(Dr. David McDowell, advisor)
"3D Finite-Element Simulation of Random & Textured Shape Memory Alloys"
The behavior of shape memory alloys (SMA’s) in the pseudoelastic regime is of considerable interest in applications ranging from biomedical to automotive to aerospace. This behavior, although nonlinear, differs significantly from that of conventional metals in which dislocation motion governs. In SMA’s, mechanically reversible phase transformations result in hysterisis. The behavior in compression differs substantially from that in tension due to the nonsymmetric nature of the displacive martensitic shear transformation.
The use of modern methods of scale transition has allowed very accurate micromechanical simulations of the behavior of shape memory alloys under complex loadings, which were difficult to perform experimentally. Very few three-dimensional finite element (3-D FE) analyses have been performed in the literature (cf. PhD thesis of T.J. Lim, Georgia Tech) to understand the morphological distribution of martensite and the autocatalytic nature of the austenite to martensite phase transformation. Furthermore, it is unknown if predictions of simple self-consistent polycrystal models are similar to those achieved using more rigorous 3D FE analyses.
In this thesis, (1) scale transition micromechanical simulations are used to determine SMA behaviour as a function of crystallographic texture. (2) 3D FE simulations are compared to results of experiments as a function of the crystallographic texture. It is found that texture plays a strong role in the anisotropy of SMA behavior.