The primary objectives of this study were: (1) to verify the effectiveness of ECAE to induce equal amounts of strain and grain refinement in bars of different cross-sectional areas, (2) to determine the effectiveness of ECAE in breaking down the as-cast macrostructure in CDA 101 Cu and in producing a homogeneous material containing micron-scale grains upon recrystallization, and (3) to determine a thermomechanical processing (TMP) schedule (from the ones examined) that produces the best microstructure in terms of grain size and uniformity. The effects of extrus ion route, levels of strain and intermediate heat treatment were investigated. To achieve the first objective, bars having square cross-sections of three different sizes, 19 mm, 25 mm and 50 mm, were processed up to eight ECAE passes through routes A, B, C and E. To achieve the second and third objectives, bars were processed up to eight ECAE passes with and without intermediate heat treatments through routes Bc, C, E and F. ECAE processing was carried out in a 90o extrusion die with sliding walls at an extrusion speed of 2.5 mm/s. Recrystallization studies were carried out on the processed material to evaluate the recrystallization behavior and thermal stability of the material. The as-worked and recrystallized materials were characterized by Vickers microhardness, optical microscopy (OM) and transmission electron microscopy (TEM). Results indicate that similar hardness values, sub-grain morphology and recrystallized grain size are generated in the three bars having different cross-sectional sizes processed through ECAE. ECAE is shown to induce uniform strain in all three billet sizes. ECAE is therefore shown to be effective in scale-up to a size of at least 50 mm, with larger billets giving better load efficiency. Results from the later parts of this study indicate that eight extrusion passes via route Bc produces the best microstructure in terms of grain size and microstructural uniformity. The routes can be arranged in the sequence Bc> E, F> C for their ability to produce a uniform recrystallized microstructure with small average grain size. Macroscopic shear bands are sometimes generated during extrusion depending upon the initial grain morphology and texture of the material.

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第21767号 / 工博第4584号 / 新制||工||1714(附属図書館) / 京都大学大学院工学研究科材料工学専攻 / (主査)教授 辻 伸泰, 教授 乾 晴行, 教授 安田 秀幸 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Shape memory alloys when deformed can produce strains as high as 8%. Heating results in a phase transformation and associated recovery of all the accumulated strain, a phenomenon known as shape memory. This strain recovery can occur against large forces, resulting in their use as actuators. The goal of this project is to lower the operating temperature range of shape memory alloys in order for them to be used in cryogenic switches, seals, valves, fluid-line repair and self-healing gaskets for space related technologies. The Ni-Ti-Fe alloy system, previously used in Grumman F-14 aircrafts and activated at 120 K, is further developed through arc-melting a range of compositions and subsequent thermo-mechanical processing. A controlled atmosphere arc-melting facility and vertical vacuum quench furnace facility was commissioned to fabricate these alloys. The facility can create a vacuum of 10-7 Torr and heat treat samples up to 977 °C. High purity powders of Ni, Ti and Fe in varying ratios were mixed and arc-melted into small buttons weighing 0.010 kg to 0.025 kg. The alloys were subjected to solutionizing and aging treatments. A combination of rolling, electro-discharge machining and low-speed cutting techniques were used to produce strips. Successful rolling experiments highlighted the workability of these alloys. The shape memory effect was successfully demonstrated at liquid nitrogen temperatures through a constrained recovery experiment that generated stresses of over 40 MPa. Differential scanning calorimetry (DSC) and a dilatometry setup was used to characterize the fabricated materials and determine relationships between composition, thermo-mechanical processing parameters and transformation temperatures.

In the drive towards higher strength alloys, a diverse range of alloying elements is employed to enhance their strength and ductility. Limited solid solubility of these elements in steel leads to segregation during casting which affects the entire down-stream processing and eventually the mechanical properties of the finished product. The work presented in this PhD shows that segregation of alloying elements during casting, particularly aluminium, leads directly to banding in the final product. It has been demonstrated that no significant homogenisation is possible in this alloy within practical time constraints of the industrial thermo-mechanical process. A through-process model was developed to design a thermo-mechanical treatment aimed at reducing the effects of segregation on the formation of banding. A new polynomial function for calculating the local phase transformation temperature (Ae3) between the austenite + ferrite and the fully austenitic phase fields during heating and cooling of steel is presented. Material was produced both with and without banding and used to study the effect upon the mechanical properties. The banded steel variants show a significant reduction in tensile strength for a similar level of ductility compared to non-banded variants. In situ measurement under uniaxial loading using high-energy synchrotron diffraction allowed direct quantification of the impact of the mechanically induced transformation of metastable austenite on the work- hardening behaviour. The results reveal that the mechanically induced transformation of austenite does not begin until the onset of matrix yielding and the experimental evidence demonstrates that the austenite to martensite transformation increases the work-hardening rate of the ferrite phase and delays the onset of Stage-III hardening until the yield point of austenite. The increase in work-hardening rate (and thus work required) supports a driving force approach to transformation induced plasticity. The transformation work required leads to an increase in the macroscopic work-hardening rate after matrix yielding which offsets the decrease in the work-hardening rate in the ferrite and martensite phases up to the UTS. Steels with a high degree of banding do not show this extra contribution due to the more dominant anisotropic effect of martensite bands on the work-hardening of ferrite coupled to increased mechanical austenite stability as a result of increased carbon content. A list of revisions as requested by the examiners is produced on pages 18 and 19 of the thesis for examination. Abstract: In the drive towards higher strength alloys, a diverse range of alloying elements is employed to enhance their strength and ductility. Limited solid solubility of these elements in steel leads to segregation during casting which affects the entire down-stream processing and eventually the mechanical properties of the finished product. The work presented in this PhD shows that segregation of alloying elements during casting, particularly aluminium, leads directly to banding in the final product. It has been demonstrated that no significant homogenisation is possible in this alloy within practical time constraints of the industrial thermo-mechanical process. A through-process model was developed to design a thermo-mechanical treatment aimed at reducing the effects of segregation on the formation of banding. A new polynomial function for calculating the local phase transformation temperature (Ae3) between the austenite + ferrite and the fully austenitic phase fields during heating and cooling of steel is presented. Material was produced both with and without banding and used to study the effect upon the mechanical properties. The banded steel variants show a significant reduction in tensile strength for a similar level of ductility compared to non-banded variants. In situ measurement under uniaxial loading using high-energy synchrotron diffraction allowed direct quantification of the impact of the mechanically induced transformation of metastable austenite on the work- hardening behaviour. The results reveal that the mechanically induced transformation of austenite does not begin until the onset of matrix yielding and the experimental evidence demonstrates that the austenite to martensite transformation increases the work-hardening rate of the ferrite phase and delays the onset of Stage-III hardening until the yield point of austenite. The increase in work-hardening rate (and thus work required) supports a driving force approach to transformation induced plasticity. The transformation work required leads to an increase in the macroscopic work-hardening rate after matrix yielding which offsets the decrease in the work-hardening rate in the ferrite and martensite phases up to the UTS. Steels with a high degree of banding do not show this extra contribution due to the more dominant anisotropic effect of martensite bands on the work-hardening of ferrite coupled to increased mechanical austenite stability as a result of increased carbon content.