Atomistic modeling of grain boundary motion as a random walk

DK Chen and Y Kulkarni, PHYSICAL REVIEW MATERIALS, 2, 093605 (2018).

DOI: 10.1103/PhysRevMaterials.2.093605

Mechanical behavior of polycrystalline materials, such as creep and microstructural evolution, is dramatically impacted by the mobility of grain boundaries. Existing methods for extracting mobility that combine atomistic simulations with conventional Brownian-like random walk model for grain boundary motion miss critical insights and are reliable only above very high (and unrealistic) temperatures. In this paper, we present a computational method based on the classical Green-Kubo relation for computing interface mobility over a wide range of temperatures. Our study makes an intriguing revelation that the severe time limitation of molecular dynamics simulations warrants the use of a generalized diffusion equation for Brownian particles not accounted for in current studies. Moreover, the method furnishes analytical expressions for the interface mobility in terms of the velocity autocorrelation functions. Taken together, our results possibly provide the first reliable estimates for interface mobility in the limit of zero driving forces. This is in sharp contrast to studies based on the widely used interface random walk approach, which extracts mobility by fitting the conventional Brownian motion equation to atomistic simulations. The efficiency of the method and applicability over a range of temperatures should open avenues for integration of computations and experiments to understand and engineer material systems with desired properties.

Return to Publications page