Simulation of grain growth in polycrystalline materials: A signal processing perspective

Polycrystalline materials are solids that are made up of crystals, also called grains, that have varying sizes and orientations. Materials of this class are used in many applications, including steels, films, and electronic interconnects, to name just a few. The factors that influence how the grains in a polycrystalline form are not well understood, although materials scientists and engineers have been studying this issue for many years. One approach to understanding how polycrystalline materials form is through simulation using Monte Carlo methods. In this paper we describe some of these simulation methods in the context of signal processing, which is much different from the materials community´s view of the work. In particular, we describe how materials scientists and engineers have modified traditional Metropolis sampling using prior information about the physics of grain growth to simulate realistic grain growth. We present results of experiments we have performed to simulate one aspect of importance in grain growth: the mobility of boundaries between neighboring grains.