國立中山大學材料科學研究所 博士論文
摩擦攪拌製程製作細晶粒鎂基合金與複材 之研發與性能分析
Development and analysis of fine-grained Mg base alloys and composites fabricated by friction stir processing
研究生：李敬仁 撰 指導教授：黃志青 博士
中華民國 九十五 年 十一 月
Table of Content
Table of Content………..i
List of Tables…….………...vi
List of Figures….………..viii
Abstract………xvi
中文摘要...xviii
謝誌………..………xix
Chapter 1 Introduction……… 1
1.1 Characteristics of magnesium alloys………3
1.2 Properties of magnesium alloys………...4
1.2.1 The classification of magnesium alloys………...…………...4
1.2.2 Influence of grain refinement on magnesium alloys………...…………4
1.2.3 Various refinement techniques………...……….6
1.2.3.1 Through severe plastic deformation……….6
1.2.3.2 Through rapid solidification……….8
1.2.4 Anisotropy of magnesium alloys…..……….9
1.3 Metal matrix composites………10
1.3.1 Processing of metal matrix composites and magnesium matrix composites…….12
1.3.1.1 Liquid-state methods………..12
1.3.1.2 Solid-state methods………15
1.3.2.3 Other processing methods………..15
1.4 Basic characters of superplastic materials………..17
1.5 Superplasticity of magnesium alloys and magnesium matrix composites……….19
1.5.1 HSRSP and LTSP of magnesium alloys…..………..………20
1.5.2 HSRSP and LTSP of magnesium matrix composites………21
1.6 Friction stir welding and friction stir processing………22
1.6.1 Introduction of friction stir welding (FSW)………..22
1.6.2 Process parameters………...………24
1.6.2.1 Tool geometry……….24
1.6.2.2 Welding parameters………25
1.6.3 Process modeling……….………...26
1.6.3.1 Material flow behavior in the stirred zone……….27
1.6.3.2 Material flow modeling………..28
1.6.3.3 Input energy and temperature prediction………30
1.6.4 Microstructure evolutions……….31
1.6.4.1 Nugget zone (or stirred zone)……….32
1.6.5 Texture……….………...35
1.6.6 Friction stir processing (FSP)………36
1.6.7 Applications of friction stir processing……….37
1.7 Motives of this research……….39
Chapter 2 Experimental methods……….41
2.1 Materials……….41
2.2 The setup of friction stir processing………...41
2.2.1 The designs of tool and fixture………..42
2.2.2 The methods of adding nano-sized powders into AZ61 alloys……….42
2.2.3 The methods of adding nano-sized powders into AZ61 alloys……….42
2.3 Microhardness measurements………43
2.4 Mechanical testing………..44
2.5 The analysis of X-ray diffraction………...44
2.6 X-ray photoelectron spectroscopy (XPS)………...44
2.7 Microstructure Characterizations………...45
2.7.1 Optical microscopy (OM)……….45
2.7.2 Scanning electron microscopy (SEM)………...46
2.7.3 Transmission electron microscopy (TEM)………46
Chapter 3 Experimental results………48
3.1 The temperature of the stirred zone of modified alloys……….…48
3.2 The macrostructure and microstructure of the stirred zone of modified alloys……….48
3.2.1 The macrostructure of the stirred zone of modified alloy after 1P and 4P FSP…...48
3.2.2 The microstructure of the stirred zone of modified alloy after 1P and 4P FSP…....49
3.2.3 The grain size of modified AZ61 alloy……….50
3.2.4 Microstructure of the modified AZ61 alloy after subsequent compressive loading50 3.3 The macrostructure and microstructure of Mg based composites made by FSP……...51
3.3.1 The appearance of the nugget zone of composites………51
3.3.2 SEM characterizations of Mg matrix composites made by FSP………...51
3.3.3 TEM phase identification………..53
3.3.4 Grain stabilization of Mg based composites made by FSP………...54
3.4 XPS analysis………...54
3.5 XRD analysis………..55
3.5.1 XRD analysis for the modified alloys………...56
3.5.1.1 XRD analysis for the modified alloys after subsequent compression…...57
3.5.2 XRD analysis for Mg based composites made by FSP………....57
3.6 Hardness measurements……….58
3.7 Mechanical tests……….59
3.7.1 Mechanical properties at room temperature………..60
3.7.1.1 Mechanical properties of the modified alloys made by FSP…………..60
3.7.1.2 Mechanical properties of Mg based composites fabricated by FSP…….61
3.7.2 Tensile behavior of the FSP Mg alloy at elevated temperatures………...62
3.7.2.1 Tensile behavior of the WD specimens of the modified alloy at elevated temperatures………..……….………...62
3.7.2.2 Topography of deformed specimens………..63
3.7.2.3 Tensile behavior of the TD specimens of the modified alloy at elevated temperatures………...63
3.7.3 Tensile behavior at elevated temperatures for the FSP Mg based composites…….64
3.7.3.1 Tensile behavior at elevated temperatures for the 1D Mg based composites………..64
3.7.3.2 Tensile behavior at elevated temperatures for the 2D Mg based composites………..65
3.7.3.3 Topography of deformed specimens……….……..66
3.8 Fractography of deformed specimens………68
Chapter 4 Discussion and Analysis on Deformation Mechanisms……..………....68
4.1 Influence of FSP pass number on the modified alloy……….68
4.1.1 Influence of FSP pass number on microstructure………..68
4.1.2 Texture analysis of the modified alloy by FSP………..69
4.1.3 Influence of FSP pass number on mechanical properties………..71
4.1.3.1 Influence of FSP pass number on mechanical properties at room temperature……….71
4.1.3.2 Influence of FSP pass number on mechanical properties at elevated temperatures………...73
4.2 Influence of subsequent-compression along ND on the modified alloy………74
4.2.1 Influence of subsequent compression on microstructure………..74
4.2.2 Influence of subsequent compression on texture………..75 4.2.3 Influence of subsequent compression on mechanical properties at room
temperature……….……….76
4.3 Influence of FSP pass number to Mg based composites………78
4.3.1 Influence of FSP pass number on microstructure and interfacial reaction………...78
4.3.2 Texture analysis of the magnesium composites made by FSP………..79
4.4 Comparison of mechanical properties at room temperature of the modified alloy and composites………..………..79
4.5 Analysis on deformation mechanism at elevated temperatures of magnesium based composites………..…………..81
Chapter 5 Conclusions………..……….84
5.1 Conclusions on the FSP modified magnesium alloys……….…………84
5.2 Conclusions on the FSP magnesium based composites………85
Chapter 6 Major achievements………...87
References………..………..88
Tables………....99
Figures………121
List of Tables
Table 1-1 Comparison among the Mg alloys, Al alloys, Ti alloys, steels and plastics…….99 Table 1-2 The standard four-part ASTM designation system of alloy and temper for the
magnesium system……….100 Table 1-3 The effect of separate solute addition on the mechanical properties……….…101 Table 1-4 Mechanical properties of magnesium matrix composites by various liquid-state processing methods……….………..102 Table 1-5 Mechanical properties of magnesium matrix composites by various solid-state
processing methods……….………..103 Table 1-6 Microstructure-Mechanical property-facture correlations for metal matrix
composites………..104 Table 1-7 HSRSP and LTSP in the magnesium alloys……….……..105 Table 1-8 HSRSP and LTSP of magnesium matrix composites………..106 Table 1-9 Superplasticity application of Al alloys modified by friction stir processing…107 Table 2-1 Chemical compositions of the AZ61 Mg alloy (in wt%)………108 Table 2-2 The sample designation for the FSP modified Mg alloys………...109 Table 2-3 The sample designation for the FSP Mg based composites………109 Table 3-1 The recrystallized grain size of the modified AZ61 Mg alloy made by FSP….110 Table 3-2 Summary of the average SiO2 cluster size and the average AZ61 matrix grain size in the 1D (with Vf~5%) and 2D (with Vf~10%) FSP specimens………...110 Table 3-3 XRD results for the 1P45, 4P45, and 4P45-cp modified alloy samples……….111 Table 3-4 XRD results for the Mg based composites of the 1D samples………112 Table 3-5 XRD results for the Mg based composites of the 2D samples………113 Table 3-6 The average hardness of the FSP modified AZ61 alloys………114
Table 3-7 The average hardness of the Mg based composites made by FSP…………..114 Table 3-8 The summary of mechanical properties for the AZ61 billet, the modified AZ61 alloy and composite made by FSP………..115 Table 3-9 The summary of the tensile properties of the WD and TD specimens for the 1P45 and 4P45 modified alloys at room temperature, tested on the Fig. 2-11 specimen size………...115 Table 3-10 The ductility of the FSP modified AZ61 alloy tested at elevated temperatures.116 Table 3-11 The ductility of the FSP 1D4P and 2D4P Mg based composites at elevated
temperatures………...118 Table 4-1 The measured apparent strain rate sensitivity (ma) value for the modified alloys and composites………120 Table 4-2 The whole comparison, considering σHP, σm, σo factors contribution for the
modified alloy and composites samples……….120
List of Figures
Figure 1-1 Schematic illustration of the ECAP facility………...121
Figure 1-2 Schematic illustration of the spray forming facility………..122
Figure 1-3 The schematic illustration of the Mg2Si, SiO and MgO formation in AZ91/5SiC……….123
Figure 1-4 Schematic illustration of (a) cavity formation due to high stress concentrations at the interface and (b) relaxation of the stress concentration by a liquid phase for metal matrix composites reinforced with particles……….123
Figure 1-5 Schematic diagram of friction stir welding………...124
Figure 1-6 Schematic diagram of entire friction stir welding……….125
Figure 1-7 Schematic drawing of the FSW tool………..126
Figure 1-8 WorlTM and MX TrifluteTM tools developed by The Welding Institute (TWI), UK………..126
Figure 1-9 Illustration of the different probes: (a) smaller contact area of headpin and (b) flared-truflute type probes………..127
Figure 1-10 Range of the optimum FSW conditions for each tool plunge downforce…..128
Figure 1-11 Three-dimensional plot of markers flow in weld during friction stir welding.129 Figure 1-12 Vertical mixing in specimens (a) before welding, (b) 0.19 mm/rev, (c) 0.35 mm/rev, (d) 0.5 mm/rev. The dashed line denote the pin diameter, which was the same for three welds………...129
Figure 1-13 (a) Metal flow patterns and (b) metallurgical processing zones developed during friction stir welding………130
Figure 1-14 Schematic extrusion zone on the advancing and retreating side showing effects of thread form on extrusion area….……….131
Figure 1-15 Schematic illustration of the welding zone in the friction stir welding process………..………..132 Figure 1-16 Typical onion ring in the nugget zone………133 Figure 1-17 Three-dimensional schema of the onion rings in the nugget zone………….133 Figure 1-18 Electron back scattered diffraction (EBSD) map from directly ahead of the tool, showing the deformation bands and the effect of the progressively increasing strain and temperature with proximity to the tool………..134 Figure 1-19 Mixed microstructure containing finer grains and elongated fibrous grains in Al alloys………135 Figure 1-20 The banded structure in the horizontal cross-section after FSW………136 Figure 1-21 Micro-hardness and strain field map in banded microstructure……….137 Figure 1-22 Schematic illustrations of (a) the basal plane tracing from the base material, transition region, to stir zone on the advancing side of the FSP Mg plate and (b) the trace surface of basal plane of FSW Mg alloy within the stirred zone…138 Figure 2-1 The microstructure of the as-received AZ61 billet………...139 Figure 2-2 TEM photographs of SiO2 particles: (a) Image and (b) diffraction pattern….140 Figure 2-3 The appearance of the horizontal-type miller………141 Figure 2-4 Schematic illustration of the tool………...142 Figure 2-5 Schematic illustration of the entire fixture design……….143 Figure 2-6 The appearance of deep grooves cut before FSP for filling nano-sized SiO2
powders………..144 Figure 2-7 Schematic illustrations for vertically adding SiO2 and surface repair methods for fabricating the Mg base composites by FSP………..145 Figure 2-8 Schematic illustration of the position for the K-type thermocouples inserted in the sample………...146 Figure 2-9 The experimental flowchart for this research………147
Figure 2-10 Schematic illustration for sampling position and dimension of the WD specimen………...148 Figure 2-11 Schematic illustration for sampling position and dimension of the TD
specimen………149 Figure 2-12 Schematic illustration for sampling position of (a) T plane, (b) L-center and
L-retreating plane, and (c) H plane………150 Figure 3-1 Typical temperature profile measured by the inserted thermocouple into the AZ61 alloy at 800 rpm with a speed of 90 mm/min………152 Figure 3-2 The macrostructure of the 1P45 and 4P45 modified alloys: (a) H plane and (b) T plane………...153 Figure 3-3 Optical micrographs taken from the H plane of 1P45 modified AZ61 alloy:(a) x100 magnification, (b) x400 magnification and (c) x800 magnification….154 Figure 3-4 Optical micrographs taken from the H plane of the 4P45 modified AZ61 alloy:
(a) x100 magnification, (b) x400 magnification and (c) x800 magnification..155 Figure 3-5 Optical micrographs on the (a) L-center plane and (b) T plane of the 1P45
modified alloy………..………156 Figure 3-6 Optical micrographs on the (a) L-center plane and (b) T plane of the 4P45 modified alloy………..………156 Figure 3-7 OM micrographs showing the variation of the recrystallized grain size in the stirred zone for the different processing parameters…….………157 Figure 3-8 OM microgaphs showing the twinning microstructure of the 4P45-cp modified alloy………...158 Figure 3-9 The optical macro- and micrographs of Mg based composites made by FSP
after two passes: (a) top view, (b) cross-sectional view and (c) high magnification………159 Figure 3-10 SEM micrographs showing the improvement of clustered silica particles in the
outside of onion ring for (a) one pass and (b) two passes……..…………...160 Figure 3-11 SEM micrographs showing the particle dispersion in the one deep groove
specimen (1D, Vf~5%) after (a) one pass,(b) two passes, (c) three passes and (d) four passes………..161 Figure 3-12 SEM micrographs showing the inserted particle dispersion in the two deep-groove specimen (2D, Vf~10%) after (a) one pass, (b) two passes and (c) four passes………..162 Figure 3-13 Statistic result of clustered silica for the various passes in the one deep groove
specimen……….163 Figure 3-14 EDS result for some larger particles………..164 Figure 3-15 SEM photograph showing the clustered silica located on grain boundaries or triple junctions and some silica embedded into grains of the AZ61 matrix….165 Figure 3-16 The bright field TEM micrographs showing the grain size of composites made by FSP: (a) 1D1P, (b) 1D2P, (c) 1D3P and (d) 1D4P….………...166 Figure 3-17 The bright field TEM micrographs showing the grain size of composites made
by FSP: (a) 2D1P, (b) 2D2P and (c) 2D4P………...167 Figure 3-18 TEM micrographs showing (a) the tangled dislocations and SiO2 particles
measuring around 20 nm within the grain interior and (b) the
[ ]
1213 zonediffraction pattern………..……168 Figure 3-19 TEM micrographs showing (a) clustered silica particles, (b) amorphous phase
silica after FSP and (c) diffraction pattern………..169 Figure 3-20 TEM/EDS results showing the presence of Mg2Si in the FSP composite
specimens………...170 Figure 3-21 TEM micrographs showing (a) dark field image of the fine MgO phase
measuring 5-10 nm and (b) ring pattern with MgO ring patterns…………..171 Figure 3-22 Variation of grain sizes of the FSP AZ61 alloy and composites at 350oC during
static annealing: (a) the variation as the function of grain size at the different annealing time, (b) the grain size of the pure modified AZ61 alloy with 4P after 10 hours, (c) the grain size of the 1D4P composite specimen after 10 hours and (d) the grain size of the 2D4P composite specimen after 10 hours…….172 Figure 3-23 XPS spectra for (a) Si-2p of 2D1P composite sample, (b) Si-2p of 2D2P composite sample and (c) Si-2p of 2D4P composite sample………….…….173 Figure 3-24 X-ray diffraction for (a) random Mg and (b) as-received AZ61 billet….….174 Figure 3-25 X-ray diffraction on (a) T plane, (b) H plane, (c) L-center plane and (d)
L-retreating plane for the 1P45 and 4P45 modified AZ61 alloy………175 Figure 3-26 X-ray diffraction on (a) T plane, (b) H plane, (c) L-center plane and (d)
L-retreating plane for the 4P45 and 4P45-cp modified AZ61 alloys……….176 Figure 3-27 X-ray diffraction on (a) T plane, (b) H plane, (c) L-center plane and (d)
L-retreating plane for the 1D composites samples with the different pass number………177 Figure 3-28 X-ray diffraction on (a) T plane, (b) H plane, (c) L-center plane and (d)
L-retreating plane for the 2D composites samples with the different pass number………179 Figure 3-29 X-ray diffraction for the 2D4P specimen showing the compound of Mg2Si after FSP……….181 Figure 3-30 Variation of the Hv microhardness distributions in the 1P45, 4P45, and 4P45-cp modified AZ61 alloy………..182 Figure 3-31 Variation of the Hv microhardness distributions in composites under the (a) one groove specimen, (b) two grooves specimen……….183 Figure 3-32 The engineering stress and strain curve for the 1P45 and 4P45 modified AZ61 alloy sampled along WD………184 Figure 3-33 The engineering stress and strain curve of the 1P45 and 4P45 modified AZ61
alloy sampled along WD and TD, with Fig. 2-11 tensile specimen size…...185 Figure 3-34 The engineering stress and strain curves comparison for the 4P45 and 4P45-cp modified AZ61 alloys……….186 Figure 3-35 The engineering stress and strain curves for the 4P45 modified AZ61 alloy and composite samples made by FSP………...187 Figure 3-36 True stress and strain curves for the WD specimens of the 1P45 modified alloy
at (a) 300oC and (b) 400oC……….188 Figure 3-37 True stress and strain curves for the WD specimens of the 4P45 modified alloy at (a) 300oC and (b) 400oC……….189 Figure 3-38 Variation of elongation at elevated temperatures of the modified AZ61 alloy
with 800 rpm FSP, sampled parallel to WD, as a function of loading strain rate for (a) 1 pass and (b) comparison between the 1P45 and 4P45 samples…...190 Figure 3-39 The appearance of the modified alloy sampled along WD for (a) 1P45 specimen at 300oC and 1x10-4 s-1, (b) 1P45 specimen at 400oC and 1x10-3 s-1, (c) 1P90 specimen at 300oC and 1x10-3 s-1 and (d) 4P45 specimen at 300oC and 1x10-4
s-1………191
Figure 3-40 SEM micrographs showing the surface topography for the WD specimen of the 1P45 modified alloy deformed at 400oC and 1x10-3 s-1: (a) lower magnification and (b) high magnification……….192 Figure 3-41 True stress and strain curves for the TD specimens of the 1P45 modified alloys
at (a) 300oC and (b) 400oC……….193 Figure 3-42 True stress and strain curves for the TD specimens of the 4P45 modified alloys at (a) 250oC, (b) 300oC, (c) 350oC and (d) 400oC………..194 Figure 3-43 Variation of elongation at elevated temperatures of the 1P45 and 4P45 modified AZ61 alloy, sampled along TD, as a function of loading strain rate………..196 Figure 3-44 Variation comparison of elongation at elevated temperatures for the WD and
TD specimens of the 1P45 and 4P45 modified AZ61 alloys as a function of loading strain rate at 300oC……….196 Figure 3-45 True stress and strain curves for the TD specimens of the 1D4P composite
samples at (a) 250oC, (b) 300oC, (c) 350oC and (d) 400oC.………..197 Figure 3-46 Variation of superplastic elongation as a function of loading strain rate for (a) 1D4P and (b) 2D4P Mg based composites……….199 Figure 3-47 The appearance of the 1D4P specimens after deformation at (a) 6x10-4 s-1, (b) 1x10-3 s-1, (c) 1x10-2 s-1 and (d) 1x10-1 s-1………..200 Figure 3-48 True stress and strain curves for the TD specimens of the 2D4P composite
samples at (a) 250oC, (b) 300oC, (c) 350oC and (d) 400oC….………..201 Figure 3-49 The appearance of the 2D4P specimens after deformation at (a) 1x10-3 s-1, (b) 1x10-2 s-1, (c) 1x10-1 s-1 and (d) 3x10-1 s-1………..203 Figure 3-50 SEM micrographs showing the surface topography of the 2D4P specimens
deformed at 350oC and 1x10-1 s-1 to (a) strain = 0.27, (b) strain = 0.79, (c) strain
= 1.20 and (d) strain = 1.37………204 Figure 3-51 SEM fractography of the 1P45 modified AZ61 Mg alloy: (a) lower
magnification and (b) higher magnification………..205 Figure 3-52 SEM fractography of the 4P45 modified AZ61 Mg alloy: (a) lower magnification and (b) higher magnification………..206 Figure 3-53 SEM fractography of the 4P45-cp modified AZ61 Mg alloy: (a) lower magnification and (b) higher magnification………...207 Figure 3-54 SEM fractography of the 2D2P composite sample: (a) lower magnification and
(b) higher magnification………..208 Figure 3-55 SEM micrograph showing the fracture surface of the 2D4P specimen deformed at 350oC and 1x10-1 s-1 to failure………209 Figure 4-1 Schematic illustration of grains structure in the stirred zone of the modified
alloy with 1 pass FSP………..210 Figure 4-2 Schematic illustration of grains orientation in the stirred zone of the modified alloy: (a) H plane, (b) L-center plane (c) T plane, (d) L-retreating plane……211 Figure 4-3 Schematic illustration of onion splitting course for the 1 pass FSP Mg alloy being tensioned parallel to the welding direction………...213 Figure 4-4 Schematic illustration of grains orientation in the stirred zone of the modified alloy after subsequent compression along the normal direction: (a) H plane and (b) L-center plane……….214 Figure 4-5 The analyses on the 2D4P composites samples for (a) flow stress against strain rate cures and (b) apparent activation energy……….215
Abstract
In this research, one solid state processing technique, friction stir processing, is applied to modify the AZ61 magnesium alloy billet and to incorporate 5-10 vol% nano-sized ceramic particles SiO2 into the AZ61 magnesium alloy matrix to form bulk composites, using the characteristic rotating downward and circular material flow around the stir pin. The microstructure and mechanical properties of the modified alloy and composite samples are examined and compared.
The FSP modified AZ61 alloy could be refined to 3-8 μm via the dynamic recrystallization during processing. However, the one-pass FSP modified alloy appeared the inhomogeneous grain structures to influence the tensile ductility along the welding direction at elevated temperatures due to the onion splitting. In contrast, the multi-pass FSP could improve the inhomogeneous grain structures to reduce the influence of the onion-splitting to the deformation at elevated temperatures. The FSP modified alloys show the lower yielding stress due to the unique texture of (0002) basal planes, with roughly surround the pin column surface of the pin tool in the stirred zone. In addition, it is suggested that the second processing of the subsequent compression along the normal direction might be necessary to alter the texture and to improve the lower yielding stress after modifying the grain size by FSP.
Friction stir processing could successfully fabricate bulk AZ61 Mg based composites with 5 to 10 vol% of nano-sized SiO2 particles. The nano-sized SiO2 particles added into magnesium matrix could be uniformly dispersed after four FSP passes. The average grain sizes of the composites varied within 0.5-2 μm, and the composites nearly double the
hardness as compared with the as-received AZ61 cast billet. This composite exhibited high strain rate superplasticity, with a maximum ductility of 470% at 1x10-2 s-1 and 300oC or 454%
at 3x10-1 s-1 and 400oC while maintaining fine grains less than 2 μm in size.
中文摘要
在本研究論文中，使用固態製程法-摩擦攪拌製程，來改質 AZ61 鎂合金半連續鑄 錠。此ㄧ製程能促使材料發生沿著凸梢工具頭外圍的塑性流動特徵，達到攪拌材料功 能。運用此ㄧ塑性流動特徵，也可在 AZ61 鎂合金半連續鑄錠中，加入體積分率 5%-10%
的奈米級 SiO2顆粒，以達到分散顆粒效果，來製作成塊狀鎂基複合材料。並且對改質
合金與複合材料作微結構與機械性質鑑定與比較。
經過摩擦攪拌製程改質的鎂合金，藉由變形過程所誘發的動態再結晶，可以將晶粒
尺寸細化至 3-8 μm。ㄧ道製程改質合金有著不均質的晶粒結構，此ㄧ微結構特徵將會
影響到沿著銲道拉伸行為的高溫機械性質，並出現一種洋蔥圈剝離現象。相對地，運用 多道數的製程，可以有效改善此ㄧ不均質的晶粒結構，以達到改善此種在高溫變形過程 所誘發的洋蔥圈剝離現象。改質合金表現出較低的降伏強度，是受到摩擦攪拌製程所誘 發的晶粒擇優取向的影響，亦即(0002)平面作面向排列於洋蔥圈圓弧上。此外，本研究 發現在摩擦攪拌改質合金後，緊接著施以二次加工的壓縮變形，此一壓縮變形沿著垂直 方向，可以藉由變形雙晶來達到轉變摩擦攪拌所誘發之特有的晶粒擇優取向，以達到改 善摩擦攪拌改質合金較低的降伏強度。
藉由摩擦攪拌製程可以成功地製作出塊狀鎂基複合材料，並且在四道的摩擦攪拌製
程後，可以有效地分散所加入 SiO2顆粒進入鎂合金基材中。所製作出的鎂基複合材料，
其晶粒可以細化到 0.5-2 μm，相對於初始的 AZ61 鎂合金半連續鑄錠，此複合材料的硬 度可以被提升至將近兩倍。此ㄧ複合材料在高溫 300oC，在應變速率 1x10-2 s-1與 3x10-1
s-1，以及高溫 400oC，下，分別有 470%與 454%的高速超塑性伸長量，且晶粒尺寸依然
可以維持在 2 μm 以下。
謝誌
能夠在這將近三年半的時間內，完成此本博士論文，首先，要獻上十二萬分的敬意 感謝我的恩師，黃志青教授。吾師的淵博學識，嚴以律己，寬以待人的處事態度以及公 正無私的行事風格，都深深影響到我，成為我學習效法的典範。另外，要感謝口試委員 高伯威教授、何扭今教授、張六文教授以及成大材料系的陳立輝教授與呂傳盛教授，在 論文審查過程中，給予學生的指導與指正，由於你們的細心審查與寶貴意見，使得本論 文更加充實，在此對老師們致上我最高的感謝。
也感謝所上的各實驗的技術員，陳貴香女士、王良珠學姊、王國強先生、林明政先 生、江宏達先生、古錦松學長、李秀月女士、施淑瑛小姐，感謝你們在各項實驗上的協 助。另外，也要感謝所辦的朱惠敏女士、顏秀芳小姐、陳秀月小姐，在行政事務上的幫 忙，還有，也要感謝熱心的華大哥在許多雜事的幫忙。
身處在黃幫的大家庭哩，有太多人要感謝了。首先，已經畢業的木城學長、建超學 長、凱琳學姊、佩汝學姊、鉉凱學長、英博學長，感謝你們在學識與生活上的開導，博 班同學的育誠、子翔、志溢、友杰、海明、炎暉學弟的協助，以及碩士班的鴻昇、致榮、
浩然、名哲，大學部的振偉、哲男學弟們帶給我美好的回憶。另外，也要感謝高張 group 的小丁、惠君、德凱學弟以及無法詳加紀錄的諸位學長姊跟學弟妹們，還有何幫的諸位 學弟們，感謝你們在實驗上的協助。還有，也要感謝已經回到大陸的王轶農博士，在實 驗上的指導。
最後在這要感謝我的父母與小妹給予我全力的支持，在我遭受挫折，仍不斷給我鼓 勵。最後非常感恩父母的全力栽培，讓我後顧之憂，全心全力的求學。在此把本論文獻 給我的家人，感謝你們的支持與鼓勵。
要感謝的人真的太多了，如果有沒記載到的諸位大人，請多多見諒。套句陳之藩所 說的話，要感謝的人太多了，一切都感謝老天，感謝上帝。
李敬仁 謹誌 於中山大學材料科學研究所 中華民國九十五年十一月
Chapter 1
Introduction
In the application of metals or alloys, magnesium alloys have attracted more and more attention lately in response to the rising eco-awareness. Due to the fact that magnesium possesses the lowest density among all light structural metals (such as Al, Mg and Ti), the use of magnesium alloys will reduce the energy consumption. In addition, magnesium also has the advantage of recyclability, further decreasing the squander of natural resources. Based on the lower density and recyclability, magnesium alloys have gradually become the highly potential metallic materials.
Metal matrix composites, MMCs, are the materials with metals as the matrix and reinforcements in the form of long fibers, short fibers, whiskers or particulates. Generally, aluminum, magnesium, titanium, and copper alloys with light weight and good ductility are often served as the matrix. With regard to the reinforcements, ceramic materials with high temperature stability, high modulus and high fracture strength are the best reinforcing phase, such as carbon fibers, glass fibers, the particulates of SiC, SiO2, Al2O3 or TiB2, etc. Due to the introduction of such second ceramic phases into the metal matrix, the composites properties are enhanced in terms of, for example, better strength, corrosion resistance, and wear resistance. Therefore, the composites with improved physical and mechanical properties will be used in higher-temperature and stricter environments than those for pure metals or alloys without any reinforcement.
Friction Stir Welding (FSW) is a newest joining method invented by The Welding
Institute (TWI) in UK, in 1991 [1]. It is a solid-state joining technique applied in joining aerospace aluminum alloys during its initial developing stage; recent progress has been extended into Ti, Mg, Cu alloys and steel. Due to the occurrence of the plastic deformation at elevated temperatures during the welding process, Mishra and his group [2,3] applied this idea on the structure modification of the Al alloy via the dynamic recrystallization [2], as well as the fabrication of the surface composite of the alloy [3] by using the complex plastic flow of the materials, and termed this process as Friction Stir Processing (FSP).
The success in fabrication of various nano-sized powders, wires or tubes has provided the new method in modifying the existing commercial materials in terms of their functional or structural characteristics. Inorganic nano oxide, nitride or metallic powders may be inserted in polymers, ceramics, metals, or semiconductors by various sorts of simple or sophisticated means. Except for few reports, most of achievements were focused on the polymer matrix modified by ceramic nano particles so as to significantly improve its mechanical or physical properties. In contrast, the reports of adding nano powders into metallic alloys were relatively much less addressed.
In this study, the uniquely plastic flow of friction stir processing has been applied to disperse the nano-sized particles into the bulk Mg alloy matrix and develop a simple and easy processing method or concept to fabricate the bulk Mg based nano-composites. The same processing parameters of FSP have also been applied on the pure Mg based alloys without the particulate reinforcements or called “the modified alloys”. Afterward, the analysis and comparison of microstructures and mechanical properties of the resulting composites and the modified AZ61 Mg alloy by FSP are explored.
1.1 Characteristics of magnesium alloys
With the global tendency toward reducing fuel consumption and CO2 emissions, the automobile industry has made the development of the light-weight vehicles [4]. In order to achieve the light-weight goal, it is necessary to adopt the light metals in structural and non-structural components on the vehicles. Table 1-1 lists the comparison among the magnesium alloys and many common materials. Because of the lowest density of Mg in all light metals (~1.7 g/cm3, similar to or only slightly above the densities of most polymers and polymer composites), it is attractive to the computer, consumer and communication (3C) electronic products that Mg alloys could be used for the shell and radiator components of these electronic products with superior electromagnetic interference (EMI) capability. The detailed advantages of magnesium alloys are listed below.
(1) Lower density: In terms of density, Mg is about 2/3 of Al, only 20% of steel and is the lowest of all metallic constructional materials.
(2) Better heat dissipation: Mg has excellent heat dissipation capability in comparison with plastics.
(3) Good damping and crash resistance: The damping capability of Mg is better than plastic and aluminum, and the crash resistance is much superior to plastics.
(4) Good castability: Mg can be easily die-cast into complicated shapes with minimum wall thickness.
(5) Excellent electromagnetic shielding capabilities: A thin Mg wall can effectively shield electromagnetic wave, meaning that Mg shell can protect human body from electromagnetic-wave damage. In contrast, the plastics cannot perform this good characteristic.
(6) Thin thickness: The thickness of Mg enclosures used for 3C procducts can be reduced to 0.5 mm for sufficient strength.
(7) Recyclabilty: Mg can be easily and economically recycled through proper re-melting procedures similar to Al. This imposes a strong benefit over plastics.
It is clear from the above advantages that Mg alloys have the potential to replace aluminum alloys or plastics on the application of the electronic products, or to be the important structural materials by these characters. Despite the numerous advantages, magnesium alloys exhibit poor workability due to the hexagonal closed-packed (HCP) crystal structure. Therefore, the shaping of magnesium alloys needs to adopt the die casting, thixomolding [5] or thixocasting [6] for the present stage. The way to improve the workability of magnesium alloys will be essential in order to promote mass production of magnesium alloys in many different engineering fields.
1.2 Properties of magnesium alloys
1.2.1 The classification of magnesium alloys
Magnesium alloys are usually designated by two capital letters followed by two or three numbers. Table 1-2 [7] presents the standard four-digit ASTM (American Society for Testing and Materials) designation system for magnesium alloys and their temper treatments. For instance, the most common “AZ31B-H24” Mg-based alloy means that the magnesium alloy contains nominal 3 wt% aluminum (A) and 1 wt% zinc (Z) and is in the B modification, distinguishing from the same AZ31 that contains the different levels of impurity. The H24 designation implies that the alloy is strain-hardened and partially annealed. The addition of some solute element, such as aluminum (A), zinc (Z), zirconium (K), rare earths (E) and silver (Q) etc., is helpful for magnesium properties on casting, mechanical and corrosion
behavior. For example, the addition of Al into Mg solvent is mainly aimed for the solution hardening and precipitation hardening through the Mg17All2 β phase, while the addition of a small amount of Zn will improve casting capability. The effects of the various solute elements on the mechanical, corrosion and casting behavior are listed in Table 1-3 [8]. Currently, the AZ, AM and ZK series Mg based alloys are extensively used in the commercial market.
1.2.2 Influence of grain refinement on magnesium alloys
It is well known that polycrystalline metals always show a strong effect of grain size hardening, except at highly elevated temperatures. The widely applied equation for the relation between grain size and strength or hardness, the Hall-Petch equation [9], is written in the form of
H = Ho + kHd-1/2, (1) or σ = σo + kσd-1/2, (2) where H and σ are hardness and flow-stress, respectively, d is the average grain size, and Ho, σo, kH and kσ are constants. Narutani et al. [10] reported this Hall-Petch relationship for magnesium, and the k value is about 280~320 MPa‧μm1/2. Our laboratory [11,12] also reported the kH and kσ are about 45~65 Hv‧μm1/2 and 300~350 MPa‧μm1/2, respectively. It is very attractive that the k slope, or called the grain size strengthening efficiency, of magnesium alloys is much higher than that of aluminum alloys (~68 MPa‧μm1/2) [13], indicating the grain refinement of magnesium can gain more effective strengthening than aluminum alloys. Except for the above strengthening, it has also been reported that the fatigue strength can be raised by grain size refinement.
The grain refinement of magnesium alloys not only raises the strength, but also
improves the formability. Bussiba et al. [14] reported that, with smaller grain size in the submicron scale, magnesium alloys can possess better superplasticity at low temperatures or high strain rates. Our laboratory also proved this theory in recent years [12,15].
Since it is well known that grain refinement can impose substantial positive influence to magnesium alloys, many researches have tried the various refinement techniques. In the next section, some of the promising techniques are introduced.
1.2.3 Various refinement techniques
The principle of refinement of grain size can be divided into two categories. Firstly, a large strain is imposed on the metallic materials by severe plastic deformation to gain fine grains via recrystallization or sub-grains by the rearrangement of dislocations. Secondly, the metallic liquid phase is solidified by a rapid cooling rate to achieve fine grains. In the following section, various techniques are reviewed.
1.2.3.1 Through severe plastic deformation
(A) Extrusion: Extrusion is a process by which a block of metal is reduced in cross section by forcing it to flow through a die orifice under a high pressure [16]. In general, this process can produce a semi-finished product of bars or hollow tubes to be convenient for final applications. Lin and Huang [12,15] have extruded the AZ31 and AZ91 magnesium alloys, which are usually manufactured by thixomolding and casting, to refine the grain size from the initial average 75 μm to 1~5 μm via the occurrence of dynamic recrystallization during extrusion. These refined magnesium alloys can perform good superplasticity and improved room temperature tensile
strength of 250~280 MPa for yielding stress and 300~350 MPa for ultimate tensile stress.
(B) Rolling: Because of lower ductility for magnesium alloys at room temperature, most rolling practices for magnesium alloys are operated at elevated temperatures, and annealing after a certain amount of strain is necessary. Otherwise, continuous deformation will result in edge buckle or wavy edge. Recently, Chang et al. [17]
refined grain from 13 μm to beyond 10 μm by rolling 2 mm thick plate to 0.5 mm for the AZ31 magnesium alloys, and the AZ31 thin sheet exhibited 300~325 MPa tensile strength. In addition, Kim et al. [18] also rolled the AZ61 sheets from a thickness of 2.15 mm to 0.5 mm after nine passes at 648 K, and effectively refined grains from 16 μm to 8.7 μm. These thin sheets also performed nice superplasticity at 623~673 K. There are many other studies on the rolling behavior of the Mg alloys.
(C) Equal channel angular pressing (ECAP): The ECAP process, performed by pressing samples into equal-diameter channel with a different exit direction, is a shear deformation maintaining the same input dimension of the pressed (or extruded) materials. The detailed illustration is shown in Fig. 1-1 [19]. This process can impose severe plastic deformation on materials after several passes to induce micro-scaled grains [19] or even to submicro- or nano-scaled extrafine grains [20].
Therefore, this process appears to be a powerful technique. Many researches have applied ECAP on the magnesium alloys. For example, Mabuchi et al. [21] refined the AZ91 alloy to 1 μm after eight passes, and the resulting alloys also showed a high elongation of 661% at a low temperature of 473 K. Lin [22] of our group also utilized this process to successfully refine AZ31 alloy, which was extruded into a bar with the 42:1 extrusion ratio. The original grain size of 12 μm was refined to 0.7 μm after eight passes, and the resulting alloys performed the elongation of 230% at 398 K (~0.44 Tm) and 1x10-4 s-1.
1.2.3.2 Through rapid solidification
(A) Spray forming: The spray forming process is an inert gas atomization of a liquid stream into variously sized droplets which are then propelled away from the region of atomization by fast flowing atomizing gas, as shown in Fig. 1-2 [23]. Droplets are subsequently deposited and collected by a substrate on which solidification takes place. Finally a coherent and near fully dense perform is produced. This process can produce bulk scale nano-grained materials via rapid solidification at a cooling rate about 102 ~103 K/s. Therefore, this process is very powerful tool, but it has an expensive issue related to patent. Chen and Tsao [24] ever spray-formed the AZ91 alloy with extra 3.34 wt% Si to result in uniform distribution of the Mg2Si and Mg17Al12 phase particles with equiaxed grain sizes measuring 2 μm. In contrast, usual cast AZ91 materials would possess much larger worm-shaped Mg17Al12 and much larger dendritic-shaped Mg2Si phases. The follow-up extrusion pressure for the as-spray-formed alloys was found to be lower than as-cast materials over the temperature of 320 ~ 420oC [24], implying that the as-spray-formed alloys possess finer structures and have better workability than the as-cast counterpart.
Except for the above refinement technologies, there are still other methods to refine the grain size. For example, Perez-Prado et al. [25] made use of accumulative roll bonding (ARB) to gain fine grains of the AZ alloys. Chang [26] applied reciprocal extrusion to refine the AZ91 alloys to 3.5 μm, and the refined alloys also exhibited superplasticity. Besides, Andrade et al. [27] ever applied shock-wave deformation to refine Cu alloys to 0.1 μm at a high strain rate (~104 s-1). Furthermore, Garces and Perez [28] could gain the columnar grain structure of the diameter of 0.2 μm of the Mg-14wt%Ti–1wt%Al–0.9wt%Mn alloys by the physical vapor deposition (PVD) technique.
Although researchers have developed numerous methods to refine the grain size, some techniques at the current stage are still limited to laboratory research trial, such as the ARB method. Some techniques are difficult to produce bulk scale materials, such as the PVD and shock-wave deformation methods. In addition, some techniques are expensive due to their patents, such as the spray forming. These factors influence their extensive industrial applications. The traditional extrusion and rolling are still the priority to be adopted in industry.
1.2.4 Anisotropy of magnesium alloys
The properties of many crystalline materials depend on the individual properties of the single crystal and also on parameters characterizing the polycrystalline state. Since preferred orientations of the grains are very common phenomena, the crystalline texture sometimes plays an important role in terms of numerous mechanical and physical behaviors. It is well known that the crystal symmetry of hexagonal closed-pack crystal structure such as Mg is poorer than that of face-centered cubic structure such as Al. Therefore, crystalline texture of the Mg alloys would often play a key factor to influence final mechanical properties. Wang and Huang [29] have reported that the {0002}<1010> texture is the main component among sharp basal {0002} texture for the AZ61 Mg alloys after warm plate extrusion. This preferred orientation has an important influence on the tensile deformation at room and elevated temperatures. The sample orientation at 45o to the extrusion direction has a lower yield strength and higher ductility at 25oC, as well as a higher superplastic elongation at 300~400oC, as compared with the specimens at 0o and 90o directions. In addition, Kim and Hong [30] reported the influence on the mechanical property from texture to be higher than that from grain refinement in the AZ61 Mg alloys after ECAP. The texture of the as-extruded
AZ61 Mg alloys gradually developed to (1010) [0111] + (1012) [1210] and )
2 1 10
( [1210] after 8 ECAP passes, following route Bc. This texture would lead the basal plane to be more favorably oriented for slip, allowing a lower stress to deform on the basal plane, even though grains were refined from 24 μm to 8 μm.
From the above two examples, it is well known that the preferred orientation of the Mg alloys often affects the final performance in mechanical properties. Except for the increase of strength or hardness induced by the grain refinement, the preferred orientation would also be another important factor influencing the mechanical properties of Mg based alloys. However, the evolution of the texture is not only affected by deformation, but also by casting, welding and heat treatment. Consequently, both primary and secondary processes for magnesium alloys that might induce texture evolution need to be taken into account in applications.
1.3 Metal matrix composites
Composite materials have long been designed to couple the weaker but more ductile phase (e.g. polymers or metals) with the stronger but more brittle one (e.g. ceramics), so as to tailor the needed properties of the resulting mixed materials. Among them, metal matrix composites have attracted attention since 1970’s.
Generally speaking, the reinforcements can be classified into the continuous and discontinuous types. The continuous reinforcements are in the form of continuous long fibers or short fibers such as carbon fibers, glass fibers, alumina fibers, and silicon carbon fibers. As for discontinuous reinforcements, they are in whisker form such as SiC or Al2O3, or in particulate form such as SiC, Al2O3, B4C and TiB2.
Continuous fibers could improve the elastic modulus and ultimate tensile strength of composites significantly. Because continuous fibers have a large aspect ratio of axial length to diametric width, such continuous fibers could effectively carry most of the load so as to enhance the elastic modulus and strength of the entire composites. From these viewpoints, continuous fibers seem to be perfect reinforcements, but they also have some disadvantages.
The strength along the radial direction is greatly lower than that along the axial direction, leading to an anisotropy problem [31]. This character will be unfavorable to some engineering applications. In order to improve this issue, most fibers are weaved into different directions to decrease anisotropy.
The composites reinforced by particulates or whiskers usually do not have a strong anisotropic character, and they sometimes can be almost isotropic. But discontinuous reinforcements would not effectively share the load. Although particulates or whiskers reinforced MMCs would not have compatible elastic modulus and strength as these reinforced by continuous fibers, the former can be processed by conventional methods such as extrusion, rolling and forging to meet the desired shapes.
It has long been known that the second phase inclusions or particles can inhibit grain growth in metallic materials. Therefore, one of the critical microstructure parameters is the particle interspacing, Ls, which can be roughly estimated from [32]
2 / 1
3 2 ⎟⎟
⎠
⎞
⎜⎜
⎝
= ⎛
f
s r V
L π
, (3)
where <r> is the average particle radius and Vf is the particle volume fraction. Previous reinforcing ceramic particles in the 1980’s usually measure around 10-50 μm in diameter,
later improvements have lead to smaller particles with a uniform size distribution in the range of 0.5-5 μm. With the typical reinforcing particles of Vf=20% (0.2) and <r>=10 μm in typical aluminum base composites, Ls will be around 30 μm. The resulting grain size after casting would also generally be in this range. Further hot extrusion may refine the grain size down to around 5~10 μm, and the toughening effects will only be moderately enhanced. With the abundant resources lately of the nano particles or nano carbon tubes (or wires) fabricated by various physical or chemical means, the reinforcements can be substantially smaller. If the reinforcement size is lowered to be submicron size or nano size, Ls can be reduced to submicron or nano range. It means that the grain size can be refined to submicron or nano grains. For example, Wang and Huang [33] ever added 1 vol % silica, 50 nm in diameter, to 6061 aluminum alloy and extruded this composite to obtain a 0.7 μm grain size.
There are various processes to produce particulates reinforced MMCs, and they could be divided into the solid-state method and liquid-state method. The detailed descriptions are presented in the following section.
1.3.1 Processing of metal matrix composites and magnesium matrix composites
1.3.1.1 Liquid-state methods
In principle, the liquid state route represents a very simple processing concept whereby particulates or whiskers are mixed into a light alloy melt, subsequently cast and then fabricated in a manner analogous to conventional unreinforced alloys, such as the stir casting [34-36] or semi-solid slurry stirring technique [37]. Sometimes, these particulates or whiskers
may be manufactured into preforms, and the melted alloy is subsequently introduced into preforms, for example, the squeeze casting [38-40] or molten metal infiltration technique [41].
Moreover, in order to uniformly disperse these particles, Lan et al. [42] reported the use of ultrasonic non-linear effects to disperse nano-sized ceramic particles in molten metal to fabricate the nano-sized SiC particle reinforced AZ91D magnesium composites. This way could effectively disperse SiC particle and reduce severe clustering occurrence. Detailed properties of Mg-matrix composites by various processes are listed in Table 1-4.
The interfacial reaction between the matrix and reinforcement in metal matrix composites systems is a very important issue. A good interfacial bond can effectively transfer load to reinforcements so as to enhance the stiffness and strength. For instance, Zheng et al.
[40] compared the difference of magnesium matrix composites mixed with SiC preforms in
the Al(PO3)3 binder between the one without the binder. If the composites are mixed into the Al(PO3)3 binder, the interfacial MgO phase would be formed mainly by the reaction between liquid Mg and Al(PO3)3 binder:
Al(PO3)3 + 9Mg→9MgO + Al +3P. (4)
Naturally, the composite with MgO to connect the matrix and reinforcement could increase the effective load transferred from matrix to reinforcement to result in a higher elastic modulus and strength. Zheng et al. [39] also observed similar interfacial reaction of MgO phase formation between the Al18B4O33 whiskers and AZ91, according to the following chemical equations:
Al18B4O33 + 33Mg→ 33MgO + 18Al + 4B, (5) ΔG1073K = -1783 kJ/mol.
Al18B4O33 + 33Mg→33MgAl2O4 +6Al +16B, (6) ΔG1073K = -1128 kJ/mol.
Both of the chemical reactions are thermodynamically possible at 800oC. However, Eq. (5)
has lower ΔG value, which implies that the formation of MgO is thermodynamically more favorable than that of MgAl2O4 (the spinel phase).
In the particulate reinforcement system, Bochenek and Braszczynska [35] observed the formation of MgO and Mg2Si compounds in the MgAl5 alloy matrix composites reinforced with SiC particles. They reported that the SiO2 film covering silicon carbide particles might have been reduced by molten magnesium according to the following two-stage reactions:
2Mg + SiO2 → 2MgO + Si, (7) ΔG1000K = -255 kJ/mol.
2Mg + Si → Mg2Si, (8)
ΔG1000K = -55 kJ/mol.
Moreover, Lan et al. [42] also found the existence of SiO2, MgO and Mg2Si after solidification of the AZ91D matrix composites reinforced with SiC particles. They used X-ray photoelectron spectroscopy (XPS) to identify the Si-2p spectra, and that of Si2p of AZ91D/5SiC consisted of three peaks at 99 eV, 100.3 eV and 102.8 eV. These were identified as those due to Si, SiC and SiO2, respectively. They also illustrated the formation of Mg2Si, SiO2 and MgO, as shown in Fig. 1-3. Moreover, Ye and Liu [43] reported that the composition of the matrix and the reinforcement materials, as well as the microstructure and porosity of the reinforcement materials would predominantly influence the interfacial reactions.
From the above examples, it needs to pay attention to the interfacial reaction between the molten metal and reinforcement. Promising chemical reactions will effectively enhance the stiffness and strength, but detrimental chemical reactions may form some extra undesired precipitates.
1.3.1.2 Solid-state methods
The solid state method utilizes higher the diffusion ability at elevated temperatures to sinter the entire composites, and the metal matrix is remained to be the solid state during processing. The most typical solid state methods are the powder metallurgy (PM) [44-47] and diffusion bonding [48]. In addition, the PM process would generally be followed by a secondary processing such as extrusion, rolling, forging and superplastic forming to form the final shapes of products as well as to reduce the porosity. Detailed properties of Mg-matrix composites made by the PM method are included in Table 1-5.
The interfacial reactions between the matrix and reinforcement prepared by the PM method were less frequently reported. Because the PM working temperature is below the melting point of the matrix, the interfacial reaction could not rapidly occur. However, once the sintering temperature reaches close to the liquid line, the interfacial reactions may occur.
For example, Li et al. [49] reported that MgO, MgAl2O4 and Mg2Si were the main reaction compounds in the Al-Mg alloy matrix composite reinforced with SiO2-based glass particles fabricated by the PM process by press sintering at 610oC. The main interfacial reactions of glass/Al-Mg alloy composite are:
SiO2(s) + 2Mg(l) → 2MgO(s) + Si(s), (9) 2SiO2(s) +2Al(s) + Mg(l) →MgAl2O4(s) + 2Si(s), (10) 2Mg(l) + Si(s) → Mg2Si. (11)
1.3.1.3 Other processing methods
In addition to these two categories of processing techniques described above, a number of other techniques have been explored for the fabrication of magnesium matrix composites, including in-situ synthesis [51] and spray forming [52,53]. The in-situ synthesis is a process wherein the reinforcements are formed in the matrix by a controlled metallurgical reaction.
During fabrication, one of the reacting elements is usually a constituent of the molten matrix alloy. The other reacting elements may be either externally-added fine powders or gaseous phases. Mabuchi et al. [51] utilized this concept and rapid solidification to fabricate the Mg-Mg2Si composite. Spray forming is also a potential processing to fabricate Mg based metal matrix composites with particulate reinforcement [52,53]. The method is that reinforcing particles are synchronously injected into the stream of the atomized matrix materials to form a bulk MMC before the solidification of an atomized matrix material dropping onto a substrate.
From the above introduction about the metal matrix composites, it is known that many microstructure factors would affect the final performance of MMCs. Table 1-6 [50] presents a list of microstructural factors that influence mechanical properties and fractures in discontinuously reinforced MMCs.
The above mentioned metal based composites, either with the conventional reinforcements ~20 μm or the more advanced ones ~0.5 μm in dimension, have micro-range grain size. Nevertheless, according to Eq. (3), Ls in the modified alloy with Vf = 3% (0.03) and <r> = 20 nm will be 167 nm. Fortunately, the success in fabrication of various nano-sized powders, wires or tubes has offered the new possibility in modifying existing commercial materials in terms of their functional or structural characteristics. However, nano-sized composites were focused on the polymer matrix modified by ceramic nano particles so as to significantly improve its mechanical or physical properties [55-59]. It is less frequently
addressed about the nano-sized particulates reinforced metal matrix composites except for a few papers [33,42,47,54]. The nano-sized particulates reinforced metal matrix composites might be able to stabilize grain size to less than 200 nm and enhance the ductility at elevated temperatures.
Although the nano-sized particulates reinforced metal matrix composites could have better properties, uniform dispersion of these nano-sized particles would be an extremely difficult task. Due to the high surface area ratio, nano-sized powders tend to cluster together, and they sometimes form micro-sized aggregates. After secondary treatments, these aggregates will act as defect to form the crack initiation to degrade the final performance.
Methods in dispersing the nano powders have been limitedly disclosed, mostly still protected by patents. How to effectively and simply disperse nano-sized particle into metal matrix will be one of the major efforts for the current proposed research work.
1.4 Basic characters of superplastic materials
Superplasticity is defined as the ability of polycrystalline materials to exhibit a large tensile elongation, typically in excess of two-hundred percent. Materials exhibiting superplastic character include metals, ceramic, intermetallics and composites [60]. This high ductility offers one possibility of one-step forming to complex components in industry to reduce spring back phenomenon and welding. A constitutive equation can describe the flow stress behavior during deformation,
Kεm
σ = & , (12)
where σ is the flow stress, K a material constant, and ε& the strain rate. The m value is the strain rate sensitivity of the material, and defined as
ε σ σ
ε ε σ
d m & d
& =
∂
= ∂ ln
ln . (13)
The m value is a representative index to superplastic materials, and it is described as the ability of resistance against necking during tensile deformation. Generally, common metals and alloys exhibit m=0.01-0.1 at room temperature, but superplastic materials possess higher values of 0.3-0.8 at superplastic working temperatures (0.6~0.8 Tm, where Tm is the melting point expressed in Kelvin).
For metals or metal matrix composites, in order to achieve fine structure superplasticity, they have to possess the following characters:
(A) Fine grain size: Typically, the grain size for metals should be less than 10 μm. The smaller grain size, and the major deformation mechanism (i.e. grain boundary sliding, GBS) will run smoothly during deformation.
(B) Second phase: The presence of a second-phase particles at grain boundaries will restrain the rapid grain growth at elevated temperatures. With a higher amount and more uniform dispersion of the second-phase particles, rapid grain growth will be more effectively restricted.
(C) Nature of the grain-boundary structure: The high-angle boundaries between adjacent grains will promote more effectively the occurrence of GBS. Low-angle boundaries do not readily slide under the appropriate shearing stress.
(D) Shape of grains: The shape of grains should be equiaxed so that the grain boundary can experience a shear stress allowing GBS to occur.
(E) Mobility of grain boundaries: During GBS, stress concentrations would develop at
the triple points to form cavities, inducing final fracture. However, the better ability for grain boundaries to migrate during GBS permits reduction of these stress concentrations.
The dominant deformation mechanism in superplastic materials is GBS under a suitable range of strain rate and temperature. In order to maintain continuous operation of GBS during deformation, appropriate accommodation mechanisms are needed for the grain compatibility during GBS. The accommodation mechanisms can be divided into three groups:
accommodation by diffusional transport [61], accommodation by dislocation motion [62,63], and combined mechanisms including both accommodation processes of dislocation motion and diffusional transport [64,65]. Moreover, Higashi and Mabuchi [66-68] have addressed a new model in which liquid phase plays a role of accommodation mechanism for metal matrix composites exhibiting high strain rate superplasticity. A schematic illustration of the accommodation mechanism by the liquid phase for metal matrix composites is shown in Fig.
1-4.
1.5 Superplasticity of magnesium alloys and magnesium matrix composites
The traditional superplastic materials exhibit good superplasticity at temperatures greater than ~0.7 Tm and strain rates lower than 1x10-3 s-1. However, it is quite uneconomic to superplastically form the materials at higher temperatures and lower strain rates in terms of the viewpoint of industry consideration. Therefore, the development of superplasticity has been focused on high strain rate superplasticity (HSRSP), which is defined as superplasticity occurring at strain rates at or greater than 10-2 s-1, and low temperature superplasticity (LTSP), which is defined as working temperatures lower than 0.6 Tm. The following sections
introduce the related reports on HSRSP and LTSP of magnesium alloys and magnesium matrix composites.
1.5.1 HSRSP and LTSP of magnesium alloys
The development of HSRSP in magnesium series materials was initially focused on the magnesium matrix composites [69-71], with the HSRSP behavior of magnesium alloys was less addressed in its early stage. However, Mabuchi et al. [75] first reported HSRSP in a ZK61 alloy. The ZK61 alloy was produced by the PM method, sintered at 250oC and then extruded at 250oC with a reduction ratio of 100:1. The grains were equiaxed with a fine grain size of 0.5 μm, and an elongation of 283% was achieved at 200oC and 1x10-2 s-1. On the other side, our laboratory also developed the method of one-step extrusion with high extrusion ratio at extrusion temperature of 280~300oC [22,76]. The grain size of AZ31 after extrusion became 1~4 μm and maximum elongation of 1000% was achieved at 300oC and 1x10-2 s-1.
As for LTSP, Mabuchi et al. [77,78] reported that AZ91 alloy, which was extruded at 250oC and then further processed by ECAP at 175oC with a total strain level of 8.05, could reveal an elongation of 661% at 200oC and 6.2x10-5 s-1. Moreover, Lin [22,79] also applied the similar method of extrusion and ECAP on AZ31 alloy to get a good elongation of 200% at 125oC and 1x10-4 s-1, as well as 461% at 150oC and 1x10-4 s-1.
Watanabe et al. [80] boldly predicted the possibility of culminating of HSRSP and LTSP in extremely fine grained magnesium alloys. They suggested that the required grain size of ≦ 0.4 μm can obtain both HSRSP and LTSP. The main factor is that the grain boundary diffusion, δDgb, for magnesium is two orders of magnitude higher than that for aluminum [81]. This will make magnesium alloys get better accommodation than aluminum alloys at
