Thermodynamic Evaluations of the Al-Li-Cu-Mg-Zr Systems
Abstract
The investigation of the AlLiCuMgZr system is accomplished in this report. The evaluations of the CuMgZr and CuLiMg systems and the experimental study of the AlLiCuMg system were carried out during the COST 507 second round. The evaluation of the CuLiMg system was based on the DSC experiments at the large range of composition. The parameters of the excess Gibbs energy obtained agreed quite well with the temperatures of the phase transformations but the fit was not so good using the experimental enthalpies. The decomposition temperature of the phase was fixed at 702.7 and was not defined earlier in the literature. The optimisation of the CuMgZr system was based on the experimental data from the literature and on one set of the DTA experiments measured. The calculated phase diagram was in good agreement with the data used. The existence of the new phase as well as the parameters of formation of the excess Gibbs energy contributed well the system studied. The phase relations near to Alrich corner containing up to 7 wt.% Li, 7 wt.% Mg and 10 wt.% Cu were investigated using microscopy analysis, quality local Xray spectral analysis and Xray phase analysis in the experimental study of the AlLiCuMg system done by Baikov Institute of Metallurgy.
Full text
COST European cooperation in the field of scientific European Commission and technical research COST 507 Definition of thermochemical and thermophysical properties to provide a database for the development of new light alloys Proceedings of the final workshop Vaals, the Netherlands 9 to 12 March 1997 Volume 1 EUR 18171 EN
EUROPEAN COMMISSION Edith CRESSON, Member of the Commission responsible for research, innovation, education, training and youth DG XII/B.1 — RTD actions: Cooperation with non-member countries and international organisations — European Economic Area, COST, Eureka and international organisations Contact: Mr Peter Lobotka Address: European Commission, rue de la Loi 200 (SDME 1/44), B-1049 Brussels — Tel. (32-2) 29-65512; fax (32-2) 29-65925
COST European cooperation in the field of scientific European Commission and technical research COST 507 Definition of thermochemical and thermophysical properties to provide a database for the development of new light alloys Proceedings of the final workshop Vaals, the Netherlands 9 to 12 March 1997 Volume 1 COST Secretariat, Brussels, June 1998 1998 EUR 18171 EN
LEGAL NOTICE Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use which might be made of the following information. A great deal of additional information on the European Union is available on the Internet. It can be accessed through the Europa server (http://europa.eu.int). Cataloguing data can be found at the end of this publication. Luxembourg: Office for Official Publications of the European Communities, 1998 Volume 1: ISBN 92-828-3901-X Volumes 1 to 3: ISBN 92-828-3900-1 © European Communities, 1998 Reproduction is authorised provided the source is acknowledged. Printed in Belgium
Preface The Final Workshop of the COST 507 Action was planned to enable the participating scientists in each project to report, either in an oral or poster presentation, on the work carried out during the 3 years of Round 2. At the same time it provided an opportunity for discussion and coordination of the final content and form for presentation of the experimental results and thermodynamic evaluations from the different participating laboratories. These Proceedings of the Workshop constitute one of 2 volumes (the other presenting evaluated data), which together summarise the results and information assembled during the entire COST 507 Action. The compiled data are presented in tabulated and graphical form such that they can be used for calculation or retrieval of information of direct relevance to alloy design and development. The Workshop was well attended by some 50 participants from all but one of the 14 signatory countries to the Action. The organisation of both the oral and poster presentations according to the 'Key System' structure used to coordinate the work of Round 2 allowed logical and self-consistent discussion of the problems associated with specific alloy categories and of the way these had been dealt with by the ' Key System' partners. A coordination plan for the COST 507 Action is presented below. As Key Speakers, Dr.Christophe Sigli (Péchiney), Mr.Colin Small (Rolls-Royce), Prof.Lazar Rokhlin (Baikov Institute) and Prof.Gunter Petzow (MPI für Metallforschung) provided stimulating reviews of the determination of phase constitution for different alloy categories and on the relevance of the results for technological applications and quality of life. Examples of the implementation of results from the COST 507 Action in new alloy development work at Péchiney and Rolls-Royce were presented. The oral and poster presentations produced many lively discussions and the informal surroundings of the Workshop venue encouraged many scientific exchanges omside the meeting rooms themselves. This was a very productive and successful final meeting for all participants in the COST 507 Action and demonstrated very clearly the highly effective interaction which has been established between the project partners in the course of the closely integrated experimental and evaluation work required to produce the final database. The scientific contacts formed in this Action will undoubtedly be of great benefit to future European research projects Last, but by no means least, I should like to acknowledge the considerable efforts made by many members of LTH, in particular Iñaki Hurtado and Tanja Jantzen, in assembling abstracts, organising posters and poster-boards, manning the information desk, compiling these Proceedings and carrying out many other time-consuming jobs associated with the organisation of meetings of this kind. The programme for the Workshop is attached. Philip Spencer (Workshop organiser) Aachen, March 1998 3 -
Programme Sunday March 9 14.00 Registration opens 18.00 Welcoming reception 19.30 Dinner Monday March 10 08.30 Welcoming remarks 08.40 Invited speaker 1 Dr. Christophe Sigli (Péchiney) 09.15 Invited speaker 2 Mr. Colin Small (Rolls Royce) 09.50 Coffee break and posters 10.20 Invited speaker 3 Prof. Lazar Rokhlin (Baikov Inst, of Met.) 11.00 Ternary phase diagram assessments 11.45 Thermophysical properties and database 12.30 Lunch 14.00 Lead System 1 (Al-Fe-Mg-Mn-Si) report 16.00 Coffee break and posters 16.30 Lead System 1 report continued 18.30 End of session 19.30 Dinner Tuesday March 11 08.30 Lead System 2 (Al-Cu-Mg-Si) report 09.45 Lead System 3 (Al-Cu-Mg-Zn) report 10.15 Coffee break and posters 10.45 Lead system 3 continued 11.30 Lead system 4 (Al-Li-Cu-Mg-Zr) report 12.30 Lunch 14.00 Lead systems 5 and 6 (Al-Ti-Me-X) report 16.00 Coffee break and posters 16.30 Lead systems 5 and 6 continued 18.30 End of session 19.30 Conference Dinner Wednesday March 12 09.15 Invited speaker 4 Prof. Günter Petzow (MPI, Stuttgart) 09.45 Database Manager's report Dr. Ibrahim Ansara (LTPCM, Grenoble) 10.15 Coffee break and posters 10.45 Summary and selected examples of practical application of the COST 507 database Dr. Tim G.Chart (Chairman COST 507) 11.30 End of session 12.30 Lunch
D9 "Thermophysical properties of light metal alloys" G. Jaroma-Weiland, R. Brandt, P. Reipert and G. Neuer Dll "Thermodynamic assessment in the Al-Cu-Mg-Zn system" P. Liang, H.L. Lukas, H.J. Seifert and F. Aldinger Fl "Heat capacity data on some Al-based alloys" C.Y. Zahra and A.-M. Zahra GR2 "Differential scanning calorimetry and thermodynamic calculations aimed at the development of Al-Mg-Si cast alloys with Cu, Ag and Sm additions" G.N. Haidemenopoulos and A.N. Vasilakos II "Thermodynamic optimization and evaluation of phase equilibria in Rare Earth alloys" R. Ferro, G. Borzone, A. Saccone, G. Cacciamani, S. Delfino, M. Giovannini, D. Maceió and N. Parodi NI "Evaluation of the Al-Mg-Mn-Fe-Si system" P. Kolby, C.J. Simensen and M.E. Seiersten "A Thermochemical Assessment of Data for the Al-rich Corner of the Al-Mn-Si System" you may find this article under UKÌ M.H. Rand, P. Kolby and T.G. Chart S3 "Extrapolations based on Ti-C-N" B. Sundman and L.F.S. Dumitrescu "A Thermochemical Assessment of Data for the Al-rich Corner of the Al-Fe-Mn System, and a Revision of Data for the Al-Mn System" you may find this article under UK1 Å. Jansson and T.G. Chart SF1 "An experimetal investigation of the Cu-Li-Mg phase diagram" L. Rokhlin, M. Hämäläinen and T. Dobatkina "An experimental investigation of the Cu-Mg-Zr phase diagram" N. Bochvar, M. Hämäläinen, L. Rokhlin and E. Lysova "Experimental investigation of the Cu-Mg-Zr system" "Thermodynamic evaluations of the Al-Li-Cu-Mg-Zr systems" M. Hämäläinen, H. Braga, N. Bochvar, T. Dobatkina, E. Lysova and L.L. Rokhlin - 12
UKI "Smith thermal analysis Studies of Al-rich regions of the Al-Mn, Al-Mn-Fe and Al-Mn-Si systems" J.A.J. Robinson, F.H. Hayes, A. Serneels, F. Weitzer and P. Rogl "A thermodynamic assessment of the Ti-Al-V system" F.H. Hayes and J.A.J. Robinson "A Thermochemical Assessment of Data for the Al-rich Corner of the Al-Mn-Si System" M.H. Rand, P. Kolby and T.G. Chart "A Thermochemical Assessment of Data for the Al-rich Corner of the Al-Fe-Mn System, and a Revision of Data for the Al-Mn System" A. Jansson and T.G. Chart 13 -
A Summary of the COST 507 Action and Examples of Practical Applications of the Database Tim G Chart Chairman, COST 507 Management Committee Chart Associates, Ashford, Middlesex, UK Abstract At the Vaals Workshop this presentation was made primarily using a series of colour OHP's summarising work progress during the course of the Action on a fairly personal basis, as a tribute to the scientists (mainly present at the meeting) who have contributed to the work over the years. A written summary of this is not possible. Here, a brief history of COST 507 is presented, together with a summary of the results achieved. For those not wishing to read the details, please refer to the Concluding Remarks - "the proof of the pudding is in the eating". Origins of COST 507 It is more than appropriate that this Final Workshop for the COST 507 project should take place in Vaals, a few kilometres from the Rheinisch Westfälische Technische Hochschule Aachen. The Action began, in its infancy, as a BRITE-EURAM Project Proposal, submitted by Philip Spencer, RWTH, on behalf of SGTE (Scientific Group Thermodata Europe) in 1986, Philip at the time living in Vaals. This proposal, at the suggestion of the COST Secretariat, led to COST 507 Round I, the original Memorandum of Understanding signed 8th December 1988 on behalf of Germany, Greece, France, Norway and United Kingdom. The project initially involved 14 signatory countries: A, B, CH, D, E, F, GR, I, N, NL, P, S, SF, and UK, indeed, most of the signatory countries at the time, and the Baikov Institute of Metallurgy, Moscow, joined during Round I. The project was given the formal, rather cumbersome, title "Measurement and Evaluation of Thermochemical and Thermophysical Properties to Provide a Database for the Development of New Light Alloys", normally abbreviated to "A Database for Light Alloys". Summary of Objectives: The principle objective of this project has been to provide a computerised thermodynamic database to permit the calculation of multicomponent phase equilibria for light alloys based on aluminium, magnesium and titanium, to aid materials scientists and engineers in the development and successful utilisation of a wide range of commercial light alloys. Experimental studies, required to provide missing information, critical assessment of ternary alloy phase diagrams and the acquisition of thermophysical properties have been included. 15 -
The database has been geared to deal with a broad range of applications, from conventional wrought and casting alloys for transport, aerospace and packaging, through to high-tech materials including titanium aluminides for gas turbine blades, aluminium-lithium based alloys, metal-matrix composites and high-strength low-density alloys. Environmental applications including recycling have been involved. The project as a whole has been targeted to provide the requirements of European industry and commerce in this field, based on discussions with representatives from industry. Organisation, Management The organisation and management of the project, commencing at RWTH Aachen, has centred around the Max-Planck-Institut für Metallforschung, Stuttgart, with Prof Dr multi Günter Petzow (known by his friends as "Petz") as the original Chairman of the 507 Management Committee, together with the National Physical Laboratory, Teddington and Günter Effenberg, Materials Services International GmbH, with immense help and support from the COST Secretariat in Brussels. Philip Spencer, RWTH, has always been a key player, as originator of the Action and Coordinator Group C (see below), and now, the Final Workshop, after some 10 years, returns in 1997 to Aachen: COST 507 1986-1997 RWTH Aachen 1986 MPI Stuttgart NPL Teddington MSI Stuttgart COST Secretariat Brussels RWTH Aachen 1997 (Vaals) Round I led to a computerised working database for 63 binary systems, provided by Himo Ansara, LTPCM Grenoble, our Database Manager [94Ans], together with constitutional data for some 35 ternary systems. In order to deal with industrial requirements within existing resource limitations, six "leading systems" were defined before the commencement of Round II during the Leuven Workshop 1991 (see Appendix) in collaboration with industry, and the work load prioritised: - 16 -
Aluminium-based systems Titanium-based systems 1 2 3 4 Al-Mg-Mn-Fe-Si Al-Si-Cu-Mg Al-Zn-Cu-Mg Al-Li-Cu-Mg-Zr 5 6 Ti-Al-metal Ti-Al-non metal The prime purpose of Round II was to move to multi-component systems of the real industrial world, as required by European industry and commerce. In addition to the continuation of the four Coordination Groups established during Round I to manage the Action: Group A Experimental phase diagram and thermodynamic investigations Peter Rogl Group B Phase diagram assessments Günter Effenberg Group C Thermodynamic database Philip Spencer Group D Thermophysical properties Greg Haidemenopoulos System Managers "volunteered" to manage, in conjunction with the Coordinators the division and timing of the workload, according to the matrix shown in Table 1. The strategy was to: 1 Compile and critically assess constitutional data prior to calculation; 2 Generate missing data by experiments; 3 Critically assess and optimise thermodynamic data; 4 Generate and compile thermo-physical data. Meetings A considerable number of formally structured meetings, progress meetings, working groups and quite informal get-togethers have taken place over the years. Some of the more important are given in the Appendix. Please note that this Appendix is not meant to suggest that COST 507 has been governed by bureaucracy. Quite the opposite. All meetings, including those of the Management Committee, consisted of people doing real work for the overall project. It is very nice to report that during the entire course of the Action, all Management Committee meeting decisions were unanimous, which for a group involving representatives from some 14 European countries, is an indication of the calibre of those involved, and the measure of success. Of note for future COST Actions, the mechanism of "Working Group" meetings, and "Short Term Scientific Missions", supported by DG XII, whereby relatively large groups of scientists or a few individuals respectively, could meet for short periods to work and exchange ideas, was particularly profitable. - 17 -
Table 1 COST 507 Round II Project Management, December 1996 Classification by Group and System and System Managers System 1A1-Mg-Mn-Fe-Si Nl Ρ Kolby 2 Al-Si-Cu-Mg D3 Ρ J Spencer 3 AI-Zn-Cu-Mg Dll H J Seifert 4A1-Li-Cu-Mg-Zr SF1 M Hämäläinen 5 Ti-Al-Me S3 Β Sundman 6 Ti-Al-NMe UKI Τ G Chart Group A Al, Bl, F3, D2, Nl, UKI A2, Bl, D2, II, UKI A2, D2, II, UKI CHI, D2, PI, RUI A4, Bl, Dl, D4, D8, UKI A4, Bl, D4, UKI Group Β Al, Bl, D12 Bl, D12, II, RUI Bl, D12, Il D12 A4, Bl, D4, D12, RUI A4, Bl, D4, D12 Group C B1,F3, Dll, NI, S3, UK1 D3, DU, II, S3, UK1 B1,D3, Dll, D12, II, S3, UK1 SF1, Bl, D3, D4, D12, F4, S3, SF3, UKI A4, Bl, D4, Dll, S3, SF2, SF3, UKI Group D D9, GR2, S2 D9, Fl, GR2S2 Fl.GRl, S2 D9, GR1 Group A Experimental measurement Group Β Critical assessment Group C Thermodynamic evaluation Group D Thermophysical properties 18
An Example using the COST 507 Database Here an example using data from "Leading System 1 is given. The quinary system AlFe-Mg-Mn-Si is important to the aluminium industry since it encompasses the 1000, 3000 and 5000 series wrought alloys. These cover, eg, major products including lithographic sheet, UBC's (universal beverage cans) body and can end material, and material for automotive applications (body panels, space frames etc). Figs 1-3 show calculated proportions of the phases present as a function of temperature for an "idealised" 5000 series alloy, and demonstrate the power of the existing database. To determine such information by experiment, whilst crucial in specific test cases, would be prohibitively expensive and time-consuming. Such information is not only of relevance to the development and improvement of the above products, but also to their recycling. Thus the database, still under continuing development and improvement, and currently including 78 binary and some 25 ternary datasets (in addition to the ternary assessments from Coordination Group B) is "up and running". Concluding Remarks During the Vaals Workshop quotations in relation to the use of the thermodynamic database were kindly provided on behalf of Alean International Ltd and Rolls-Royce pic. The former is reproduced here, together with an update of the latter: Please refer to Figs 4 and 5. "The COST 507 database for aluminium alloys has been widely used in model alloy work: knowledge of equilibrium conditions is a prerequisite for kinetic analyses of solidification, homogenisation and annealing. The final quinary database will be an invaluable tool for the European aluminium industry, applicable to alloy design and optimisation, process modelling and efforts to improve recyclability through increasing iron and silicon tolerance." Dr Ρ V Evans Alean International Ltd, Banbury Laboratory Please refer to Fig 6. "Titanium aluminides based on the TiAl phase are a relatively new class of lightweight high temperature materials that will find application in the gas turbine in the near future. The 1st generation alloys were based on Ti - 48 at% Al with additions of 2 - 4 at% transition metals. However the alloy chemistry was not optimised for gas turbine applications and the chemistry microstructure relationships were poorly understood. The titanium aluminide data from COST 507 has been used and extended by RollsRoyce to assist in solving and understanding these problems. This has enabled a viable production process for small components to be identified and initial production for testing in advanced engine demonstrators is underway". Rolls-royce pic 19
Thanks Especial thanks are due to the sponsoring organisations, without which COST 507 could not have taken place. It is a pleasure to thank sponsors' representatives, many of whom took a very active part in the work planning and execution, for example, Colin Small of RollsRoyce pic, Paul Evans of Alean International Ltd (Banbury Laboratory) and Christophe Sigli of Pechiney. Thanks are due to the many scientists involved, also those concerned with organisational aspects over the years particularly Prof Petzow and his staff at the MaxPlanckInstitute für Metallforschung and Monsieur Pi than and his staff of DG XII, including especially our various Scientific Secretaries, particularly Martin Kedro and Peter Lobotka. On behalf of everyone I take pleasure in thanking our four Coordinators Peter Rogl, Günter Effenberg, Phil Spencer and Greg Haidemenopoulos, and our Database Manager, Himo Ansara. Also Dr Alan Prince and Prof Björn Uhrenius who accepted invitations to act as external evaluators for the Action. Günter Effenberg needs a special mention, since he not only coordinated Group Β but throughout the years took a major leading role in the overall management of the project. I hope to be forgiven for not mentioning many other names, all having made important contributions. Acknowledgements Figs 4, 5 and 6 are reproduced here with kind permission of Alean International Ltd (Banbury Laboratory) and RollsRoyce pic. Reference 94Ans I Ansara, Thermochemical Database for Light Metal Alloys, COST 507, Concerted Action on Materials Sciences, European Commission, DG XII, Luxembourg, 1995 20
Appendix COST 507: Some Key Meetings RWTH Aachen Irsee RWTH Aachen Brussels Brussels MPI Stuttgart Brussels MPI Stuttgart Brussels NPL Teddington RWTH Aachen Hofburg Vienna Leuven Haldensee Tirol Stockholm Schloß Ringberg Paris Brussels MPI Stuttgart NPL Teddington St Margherita Ligure NPL Teddington Skiathos Brussels RWTH Aachen Schloß Ringberg Schloß Ringberg MPI Stuttgart CA Ashford NPL Teddington Brussels Brussels Vaals Sep '86 May '87 Nov '87 Dec '87 Feb '89 May '89 Sep '89 Oct '89 Sep '90 Dec '90 Jun '91 Nov '91 Dec '91 May '92 Jun '92 Nov '92 Jun '93 Dec '93 Jan '94 Mar '94 Apr '94 May '94 Sep '94 Dec '94 Jan '95 Feb '95 Mar '95 Nov '95 Jul '96 Sep '96 Dec '96 Jan '97 Mar'97 Completion of Brite-Euram Proposal Framework for an Al-Project Initial plans for 507 COST New Projects Group First Management Committee Meeting Coordination Framework Established Management Committee Coordination Meeting Management Committee Coordination Meeting Coordination Meeting COST Ministerial Conference Leuven Workshop Coordination Meeting COST Senior Officials, Round II Seminar and MC Coordination Meeting EUROMAT'93 Management Committee Coordination Meeting Coordination and Modelling Meeting Workshop and Management Committee Coordination Meeting Coordination Meeting Management Committee Coordination Meeting Coordination Meeting Coordination Meeting Management Committee and Working Group Budget and Workshop Planning Working Group Meeting Management Committee and Working Group Working Group (Final Workshop) Final Workshop and Management Committee 21
Plenary Lecture MATERIALS, TECHNOLOGY AND QUALITY OF LIFE G. Petzow Max-Planck-Institut für Metallforschung, Stuttgart (D) ABSTRACT Technological progress is closely linked with the evolution of materials. An assessment of the availability of materials for future developments shows a huge potential of unrealized materials compared with those already available. Today's tailor-made materials aie the consequence of a gradually improving understanding of the architecture of matter. Materials science which explores the structure, properties, preparation and processing of materials is supported in many cases by models and concepts of physics, chemistry and crystallography. The great significance of materials science in technological progress is that it can lead to a basic understanding of internal structure, so that new materials can be developed consequently for specific applications. In this connection computational phase studies and thermochemical data play an important role. The topic chosen reflects an attempt to illustrate the interplay between many materials and technologies. With less resources, less energy and smaller environmental impact, and based on more innovative materials and technologies the today's standard living must be kept constant and made accessible to all people. 29
INTRODUCTION All of us are aware of the rapid changes in our world caused by scientific discoveries and technological developments. Scientists and technologists contribute decisively to our modern society and have to master not only the advantages but also the disadvantages of present and future discoveries and technologies. This is in full accordance with the words of Albert Einstein: "Awareness of the human being and his fate should always be the ultimate concern in specialized scientific endeavors. One should never lose sight of this among al diagrams and equations". Since all of us are more or less concerned with materials in some way, I would like to focus in my presentation especially the role of materials in the socio-economic-context of technology, especially in view of the increasing meaning of the ecological situation. The topic chosen reflects an attempt to illustrate one of the many facets of materials science. HISTORICAL DEVELOPMENT The implementation of technological ideas and concepts relies on the existence of suitable materials which in turn are constantly being improved by the demands of these new technologies. This reciprocal relation has always had a decisive effect on human progress. Between man, materials and technology is a steady interaction ongoing for millions of years. The base of progress is an unalterable interconnected three-way relationship as simplified shown schematically in Fig. 1 [1]. The earth's population has been continuously multiplying since early man first appealed about two million years BC. E.g. the population grew by a factor of 100 in about 100 000 years, from estimated 100 000 to 10 million, and this was reached about 4000 BC. Despite natural catastrophes, plagues and decimating wars, by 1820 the population had grown again by a factor of 100 to 1 billion people, this time in less than 6000 years.
Today there are about 6 billion people and the doubling rate is only about 33 years. Accordingly, it could be possible that in not quite 200 years time there will already be 100 billion (IO11) people on earth. The period for the factor 100 in population increase is less than 400 years. Compared to 6000 and 100 000 years the population increase becomes alarming fast. All of us are aware, those extrapolations often are not reasonable. But no matter whether the steep increase will continue or come into a final state mankind is faced with drastic increase of population during the upcoming decades. Overpopulation will more and more influence our life and our thinking. The conditions for the flourishing of the species Homo Sapiens have improved over time such that man has made the earth his subject. Through his materials and their use in technical devices and processes, man has made favorable living conditions which he modifies continually and to which he in turn continuously adapts. Materials are one of our oldest cultural assets. Historical eras aie named after the materials that dominated at that time: the Stone Age. the Copper and the Bronze Age and the Iron Age possibly the end of which we are living through at this time. New materials such as polymers, semiconductors and superconductors, advanced alloys and ceramics, the amorphous metals, and increasingly the composite materials, are appearing on the scene and providing an impetus for technological developments, often with far-reaching consequences. One can expect the discovery of more materials in the future. Every chemical compound and every alloy is a material that could potentially revolutionize our lives to the extent that the first stone tools revolutionized the life of early man. Materials are turned into tools, devices, machines, houses and streets. Revolutionary technological developments have, as in the case of materials, followed in quick succession in modem times. Thus important developments such as printing, radar, radio, telephone, satellites, rockets etc. that enhance man's favorable living conditions, aie not shown in Fig. 1 for reasons of clarity. In Fig. 1, the rise of the evolutionary curves for materials and technology includes the number of discoveries and technological events. - 31
In the interaction with materials and technology man is the decisive partner, of course. Materials and Technologies are ambivalent. They are, a priori, neither good nor bad. Only man decides about their applications. Considering the importance of materials for the evolution of mankind it is necessary to have an idea on the availability of materials. AVAILABILITY OF MATERIALS Today there is a broad spectrum of materials available, as schematically expressed in Fig. 2. Just counting the materials which have been stored in databanks, over ten million different materials are already in application. This may seem to be a large number, but the number of unrecorded materials is by far greater, as will be shown later. Materials are prerequisite for technology and human life. But even today materials are mostly taken for granted - they are self evident to most people. Without materials men immediately would fall in great trouble and the statement by Georgius Agricola [2], given more than 500 years ago. would become true. He wrote: "... If mankind ceased to use metals, all the possibilities to guard and preserve health, as well as to lead a life corresponding to our cultural values would be taken away. People would lead the most detestable and most miserable life among wild animals..." Everything said about metals can be generalized to all materials. Agricola addressed the social aspects of materials: problems of public health and culture as well as general questions of standards of living. He made the connection between materials technology and society in context of his time. In our days the increasing needs for materials are obvious from Fig. 3 showing the trends in application of materials in the U.S. [3]. The steep increase in materials consumption caused not only by the fast growing of population but also by higher demands of people. 32 -
The expected increase in the consumption according to Fig. 3 would be very good to improve our economies but at the same time with a heavy burden on our anyhow limited resources. And indeed the growing demand of materials imply the "looting" of our planet on nonrenewable sources, which becomes directly evident by Fig. 4 [3]. It has taken a long time for developments to reach their current rate (Fig. 1). But today the question arises, whether the steeply increasing population curve and the pace of technological innovations really present true progress! Our earth, often compared with a spaceship, has limited resources and as a consequence its loading capacity is limited. The "spaceship" earth with limited resources and increasing population is confronted with serious problems. In this universal consideration, the planet earth is a part that may give up energy to its surroundings, but can replace it again from outside, from the sun. The energy losses can be balanced, as shown schematically in Fig. 5 [4]. But it is another matter for material converted by the economy under application of energy. Out of raw products arise commodities and foodstuffs. If the consumption of materials out of the order of the concentrated storage places into scattering by mass production and to refuse dumps is regarded as an increase in the entropy (measure of irreversibility), then this fits with the generalization of the thermodynamic laws. The material entropy in the global ecosystem of the world causes an increase in the disorder, or rather in the uniform intermixture [5, 6]. This economic occurrence, which increases on dramatic by population growth, has a direction. It goes from the concentration of the material towards distribution and scattering. The potential of materials decreases as they are used technically. They go from a certain concentration, that one can also call order, over to an even distribution called disorder of chaos. Entropy is also therefore defined as a measure of disorder or a measure of probability of a state. Disorder is always more probable than order. It should be noted here that Albert Einstein described the entropy law as the most important law of science. - 33 -
There was a first warning more than twenty years ago by a report to the Club of Rome [7], Its authors predicted a catastrophic situation early in the next millennium. In case the conditions of that time would remain, the increase of population would cause a need for more food, more industrial production, more energy and raw materials and as a consequence an increase of pollution, as schematically expressed in Fig. 6. All that would have yield to an unbalanced situation on earth in the next century creating the catastrophe. But fortunately today, about 20 years after the first report to the Club of Rome, a new prediction with updated information has been given [8]. It is realized that the situation has changed to a more positive side: - Population rate is decreased since 1971 from 2.1 % to 1.7 % per annum. The energy and materials consumption became more reasonable; recycling. - Better materials and technologies are available. Therefore, the catastrophe is shifted further to the future and there is even hope for avoidance by achieving a steady state on our planet by further improving of the mentioned three general points of view. This new prediction expresses the first success of ecological renewal, which will more and more influence our life and our thinking. A similar situation to present times has happened in history already twice: around 4.000 BC when early man settled from a nomadic behavior as hunters and began with agriculture. And then around the middle of the 18th century: the industrial revolution. In both cases a change in the habits of living was essential because of the population growth and in both cases materials and technologies have reached a standard which allows such drastic changes in human being. We are now obviously in the beginning of a third renovation: the ecological renewal. And again, there is no alternative to technology. Ecology cannot be realized besides technology and not against technology. But there is only one choice for industry and that is to adapt ecologically [9], New materials and innovative technologies offer a means just for that. There does not seem to exist another solution of our ecological problems than a broadly distributed development of technologies and materials. So far materials concerned, three directions are important for approaching the steady state on earth: Recycling - Optimizing of materials - New materials. 34 -
RECYCLING OF MATERIALS Fortunately, consumption of materials does not have to be a synonymous with an irrecoverable loss, such as is the case with energy use, but - at least in the most favorable case - must be considered as a stage within the cycle of materials shown schematically in Fig. 7 [10]. The path taken by the substances leads from the resources and raw materials to the materials themselves, which become waste after their use in products of various technological areas. At best the waste products can be recycled; in less favorable cases they have to be stored in refuse dumps. But even those waste products not suitable for reuse are not lost from this cycle, unless they cannot be decomposed by chemical processes or micro-organisms and thus cannot be fed back as resources. Even waste must be considered as a products that cannot simply be thrown away but must be utilized to stretch the available resources. Many successful examples support this point. This might be underlined by Fig. 8 showing the recycling rates for some important classes of materials [3]. The values achieved are a beginning and will certainly be improved in the near future. Much material can be won back through recycling. Recycling is therefore correct as a concept and as a responsibility absolutely necessary. Non-renewable resources must be recycled if the industrial society wants to retain its standard of living at roughly the same level. However, a complete recovery cannot be achieved as the second law of thermodynamics demands a tribute in the form of entropy. A more or less large remainder disappears from exploitability and provides storage and environmental problems, as it doesn't disappear as a substance. OPTIMIZING OF MATERIALS As shown in Fig. 2 there are many classes of available engineered materials, metals, ceramics, polymers etc. with manifold, very different properties, which are intensively studied, initially strictly separated in the classes. Since about 40-50 years, however, the overlap between these classes has become apparent. Materials science as a scientific discipline began to grow. Materials science is supported in many cases by concepts of chemistry, physics and crystallography. - 35
Some time ago, the well-known physicist John M. Ziman said that the coming decades belong to materials science. The considerable success of materials science was its explanation of empirical findings accumulated in large numbers and the resultant improvements and extensions. The great significance of materials science in technological progress is that it can lead to a basic understanding of internal structure, so that new materials can be invented and tailor-made for specific applications, literally by microstructural and molecular design. As can be seen from Fig. 9 the structure and properties of materials aie determined by a whole range of characteristics which extend across a very wide range, from atomic dimensions in the tenth of a nanometre range to the dimensions of structures in the centimeter or meter range [11]. All of the characteristics in this range of scale of several magnitudes contribute, in their own particular way, to the characteristic profile of a given material. In addition to the structure, determined by the interaction of the various types of atomic bonding, the microstructure of a material also plays a significant role in determining its characteristic properties, from nanoto macrostructures. The microstructure is an important domain within the science of materials. More and more often it bridges the gaps in communication between scientists, who seldom enough venture outside their atomic field of interest, and engineers, who show little interest in leaving their safe macroscopic ground of their continuum conception. It is surely immediately apparent that the enormous range within which microstructures occur, also encompasses an extensive and fascinating world which even today cannot as yet be continuously observed to modeled because the effective parameters are too numerous. There are, however, many rules which permit such microstructures to be generated precisely and reproducibly. Even though materials are an ancient cultural inheritance of man, their scientific exploitation began only at the beginning of this century. Today we know that the internal architecture of a material is determined by the type of atoms it contains and their three dimensional arrangement in accordance with certain degrees of order: from strictly ordered arrangements, such as in crystals, to extremely disordered or chaotic arrangements such as occur in some solidified melts. As an example of a material of the highest order. Fig. 10 shows a section from a copper single crystal. The copper atoms are strongly arranged in the cubic face centered structure. Each light colored fleck is produced by a whole column of atoms. The image was produced by transmission 36 -
of a nanometre thick single crystal copper-foil in a high resolution high voltage electron microscope. The direct resolution of this instruments is 0.105 nm, which is less than the distance between the adjacent copper atoms and can thus be resolved [12]. The enormous power of this microscope is best demonstrated by comparison with the human eye. Were our own eyes to have the same ability, we would be able to see a tennis ball on the surface of the moon. In Fig. 11, a direct lattice image of a silicon nitride alloy, areas of higher order are clearly visible from the periodic contrast [13]. Between these areas there exists only a thin amorphous film 0.1 nm thick, .which markedly differs in its composition from the crystalline areas, as can be seen from the electron energy loss spectrum inserted in the upper right of Fig. 11. The Yb203 sintering aid has become concentrated in the glass phase. Two ordered atomic arrangements of differing orientation are apparent, separated by an amorphous phase. The symmetry of the atomic arrangement, in contrast to Fig. 10 exists in various different areas separated by a disordered phase. A microstructure has thus to some extent been formed from both the elements of order and chaos. Whilst symmetry, as the building plan of the structure of a material, plays an important role, it is the deviations from it which make the reality. The type, amount, arrangement, size, shape and orientation of the various phases in their respective ordered conditions all go to form the actual microstructure of the material which thus results from the combination of each of all phases and the defects they each contain. Such defects can be from nil to three dimensional, and in size from vacancies and dislocations to grain boundaries, pores and shrinkage cracks. The various different combinations of these factors result in the fascinating multiplicity of possibilities. The microstructural parameters strongly influence many of the properties of a material. Because of this, a great deal of attention is paid to the microstructure in science, development and testing of materials. As a rule, each material contains many million microstructural features in each cubic centimeter. Even a single crystal, in which such an important feature as the grain boundaries is missing, still contains various different ordered conditions and thus has a microstructure. These microstructural features can exist in sizes spanning more than ten orders of magnitude (Fig. 9). 37
Trends in Application of Materials in the U.S. (Turnover in Billions of US-$) Materials Metals Polymers Anorganic Materials leg ceramic, glass) Others (especially wood) Total 1970 120 (49 %) 36(15%) 38(16%) 49 (20 %) 243 1980 132(46%) 53(19%) 45(16%) 55(19%) 285 1990 135(41%) 76 (23 %) 53(16%) 66 (20 %) 330 2000 141(38%) 96 (26 %) 63(17%) 70(19%) 370 Fig. 3: Trends for consumption of materials in the USA (Turnover in billions of US$) Metal World Resources 1992/93, Production 1992 and Expected Lasting Period Raw Material Iron Bauxite Copper Zinc Lead Nickel Tin Resources Mio. t 68 880 22 983 328 149 70 36 5 Production lOOOt 505 422 108 669 9 405 7 245 2991 905 176 Lasting Period years 136 211 35 21 23 40 28 Source: Bundesanstalt fur Gcowisscnschaften und Rohstoffe (BGR), Oct 1994 Fig. 4: Metal world resources 1992/93, production 1992 and expected lasting period - 44 -
Global Ecosystem Heat Loss into Space Fig. 5: Population and capital in a global system ; r/a '3 Ρ & Ja •s <: Raw Material Supply ƒ Population / ^^ Food Needs / Industrial Production / Pollution ^S 2000 Years ·> *" ^ χ Sw "" ^ >· Fig. 6: Change of the situation in "spaceship" earth (schematically) - 45 -
Raw Materials The Cycle of Materials Elements _J · Metals ^ · Chemicals · Cement ) • Paper Engineering Materials ^^Dispoaall"^^ Scrap f^\ „ Nrjrjdecomposing Waste Fig. 7: Cycle of materials A / Λ 65% y> ; "—>—Í ^ 4B% ) J ! ' S, 36% ) ¿ ' "~Λ 27% 1 10% Plas tics Alum Paper inum Glass Iron a ndStc el ¿ ¿_ / 10 20 30 40 SO 60 70 80 90 100 Fig. 8: Recycling rates of several materials 46
atomic bonding crystal structure component natural sciences basic research engineering sciences ^^~ application Fig. 9: Microstructural characteristics ranging from the atomic to component size Fig. 10: Direct image of atomic arrangement in a copper single crystal - 47 -
i LÌ i.'»Af·· iù^ VV.V.X. iTTaTtV« ·'-'»'<.'<.*»"...,.,.. ι. ..... ,t <t trit It it It II lililí fl fl fi fl .. . *·>*»><■■ ι.'»."»» ». », ,, ι, ir <<-<«<< ititi It il It lllllllllll . .. . - <.».-. Η ·. ~ ...........,, .(.. li í<¿* li tililiti Itfi . . -. #»/#w*»J*M*»r*»ï*» .»"»«:»»'>.'.'.. r,., |.|# %# It It It it II ti 11 ι i 11 U lii -.. . ..-..»».¿.. A. ...»,.».'.»»'..:».», _'.".· *r.» ...» 4.4* .» 4/ il lili Il ti ti tifili i. -..-*»»».-.■. V»» >t ...... »... ...».»,..».»<< i. Il il lili ti It till fit . ·.. ..-.....'.....·..·'. '. ». .... .. .. ..... rf-tfetéflruttliliitiilliillilllil . ·..'.,■». ;'-»;»..■■.'.»» ». »......».»·.». .#■«. rr trtt it if tt It it lllltltl il II f I f »...i.». .. », ™ ». .... ». ¿f srit tr it tt If It ti If li tl il tl lifl il f t ·». .. '. -..A, ». .. .. »»al· .. ». i. .. .. I. ». »...-.♦ lritirittll±iAj'"'"'l'l''·' .'V -^ >,. ... ... ...'..». »..V........ »... tr.lr.i-* rr-..r.rr*t it tf if It li ti ti "^ ... -..».. ..»,.'..».. i. /.»..», ^»....J. *».«»;»» tr^rς* .ir #V.,> »V /# IV ^tf | |Yb203-fluxed Yb at the grain boundaries Fig. i l : High resolution electron microscopic image of a Silicon Nitride (Si3N4) alloy with and amorphous grain boundary phase in which the sintering additive Ytterbium Oxide (Yb203) has become concentrated (compare with electron energy loss spectrum in the insert) Al203 + 15 vol % Zr02 Sintered 1550°C,2h Strength : 570 MPa Sintered U50°C, 2h HIP 1600 °C. 10 min Strength: 1050 MPa Fig. 12: Example of a microstructural optimization of a dispersion-strengthened alumina-zirconia (AKOi-ZrOi) ceramic - 48 -
(Number of Systems) 30 25 20 15 10 ■---'-■--■ I .... I .... I ζ = '86Ί Ν J unknown ++ known ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' 0 20 40 60 80 Number ot Components Ν Fig. 13: The "mountains" of materials - 49 -
1600 OJ 1200 .« 1100 d 1000 900 800 1960 SiC alloys Thermal barrier coated superalloys Directional superalloys Nonmetallics Intermetallic compounds Fiber-reinforced (FR) superalloys Rapidly solidified (RS) superalloys Oxyde-dispersion-strengthened (ODS) superalloys Single crystal (SC) superalloys Directionally-solidiRed (DS) superalloys Conventional superalloys 1970 1980 1990 Year of initial engine use 2000 2010 Fig. 14: Evolution of hightemperature materials 400 E •5 JC — 320 Ι Ο 3 α DC Ζ40 α. >- 13 2Í 160 ■ζ ω 2 3 BO S χ < 2 °I9 ι 1 _ Steel CoFerrife «_ o rJ>= O 1920 1930 1 Alnico 1 1940 ι ι ι ι Τ t Nd2F*|4B Alloyi of StnCo. »__ y^ Sintered SmCo Columnar Alnico Î o IsrnCo, |BaSrFerrile r> 1 1 1 " . 1950 I960 1970 1980 199 50 s o CD 2 *■* 40 H CJ 3 a o .„ o: 30 £ ΐΟ tr 20 ζ UI 3 io s χ < ό° YEAR Fig. 15: Evolution of the maximum energy product of permanent magnets - 50 -
uu 120 no too * 90 g έ· 80 Q. J 60 I 50 I» 30 20 10 η Liquid Ν; Liquid H_ HgP^J Nb£ 1 February 1988 TICa -Ba-Cu-0 January 1988 BiCaSrCu-0 April 1987 YBa-Cu-0 | March 1987 YBa-Cu-0 B1-era . A15-era March 1987 La-Sr-Cu-0 ι January 1987 La-Ba-Cu-0 January 1987 La-Sr-Cu-0 ι April 1986 La-Ba-Cu-0 1 I / Oxides NbNj *%£ *'-*-^ Lo-Bo-Cu-O S BoPbir.'.Bi& BŒ Fig. 16: Evolution of high Tc-Superconductors a Cellular Structure Anisotropy GO < -10 -10' -io; -10< -iof -ioÉ Polymers Metals Ceramics Stone Wood/Bone Fig. 17: Challenges and trends in materials development - 51 -
Al Phase Relations in the Aluminium-rich Part of the System: Aluminium-Iron-Manganese F. Weitzer*, P. Rogl* and M. Bonn** *lnstitut für Physikalische Chemie, Universität Wien, A-1090 Wien, Währingerstraße 42, Austria *CNRS-URA 1278, Centre de la Microsonde Electronique de l'ouest, ¡FREMER, F-29263 Plouzané, Brest, France Abstract Phase relations in the manganeserich part of the ternary system AlFeMn have been established for an isothermal section at 550 °C for Al concentrations higher than 70 at.%. Experimental techniques employed were optical microscopy, ΕΡΜΑ and Xray powder diffraction analysis of arcmelted alloys. The samples were subsequently annealed in Al203crucibles sealed in evacuated quartz capsules for up to 1500 h and were finally quenched in cold water. Phase equilibria in the AlFeMn are characterized by the formation of extended solid solution ranges extending from the binary far into the ternary system. The homogeneity regions determined at 550 °C were: (FexMn._x)Al6 with the MnAl6type for 0 < χ < 0.6; (FexMn..x)Ai4,15 (λ-ΜηΑΐ4+χtype) for 0 < χ < 0.08 and (FexMni.x)Al4 (μ-ΜηΑ14-Ιγρε) for a ternary compound τι with a small range at 550 °C for 0.08 < χ < 0.20. 1 Introduction The various degrees of beneficial influence of alloying elements on grain refinement and precipitation hardening in aluminium spurred an early interest in aluminium-ironmanganese alloys. Despite numerous investigations, which dealt with the constitution of the Al-Fe-Mn system, a complete phase diagram has not yet been established. A critical assessment of all the information available in literature up to 1990 was provided by [92Ran]. For a better understanding of alloy behaviour in technical applications as well as a basis for a proper thermodynamic description of the ternary diagram, work throughout COST 507-11 attempted to determine the phase relations for the Al-rich region of the ternary system Al-Fe-Mn within a partial isothermal section at 550 °C including a reinvestigation of the isothermal reactions in the Al-Mn system by high resolution Smith thermal analysis. 2 Experimental Samples, each of a total weight of about lg, were prepared by arc melting 4N ingots of Al (Alfa Ventrón, Karlsruhe, FRG), platelets of Mn (99.9% pure, South Africa) and lumps of Fe (claimed purity of 99.9%, supplied by J.Matthey & Sons, UK). To ensure maximum homogeneity the alloy buttons were remelted several times under as low an electric current as possible in order to keep the total weight losses below 0.5 mass%. A part of each button was annealed at 550 °C for up to 1500 h. Each specimen was contained in a small alumina crucible and sealed under vacuum in a quartz tube which - 53 -
three different sections starting from binary liquid Cu-Mg alloys (50, 75, and 90 at% Mg), and one section starting from a binary liquid Mg-Y alloy (8 at% Y). The obtained experimental results for ternary liquid alloys were interpreted using an association model [82Som2, 90Som]. It was found that a good description was possible with the assumption of three binary associates (Cu2Mg, CuY, Mg2Y). With the model parameters obtained from the three limiting binary systems it was possible to calculate thermodynamic properties for ternary liquid alloys. Although there are some indications for ternary interactions, the existing experimental data are still not extensive enough to warrant a corresponding evaluation. All experimental details and the final results are summarized in Appendix 2 and are published as [97Gan2]. 4 The Ternary Cu-Mg-Si System 4.1 Thermodynamic Properties Magnesium vapor pressures were determined along an isopleth with a constant concentration ratio of Xcu/xsi = 7/3 using again the described isopiestic method [93Gna]. This particular composition was selected because of a minimum melting temperature of 802°C at the eutectic point (30 at% Si) in the Cu-Si binary system [8601e]. It was supposed to guarantee a maximum stability range of the liquid phase in the corresponding ternary section which, however, was reduced considerably by the appearance of a melting point maximum of about 950°C around 33 at% Mg and 13 at% Si [39Wit, 60Asc]. From the vapor pressure measurements, partial thermodynamic properties of magnesium were derived; they are presented for a temperature of 1173 K. A GibbsDuhem integration was performed based on the method of Peiton and Flengas [69Pel], using the obtained magnesium activities to calculate integral Gibbs energies of mixing for liquid alloys in the investigated isopleth. The enthalpy of mixing in this isopleth with Xcu/xs¡ = 7/3 was determined by solution calorimetry [80Som]. Measurements were performed starting both from liquid binary Cu0 7Si0 3-alloys (and adding solid magnesium) and from liquid magnesium (adding solid Cu07Si03-samples). Partial enthalpies of mixing of magnesium were derived from the obtained data and were found to be in very good agreement with those derived from the temperature dependence of the magnesium activities. The obtained thermodynamic results are again interpreted by the described association model [82Som2, 90Som]. All experimental details as well as the final results can be found in Appendix 3 and will be published soon [97Gan3]. 4.2 Phase Diagram The Cu-Mg-Si phase diagram along the isopleth with XcAsi = 7/3 was studied by means of differential thermal analyses (DTA) and X-ray powder diffraction methods. With these it was possible to derive the shape of the liquidus curve. The phase relationships at lower temperatures are very complicated (see also [60Asc]), and it was found that additional measurements will be necessary outside the investigated section - 60 -
(especially around 33 at% Mg) in order to be able to interpret all experimental obser vations unequivocally and to construct a correct and consistent phase diagram. These experiments are still in progress. 5 Acknowledgement The financial support of this study by the Austrian „Bundesministerium für Wissenschaft, Verkehr und Kunst" (contract No. GZ 49.888) and the German „Bundesministerium für Wissenschaft, Forschung und Technologie" (contract No. 03K72025) within the framework of the COST 507 project is gratefully acknowledged. 6 References 39Wit H Witte, Metallwirtschaft, 1939,18, 459463. 60Ase L J Asehan, Acta Polytechn Scand, Chem Includ Metall Ser No 11, 1960, 163 69Pel A D Peiton, S N Flengas, Can J Chem, 1969, 47, 22832292. 73Roc E G Rochow: Comprehensive Inorganic Chemistry, vol. 1, J C Bailor, Jr, H J Emeleus, R Nyholm, and A F TrotmanDickenson, eds, Pergamon Press, Oxford, 1973, p. 1361. 80Som F Sommer, J J Lee, Β Predel, Ζ Metallkd, 1980, 71, 818821. 82 Sonil F Sommer, Materials and Physical Chemistry in Liquid Metal Systems, H M Borgstedt, ed, Plenum Press, New York, 1982, 387393. 82 Som2 f Sommer, Ζ Metallkd, 1982, 73, 7276; 7786. 8601e R W Olesinski, G J Abbaschian, Bull Alloy Phase Diag, 1986, 7, 170178. 89 Ips H Ipser, R Krachler, K L Komarek, Thermochemistry of Alloys, H Brodowsky and ΗJ Schaller, eds, Kluwer Academic Publishers, Dordrecht, 1989, pp. 293306. 90Heh F Hehmann, F Sommer, Β Predel, Mater. Sci. Eng., 1990, A125, 249. 90Som F Sommer, J NonCryst Solids, 1990,117/118, 505512. 920ka H Okamoto, J Phase Equil, 1992,13, 102103. 94Feu H Feufel, M Krishnaiah, F Sommer, Β Predel, J Phase Equil, 1994, 15, 303309. 94Gna Τ Gnanasekaran, H Ipser, Met Mater Trans Β, 1994, 25B, 6372. 95Feu H Feufel, F. Sommer, J Alloys Comp, 1995, 224, 4254. 97Ganl V Ganesan, H Ipser, J Chim Physique, 1997, in print. 97Gan2 V Ganesan, F Schuller, H Feufel, F Sommer, H Ipser, Ζ Metallkd, 1997, submitted for publication. 97Gan3 V Ganesan, H Feufel, F Sommer, H Ipser, Met Mater Trans, 1997, publication in preparation. 61
A4 Summary of Final Report: Project: A4 Experimental Investigations and Thermodynamic Modelling of the Constitution of Quaternary Al-Ti-Metal-Nonmetal Systems Julius Schuster and Peter Rogl Institut für Physikalische Chemie Univ. Wien, Währingerstr. 42, A-1090 Wien, Austria The results achieved have been presented in form of several manuscripts and publications. Active research cooperation existed throughout the project with groups D4, Dll.Il.andUKl. l.System Ti-Al-C-N. Thermodynamic calculations of the boundary systems Ti-Al-C and Ti-Al-N revealed consistency of the observed phase equilibria with the experimental results obtained in COST 507-round I. For completion of the investigations started in Cost507-I phase equilibria were investigated experimentally at 1375°C. For alloys up to 50 at.%Ti 16 four-phase equilibria were observed or are likely to exist. A scheme of the existing threeand four-phase was established. The two N-phases showed no solubility for the corresponding nonmetal. Although the two Hphases form a complete solid solution at 1480°C, there is a miscibility gap in the nitrogen-rich region. For the region with more than 50 at% Ti a vertical section Ti3AlCxNi.x was investigated revealing a novel quaternary perovskite phase P(t) with tetragonal lattice distortion. At temperatures below 1200°C a Re3B type phase Ti3AIN was discovered which proved to be isotypic with Zr3AlN. These results were discussed in detail in publications 1,2. 2.System Ti-Cu-AI-N. As there were no data available on the Ti-Cu-N boundary system a reinvestigation was necesssary in terms of an isothermal section at 850°C revealing a novel phase Ti3CuN with hitherto unknown crystal structure (Ref. 3). Based on these results as well as on new critical assessments of the Ti-Cu-Al (Ref. 4) and Ti-Al-N (Ref. 5) ternaries, the major part of the Ti-Al-Cu-N quaternary was investigated experimentally at 850°C employing XRD, SEM-EDX and LOM (Ref. 6). The composition of the quaternary eta phase was characterized with respect to its metal ratio by SEM-EDX. A slight deficiency in N is likely (Ref. 7), From the thermodynamic data of the binary boundary systems as well as from the evaluated phase equilibria in the ternary boundary systems thermodynamic data could be extracted for the ternary boundary phases and the quaternary eta-phase (Refs. 8, 9). Employing these data we could prove and amend consistency in the entire quaternary system Ti-Al-Cu-N. 62 -
3.System Ti-Sn-Al-N. The investigation of the phase equilibria of the boundary system Ti-Al-Sn in terms of an isothermal section at 900°C was extended into the melting range. The existence of the ternary phase Ti5AlSn2 was also observed in as cast alloys and isotypic phases were also synthesized for the combinations (Ti.Zr.Hf)- (Sn,Pb)-Al (Ref. 10). The ternary boundary system Ti-Sn-N was studied at 900°C employing XRD, SEM-EDX and LOM. No ternary phase was found. Similarly no quaternary phase was encountered in the experimental investigations of the Ti-Al-Sn-N quaternary. The final results are presented in two papers, (Ref.l 1) and (Ref. 12). 4.System Ti-Ni-Al-N. Detailed investigations concerned the ternary phase equilibria Ti-Ni-N at 900°C, as only limited information existed for the join TiN-Ni. No ternary compound exists, however, a large N-uptake up to 11 at%N has been observed from quantitative ΕΡΜΑ for Ti2NiNi.x (x < 0.5) (Ref. 13). Based on the thermodynamic calculation of the quaternary system in cooperation with group D4 (TU-Clausthal), specific regions of interest have been investigated in detail: (a) precise solid solubility limits of Ni in a2Ti3Al and γΤίΑΙ and (b) isothermal reactions and solidus/liquidus for selected ternary alloys employing Smith calorimetry. Phase Equilibria in the Ti-Al-Ni isothermal section at 900°C and 1000°C have shown several discrepancies to earlier data in the literature. A compound Ti3NiAl2N has been identified and characterized by XRD (Ref. 14). 5.System Ti-AI-Mn-N. The investigation of the phase equilibria in the Ti-Mn-N ternary at 900°C revealed the existence of hitherto unknown new compounds in the TiMn binary. A reinvestigation of the binary system in cooperation with group II (Genua) has been performed with XPD, LOM, EMPA and DTA. A complete constitutional diagram χ vs. Τ from 600°C to melting was presented (Ref. 15). Three incongruently melting compounds exist in the region TL^Mnss to TÍ55M1145, and four peritectic phases have been identified in the region between TiMn2 and Mn. The ternary phase equilibria studies were also extended up to 1 bar of N2 at 900°C (Ref. 16 and Ref. 17). No ternary compound exists. Ó.System Ti-Al-B-C. With respect to the inconsistencies in literature data concerning the calculation of the quaternary system, the section TiAlB was reinvestigated at 1000°C. Quantitative EMPA proved the coecistence of TiB, TÍ3B4 and TiB2 with a2 Ti3xAll+x (35.5 at.%Al). Based on this new experimental results a thermodynamic calculation of the System TiAlB was attempted by our partner group DU (MPI Stuttgart). Results are summarized in Ref. 18. 7. Further contributions to COST 507Π. Within the regularly held meetings among the partner groups in Cost 50711, latest results were reported and ad hoc cooperations were carried out with Cost 50711 partners in further systems of interest. (Ti,V,Ta)-AI-N with COST507II partner D4 (TUCIausthal): Logistical assistence was given regarding the description of these three ternary systems. WB, W-B-C with COST507 Partner DU (MPI Stuttgart): As part of the quinary system the subsystems WB and WBC have been assessed and modelled by 63
thermodynamic optimization. A full set of thermodynamic data has been published for W-B (Ref. 19) and is provided for W-B-C (Ref. 20). Hf-C, Zr(Hf)-B-C with group UK1 (NPL-London): The subsystems Hf-C Ti-Zr-B, Ti-Hf-B, Zr-B-C, and Hf-B-C have been assessed and calculated by thermodynamic optimization (Ref.21 to 25). A full set of thermodynamic data has been provided. Difficulties in the binary boride evaluation have forced a calculation and optimization of the data sets on the basis of ternary boride systems (26, 27). Duration of project: 3 years Total amount of funding received (in ECU): 147.500.- Total manpower involved (in man years) 3.5 Number of publications from COST 507-11: 27 (a) number of contributions to scientific journals: 7 (b) number of contriubutions to conferences: 9 (c) other: 11 Publications (Manuscripts or copies of References Nr. 1-6, 13,16,17,19 and 20 have already been submitted as part of the annual reports 1994 and 1995. Manuscripts of the remaining publications and contributions covering the period following the last annual report are enclosed below). 1) M.Pietzka and J.C.Schuster "Das Perowskitsystem Ti3AlCi.x-Ti3AlNi.x", in "Nichtmetalle in Metallen '94", D.Hirschferld ed., (DGM Informationsges. Verl., Oberursel, FRG, 1995) 227-2136 2) M.Pietzka and J.C.Schuster "Phase Equilibria in the Quaternary System Ti-AlC-N" J.Amer.Ceram.Soc, 79 (1996) 2321-2330 3) N.Durlu and J.C.Schuster "The Ternary System Titanium-Copper-Nitrogen" δ"1 Int.Conf. on High Temp.Mater.Chemistry, Wien, Austria, 1994. 4) J.C.Schuster "The System Ti-Cu-Al", manuscript submitted to COST507-II, Group Β (June 1995) 5) J.C.Schuster "The System Ti-Al-N", manuscript submitted to COST507-II, Group Β (June 1995) 64
6) N.Durlu, U.Gruber, M.Pietzka, H.Schmid and J.C.Schuster "Phases and Phase Equilibria in the Quaternary System TiCuAlN" Z.Metallk., accepted 7) J.C.Schuster "Quaternary Nitrides Having ηType Crystal Structures", paper presented at "Journée de Etudes sur les Nitrures" (St.Malo,France,1996) 8) M.APietzka and J.C.Schuster "A Simple Method to Estimate the Stability of Ternary and Quaternary Phases" paper presented at "CALPHAD XXV" (Erice,Italy,1996) 9) M.A.Pietzka and J.C.Schuster "Estimation of the Thermochemical Stability of Ternary and Quaternary Phases Within the System TiCuAlN at 850°C" manuscript in preparation 10) , M.APietzka and J.C.Schuster "New Ternary Aluminides T5M2AI Having W5Si3Type Crystal Structure" J.Alloys and Compounds, 230 (1995) .L10L12 11) J.C.Schuster and M.A.Pietzka "On the Constitution of the System TiAlSnN", paper presented at "Thermodynamics of Alloys" (Marseille, France, 1996) 12) M.APietzka and J.C.Schuster "Phase Equilibria of the Quaternary System Ti AlSn N", J.Alloys and Compounds, in print 13) Y.LeFrieck, J.Bauer, P.Rogl and M.Bohn "The NNiTi System", J.Phase Equilibria, to be submitted 14) B.Huneau "Etude du Systeme Ternaire AlNiTi" INSA Rennes, Diploma Thesis (1996) 15) R.Krendelsberger, P.Rogl, J.Bauer, A.Saccone and R.Ferro "The TitaniumManganese System", paper presented at the Hauptversammlung DGM (Stuttgart, FRG, 1996) 16) .J.Bauer, R.Krendelsberger, P.Rogl and D.Antoine "The MnTiN System" Met.Trans,Ser.A., in preparation 17) P.Rogl, D.Antoine and J.Bauer "The ManganeseTitaniumNitrogen System", paper presented at Hauptversammlung DGM (Bochum, FRG, 1995) 18) C.Bätzner, G.Effenberg, P.Rogl, J.Bauer and M.Bohn "Experimental Investigation and Thermodynamic Modelling of the System AlBTi", paper presented at " 12th Intl.Symp.Boron, Borides and Rel.Compounds" (Baden, Austria, 1996) 19) H.Duschanek and P.Rogl '"Critical Assesment and Thermodynamic Calculation of the Binary System BoronTungsten", J.Phase Equilibria 16 (1995) 150161 65
20) H.Duschanek and P.Rogl "Critical Assessment and Thermodynamic Calculation of the Ternary System Boron-Carbon-Tungsten", to be published in "Ternary Alloys. Special Issue" *21) HBittermann and P.Rogl "The System Hafnium-Carbon" J.Phase Equilibria, to be submitted *22) H.Duschanek and P.Rogl "Critical Assessment and Thermodynamic Modelling of the Ternary System Zirconium-Boron-Carbon" manuscript submitted to COST-II Groups B,C (Nov. 1996) *23) H.Bittermann and P.Rogl "A Thermodynamic Modelling of the Ternary System Hf-BC", paper presented at "12th Intl.Symp.Boron, Borides and Rel.Compounds" (Baden, Austria, 1996) *24) HBittermann and P.Rogl "Critical Assessment and Thermodynamic Calculation of the Hf-B-C System", manuscript submitted to COST-II Group ... (Dec. 1996) 25) C.Bätzner, G.Effenberg and P.Rogl "Metal BoroCarbides in the Electronic Materials Science International Workplace", paper presented at "12th Intl. Symp.Boron.Borides and Rel.Compounds" (Baden,Austria, 1996) 26) HBittermann, H.Duschanek and P.Rogl "A Thermodynamic Modelling of the Ternary System Ti-Zr-B and Ti-Hf-B", paper presented at CALPHAD XXV (Erice,Italy,1996) 27) HBittermann and P.Rogl "Critical Assessment and Thermodynamic Calculation of the Ternary System Boron-Hafnium-Titanium", J.Phase Equilibria, in print 66 -
CHI Summary of final report Directional solidification and phase equilibria in ΑΙ-Ni system O. Hunziker and W. Kurz Laser surface resolidification and Bridgman directional solidification experiments have been performed in order to determine the microstructure selection map for Ni-rich Ni-ΑΙ alloys and Ni-Al-Ti alloys. It has been shown that taking into account disorder trapping is necessary in order to explain why does γ'-Νί3Α1 only grow as primary phase from the liquid at low velocity (below 100 μιη/s). The effect of disorder trapping in γ' has been evaluated by laser experiments on Ni3(Al,Ti), where growth of γ' is thermodynamically favoured by the presence of Ti. An interactive program calculating the microstructure map has been developed on the basis of models for dendritic, eutectic and plane front growth. This program is an efficient tool in order to optimise the phase diagram by fitting the calculated microstructure map to the experimental map. The phase diagram determined by this optimisation method displays the following characteristics, necessary to fit the solidification results: - A very small temperature difference between the stable β-γ' and the metastable β-γ eutectics (<0.05K). This temperature interval is probably overestimated in all recent versions of Ni-Al phase diagram. - Similar liquidus slopes for γ-Ni and γ'-ΝΪ3 Al near the eutectic temperature. - Solid composition of γ'-Νί3Α1 close to 76 at% Ni. - Solidification interval for γ'-Νί3Α1 between 1 and 4 K, determined at 2.3 Κ in the optimised version. 1646 r te & 1643 ε Ol H 1642 74.5 75 75.5 76 76.5 Composition [at% Ni] Optimised phase diagram with metastable extensions near the melting point of γ'-Νί3Α1. Duration of project: 2 years Total amount of funding received (in ECU): 53ΌΟΟ.- (79Ό00.- SFr) Total manpower involved (in man-years) : 2 Number of publications from action COST 507-Π : 3 (a) Number of contributions to scientific journals : 1 (accepted) (b) Number of contributions (posters, present at conf.) : 2 (1 accepted) 67
D2 Thermodynamic Investigations of Cu-Mg-Si, Cu-Mg-Y, AI-MgZn, Al-Cu-Mg, Al-Mg-Si, Al-Cu-Li and Al-Cu-Zr Alloys Y.B. Kim, H. Feufel, U. Stolz and F. Sommer Max-Planck-Institut für Metallforschung Seestr. 75, D-70174 Stuttgart, Germany Abstract The enthalpy of mixing of liquid ternary aluminium and magnesium alloys has been determined to understand the thermodynamics and the phase equilibria of these light alloys. The experiments were performed using a high-temperature isoperibolic type of solution calorimeter. Results for four sections with constant composition ratios of two components are obtained for liquid Al-Cu-Mg alloys at 986 K. Within the range of experimental accuracy, the measured enthalpies of mixing of a single series and of the points of intersection are self-consistent. An association model is used to calculate the thermodynamic mixing functions of ternary liquid alloys based on model parameters which are determined from thermodynamic results of the base binary systems. The calculated enthalpy of mixing shows systematically less negative values than the measured values. The existence of ternary phases and the possibility to obtain metastable quasicrystals by rapid solidification show that these deviations are caused by additional ternary interactions in the liquid state. These ternary interactions are correlated with a value of 1.8 of the mean number of conduction electrons per atom. The enthalpy of mixing of liquid Al-Mg-Zn ternary liquid alloys was determined in the temperature range between 883 K and 933 K. An association model is used to 68 -
calculate the thermodynamic mixing functions of the ternary alloys based on the thermodynamic properties of the binary bordering systems. The calculated values of the enthalpy of mixing show a good agreement with measured values within the range of experimental accuracy. The presence of additional ternary interactions or ternary associates in the liquid state could not be observed. Enthalpies of mixing of ternary liquid Cu-Mg-Y alloys were determined along five different sections. For one of these sections (XGAY = 0.5), magnesium vapor pressures were measured by means of an isopiestic method (A2), and the corresponding values of the magnesium activity at 1173 K were derived as a function of composition. The Gibbs energy of mixing was computed by a Gibbs-Duhem integration. An association model was applied to calculate the thermodynamic functions of mixing for ternary liquid alloys using the model parameters of the base binary systems. Thermodynamic properties of ternary liquid Cu-Mg-Si alloys with a constant composition ratio xci/xs¡ = 7/3 were determined using a combination of different experimental methods: Enthalpies of mixing were measured by isoperibolic calorimetry and magnesium vapor pressures were obtained by an isopiestic method (A2). Partial thermodynamic properties of magnesium were derived from the vapor pressure data, and the composition dependence of the magnesium activities is given for 1173 K. Gibbs energies of mixing for liquid alloys were calculated by a Gibbs-Duhem integration. Finally, an association model was applied to provide the thermodynamic functions of mixing for ternary liquid alloys using the model parameters of the three limiting binary systems. The enthalpy of mixing of liquid Al-Cu-Zr alloys were determined along eigth different sections at 1480 ± 5 K using two different measurement procedures. The enthalpy of mixing exhibit very negative values with a rninimum of about -53 kJ mol"1 at the composition Al50Cu2oZr3o. 69 -
and have been independently described by different sublattice models in the first round of the COST-507 project, Table 3. The presence of a non-first order transition from D83 to D82 seems necessary to make compatible both structures, although there is lack of experimental data to corroborate it. Thus, it may be essential to make systematic measurements of Cp along the γ-phase region in order to determine the possible order-order transition boundary. Table 3. Structures of the binary γ-phases and sublattices used to describe them. γ phase Structure Sublattice model in binary Al-Cu D83(cP52), Ρ 4 3m (Al)16:(Al,Cu)4:(Cu)32 in binary Cu-Zn D82(cI52), 14 3m (Cu,Zn)8:(Cu,Zn)8:(Cu)12:(Zn)24 From the point of view of the sublattice models, two alternatives are possible. Either to build an „hypothetical intermediate" sublattice to connect both structures or to modify the binary descriptions of the binary γ-phases. The first alternative has some advantages, such as its rapid implementation and the lack of necessity for a change in the binary descriptions already in use. Its main disadvantage is related to the complications it introduces in the phase description, which will increase for higher order systems. Thus, even if this „provisional" solution is first used, a modification of the binary descriptions seems to be inevitable in the long term. The new model should account for the experimentally observed structures and atomic occupancies for the whole range of the γ-phase. Experimentally, eight crystallographically different positions are observed, which in a simplified model could be reduced to four. In this case a possible sublattice model for describing the γ-phase would be: (Al,Cu,Zn)8 : (Cu)8 : (Cu,Zn)12 : (Al,Cu,Zn)24 The ternary Al-Cu-Zn system also displays one ternary phase, τ, with a large range of solubility. This phase is present below 740°C at compositions close to Cu4oZn2oAl34. Its main feature is that at low temperatures its range of existence is split into two non-interconnected zones. In one of these regions, denoted as τ', the experimental results show that a superstructure is created by the presence of structural vacancies, fig. 11. The vacancies produce a loss of symmetry in this τ' phase (rombohedral) with respect to the high temperature τ phase (B2). Crystallographically, both τ and τ', could be described as a single phase using a two sublattice model containing vacancies: (Al,Zn)i : (Cu,Va)i However, such a model, although simple, would not be compatible either with the present descriptions of the B2 structures or with the description of the A2B2 ordering transition. Thus, the actually compatible model should be (Al,Cu,Zn,Va)i : (MCaZn^Va)! - 76 -
There is a complete lack in the literature of thermodynamic data for the τ and τ' phases. Thermodynamic measurements for this phase, such as enthalpies of formation and Cp, are badly needed. 2.4 The Cu-Mg-Si system The CuMgSi system displays one Laves phase, C15 and two near stoichiometric ternary phases σ (CuióMgöSi?) and τ (Cu3Mg2Si). Published experimental information for the system is limited to the Cucorner, where DTA and Xray measurements have been reported in [60Asc]. For the thermodynamic assessment of the CuMgSi system, the partial and integral enthalpies of mixing, the activities of Mg in the liquid phase and DTA detenriinations of phase boundaries reported in [97Gan] were used. The calculations are compared with experimental data in Figures 12 and 13. The invariant reactions which were used for the optimisation of the system are presented in Table 4. Table 4. Invariant reactions Phase equilibria liquid = (Si) + CU19SÌ6T1 + σ liquid = CuMg2 + Mg2Si + τ liquid = (Si) + Mg2Si + τ liquid = σ liquid = τ Literature 1015 [97Gan] 1013 róOAscl 797 [97Ganl 1043 f39Witl 1099 T97Ganl 1203 f97Ganl present work, Τ (Κ) 1015 798 1037 1102 1206 The measurements were carried out for the isopleth Cu7Si3 + Mg only. A complete assessment of the system requires additional experimental data. 2.5 The Cu-Mg-Y system For the thermodynamic assessment of the system use has been made of available thermodynamic evaluations of the binary subsystems. The basis of the evaluation is provided by the thermodynamic measurements carried out by [97Gan]. Figures 14 and 15 show the calculated partial enthalpies of Y in the liquid phase compared with the experimental data. 2.6 The Cu-Mg-Ni system In the system CuMgNi experimental information concerning the solid phases is lacking. Therefore the evaluation of the system is restricted to the tempera ture and concentration range for which experimental data are available. The activity data of Mg in the liquid phase as reported in [93Cmal] and [93Gna2] for several isopleths are reproduced well. In Fig. 16 the experimental activity data are plotted together with the result of the calculation for the 77
isopleth XCU/XNÌ=0.5. fri Figure 17 the results of the calculation of this isopleth are compared with the experimental data. Acknowledgement The financial support of the German Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie, project 03K7203 6, within the framework of COST Action 507, is gratefully acknowledged. 3 References [34Mat] K.Matsuyama, Kizoki no Kenkyu, 1934, 11, 461-190. [36His] C.Hisatsune, Mem.Coll.Eng.Kyoto to Hagane, 1936, 22, 597-622. [39Wif] H. Witte, Metallwirtschaft, 1939, 18, 459-463. [40Smi] C.S.Smith, Trans. AIME, 1940, 137, 313-329. [46Pet] D.A.Petrow,GS.Berg, Zhur. Fiz. Khim., 1946, 20, 1475 [49Ura] GG.Urazov, M.S.Mirgalovskaya, Izv. Sekt. Fiz. Khim. Anal., 1949, 19,514. [53Phi] H.W.L.Philips, J. Inst. Metals, 1953, 82, 9-15. [55Zam] M.I.Zamotorin, Lenin, polit, instit., Trudy, 1955, 180, 32. [59Phi] H.W.Phillips, J. Inst. Metals, 1959, 25, 27. [60Asc] L.J.Aschan, Acta Polyt.Scan., Ch 11, 1960, 285. [72Pre] B.Predel, H.Ruge, Mater. Sei. Eng., 1972, 9, 141. [73Hul] R.Hultgren, D.D.Desai, D.T.Hawkins, M.Gleiser and K.K.Kelley, Selected Values of Thermodynamic Properties of Binary Alloys, American Society of Metals, 1973. [81Mel] E.V.Mel'nik, V.V.Kinzhibalo, Russ.Metall., 1981, 3, 154. [83Bel] A.I.Beljaew, Metalloved. alumin., Metallurgia, Moscow, 1983, 73. [86Rod] E.K.Rodionova, N.M.Martynova, Zh.Fiz.Khim., 1986, 6, 1382. [86Not] M.Notin, M.Dirand, D.Bouaziz, C.R.Acad.Sci.Paris, 1986, 302, 63 [90Kuz2] G.M.Kuznetsov, L.N.Kalkulova, O.B.Mamzurin, Izvest.Vyssh. Zaved., 1990, 2, 94-100. [91Pri] A.Prince, G.Effenberg "Aluminium-Copper-Magnesium", inTernary Alloys, ed. GPetzow, G.Effenberg, Vol.4, VCH Verlagsges., Weinheim, 1991, 547. [93Cmal] T.Gnanasekaran, H.Ipser,J.Chim.Phys.,1993, 90, 367-372. [93Cma2] T.Gnanasekaran, H.Ipser,J.Non-Cryst.Solids, 1993,156-158,384-387. [95Kim] Y.B.Kim, F.Sommer, B.Predel, Z.Metallkd., 1995, 86, 597. [95Soa] D.Soares, L.F.Malheiros, M.Hämäläinen, F.Castro, J.Alloys and Compounds, 1995, 220, 179. [97Gan] V.Ganesan, F.Schuller, H.Feufel, F.Sommer, H.Ipser, 1997, to be published. 78
a.B5 ø.ia 0.15 0.20 0.25 0.30 0.35 0.40 mo Ir frae t i On ΞΙ Fig. 1. The calculated phase diagram CuSi compared with experimental data. 0.12 01 0.02 0 0.2 0.4 0.6e15 ø.a 9 mole fraction Cu Fig.2.Isothermal section at 785 Κ in the AlCuMg system. 0. 16 1.0 0 5 10 15 20 25 30 35 40 E3 uj(f ccfil, Cu) Fig. 3. Solubility ranges of Cu and Mg in (Al) in the AlCuMg system. Calculation or experiment i59Phil ternary L(fcc) used no ternary L(fcc) used solvus surface 0 ^— '-' Invari ants • ♦ ■ Fig.4.Solvus surface in the Alcorner in the AlCuMg system. 0.02 0.04 0.0S weight fraction Cu Ø.øa 79
0 Mg 0.2 0.4 0.6 ø.a 1.0 mole fraction Cu ^ (Bl)»T*AlHg( 0 0.05 0.10 0.15 0.20 0.25 0.30 uielght fraction Mg Fig.5.Liquidus surface in the AlCuMg system. Fig.6.Isopleth at 70 wt.% Al in the AlCuMg system. A τ Γ 0 25 50 75 100 125 150 ε3 weight fraction Si Fig.7.Liquidus surface in the AlCuSi system. Fig.8.Isopleth at 4 wt.% Cu in the AlCuSi system. 80
700 [53Phi] ]iquld a » » ■*h·^» f t » (fil)+fil2Cu+(Sl) 800 +[90Kuz21 04A 0.05 0.10 0.15 0.20 0.25 lueight fraction Si 0.1 0.2 0.3 ujeight fraction Cu Fig.9.Isopleth at 1 wt.% Si in the AlCuSi Fig.10. Isopleth at 2 wt.% Cu in the AlCuSi system. system. <111> τphase #0·0·0·0·0· τ'phase #0·0·0·0·0# φ Al, Zn, [Va] 0Cu'Va 0Va Fig. 11. Schematic representation of the atomic occupancy in the τ and τ' phases for the AlCuZn system. The direction <111> in the cubic lattice is considered. 0 r -5 -10 -15 3+x(Cu)7*x(Sl)=0 A [97Gan],T = 1100 Κ Q[97Gan),T970 Κ 25 A 0 0.2 0.4 0.6 0.8 mole fraction Mg Fig. 12.Integral enthalpy of mixing in the CuMgSi system. 1.0 900 800 3 700 600 500 400 A 300 Mg2Si+ +CuMg2+T CuMg2+(Mg) +Mg2Si 0.2 0.4 0.6 0.8 mole fraction Mg 1.0 Fig. 13.Isopleth Cu7S¡3+Mg compared with experimental data [97Gan]. 81
-15- -20A -50E3 -559»x(Cu)-x(Mg)-0 T=1023 K -21- -240 „-3Bα. «-33X ti c -36II --39- * -42- « °--45- -4BE3 -51i 3lx(Cu)-x(Mg) T=1023 Κ */ * / * 1 ι I -0 / * / 1 1 - - 0 0.05 0.10 0.15 0.20 A 0 0.05 0.10 0.15 0.20 mole fraction Y mole fraction Y Fig. 14. Partial enthalpy of Y in the Cu-Mg-Y Fig. 15. Partial enthlapy of Y in the Cu-Mg-Y system for a constant Cu/Mg ratio of 9/1. system for a constant Cu/Mg ratio 3/1 V 1500 XCU/XNI - 0-5 Fig. 16. Calculated activity of Mg in the liquid Fig. 17.Calculated isopleth of the system state together with experimental data Cu-Mg-Ni together with experimen- [93Gnal] and [93Gna2], tal data [93Gnal] and [93Gna2], - 82 -
D4 Thermodynamic Assessments, Experiments and Applications Related to the Ti-Al-V(Ni)-N System Kejun Zeng and Rainer Schmid-Fetzer AG Elektronische Materialien, Technische Universität Clausthal Robert-Koch-Str. 42, D-38678 Clausthal-Zellerfeld, Germany Abstract This research forms part of the COST507-II programme on the Ti-Al-metalnonmetal system (Leading systems 5 and 6). It includes assessments and thermodynamic modeling of the Ti-N, V-N, Ti-Al-N, Ti-V-N and Al-V-N systems. Missing or inconsistent experimental data on the ternary systems have been identified. These special experiments were performed partly in the own lab and partly by our COST partners at the Universität Wien. Substantial cooperation on assessment involved especially the LTPCM-ENSEEG, Grenoble, and the University of Manchester/UMIST. Our accomplishments are documented in seven publications, given as appendices to this report. Further work on the Ti-Ni-N and Ti-Al-Ni systems is currently under way, involving thermodynamic modeling as well as the performance of key experiments, again in cooperation with the Universität Wien. This work is depicted here. It will be finalized and published with the termination of our project in May 1997. The application of the generated thermodynamic datasets to a variety of technological problems in materials design and processing of materials is emphasized. Exciting applications are given, covering the span between metallic titanium alloys and the nitride ceramics. 1. Introduction The overall system currently under study is the Ti-Al-metal-nonmetal system. This report addresses especially the quaternary Ti-Al-V-N, which is essentially described by its ternary edge systems. An extension into important subsystems of the quaternary Ti-Al-Ni-N is also given. Most of our work has been finalized to a publication level. The second section provides a concise description of the specific systems which were part of our study. The third section emphasizes applications. Details are given in an appendix composed of the matching publications which emerged from this project. Work on the systems involving Ni is in progress and will be finalized at the end of our still ongoing project within COST507-II by May 1997. 2. Data Assessment and Experimental Work 2.1 Ti-N system [96Zenl] - 83 -
A comprehensive critical assessment has been made of the experimental constitution data of the TiN system. The result differs from previous evaluation of the solid state equilibria [93Jon] and is supported also by recent experimental data. Based on the selected original experimental phase diagram and thermodynamic data from the literature, a set of thermodynamic functions for the TiN system has been chosen and the parameters were optimized by the least squares method. A stepwise optimization procedure is useful in this system and described in some detail. Four different analytical descriptions were used to model the four different types of stable phases in the TiN system: gas, liquid, solid solution phases aTi, ßTi, and 5TiN..x, and stoichiometric compounds: ε-Τί2Ν, η-Τϊ3Ν2, and ζ-Τί4Ν3. Most of the experimental information is in accordance with the modeling, especially the invariant equilibria. The identification of reliable experimental data is demonstrated to be a point of distinction in this system. No derived values like the tabulated Gibbs energy of formation were used in optimization. The phenomenon of congruent vaporization or sublimation is found to be restricted to the liquid and δ phases. A phase diagram including also this information on the gas phase equilibria is presented and used to explain the discrepancies arising in experimental data due to the composition shift of samples during evaporation. Details are given in a publication [96Zenl] (see appendix I). 2.2 V-N system [97Dul] A consistent thermodynamic data set for the V-N system is obtained by a computeraided least squares method applied to all of the experimental phase diagram and thermodynamic data available from the literature. The sublattice model V.(N,Va)a is used to model the phases: fee (a=l), bec (a=3), and hep (a=0.5). The liquid phase is described by the Redlich-Kister formula, and the gas phase is treated as an ideal gas. Special attention is paid to the modeling of the fee phase with its exceptional bulk of experimental data. This phase is first analyzed by an ideal solution, then by a regular, and finally by a subregular interaction in the nitrogen sublattice. The other solution phases are analyzed with similar modeling procedures. This step-by-step analysis procedure permits insight into reliable estimations for the parameters at each of the higher level models. Comparisons between the calculated and measured phase diagram and thermodynamic quantities show that most of the experimental information is satisfactorily accounted for by the thermodynamic calculation. Inconsistent experimental information is identified and ruled out. The thermodynamic properties of the fee and hep phases in the V-N system are compared with those in the Cr-N and Ti-N systems and related to Neumann-Kopp's rule. Details are given in a publication [97Dul] (see appendix II). 2.3 Ti-Al-N system [97Zenl, 96Zen2] The Ti-Al-N phase diagram has been assessed and a consistent set of thermodynamic functions has been developed. Three ternary line compounds, Ti-Ti3AlN0.56» χιTi2AlN0.82> x3-Ti3Al2N2, and the ternary solubilities of aTi and ßTi have been modeled. An estimated model for the metastable mutual solubilities of the nitrides - 84
AIN and TiNi_x is also given. Various combinations of different versions of the Ti-Al and Ti-N binary data have been tried. Finally, two versions of the present modeling are offered, based on two versions of the Ti-Al system. The experimental phase equilibria at 1300°C can be well reproduced. Inconsistencies are detected at lower temperatures, which are also related to the observed melting behavior of the ternary compounds. Additional key experiments have been performed by our COST partner (Universität Wien) and the results confirm that previous data refer to insufficient equilibration. These inconsistencies and the current approach to determine the Gibbs energies of the ternary phases in such a complex situation are discussed in detail. Applications of the developed thermodynamic modeling include: (1) the joining of A1N with Ti-braze, where also our own thin film experiments on the Ti/AIN interface reaction are reported, and an interpretation of the diffusion path is given; (2) diffusion barriers in semiconductor contacts, where an interpretation of the Al/TiN interface reaction is given; (3) the in-situ synthesis of Al-matrix composites, strengthened with (TiAl3+AlN) reaction products during sintering of Al+TiN powder mixtures; (4) the formation of mixed AIN-TiN hard coatings and their stabilities with respect to the pure nitrides; and (5) the nitridation of Ti-Al alloys. Details are given in a publication [97Zenl] (see appendix Ilia). Preliminary data were published in [96Zen2] (see Appendix Illb). 2.4 Ti-V-N system [97Zen2] Experimental Ti-V-N phase diagram data have been thermodynamically assessed and a consistent set of thermodynamic functions has been developed. Calculations have been performed to reveal some important features of the system. Problem areas in experimental data are indicated. The thermodynamic calculation is linked with limiting cases for solid state diffusion kinetics. This is applied to a numerical simulation of ternary solidification processes of Ti-V-N alloys and to a discussion of nitriding behavior of Ti-V alloys. Details are given in a publication manuscript [97Zen2] (see appendix IV). 2.5 Al-V-N system [97Du2,3] Thin film diffusion couples V/A1N have been used to investigate the phase relationships in the Al-V-N system. The couples were prepared by electron beam evaporation and annealed at 1000 and 1300°C for 144 and 69 hr., respectively. The phases were identified by X-ray diffraction. The Al-N-V ternary phase diagram is calculated from the thermodynamic parameters of the corresponding binary systems. It is shown that the direct extrapolation from the binary edges can describe the experimental isothermal section at 1300°C from the literature. For the isothermal section at 1000°C, however, the extrapolation does not reproduce the reported A1N+(V)+VNX phase equilibrium. The calculated Al-N-V phase diagram is also compared with the phase assemblage data resulting from our own thin film diffusion couples V/A1N. It is plausible that these assemblage data are associated with intermediate states approaching true equilibrium states. Details are given in two publications [97Du2, 97Du3] (see appendices Va and Vb) 85 -
3 Results and Discussion 3.1 Measurement of density A pygnometer was used to determine the density of the fabricated alloys. The following tables show the various values in dependence on a constant Ti and Al proportion and a variable Cu and Ni content. The determined values of the alloys range roughly between 4 and 5 g/cm3 compared with 4,5 g/cm3 for CpTi. The results show for all TixxAl-base systems a constant increasing of the density after increasing the Cuand/ or the Ni-proportion. Furthermore there is no connection between density and phase transition evident. Because of its higher Al-amount the TÌ20A1and Ti25Al-based alloys have slightly smaller values than the TilOAland Til5Al-based alloys. 3.2 Measurement of characteristic temperatures The problem is that especially the titanium will react with the A1203 crucible, if there is a liquid phase. Therefore it is possible to measure phase transitions in solid state, but no temperatures after the reaction with the crucible because the chemical reaction changes the composition of the alloys. The plots of coolingand reheating-cycle demonstrate that the composition of the alloys have changed, hence the measurement of characteristic temperatures after the reaction with the crucible is impossible. Table 5 shows the temperatures of phase transitions of the investigated alloys (heating rate: 20 K/min, atmosphere: Argon). Table 5 depicts the changing of characteristic temperatures determined for alloys in all base systems. Annealing the samples at 800°C -12 hours cause no evident changing of the characteristic temperature. 3.3 Oxidation behaviour The investigation of the oxidation behaviour indicates no significant increasing of weight after a heat treatment at 600°C for 15 hours in atmosphere. Alloys with solidus temperatures above 900°C show no systematic behaviour after heat treatment at 825°C for 15h in atmosphere. Furthermore it is obvious, that a higher Al-portion causes a less increasing of oxidation weight. The value of the oxidation mass of the annealed samples reaches ca. 50% of those, which were casted. It is determined that a phase system in an equilibrium state has a much better oxidation behaviour than a system in the casted state. The absolute oxidation values are shown in Table 6. 92 -
3.4 Electronmicroscopy (SEM) and X-ray diffraction The microstructure of nearly all investigated quaternary alloys consists of only three phases. In all investigated alloys the binary phase Ti3Al was found. Furthermore complex binary and ternary phases like TixAlyCuz and TiNi3 were determined. Phases containing all four system elements (Ti, Al, Cu, Ni) were also detected. The results of the Xray diffraction are lines from binary and ternary compounds and some unidentified lines. The lack of data from possible quaternary phases prevents the assignment. The comparison of these two methods explain that there is no agreement between the results. The figures 36 in the appendix indicate the results for some selected alloys. The element distribution is originated from a SEManalysis and the phase determination is based on the Xray diffraction. 3.5 Wetting and Joining tests The wetting and joining tests were carried out at 1100°C for the systems with a lower melting temperature e.g. Til5AlCuNi. Systems with a higher melting temperature e.g. Ti20AlCuNi, are investigated at 1250°C. In general, there are three different cases of wetting behaviour obvious: 1. a complete melting and spreading with a small wetting angle (less than 30°) 2. no complete melting but some reactions between filler metal and substrate 3. no melting The behaviour in cases two and three originate from a possible difference between melting and brazing temperature of these systems mainly Ti20Albased alloys. Due to high activity between Titanium and the ceramic crucible a thermodynamic reaction occurs, with the result that the melting temperature of some alloys could not be determined precisely. In comparison to the TÌ25A1 and Ti20Alsystems the Ti 15Al based alloys show bigger wetting angles (Fig. 7) and a welldefined diffusion zone between the Ti3Alsubstrate and the filler matrix (Fig. 8) with a thickness of ca. 30μπι. The SEManalysis reveals that this zone consists of mainly TiAlCu and TiNi compounds. The Ti25Albased systems show some erosion of the substrate (Fig.9) and a few pores within the filler matrix (Fig. 10). It appears from the SEM analysis that there is a decrease of Cu and Nidiffusion within the interface. Furthermore there is a tendency for TiAl phase formation during the braze process. 93 -
4 Appendix Tab. 1: Density of the TilOAlbased alloys 5,20 5,05 4,90 5 4,75 4,60 1 ρ [g/cm3] TMOAlxxCuyyNi / / / ^/* ~~"' / 10 15 20 «5 wt.% Cu ■10 wt.% Cu *15wt.%Cu x20wt.%Cu *25wt.%Cu 25 [wt.% Ni] Tab. 2: Density of the Ti 15 Albased alloys 5,10 < 4,95 4,80 4,65 , 4,50 ' j Ρ [Q/cm3] ι Τ¡15AI xxCu yyNi / ^ sí ■/ ^^<//^^ V _/ / ) 10 15 20 25 »5 wt.% Cu ■10 wt.% Cu A15 wt.% Cu *20wt.%Cu ♦25 wt.% Cu [wt.% Ni] - 94 -
Tab. 3: Density of the Ti20Albased alloys 4,90 < 4,75 4,60 4,45 4,30 < ι Ρ [g/cm3] y TÌ20AI xxCu yyNi ^ ^ ^ ^/^^^ < ^* 5 10 15 20 25 »5wt.% Cu »10wt.%Cu *15wt.%Cu *20wt.%Cu «25wt.%Cu [wt.% Ni] Tab. 4: Density of the Ti25Albased alloys 4,60 1 4,50 4,40 ' 4,30 ρ [g/cm3] > *s Η.Δ.Ό * 5 10 TÌ25AI xxCu yyNi ^* ^^ / 15 -•-5 wt.% Cu -»-10 wt.% Cu -*-15wt.%Cu -•-20 wt.% Cu 20 [wt.% Ni] 95 -
Tab.5: Characteristic temperatures of the investigated alloys (*) reaction with the crucible Alloy [wt.-%] TilOA15Cu5Ni TilOA15CulONi TilOAl 5Cu 15Ni TilOA15Cu20Ni TilOAl 5Cu25Ni TilOAl 5Cu30Ni TilOAl 10Cu5Ni TilOAl lOCu lONi TilOAl lOCu 15Ni TilOAl 10Cu20Ni TilOAl 10Cu25Ni TilOAl 15Cu5Ni TilOAl 15CulONi TilOAl 15Cu 15Ni TilOAl 15Cu20Ni TilOAl 20Cu 5Ni TilOAl 20Cu lONi TilOAl 25Cu5Ni TilOAl 25Cu lONi TilOAl 30Cu5Ni TÌ15A1 5CulONi Ti 15Al 5Cul5Ni Til5A15Cu20Ni Til5A15Cu25Ni Ti 15 Al 10Cu5Ni Ti 15Al lOCulONi Ti 15 Al 10Cul5Ni Ti 15 Al 10Cu20Ni Ti 15 Al 15Cu5Ni Ti 15 Al 15Cul0Ni Ti 15 Al 15Cul5Ni Til5A120Cu5Ni Til5A125Cu5Ni Til5A130Cu5Ni TÌ20A1 5Cu5Ni characteristic temperatures [°C] 821,972 835, 893, 1035 835, 890, 945 877,911,949 733.910.950 690, 888, 969 818 843,927 823, 936, 975 823,940,974, 1001 690, 850, 939 861,976 804,826,848,917,946 824,847,921,936 719,804,930 725, 865, 955 859, 899, 947, 971 868, 976 860, 989 876, 976 924,975, 1071 907,975, 1057 916,969,1054 726,908.973. 1040 843.957, 1052 973, 1016 968, 1074 956, 994, 1084 863, 938 842,987, 1020 974, 1078 859, 954, 1048 867,945,1091 865,944, 1044 750, 1033, 1059, 1188 * [°C] 1174 1192 1017 1029 944 1104 1175 970 996 998 1025 1060 1146 1023 1187 1178 1094 1118 1137 1133 1148 1129 1134 1124 1139 - 96 -
Tab.5. continued TÌ20A1 5CulONi Ti20A15Cul5N¡ TÌ20A1 5Cu20Ni TÌ20A1 5Cu25Ni Ή20Α1 10Cu5Ni TÌ20A1 lOCulONi TÌ20A1 10Cul5Ni TÌ20A1 10Cu20Ni TÌ20A1 15Cu5Ni TÌ20A1 15CulONi TÌ20A1 15Cul5Ni TÌ20A1 20Cu5Ni TÌ20A1 20CulONi TÌ20A1 25Cu5Ni TÌ25A1 5Cu5Ni TÌ25A1 5CulONi TÌ25A1 5Cul5Ni TÌ25A1 5Cu20Ni TÌ25A1 10Cu5Ni TÌ25A1 lOCulONi TÌ25A1 10Cul5Ni TÌ25A1 15Cu5Ni TÌ25A1 15CulONi TÌ25A1 20Cu5Ni 1022, 1115, 1140 1027,1108,1153 1032, 1150, 1173 1033,1162, 1207 1023,1153 1036,1148, 1184 1020,1112, 1155 1028,1172, 1214 989, 1020, 1157 1017, 1133, 1175 1019, 1110, 1162 1025, 1184 1018,1124,1174 1015,1169,1198 1087, 1218 1217 1102, 1230 1086, 1203, 1229, 1250 1234, 1245 1223,1238 1222, 1267 1075, 1226 1089, 1222, 1234 1229, 1243 1269 1166 1203 1233 1338 1266 1188 1220 1268 1203 1303 1322 1273 1269 1342 1263 1308 1340 97
Tab. 6: Oxidation behaviour for selected alloys (moltenand annealed state) *) commercial material Alloy [wt.-%] Ti 99,9 Ti 99,9 Ti 15 Al Til5A15CulONi Ti 15Al lOCulONi Ti 15 Al 10Cul5Ni Ti 15 Al 10Cu20Ni TÍ15A1 15Cul5Ni TÌ20A1 5CulONi TÌ20A1 5Cul5Ni TÌ20A1 5Cu20Ni TÌ20A1 5Cu25Ni TÌ20A1 10Cu5Ni TÌ20A1 lOCulONi Ή20Α1 10Cul5Ni TÌ20A1 10Cu20Ni TÌ20A1 15Cu5Ni TÌ20A1 15CulONi TÌ20A1 15Cul5Ni TÌ20A1 20Cu5Ni TÌ20A1 20CulONi TÌ20A1 25Cu5Ni TÌ25A1 5Cu5Ni TÌ25A1 5CulONi Ti25A15Cul5Ni TÌ25A1 5Cu20Ni TÌ25A1 10Cu5Ni TÌ25A1 lOCulONi TÌ25A1 10Cul5Ni TÌ25A1 15Cu5Ni TÌ25A1 15CulONi TÌ25A1 20Cu5Ni Heat-treatment 600°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h 825°C/15h Oxidation mass [mg/cm2] molten-state 0,135* 10,48* 2,82* 4,51 3,80 4,48 2,9 3.07 3,89 4.06 1,01 4,18 5,05 4,50 5,05 3,61 3,90 4.19 3,41 5,80 3,64 3,28 3,62 1,77 1,42 1,37 3,43 2,28 3,03 3,97 3,46 3,60 annealed-state 2,34 3,48 1,42 2,68 1,94 1,67 2,13 1,67 1,57 3.47 - 98 -
Fig. 3: ESMA and Xray diffraction of the alloy Ti 10A1 5Ni 5Cu XXOOO M5463 ZOpa·— 1 1 ? 2 y . 41,5 2,53 45,7 2,31 47,2 2,24 47,9 2,21 —™* ■ 2ΘΠ d [A] 20KU 2EB 6FE AACHEN 95 1Ti 2I.3AI Ti10AI05Ni05Cu molten by arc A B1 B2 Ti 63,8 74,5 70,7 Al 14,4 18,6 18,1 Ni 6,2 3,9 6.7 Cu 15,6 3,0 4.5 Composition of the phases [at%] Fig. 4: ESMA and Xray diffraction of the alloy Ti 15A1 20Ni lOCu Ti 15AI20NM0CU molten by arc A Β C D Ti 68,7 64,4 41,8 48,6 Al 25,1 21,7 28,4 21,6 Ni 4,4 9.0 20,5 21,5 Cu 1,7 4,9 9,3 8,1 Composition of the phases [at%] — 42.1 2,49 43,6 2.41 \ 45,4 2,32 ^"""^^r———. 48,0 2,20 49,5 2,14 50,6 2,10 2ΘΠ d [A] 1TÌ3AI 2T¡AICu/T¡4AI5Cu3 3Ti/TiAI 4TiAICu/T¡4AI5Cu3/T¡AICu2 99
Fig. 5: ESMA and Xray diffraction of the alloy Ti 20A1 20Cu lONi 1,2 v—7—y 45,0 2,34 49,8 2,13 55,7 1,92 2ΘΠ d [A) Ti20AI10Ni20Cu molten by arc A Β C Ti 63,1 38,0 36,8 Al 27,0 33,5 28.1 Ni 7,6 10,6 10,5 Cu 3,7 17,9 24,6 Composition of the phases [at%] 1TiAICu/Ti4AI5Cu3/TiAICu2 2TÌNÌ3 3T¡AICu/Ti4AI5Cu3 Fig. 6: ESMA and Xray diffraction of the alloy Ti 25A1 5Cu lONi TÍ25AI 10Ni05Cu molten by arc A Β C D Ti 60,3 60,7 37,2 40,5 Al 36,1 34,6 40,4 40,1 Ni 2.2 2,8 16,3 13.9 Cu 1,5 1,9 6,1 5,5 Composition of the phases [at%] 1TÌ3AI 2T¡AICu/T¡4AI5Cu3 3Ti/TiAI 4TiAICu/Ti4AI5Cu3/T¡AICu2 5TÍNÍ3 100
Fig.7: Wetting angle of Ti 15A1 lOCu 20Ni on Ti3Al-substrate Fig.8: SEM-micrograph of the joining zone of the sample above [at.-%l Pos.l Pos.2 Al 24,6 23,8 Ti 60 57 Nb 7 1,4 Cr ο,ι ο,ι Cu 4,8 10,3 Ni 3,8 6,8 101 -
KS 1275.1 0.14-4— —ΐΟ. 12 55 0.10 _ω o E 0.080.06 0.040.02S¡ Mg;S¡ \ AI2Cu0 0 100 200 300 400 500 600 Temperature CC KS 1295 0.14J— —·— —1 0.12 ν 0.10 o E 0.08 0.06 Η 0.04 0.02 A o o 100 200 300 400 500 Temperature C Fig.2 Phase amount (in mole%) as a function of a temperature for alloys (a)KSl275.l and(b)KSl295 108
4.2 Thermal conductivity of industrial AlSialloys Thermal conductivity results of selected AlSialloys are presented at Fig.3. For alloys with varying Sicontent and constant amount of other elements the almost linear decreasing of thermal conductivity with increasing of Si has been observed (ca. 2 Wrrï'K"1 per each weight % of Si). It is caused by the low solubility of Si in the ΑΙmatrix and its precipitation as free silicon. Copper builds with aluminium the intermetallic compound AI2Cu and its Influence on thermal conductivity is more intensive: thermal conductivity of KS1295 (4%Cu) is 16% lower then KS1275 (1%Cu). 170 G ...<>·· Q· _.^. •KS1275.1 •KS1275 •KS281.1 ■KS1295 ! ■·.·.%, > 100 150 200 250 300 350 Temperature [°C ] 450 500 Fig.3 Thermal conductivity of AlSibased industrial alloys The influence of the thermal treatment on the thermal conductivity can be observed in Fig.4. Specimens of alloy KS1295 were aged at different conditions (time and temperature) resulting in different hardness of the material. Investigations of thermal conductivity and hardness as a function of a temperature and time [96Jar/Rei] allowed us to find a correlation between the both properties for the examined alloys (Flg.5). Three curves obtained for KS1295 at 100, 250 and 300°C show approximately a linear correlation between thermal conductivity and hardness up to 150HB. The slope of the curves decreases with increasing temperature. At 100°C the curves for KS270 and KS1295 are nearly parallel in the almost whole range of the hardness. 109
145 140 Cl·. 135 130 E i. '3 o ■g 125 c o ü 1 120 115 110 ! ! ""•■o· ;... ■ ■ ***». ' ■ ' **< —* r—* :. : ·· ; —~·>:.ν .* ; A ; + j„ · r* p. -¿.;*Í¡f'·'-'-'- · τ r '"Λ ^ Τ.*Φ Τ , , , , , ! ! ! ! ! hardness Π 77ΗΒ Δ 88ΗΒ ► 106 HB • 152 HB ♦ 169 HB :&^.. : l'I.*···. ■ ' 50 100 150 200 250 300 350 temperature [°C] 400 450 500 Fig.4 Thermal conductivity of alloy KS1295 with different hardness o D "D C O O ro E 160 155 150 145 140 135 130 125 120 115 70 Ss, \ "S •^ "•»s. ··'. •^ KS1295 · 100°C 300°C 250°C V., 90 110 130 150 HB ♦'* 170 Fig.5 Correlation between thermal conductivity and hardness for KS270 and KS1295 110
4.3 AlSiZn alloys Electrical resistivity of AlSiZn alloys was measured in as cast condition. Fig.6 represents the temperature dependence of three specimens: alloys 3 and 6 with high Zn concentrations undergo at around 290°C a phase transformation. On specimens with Zn content < 20wt% (e.g. alloy 7) this transformation was not observed. This behaviour confirms the DSC investigations of Zahra [96Zah]. The electrical resistivity values at room temperature are in good agreement with results obtained during 1st round [92Hol]. E υ * a 3. > ω ι— ro υ 'ιΟ ü> 0) 26 24 22 20 18 16 14 12 10 ; ; ; ; ; ; '~' ' l^5 ; '-- ψΑ^ , —■— alloy 3 A alloy 6 —β— alloy 7 Li Ltfi^ ' ':. .^.-pf.. ■ lit?- .jrfâ?T ' ití»x»M»d*e^H^^ í^«^swwf , , , I , , , . ι , , , , ι , , , , ι , , , , ι , , , , I , , , , 50 100 150 200 250 300 temperature [°C] 350 400 450 Fig. 6 Electrical resistivity of AlSiZn alloys (specimens 3, 6, and 7) 4.4 TiAICuNi alloys Specific heat capacity measurements of quaternary Tibased alloys were carried out under atmosphere of a flowing argon. Nevertheless it was not possible to protect the specimens against the oxidation. For alloys Ti1 and Ti4 the oxidation was observed already around 500 K. Therefore the specific heat capacity was measured after the samples were heated up to 1000K. Fig.7 presents the results of the measurements on alloys Ti1, 4, and 5 as well as on the alloy TiAI10Cu5Ni25 prepared and investigated at RWTH, Aachen (project D8). DTA investigations made by D8 [95l_ug/Kot] show that for alloys with ΑΙcontent lower than 15wt% the first phase transition occurs between 700 and 900°C, while for alloys with greater ΑΙcontent above 1000°C. From our DSC curves can be seen, that the first transition depends on Ni amount also. For alloys with Nicontent greater than 15wt% a rapid increasing of specific heat capacity above 500°C was observed. m
0.52 300 400 500 Fig 600 700 800 Temperature °K 7 Specific heat capacity of quaternary Ti-AI-Cu-Ni alloys 900 1000 5 Conclusion THERSYST data collection on thermophysical properties of light metal alloys comprises within about 1500 data-sets on 158 alloys and metal matrix composites. Data taken from publications were critically analyzed respecting to their reliability. Data base is installed as a standalone version at our institute and the data can be inquired of there. Extensive investigations of thermal conductivity and hardness of industrial aluminiumsilicon-based alloys have been performed in order to find the correlation between both properties. The hardness belongs to the important parameters describing the metallurgical state of the alloy. It can be quickly and readily determined. The measurements of the thermal conductivity are expensive and time consuming. The correlation obtained makes it possible to replace in some cases the thermal conductivity investigations by simpler hardness measurements. References 85Sun/Jan 89Neu/Bra 92Hol 94Ans B.Sundman, B.Janson, J.-O.Anderson, CALPHAD 1985, 9, 153 G.Neuer, R.Brandt, G.Jaroma-Weiland, G.Pflugfelder, Int.J.Thermophys. 1989,10,749-763 S.Holtz, Untersuchungen zur Thermodynamik und Konstitution des Legierungssystems AlSiZn; Diplomarbeit, Inst, für Metallkunde und Metallphysik der Universität Clausthal, 1992 I.Ansara, COST507 Data Base for Light Metal Alloys, Round 1, CEC, Brussels (1994) - 112
94Jar/Bra G.Jaroma-Weiland, R.Brandt, G.Neuer, 1994, IKE-5-238, COST507 Thermophysical Properties of Light Metal Alloys (Institut für Kernenergetik und Energiesysteme der Universität Stuttgart) 96Jac/Spe M.H.G.Jacobs, P.J.Spencer, CALPHAD 1996, 20, 307 96Jar/Rei G.Jaroma-Weiland, P.Reipert, R.Brandt, G.Neuer, 1996, 14th ECTP, Lyon, France, 16-19 Sept. 96Lug/Köt E.Lugscheider, B.Kötzing, LMartinez, M.Koschlig, 4th Colloq.: Brazing, High Temperature Brazing and Diffusion Welding, 1995 Aachen, 27-29 July 1995 960la/San P.OIafsson, R.Sandström, A.Karlsson, Materials Sci.Forum 1996, 217-222, 981 96Zah C.Y.Zahra, A.-M.Zahra, COST507/II, Final Report 1996, CNRS,Marseille - 113 -
Dll Thermodynamic Assessment in the AlCuMgZn System P. Liang, H. L. Lukas, H. J. Seifert and F. Aldinger MaxPlanckInstitut für Metallforschung, Pulvermetallurgisches Laboratorium, Heisenbergstr. 5, D70569 Stuttgart, Germany Abstract The main part of this research is the assessment of the AlCuMgZn system, which is one of the key systems of the COST 507 project. Our contribution to it includes full optimization of the CuMgZn and AlMgZn ternary systems and revision of the thermodynamic descriptions of the AiMg, MgSi, MgZn and AlMgSi systems, us ing experimental results planned after the assessments of round I and carried out by our partners of the COST 507 project. The optimization of the AlMgZn ternary system incorporates, in addition to the available published data, the results of exper imental measurements carried out in a collaboration between CNRS at VitrysurSeine, UMIST Manchester and MPI Stuttgart. The experiments on ternary AlMgZn alloys were specifically performed to provide missing data of the ternary solubilities of the ΑΙMg and MgZn phases as well as to improve the knowledge of the extensions of the homogeneity ranges of the ternary rand ^phases. In the AlMgSi system the technically most important part, the solvus surface of the (Al) solid solution could be improved, on one hand due to a better description of the Gibbs energy of the Mg2Si binary phase, derived from enthalpy of formation and melt ing as well as from heat capacity measurements, on the other hand by DTA, dilatometrie and metallographic investigation of Alrich alloys, which were carried out by F. Som mer' s group at MPI Stuttgart. 1 Introduction The phase relationships in the quaternary AlCuMgZn system are very complex and experimentally not well established. Only a partial phase diagram in the Alrich corner was investigated [47Str]. No quaternary phase has been found. I. e. the phases encoun tered are the same as in the ternary subsystems AlCuMg, AlCuZn, AlMgZn and CuMgZn or quaternary extensions of them. In the AlCuMgZn system, several nonstoichiometric phases with the same crystal structures exist in different ternary subsystems with quaternary ranges of homogeneity (TPhase in the AlCuMg system and r Phase in the AlMgZn system, Mg^Znn and Mg2Cu5Al6, the Laves phases in the CuMgZn and AlCuMg systems). In modelling these phases, the model descriptions have to be compatible with regard to possible mix ing in the quaternary system. Therefore, the Tphase, Laves phases and Mg2Cur,AlG in the AlCuMg system have been modelled so as to allow combination with the rphase, 114
MgZn9 (C 14) and Mg2Znu in the AlMgZn system and with the Laves phases in the CuMgZn system. 2 Data Assessment Due to the experimental measurements planned and carried out after round I of COST 507, revision of some already evaluated systems was desirable. 2.1 The ΑΙMg System The central part of the ΑΙMg phase diagram was not yet well established until now. Therefore and because of interest in the formation of quasicrystals in rapidly quenched samples, the constitution between the /3(Al,3Mg2) and 7(Ali2Mgr7) phases was re investigated in a collaboration with CNRS at VitrysurSeine [97Su]. The tempera ture range of stability of the ephase is shifted to lower temperatures than reported by Schürmann [81 Sch]. The phase denoted as ζ by Schürmann seems to be only a mod ification of the 7phase inside the 7 homogeneity range. The 7phase model in the compound energy formalism was changed to satisfy the suggestion made by I. Ansara et al. [97Ans]. A new optimization was made keeping for the phases fee, liquid, hep the description of N. Sauders [90Sau]. The calculated central part of the phase diagram after this descrip tion is shown in Fig. 1. 2.2 The Cu-Mg System The parameters of the Laves phase MgCu2 were modified to accept the lattice stabilities proposed to be used for all pure elements in the fictitious state of the Laves structure [92Sau, 93Cos, 94Coe, 94Zen]. GCu:Cu_C15(T) 30Gf£R(T) = 15000 J/mol. GMgV:eM¡C15m 30GffgR(7/) = 15000 J/mol. Rerunning the least squares program BINGSS with the set of experimental data used by [91 Cou] yielded new parameters. These parameters virtually give the same phase diagram as those reported by [91 Cou] with a maximal deviation of 0.02 mol% in the solubility limits of MgCu2. 2.3 The Mg-Zn System In the MgZn system measurements of C1^0 at 28.1 mol% Zn in the temperature range 650 and 750 Κ were carried out [95Som], which confirm the magnitude of the positive 115
deviation from KoppNeumann' s rule derived by [92Aga] from the temperature depen dence of the enthalpy of mixing of the liquid, but the numerical value is refined to about 70% of that assessed by [92Aga]. Kowalski and Spencer [93Kow] in the CuZn system assumed different lattice stabilities for pure Zn in its stable state (hep with c/a = 1.856) and as hep phase with axial ratio c/a = 1.556 (cCuZn phase), which lies close to the ideal axial ratio c/a = 1.633. To get compatibility, in MgZn the Mg(hcp) solid solution had to be reformulated to show the same lattice stability for the end member pure Zn as [93Kow] used for the tCuZn phase. Furthermore the MgZn2 Laves phase should be modelled similarly as MgCu2 with an tistructure atom formation on both sublattices, although the only information for that is the vague statement: "The homogeneity range of MgZn2 is about 1 at.%" [90Mas]. The MgZn system was reoptimized using the experimentaldata file of [92Aga] and the new Cp^u'd data of [95Som] (Fig.2). The Laves phase parameters °¿Mgfzn,Mg and ^Mg.ziîrZn were estimated by trial and error to reproduce a maximum homogeneity range of 1 at.% for MgZn2. 2.4 The MgSi System The enthalpy of formation and melting as well as heat capacity of Mg2Si were measured calorimetrically by [97Feu]. The MgSi system was reoptimized using the experimental data file of [92Cha] together with the new data of Mg2Si [97Feu]. 2.5 The AlMgSi System New measurements on this system were carried out by F. Sommer's group at MPI Stuttgart. The Al corner of the AlMgSi ternary system was investigated by DTA and optical micrography. A few points of the solvus of the (Al) solid solution were precisely determined by dilatometry. The results of these experiments and the updated MgSi system together with literature data were used to redetermine a complete set of analytical descriptions of the Gibbs energies of all stable phases of the AlMgSi sys tem. The technically most important part, the solvus surface of the (Al) solid solution could be improved (Fig. 3). A publication of this reoptimization is in press [97Feu] . 2.6 The Cu-Mg-Zn System The assessment of the CuMgZn system is mainly based on the critical review of ex perimental literature provided by COST 507, Coordination group B [94Gho]. The Cu MgZn phase diagram is characterised by the formation of three Laves phases MgCu2, MgZn2 and the MgNi2type phase Mg2CuZn,3 (τ), which have large solubility ranges. 116
The three Laves phases were described by the "compoundenergyformrlism" with Cu Zn exchange, Mg(Cui_IZnr)2 and slight antistructure atom formation (Cu and Zn on the Mg sublattices, Mg on the CuZn sublattices). The Gibbs energy descriptions of the three quasibinary Laves phases were optimized using published liquidus, solidus and enthalpy of mixing data of the quasibinary system MgCu2MgZn2 [48Koe, 52Lie, 64Kin, 79Pre]. The homogeneity ranges with respect to excess or deficient Mg were interpolated between the binary end members, since no reliable experimental data for the range of deviation from the stoichiometric MgX2 are available in the ternary. The solubilities of Mg in the binary CuZn phases were interpolated from the binary CuMg fee and ZnMg hep parameters respectively, as no experimental data for these solubilities were found in literature. To satisfy the BraggWilliams description of order ing in the ternary range of the /3CuZn phase, its formulation in the compound energy formalism was extended into ternary and quaternary systems. The liquid phase was described as substitutional solid solution after the RedlichKister Muggianu formalism. No ternary parameters were introduced. Using the optimized quasibinary parameters and the estimated ternary parameters, to gether with the parameters of the binary subsystems, the ternary CuMgZn system was calculated. The results are shown in Fig. 4 to 9 and compared with the experimental values. This optimization was presented at the conference " Thermodynamics of Alloys" at Marseille, Sep. 1996 and is in preparation for publication in Calphad. 2.7 The AlMgZn System The AlMgZn ternary system is a relatively complex system which includes two ternary phases, τ and φ. The τphase has a large homogeneity region. Its ideal formula is Mg32(Zn,Al)49. It is cubic, space group Im3, 162 atoms to the unit cell [52Ber]. The ternary τphase was modelled according to its crystal structure with cubic symmetry as (Mg)26(Mg,Al)6(Al,Zn,Mg)48(Al)i in the compound energy formalism. The unit cell of the ternary (¿phase was at first time determined in a collaboration between CNRS/ONERA in Châtillon, CNRS in VitrysurSeine and MPI in Stuttgart using trans mission electron microscopy [97Don]. The unit cell of the (¿phase is orthorhombic, space group Pbc2, or Pbcm with large lattice parameters (a=897.9, b=1698.8, c=1934 pm). The (¿phase was approximated by the sublattice formula Mge(Al,Zn)s. The optimization of the AlMgZn ternary system incorporates, in addition to the avail able published data, the results of experimental measurements carried out in a collabo ration between CNRS at VitrysurSeine, UMIST at Manchester and MPI at Stuttgart. The experiments on ternary AlMgZn alloys were specifically performed to provide missing data of the ternary solubilities of the ΑΙMg and MgZn phases as well as to improve the knowledge of the extensions of the homogeneity ranges of the ternary r and (¿phases. The alloys were investigated using Xray diffraction, differential scan ning calorimetry and DTA in the composition range around the τand (¿phases. The compositions of the constituent phases were determined by Electron Probe Micro Anal ysis (ΕΡΜΑ) and Energy Dispersive Xray Spectroscopy (EDX). 117
Y,Cu5ZnB CuZn (Cu) O 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 M9 mole fraction Cu Cu Fig. 8. Isothermal section at 600 K of the Cu-Mg-Zn ternary system. 0.7 MgZn2 y 0.6 V? 0 <?° 0.4 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 M9 mole fraction Cu Cu Fig. 9. Liquidus surface of the Cu-Mg-Zn ternary system. 124 -
MggZn,, this work □ Single phase, nominal composition Φ Single phase, ΕΡΜΑ Analysis γ Two phase equilibrium, ΕΡΜΑ Δ Three phase equilibrium, ΕΡΜΑ ^ 0.6 i? (Al)" Jr 0.5 ^ 0.4 0 0.1 0.2 0.3 10.4 e· 0.5 0.6 0.7 0.8 0.9 1.0 mole fraction Mg ^ Fig. 10. Isothermal section at 608 K of the Al-Mg-Zn ternary system. 900 800 *: c 0) l_ _3 m i_ CI) u. E (1) 700-J 600 J 500- < + crJ_ 400300 mole fraction Zn Fig. 11. Vertical section of the Al-Mg-Zn ternary system at 36 mol% Mg. 125 -
0.2 0.3 0.4 Al mole fraction Zn M9Zn2 Fig. 12. Vertical section of the Al-Mg-Zn ternary system from Al to MgZn2. Temperatures in °C mole fraction Mg Fig. 13. Liquidus surface of the Al-Mg-Zn ternary system. - 126
Temperatures in °C 0.05 0.10 mole fraction Mg 0.15 Fig. 14. Solvus of the (Al) solid solution of the Al-Mg-Zn system. .030- .025c vi : . ι on u m £ .0156 .010- .0050. Mgzfr ^ •^ I 4&Λ J^ % φ ^ ^ ■ ■ ■ ι ■ ■ ■ ' 1 $ χ 7s Temperatures i n °C \^ \ * V\ \ \^ \ \. -" : —^ : ^—- τ ■ ·*■■· - 0 .01 .02 .03 .04 .05 .06 .07 .OB .09 .10 .11 .12 Mg mo 1 e f rad i on Al Fig. 15. Solvus and solidus of the (Mg) solid solution of the Al-Mg-Zn system. - 127 -
Fl Heat Capacity Data on some Al-based Alloys C.Y.Zahra and A.-M. Zahra Centre de Thermodynamique et de Microcalorimétrie, C.N.R.S. 26, rue du 141e R.I.A., F-13331 MARSEILLE CEDEX 3 Abstract Specific heat capacities were determined on 2 industrial ΑΙ-Si based alloys and on 11 ternary Al-Si-Zn alloys. The data were obtained with the help of a heat flux DSC apparatus and have a reproducibility of ± 1.5 %. In the case of the industrial alloys, the Cp-values may be interpolated linearly between 50 and 250 °C. The values for the ternary alloys lie in general below those derived from the application of the additivity rule. 1 Introduction The present study covers Crj-measurements on Al-based alloys made available from COST 507/Π partners D9 (Dr. G. Neuer) and destined for the databank THERSYST. Two groups of alloys were studied : - industrial ones already examined by partners D9. In this way, results of independent measurements may be compared; - ternary alloys for which COST partners D3 (Dr. P.J. Spencer) calculated the thermodynamic properties starting from known data on the binary systems involved. The ultimate aim is to confront their predictions with experimental data. 2 Experimental Two technical ΑΙ-Si based alloys prepared by Kolbenschmidt AG were selected for a comparative study; their chemical compositions (in mass%) and Brinell hardnesses before the Cp-measurements are given in table 1. The stability of their structural states was evaluated from DSC runs (fig. 1). Dissolution sets in above 280 °C. Table 1 : Alloy compositions KS-4 8.44% Si KS-5 13.46% Si 3.09 % Cu 3.71 % Cu 0.49% Mg 0.87% Mg 0.05% Ni 2.18% Ni 0.60% Fe 0.55% Fe 0.39 % Mn 0.45 % Mn 0.46% Zn 0.08% Zn 0.09 % Ti 0.05 % Ti 0.08 % Pb 0.01 % Pb 0.02 % Sn 0.01 % Sn HB: 99 HB: 152 The eleven ternary Al-Si-Zn alloys were cast by Vereinigte Aluminiumwerke Bonn. Their chemical compositions (in atom and mass %) are indicated in table 2 and also represented in fig. 2. From each alloy including the technical ones, 2 disks 5.9 mm in diameter and 1 mm thick were machined. - 128
Alloy 1 2 3 6 7 S 9 10 11 12 13 Al 76.46 72.73 60.82 39.13 84.83 79.89 11.73 90.19 58.51 55.84 53.28 Table 2 atom% Si 4.40 9.08 18.69 1.02 5.04 9.97 0.05 4.71 1.00 4.66 10.64 : Alloy compositions Zn 19.14 18.19 20.49 59.85 10.13 10.14 88.22 5.10 40.49 39.50 36.08 Λ1 60.01 57.61 46.81 21.13 74.02 69.57 5.20 83.93 37.11 35.70 35.10 mass% Si 3.59 7.49 14.98 0.57 4.58 9.04 0.02 4.56 0.66 3.10 7.30 Zn 36.40 34.90 38.21 78.30 21.40 21.39 94.78 11.51 62.23 61.20 57.60 Alloys 6 and 9 were kept at 360 °C for 7 h before oven cooling to room temperature; the other alloys were annealed for 5 h at 400 °C and maintained a week at 265 °C (1,2,3,11,12,13) or 240 °C (7,8) or 175 °C (10) before oven cooling. The hope of retaining stable structures at room temperature was vain, as successive DSC experiments at 5 °C/min still showed endothermal dissolution effects at low temperatures (fig. 3). As expected, the specimens 7, 8 and 10 with low Zn concentrations do not undergo a transformation around 277 °C. Due to complications arising from the existence or metastable phases and to the fact that Cpmeasurements are very time consuming and tedious, it was decided to carry them out at selected temperatures only. A PerkinElmer thermal analyser DSC 7 series 1020 was used in a constant temperature environment. The recommendations given by Höhne et al. in their book on differential scanning calorimetry [96Hoh] were observed and the following procedure was adopted in order to obtain a high accuracy of the Cpdata : The discontinuous 3step technique which measures heats was employed, as it needs no correction for thermal lag after precise temperature calibration. Steps of 20 °C were chosen at a rate of 5 °C/min. In order to check reversibility, at least three heating and cooling cycles under argon atmosphere were performed after establishment of isothermal steady state conditions. AI was used as reference sample, as the alloy and calibrant should have similar heat capacities and similar thermal conductivities. Cpdata for Al were calculated according to a formula given in Dinsdale's paper "SGTK data for pure elements" [91 Din]. A critical assessment of the thermodynamic properties of AI was also published by Desai [87Des]. Specific heat values of interest proposed by Dinsdale and Desai are confronted in table 3. They agree within 0,5 % except at 450 °C. Also are given some Cpdata on Si and Zn according to [91 Din]. Table 3 : Cp in J/gK 50 °C 100 °C 150°C 250 °C 350 °C 400 °C Al SGTE 0.91604 0.94297 0.96649 1.00954 1.05180 1.07346 Desai 0.91367 0.94186 0.96485 1.00543 1.04987 1.07546 % 0.25 0.12 0.17 0.41 0.18 0.08 Si SGTE 0.73656 0.86572 0.87721 Zn SGTE 0.39156 0.44023 0.45037 129
3. Results 3.1. Industrial Al-Si based alloys Table 4 summarises the results of Cp-determinations on alloys KS-4 and KS-5 as well as their repeatability (the maximum Öeviation is ± 0.2 %). Originally four temperatures were selected : 50, 150, 250 and 450 °C; at the latter, the alloy had been kept for an hour before the actual measurements, but results had to be discarded, as they did not give a unique value on heating and on cooling. Also indicated are the Cp-results of Jaroma-Weiland [96Jar] on the same alloys, with extrapolations to 50 °C. She used sapphire as reference material in conjunction with the normal continuous 3step procedure. The reproducibility of the two independent determinations, i.e. the deviation of the mean values of the results of an instrument from the total mean value [96 Höh] lies within ± 1.5 %. The agreement is very good, as Jaroma-Weiland et al. evaluate the inaccuracy (band) of their measurements to lie below 3 % [94Jarl. The present Cp-data may be connected linearly between 50 and 250 °C. The higher Si content of KS-5 is responsible for lower Cp-values. KS-4 Present study repeatability Jaroma-Weiland reproducibility ± ± Tal 50 °C 0.879 0.2 % 0.889 1.1 % ble 4 : Cp inJ/gK 150 °C 0.928 0.1 % 0.928 -0% 250 °C 0.9715 0.1 % 0.985 1.4% KS-5 Present study 0.852 repeatability ± 0.1 % Jaroma-Weiland 0.865 reproducibility ± 1.5 % 3.2 Al-Si-Zn alloys 0.900 0.2 % 0.901 0.2 % 0.944 0.2 % 0.954 1 % Having gained confidence in the adopted measuring technique, the Cp-data on the ternary alloys were determined on heating each specimen three times over the temperature intervals 40 - 60 and 390 - 410^C (340 - 360 °C in the case of alloys 6 and 9). The mean values (Cpexp in J/gK), often obtained on two different specimens, are given in table 5; their reproducibility is estimated to be the same as for the industrial alloys (± 1.5 %). The results are consistent with the compositions of the alloys : increasing Si and especially Zn concentrations decrease the Cp-values. For comparison, the Cpadd-data derived according to the additivity rule are indicated; the values for the pure elements are taken from table 3 [91 Din]. At 50 °C, the calculated values are more important than Cpexp. The alloys 7, 8 and 10 in the Al rich corner possess Cp-values at 400 °C which correspond to additivity within the experimental inaccuracy. - 130
Table 5 : Cp in J/gK All. 1 2 3 6 7 8 Q 10 11 12 13 4 Dy 50 Cpexp 0.737 0.716 0.706 0.517 0.814 0.797 0.428 0.848 0.607 0.596 0.601 Conclusion °C Cpadd 0.808 0.804 0.775 0.600 0.845 0.845 0.453 0.881 0.702 0.700 0.708 350 Cpexp 0.602 0.530 °C Cpadd 0.684 0.512 400 Cpexp 0.891 0.886 0.871 1.003 0.982 1.046 0.751 0.736 0.743 °C Cpadd 0.946 0.942 0.909 1.000 0.991 1.032 0.819 0.818 0.828 Specific heat capacities were measured with the help of a PerkinElmer thermal analyser DSC7 series 1020, applying the discontinuous 3step method. The repeatability of the data is of ± 0.2 %. Their reproducibility as derived from a comparison with independent determinations on some alloys is of± 1.5 %. The Cpdata obtained on two technical ΑΙSi based alloys may be connected linearly between 50 and 250 °C. Ternary AlSiZn alloys possess Cpvalues at 50 and 400 °C which lie in general below those corresponding to the NeumannKopp rule. 5 References 87Des Ρ D Desai, Internat J Thermophys, 1987, 8, 621638. 91Din A Τ Dinsdale, Calphad, 1991, 15, 317425. 95Höh G Höhne, W Hemminger, ΗJ Flammersheim, Differential Scanning Calorimetry, Springer, 1995. 96Jar G JaromaWeiland, Progress Report for Period July-December 1995, COST 507 Thermophysical Properties of Light Metal Alloys. University of Stuttgart. 94Jar G JaromaWeiland, R Brandt, G Neuer, Final Report 1994, COST 507 Thermophysical Properties of Light Metal Alloys. University of Stuttgart. 131
3.75 2.5 KS4 áo.oo ilo.oo ido. oo 2ÍõTõõ ÉõM 310.00 Temperature (CI i.ÕÕ iWM 460.00 3 75 2.5 1.25 - Temperature (C) Fig. 1 : DSC curves, at 5 °C/min, on the industrial alloys KS-4 (upper curve) and KS-5 (lower curve) - 132 -
Zn ¿Kl Si Fig. 2 : Position of the 11 alloys studied '¿■I ID ■■ 2J0.OO Temperature (C) Fig. 3 : DSC curves, at 5 °C/min, of the AlZnSi alloys 113 (n° indicated in upper left hand corner) 133
2. Experimental Procedures Four alloys have been prepared for the present study. Alloy A357 used as reference, alloy A357 with Cu addition, designated as A357+Cu, alloy A357+CU with Ag addition designated as A357+Cu+Ag and finally alloy A357+CU with Sm addition designated as A357+Cu+Sm. The chemical compositions of the alloys are given in Table 1. Conventional precision casting of the four alloys was performed by ALPHA Spain in the form of bars with a length of 20 cm and with 2.3x2.3 cm2 cross section. The bars received a solution heat treatment as follows : A357 : 538° C/22h, water quench (<10° C), A357+CU, A357+Cu+Sm : 535° C/40h, water quench (<10° C), A357+Cu+Ag : 525° C/40h, water quench (<10° C). Table 1: Chemical composition (in Alloy A357 A357+CU A357+Cu+Ag A357+Cu+Sm Si% 6,42 6,06 6,40 6,19 Fe% 0,036 0,054 0,070 0,043 Cu% 0,021 0,828 1,142 0,765 wt%) of the alloys used in the Mg% 0,60 0,53 0,77 0,55 Ni% 0,007 0,006 0,009 0,009 Zn% 0,008 0,009 0,018 0,026 Ti% 0,160 0,150 0,145 0,139 present study. Ag% - - 0,7 - Sm% - - - 0,5 Al% 92,74 92,36 91,44 92,27 Artificial ageing has been performed in all four alloys. The temperature of artificial ageing was determined by computational thermodynamics (discussed below), while the time of artificial ageing was determined from hardness measurements (also discussed below). Differential Scanning Calorimetry (DSC) was performed in the four alloys, in order to determine the melting and solidification regimes, as well as the enthalpies of fusion and solidification. These data were used to validate corresponding predictions of the computational thermodynamics program. The DSC used was a Rheometric Scientific. Specimens were pre-weighted (in the range of 3-4 mg) and heated from room temperature to 500° C with a rate of 20°C/min while from 500 to 700° C the heating rate was 1° C/min. After holding at 700° C for 5 min the specimens were cooled with rate 1° C/min. Standard Brinell hardness measurements were performed as a function of time at the artificial ageing temperature. Tensile testing was performed on an INSTRON servo-hydraulic machine. Tensile specimens were prepared from the heat-treated bars according to DIN 50125. 3. Results and Discussion 3.1 DSC Results Differential Scanning Calorimetry was carried out for the alloys. Characteristic DSC thermograms are shown in Fig.1 for alloy A357+Cu ,for alloy A357+Cu+Ag and for alloy A357+Cu+Sm with the characteristic temperatures shown in Fig.2. The DSC data are summarized in Table 2 and compared with the Thermo-Calc prediction for Tstart and Tend temperatures. It can be seen that the agreement between experimental and calculated values is good for A357+CU and becomes even better for A357+Cu+Ag and A357+Cu+Sm. - 140 -
Fig. 1. Time (x103 sec) DSC Thermograms of alloys A357+Cu, A357+Cu+Ag, A357+Cu+Sm. Fig. 2: Schematic drawing of DSC run showing the determination of characteristic temperature of 'twin peak' during melting and solidification. 141
Table 2: Heating run Cooling run Results of the DSC measurements and comparison Thermo-Calc (T.C.) predictions. Tslart [°C] Tpeakdl [°C] Tpeak(2) [°C] Tend [°C] AHf [mJ/mg] Tstart [°C] Tpeakdl [°C] Tpeak(2) [°C] Tend [°C] ΔΗ8 [mJ/mg] A357 + Cu DSC 562.83 573.24 605.52 609.06 277.67 604.62 603.63 567.52 561.83 224.19 T.C. 619.2 619.2 A357 + Cu + Ag DSC 557 570.39 610.88 613.2 294.21 610.08 607.3 567.44 557.63 138.19 T.C. 616.4 616.4 with corresponding A357 + Cu + Sm DSC 558.42 570.02 614.03 615.93 218.35 614.74 612.29 566.39 558.94 164.96 T.C. 619.9 619.9 3.2 Thermodynamic Analysis Computational thermodynamics was performed with the ThermoCalc software package [85Sun] and the recently developed COST 507 database for light alloys [94Cos]. The isopleth section (with varying Si content) for alloy A357+Cu+Sm is shown in Fig.3. Sm precipitates as the rhombohedral phase below 250° C. An important feature of this diagram is that addition of Sm lowers the liquidus temperature to 420° C compared to 540560° C in the alloys without Sm. 8. 300 E 200 100 (AlWSm^MgjSi (AI)»(Sm)*Mg,Si*Si (Alfr+Sm+MOíSi+Si+AljCije 0 2 6 8 10 12 14 Weight Percent Si 16 18 20 Fig. 3: Isopleth section for A357+Cu+Sm alloy. Scheil solidification diagrams have been constructed, with the assumptions that no diffusion is taking place in the solid and that local equilibrium exists at the solid / liquid interface during solidification. Fig. 4 depicts the fraction of liquid phase as a function of temperature during solidification for the four alloys investigated. The kinks in the curves are due to the formation of solid phases ((Al), Si, Mg2Si, Θ) during solidification. From these diagrams, the temperature range for solidification can be determined. This range is of the order of 70° C for all alloys. Fig. 5 depicts the content of the remaining liquid phase during 142
solidification. It can be seen that Si content rises in the liquid as solidification proceeds. With the assumption of limited diffusion in the solid, this diagram then depicts the resulting microsegregation. A357 A357+CU A357*Cu*Ag A357+Cu+Sm 0.2 0.4 0.6 0.8 1.0 Fraction Liquid Fig. 4: Scheil solidification diagram depicting the fraction of liquid phase as a function of temperature during solidification. Fraction Liquid Fig. 5: Scheil solidification diagram depicting the Si content in the liquid phase as a function of fraction of liquid phase during solidification. 3.3 Determination of Ageing Temperature and Time The temperature of artificial ageing was determined by considering the driving force for precipitation of the strengthening phases Mg2Si and θ during ageing. The driving force AG (in J/mole) was calculated with Thermo-Calc. Results are shown in Fig. 6 for the precipitation driving force of Mg2Si and θ as a function of Cu content in A357+Cu for two ageing temperatures 155 and 170° C. It can be seen that AG is higher for the lower ageing temperature and that increasing the Cu content has the effect of increasing the driving force for precipitation of θ-phase while decreasing the driving force for precipitation of Mg2Si. IS^CMQjSi . !70"CMgjSi · 155° C EHUIaM . 1.S 2 2.5 Weight Percent Cu Fig. 6: Driving force for precipitation ofMg2Si and θ (AI2Cu) phases as a function of temperature and Cu content in A357+Cu alloy. 143
Thermodynamic calculations have also shown that the driving force AG is not affected by Ag or Sm. For supersaturated metastable solutions, the finer precipitate size scales with the initial nucleus size which in turn depends on the magnitude of AG. The higher the driving force for precipitation, the finer the dispersion and the higher the strength. For this purpose the ageing temperature was selected to be 155° C, in contrast to 170° C conventionally used in the A357 alloy. The time of artificial ageing was determined from hardness measurements as a function of time at 155° C. Peak hardness is obtained at 25h for A357+Cu and A357+Cu+Ag, and at 20h for A357+Cu+Sm. Artificial ageing was then performed at 155° C for the times discussed just above. 3.4 Mechanical Properties The results of the tensile testing of the heat-treated alloys are summarized in Fig. 7. The elongation values are lower than expected, however they are comparable since identical casting conditions were used in all alloys. The three alloys with Cu, Ag and Sm additions possess better combinations of strength and ductility than the reference conventional A357 alloy. Since all alloys have been casted in the same way it can be deduced that the improvement in mechanical properties is due to the alloying elements and the artificial ageing followed. In comparison to the reference A357 alloy, the alloys containing Cu posses higher strength and ductility due to the formation of fine dispersion of Θphase. In addition the lower temperature ageing treatment (higher AG) produces also a finer dispersion of Mg2Si precipitate. Silver on the other hand enters the (Al) solid solution providing solid solution strengthening, while Sm forms fine particles of the rhombohedral phase. Due to the fact that the strengthening dispersions are very fine, both strength and ductility in these alloys increase. A357 A357+CU A3S7+Cu+Ag A357+Cu+Sm Fig. 7: Yield strength, ultimate tensile strength and elongation for alloys A357 (reference), A357+Cu, A357+Cu+Ag and A357+Cu+Sm. - 144
4. Conclusions Differential scanning calorimetry was conducted for alloys of the system Al-MgSi with additions of Cu, Ag and Sm in order to determine the melting and solidification regimes as well as melting and solidification enthalpies. The experimental results are in good agreement with Thermo-Calc predictions. Thermo-Calc was then used to calculate the precipitation driving forces for AI2Cu and Mg2Si as a function of temperature and composition. It was found that a lower ageing temperature provides a higher precipitation driving force. Mechanical testing of the alloys investigated really showed that the modified alloys exhibited enhanced strength/ductility combinations than the conventional A357 alloy. 5. References 83Bac Fr. -W. Bach, H. Haferkamp, Β. Li, METTAL 37, (1983),Heft 12, p.1202. 83Nag G. E. Nagel, J. P. Mouret, J. Dubruelh: "A357 Type Alloy with Improved Properties", AFS Transactions Vol. 91 Rosemont, Illinois, USA, (1983). 84Gra D. A. Granger, R. R. Sawtell, Μ. M. Kersker: "Effect of Beryllium On the Properties of A357.0 Castings", AFS Transactions Vol. 92, St. Louis, Missouri, USA, (1984). 85Sun B. Sundman, Β. Jansson, J-O. Anderson, CALPHAD, 9, (1985), p. 153. 86Mai E. Maier, G. Lang, ALUMINIUM, 62 (1986), Heft 3, p. 193. 880ze M. W. Ozelten, G. R. Turk, P. G. Porter: "Relationships Between Mechanical Properties, Composition and Microstructure of the Aluminum Casting Alloy A357-T6", TMS Proc. on Technology for Premium Quality Castings, (1988), p. 81. 89Ape D. Apelian, S. Shivkumar, G. Sigworth, AFS Transactions Vol. 97, San Antonio, Texas, USA, (1989), p. 727. 89Bru Ernst Brunhuber, AFS Transactions Vol. 97, San Antonio, Texas, USA, (1989), p. 21. 96Bri Brite-Euram Project BE4084, "Advanced Aluminum Precision Casting for Integrally Stiffened Net Shape Components", Final Report, CEC, Brussels, 1996. 94COS COST 507 Database for Light Alloys, Round I, COST 507 Program, CEC, Brussels, (1994). Acknowledgment This work has been performed within the framework of COST 507 program and was partially supported by the Research Committee of the University of Thessaly. This support is greatly appreciated. - 145
Il COST 507 - II Project group: Italy 1 Project title Thermodynamic Optimization and Evaluation of Phase Equilibria in Rare Earth Alloys R. Ferro, G. Borzone, A. Saccone, G. Cacciamani, S. Delfino, M. Giovannini, D. Maceió, N. Parodi Dipartimento di Chimica e Chimica Industriale Sezione di Chimica Inorganica e Metallurgia via Dodecaneso, 31,1-16146 Genova, Italy The Italian team is involved in activity concerning both the coordination groups A and C, as summarised in the first and second part of the following text. - 146 -
COST 507 - ROUND Π UNIT II SEZIONE DI CHIMICA GENERALE, INORGANICA E METALLURGIA (former Istituto di Chimica Generale) OF THE DIPARTIMENTO DI CHIMICA E CHIMICA INDUSTRIALE UNIVERSITÀ' DI GENOVA, ITALY 1996 REPORT Is' PART: ACTIVITY CARRIED OUT WITHIN THE FRAMEWORK OF THE COORDINATION GROUP A EXPERIMENTAL INVESTIGATIONS ON BINARY AND TERNARY RARE EARTH ALLOYS (WITH Mg or Al) Summary Experimental investigation of binary and ternary Aluminium (or Magnesium) alloys with the rare earth metals has been carried out. The investigation lines concerned: Phase equilibria determination, Thermodynamic measurements (formation enthalpy and heat capacity) In cooperation, mainly with other units, structural and magnetic properties were also studied. The different systems considered are the following: -Aluminium Systems Binary Systems: Phase equilibria investigation in the rare earth-rich regions of R-Al systems. Measurements of formation enthalpy of La-Al and Yb-Al alloys. Measurements of molar heat capacity of R-Al alloys (R=La, Ce, Pr, Nd). Ternary Systems: Enthalpies of formation and phase equilibria in selected regions of Ce-Al-Ni and Y-Al-Ni systems. -Magnesium Systems Binary Systems: Thermochemical investigation of selected Ce-Mg, Nd-Mg, Sm-Mg alloys. Ternary Systems: Structural, magnetic properties in the Ce-Mg-Y system. - 147 -
Introduction The project of our unit concerns the study of "Thermodynamics and phase equilibria in the system Al-Mg-Cu-Zn-R-(Ni. Mn) with R=rare earth. Effects of rare earth additions to Al-based or Mg-based alloys). Within the coordination group A, work is being carried out on phase equilibria determination and thermochemical measurements of specific rare earth containing subsystems. On the same systems work is also in progress in the framework of coordination group C (see the corresponding report). In the following a summary is reported of the results obtained in the investigation of Al and Mg alloys. Aluminium Alloys [96Borl] The standard molar enthalpies of formation for the different solid La-AJ and Yb-Al alloys have been measured by means of direct calorimetry. The composition and equilibrium state of samples were checked by micrographie and X-ray diffraction techniques. The following values have been obtained (kJ/mol of atoms): LaAl, AformH°= -46.0±2; LaAl2, AforalH°= -50.5.012; LaAl3, AformH°= -44.012; La3Aln, Af0rinHo= -41.012; YbAl2, Af0nnHo= -39.5.012; YbAl3, AforalH°= -32.5.012. Experimental results are discussed and compared with literature data. [96Bor2] Several applications of selected R-Al-M (R=rare earth metals, M=transition metal) alloys for their typical characteristics such as magnetic properties have been discussed in order to gain information on stable and metastable phases of these systems. A systematic study on the reactivity of the R-Al and R-Al-Ni alloys has been started. To this end calorimetrie techniques, X-ray diffraction and microscopy analyses are used. The results so far obtained for selected Ce-Ni and Ce-Al-Ni alloy compositions have been illustrated and discussed. [96Bor3] The results obtained for Ce-Al-Ni system have been presented. The thermochemical investigation was performed with an isoperibol aneroid calorimeter. The ternary phases were synthetised directly in the calorimeter, starting from a mixture of the finely powdered components enclosed in a gas-tigth inox crucible sealed by electric welding. Metallographic analysis, electron probe microanalysis (ΕΡΜΑ), X-ray diffraction analysis (powder diffraction) were used to characterised all samples, to ascertain the absence of unreacted metals and to assess their equilibrium state. The data so far obtained for characteristics sections have been reported, discussed and compared with those relevant to the involved binary alloys. [96Gam] The R-rich regions of the [Rare Earth + Al] systems (with R=La, Ce, Pr, Nd) have been recently reinvestigated by using the differential thermal analysis techniques and the melting behaviour of the (R+Al) alloys with a particular attention to the thermodynamic properties has been undertaken. In this paper the first results of the measurements of the molar heat capacity (Cp°) of some [La+Al], [Ce+Al], [Pr+AJ] and [Nd+Al] alloys are presented and discussed. - 148 -
[96Sac] The Rrich regions of the RAl systems (R=La, Ce, Pr, Nd) have been studied by using thermal analysis (DTA), metallographic analysis, quantitative electron probe microanalysis (ΕΡΜΑ) and Xray examinations (XRD). The coordinates of the Rrich eutectics, the lowering of the allotropie transformation temperatures of the rare earth metals by additions of Al and melting behaviours of the Rrich compounds have been particularly investigated. The terminal aluminium solubilities in the different rare earth metals and the relative stabilities of the catatectic and eutectoid equilibria have been discussed. Magnesium Alloys [96Cac] A systematic study of the enthalpy of formation of binary RMg alloy considering, to begin with, the systems formed with La, Ce, Pr, Nd, Sm and Yb, have been started. These systems were investigated by high temperature direct reaction drop calorimetry by using an instrument built in our laboratory and previously described. Other experimental methods employed in this research were Xray diffractometry, optical and electron microscopy. Preliminary results concerning the enthalpy of formation of some alloys at the compositions RMg, RMg2 and RMg3 in the mentioned systems have been presented and discussed. [96Fla] The structural and magnetic properties of a series of novel compounds in the ternary CeMgY system have been studied by Xray diffraction, SQUID magnetometry and XAS spectroscopy. The crystal structures of the compounds Ceo.33Yo.67Mg2 and Ceo.49Yo.5iMg4.7 have been refined by the Rietveld technique. Magnetic data are consistent with a tripositive ground state 2F¡a for the cerium atom. The XAS measurements confirm this result. Crystal field effects are apparent in the magnetic data at temperatures below about 50 Κ in the compounds with higher Ce concentration. 149
Isopleth, Al Mn, 4 wt % Si 1100 1050 1000 43 PHI (Liquidus/solidus) 93 CHA (Liquidus) 96 HAY/ROB (Liquidus/solidus) 96THO (Uquld) 96THO (Liquid + C alpha) 96THO (Uquid + beta) 51 PRA'RAY (Uquid + beta) ô 96 SIM (Liquid* C alpha) Ο Θ O 96 HAY/ROB (Liquidus/solidus, Alean san + + Q Liquid + Beta AIMiiSi O O O 950 _ 900 850 800 Liquid + C alpha_AIMnSi + lcc_A1 C alphaAIMnSi + fee A1 + diamond Si I ! L litsid 4 Mn wt % 6 Fig 1 Calculated isopleth for the AlMnSi system at 4 wt % Si. Experimental data from [43Phi], [51 Pra/Ray], [93Cha], [96Hay/Rob], [96Sim] and [96Tho] are superimposed. 252
*: ω .— 2 0) Q. ε Η 1050 1000 950 900 850 800 Isopleth, Al - Si - 4 wt % Mn IIII Liquid Θ 96ΤΗΟ (Liquid) Liquid + AI6Mn O Q 96 THO (Liquid + bela) / Uquid + beta AIMnSi Δ 96 THO (Liquid + C alpha) / ;T^ A \\ /VQ Δ Δ Δ Δ \^Η Liquid + C alpha AIMnSi ^r ——^^^ \ ^^^^^^ \ Liquid + C alpha AIMnSi + fcc \ / C alpha AIMnSi + Ice + Si ι / ι ι ι ISOT 4 6 Si wt % Fig 2 Calculated isopleth for the Al-Mn-Si system at 4 wt % Mn. Experimental data from [96Tho] are superimposed. - 253 -
Fig 3 Calculated liquidus projection for the Al-rich corner of the Al-Mn-Si system (full line). The data of Phillips [43Phi] are superimposed (dotted line). - 254 -
Al-Mn-Si 873 K 0.90 Liquid 0.80 . WAI 0.70 0.60 ΓΓΒφΠ Fig 4 Calculated partial isothermal section for the Al-rich corner of the Al-Mn-Si system at 873 K. Experimental data of Thornton et al [95Tho] are superimposed. - 255 -
Τ=823 Κ 1 r 94Kol/S¡g Δ fee + cubic alpha + ΑΙ6Μη □ fcc + cubic alpha O fcc + cubic alpha + Si 0.5 1.0 Wt % Si Fig 5 Calculated partial isothermal section for the ΑΙrich corner of the AlFeMn system at 823 K, showing the fcc solvus. Experimental data of Kolby et al [94Kol/Sig] are superimposed. 256
A Thermochemical Assessment of Data for the Al-rich Corner of the Al-Fe-Mn System, and a Revision of Data for the Al-Mn System Åke Jansson' and Tim G Chart* "Royal Institute of Technology, Stockholm, Sweden *Chart Associates, Ashford, Middlesex, UK Abstract This research forms part of the COST 507 programme on the Al-Fe-Mg-Mn-Si system (Leading System 1). It includes a revision of data for the binary Al-Mn system, based upon the original COST 507 assessment for this system [92Jan, 94Ans] and an optimisation of data for the Al-rich corner of the Al-Fe-Mn ternary system, which in addition to the available published data, incorporates the results of experimental measurements carried out during the Action by a combination of Alean International Ltd (Banbury Laboratory), Katholieke Universiteit Leuven, Pechiney, SINTEF, Universität Wien and University of Manchester/UMIST; these experiments specifically performed to provide missing data, as described in other papers of the proceedings of the COST 507 Final Workshop, eg [97Rob/Hay]. The overall system currently under study is the Al-Fe-Mg-Mn-Si system. The phase relationships in the base Al-Fe-Mn-Si system itself are very complex and not wellestablished. In particular, during Leading System 1 meetings held on the quinary system and its constituent sub-systems, it became clear that the phase diagram data for some of these key sub-systems, largely based on the early work of Phillips [43Phi] may be subject to considerable experimental uncertainty, especially data for the Al-Mn binary system and the Al-Fe-Mn and Al-Mn-Si ternary systems, due, eg, to difficulties of nucleation, the presence of metastable phases and the effects of impurities. The results of the assessments are summarised. The experimental phase equilibrium studies carried out in combination by the above groups have proven to be essential to the project. It is clear, however, that some further careful experimentation is required. 257
1 Introduction During the last two years new experimental measurements on the Al-Mn and Al-Fe-Mn systems have been carried out at Alean International Ltd (Banbury Laboratory) by Thornton and Evans, and at University of Manchester/UMIST by Hayes et al, specifically to elucidate the problems, following recommendations made COST 507 progress meetings. Work carried out as part of the COST 507 Action and under the present project, together with supplementary work has incorporated these new data, and (a) has provided a revised and improved dataset for the Al-Mn binary system which is consistent with our current understanding of experimental data for this system, and (b) has provided an updated dataset for the Al-Fe-Mn ternary system which is considered very satisfactory for present purposes. The effort has been considerable, and the problems, now largely resolved, have understandably delayed the development of the Al-Fe-Mg-Mn-Si database, also the addition of Cu. Although further work is required on the Al-Mn-Si system, which suffers related experimental uncertainties, especially the data of [43Phi], the current situation is good, very much improved compared with that of a year ago. The results of experimental measurements on the Al-Mn and Al-Fe-Mn systems carried out at Alean International during the course of the project, have proven essential to our understanding of the experimental uncertainties and to the development of the database. 2 Data Assessment 2.1 Al-Mn system The calculated phase diagram for this system is shown in Fig 1. The phase diagram data for this binary system, crucial to the present project, has caused great concern due to experimental inadequacies. Much painstaking work and careful analyses of the available information has been carried out both in the present work and by others in the field. The experimental data, particularly for equilibria involving Al4Mn, are fraught with problems due to the appearance of metastable phases, rather than the stable phase, see eg, [87Mur/McA, 87 McA/Mur, 92Jan, 96Wei/Rog]. This leads to low, nonequilibrium liquidus temperatures, which, however, are consistently reproduced by different experimental studies. For the present project the most important part of the system is the range 0 to 4 or 6 wt % Mn, although for multicomponent calculations it is important to have a good overall representation of the system. For the Al-rich region the prime data source is due to Phillips [43Phi], coupled with recent data from University of Manchester/UMIST [95Hay/Rob], now more generally available [96Wei/Rog], carried out using a sophisticated calorimetrie technique, and data of Thornton [96Tho4]. In addition Sigli [95Sig] has redetermined the solid solubilities of Mn in fcc Al. The current analysis takes into account a revision of the original [43Phi] thermal analysis data by Phillips himself, reported obscurely in an evaluation of the Al-Mn-Si system [59Phi] (see also [61Phi]), and includes information obtained on heating, rather than cooling, by [33Dix/Fin, 87McA/Mur], which avoids the problem of undercooling. The data of [96Wei/Rog] and [96Tho4] have been incorporated. Figs 2-4 show calculated phase equilibria for the Al-rich part of the system. Fig 2 shows the undercooling effects observed over the composition range 10 - 30 wt % Mn, almost - 258 -
certainly due to the appearance of metastable phases, and Fig 3 shows the corrected [59Phi] information together with a single value of [33Dix/Fin] at = 10 % Mn, and the recent data of [96Tho4]. The fcc solvus is shown in Fig 4. The current re-assessment, which is a modified and improved version of that in the COST 507 database [92Jan], represents a very satisfactory outcome. 2.2 Al-Fe-Mn System During the course of the current project new experimental information concerning the solid-state equilibria has been determined by Thornton [95Thol, 95Tho2] at 570 and 600 °C, and liquidus data have been determined at University of Manchester/ UMIST [95Hay/Ser] to corroborate the 2 wt % Mn isopleth due to [43Phi]. In addition solidstate equilibria at 550 °C by Rogl, Universität Wien [95Wei/Rog], [95Rog], have been made available to the project. These latter data are now more generally available [96Wei/Rog]. An optimisation has been carried out whereby substantial consistency between the lowtemperature (570 and 600 °C) solid-state data of [95Thol, 95Tho2], the solid-state data of Rogl (550 °C) [96Wei/Rog], and the liquidus data of Phillips [43Phi] (allowing for expected undercooling, see Section 2.1), and Hayes [95Hay/Ser], has been achieved. The revised data for the Al-Mn system have been incorporated. Figs 5 and 6 show calculated partial isothermal sections for 873 and 843 K, with the results of [95Thol, 95Tho2] superimposed. Fig 7 shows a calculated partial isothermal section for 823 K, with the results of [96Wei/Rog] superimposed. Fig 8 shows the calculated isopleth for a constant 2 wt % Mn. Experimental data of [43Phi], [95Hay/Ser] and [96Thol, 96Tho2, 96Tho3] are superimposed. Figs 9 and 10 show the calculated fcc solvus at 823 and 873 K respectively, with experimental data made available by SINTEF [95Sim/Kol] and due to Hamerton [96Ham] superimposed. Figs 5-10 show that substantial agreement between the experimental data and the current assessment has been achieved. Fig 8 indicates the undercooling experienced by [43Phi] (these experimental values were not corrected by Phillips), and probably reflects the true situation. It has not yet been possible to obtain agreement with all of the data due to Thornton without unreasonable stabilisation of the Al6(Mn,Fe) phase and an accompanying unreasonable raising of the liquidus curve. The data for 3, 3.25 and 3.5 wt % Fe are currently under review by Thornton; with respect to Fig 8 it may be that very small amounts of Al]3(Fe,Mn)4 were present, but extremely difficult to detect. 3 Conclusions and Recommendations for Further Work The current thermodynamic data assessments for the binary system Al-Mn and the ternary system Al-Fe-Mn represents a significant milestone for the overall project, and a considerable advancement in our knowledge of phase diagram information for these two systems. Subject to further limited confirmatory experimental studies, the assessment of data for these two systems can now be regarded as substantially complete. Completion of the above paves the way for completion of data for the Al-Mn-Si system, both experimental studies (in progress at Alean International Ltd, Banbury - 259 -
Laboratory, as part of this COST Action) and the subsequent assessment of data. Furthermore, major emphasis can now be placed on the assessment of data for Cucontaining systems, very important to the overall project. The results of experimental measurements on the Al-Mn and Al-Fe-Mn systems carried out by the combination of Alean International Ltd, Katholieke Universiteit Leuven, Pechiney, SINTEF, Universität Wien and University of Manchester/UMIST, during the course of the project have proven essential to our understanding of the experimental problems associated with these systems, and to the development of the database. 4 Acknowledgements The authors gratefully acknowledge support from NUTEK, Swedish National Board for Industrial and Technical Development (Åke Jansson), National Physical Laboratory, UK (Tim G Chart) and the COST Secretariat, DGXII. Valuable discussions with Dr Paul V Evans, Martin C Thornton and Dr Richard Hamerton (Alean International Ltd), Dr Christophe Sigli (Pechiney) and Prof Peter Rogl (Universität Wien) are gratefully acknowledged. 5 References 33Dix/Fin E H Dix, W L Fink, L A Willey, Trans AIME, 1933, 104, 335352. 40Fah/Hof E Fahrenhorst, W Hofmann, Metallwirtschaft, 1940,19, 891 -893. 43Phi H W L Phillips, J Inst Metals, 1943, 69, 275-350. 45But/Hum E Butchers, W Hume-Rothery, J Inst Metals, 1945, 71, 78-91. 530bi/Hat I Obinata, E Hata, K Yamaji, Jpn J Inst Met, 1953, 17, 496-501. 58Liv/Voz V A Livanov, V M Vozdvizhenskii, Tr Mosk Aviats Tekhnol Inst, 1958, (31) 84-99. 59Phi H W L Phillips, Annotated Equilibrium Diagrams of Some Aluminium Alloys Systems, The Institute of Metals, London, (1959), Monograph and Report Series, No 25, pp 77-83. 61Phi H W L Phillips, Equilibrium Diagrams of Aluminium Alloys Systems, The Aluminium Development Association, London, (1961), pp 91-96. 64Dri/Kad M E Drits, E S Kadaner, E M Padezhnova, N R Bochvar, Russ J Inorg Chem, 1964, 9, 759-762. 71God/Kos T Gódecke, W Koster, Z Metallkde, 1971, 62, 727-732. 87Mur/McA J L Murray, A J McAlister, R J Schaefer, L A Bendersky, F S Biancaniello, D L Moffat, Metall Trans A, 1987, 18A, 385-392. 87McA/Mur A J McAlister, J L Murray, Bull Alloy Phase Diag, 1987, 8, 438447. 91Min/Yam Y Minamino, T Yamane, H Araki, N Takeuchi, Y-S Kang, Y Miyamoto, T Okamoto, Met Trans A, 1991, 22A, 783-786. 92Jan A Jansson, Metall Trans A, 1992, 23A, 2953-2962. 94Ans I Ansara, Thermochemical Database for Light Metal Alloys, COST 507, Concerted Action on Materials Sciences, European Commission, DG XII, Luxembourg, 1995 95Hay/Rob F H Hayes, J A J Robinson, COST 507 II Project, Report on Smith Thermal Analysis Studies of Al-Mn (-Fe) Alloys, Manchester - 260 -
Materials Science Centre, November 1994. 95Hay/Ser F H Hayes, A Semeeis, J A J Robinson, COST 507 II Project, Report on Smith Calorimeter Experiments in the Ternary Al-FeMn System, Manchester Materials Science Centre, April 1995. 95Rog Ρ Rogl, Universität Wien, Private Communication, July 1995. 95Sig C Sigli, CALPHAD XXIV Conference, Kyoto, Japan, May 1995. 95Sim/Kol C J Simensen, Ρ Kolby, A contribution to the ternary phase diagram of aluminiummanganeseiron, SINTEF Report, STF24 F95085, December 1995. 95Thol M C Thornton, Alean International Ltd, Banbury Laboratory, Private Communication, March 1995. 95Tho2 M C Thornton, Alean International Ltd, Banbury Laboratory, Private Communication, April 1995. 95Wei/Rog F Weitzer, Ρ Rogl, Deutsche Gesellschaft für Materialkunde meeting, Bochum, June 1995. 96Ham R G Hamerton, Alean International Ltd, Banbury Laboratory, Private Communication, June 1996. 96Thol M C Thornton, Alean International Ltd, Banbury Laboratory, Private Communication, January 1996. 96Tho2 M C Thornton, Alean International Ltd, Banbury Laboratory, Private Communication, April 1996. 96Tho3 M C Thornton, Alean International Ltd, Banbury Laboratory, Private Communication, May 1996. 96Tho4 M C Thornton, Alean International Ltd, Banbury Laboratory, Private Communication, September 1996. 96Wei/Rog F Weitzer, Ρ Rogl, F H Hayes, J A J Robinson, Phase Relations in the Aluminium-Rich Part of the System: Aluminium-IronManganese, in "Werkstoff Woche '96, Materialwissenschaftliche Grundlagen", ed. F Aldinger and H Mughrabi, 1997, Deutsche Gesellschaft für Materialkunde, pp 185190. 97Rob/Hay J A J Robinson, F H Hayes, A Semeeis, F Weitzer, Ρ Rogl, Smith Thermal Analysis Studies of the Al-rich Portion of the Al-Mn and Al-Mn-X (X = Fe, Si) Systems, Proceedings of COST 507 Final Workshop, Vaals, The Netherlands, March 1997, European Commission, DG XII, Luxembourg. 261
Al-Fe-Mn 823 K 0.40, AI13(Fe,Mn)4/ 0.30 Fe 0.10, 0.20 to ——— Al W 0.90 y / βΛ 0.80 C "Al \V N/S^ \ 0.70 0.60 AI,,Mn4 Fig 7 Calculated partial isothermal section for the Al-rich comer of the Al-Fe-Mn system at 823 K. Experimental data of Weitzer et al [96Wei/Rog] are superimposed. 268 -
850 co co _l LU O I LU rr I < tr LU o_ 800 750 700 ω 650 600 AlFeMn section at 2 wt% Mn _J I I I L 43Phl: Δ L/L+AI6(Mn,Fe) □ L/L+AI13(Fe,Mn)„ φ 95Hay/Ser X 96Tho1,2,3 Uq+AI6(Mn,Fe) V 96Tho1,2,3 Uq+AI6(Mn,Fe)+AI,3(Fe,Mn)4 WEIGHT PERCENT FE Fig 8 Calculated isopleth for the AlFeMn system at 2 wt % Mn. Experimental data from [43Phi], [95hay/Ser] and [96Thol,96Tho2,96Tho3] are superimposed. 269
0.1 0.2 0.3 0.4 WEIGHT PERCENT MN Fig 9 Calculated partial isothermal section for the ΑΙ-rich comer of the Al-Fe-Mn system at 823 K, showing the fcc solvus. Experimental data of Simensen and Kolby [95Sim/Kol] are superimposed. 270 -
0.50 0.45 0.40-1 LU £ 0.35 LU 0.30ü LU 0.25 H CL H1 0.20 - 2 0.15 LU ^ 0.10 0.05 H 0 WEIGHT PERCENT MN Fig 10 Calculated partial isothermal section for the Al-rich comer of the Al-Fe-Mn system at 873 K, showing the fcc solvus and three-phase region fcc + Al13(Fe,Mn)4 + AI6(Mn,Fe). Experimental data of Hamerton [96Ham] are super-imposed (diamond = three-phase). - 271 -
Θ CORDIS The Community Research and Development Information Service Your European R&D Information Source CORDIS represents a central source of information crucial for any organisation - be it industry, small and medium-sized enterprises, research organisations or universities - wishing to participate in the exploitation of research results, participate in EU funded science and technology programmes and/or seek partnerships. CORDIS makes information available to the public through a collection of databases. The databases cover research programmes and projects from their preparatory stages through to their execution and final publication of results. A daily news service provides up-to-date information on EU research activities including calls for proposals, events, publications and tenders as well as progress and results of research and development programmes. A partner search facility allows users to register their own details on the database as well as search for potential partners. Other databases cover Commission documents, contact information and relevant publications as well as acronyms and abbreviations. By becoming a user of CORDIS you have the possibility to: • Identify opportunities to manufacture and market new products Identify partnerships for research and development Identify major players in research projects • Review research completed and in progress in areas of your interest The databases - nine in total - are accessible on-line free of charge. As a user-friendly aid for on-line searching, Watch-CORDIS, a Windows-based interface, is available on request. The databases are also available on a CD-ROM. The current databases are: News (English, German and French version) - Results - Partners - Projects - Programmes - Publications - Acronyms - Comdocuments - Contacts CORDIS on World Wide Web The CORDIS service was extended in September 1994 to include the CORDIS World Wide Web (WWW) server on Internet. This service provides information on CORDIS and the CORDIS databases, various software products, which can be downloaded (including the above mentioned Watch-CORDIS) and the possibility of downloading full text documents including the work programmes and information packages for all the research programmes in the Fourth Framework and calls for proposals. The CORDIS WWW service can be accessed on the Internet using browser software (e.g. Netscape) and the address is: http://www.cordis.lu/ The CORDIS News database can be accessed through the WWW. Contact details f or further Information If you would like further information on the CORDIS services, publications and products, please contact the CORDIS Help Desk : CORDIS Customer Service Telephone: +352-441012-2240 B.P. 2373 Fax: +352-441012-2248 L-1023 Luxembourg E-mail: [email protected]u WWW: http://www.cordis.lu/ 273
European Commission EUR 18171 —COST 507 Definition of thermochemical and thermophysical properties to provide a database for the development of new light alloys (Volume 1) Luxembourg: Office for Official Publications of the European Communities 1998 — 273 pp. — 17.6 χ 25 cm Volume 1: ISBN 92-828-3901-X Volumes 1 to 3: ISBN 92-828-3900-1