Computational Study of Microstructure Evolution during Phase Transformations.

다국어 초록 (Multilingual Abstract)

Phase transformation is always a critical topic in the study of materials science. Most people have been familiar with some transformations between solid and liquid, such as ice to water, or transformations between liquid and gas, such as water to vapor. Besides, the phase transformations in solids also occur everywhere. Some solid phase transformations occur due to temperature variations. Those transformations may also be affected by external stress or strain, as seen in shape memory alloys (SMAs). The solid-solid transformation is considered to be the one of the most effective ways to tailor the microstructure and properties of the alloys, moreover, it sometimes strengthens the structural materials. There are some types of solid-state phase transformations that are hard to characterize in the traditional experiments. The difficulty mainly comes from two aspects. Firstly, some of the phase transformations happen too fast, such as martensitic transformation. The speed of the martensitic transformation is close to the speed of sound traveling in solids (~1000m/s), which makes it difficult to know how it starts and evolves. Secondly, some of the phase transformation processes are too slow, such as oxidation. It could take years to form a continuous layer of oxides in microns.With the fast development of high-performance computing, the study of phase transformations through computational tools attracts more and more attention. The objective of this thesis is to apply computational tools to study the two types of phase transformations and their corresponding mechanical properties: precipitation and martensitic transformation.As one of the most important structural phase transformations discovered in metallurgy and materials science, martensitic transformation (MT) has been attracting continued attention since its discovery in the late nineteenth century till today because it relates closely to the functional properties of NiTi-based alloy such as the superelasticity and shape memory effect. Most importantly, MT can be tailored through nano-scale defects in materials. Firstly, nano-scale defects in the B2 parent phase are known to have profound impacts on the properties of NiTi-based shape memory alloys. We employed the phase field models (PFM) to study the effects of two typical nano-scale defects, nano-scale precipitates and voids, on MT. The simulation of precipitation unveiled the mechanical and chemical effects on the behavior of MT in NiTi-Hf alloys. Moreover, the simulation of MT with the coexistence of precipitates explained the mechanism of two typical patterns of martensite. The results indicates that the stress-strain response has great dependence on the concentration heterogeneity in the matrix as well as precipitate microstructures. Through the simulation we proved the feasibility to achieve linear or quasi-linear superelasticity with high recoverable strain (up to 4%) in NiTi-Hf alloys after the precipitation. In the simulation of MT under the effects of nano voids in NiTi, we observed that martensite could be confined in the interspacing area between voids. Besides, MT could be triggered at lower critical stress with larger volume fraction of voids. This simulation may shed lights on the design of the porous NiTi alloys for the biomedical application.In superalloys, the microstructure of precipitates can be altered by the formation of an oxide layer on the surface. It is observed that the γ' precipitates dissolve at the near-surface region with the formation of the oxide layer in the alloy. We employed DICTRA module in Thermo-calc Software to solve the multicomponent diffusion equations in alloy H282 with an outward flux of chromium or aluminum due to oxidation and applied PFM to simulate the dissolution of precipitates. The local variation of precipitates? volume fraction as a function of oxidation time has been quantitatively determined. The calculation of precipitates depletion depth shows good agreement with the experiments. The highly heterogeneous structure of γ' precipitates is expected to have a significant effect on the creep behavior of the alloy.