Open Collections will be undergoing maintenance Monday June 8th, 2020 11:00 – 13:00 PT. No downtime is expected, but site performance may be temporarily impacted.

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Microstructure evolution during extrusion of AA3xxx aluminum alloys Yuanyuan, Geng 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2011_spring_geng_yuanyuan.pdf [ 5.35MB ]
Metadata
JSON: 24-1.0071658.json
JSON-LD: 24-1.0071658-ld.json
RDF/XML (Pretty): 24-1.0071658-rdf.xml
RDF/JSON: 24-1.0071658-rdf.json
Turtle: 24-1.0071658-turtle.txt
N-Triples: 24-1.0071658-rdf-ntriples.txt
Original Record: 24-1.0071658-source.json
Full Text
24-1.0071658-fulltext.txt
Citation
24-1.0071658.ris

Full Text

 MICROSTRUCTURE EVOLUTION DURING EXTRUSION OF AA3xxx ALUMINUM ALLOYS by YUANYUAN GENG B.Eng., ZHENGZHOU UNIVERSITY, 2008 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Materials Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   March 2011 © Yuanyuan Geng, 2011 ii  Abstract The through process microstructural evolution of aluminum alloys AA3003 (1.27wt% Mn) and AA3102 (0.26wt% Mn) during high temperature extrusion was investigated. The Direct Chill(DC) cast microstructure of both alloys showed interdendritic  eutectic regions consisting of rod/plate like constituent particles in the α-Al. Prior to extrusion, the as-cast materials were homogenized for three different conditions, i.e. 500°C for 8h, 550°C for 8h and 600°C for 24h. Optical Microscopy was used to examine the behavior of constituent particles and dispersoids during homogenization. Back Scattered Scanning Electron Microscopy and Image Analysis with Clemex were conducted to study the evolution of the size, aspect ratio, area fraction and number density of constituent particles. Moreover, to investigate the evolution of constituent particles during homogenization at 600oC, samples heated to and soaked at 600oC for different times were studied. It was found that different mechanisms of microstructure evolution occurred during homogenization including the breaking up, growth and coarsening of particles. Extrusion trials were conducted on a laboratory scale extrusion press located at Rio Tinto Alcan’s Avidia R&D Center in Jonquiere, Quebec. Various AA3003 and AA3102 DC billets with different homogenization treatments were extruded at 400oC and 550oC with an extrusion ratio of 130. The extrudates were examined in different orientations to study the through-thickness microstructure profile. Grain structures were revealed by Polarized Light Optical Microscope and Electron Back Scattered Diffraction. Back Scattered Scanning Electron Microscopy and Image iii  Analysis were conducted on the extrudates to quantify the characteristics of the extruded material. It was found that the homogenization treatments have a significant effect on the as-extruded microstructure. In particular, the presence of dispersoids suggested significant pinning effects on the recrystallization behavior. In addition, Gleeble tests were conducted on a low iron AA3102 alloy using a Gleeble 3500 Thermo-mechanical Simulator at strain rates of 0.1s-1, 1s-1 and 10s-1 and deformation temperature of 400oC, 500oC, and 600oC. The yield stress, work hardening and flow stress results were fit into a physically based flow stress model of Kocks and Chen. It was found that the constituent particles had a minimal effect on the flow stress exponent in the model. iv  Table of Contents Abstract .......................................................................................................................................................... ii Table of Contents....................................................................................................................................... iv List of Tables ............................................................................................................................................. vii List of Figures .......................................................................................................................................... viii List of Symbols ......................................................................................................................................... xii Acknowledgements ............................................................................................................................... xiii Chapter 1 Introduction ............................................................................................................................ 1 1.1 Homogenization .............................................................................................................................. 2 1.2 Extrusion ............................................................................................................................................ 3 1.2.1 Extrusion process .................................................................................................................. 4 1.2.2 Process parameters for extrusion .................................................................................. 4 1.3 Microstructure evolution ............................................................................................................ 5 1.4 Project objective ............................................................................................................................. 6 Chapter 2 Literature Review ................................................................................................................. 7 2.1 Characteristics of the AA3xxx aluminum system ............................................................. 7 2.1.1 Phase diagrams ....................................................................................................................... 8 2.1.2 Microstructure ..................................................................................................................... 10 2.2 The evolution of second phase particles during homogenization ......................... 11 2.2.1 Constituent particles ......................................................................................................... 12 2.2.2 Dispersoids ............................................................................................................................ 15 2.3 Recrystallization .......................................................................................................................... 17 2.3.1 Basic concepts in recrystallization .............................................................................. 17 2.3.2 Effect of second-phase particles ................................................................................... 19 2.3.3 Effect of thermo-mechanical factors ........................................................................... 21 2.3.4 Dynamic recrystallization in hot-worked aluminum .......................................... 22 2.3.5 Microstructure development during extrusion ..................................................... 25 2.4 Constitutive modeling ............................................................................................................... 27 2.4.1 Deformation mechanisms of aluminum alloy ......................................................... 28 2.4.2 Kocks and Chen Model ...................................................................................................... 29 v  Chapter 3 Scope and Objectives ........................................................................................................ 31 3.1 Scope ................................................................................................................................................. 31 3.2 Objectives ........................................................................................................................................ 32 Chapter 4 Experimental Methodology ........................................................................................... 33 4.1 Initial materials ............................................................................................................................ 33 4.2 Homogenization ........................................................................................................................... 34 4.3 Extrusion trials ............................................................................................................................. 36 4.4 Microstructure characterization .......................................................................................... 39 4.4.1 Sample preparation............................................................................................................ 39 4.4.1.1 Billet sample .................................................................................................................. 39 4.4.1.2 Extruded sample ......................................................................................................... 40 4.4.2 Metallographic preparation procedure ..................................................................... 42 4.4.3 Microscopy ............................................................................................................................. 44 4.5 Image analysis .............................................................................................................................. 45 4.6 Compression tests ....................................................................................................................... 46 Chapter 5 Experimental Results ....................................................................................................... 49 5.1 Microstructure characteristics in as-cast billets ........................................................... 49 5.1.1 As-cast AA3003 .................................................................................................................... 49 5.1.2 As-cast AA3102 .................................................................................................................... 51 5.2 Microstructure evolution during homogenization ....................................................... 52 5.2.1 Evolution of constituent particles and segregation in the matrix ................. 52 5.2.2 Observations on dispersoids for different homogenization temperature . 57 5.2.3 Optical micrographs of as-homogenized AA3102 ................................................ 62 5.3 Microstructure characteristics of extruded Samples .................................................. 63 5.3.1 As-extruded AA3003 ......................................................................................................... 63 5.3.1.1 Constituent particles ................................................................................................. 63 5.3.1.2 Grain structure ............................................................................................................. 69 5.3.2 As–extruded AA3102 ........................................................................................................ 73 5.4 Effect of constituent particles on flow stress .................................................................. 76 5.4.1 SEM micrograph of low iron AA3102......................................................................... 76 5.4.2 Stress-Strain curves ........................................................................................................... 77 Chapter 6 Discussion ............................................................................................................................. 80 6.1 Evolution of second phase particles during homogenization and extrusion .... 80 vi  6.1.1 Constituent particles ......................................................................................................... 80 6.1.2 Dispersoids ............................................................................................................................ 85 6.2 Effect of extrusion conditions on as-extruded microstructure ............................... 85 6.2.1 Time scheme for extrusion and the through thickness variation in processing ......................................................................................................................................... 86 6.2.2 Coarse surface grain region ............................................................................................ 87 6.2.3 Effect of homogenization treatment ........................................................................... 89 6.3 Physically-based flow stress model .................................................................................... 93 6.3.1 The model ............................................................................................................................... 93 6.3.2 Constitutive modeling of low iron AA3102 ............................................................. 94 6.3.3 Discussion on the fit parameters for the flow stress model ............................. 96 Chapter 7 Summary and Conclusions ............................................................................................ 99 7.1 Summary ......................................................................................................................................... 99 7.2 Future work ................................................................................................................................ 102 References ............................................................................................................................................... 103 Appendices .............................................................................................................................................. 113 Appendix A Repeatability of compression results and determination of yield stress and flow stress .................................................................................................................................. 113 A.1 Repeatability of stress-strain curves .......................................................................... 113 A.2 Determination of yield stress and steady state flow stress .............................. 114 Appendix B Summary of compression tests results ......................................................... 116 Appendix C Image analysis method and sensitivity test for constituent particles quantification .................................................................................................................................... 117 C.1 Clemex Routine ..................................................................................................................... 117 C.2 Sensitivity test on threshold value ............................................................................... 118 Appendix D Calculation of Burger’s vector and Young’s Modulus ............................. 120 D.1 Temperature dependent Young’s Modulus ............................................................. 120 D.2 Temperature dependent Burger’s vector ................................................................. 121 Appendix E Strain and strain rate calculation .................................................................... 123  vii  List of Tables  Table 2-1 Wrought aluminum alloy composition limits, reproduced with permission from (Woldman and Frick 2000)   ......................................................................................................... 8Table 2-2 Crystal structure and density of the main phases in Al-Mn-Si alloy   ................ 9Table 4-1 Chemical composition of experimental alloys (wt %)   ....................................... 33Table 4-2 Parameters for the extrusion trials   ............................................................................ 37Table 4-3 Metallographic procedure for grinding and polishing   ....................................... 43Table 6-1  The measured fraction of constituent particles and equilibrium phase fraction from Thermo-Calc   ................................................................................................................. 83Table 6-2  The SEM data and simulated (Du et al. 2011) volume fraction of the constituent particles   .............................................................................................................................. 84Table 6-3 Comparison of fit parameters for low iron AA3102 alloys with those from previous AA3003 and AA3102 alloys (Kubiak 2009); note the data from Kubiak is represented by the shadowed boxes.   ............................................................................................. 97Table B-1 Flow stress and yield stress value (MPa) for as cast low iron AA3102   ... 116Table B-2 Stress value (MPa) for homogenized low iron AA3102   ................................. 116Table C-3 Selected range of threshold value for one micrograph from homogenized sample (600oC, 48h)   ........................................................................................................................... 119Table C-4 Selected range of threshold value for all micrographs from homogenized sample (600oC, 48h)   ........................................................................................................................... 119Table D-5 Parameters meaning and references for Modulus equations   ..................... 120Table D-6 Temperature dependent Shear Modulus and Young’s Modulus values   .. 121Table D-7 Parameters meaning and references for Burger’s vector equations   ........ 121Table D-8 Coefficient of thermal expansion data from (Fickett 1971)   ......................... 122Table D-9 Calculated temperature dependent lattice parameter and Burger’s vector values   ......................................................................................................................................................... 122 viii  List of Figures  Figure 1-1  Industrial manufacturing process of AA3xxx for heat exchanger application   ..................................................................................................................................................... 2Figure 1-2 Schematic of direct extrusion press, reproduced with permission from (Davis 1993)   ................................................................................................................................................. 4Figure 2-1 Sections of Al-Mn-Si phase diagram with a) 1.5%Mn; b) 0.5% Mn, reproduced with permission from (Belov et al. 2005)   .............................................................. 9Figure 2-2 Area fraction and number density of constituent particles during homogenization at 600oC, reproduced with permission from (Li and Arnberg 2003)   ......................................................................................................................................................................... 13Figure 2-3 Effect of deformation on PSN, reproduced with permission from (Humphreys and Hatherly 2004)   ..................................................................................................... 20Figure 2-4 Polarized Light Optical Micrograph of partly extruded 6060 billet, reproduced with permission from (Kayser et al. 2010)   ......................................................... 26Figure 4-1 Recirculation air furnace   ............................................................................................... 35Figure 4-2 Homogenization treatment profiles   ......................................................................... 35Figure 4-3 Extrusion trials in ARDC 1) extruded strip exits here, 2) billet position for extrusion, 3) mechanical billet loading system, 4) ram   ......................................................... 36Figure 4-4 Water quench unit located at die exit 5) Extrudate being water quenched   ......................................................................................................................................................................... 37Figure 4-5 Principle extrusion variables   ...................................................................................... 38Figure 4-6  I-beam Extrudate a) dimension b) actual extrudate   ........................................ 38Figure 4-7 Position indication of billet samples   ........................................................................ 40Figure 4-8 Dimension of a) AD sample with the observation plane indicated as the red color b) tapered sample with the observation plane indicated as the blue color   ......................................................................................................................................................................... 40Figure 4-9 Dimension of Tapered samples   .................................................................................. 41 ix  Figure 4-10 AA3102 as-cast, and extruded at 400oC a) schematic representation of the combined 3-D profile, b) picture from AD sample, and c) picture from tapered sample   .......................................................................................................................................................... 42Figure 4-11 Picture of anodizing equipment   .............................................................................. 44Figure 4-12 Example of manually plotting on as-cast AA3003   ........................................... 46Figure 4-13 Gleeble Thermal-mechanical Machine for compression tests   ................... 47Figure 4-14 Jaw setup 1) jaws, 2) thermocouples, 3) sample, 4) load cell   .................... 47Figure 5-1 As-cast AA3003 a) centre of the billet,  b) half radius section of the billet,  c) outer section of the billet   ................................................................................................................ 50Figure 5-2 Grain structure of as-cast AA3003   ............................................................................ 50Figure 5-3 Higher magnification of Back Scattered SEM micrographs showing views of constituent particles in as-cast AA3003 a) x2000 b) x7000   ........................................... 51Figure 5-4 OM micrograph of as-cast AA3102   ........................................................................... 52Figure 5-5 SEM micrograph of as-cast AA3102   ......................................................................... 52Figure 5-6 As-cast AA3003: a)backscattered electron image, b)x-ray map for Mn, c)x-ray map for Fe, d)x-ray map for Si   ........................................................................................... 53Figure 5-7 AA3003 homogenized at 600oC for 24h: a)backscattered electron image, b)x-ray map for Mn, c)x-ray map for Fe, d)x-ray map for Si   ................................................ 53Figure 5-8 Backscattered electron micrographs for samples with a heating rate of 150oC/h and quenched from : a) 200oC, b)350oC, c) 400oC, d) 500oC and e) 550oC   . 54Figure 5-9 Backscattered electron micrographs for samples soaked at 600oC for : a)0min, b)10min, c)20min, d)40min,e)1h,f)12h,g)24h, and f)48h   ................................... 55Figure 5-10 Size and aspect ratio of constituent particles in AA3003 soaked at 600oC   ......................................................................................................................................................................... 56Figure 5-11 Area fractions and number density of constituent particles in AA3003 during homogenization at 600oC   ..................................................................................................... 57Figure 5-12 a) OM micrographs of AA3003, 500oC, 8h homogenized   ............................. 58Figure 5-13 Backscattered electron micrographs of homogenized AA3003 for a) 500oC, 8h b) 550oC, 8h and c) 600oC, 24h   .................................................................................... 60 x  Figure 5-14 Area fractions of constituent particles in AA3003   .......................................... 61Figure 5-15 Number density and circular diameter of constituent particles in AA3003 after different homogenization treatments   ............................................................... 61Figure 5-16 OM Micrographs of Homogenized AA3102 a) 500oC, 8h b) 550oC, 8h c) 600oC, 24h   .................................................................................................................................................. 62Figure 5-17 SEM micrographs of constituent particles in AA3003, extruded at 400oC a) as-cast, and homogenized for b) 500oC, 8h c) 550oC, 8h d) 600oC, 24h   ..................... 64Figure 5-18 SEM view of constituent particles in AD sample from AA3003, 600oC and 24h homogenized and extruded at 400oC a) surface b) center of the extrudate   65Figure 5-19 the number density variation of the constituent particles of the entire AD plane (AA3003, 600oC and 24h homogenized and extruded at 400oC)   ................... 65Figure 5-20 Average circular diameters of constituent particles a) before extrusion b)after extrusion at 400oC   ................................................................................................................... 66Figure 5-21 Area fraction of constituent particles before and after extrusion (400oC)   ......................................................................................................................................................................... 67Figure 5-22 Number density of constituent particles a) before extrusion b)after extrusion at 400oC   .................................................................................................................................. 68Figure 5-23 Microstructure of AA3003 I-Beam extruded at 400oC   .................................. 70Figure 5-24 EBSD mapping of 400oC as-extruded (24h@600oChomogenized)   .......... 71Figure 5-25 Microstructure of AA3003 I-Beam extruded at 550oC   .................................. 72Figure 5-26 Microstructure of AA3102 I-Beam extruded at 400oC   .................................. 74Figure 5-27 Microstructure of AA3102 I-Beam extruded at 550oC   .................................. 75Figure 5-28 SEM micrograph of as-cast low iron AA3102   .................................................... 76Figure 5-29 Stress-strain curves from as-cast low iron AA3102 for selected compression conditions   ....................................................................................................................... 77Figure 5-30 Stress-strain curves from 24h@600oC homogenized low iron AA3102 for selected compression conditions   .............................................................................................. 78Figure 5-31 Selected stress-strain curves of as-cast and homogenized low iron AA3102   ........................................................................................................................................................ 78 xi  Figure 5-32 Effect of homogenization on flow stress and yield stress of low iron AA3102   ........................................................................................................................................................ 79Figure 6-1 Particles with different contrast from SEM micrograph of a 550oC, 8h as-homogenized AA3003 sample   ........................................................................................................... 81Figure 6-2 Time scheme for extrusion   ........................................................................................... 86Figure 6-3 Polarized Optical Micrographs of AD view of AA3003 and AA3102 extrudates   ................................................................................................................................................... 88Figure 6-4 Grain size for AA3003 alloy with different homogenization treatments and extruded at 400oC (μm)   ............................................................................................................... 91Figure 6-5 Grain size for AA3003 alloy with different homogenization treatments and extruded at 550oC (μm)   ............................................................................................................... 91Figure 6-6 Grain size for AA3102 alloy with different homogenization treatments and extruded at 400oC(μm)   ................................................................................................................ 92Figure 6-7 Grain size for AA3102 alloy with different homogenization treatments and extruded at 550oC (μm)   ............................................................................................................... 92Figure 6-8 Constitutive modeling for low iron AA3102, as-cast   ........................................ 95Figure 6-9 Constitutive modeling for low iron AA3102, homogenized 8h@500oC   ... 95Figure 6-10 Constitutive modeling for low iron AA3102, homogenized 24h@600oC   ......................................................................................................................................................................... 96Figure A-1 Repeated tests on AA3003 alloys   ........................................................................... 113Figure A-2 Repeated tests on as-cast low iron AA3102 alloys   ......................................... 114Figure A-3 Determination of steady state flow stress for a 600oC, 24h homogenized low iron AA3102, at compression condition of 600oC, and 10s-1   .................................... 115Figure C-4 SEM image being processed and threshold peak   ............................................ 118 xii  List of Symbols  Latin Symbols Description      Units  a0   lattice parameter at room temperature  m aT   lattice parameter at a given temperature   m a1   fitting parameter     - a2   fitting parameter     °C-1 A   pre-exponential diffusion constant   s-1 AC    circular area of the billet    m2 Af    area of the extruded strip flange    m2 Aw    area of the extruded strip web   m2 AE    area of the extruded strip profile   m2 b   temperature dependent Burger’s vector  m D   diffusion coefficient     m2s-1 E   Young’s modulus     GPa ER   Extrusion ratio     - k   Boltzman constant     Pa·m3K-1 n   stress exponent     - Q   activation energy   J·mol-1 QD   activation enthalpy for self-diffusion  J·mol-1 R   universal gas constant    J·K-1mol-1 T   temperature      K Tm   melting temperature     K v   dislocation velocity     m·s-1 𝑉𝐸                                     extrusion speed                                                                mm·s-1 Z                                       Zener Hollomon parameter                                          s-1  Greek Symbols Description      Units α               CTE at the given temperature   K-1 αsd                                                     Semi-die angle                                                                   o ε   true strain      - 𝜀𝑒𝑓𝑓                                effective strain rate                                                          s-1 ε    strain rate      s-1 μ   temperature dependent shear modulus  GPa μ0   shear modulus at room temperature   GPa σ   true stress      MPa σflow    steady state flow stress    MPa σy   yield stress      MPa  xiii  Acknowledgements First and foremost, I would like to express my sincere gratitude to my supervisor, Prof. Warren Poole and my co-supervisor, Dr. Qiang Du for their continuous support of my Master study and research. I am grateful to Prof. Warren Poole for his patience, motivation, and immense knowledge in assisting my research. I would like to thank Dr. Du for his many professional suggestions and help all along during my study. Gratitude is extended to the faculty and staff of the Department of Materials Engineering, in particular Dr. Fateh Fazeli for his help with Gleeble machine and Ms. Mary Fletcher for her kindly assistance in Scanning Electron Microscopy. I would like to thank Dr. Nick Parson for his valuable suggestions and thank Rio Tinto Alcan for funding this project. My sincere thanks also go to the fellow colleagues working on this project, Prof. Mary Wells, Payman Babaghorbani and Yahya Mahmoodkhani. I extend my thanks to my colleagues and my friends, especially Leo Colley for his professional suggestions on a wide range of topics, e.g. should I learn to ski or snowboard. I am indebted to my parents and my brother for their love and support. Also I would like to show my special gratitude to my aunt for her endless love, encouragement and care throughout my life. 1  Chapter 1 Introduction Aluminum is one of the most important and widely used metals in the transportation, construction, packaging and electrical sectors. Since its introduction in 1901, the aluminum industry has assumed an important role in the Canadian economy. The Canadian aluminum industry now produces 2.8 million tons per year representing approximately 10% of the worldwide aluminum production (Aluminum Association of Canada 2010). In October 2007, world leading mining organization Rio Tinto acquired Canada’s existing largest aluminum company Alcan. Alcan integrated with Rio Tinto's existing aluminum business, resulting in Rio Tinto Alcan, the aluminum industry's new global leader (Rio Tinto Alcan 2010). The materials which will be investigated in this study are the commercial AA3xxx aluminum alloys, which is one of the value-added products that Rio Tinto Alcan is seeking to expand its applications. AA3xxx aluminum alloys are aluminum alloys with manganese as the major alloying element and silicon and iron as minor alloying elements. AA3xxx alloys are non-heat treatable alloys but may be strengthened by cold work. They have moderate mechanical strength, high ductility and excellent corrosion resistance (Davis 1993; Kaufman 2000), and are widely used for sheet and extruded products. Among all 3xxx series alloys, AA3102 and AA3003 are the two most popular alloys for heat exchanger components. As shown in Figure 1.1, the industrial manufacturing process for AA3xxx for heat exchanger application includes casting, homogenization, hot extrusion, cold work and brazing.  2  One of the key requirements of the extruded product is a fine grain size to improve mechanical properties. The challenge for the manufacturers is therefore to produce extruded materials at the highest possible productivity with the required quality.  Figure 1-1  Industrial manufacturing process of AA3xxx for heat exchanger application 1.1 Homogenization For industrial practice homogenization of as-cast billets is done prior to extrusion. It is a three-step process which includes heating up, soaking at high temperature, and subsequent cooling. Normally the main purposes of homogenization are 1) to remove particles and microsegregation that will give areas with low melting temperatures, in order to avoid tearing during extrusion; 2) to round off hard particles with sharp edges, which would give poor ductility and holes in thin walled products; 3) to form dispersoids that would control grain size during  3  extrusion; 4) to obtain a uniform distribution of alloying elements in solid solution before extrusion (Dons 2001). From the microstructure perspective, when the heating starts at room temperature, the alloying elements are supersaturated in a solid solution. Thus, nucleation and precipitation of second phase particles tend to occur. As the alloy is heated, these second phase particles may subsequently redissolve at higher temperature or during the long soaking stage. Further, the second phase particles could behave as nuclei or exert pinning for future microstructure evolution (e.g. recrystallization or grain growth) during high temperature deformation and its subsequent annealing. The industrial homogenization treatments selected for the different aluminum alloys are related to the extrusion technology of the material in question. Commercial heat treatments for AlMn alloys are typically 6 hours to 12 hours of homogenization at temperature of 600oC to 620oC (Laue and Stenger 1981), but there is interest to examine lower temperatures and shorter times as well, considering the energy consumption involved. 1.2 Extrusion  Extrusion is a deformation process that is commonly used to produce various shapes of sections, tubes, wires or strips.   4  1.2.1 Extrusion process Direct non-lubricant hot extrusion as studied in this work has a typical sequence of operations as shown in Figure 1.2(Davis 1993): A heated billet and a dummy block are loaded into the container; the billet is extruded by the force of the ram being pushed against it through the die. During extrusion, a thin shell of material may be left on the container walls. Extrusion is halted in order to leave a thin disk of material (butt) in the container; the container is separated from the die, the extruded section with the butt, and the dummy block; the butt is sheared off; the shear die, the container and the ram are returned to their initial loading positions (Saha 2000).  Figure 1-2 Schematic of direct extrusion press, reproduced with permission from (Davis 1993) 1.2.2 Process parameters for extrusion Process parameters for extrusion process include the homogenization treatments prior to extrusion, extrusion ratio/extrusion speed, and extrusion  5  temperature. The total work in extrusion can be divided into a number of components: the work of deformation required for homogeneous reduction of the area from the initial to final cross-section, redundant or internal frictional work which is required to overcome internal friction and finally, the work to overcome the frictional resistance at the container-billet interface and die bearing-billet material interface (Saha 1998). The heat in the extrusion is generated due to the plastic deformation and friction between the billet and the die. The surface layer of an extrudate undergoes more severe shear deformation than the regions in the center. The friction at the die bearing interface and the increased localized deformation at the surface both contribute to higher temperature rise on the surface compared to the center. As a result, there is a thickness-dependent thermo-mechanical profile along the surface to center from the extrudates, which then leads to a thickness-dependent microstructure. 1.3 Microstructure evolution  Each manufacturing process from casting to extrusion may influence the final microstructure of the as-extruded material, and thus the final property of the products after cold work and brazing. The as-extruded microstructure depends on the starting chemistry, homogenization treatments and also the through profile thermo-mechanical history experienced during the extrusion. Although there is a wide application for extruded AA3xxx alloys, investigations on the microstructure of the extrudates are very limited (Parson and Ramanan 2008).  6  1.4 Project objective In 2006, Department of Materials Engineering at the University of British Columbia started a project with Alcan with the objective of developing a process-model for AA3xxx aluminum alloys to improve the properties and quality of these alloys.  The ongoing research project includes the solidification & homogenization model for AA3xxx alloys by Du et al. (2011), a constitutive model to describe the steady state flow stress during hot compression for AA3xxx by Kubiak (2009), current work on through process microstructure characterization, an FEM model for thermo-mechanical simulation during extrusion and an incorporated microstructure model by Mahmoodkhani et al. (2010) and the recrystallization kinetics study for cold worked extrudates by Babaghorbani. Knowledge of through-process microstructure change could help us to understand the microstructure evolution in AA3xxx alloys during extrusion, and also to provide experimental validation to simulation results, and hopefully to optimize the manufacturing process. Current work studies the microstructure evolution, which covers the evolution of grain structure and second-phase particles of AA3003 and AA3102 in as-cast, as-homogenized and as-extruded states to evaluate the effect of homogenization, extrusion conditions and alloying element content on microstructures of the extrudates, and it also covers a continuation work on previous constitutive study by Kubiak(2010) on A3xxx alloys during hot compression.   7  Chapter 2 Literature Review To understand the microstructure development and its effect on recrystallization behavior through the extrusion process, the microstructure evolution at each stage, from as-cast to homogenization and through extrusion, needs to be characterized. Meanwhile, the interaction between the processes need to be clarified, for example the evolution of the second phase particles during homogenization is important itself for understanding materials’ response to the homogenization procedure and also important for understanding their effect on the recrystallization in the subsequent extrusion process.  This chapter reviews the relevant literature to provide the background for the current study. First, the phase diagram and microstructure characteristics of AA3xxx aluminum alloys are introduced. The microstructural evolution of constituent particles and dispersoids during the homogenization heat treatment are then reviewed. The static and dynamic recrystallization mechanisms relevant to extrusion of aluminum alloys are reviewed. Finally, the constitutive behavior of 3xxx during hot compression is summarized with a particular emphasis on the Kocks and Chen constitutive model. 2.1 Characteristics of the AA3xxx aluminum system The phase diagrams and the microstructure characteristics of 3xxx aluminum alloy are presented in this section to summarize current knowledge on the phase component in these alloys.   8  2.1.1 Phase diagrams The nominal chemical compositions and its alloying limits for commercial 3003 and 3102 alloys are shown in table 2.1(Woldman and Frick 2000). The 3003 and 3102 alloys used in this study have chemical compositions of Al-1.27Mn-0.54Fe-0.1Si and Al-0.26Mn-0.5Fe-0.1Si, respectively. The precipitation characteristics of AA3xxx alloys are affected considerably by the presence of iron and silicon. Studies have shown that not only do iron and silicon have low solubilities in aluminum (0.01~0.15wt% for silicon and approximately 0.04wt% for iron, from 20oC to 600oC), but they also have an effect on reducing the solubility of manganese in aluminum (Davis 1993).  Table 2-1 Wrought aluminum alloy composition limits, reproduced with permission from (Woldman and Frick 2000) Alloy Chemistry(wt%) Limits Al Si Fe Mn Cu Ti Each Total 3003 0.6 0.7 1.0-1.5 0.05-0.20  0.05 0.15 Bal. 3102 0.4 0.7 0.05-0.40 0.10 0.10 0.05 0.15 Bal.  The equilibrium phase diagram is an essential tool in determining solidification and melting temperature, and also in understanding the solid state phases and the temperature range over which they occur. Phase diagram sections based on Al-Mn-Si system are displayed in Figure 2.1(Belov et al. 2005) to help understand the Al-Mn-Fe-Si quaternary system. It can be seen from Figure 2.1 that the Al-Mn-Si alloy is characterized by main Al phase with Al6Mn and α-Al15(Mn)3Si2  9  (see table 2.2, data taken from (Belov et al. 2005)). The melting point of the system is approximately 650oC (for <<1% Si). Higher silicon content favors the formation of α-Al15(Mn)3Si2 phase and reduces the melting point of the system.   Figure 2-1 Sections of Al-Mn-Si phase diagram with a) 1.5%Mn; b) 0.5% Mn, reproduced with permission from (Belov et al. 2005)  Table 2-2 Crystal structure and density of the main phases in Al-Mn-Si alloy Phases Crystal structure Lattice Parameters Density α-Al15(Mn)3Si2 Cubic a=1.260nm 3.55g/cm3 Al6Mn Orthorhombic a=0.650nm b=0.755nm c=0.866nm 3.18g/cm3 Al Fcc a=0.405nm 2.70g/cm3  10  Moving from a ternary system to quaternary alloy, there is a debate regarding the presence or absence of a quaternary (AlFeMnSi) phase in Al-Fe-Mn-Si system (Belov et al. 2005). To simplify the problem current work accepts the assumption that iron and manganese are substitutional for each other in either phases(Alexander and Greer 2002; Li and Arnberg 2003). However, there is a complication since the addition of iron further decreases the solubility of manganese in solid solution(Davis 1993), and increases the probability of forming second phases.  Therefore, depending on the chemistry of the alloy, the following eutectic and peritectic reactions, as shown in Eq.2.1~2.4 may occur during cooling. 𝐿 → 𝐴𝑙(𝑠) + 𝐴𝑙6(𝐹𝑒,𝑀𝑛)                                                      (2.1) 𝐿 + 𝐴𝑙6(𝐹𝑒,𝑀𝑛) → 𝛼⎯ 𝐴𝑙(𝐹𝑒,𝑀𝑛)𝑆𝑖                                  (2.2) 𝐿 → 𝐴𝑙(𝑠) + 𝛼⎯𝐴𝑙(𝐹𝑒,𝑀𝑛)𝑆𝑖                                               (2.3) 𝐴𝑙6(𝐹𝑒,𝑀𝑛) + 𝑆𝑖 →  𝛼⎯𝐴𝑙(𝐹𝑒,𝑀𝑛)𝑆𝑖 + 𝐴𝑙(𝑠)                (2.4) 2.1.2 Microstructure  During solidification of commercial ingots, some of the manganese forms Al6(Mn,Fe) and cubic α-Al(Fe,Mn)Si by reactions in Eq.2.1~2.3. Those rod/plate-like primary particles, referred to as “constituent particles” in this work, make up the inter-dendritic eutectic network in the cast microstructure. DC-cast AA3xxx alloy is composed by equiaxed grains with a second dendrite arm spacing of 7-25μm dependant on the distance from the surface and solidification rate(Tibballs et al. 2001). Constituent particles with the Al6(Mn,Fe) phase predominates in the as-cast  11  structure, and  heat treatment results in its transformation to α-Al(Fe,Mn)Si if silicon is high enough(Hatch 1984; Alexander and Greer 2002; Warmuzek and Sieniawski 2004).  Furthermore during solidification manganese atoms are segregated in the primary Al dendrite, which results in a supersaturated solid solution after solidification. During the subsequent heating, this supersaturated solid solution decomposes via the precipitation of secondary particles, which are called “dispersoids”, with the α-Al(Mn,Fe)Si or Al6(Mn,Fe) phase (Hatch 1984; Lodgaard and Ryum 2000; Li and Arnberg 2003; Lacaze et al. 2005).  The evolution of both constituent particles and dispersoids during homogenization is discussed in the next section. 2.2 The evolution of second phase particles during homogenization  Homogenization is a three-step process that includes heating up to selected homogenization temperature, soaking for certain period of time, and subsequent cooling down to room temperature. During homogenization, the constituent particles grow while simultaneously the dispersoids nucleate, grow and coarsen. Since the second phase particles in the material may have significant effect on the final mechanical property and subsequent processing behavior (Vatne et al. 1997; Humphreys 1977; Vandermeer and Juul Jensen 2001; Zhang et al. 2001; Jones and Humphreys 2003; Vandermeer and Juul Jensen 2003; Humphreys and Hatherly 2004; Vandermeer 2005; Zhu et al. 2009), the evolution of both constituent particles and dispersoids in 3xxx series alloys in response to a homogenization process is  12  reviewed. 2.2.1 Constituent particles The constituent particles are originally eutectic particles that distributed in the interdendritic regions during solidification. Upon the homogenization treatment, morphology of those constituent particles changes with a possible phase transformation of Al6(Mn,Fe) phase to α-Al(Mn,Fe)Si.  Different methods have been adopted by researchers to investigate constituent particles in this series of aluminum alloy (Hatch 1984; Alexander and Greer 2002; Li and Arnberg 2003; Warmuzek and Sieniawski 2004; Dehmas et al. 2005). Etching with 10% phosphoric acid is considered to be able to reveal the Al6(Mn,Fe) in light color and α-Al(Mn,Fe)Si in dark color under Optical Microscope, but the accuracy of the measurement may be influenced by etching time(Hatch 1984). Since the α-phase has a brighter contrast than Al6(Mn,Fe) in Backscattered Electron imaging mode(due to its higher average atomic number), BE-SEM(Back-scattered Scanning Electron Microscopy)  and image analysis based on these micrographs has been adopted to differentiate these two phases(Li and Arnberg 2003). Electron Microprobe Analysis technique is also used to study the phase transformation of constituent particles (Alexander and Greer 2002).  Li and Arnberg (2003) observed the evolution of the constituent particles in a 3003 alloy (Al-1.15Mn-0.58Fe-0.2Si, wt%) by SEM. During heating, the eutectic phase starts to break up at about 400oC. The number density reaches its peak at 560oC when most of the network breaks down to smaller particles. While soaking at  13  600oC, the area fraction of constituent particles slowly increases until it reaches 4.0% after long time homogenization (as shown in Figure 2.2). An average diameter of 1~1.2μm for constituent particles is reported in their work. The fraction of α-Al(Mn,Fe)Si transformed from Al6(Mn,Fe) increases with temperature and homogenization time.  Figure 2-2 Area fraction and number density of constituent particles during homogenization at 600oC, reproduced with permission from (Li and Arnberg 2003) The SEM investigation by Dehmas et al. (2005) on an Al-1.18Mn-0.58Fe-0.2Si-0.08Cu alloy reported an area fraction of 3.0±0.2% for the as cast material, and slightly changed 3.1±0.2% for material heated up to 600oC. The transformation of Al6(Mn,Fe) to α-Al(Mn,Fe)Si starts around 400oC , and fraction of α  phase reaches around 30% of total constituent particles at 600oC. In contrast, the work by Alexander and Greer (2002) on Al-0.5Fe-1.0Mn-0.2Si alloy using SEM shows that 5% of the constituent particles in as cast material are α-Al(Mn,Fe) and this fraction  14  increases to 60% when heated up to 500oC.  With small variations, these experimental findings have suggested that the Al6(Mn,Fe) would gradually transform into the α-phase during heating and soaking. Long-term heat treatments appear to permit manganese to substitute for iron in the α-Al(Mn,Fe)Si phase indicated by an increased Fe/Mn ratio (Alexander and Greer 2002; Li and Arnberg 2003; Warmuzek and Sieniawski 2004; Dehmas et al. 2005). The difference between data reported by different researchers may be related to the chemistry, the accuracy of the measurement and difference in heating rate. The Al6(Mn,Fe) to α-Al(Mn,Fe)Si transformation has been extensively studied in can body stock material AA3004(Al0.25Cu0.7Fe1Mn0.3Si) since the α-Al(Mn,Fe)Si phase is more favorable for galling resistance and therefore it is desirable to control the transformation of the Al6(Mn,Fe) phase to α-Al(Mn,Fe)Si. This transformation requires silicon and furthermore the weight percentage of (Fe+Mn) in α-Al(Mn,Fe)Si is higher than in Al6(Mn,Fe) which indicates manganese is also in demand. There have been discussions on whether the diffusion of silicon or the diffusion of manganese is the rate-controlling step of this phase transformation (Alexander and Greer 200). Upon using Energy-filtered TEM to track the local chemistry changes during the transformation, Alexander and Greer (2002) reported that the Al6(Mn,Fe) particles are lath-shaped and α-Al(Mn,Fe)Si nucleates at various points on their surfaces. The growth of α-Al(Mn,Fe)Si phase preserves the local contents of iron and manganese but requires both intake of silicon and diffusion of manganese from phase boundary and solid solution, leading to the conclusion that the transformation  15  rate is not simply related to either solute. As to the composition of the α-Al(Mn,Fe)Si phase, the relevant literature indicates its stoichiometry as either α-Al12(Mn,Fe)3Si (Hatch 1984; Rouns 1998; Alexander and Greer 2002; Warmuzek and Sieniawski 2004) or α-Al15(Mn,Fe)3Si1- 2(Mathew, Ramachandran et al. 1984; Martins et al. 2009). These differences are probably due to Al atoms substituting for five Si atoms in the α unit cell(Alexander and Greer 2002). 2.2.2 Dispersoids The nucleation, growth and coarsening of the small dispersoids in the intra-granular region also occur simultaneously with the evolution of constituent particles during homogenization. Optical Microscopy using 0.5% hydrofluoric acid etched sample is commonly used to qualitatively evaluate the spatial distribution of dispersoids(Hatch 1984; Kubiak 2009). Resistivity measurement is considered as a simple method for indirect estimation of the dispersoids fraction since the electrical conductivity can be related to the manganese in solid solution (Li and Arnberg 2003; Kubiak 2009). Transmission Electron Microscopy is found useful to quantitatively study the size, morphology, phase type of dispersoids in this alloy (Li and Arnberg 2003; Dehmas et al. 2005), although determination of number density or volume fraction is experimentally challenging. Previous work by Kubiak (2009) and Li and Arnberg (2003) reported similar conductivity evolution trend during heating and soaking of AA3003 alloys at 600oC. Work by Mathew et al. (1984) also reported the same trend of conductivity change  16  during soaking period based on Al-1Mn alloy. At approximately 300oC, the supersaturated solution starts to decompose and the dispersoids precipitate. After reaching a peak at about 500oC, the conductivity drops due to dissolution of dispersoids back into the solid solution at this temperature (since the solubility of the manganese in solid solution increases with temperature). With soaking at 600oC, the conductivity increases again slowly due to losing manganese solute by coarsening of the constituent particles and remains constant after long time soaking. TEM observations by Li and Arnberg (2003) showed an equiaxed shape of dispersoids for low temperature (300oC) and elongated plate-like dispersoids for high temperature (600oC), which can grow to sizes of several nm to around 200 nm with number density ranged from 0 to 1500/μm3. Another important feature in the microstructure is the Precipitate Free Zone (PFZ) that appears in 3xxx alloy during homogenization treatment. The origin of PFZ formation is not completely understood (Gandin and Jacot 2007); however, it may be important for further microstructure evolution (Morris and Duggan 1978; Boubertakh et al. 2009). Both optical observation (Kubiak 2009) and TEM investigation (Li and Arnberg 2003) suggest that smallest PFZ area fraction is first found in the area adjacent to the interdendritic region when heated to 400oC~500oC. At 600oC with increasing soaking time PFZ increases towards center of interdendritic region and reaches its maximum when the dispersoids are mostly dissolved.  17  2.3 Recrystallization The recrystallization behavior in hot worked aluminum is complicated. There is a long history of debate on the recrystallization mechanisms (static/dynamic, continuous/discontinuous dynamic recrystallization) in aluminum during hot deformation (Sidor et al. ; Humphreys 1977; Yamagata 1992; Doherty et al. 1997; Ponge et al. 1997; Vandermeer and Juul Jensen 1998; Vandermeer and Juul Jensen 2003; Humphreys and Hatherly 2004; Kaibyshev et al. 2005; Kassner and Barrabes 2005; Liu and Morris 2005; Vandermeer 2005; Kayser et al. 2010). To simplify the problem, concepts behind classic static recrystallization and related recovery process are introduced in section 2.3.1. Factors affecting recrystallization, which may be relevant to this work, are briefly reviewed in section 2.2.2 and 2.2.3. Dynamic recrystallization, especially the Geometric Dynamic Recrystallization (GRX) in hot worked aluminum alloys is reviewed in section 2.3.4. Review on investigations of material flow, recrystallization and coarse surface grain structure in aluminum extrusion is given in section 2.3.5. 2.3.1 Basic concepts in recrystallization  Recrystallization and recovery are two competing softening processes in deformed materials. Recrystallization is a process that involves the formation of a new grain structure in a deformed material by the formation and migration of high angle grain boundaries driven by stored energy from the deformation. High angle grain boundaries in aluminum are those with misorientation angles above 15o (Heilmann et al. 1983). Recovery can be defined as all annealing processes occurring  18  in deformed materials without the migration of high angle grain boundaries (i.e. dislocations rearrange to lower the energy through formation of sub-grains).  Nucleation of recrystallization is typically found when high angle grain boundaries are produced, i.e. regions with lattice misorientation, deformation zones and shear bands. The new grains continue to grow with the rate depending on the grain boundary mobility. Grain growth, also known as grain coarsening normally appears as normal grain growth or abnormal grain growth(Doherty, Hughes et al. 1997; Humphreys and Hatherly 2004). The former one is a uniform and continuous process with narrow size distribution while in the latter one a few grains grow under consumption the matrix of smaller grains and a bimodal size distribution is formed (Hillert 1965; Humphreys and Hatherly 2004).  It is also worth emphasizing that the concepts of continuous recrystallization and discontinuous recrystallization are purely phenomenological, referring to spatial and temporary heterogeneity of the microstructural evolution whether there is clear recrystallized region or none during the process (Jazaeri and Humphreys 2004). Investigation on the deformed state is considered critical for studying recrystallization. Successful cases have been reported by many researchers (Hasegawa and Kocks 1979; Hansen and Bay 1981; Ferry and Munroe 1995; Gourdet and Montheillet 2000; Huang and Humphreys 2000; Holm et al. 2003; Hu et al. 2008) in which characterization of subgrain structure in the deformed state are carried out to investigate the recrystallization mechanisms. TEM, SEM, X-ray  19  diffraction (XRD), Optical Microscopy and especially Electron Backscattered Diffraction (EBSD) technique are found useful for this study. To summarize, all the factors that would have effect on the nucleation or growth process will influence the recrystallization behavior in the material (Humphreys and Hatherly 2004). 2.3.2 Effect of second-phase particles The effects of small particles, large particles and bimodal particle distribution on recrystallization kinetics are discussed respectively below. Small Particles/Close-Spaced Cluster (Zener Pinning): The retardation effect that small particles exert on grain boundaries motion is called Zener pinning (Humphreys and Hatherly 2004). This effect has been investigated by many researchers through experimental and modeling approaches (Vatne et al. 1997; Gladman 1966; Humphreys 1977; Gladman 1992; Kim and Kishi 1999; Zhang et al. 2001; Jones and Humphreys 2003; Humphreys and Hatherly 2004; Liu and Morris 2005; Harun et al. 2006). For a random distribution of spherical, mono-sized particles, the pinning pressure/force 𝑃𝑧 exerted on the boundary is given by Eq. 2.5(Humphreys and Hatherly 2004). 𝑃𝑧 = 3𝛾𝐹𝑉2𝑟                                            (2.5) Where γ is the specific boundary energy, Fv  is the particle volume fraction and r is the particle radius. The growth of recrystallized grain is controlled by the  20  competition of the two opposing pressures, i.e. driving pressure for growth and Zener Pinning, therefore the sum of these two parts(positive or negative) is sometimes used to roughly determine whether recrystallization would occur or not(Humphreys and Hatherly 2004). It is clearly shown in Eq.2.5 that the ratio of Fv/r is a critical parameter. In fact, the volume fraction, morphology, and size/spatial distribution of the particles also affect the recrystallization (Hillert 1984; Nes et al. 1985; Li and Easterling 1990; Liu and Patterson 1993; Humphreys and Ardakani 1996; Rabkin 1998; Harun, Holm 2006; Gottstein and Shvindlerman 2010). Large Particles (Particle Stimulated Nucleation):  Figure 2-3 Effect of deformation on PSN, reproduced with permission from (Humphreys and Hatherly 2004) The deformation zones formed around large particles (1~2μm) can act as nucleation sites for recrystallization because they have a high dislocation density  21  and large lattice orientation (Doherty et al. 1997). This phenomenon is termed as particle stimulated nucleation (PSN). The critical particle parameter for PSN to occur is temperature and strain dependent. As shown in Figure 2.3 (Humphreys and Hatherly 2004), during high temperature deformation, the PSN may not occur (i.e. deformation zone around it doesn’t form) and the process is nucleation limited while during low temperature deformation it’s growth-limited. Bimodal Distribution: When there is a bimodal size distribution of particles in the material, both the pinning effect and PSN may operate concerning both nucleation and recrystallization kinetics (Song and Rettenmayr 2007). This is usually the case in the heat treatment for supersaturated solid solution where the amount of second phase particle varies with time and the particles are inhomogeneously distributed. Al-Si alloy provides a microstructure with two distinguished distributions of second phase particles. Chan and Humphreys (Chan and Humphreys 1984) showed that although there were PSN by very large particles (>5μm) in this alloy, but the recrystallization kinetics were determined by the significant pinning effect of small particles. 2.3.3 Effect of thermo-mechanical factors Recrystallization is a thermally activated process; therefore temperature greatly influences the recrystallization rate. Heating rate can also be important and rapid heating may result in less time for recovery to occur (since recovery and  22  recrystallization are two competing processes to reduce the free energy of the system), which means there is a lower reduction in internal energy from recovery and hence recrystallization is more rapid (Doherty et al. 1997). The amount of strain affects the rate of recrystallization by changing the amount of stored energy. Increasing the strain rate of deformation prior to recrystallization increases the rate of recrystallization. The mode of deformation, including uniaxial compression, channel die, plane strain compression/tension, and torsion and rolling, also influences recrystallization through strain path variation. Work has been carried out on strain path effect on recrystallization and they are primarily focused on cold worked aluminum (Davenport and Higginson 2000). It is noted by Embury et al.(1992) that multi-axial deformation shows slower recrystallization kinetic then uniaxial deformation for the equivalent strain, while with equivalent stress, the kinetics are identical. He concludes that the change in strain path promoted the dynamic recovery and hence lower driving pressure for recrystallization. 2.3.4 Dynamic recrystallization in hot-worked aluminum Dynamic recrystallization refers to the occurrence of recrystallization during hot deformation (McQueen 1988; Humphreys and Hatherly 2004). When dynamic recrystallization occurs, both nucleation and growth take place while the strain is applied rather than afterwards as part of a separate heat treatment. Some authors have used the term 'postdynamic' or 'metadynamic' to describe recrystallization that occurs during the cooling phase of a hot-working process or between successive  23  passes(Fernández et al. 2003), which is still considered as static recrystallization in this study due to the fact that there is no concurrent deformation. Compared to the systematic study on static recrystallization, dynamic recrystallization is not as well understood due to the inherent complexity of the phenomenon and the difficulty for experimental studies. It is, however, believed that the physical mechanisms accounted for dynamic recrystallization are similar in many respects as static recrystallization, and that dynamic recrystallization is more sensitive to strain rate and less sensitive to temperature than static recrystallization (Humphreys and Hatherly 2004). It was believed that dynamic recovery was the sole softening mechanisms during hot deformation in aluminum due to its high stacking fault energy (SFE) (Ravichandran and Prasad 1991). The high SFE in aluminum and aluminum alloys, in particular at high temperature, may favor dislocation cross-slip, dynamic recovery, and subgrain formation over dynamic recrystallization. However, there is now direct microstructural evidence to support the hypothesis that dynamic recrystallization does occur in aluminum under deformation with large strain (>3) at elevated temperature both in polycrystalline and single crystal aluminum in various deformation modes including hot rolling, extrusion, hot torsion (Kassner et al. 1989; Yamagata 1992; Blum et al. 1996; Huang and Humphreys 1997; Ponge et al. 1997; Kassner and Barrabes 2005). Dynamic recrystallization has also been observed in Al-Cu-Mg-Zr alloy during hot compression (Huang et al. 2010).  24  It is noted that during large strain deformation of aluminum alloys at elevated temperature, equiaxed grain structures evolve under certain conditions, and the microstructure remains reasonably homogeneous with no clear distinction between unrecrystallized region and recrystallized region. Therefore it is considered as continuous dynamic recrystallization(Gholinia et. al 2002). McQueen and his coworkers first brought in the concept of geometric dynamic recrystallization (GRX) in the 1980s to explain certain dynamic continuous recrystallization cases in hot-worked aluminum (Kassner et al. 1989; Doherty et al. 1997). In geometric dynamic recrystallization, the microstructure is elongated; the adjacent grain boundaries are pushed towards each other; grains become thinner and thinner and eventually pinch off. Since the dramatically increase of high angle grain boundaries are not the same in GRX as in classic (discontinuous) dynamic recrystallization, for GRX to occur, the grain thickness has to decrease to a critical value which is of the order of sub-grain size(Humphrey and Hatherly 2004). Gholinia et al. (2002) studied the microstructure of hot-rolled Al3Mg0.2Cr0.2Fe alloy and Al3Mg0.2Sc0.1Zr alloy. The GRX mechanism was used to explain the dynamic recrystallization in these two alloys. Their study shows that a strain above 3 is needed for dynamic recrystallization; a minimum deformation temperature of 300oC is required however a temperature higher than 400oC leads to grain growth. The Al3Sc dispersoids in Al-Mg-Sc-Zr alloy increases the critical strain or the critical temperature for dynamic recrystallization to occur. The mechanisms of GRX has also been proposed by other researchers to successfully explain the grain  25  refinement in pure aluminum(Chang et al. 2000), Al-Mg(Doherty et al. 1997), Al-Mn(Gourdet and Montheillet 2002), Al-Cu(Kaibyshev et al. 2006), Al-Li-Mg-Sc(Kaibyshev et al. 2005) alloys during large strain hot deformation. 2.3.5 Microstructure development during extrusion Hot extrusion is a deformation process in which materials undergo complicated thermal and mechanical history. During the entire extrusion process, different metallurgical phenomena take place: static recrystallization, potentially dynamic recrystallization and grain growth (Pettersen et al. 2003). Physical parameters that usually determine the recrystallization kinetics, such as temperature, strain, and strain rate are now related to one or several process parameters including extrusion ratio, die geometry and temperature of container and extrusion.  Among all factors, extrusion ratio seems to be most crucial for occurrence of recrystallization in hot extrusion of 6xxx aluminum alloys. Recrystallization at the surface or no recrystallization occurs at extrusion ratio of 20 or lower, while complete recrystallization is observed at extrusion ratio of 40 or higher (Van Geertruyden et al. 2005; Ishikawa et al. 2006; Schikorra et al. 2007). In extrusion under high temperature and high strain, dynamic recrystallization may occur(Van Geertruyden et al. 2005). Moreover, static recrystallization and grain growth may act together when deformation ends after exit from the die, making the contribution of each phenomenon difficult to separate (Schikorra et al. 2008).   26   Figure 2-4 Polarized Light Optical Micrograph of partly extruded 6060 billet, reproduced with permission from (Kayser et al. 2010) Studies have been reported on both half-extruded and post-extruded aluminum alloys (Kaibyshev et al. 2005; Van Geertruyden et al. 2005; Ishikawa et al. 2006; Jäger et al. 2009; Zhu et al. 2009; Kayser et al. 2010). Kayser et al. (2010) investigated the microstructure of Al-Mg-Si-Fe by studying the half-extruded materials. The microstructure of the extruded rod gives an apparent three different zones (labeled as A, B and C in Figure 2.4). Zone A has a fine grain structure, which may be caused by the severe shearing in the die direction. Zone B comes from shear intensive zone (SIZ) and has a coarse grain structure caused by static recrystallization. Zone C comes from metal flow zone (MFZ) and maintains the fibrous structure. Moreover, the EBSD study indicates the occurrence of GRX, where the grain boundary angle gradually increases and the grain shape becomes serrated thus finally resulting opposing grain boundaries pinched to form new and smaller  27  grains. Consistent results for investigation on material flow from the die are also reported by Van Geertruyden et al. (2005) for the half-extruded Al-Mg-Si-Cr-Fe alloy. Another important microstructural feature in the extrudate is the peripheral coarse grain structure (PCG) formed at the surface of the extrudate. It is a common phenomenon in extrusion. Both mechanical properties and surface quality of the extrudate can be affected by this structure, and the machinability and surface appearance may also be reduced by PCG (Van Geertruyden et al. 2005; Birol 2010). Although the origin of the PCG structure is not clear, it is suggested by many researchers that the formation of the PCG is due to the abnormal grain growth after the extrudate exit from the die before/during the water quench (Van Geertruyden et al. 2005; Jäger et al. 2009; Birol 2010). To conclude, the evolution of the grain structure during extrusion is a result of an extremely complex set of mechanisms that superimposed in a complicated way, including dynamic recovery, static/dynamic recrystallization. The microstructure profile of the as-extruded material is dependent on the second phase particle, cooling condition, and extrusion ratio and temperature profile. 2.4 Constitutive modeling  This section includes a brief review on deformation mechanisms of aluminum alloy during high temperature and an introduction to the Kocks and Chen constitutive model for flow stress.   28  2.4.1 Deformation mechanisms of aluminum alloy The behavior of crystalline materials undergoing plastic deformation is controlled by different dislocation mechanisms except twinning. In general the strength of the material is determined by the kinetics of many processes at the atomic scale: the glide and climb of the dislocations, the diffusive flow of individual atoms, grain boundary sliding by diffusion and defect-motion and etc. Deformation mechanisms mainly include elasticity, low temperature plasticity by dislocation glide or twinning, power-law creep by dislocation glide/glide and climb, and diffusion flow (Frost and Ashby 1982). Normally more than one mechanisms contribute to the deformation of crystalline solids, with the dominant one depending on the stress and temperature to which the solid is exposed and its properties. For any specific strain rate and grain size, deformation temperature becomes the dominant factor in determining the overall behavior (Ashby 1972; Frost and Ashby 1982). Therefore based on the dominating mechanism, different constitutive equations have been developed and applied to predict the stress of the material (Nes 1997; Wright and Paulson 1998; Kocks 2001; McQueen and Ryan 2002; Meyers et al. 2002; van Haaften et al. 2002). During high temperature deformation, the climb of dislocations governed by lattice diffusion can be considered as the main process (Frost and Ashby 1982), where the temperature dependence is governed by the activation energy of diffusion. Activation energy of 140kJ∙mol-1 is obtained for pure aluminum (Bakshi and Kashyap 1995). The value may vary for aluminum alloys depending on the  29  chemistry, e.g. activation energies of 170 and 212 kJ∙mol-1 are reported for 0.6% and 1.3% Fe content of AlFe alloys (Bakshi and Kashyap 1995). 2.4.2 Kocks and Chen Model Kocks and Chen (1993) developed a physically based constitutive equation for prediction of yield stress and steady state flow stress in temperature/strain rate regime where solute drag mechanism dominates, i.e. at high temperature and low strain rate regime where the deformation is achieved by continuous motion of dislocations accompanied by diffusion of solute. The derivation of the model can be found elsewhere (Kocks and Chen 1993). The final equation is shown in Eq.2.6. 𝜀̇ = 𝐴 �𝜎 𝜇 � 𝑛 𝜇𝑏3 𝑘𝑇 exp �−𝑄𝐷 𝑅𝑇 �                      (2.6) Where 𝜀̇ is the strain rate, μ is the temperature dependent shear modulus, k is Boltzmann’s constant, b is magnitude of the Burger’s vector, T is deformation temperature and QD is the activation energy for diffusion of the main solute species (e.g. Mg in AlMg alloy and Mn in AlMn alloy), and R is the gas constant, A is a pre-exponential constant and n is the stress exponent. It’s a physical based model in which the stress is expressed as a function of strain rate and deformation temperature. At different strain rates and temperatures, the model will give a stress exponent n around 3 if solute drag is the operative deformation mechanism(Chen et al. 1998).  30  This model has been applied to fit the experimental steady state flow stress of AA5182 alloy from hot compression when the yield stress is below 70MPa (Chen et al. 1998) and  the work-hardening  curves are for practical purposes flat. In their study a stress exponent n value of 3.65 is found by using activation energy of 131kJ/mol for diffusion of Mg in aluminum. In contrast, Kubiak (2009) used the flow stress results from hot compression tests of AA3003 (1.27wt%Mn), AA3102 (0.26wt%Mn) and a variant Medium Mn alloy (0.75wt%Mn) over regime with strain rate of 0.1s-1, 1s-1 and 10s-1 and deformation temperature of 400oC, 500oC, and 600oC. High stress exponent values up to 8 were obtained by adopting activation energy of 211kJ/mol for diffusion of Mn solute in aluminum. The model was able to accurately describe the data but its physical basis was lost. Hypothesis was brought up by Kubiak that there may be some microstructural reason for the high stress exponent value. Therefore, further validation is suggested.   31  Chapter 3 Scope and Objectives 3.1 Scope  The through-process microstructure characterization during extrusion was implemented in this work. Various microscopy methods were adopted for microstructure characterization on the AA3003 (1.27Mn0.54Fe0.10Si) and AA3102 (0.26Mn0.50Fe0.10Si) alloys for as-cast, as-homogenized, as-extruded states.  Selected homogenization treatments included 8 hours soaking at 500oC, 8 hours soaking at 550oC, and 24 hours soaking at 600oC. In addition, to investigate the evolution of constituent particles during homogenization, samples were heated up and soaked at 600oC for different times. The extrusion trials were performed at extrusion temperature of 400oC and 550oC at extrusion ratio of 130.  Optical Microscopy was performed to qualitatively evaluate the distribution of dispersoids in those alloys. The quantification of constituent particles during homogenization was examined by Scanning Electron Microscopy and Clemex Image Analysis. Polarized Light Optical Microscopy was adopted on anodized samples to reveal the through-process grain structure variation. Electron Backscattered Diffraction technique was also used to analyze recrystallization behavior.  As continuation work of previous constitutive modeling work on AA3xxx alloys during hot compression by Kubiak, Gleeble tests were performed on a low iron AA3102 (0.26Mn0.10Fe0.10Si) alloy. The flow stress and the yield stress data were then fitted to Kocks and Chen’s solute drag model to clarify the effect of the constituent particles fractions on the constitutive behavior of AA3xxx alloy.  32  3.2 Objectives Knowledge of the microstructure evolution of 3xxx aluminum alloys during extrusion manufacturing process is important to understand the effect that the extrusion conditions have on microstructure development and ways the process can be optimized. Quantification on constituent particles could be used to validate the solidification and homogenization model by Du et al. (2010). The continuation work on constitutive study of low iron AA3102 is to clarify the effect of constituent particles on the model parameters and to complete the establishment of the AA3xxx constitutive model from previous work by Kubiak (2009). The grain structure information can be useful to justify the FEM and microstructure model predication by Mahmoodkhani et al.(2010).  The objectives of this research are: 1) to study the through process microstructure evolution; 2) to quantify the evolution of constituent particles during homogenization, including the number density, size and area fraction; 3) to study the effect of different homogenization conditions on the as-extruded microstructure; 4) to study the constitutive behavior for the stress of a low iron AA3102 alloy during hot compression.  33  Chapter 4 Experimental Methodology In this chapter, a brief description of the materials is given, followed by the procedure for homogenization treatment and extrusion trials. Microstructure characterization methods are introduced, including optical microscopy, polarized light optical microscopy, scanning electron microscopy, and electron backscattered diffraction. Finally, the methodology for compression tests with the Gleeble thermomechanical simulator is described. 4.1 Initial materials All the experimental alloys, either in billet form or as extrudates were provided by the Alcan Research and Development Center (ARDC) of Rio Tinto Alcan located in Jonquiere Quebec. The Chemical Composition of AA3003 and AA3102 for microstructure characterization at as-homogenized and as-extruded state and a low iron AA3102 alloy for flow stress behavior study are listed in Table 4.1. Table 4-1 Chemical composition of experimental alloys (wt %) Alloys Mn Fe Si Ti Cu Al AA3003 1.27 0.54 0.10 0.02 <.01 Bal. AA3102 0.26 0.50 0.10 0.02 <.01 Bal. Low Iron AA3102 0.26 0.10 0.10 0.02 <.01 Bal. The composition of AA3003 and AA3102 alloys was selected in such way that it allowed us to identity the main effect of the alloying element manganese on  34  microstructure evolution. The low iron AA3102 which has a low volume fraction of constituent particles was chosen to study its stress/strain response as a function of strain rate and temperature. 4.2 Homogenization Specimens taken from as-cast billets were homogenized in a Carbolite TM (HRF) circulating air furnace. Thermocouples were welded onto the sample to monitor the temperature changes during the homogenization, and thermocouple data was acquired using NI-9219 Data Acquisition (DAQ) module which was then connected to computer and recorded by LabView Express software. K type thermocouples were chosen for their temperature range up to 1100oC, and tolerance of ±1.5°C (ASTM Standard E235 - 06). The NI -9219 DAQ module has four 6-terminal connectors that provide connections for four analog input channels. For thermocouple output mode of this DAQ module, the positive signal of the signal source was connected to signal terminal HI, and the negative to the terminal LO. A photo of the recirculating air furnace is given in Figure 4.1. The heating rates were designed to be 150̊C/h up until 50 oC below the soaking temperature, and a heating rate of 50̊C/h for the last 50 oC to soaking temperature (see Figure 4.2). After soaking, samples were immediately water quenched to room temperature.  35   Figure 4-1 Recirculation air furnace  Figure 4-2 Homogenization treatment profiles  36  4.3 Extrusion trials Rio Tinto Alcan’s state-of-the-art research extrusion press located in ARDC was used for the extrusion trials. Photos of the press and the water quench are given in Figure 4.3 and Figure 4.4. The billets were heated up prior to extrusion to the deformation temperature in an induction heater, and then moved by a mechanical billet loading system into position in the press. The ram started pushing the billet through the die immediately after the billet is in position. The extrudate exited the die and travels through a water quench. The end of the extruded strip was then pulled gently by a technician to ensure that the strip did not back up in the press.  Figure 4-3 Extrusion trials in ARDC 1) extruded strip exits here, 2) billet position for extrusion, 3) mechanical billet loading system, 4) ram  37   Figure 4-4 Water quench unit located at die exit. 5) Extrudate being water quenched Table 4-2 Parameters for the extrusion trials Trials Billet T/oC Container T/oC Die T/oC Extrusion Ratio Ram Speed /mms-1 Exit Speed/m s-1 1 400 450 450 130 14 1.8 2 550 480 480 130 14 1.8 The principle extrusion variables are shown in Figure 4.5(Saha 2000). Ac and AE are cross-sectional areas of the container bore and extruded shape respectively, VR is the ram speed, VE is the extrusion speed (i.e. the extrudate exit speed). Billet temperature is also called extrusion temperature. The dimension of the I-beam extrudate is illustrated in Figure 4.6.  38    Figure 4-5 Principle extrusion variables   a)  b) Figure 4-6  I-beam Extrudate a) dimension b) actual extrudate The extrusion ratio (ER) and the effective strain (Ɛ eff) are calculated in Eq. 4.1 ~ Eq.4.2. 𝐸𝑅 = 𝐴𝐶 𝐴𝐸 = 𝐴𝐶2𝐴𝑓 + 𝐴𝑤 ≈ 130                             (4.1) 𝜀𝑒𝑓𝑓 = 𝑙𝑛(𝐸𝑅) = 𝑙𝑛130 ≈ 4.9                            (4.2)  39  Where Af  is the area of the flange, and Aw is the area of the web. The extrusion speed VE is calculated in terms of ram speed by using simple mathematical relations of volume constancy by Eq. 4.3(Saha 2000): 𝑉𝐸 = 𝑉𝑅𝐴𝐶𝐴𝐸 = 14𝑚𝑚 ∙ 𝑠−1 × 𝜋 ∙ (50𝑚𝑚)260.5𝑚𝑚2 ≈ 1.8𝑚 ∙ 𝑠−1      (4.3)  4.4 Microstructure characterization The microstructure characterization techniques include Optical Microscopy, Polarized Light Optical Microscopy, Scanning Electron Microscopy (SEM) and Electron Backscattered Diffraction (EBSD). They are briefly introduced in the following subsections. 4.4.1 Sample preparation  Sample sectioning from the billets and the extrudates are explained in this section. 4.4.1.1 Billet sample  Figure 4.7 shows three samples taken from center of the billet, middle of the radius of the billet and outer of the billet respectively for microstructure study, e.g. samples 1 to Sample 3 are used to investigate whether there is any significant microstructure variation across the radius of the billet due to solidification in casting.  40   Figure 4-7 Position indication of billet samples 4.4.1.2 Extruded sample It was expected that the microstructure of the extrudate varied through its thickness, therefore two types of samples were made from the extrudates, which were named as “AD samples” and “tapered samples”.  Figure 4-8 Dimension of a) AD sample with the observation plane indicated as the red color b) tapered sample with the observation plane indicated as the blue color The location in the extrudate from where the samples were taken is shown in Figure 4.8. The tapered samples were made by cold mounting a sample which was Casting Direction  41  purposely inclined around 2o from the observation plane. A screw was inserted into the mount to form conductive path to each sample which is necessary for anodizing. In order to examine the through-thickness microstructure variation for each homogenization condition, micrographs were taken from different positions from the tapered samples (See Figure 4.9), and were labeled as “surface”, “¼ thicknesses” and “½ thicknesses”(center). By combining the observation from AD sample and the Tapered sample, the experimental method adopted allows a 3-D profile microstructure. An example of the microstructure interpretation is displayed in Figure 4.10 to explain the through-thickness profile of the extrudate material. As shown in Figure 4.10, the surface grains are coarse and elongated along the extrusion direction, and center grains are finer and more equiaxed.  Figure 4-9 Dimension of Tapered samples  42   Figure 4-10 AA3102 as-cast, and extruded at 400oC a) schematic representation of the combined 3-D profile, b) picture from AD sample, and c) picture from tapered sample 4.4.2 Metallographic preparation procedure ColdCure Resin and Hardener were used as mounting media (2:1 mixed, and cured for 24h in room temperature). Samples were grinded and polished according to the metallographic preparation procedure as shown in Table 4.3. Polarized light optical microscopy requires a high quality surface finishing, therefore samples were more finely polished using 0.05 μm diamond suspensions. However, Back-scattered Scanning Electron Microscopy doesn’t have such high requirement on surface quality and more importantly, test showed that the colloidal silica would disturb the EDX analysis results, so samples for SEM were polished down to 1μm for micrographs collection and element mapping analysis.  43  Table 4-3 Metallographic procedure for grinding and polishing Type of Paper/cloth Grit or μm Speed(rpm) Time (min) Lubricant SiC 240/600/800 grit 200 2 for each water Texmet 6 μm 200 7 diamond suspension Microcloth 1 μm 200 5 diamond suspension Chemotextile 0.05 μm 200 5 colloidal silica In order to get desired microstructural results, different etchants were prepared and tested. Etchant of a mixture of distilled water and 0.5% hydrofluoric acid (48%HF) was used to qualitatively reveal the dispersoids and the dendritic structure of the material. Samples were immersed in the etchant for 60s and then water rinsed. Barker’s reagent, i.e. 3%HBF4 (48%), was employed to anodize the material for revealing grain information. The anodizing equipment was set up as Figure 4-11. Pure aluminum sheet was used as the cathode, and the specimen acted as the anode. The voltage was set to be 30-35volts. Depending on the size of the sample, i.e. cross-sectional area that electrical current passes through, it took 200-250 seconds to anodize each sample at room temperature.   44   Figure 4-11 Picture of anodizing equipment As to the preparation for Electron Back-Scattered Diffraction (EBSD), Samples were electro-polished. 25% Perchloric acid (HClO4) and 75% Methyl alcohol (CH3OH) was chosen as the electrolyte. The electro-polishing was performed at -25oC (provided by liquid nitrogen bath) for 40s. 4.4.3 Microscopy Optical microscopy was implemented by a Nikon Epiphot 300 Optical Microscope. Optical images with magnification of x50, x100, and x200 were studied. Polarized light optical microscopy was implemented on Nikon Epiphot 300 Optical Microscope as well with polarizer slide and analyzer slide crossed inserted in the light path. Scanning Electron Microscopy (SEM) was performed on a Hitachi S-3000N Scanning Electron Microscope with light element EDX to investigate a)  45  microstructure features of the constituent particles and b) elemental distribution in the as cast and homogenized samples and c) micrographs collection for further image analysis and quantification. Images were collected in the back-scattered imaging mode with working voltage of 20kV and working distance of 15mm. Electron Back-Scattered Diffraction (EBSD) was done on a Hitachi S-570 SEM. The HKL Channel 5 software was used to obtain the EBSD patterns using a step size of 2μm. The boundary between low and high angle boundaries was selected to be 15o. 4.5 Image analysis Quantitative analysis of constituent particles, which included the size, aspect ratio, area fraction and number density, was done using the Clemex Image Analysis Software with a built-in routine. An example of image analysis process is shown in Appendix C along with reliability test of the method. To quantify the grain structure of the as-cast and as-homogenized samples, micrographs from Polarized Light Optical Microscopy were printed out. The grain boundaries were manually traced on a transparent plastic sheet, which was then scanned into computer and analyzed by Clemex Image Analysis Software. Figure 4.12 shows an example from the as-cast AA3003. For the grain structure of the extrudate, due to the complexity of EBSD preparation and large grain size variation, grain size estimation was done by Clemex Image Analysis with routine based on equal area diameter method.  46   Figure 4-12 Example of manually plotting on as-cast AA3003 4.6 Compression tests Compression tests were performed at strain rates of 0.1 s-1, 1 s-1, and 10s-1 at deformation temperatures between 400°C to 600°C using the Gleeble® 3500 thermal-mechanical testing machine as shown in Figure 4.13. The dimension of each compression sample was measured prior to test, which was approximately 8mm in diameter and 12mm in length. A thermocouple (K type) was spot welded to each sample at approximately mid-point along the length. Compression samples were lubricated using nickel paste which is applied manually. The jaw set-up inside the Gleeble machine is shown in Figure 4.14. A 5,000 lb (22.2x103N) load cell was used to for all the tests.  Strain measurements were made using a c-gauge that is located at the mid-point of the sample length. The ram was moved to a position in which it was exerting a small force on the sample (<200 N) to hold the sample in place as the weight of the C-gauge would otherwise moved the sample; when this position was reached the ram was set to what would be the zero  47  position for the subsequent test. The rough vacuum was used to evacuate the experimental chamber to a pressure of 2.7^10-1 torr(36 Pa).  Figure 4-13 Gleeble Thermal-mechanical Machine for compression tests  Figure 4-14 Jaw setup 1) jaws, 2) thermocouples, 3) sample, 4) load cell Compression samples were heated up at a rate of 5°C/s to the required deformation temperature. The samples were then held at temperature for three  48  seconds before deformation at the required strain rate. Gleeble output data included: time, diameter, force, strain, stress, stroke, and temperature. Stress-strain curves were constructed using the Gleeble strain-stress data. The yield stress values were determined using the 0.2% offset method by finding the intersection of the stress-strain curve with a line parallel to the slope of Young’s modulus and which intercepted the abscissa at 0.2%. The calculation for the temperature dependant Young’s modulus is shown in Appendix D. The repeatability of tests and methods for determining flow stress are explained in Appendix A.   49  Chapter 5 Experimental Results This chapter presents the experimental results produced from the microstructure characterization and image analysis methods described in the previous chapter. The results are categorized according to the sequence of the industrial manufacturing route. At first, the microstructure characterization of the as-cast materials is given. Results of the microstructure characterization on the as-homogenized materials are then presented, followed by ones on the as-extruded materials. Finally, the results from Gleeble tests of a low iron AA3102 alloy are presented. 5.1 Microstructure characteristics in as-cast billets The results on microstructure characteristics in as-cast AA3003 and AA3102 billets are displayed in this section. 5.1.1 As-cast AA3003  Optical images of samples taken from different positions within the billet are shown in Figure 5.1. It can be seen there are only minor variations in the grain size and dendrite arms spacing over the cross-section of the billet. To evaluate the grain size change through the entire manufactory process, grain size characterization on the as-cast AA3003 was performed. The average grain size was found to be 72μm. A polarized light optical micrograph of as-cast AA3003 is shown in Figure 5.2.   50    a) b)  c) Figure 5-1 As-cast AA3003 a) centre of the billet,  b) half radius section of the billet,  c) outer section of the billet  Figure 5-2 Grain structure of as-cast AA3003  51   a)  b) Figure 5-3 Higher magnification of Back Scattered SEM micrographs showing views of constituent particles in as-cast AA3003 a) x2000 b) x7000 Another important microstructural feature are the constituent particles. It can be observed in Figure 5.3 that the interdendritic constituent particles have a eutectic structure. 5.1.2 As-cast AA3102 Microstructure characterization was also performed on AA3102 alloys. Figure 5.4 shows the Optical Micrograph, and Figure 5.5 shows the SEM micrograph of the as-cast AA3102. Similar to AA3003 alloys, the interdendritic constituent particles have a eutectic structure.  52    Figure 5-4 OM micrograph of as-cast AA3102 Figure 5-5 SEM micrograph of as-cast AA3102 5.2 Microstructure evolution during homogenization Upon homogenization the redistribution of the alloying elements occurs, and a potential phase transformation of Al6(Mn,Fe) to α-Al(Mn,Fe)Si may occur together with the evolution of the constituent particles. The grain size from the as-homogenized material is consistent with the result from the as-cast material, which indicates that there is no change in grain size during homogenization in these two alloys. 5.2.1 Evolution of constituent particles and segregation in the matrix Element mapping using quantitative X-ray mapping from the EDX in the SEM was conducted on the as-cast AA3003 and homogenized sample. Figure 5.6 and 5.7 show backscattered images and corresponding element maps for manganese, silicon and iron in the as-cast and homogenized (600oC for 24h) samples respectively. The distribution of manganese atoms changes after homogenization as Mn is segregated from the primary Al to the manganese-rich constituent particles.  53   Figure 5-6 As-cast AA3003: a)backscattered electron image, b)x-ray map for Mn, c)x-ray map for Fe, d)x-ray map for Si  Figure 5-7 AA3003 homogenized at 600oC for 24h: a)backscattered electron image, b)x-ray map for Mn, c)x-ray map for Fe, d)x-ray map for Si  54  The SEM micrographs of AA3003 samples heated and quenched from different temperatures without soaking are shown in Figure 5.8. It is shown that the eutectic microstructure starts to break up into smaller particles around 400oC. The SEM micrographs of AA3003 samples homogenized at 600oC for different soak times are shown in Figure 5.9 to reveal the evolution of the constituent particles. It is shown that the constituent particles become coarser and are spheroidized with increasing soaking time.  Figure 5-8 Backscattered electron micrographs for samples with a heating rate of 150oC/h and quenched from : a) 200oC, b)350oC, c) 400oC, d) 500oC and e) 550oC  55   Figure 5-9 Backscattered electron micrographs for samples soaked at 600oC for : a)0min, b)10min, c)20min, d)40min,e)1h,f)12h,g)24h, and f)48h  56  Quantitative measurements of the number density and area fraction of constituent phases on those SEM micrographs have been done as described in section 4.5.1 and the results are shown in Figure 5.10 and Figure 5.11. It can be observed in Figure 5.10 that the circular diameter increases during the first 1 hour and it remains almost constant at a size of 1.2μm. Note the size distribution of constituent particles was measured in all cases, the circular diameter displayed here are average circular diameter. Homogenization time at 600oC /h 0 10 20 30 40 50 60 C irc ul ar  D ia m et er / µ m .5 1.0 1.5 2.0 2.5 A sp ec t R at io 0.0 .5 1.0 1.5 2.0 2.5 Circular Diameter Aspect Ratio Figure 5-10 Size and aspect ratio of constituent particles in AA3003 soaked at 600oC  57  soaking time at 600oC/ h 0 10 20 30 40 50 60  A re a fr ac tio n( % ) 2 3 4 5 6 N um be r d en si ty  (# /m m 2 ) 20000 25000 30000 35000 40000 45000 50000 55000 Area Fraction Number density  Figure 5-11 Area fractions and number density of constituent particles in AA3003 during homogenization at 600oC 5.2.2 Observations on dispersoids for different homogenization temperature OM micrographs of samples etched with 0.5% hydrofluoric acid are shown in Figure 5.12. A large number density of dispersoids are found in the samples homogenized at 500oC for 8h (Figure 5.12a). Fewer dispersoids are observed in the samples homogenized at 550oC for 8h (Figure 5.12b). It is also noted that in Figure 5.12(b), precipitation free zone appears near the constituent particles. Almost no dispersoids are found in the samples homogenized at 600oC for 24h (Figure 5.12c).  58   Figure 5-12 a) OM micrographs of AA3003, 500oC, 8h homogenized  Figure 5-12 b) OM micrographs of AA3003, 550oC, 8h homogenized  59   Figure 5-12 c) OM micrographs of AA3003, 600oC, 24h homogenized SEM micrographs of constituent particles are shown in Figure 5.13. As the soaking temperature goes up, the network of the constituent particles breaks up and the particle size becomes larger while the number density drops.  60   Figure 5-13 Backscattered electron micrographs of homogenized AA3003 for a) 500oC, 8h b) 550oC, 8h and c) 600oC, 24h Quantitative analysis on the area fraction and number density of the constituent particles in those conditions is performed and results are shown in Figure 5.14 and Figure 5.15.  61      Homogenization treatment as cast 500C, 8h 550C, 8h 600C, 24h A re a fr ac tio n% 2.4 2.78 3.45 3.9  Figure 5-14 Area fractions of constituent particles in AA3003 C irc ul ar  D ia m et er / µ m .8 .9 1.0 1.1 1.2 1.3 1.4 1.5 Pa rti cl e/ m m 2 20000 30000 40000 50000 60000 Circular Diameter Number Density 500oC, 8h              550oC, 8h             600oC, 24h Figure 5-15 Number density and circular diameter of constituent particles in AA3003 after different homogenization treatments  62  5.2.3 Optical micrographs of as-homogenized AA3102 Figure 5.15 shows the OM micrographs of as-homogenized AA3102. A few dispersoids are shown in Figure 5.16 a) and b), and no dispersoids exist in Figure 5.16 c), i.e. samples homogenized for 24h at 600oC. Qualitative observations show that this alloy has fewer dispersoids compared to AA3003 alloy, though quantification of this effect is still needed. Figure 5-16 OM Micrographs of Homogenized AA3102 a) 500oC, 8h b) 550oC, 8h c) 600oC, 24h   63  5.3 Microstructure characteristics of extruded Samples Microstructure observation on extruded samples includes the SEM micrographs on constituent particles, polarized light optical microscopy and EBSD images on grain structure study. Those micrographs were then analyzed by Clemex to quantify the volume fraction, circular diameter, number density of constituent particles and to estimate the grain size of the extrudates. 5.3.1 As-extruded AA3003               Both grain structure and constituent particles characteristics are investigated from the as-extruded AA3003. 5.3.1.1 Constituent particles SEM micrographs were taken from the AD sample of extrudates. Figure 5.17 shows BSE micrographs of constituent particles in the extrudates. In Figure 5.17(a) the original eutectic rod-like constituent particles have been fractured during extrusion and become distributed randomly in the matrix. In Figure 5.17(b, c & d), it is evident that the constituent particles tended to align preferentially along the extrusion direction in the homogenized cases. During the extrusion, the surface and centre of the extrusions go through different thermo-mechanical histories. The entire AD cross-thickness area was observed by SEM to study the potential effect of this on the distribution of constituent particles. Figure 5.18 shows an example of the constituent particle distribution at the surface and the centre of cross thickness view on the AD sample.  It appears that slightly more constituent particles exist at the surface compared to centre. Figure 5.18 a) and b) show the SEM micrograph for  64  the constituent particles at the surface and center of AD sample from the extrudate. Figure 5.19 shows the quantified number density variations across the AD section. It is seen that there appears to be more particles at the surface than center of the extrudate.  Figure 5-17 SEM micrographs of constituent particles in AA3003, extruded at 400oC a) as-cast, and homogenized for b) 500oC, 8h c) 550oC, 8h d) 600oC, 24h Extru sion  D irection   65   Figure 5-18 SEM view of constituent particles in AD sample from AA3003, 600oC and 24h homogenized and extruded at 400oC a) surface b) center of the extrudate from surface to surface cross the section of the AD sample #/ m m 2 16000 17000 18000 19000 20000 21000 22000 23000 24000  Figure 5-19 the number density variation of the constituent particles of the entire AD plane (AA3003, 600oC and 24h homogenized and extruded at 400oC)  66  500oC, 8h              550oC, 8h             600oC, 24h C irc ul ar  D ia m et er /µ m 0.0 .2 .4 .6 .8 1.0 1.2 1.4 Before extrusion  Figure 5-20 Average circular diameters of constituent particles a) before extrusion 500oC, 8h              550oC, 8h             600oC, 24h C irc ul ar  D ia m et er /µ m 0.0 .2 .4 .6 .8 1.0 1.2 1.4 1.6 After extrusion at 400oC  Figure 5-20 Average circular diameters of constituent particles b) after extrusion at 400oC The evolution of the size of the constituent particles in homogenized materials before and after extrusion is presented in Figure 5.20 a) and b) respectively. It can be observed that homogenization at higher temperatures (and in  67  the case of 600oC, longer times) leads to slightly larger average circular diameters. Turning to the volume fraction of constituent particles, it is noted that the area fraction of the phase on the plane of a two dimensional image is equal to the volume fraction of the phase in the solid(Russ 1990). The area fractions before and after extrusion are compared in Figure 5.21. It can be seen from Figure 5.21 that considering the error bars, the area fraction of homogenization treatments of 8h at 550oC and 24h at 600oC appears unchanged after extrusion. However, there is what appears to be a small difference between the area fraction before and after extrusion in the case of the homogenization of 8 hours at 500 oC. The reason for this difference is unclear.    500oC, 8h                   550oC, 8h                600oC, 24h Ar ea  F ra ct io n/ % 0 1 2 3 4 5 Before extrusion After extrusion at 400oC  Figure 5-21 Area fraction of constituent particles before and after extrusion (400oC) Turning to the number density of constituent particles, one needs to be careful when one compares the number density before and after extrusion since there is a noticeable change in the alignment of constituent particles after extrusion.  68  This has an effect on the stereology of the measurement(Russ 1990). Based on current experimental method, it would be inappropriate to compare the number density before and after extrusion due to this effect. Figure 5.22 a) and b) show the number density of constituent particles before and after extrusion, respectively. The trends in the evolution of number density for the constituent particles are similar, i.e. homogenization at 600oC for 24h gives lowest number density among all. Qualitative observations from optical metallography suggest that the morphology of the constituent particles does not strongly evolve during extrusion. Fragmentation of the constituent particles occurred only in the as-cast case where the eutectic network was broken down into small rod like particle, but rarely occurred in the homogenized cases. #/ m m 2 20000 25000 30000 35000 40000 45000 50000 Before extrusion 500oC, 8h              550oC, 8h             600oC, 24h  Figure 5-22 Number density of constituent particles a) before extrusion  69  #/ m m 2 20000 25000 30000 35000 40000 45000 After extrusion at 400oC 500oC, 8h              550oC, 8h             600oC, 24h  b) after extrusion at 400oC 5.3.1.2 Grain structure  Through-thickness microstructure profiles taken from the taper mounted samples are shown for AA3003 with four homogenization conditions (i.e. as cast, homogenized for 8h@500oC, for 8h@550oC and for 24h@600oC).  Figure 5.23 shows the as-extruded microstructure for an extrusion temperature of 400oC. Thickness-dependent profiles are observed for all extrudate. For the extruded as-cast material, a fine grain size layer exists at the surface. The most inhomogeneous structure is shown in materials which had been homogenized at 500oC for 8h. Grains of irregular shape are found in material which has been homogenized at 550oC for 8h. The finest and equiaxed grain structures are found for 600oC, 24h homogenized one. Except for the homogenization treatment of 500oC for 8h, grain size decreasing from the surface of the extrudate to the center is observed.  70   Figure 5-23 Microstructure of AA3003 I-Beam extruded at 400oC OIM (Orientation Imaging Microscopy) image from EBSD on 400oC as-extruded sample (600oC, 24h homogenized prior to extrusion) is shown in Figure  71  5.24. The image is presented in the inverse pole figure color. Each color represents a crystal orientation. Black lines are the grain boundaries with misorientation angles greater than 15o, and gray lines are the grain boundaries with misorientation angles between 5o and 15o. In the texture near surface region, crystal orientations are distributed more randomly while there is a strong cube recrystallization texture in the interior of the extrusion.   Figure 5-24 EBSD mapping of 400oC as-extruded (24h@600oChomogenized)  ND ED  72   Figure 5-25 Microstructure of AA3003 I-Beam extruded at 550oC             Grain Structure of AA3003 extruded at 550oC is shown in Figure 5.25. Different from previously showed microstructures in Figure 5.23 that were extruded at 400oC, it is found in Figure 5.25 that materials from this extrusion trial show some very fine grains at the surface of the samples while large grain regions beside them. Other than that, same trend of microstructure evolution from surface  73  to center is observed. The most inhomogeneous microstructure is also found for the materials homogenized for 8h@500oC, which is also the only condition in this work where incomplete recrystallization is observed. 8h@550oC homogenized material shows large grain at the surface and smaller grains in the center. 24h@600oC homogenized material shows fine and equiaxed grains, being the most homogenized microstructure among these four. 5.3.2 As–extruded AA3102 The grain structure of AA3102 extruded at 400oC, 550oC is shown in Figure 5.26 and Figure 5.27 respectively. Similar to what was observed for AA3003, an extrusion temperature of 550oC also gives very fine surface grain layers outside of coarse grain region compared to those extruded at 400oC. Fully recrystallized structures are shown for each case. The most inhomogeneous grain structures are observed for homogenization condition of 24h@600oC.  74   Figure 5-26 Microstructure of AA3102 I-Beam extruded at 400oC  75   Figure 5-27 Microstructure of AA3102 I-Beam extruded at 550oC   76  5.4 Effect of constituent particles on flow stress  The results from the Gleeble tests are presented in this section to investigate the flow stress behavior of a low iron AA3102 alloy with a low volume fraction of constituent particles. 5.4.1 SEM micrograph of low iron AA3102 The SEM micrograph of low iron AA3102 is shown in Figure 5.28. It is found that there is much lower number density (5673.0 mm-2) and area fraction (0.7%) of constituent particles in this alloy compared to AA3003 (number density of 38597.6 mm-2 and area fraction of 2.0%), due to its low iron and low manganese content.  Figure 5-28 SEM micrograph of as-cast low iron AA3102  77  5.4.2 Stress-Strain curves Compression tests on low iron AA3102 specimens were performed on the Gleeble machine. Stress-strain curves of as-cast and as-homogenized AA3102 specimens subject to various strain rate and deformation temperature are shown in Figure 5.29 and Figure 5.30 respectively. The flow stress increases with decreased deformation temperature and increases with increased strain rate as expected. Figure 5.31 plots the selected stress-strain curves of cast and differently homogenized specimens under the same compression condition. Yield stress and flow stress data achieved from the stress-strain curves are plotted in Figure 5.32, which shows that the flow stress does not vary much for homogenized material for this alloy. ε 0.0 .1 .2 .3 .4 .5 .6 σ / M P a 0 20 40 60 80 600 oC,10/s 600oC,1/s 600oC,0.1/s 500oC,10/s 400oC,10/s 400oC,1/s  Figure 5-29 Stress-strain curves from as-cast low iron AA3102 for selected compression conditions  78  0.0 .1 .2 .3 .4 .5 0 20 40 60 80 400oC, 1/s 400oC, 10/s 500oC, 1/s 500oC, 10/s 600oC, 0.1/s 600oC, 1/s 600oC, 10/s ε σ / M Pa  Figure 5-30 Stress-strain curves from 24h@600oC homogenized low iron AA3102 for selected compression conditions  Figure 5-31 Selected stress-strain curves of as-cast and homogenized low iron AA3102 ε 0.0 .1 .2 .3 .4 .5 σ/ M pa 0 20 40 60 80 400oC, 10/s (500oC, 8h homogenized) 400oC, 10/s (600oC, 24h homogenized) 400oC, 10/s (as cast) 600oC, 0.1/s (500oC, 8h homogenized) 600oC, 0.1/s (600oC, 24h homogenized) 600oC, 0.1/s (as cast)  79  400oC, 10/s;     500oC, 10/s;    600oC,10/s;   600oC,1/s;    600oC,0.1/s σ /M Pa 0 10 20 30 40 50 60 Flow stress for as-cast Flow stress for 600oC, 24h homogenized Yield Stress for as-cast Yield Stress for 600oC, 24h homogenized  Figure 5-32 Effect of homogenization on flow stress and yield stress of low iron AA3102   80  Chapter 6 Discussion In this chapter, the results from the previous chapter are discussed. The discussion starts with the evolution of the second phase particles during homogenization and extrusion, and then considers through-thickness profiles of the extrudates. The effects of the homogenization conditions, extrusion temperature and alloying chemistry on the microstructure of the as-extruded 3xxx material are considered. Finally, the constitutive behavior of the low iron 3102 alloy is evaluated using Kocks and Chen solute drag model. Comparison is made among the modeling parameters to investigate the effect of constituent particles. 6.1 Evolution of second phase particles during homogenization and extrusion This section provides discussion on evolution of constituent particles and dispersoids during homogenization and extrusion. 6.1.1 Constituent particles The evolution of constituent particles in DC cast 3003 during homogenization has been investigated quantitatively. It is found that different mechanisms including break-up and coarsening compete during the process. As homogenization goes on, the eutectic phase spheriodizes, and the size and area fraction of the constituent particles increase, while the aspect ratio and number density decrease. When soaking at 600oC the size and area fraction of the constituent particles go through a large increase for the first 1h while slower increases for long time  81  soaking. It is noted that there are a large number of dispersoids in the dendrite arm region when soaking starts. Those dispersoids disappear during the soaking indicated by increased PFZ zone. Therefore the increase in the area fraction and the coarsening of the constituent particles is a complex process, caused by both diffusion of Mn from the supersaturated solid solution toward the constituent particles and also concurrent nucleation/growth/dissolution of dispersoids.  Figure 6-1 Particles with different contrast from SEM micrograph of a 550oC, 8h as-homogenized AA3003 sample As suggested by the literature, the α-Al(Mn,Fe)Si is a silicon-rich phase and it is supposed to show as brighter phase under backscattered SEM. EDX was performed with a working distance of 15mm and a 20.0KV accelerating voltage on the as-cast and homogenized samples to investigate the phase of the constituent particles in the current alloy. One example is shown in Figure 6.1 for a sample homogenized 8h@550oC. EDX for the dark particle shows 3.94wt% Mn and Bright Particle Dark Particle  82  1.37wt% Fe while the bright particle shows 8.35wt% Mn, 6.42wt% Fe and 0.15wt% of Si(such low silicon content is negligible and unreliable because of the limitation of the detection method). Other EDX results reveal that there are no silicon-rich particles in either cast material or homogenized material (theoretically, the silicon content in a α-Al(Mn,Fe)Si particle is >5.3wt%), further suggesting that very few α-Al(Mn,Fe)Si phase exists. This is assumed to be caused by the low silicon content in this alloy (0.1wt%) compared to other reported work (where silicon content⩾0.2wt%)( Li and Arnberg 2003, Dehmas et al. 2005, Alexander and Greer 2002), which is highly possible because silicon is essential for phase transformation from Al6(Mn,Fe) to α-Al(Mn,Fe)Si. According to the ASTM E-562 standard for fraction determination of secondary phase, area fraction measured by backscattered electron SEM technique could be used as the volume fraction. Comparison of experimental area fractions were compared with equilibrium predictions from Thermo-Calc(2010). Thermo-Calc is a software package used to perform thermodynamic and phase diagram calculations for multi-component systems. TTAl6 database is a comprehensive database for all commercial Al-alloys. The mole fractions of equilibrium phases were found for current temperatures (i.e. 500oC, 550oC, and 600oC) from Thermo-Calc upon using TTAl6 database. The densities of the matrix, the Al6Mn phase, the α-Al(Mn,Fe)Si are taken as 2.70 g/cm3, 3.18 g/cm3 and 3.55g/cm3 (from Table 2.2) to convert mole fractions to volume fractions. Table 6.1 shows the comparison between the experimental data and the equilibrium predication from Thermo-Calc.  83  Optical observations have shown that the presence of dispersoids in materials homogenized for 8h at 500oC and 8h at 550oC are evident, as in Figure 5.12(a) and (b), which explain the discrepancies between measured and predicted data for these two temperatures. It is also acceptable in terms that the measured volume fraction of constituent particles at 600oC is the closest to the Thermo-Calc predication since optical observation does show that long time soaking at high temperature almost completely eliminates the dispersoids in the system, as in Figure 5.12(c). Table 6-1  The measured fraction of constituent particles and equilibrium phase fraction from Thermo-Calc Homogenization  Volume Fraction Measured fraction of constituents (%) Thermo-Calc Predication (%) Al6Mn α-Al15(Mn,Fe)3Si Sum 500oC,8h 2.8±0.3 4.2 1.1 5.3 550oC,8h 3.5±0.3 4.1 0.7 4.8 600oC,24h 3.9±0.2 4.1 0.2 4.3 It is evident from the experimental observations that the increase of the fraction of constituent particles in AA3003 alloy during homogenization is accompanied by a decrease in volume fraction of dispersoids. The growth and transformation of constituent particles occur through diffusion of the alloying elements at the scale of the secondary dendrite arm spacing, while the nucleation, growth and coarsening of dispersoids occur intragranularly. Simulations have been  84  implemented by Gandin and Jacot (2007) and Du et al. (2011) to model the evolution of the constituent particles and dispersoids during homogenization in AA3003 alloy. The latter work used previously established solidification model to simulate the solidification behavior and the microsegregation of the as cast material and it adopted a 1-D Pseudo Front Tracking (PFT) model for intergranular phase transformations, and Kampmann-Wagner Numerical (KWN) type model for intragranular precipitate kinetics. As shown in Table 6.2, this experimental results on area fraction of the constituent particles during soaking at 600oC are compared with simulation results by Du et al. (2011) for the same alloy and same thermal history (i.e. heating rate, soaking temperature and time). Both of the simulation and the measurement reveal the trend that the fraction is progressively increasing with the increase of the soaking time. The agreement between the measured and simulated values is satisfactory considering the error bar associated with the measurement. Table 6-2  The SEM data and simulated (Du et al. 2011) volume fraction of the constituent particles Sample As-cast 0min  at 600oC 20 min  at 600oC 1 h  at 600oC 24h  at 600oC SEM data (%) 2.0±0.5 2.8±0.6 2.8±0.4 3.4±0.4 3.9±0.2 Simulation (%) 2.0 2.4 2.6 2.9 3.8 During extrusion, the constituent particles in the as-cast materials which are originally rod/plate like eutectic phase are fragmented and distributed randomly in  85  the Al-matrix while the constituent particles in the homogenized materials are aligned along the extrusion direction with no or very few fragmentation of the particles. 6.1.2 Dispersoids There was no evidence of dispersoids in the as-cast material. A homogenization condition of 8h@500oC gives a large density of fine dispersoids. For a homogenization condition of 8h@550oC, there are fewer dispersoids and a Precipitate Free Zone adjacent to the dendrite arm appears. A homogenization condition of 24h@600oC shows very few dispersoids. The effect of these dispersoids on the recrystallization behavior during extrusion will be considered in section 6.2. Optical observations show that those dispersoids still exist in extruded material. The optical observations in the current work should be considered preliminary. Further detailed work using TEM is currently underway. Preliminary result has shown that the dispersoids are α-Al(Mn,Fe)Si phase with a bcc crystal structure with a size distribution of 50nm-225nm(Roumina 2010). 6.2 Effect of extrusion conditions on as-extruded microstructure To investigate the microstructure evolution during extrusion, it is useful to summarize the time scheme for extrusion, then the formation of the surface coarse grain structure. The effect of homogenization treatment on grain structure will also be discussed.   86  6.2.1 Time scheme for extrusion and the through thickness variation in processing The time scheme for extrusion is shown in Figure 6.2. It can be seen from this figure that the front and the butt part of the extrudate has been avoided when sampling. After exited from the die, it took approximately two seconds for the extrudate to reach the water quench. The possibility of significant microstructural evolution during this time cannot be ignored.  Figure 6-2 Time scheme for extrusion Investigation by Mahmoodkhan (2010) using a finite element model has found that for extrusion ratio of 130 and current die geometry the strain at the surface of the extrusion can be several times larger than the average strain. Further, the strain rate and temperature of the surface layer can be significantly different from the center. Polarized Light Optical Micrographs have confirmed that all extrudates exhibit thickness-dependent microstructure profile, which is due to the  87  inhomogeneous deformation and temperature history the extrudates have experienced. The homogenization treatments before extrusion are also found to have significant effect on the as-extruded microstructure. 6.2.2 Coarse surface grain region Micrographs from AD samples (i.e. the surface perpendicular to the extrusion axis) for all extrudates are displayed in Figure 6.3. By combining those micrographs and the micrographs displayed in section 5.3 (Figure 5.23, Figure 5.25, Figure 6.26 and Figure 6.27), it is found that except for cases that had been homogenized at 600oC for 24h before extrusion, they all reveal coarse surface grains, which has been called peripheral coarse grain (PCG) structure by some researchers(Van Geertruyden et al. 2005; Jäger et al.2009; Birol 2010). As shown in Figure 6.3, it is found that, in general, the extrusion temperature of 550oC gives larger depth of coarse surface grain regions than 400oC, and also an overall larger portion of coarse surface grain region is observed for AA3003 alloy than AA3102 alloy.  88   Figure 6-3 Polarized Optical Micrographs of AD view of AA3003 and AA3102 extrudates  89  It is also noticed that very fine (in the order of microns), equiaxed grains are found outside of the coarse grain surface region(as shown in Figure 5.25 and Figure 5.27, homogenized 8h@500oC and 8h@550oC, for extrusion temperature of 550oC of AA3003 and AA3102 respectively). The reason for this fine grain layer is not yet completely understood, however investigation from FEM simulation (Mahmoodkhan 2010) does show that the temperature and strain peak for extrusion trial implemented at 550oC, is actually located at the subsurface, due to the lower temperature of the die(i.e. 530oC) than billet(i.e. 550oC). To summarize, the surface coarse grain structure is manifested as a result of a combination of processing parameters such as strain, strain rate and temperature as well as alloy chemistry. No decisive conclusion for the mechanisms of surface grain formation could be made from the current work. Further work is proposed on extrudate from low extrusion ration trials to investigate the recovered subgrain structure and to investigate the half-extruded material to justify whether they are results of dynamic or static recrystallization. A practical solution can be found to avoid the formation of surface coarse grain structure in these alloys by modifying the chemistry and allowing long time homogenization at high temperature. 6.2.3 Effect of homogenization treatment The effect of homogenization treatment on the as-extruded microstructure is to a large extent due to the effect of second phase particles. A bimodal size distribution of particles is found in this material, i.e. constituent particles with an average size around 1μm and dispersoids of 50nm~300nm. As discussed in section  90  6.1, materials with different homogenization give different size distribution for the two types of particles. Micrographs in section 5.3 show that significant differences are found for different homogenization treatment in the size and morphology of the grains in the extrudate. Figure 6.4-Figure 6.7 show the grain size (with an error of ±10μm) estimated from micrographs of taper mounted samples. It is found except for the homogenization condition of 8h@500oC, all extrudates show a similar trend of grain size decrease from surface to center. Coarse surface grain structures are shown in the as-cast and homogenization condition of 8h@550oC. Grain structure in AA3102 alloy is more homogenous than AA3003 probably due to its lower dispersoid density (as suggested in Figure 5.12 and Figure 5.16). To summarize, homogenization treatment and Mn composition affect the as-extruded structure significantly. Although it is unclear whether the recrystallization occurred during the extrusion process or during the water quench, it appears that the fine dispersoids impeded the recrystallization and grain growth. Materials with large density of dispersoids show an irregular grain structure, partial recrystallization or very large grains in the material, while materials without dispersoids reveal an equiaxed and fine structure.  91   Figure 6-4 Grain size for AA3003 alloy with different homogenization treatments and extruded at 400oC (μm)  Figure 6-5 Grain size for AA3003 alloy with different homogenization treatments and extruded at 550oC (μm)  92   Figure 6-6 Grain size for AA3102 alloy with different homogenization treatments and extruded at 400oC(μm)  Figure 6-7 Grain size for AA3102 alloy with different homogenization treatments and extruded at 550oC (μm)   93  6.3 Physically-based flow stress model It is found by previous work (Kubiak 2010) that the Kocks and Chen constitutive model for flow stress is empirically successful in evaluating the effect of homogenization treatment on the flow stress of 3xxx alloys, and could be validated with extrusion trials results. However, the derived stress exponent values, n, are much larger than the theoretical value of 3. This part of the current work is continuation work of the constitutive behavior study of AA3xxx alloys. The main goal of the work is to examine the effect of the constituent particles on the constitutive behavior of AA3xxx alloys by using a low iron AA3102 alloy which has very low constituent particles fraction, thus to further verify the effect of constituent particles on the model parameters if there is any. Model calculations for low iron AA3102 are presented in section 6.3.1 and solved modeling parameter are compared with those from other AA3xxx alloys and the discussion is shown in section 6.3.2. 6.3.1 The model Kocks and Chen’ constitutive equation(Kocks and Chen 1993) is shown in Eq.6.1. 𝜀̇ = 𝐴 �𝜎 𝜇 � 𝑛 𝜇𝑏3 𝑘𝑇 exp �−𝑄𝐷 𝑅𝑇 �                                                      (6.1) Where, σ is the stress, μ is the temperature dependent shear modulus, n is the stress exponent, ε  is the strain rate, k is the Boltzmann’s constant, T is the deformation temperature, b is the temperature dependent magnitude of the Burgers  94  vector, QD is the activation energy for diffusion of the diffusing species, R is the gas constant, and A is a pre-exponential constant. Upon rearrangement of the equation to Eq.6.2, the x-axis component is given as in Eq. 6.3. log �𝜎 𝜇 � = 1 𝑛 log � 𝜀̇ 𝑠−1 𝑘𝑇 𝜇𝑏3 exp𝑄𝐷 𝑅𝑇 � − 1 𝑛 log𝐴                             (6.2) log � 𝜀̇ 𝑠−1 𝑘𝑇 𝜇𝑏3 exp𝑄𝐷 𝑅𝑇 � = 𝑥                                                             (6.3)               In the current work, it is assumed that the diffusion of Mn is the rate controlling parameter. The activation energy QD for diffusion of Mn in aluminum is taken as 211.4 kJ/mol (Du and Jacot 2005). The calculation method and results for the temperature dependent shear modulus and Burger’s vector are shown in Appendix D. 6.3.2 Constitutive modeling of low iron AA3102 The fit of the current experimental data for flow stress and yield stress of the as-cast and homogenized low iron AA3102 alloy are plotted in Figure 6.8, Figure 6.9 and Figure 6.10. It is seen that a good fit to the data is achieved for all the three cases.  95  9 10 11 12 13 14 15 16 10-3 10-2 10-1 flow stress, as-cast yield stress, as-cast model fit µ σ       − RT Q b kT s Dexplog 31 µ ε  Figure 6-8 Constitutive modeling for low iron AA3102, as-cast 9 10 11 12 13 14 15 16 10-4 10-3 10-2 flow stress, 8h@500oC yield stress, 8h@500oC model fit µ σ       − RT Q b kT s Dexplog 31 µ ε  Figure 6-9 Constitutive modeling for low iron AA3102, homogenized 8h@500oC  96  9 10 11 12 13 14 15 16 10-4 10-3 10-2 flow stress, 24h @600oC yield stress, 24h @600oC model fit µ σ       − RT Q b kT s Dexplog 31 µ ε  Figure 6-10 Constitutive modeling for low iron AA3102, homogenized 24h@600oC 6.3.3 Discussion on the fit parameters for the flow stress model                           The model fitting parameters (A and n) are determined by the slope and the intercept from linear fitting       µ σlog versus       − RT Q b kT s Dexplog 31 µ ε . Table 6.3 displays the comparison of model parameters for different alloys at different homogenization condition.   97  Table 6-3 Comparison of fit parameters for low iron AA3102 alloys with those from previous AA3003 and AA3102 alloys (Kubiak 2009); note the data from Kubiak is represented by the shadowed boxes. Alloys Homogenization Treatment Constituent Particles (Volume %) Dispersoids Mn in SS1(wt%)   n A T (oC) t(h) AA3003 As-cast 2.5(SEM) None N.A 8.2 4.7E+34 AA3003 500 8 2.8(SEM) High 0.24 10.2 9.0E+40 AA3003 600 24 3.9(SEM) Low/Few 0.51 8.1 1.1E+36 AA3012 As-cast N.A None N.A 7.1 8.9E+32 AA3102 500 8 1.8 Low/Few 0.057 8.3 1.0E+37 AA3102 600 24 N.A None 0.13 7.0 3.2E+33 Low iron 3102 As-cast 0.6 None N.A 7.3 3.0E+41 Low iron 3102 500 8 0.5 None 0.14 7.6 1.5E+35 Low iron 3102 600 24 0.5 None 0.22 7.2 1.1E+34 The homogenization treatment affects the flow stress of the material by modifying the solute in solid solution and the dispersoids number density. Stress exponent “n” and pre-exponential “A” values quantify the effect of the homogenization. Therefore the application of this physically-based constitutive model provides a means of quantifying and comparing the effect of homogenization on flow stress over a range of temperature and strain rates. It can be seen from  1Value of Mn in SS for each condition is equilibrium solubility at selected temperature calculated from Thermo-Calc Software.  98  Table 6.3 that higher “n” values (n=10.2, 8.3, 7.6) are always found for 500oC, 8 hour homogenization conditions which have comparably higher dispersoids number density. For materials with no/few dispersoids, i.e. 24h@600oC homogenized AA3003, AA3102, low iron AA3102, “n” value slightly increases with the Mn content in solid solution. Previous work has also found that A-1/n is linearly related to the solute concentration in solid solution (Kubiak 2009). The results show that the stress exponents (n) for the low iron 3102 are actually quite close to those gained from 3102 alloy, which suggests that the constituent particles do not have a large influence on the stress dependence of AA3xxx alloy. Current work and previous work all present “n” values that are well above 3. Therefore further investigation is required to understand the high stress exponent value. Nevertheless, the role of constituent particles has been eliminated as a source to the large n value.   99  Chapter 7 Summary and Conclusions 7.1 Summary Microstructure development during extrusion manufacturing of commercial AA3003(Al1.27Mn0.54Fe0.1Si) and AA3102(Al0.75Mn0.52Fe0.1Si) alloys were investigated in as-cast, as-homogenized, and as-extruded state to understand the effect of process parameters on as-extruded microstructure. Microstructure characteristics studied included the grain structure, inter-granular constituent particles and the intra-granular dispersoids. The size and area fraction of the constituent particles were quantitatively measured using back-scattered SEM micrographs. The phase of the constituent particle was found to be Al6(Mn,Fe). Optical microscopy was applied to qualitatively reveal the dispersoids in those conditions. The largest number density of dispersoids was found at homogenization conditions of 8h@500oC. There were very few/none dispersoids at homogenization conditions of 24h@600oC. The effect of heating and soaking on constituent particles evolution until 600oC was also investigated on samples heated to and  soaked at 600oC for different times(from 0h to 48h). The eutectic structures started to break up into smaller particles around 400oC during heating. When soaking at 600oC, the area fraction increased while number density dropped rapidly at the first 1 h, due to the increase of Mn/Fe in solid solution which was mainly contributed by fast dissolution of the fine dispersoids. After long time homogenization, when the dispersoids were mostly  100  dissolved, the size and number density change was mainly contributed by Mn/Fe diffused from solid solution and thus slowed down. Extrusion trials on the as-cast and as-homogenized (i.e. 8h@500 oC, 8h@550 oC and 24h@600oC) AA3003 and AA3102 billets were performed at the Rio Tinto Alcan Research and Development facility in Jonquiere, Quebec. Extrusion temperatures of 400oC and 550oC and extrusion ratio of 130 were chosen. Both Tapered samples and AD samples were cut from the extrudates to give a 3D microstructure profile. The through-thickness profile was observed for each extrudate due to inhomogeneous thermo-mechanical history that the material has gone through. Furthermore, the homogenization treatments showed significant effects on the as-extruded microstructure. It was observed that the most inhomogeneous structures were those had been homogenized for 8h@500oC, which had a high density of dispersoids in the material.  The finest and most homogeneous structures were found for samples homogenized 24h@600oC. Compression tests were performed on the as-cast and as-homogenized low iron AA3102 using the Gleeble 3500 thermomechanical simulator at various conditions (strain rate from 0.1/s to 10/s, temperature from 400oC to 600oC). Flow stress and yields stress were measured and fit to the Kocks and Chen model with a good correlation. The values for the flow stress exponent, n and the constant, A were compared with previous work to determine the effect of constituent particles on model parameters. It was found that the constituent particles do not affect the flow stress of 3xxx alloy and the stress exponent, n was still higher than theoretically  101  expected.  The main conclusions of this work are as follows: 1. During homogenization, Mn is transported from the primary Al to the constituent particles. Break-up of the eutectic phase, spheroidization and coarsening of constituent particles take place. The constituent particles are found to be the Al6(Mn,Fe) phase. Except for the extrusion with the as-cast billet, no fragmentation of the constituent particles occurs during extrusion.  2. The homogenization treatment, thermo-mechanical history, and Mn content all affect the as-extruded microstructure in terms of grain morphology and grain size distribution. In presence of a high density of dispersoids, recrystallization was either incomplete or showed a large grain size. Long time homogenization time at 600oC is recommended to obtain homogeneous and fine as-extruded microstructure.  3. A surface coarse grain layer is found at all extrudates except for those that have been homogenized for 24h@600oC.  The formation of the surface coarse grain structure is a result of the combination of processing parameters such as strain, strain rate and temperature as well as alloy chemistry. Very fine grains outside of surface coarse grain are also found in most extrudates for extrusion temperature of 550oC.  4. The compression behavior of a low iron 3102 alloy is studied and the results fit well with Kocks and Chen model. The effect of fraction of constituent particles on the flow stress is negligible and thus is eliminated as an explanation for  102  the high stress exponent. 7.2 Future work Further microstructure studies on extrudates from low extrusion ratio trials are suggested. If the conditions can be found to produce unrecrystallized fibrous structure, this would be a good way to study 1) subgrain structure, 2) the way that grains deforms during extrusion process and 3) the recrystallization mechanism during the extrusion (dynamic/static). Optical observation only provided qualitative examination of the dispersoids. Therefore quantitative investigation on size/fraction of dispersoids with TEM for both as-homogenized and as-extruded samples is suggested to evaluate if there is any potential precipitation/dissolution of dispersoids during the extrusion. Size distribution of the dispersoids could also be used to quantify their effect on the recrystallization kinetics and to study the relation between dispersoids particle size distribution and the recrystallized grain size from the extrudate.  103  References  Alcan Company Website(2010), 'History of Rio Tinto Alcan', <http://www.riotintoalcan.com/ENG/whoweare/28.asp>, accessed 11-26. Alexander, D. T. L. and Greer, A. L. (2002), 'Solid-state intermetallic phase tranformations in 3XXX aluminium alloys', Acta Materialia, 50 (10), 2571-83. Aluminum Association of Canada (2010), 'Aluminum Association of Canada', <http://www.aia.aluminium.qc.ca>, accessed 10-13. Ashby, M. F. (1972), 'A first report on deformation-mechanism maps', Acta Metallurgica, 20 (7), 887-97. Bakshi, P. K. and Kashyap, B. P. (1995), 'Stress-strain rate relations for high-temperature deformation of two-phase Al-Cu alloys', Journal of Materials Science, 30 (20), 5065-72. Belov, N. A., Eskin, D. G. and Aksenov, A. A. (2005), 'Alloys of the Al-Fe-Mn-Si System', Multicomponent Phase Diagrams (Oxford: Elsevier), 1-46. Birol, Y. (2008), 'Recrystallization of a supersaturated Al-Mn alloy', Scripta Materialia, 59 (6), 611-14. Blum, W., Zhu, Q., Merkel, R., and McQueen, H. J. (1996), 'Geometric dynamic recrystallization in hot torsion of Al-5Mg-0.6Mn(AA5083)', Materials Science and Engineering: A, 205 (1-2), 23-30. Chan, H. M. and Humphreys, F. J. (1984), 'Effect of particle simulated nucleation on orientation of recrystallized grains', Acta Materialia, 32 (1984), 235. Chang, C. P., Sun, P. L., and Kao, P. W. (2000), 'Deformation induced grain boundaries in commercially pure aluminium', Acta Materialia, 48 (13), 3377-85. Chen, S. R., Stout, M. G., Kocks, U. F., MacEwen, S. R., and Beaudoin, A. J. (1998), 'Constitutive modeling of a 5182 aluminum as a function of strain rate and temperature', Conference: 1998 Minerals, Metals and Materials Society (TMS) fall meeting, Rosemont, IL (United States). Davenport, S. B. and Higginson, R. L. (2000), 'Strain path effects under hot working: an introduction', Journal of Materials Processing Technology, 98 (3), 267-91.  104  Davis, J. R. (1993), ASM Specialty Handbook: Aluminum and Aluminum Alloys (American Society for Metals). Dehmas, M., Weisbecker, P., Geandier, G., Archambault, P., and Aeby-Gautier, E. (2005), 'Experimental study of phase transformations in 3003 aluminium alloys during heating by in situ high energy X-ray synchrotron radiation', Journal of Alloys and Compounds, 400 (1-2), 116-24. Doherty, R. D., Hughes, D. A. Humphreys, F. J., Jonas, J. J., Juul Jensen, D., Kassner, M. E., King, W. E., McNelley, T. R., McQueen, H. J., and Rollett, A. D. (1997), 'Current issues in recrystallization: a review', Materials Science and Engineering: A, 238 (2), 219-74. Dons, L (2001), 'The Alstruc homogenization model for industrial aluminum alloys', Journal of Light Metals, 1 (2), 133-49. Du, Q., Geng, Y., Poole, W. J., and Wells, M. A. (2011), 'Microstructure evolution during homogenization heat treatment of an AA3003 Alloy', (Department of Materials Engineering; Vancouver: The University of British Columbia), unpublished manuscript, 51. Du, Q. and Jacot, A. (2005), 'A two-dimensional microsegregation model for the description of microstructure formation during solidification in multicomponent alloys: Formulation and behaviour of the model', Acta Materialia, 53 (12), 3479-93. Embury, J. D., Poole, W. J., and Koken, E. (1992), 'Some views on the influence of strain path on recrystallization', Scripta Metallurgica et Materialia, 27 (11), 1465-70. Fernández, A. I., Uranga, P., López, B., and Rodriguez-Ibabe, J. M. (2003), 'Dynamic recrystallization behavior covering a wide austenite grain size range in Nb and Nb-Ti microalloyed steels', Materials Science and Engineering: A, 361 (1-2), 367-76. Ferry, M. and Munroe, P. R. (1995), 'The effect of subgrain size on the static recrystallization behaviour of an aluminium-based metal-matrix composite', Scripta Metallurgica et Materialia, 33 (6), 857-62.  105  Fickett, F. R. (1971), 'Aluminum-1. A review of resistive mechanisms in aluminum', Cryogenics, 11 (5), 349-67. Frost, H. J. and Ashby, M. F. (1982), Deformation mechanism maps: the plasticity and creep of metals and ceramics (Oxford: Pergamon Press). Gandin, Ch A. and Jacot, A. (2007), 'Modeling of precipitate-free zone formed upon homogenization in a multi-component alloy', Acta Materialia, 55 (7), 2539-53. Gholinia, A., Humphreys, F. J., and Prangnell, P. B. (2002), 'Production of ultra-fine grain microstructures in Al-Mg alloys by coventional rolling', Acta Materialia, 50 (18), 4461-76. Gladman, T. (1966), 'On the Theory of the Effect of Precipitate Particles on Grain Growth in Metals', Proceedings of the Royal Society of London. Series A, Mathematical and Physical Sciences, 294 (1438), 298-309. Gladman, T. (1992), 'Second phase particle distribution and secondary recrystallisation', Scripta Metallurgica et Materialia, 27 (11), 1569-73. Gottstein, G. and Shvindlerman, L. S. (2010), 'On the retardation of grain boundary motion by small particles', Scripta Materialia, 63 (11), 1089-91. Gourdet, S. and Montheillet, F. (2000), 'An experimental study of the recrystallization mechanism during hot deformation of aluminium', Materials Science and Engineering: A, 283 (1-2), 274-88. Gourdet, S. and Montheillet, F. (2002), 'Effects of dynamic grain boundary migration during the hot compression of high stacking fault energy metals', Acta Materialia, 50 (11), 2801-12. Hansen, N. and Bay, B. (1981), 'Initial stages of recrystallization in aluminium containing both large and small particles', Acta Metallurgica, 29 (1), 65-77. Harun, A., Holm, E. A., Clode, M. P., and Miodownik, M. A. (2006), 'On computer simulation methods to model Zener pinning', Acta Materialia, 54 (12), 3261-73. Hasegawa, T. and Kocks, U. F. (1979), 'Thermal recovery processes in deformed aluminum', Acta Metallurgica, 27 (11), 1705-16.  106  Hatch, J. E. (1984), Aluminum: properties and physical metallurgy (American Society for Metals). Heilmann, P., Clark, W. A. T., and Rigney, D. A. (1983), 'Orientation determination of subsurface cells generated by sliding', Acta Metallurgica, 31 (8), 1293-305. Hillert, M. (1965), 'On the theory of normal and abnormal grain growth', Acta Metallurgica, 13 (3), 227-38. Hillert, M. (1984), 'On the estimation of the zener drag on grain boundaries', Scripta Metallurgica, 18 (12), 1431-32. Holm, E. A., Miodownik, M. A., and Rollett, A. D. (2003), 'On abnormal subgrain growth and the origin of recrystallization nuclei', Acta Materialia, 51 (9), 2701-16. Hu, H., Zhen, L., Zhang, B., Yang, L., and Chen, J. (2008), 'Microstructure characterization of 7050 aluminum alloy during dynamic recrystallization and dynamic recovery', Materials Characterization, 59 (9), 1185-89. Huang, X., Zhang, H., Han, Y., Wu, W., and Chen, J. (2010), 'Hot deformation behavior of 2026 aluminum alloy during compression at elevated temperature', Materials Science and Engineering: A, 527 (3), 485-90. Huang, Y. and Humphreys, F. J. (1997), 'Transient dynamic recrystallization in an aluminium alloy subjected to large reductions in strain rate', Acta Materialia, 45 (11), 4491-503. Huang, Y. and Humphreys, F. J. (2000), 'Subgrain growth and low angle boundary mobility in aluminium crystals of orientation {110}<001>', Acta Materialia, 48 (8), 2017-30. Humphreys, F. J. (1977), 'The nucleation of recrystallization at second phase particles in deformed aluminium', Acta Metallurgica, 25 (11), 1323-44. Humphreys, F. J. and Ardakani, M. G. (1996), 'Grain boundary migration and Zener pinning in particle-containing copper crystals', Acta Materialia, 44 (7), 2717-27. Humphreys, F. J. and Hatherly, M. (2004), Recrystallization and Related Annealing Phenomena (Second Edition) (Elsevier).  107  Ishikawa, T., Sano, H., Yoshida, Y., Yukawa, N., Sakamoto, J., and Tozawa, Y. (2006), 'Effect of Extrusion Conditions on Metal Flow and Microstructures of Aluminum Alloys', CIRP Annals - Manufacturing Technology, 55 (1), 275-78. Jäger, A., Heilmann, M., Misiolek, W. Z., Schikorra, M., and Tekkaya, A. E. (2009), 'Influence of cooling rate on distortion and microstructure in extrusion of Al-Mg-Si alloys', International journal of material forming, 2(1), 81-84. Jazaeri, H. and Humphreys, F. J. (2004), 'The transition from discontinuous to continuous recrystallization in some aluminium alloys: II - annealing behaviour', Acta Materialia, 52 (11), 3251-62. Jones, M. J. and Humphreys, F. J. (2003), 'Interaction of recrystallization and precipitation: The effect of Al3Sc on the recrystallization behaviour of deformed aluminium', Acta Materialia, 51 (8), 2149-59. Kaibyshev, R., Mazurina, I., and Gromov, D. (2006), 'Mechanisms of grain refinement in aluminum alloys in the process of severe plastic deformation', Metal Science and Heat Treatment, 48 (1), 57-62. Kaibyshev, R., Shipilova, K., Musin, F., and Motohashi, Y. (2005), 'Continuous dynamic recrystallization in an Al-Li-Mg-Sc alloy during equal-channel angular extrusion', Materials Science and Engineering: A, 396 (1-2), 341-51. Kassner, M. E. and Barrabes, S. R. (2005), 'New developments in geometric dynamic recrystallization', Materials Science and Engineering: A, 410-411, 152-55. Kassner, M. E., Myshlyaev, M. M., and McQueen, H. J. (1989), 'Large-strain torsional deformation in aluminum at elevated temperatures', Materials Science and Engineering: A, 108, 45-61. Kaufman, J. G. (2000), Introduction to Aluminum Alloys and Tempers (American Society for Metals). Kayser, T., Klusemann, B., Lambers, H. G., Maier, H. J., and Svendsen, B. (2010), 'Characterization of grain microstructure development in the aluminum alloy EN AW-6060 during extrusion', Materials Science and Engineering: A, 527 (24-25), 6568-73.  108  Kim, B. N. and Kishi, T. (1999), 'Finite element simulation of Zener pinning behavior', Acta Materialia, 47 (7), 2293-301. Kocks, U. F. (2001), 'Realistic constitutive relations for metal plasticity', Materials Science and Engineering: A, 317 (1-2), 181-87. Kocks, U. F. and Chen, S. R. (1993), 'Constitutive Laws for Deformation and Dynamic Recrystallization in Cubic Metals’, Aspects of high temperature deformation and fracture in crystalline materials (Nagoya, Japan: Japan Institute of Metals), 593-600. Kubiak, A. D. (2009), 'Effect of homogenization on high temperature deformation behavior of AA3xxx aluminum alloys', (University of British Columbia). Lacaze, J., Tierce, S., Lafont, M.C., Thebault, Y., Mankowski, G., Blanc, C., Robidou, H., Vaumousse, D., and Daloz, D. (2005), 'Study of the microstructure resulting from brazed aluminium materials used in heat exchangers', Materials Science and Engineering: A, 413-414, 317-21. Laue, K. and Stenger, H. (1981), Extrusion: processes, machinery, tooling (American Society for Metals). Li, W. B. and Easterling, K. E. (1990), 'The influence of particle shape on zener drag', Acta Metallurgica et Materialia, 38 (6), 1045-52. Li, Y. J. and Arnberg, L. (2003a), 'Evolution of eutectic intermetallic particles in DC-cast AA3003 alloy during heating and homogenization', Materials Science and Engineering: A, 347 (1-2), 130-35. Li, Y. J. and Arnberg, L. (2003b), 'Quantittive study on the precipitation behavior of dispersoids in DC-cast AA3003 alloy during heating and homogenization', Acta Materaialia, 51 (12), 3415-28. Liu, W. and Morris, J. (2005), 'Evolution of recrystallization and recrystallization texture in continuous-cast AA 3015 aluminum alloy', Metallurgical and Materials Transactions: A, 36 (10), 2829-48. Liu, Y. and Patterson, B. R. (1993), 'Particle volume fraction dependence in Zener drag', Scripta Metallurgica et Materialia, 29 (8), 1101-06.  109  Lodgaard, L. and Ryum, N. (2000), 'Precipitation of dispersoids containing Mn and/or Cr in Al-Mg-Si alloys', Materials Science and Engineering: A, 283, 144-52. Lubarda, V. A. 'On the effective lattice parameter of binary alloys', Mechanics of Materials, 35 (1-2), 53-68. Mahmoodkhani, Y., Wells, M. A., Parson, N., Geng, Y., and Poole, W. J. (2010), 'Mathematical Modelling of the Extrusion of AA3xxx Aluminum Alloys', Proceedings of the 12th International Conference on Aluminum Alloys (Yokohama, Japan: The Japan Institute of Light Metals). Martins, J., Carvalho, A., and Padilha, A. (2009), 'Microstructure and texture assessment of Al–Mn–Fe–Si (3003) aluminum alloy produced by continuous and semicontinuous casting processes', Journal of Materials Science, 44 (11), 2966-76. Mathew, E. V., Ramachandran, T. R., Gupta, K. P., and Das, S. (1984), 'Homogenization of commercial Al-Mn alloys', Journal of Materials Science Letters, 3 (7), 605-10. McQueen, H. J. (1988), 'Initiating nucleation of dynamic recrystallization, primarily in polycrystals', Materials Science and Engineering: A, 101, 149-60. McQueen, H. J. and Ryan, N. D. (2002), 'Constitutive analysis in hot working', Materials Science and Engineering A, 322 (1-2), 43-63. Meyers, M. A., Benson, D. J., Vöhringerb, O., Kad, B. K., Xue, Q., and Fu, H. H. (2002), 'Constitutive description of dynamic deformation: physically-based mechanisms', Materials Science and Engineering: A, 322 (1-2), 194-216. Morris, P. L. and Duggan, B. J. (1978), 'Precipitation and recrystallization in an Al-1-8%Mn alloy', Metal Science, 12, 1-7. Nes, E., Ryum, N., and Hunderi, O. (1985), 'On the Zener drag', Acta Metallurgica, 33 (1), 11-22. Nes, E. (1997), 'Modelling of work hardening and stress saturation in FCC metals', Progress in Materials Science, 41 (3), 129-93.  110  Parson, N. and Ramanan, R. (2008), 'Optimising AA3003 for Extrudability and Grain Size Control', The Conference for Innovations in Aluminum Extrusion (Orlando, Florida USA.). Pettersen, T., Holmedal, B., and Nes, E. (2003), 'Microstructure development during hot deformation of aluminum to large strains', Metallurgical and Materials Transactions: A, 34 (12), 2737-44. Ponge, D., Bredehoft, M, and Gottstein, G. (1997), 'Dynamic recrystallization in high purity aluminum', Scripta Materialia, 37 (11), 1769-75. Rabkin, E. (1998), 'Zener drag in the case of anisotropic grain boundary energy', Scripta Materialia, 39 (12), 1631-37. Ravichandran, N. and Prasad, Y. (1991), 'Dynamic recrystallization during hot deformation of aluminum: A study using processing maps', Metallurgical and Materials Transactions: A, 22 (10), 2339-48. Roumina, R. (2010), 'TEM Study of AA3003 Alloy', unpublished report(Vancouver: Department of Materials Engineering,  The Univeristy of British Columbia). Rouns, T. N. (1998), 'Aluminum alloys for packaging III', in S.K.Das (ed.), TMS Annual Meeting (San Antonio, Texas), 3-20. Russ, C. J. (1990), 'Computer assisted microscopy : the measurement and analysis of images'(Newyork and London: Plenum Press),453. Saha, P.  K. (2000), 'Aluminum Extrusion Technology' (American Society for Metals). Saha, P. K. (1998), 'Thermodynamics and tribology in aluminum extrusion', Wear, 218 (2), 179-90. Schikorra, M., Donati, L., Tomesani, L., and Tekkaya, A. (2007), 'Microstructure analysis of aluminum extrusion: grain size distribution in AA6060, AA6082 and AA7075 alloys', Journal of Mechanical Science and Technology, 21 (10), 1445-51. Schikorra, M., Donati, L., Tomesani, L., and Tekkaya, A. E. (2008), 'Microstructure analysis of aluminum extrusion: Prediction of microstructure on AA6060 alloy', Journal of Materials Processing Technology, 201 (1-3), 156-62.  111  Sidor, J., Petrov, R. H., and Kestens, L. A. I. (2010)'Deformation, recrystallization and plastic anisotropy of asymmetrically rolled aluminum sheets', Materials Science and Engineering: A, In Press, Corrected Proof. Somani, M. C., Birla, N. C., Prasad, Y. V. R. K., and Singh, V. (1995)  'Microstructural validation of processing maps using the hot extrusion of P/M Nimonic AP-1 superalloy', Journal of Materials Processing Technology, 52 (2-4), 225-37. Song, X and Rettenmayr, M. (2007), 'Modeling recrystallization in a material containing fine and coarse particles', Computational Materials Science, 40 (2), 234-45. Thermo-Calc Software Website(2010), 'Welcome to Thermo-Calc Software', < http://www.thermocalc.com>, accessed 01-12. Tibballs, J. E., Horst, J. A., and Simensen, C. J. (2001), 'Precipitation of α-Al(Fe,Mn)Si from the melt', Journal of Materials Science, 36 (4), 937-41. Van Geertruyden, W., Browne, H., Misiolek, W., and Wang, P. (2005), 'Evolution of surface recrystallization during indirect extrusion of 6xxx aluminum alloys', Metallurgical and Materials Transactions: A, 36 (4), 1049-56. Van Haaften, W., Magnin, B., Kool, W., and Katgerman, L. (2002), 'Constitutive behavior of as-cast AA1050, AA3104, and AA5182', Metallurgical and Materials Transactions: A, 33 (7), 1971-80. Vandermeer, R. A. (2005), 'Microstructural descriptors and the effects of nuclei clustering on recrystallization path kinetics', Acta Materialia, 53 (5), 1449-57. Vandermeer, R. A. and Juul Jensen, D. (1998), 'The Migration of High Angle Grain Boundaries during Recrystallization', Interface Science, 6 (1), 95-104. Vandermeer, R. A. and Juul Jensen, D. (2001), 'Microstructural path and temperature dependence of recrystallization in commercial aluminum', Acta Materialia, 49 (11), 2083-94. Vandermeer, R. A. and Juul Jensen, D. (2003), 'Recrystallization in hot vs cold deformed commercial aluminum: a microstructure path comparison', Acta Materialia, 51 (10), 3005-18.  112  Vatne, H.E., Engler, O., and Nes, E. (1997), 'Influence of particles on recrystallisation textures and microstructures of aluminium alloy 3103', Materials Science and Technology, 13, 93-102. Warmuzek, M. and Sieniawski, J. (2004), 'Influence of the heat treatment on the precipitation of the intermetallic phases in commercial AlMn1FeSi alloy', Journal of Materials Processing Technology, 157-158, 624-32. Woldman, N. E. and Frick, J. P. (2000), Woldman's engineering alloys (American Society for Metals). Wright, R. N. and Paulson, M. S. (1998), 'Constitutive equation development for high strain deformation processing of aluminum alloys', Journal of Materials Processing Technology, 80-81, 556-59. Yamagata, H. (1992), 'Dynamic recrystallization of single-crystalline aluminum during compression tests', Scripta Metallurgica et Materialia, 27 (6), 727-32. Zhang, H., Konopleva, E. V., and McQueen, H. J. (2001), 'Effects of Mn dispersoid on hot working of Al-1Mn', Materials Science and Engineering: A, 319-321, 711-15. Zhu, H., Zhang, X., Couper, M. J., and Dahle, A. K. (2009), 'Effect of primary intermetallic particles on surface microstructure and appearance of aluminium extrusions', Materials Chemistry and Physics, 113 (1), 401-06.   113  Appendices Appendix A Repeatability of compression results and determination of yield stress and flow stress A.1 Repeatability of stress-strain curves Before starting any Gleeble tests on the experimental low iron AA3102 alloy, tests were repeated from the previous study of Kubiak (2009). This was done to ensure that the current Gleeble set up was stable and would give consistent results to those of Kubiak (2009). The stress-strain curves from the current study and that of Kubiak were shown in Figure A.1. The maximum difference was below 3MPa. ε 0.0 .1 .2 .3 .4 .5 .6 σ /M Pa 0 20 40 60 80 100 120 140 Test Data(400oC, 10/s ), as-cast Previous Data(400oC, 10/s), as-cast Test Data(600oC, 0.1/s), 630oC homogenized for 8h Previous Data(600oC, 0.1/s), 630oC homogenized for 8h  Figure A-1 Repeated tests on AA3003 alloys  114  Multiple tests were also conducted on selected compression condition to ensure the repeatability of the Gleeble tests. For example Figure A.2 shows two tests on as cast low iron AA3102. ε 0.0 .1 .2 .3 .4 .5 .6 σ /M pa 0 5 10 15 20 25 30 1st run, ( 600oC, 1/s ) 2nd run, ( 600oC, 1/s )  Figure A-2 Repeated tests on as-cast low iron AA3102 alloys A.2 Determination of yield stress and steady state flow stress By collecting the stress-strain data from the compression tests, the stress-strain curves are achieved.  Steady state flow stress is solved by taking the average of stress data from 𝜀=0.4 to 0.6. An example is shown in Figure A.4. Given the method and nature of Gleeble tests, it gives a more accurate number for flow stress than yield stress.  115  ε 0.0 .2 .4 .6 .8 σ /M Pa 0 5 10 15 20 25 30  Figure A-3 Determination of steady state flow stress for a 600oC, 24h homogenized low iron AA3102, at compression condition of 600oC, and 10s-1    116  Appendix B Summary of compression tests results Table B-1 Flow stress and yield stress value (MPa) for as cast low iron AA3102  400oC 500oC 600oC Flow stress Yield Stress Flow stress Yield Stress Flow stress Yield Stress 0.1/s 27.6 13.7 13.4 7.7 7.9 3.4 1/s 38.3 18.3 16.3 9.2 13.0 6.1 10/s 52.1 27.5 29.0 12.6 16.0 7.9 Table B-2 Stress value (MPa) for homogenized low iron AA3102 Homogenization Condition Compression Condition Flow Stress Yield Stress 500oC, 8h 400oC, 10/s 49.9 25.7 500oC, 1/s 20.4 10.0 500oC,10/s 29.7 12.2 600oC, 0.1/s 8.1 5.9 600oC, 24h 400oC, 10/s 52.1 19.2 500oC, 1/s 20.2 9.6 500oC,10/s 27.5 12.3 600oC, 1/s 13.0 5.7 600oC, 10/s 15.8 6.2 600oC, 0.1/s 6.7 3.4   117  Appendix C Image analysis method and sensitivity test for constituent particles quantification C.1 Clemex Routine The following steps show the original routine file for constituent particles quantification. Figure C-5 shows a SEM image being processed and threshold peak information. Routine developed: Clear== All Grey threshold BPL1 range 158..255  (// picked up particles which has a grey degree between 158 to 255) Closing CIRCx1==BPL1 Extend   (//close the particle edge) Erode CIRC x1==BPL1 Extrend    ( //it extended circx1 in the last step, thus deduct circx1 here) Oject Transfer BPL1==None    (//ignore the small particle) Circular Diameter less than 0.4um Bridge Removal==BPL1   (//seperate two particles that overlaped) Pause Edit Kill BPL1   ( //kill the micro marker) Relative Measure==RELM2  (//give the area fraction of the constituent particles) Area Percent, BPL1 relative to field  Object Measures(BPL1)==OBLM1 (Aspect Ratio, Roundness, Length, Circular Diameter, Area Export Data to excel (//need to manually calculate the number density by dividing the total counts over field area)  118   Figure C-4 SEM image being processed and threshold peak C.2 Sensitivity test on threshold value  Error for this method raised from a) the nature of the particles that they were spatially distributed and buried in sub-surface, which cause the edge of the particle blurred. b)  Threshold values picked since they could result in obvious difference upon the result. c) Observer’s subjective judge on choosing the threshold values. To minimize the error and get trustful results, sensitivity tests have been done on a series of micrographs for selected samples for liable results and error bars. The threshold range is selected as it is able to pick up all the particles yet still doesn’t exceed the edge of the particle. Around ten images for each sample are taken from SEM and each of the micrographs is analyzed using method illustrated in Table B.1. Table B.2 shows an example of the sensitivity test result for 3003 sample homogenized at 600oC for 24h. For this condition 4.07 is taken as area fraction with an error bar of ±0.18, gained from (4.27-3.91)/2. 119  Table C-3 Selected range of threshold value for one micrograph from homogenized sample (600oC, 48h) Threshold Value 146 144 142 140 Area Fraction Result 3.66 3.79 3.91 4.05  Table C-4 Selected range of threshold value for all micrographs from homogenized sample (600oC, 48h) No. of images 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Mean* Average* 4.62 3.85 4.43 4.45 3.91 3.79 3.83 3.79 4.02 3.89 4.21 4.19 4.1 4 4 3.94 4.23 4.07 Min* 4.41 3.66 4.22 4.25 3.72 3.62 3.67 3.63 3.82 3.69 3.97 4 3.93 3.92 3.9 3.88 4.13 3.91 Max* 4.82 4.05 4.64 4.66 3.96 3.96 4 3.94 4.23 4.09 4.46 4.39 4.29 4.22 4.26 4.25 4.36 4.27 *with the Average as the average from area fraction results from Table C-3 of single micrograph, Min as the minimum area fraction, Max as the maximum one, and Mean is the average of data in each row in Table C-4.  120  Appendix D Calculation of Burger’s vector and Young’s Modulus D.1 Temperature dependent Young’s Modulus 𝜇 = 𝜇0 �1 + �𝑇 − 300𝑇𝑚 �𝑇𝑚𝜇0 𝑑𝜇𝑑𝑇�                        (𝐷. 1) 𝐸 = 2.6𝜇                                                               (𝐷. 2) Table D-5 Parameters meaning and references for Modulus equations Symbol Meaning Values Unit References T temperature -- -- -- 𝜇0 Shear modulus at 300K 25.4 GPa Frost and Ashby(Frost and Ashby 1982) 𝑇𝑚 Meting point 933.47 K From Al-Mn-Si phase diagram(Belov, D.G.Eskin et al. 2005) E Young’s Modulus -- GPa -- 𝜇 Shear Modulus -- GPa -- 𝑇𝑚 𝜇0 𝑑𝜇 𝑑𝑇  Temperature dependence of the modulus -0.5 1 Frost and Ashby(Frost and Ashby 1982)    121  Table D-6 Temperature dependent Shear Modulus and Young’s Modulus values Temperature/oC 300 400 500 550 600 μ(Gpa) 21.7 20.3 19.0 18.3 17.6 E(GPa) 56.4 52.8 49.3 47.5 45.8  D.2 Temperature dependent Burger’s vector Eq.D.3 and Eq. D.4 are used for calculation of Burger’s vector. 𝑎 = 𝑎0 + 𝛼𝑇𝑎0                                    (D. 3) 𝑏 = 𝑎 √2                                                  (D. 4) Table D-7 Parameters meaning and references for Burger’s vector equations Symbols Meaning Values Unit References 𝑎 Lattice Parameter at given temperature -- M -- 𝑎0 Lattice Parameter at 300K 4.049x10-10 M For pure aluminum(Lubarda) 𝛼 Thermal Expansion Coefficient Table 4 K-1 For pure aluminum(Fickett 1971) 𝐵 Burger’s vector -- M --  122   Table D-8 Coefficient of thermal expansion data from (Fickett 1971) T (oC) 300 400 500 550 600 𝜶(x10-6K-1) 25.9 27.9 30.7 32.2 34.3  Table D-9 Calculated temperature dependent lattice parameter and Burger’s vector values Temperature/oC 300 400 500 550 600 𝒂(x10-10m) 4.08 4.13 4.15 4.16 4.17 b(x10-10m) 2.89 2.92 2.93 2.94 2.95     123  Appendix E Strain and strain rate calculation Effective strain during extrusion is given in Eq.E.1(Saha 2000). 𝜀𝑒𝑓𝑓 = 𝑙𝑛𝐸𝑅 = 𝑙𝑛130 ≈ 4.87                                  (E. 1) The time average mean strain rate 𝜀?̇?𝑣𝑒𝑟𝑎𝑔𝑒 during extrusion is expressed in Eq. E. 2 (Somani et al.) 𝜀?̇?𝑣𝑒𝑟𝑎𝑔𝑒 = 6𝑉𝑅𝐷𝐶2𝑙𝑛𝐸𝑅 ∙ 𝑡𝑎𝑛𝛼𝑠𝑑𝐷𝐶3 − 𝐷𝐸3                             (E. 2) Where 𝑉𝑅 is the ram speed; 𝐷𝐶  is the diameter of the cast billets; ER is the extrusion ratio; 𝐷𝐸  is the diameter of the extrudate; 𝛼𝑠𝑑 is the semi-die angle. It only gives an approximate indication of the average strain rate. In practice, there is a wide variation of strain rate from point-to-point in the extruded product. Normally 𝛼𝑠𝑑  is taken as 45oC for estimation. For current I-beam geometry and ram speed, an average strain rate of 8.2s-1. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0071658/manifest

Comment

Related Items