Multiscale Simulations of Anisotropic Grain Growth Using Wavelet Based Multiresolution Analysis

Abstract

The present work serves to document the development and findings associated with a wavelet-based multiscale simulation analysis for anisotropic grain growth of a two-dimensional polycrystalline material. In particular, lattice-based Monte Carlo and atomically-based Molecular Dynamics simulations are used to compute the grain boundary energies over their respective spatial domains. Serial coupling is performed utilizing an orthonormal set of Haar wavelet transforms embedded within a corresponding multiresolution analysis. For the Monte Carlo approach, anisotropies in grain boundary energies, caused by differences in grain orientation (texturing), are examined using two distinct methods, while the molecular dynamics simulations, offering inherent anisotropy, are conducted assuming the interatomic Lennard Jones potential. Among other findings, under the present context, the results confirm the viability of the wavelet-based multiresolution analysis (MRA) method for use as a potential coupling agent, and provide substantiation for its use with other applications. The results further offer quantitative comparisons between isotropic and anisotropic modeling results, and demonstrate their range of applicability.