Executive Summary : | Martensitic phase transformations (MPTs) are central in exhibiting extraordinary properties such as superelasticity and shape memory effects in shape memory alloys (SMAs), for example, nitinol, CuNiAl, InTh alloys, etc. These alloys find wide applications in control systems, medical appliances like dental braces, stents, etc. In other industrially important materials such as dual-phase steel, ferromagnetic alloys, caloric materials, graphite/diamond, etc., MPTs play an important role in exhibiting enhanced mechanical and other properties. During the MPTs a parent phase austenite (high-temperature phase) transforms to the product phase martensite (low-temperature phase) which in general has multiple variants and they are crystallographically equivalent. In the realistic single grains of such phase-changing materials, the heterogeneous sites including the shear bands, dislocations, crack tips and crack surfaces, external surfaces, etc., are the regions where the martensitic phase nucleates. The polycrystalline solids of these materials have additional heterogeneous sites including the grain boundaries and triple junctions which are also favorable to M nucleation. While the criteria for homogeneous nucleation, which is generally driven by the thermal fluctuations, of the martensitic phase in crystalline solids are well known, only a few very oversimplified models for the heterogeneous nucleation neglecting mechanics are present in the literature. Notably, the stress fields are important as MPTs involve significant volumetric and shear strains in the materials, and they must be taken into consideration in obtaining the robust criteria. The aim is to conduct the theoretical and computational studies in order to derive the general criteria for the heterogeneous nucleation from each of these heterogeneities considering mechanics and also to understand the growth of the phases. The extended variational thermodynamic framework developed by Gibbs would be used, and the finite element method would be used for the numerical studies. |