in the annealed state, whereas alloy steels are always bath-nitrided in the tempered or hardened state. After the usual treatment times of 90 to 120min, the compound layer reaches a thickness of 16µ in the case of alloy steels. Wear strength is thus mainly determined by the quality of the compound layer, though the diffusion layer also has wear-inhibiting properties. The thickness and quality of the compound layer are decisive for the service life of the tools. Gas nitrided steel surfaces should not be as good as bath-nitrided surfaces with regard to wear. It was found that these differences result from different composition of the layers produced in the gas stream and in the salt bath. 3.2
Resistance to scuffing
During sliding processes, hardened components exhibit scuffing and welding together under certain conditions of loading. The feared welding with opposing materials in sliding joints does not occur with bath-nitrided components. Furthermore, the compound layer has anti-frictional properties which prevent or greatly postpone scuffing if lubrication fails. 3.3
Corrosion resistance
Being formed of stable nitrides, the compound layer has quite good chemical stability, and so greatly improves the corrosion resistance of structural steel. 3.4
Good tempering properties
On account of the good tempering properties and great hardness at high temperature, the wear resistance and good sliding properties of the surface are also maintained at high temperature. Heat cracking, caused by continual changes of thermal stressing, is greatly reduced by bathnitriding. The nitrided surfaces do not tend to weld to the solid or liquid materials being processed. 3.5
Dimensional stability
The fact that the absolute amount of dimensional changes are actually very slight can be attributed to absence of structural changes at the nitriding temperature of 570°C. 3.6
Rigidity
The considerable increase of boundary hardness also provides much greater rigidity. The resultant increased resistance to bending has a beneficial influence especially for long and thin die parts, like ejector pins and cores. 3.7
Application in pressure die casting
Because of the more stringent requirements of the pressure die casting industry and its customers, notably the automotive industry, relating to quality, complexity and size of the castings produced, interest is now being focused on the casting die and its various accessories.
420 It is a question of finding cost-reducing solutions offering optimum fulfilment of all the requirements. Bath-nitriding has found wide application. The filling chambers, plungers, internal surfaces of the die, cores, gate system, slides and ejector pins are subject to service temperatures of about 250°C; and in the case of cores up to 500°C, under alternating loading. Obviously the parts that come into direct contact with the molten metal, with a temperature up to 700°C, are subject to extreme loading. Chromium-plating has been used in the past in an attempt to increase the service life. The repairs and servicing required after a certain production time are then very time-consuming and costly, and require the use of spark-erosion machines. Bath-nitriding of these tool components gives a cost saving of about 50%, not only because the method of surface treatment is less expensive, but also because subsequent work is cheaper. In the case of chromium plating, the residual chromium layer must be removed and then reapplied after machining. With nitrided surfaces, cost savings are possible by using spark erosion to work directly on the surface. Tools made of hot work die steel owe their considerable improvement in service life to the long-term stability of the compound layer at elevated temperature. Even when the compound layer has worn away after a long time in use, as a result of thermal decomposition or mechanical erosion, the diffusion layer in these materials, with hardness greater than 900 HV, still exhibits notably better wear resistance than in the non-nitrided state. The compound layer, with very high nitrogen content, prevents adherence of the casting material on the plunger and in the sleeve. The same applies to the core and ejector pins and the core-slides. Since the guide tolerances of the moving parts of the die are not great (if possible, not exceeding max. 0.08 mm), pin sticking occurs if casting material adheres to the lateral surfaces. Especially on the die surfaces opposite the gate there is severe stressing and erosion of the die steel by the molten metal, entering at a pressure of several hundred atmospheres. Because of the very nitrogen-rich compound layer and because of the diffusion layer, which is harder and, therefore, more wear-resistant than the core zone, this erosion process is greatly hindered, and this results in a considerable increase of service life. In pressure die casting of aluminium alloys with low silicon content, after a treatment time of just 2 hr, increases of service life by about 200% are normal. For the casting of aluminium alloys with higher silicon content, which exert a stronger wearing action, it is recommended to use a treatment time of about 4 hr for sleeves and die components [3]. Microporosity of the bath-nitrided surfaces ensures good lubrication and antifrictional properties, MoS2, for example on the moving parts such as core-slides and ejectors, which are particularly stressed, but also cores, prevents fretting and makes a decisive contribution to fault-free operation of the pressure die casting machines. With other antifriction coatings, e.g. bonderising, a similar, though less stable, effect is achieved, but this tends to decrease with rising temperature. 3.8
Conclusion
The requirements imposed on pressure die casting, and therefore their cost-effective, qualitative and quantitative production, have risen. Salt-bath nitriding offers a large number of advantageous properties, in conjunction with the use of appropriate steel grades. The service
421 life of the tools and their elements is increased as a result of improvement of wear resistance, heat resistance, hardness, safety against fracture, resistance to scoring, dimensional stability, antifrictional properties corrosion resistance, temper stability and rigidity. The result is a substantial reduction of costs, so that pressure die castings experience a further upturn. A newly developed surface treatment is called ‘Plasma Nitriding’. It is the most modern and most environmentally safe process known today. Still being more costly, it will be applied in a progressive way thus coming down to acceptable costs. Its principle is to diffuse nitrogen into the surface of the workpiece by way of an ionised gas atmosphere. Depending on pressure, temperature process time and relevant gas composition excellent results are achieved. Pressure of 0.3 to 10 mbar, temperature between 400 to 600°C, process times 10 min. up to 36 hours are available to obtain a hardness of 700 - 1.500 HV. This hard layer has a thickness up to 30 µm, under which we have a diffusion zone of about 0.8 mm where the nitrogen is found in the lattice of iron. As the process temperature is 50 - 100°C lower than the bath nitriding process, there is less dispersion and more core stability to endure a higher rate of stress and fatigue relief. After all this process of plasma nitriding will most likely replace step by step the other nitriding processes in the future. Out of the many coating processes to prolong tool life let me just mention TIN coating (vacuum process; process temperature: approx. 450°C). It can be used for high-performance die casting dies but it is rather costly. A fine layer of 3 microns provide a hardness of approx. 2,300 HV. Under no circumstances high-performance dies for Al-Mn-Zn pressure casting should be equipped with ejector pins made of through hardened tool steel, because they will not perform for the required service life. For this application exclusively nitrided ejectors made of hot work steel are to be used. The various kinds of treatments for improving the surface on sleeves or ejectors, mentioned so far, have one thing in common, the wear effect is transferred to the ejector bore. No doubt this rework is much more costly! 3.9
New ejector pin/ejector hole concept
During diecasting process a certain amount of wear on ejector pins and particularly on the matching guide holes is unavoidable. Therefore the suppliers of standard components offer pins with oversize to fit the reground hole. This procedure goes contrary to the principle of economy, because the pin surface has a higher wear resistance due to nitriding. Since the weaker part is on the ejector pin holes, costly rework is necessary to regrind these holes. For reasons of cost it should be considered to break with proven practice and to transfer the wear to the pin. This would mean to reduce the surface hardness respectively in order not to touch the holes but quickly to replace worn-out pins. In order to put the wear effect from the die bore to the ejector pin it was found by test series in various die casting companies that a coating with electroless-nickel Boronnitride gives the best results. This layer of about 20 microns and a hardness of about 40 HRC withstands temperatures of 650°C. It ensures a good sliding even without greasing for a controlled period. Moreover it can be applied even for lower grade tool steels, as the process temperature is only max. 90°C. With a preventive maintenance concept for diecasting moulds, it is recommended to replace ejector pins with new plated pins at fixed periods.
422
4
Conclusion
The foregoing statements of this technical report were intended to ensure die casters that bath and plasma nitrided tool elements are reliable partners for high production level. The new repair concept keeps costs low and saves costly down time in production. Finally let me explain you one more cost saving idea: Core pins are widely used with a draft angle of 2.5°. This conicity requires stamping after casting to ensure a cyl. hole for threading. Tests with a concentricity of only 0.5° which does not require the stamping operation have shown exciting results. With specially designed core pins down time of casting machines was drastically reduced. Summing up all the demonstrated means of upgrading die cast dies there is a huge field of economising die casting operation by cutting down time. The leading manufacture of die elements like ejector pins and core pins Drei-S-Werke, Germany offer all these profitable tool elements from one source. It’s the result of a longlasting partnership between automobile companies and their subcontraction die casters, as well as die manufacturers with the specialists of the Drei-S-Werk for the beneficial future of the die casting industry.
5
References
[1] B. Finner, Development and practical application of the Tenifer process, old and new, ZWF, 1975, 70 ,12, 653-664. [2] F.W. Eysell, Bath nitriding of tools, Werkstatt Betrieb 1965, 5, 273-277. [3] H.R. Schmauser, Bath nitriding for pressure diecasting ejector pins and other tool components, Giesserei-Praxis 1971, 10, 181-189
Corrosion and Surface Treatment
Corrosion Behaviour of the Microstructural Constituents of AZ Alloys Guangling Song, Andrej Atrens, David St. John and Li Zheng CRC for Cast Metals Manufacturing (CAST), Department of Mining, Minerals and Materials Engineering, The University of Queensland, Brisbane, Qld 4072, Australia
1
Introduction
The corrosion performance of AZ91 is determined by its microstructure. It has been shown [1,2] that the α and β phases are the two most important components in AZ alloys. It has been reported that increasing the amount of aluminium leads to improvement of corrosion resistance of cast magnesium alloys [1]. However, the addition of aluminium in AZ alloys results in the formation of more β precipitates along the grain boundaries and a higher aluminium content in the α matrix. Which of these microstructural features is mainly responsible for the improved corrosion resistance is unknown. There are contrary opinions. For example, some researchers [1,2] believe that the β=phase in AZ91 acts as an inert corrosion barrier, and the presence of this phase improves the corrosion resistance. However, other researchers [3,4] have proposed that the β phase increases the corrosion rate though a galvanic effect between the intermetallic β and the magnesium alloy matrix, and therefore the presence of β in an alloy decreases the corrosion resistance. With regard to the role of the α phase in AZ alloys, it was reported [5,6] that the corrosion resistance of α increased with increasing aluminium content. However, increased anodic activity with increasing aluminium content has also been reported [2]. There may be clear explanations for the contradictory phenomena observed. This paper is aimed at further clarifying the roles of the α and β phases in the corrosion of AZ91.
2
Experimental Results
The methods of assessment used in this study are polarisation curve measurement, hydrogen evolution collection and corrosion morphology observation, which have been described previously [7]. Five AZ alloys were used for the experiments in this paper: AZ21, AZ501 and AZ91, AZ91E, and AZ91E(Ca). Their chemical compositions are listed in Table 1. AZ21 and AZ501 are single phase, α=and β respectively, as confirmed by XRD. AZ91, AZ91E and AZ91(Ca) have a microstructure consisting of α grains with β=precipitates at the grain boundaries. In this paper, the AZ21 represents α phase only and AZ501 represents β phase, while AZ91 should be regarded as a simple α+β alloy. Figure 1 presents the hydrogen evolution rates for AZ21, AZ91 and AZ501 alloys at their corrosion potentials in 1N NaCl solution (pH11). It is clear that AZ501 (β) is inert in the
Magnesium Alloys and their Applications. Edited by K. U. Kainer. © WILEY-VCH Verlag GmbH, Weinheim. ISBN: 3-527-30282-4
426 solution. There was no hydrogen evolution at all. AZ91 had the highest hydrogen evolution rate.
hydrogen evolution rate (mol/s/cm2)
Table 1: Chemical compositions of AZ21, AZ501 and AZ91 alloys samples Al Zn Mn Fe Cu [wt%] [wt%] [ppm] [ppm] [ppm] AZ21(α) 1. 33 0. 59 11 11