Bi-region kinetic monte carlo simulation of grain growth

There exist deviations between the results of most kinetic Monte Carlo (kMC) simulation of grain growth and the theoretical value of grain growth. An important reason is that the Read-Shockley formulation in most kMC simulation is only suitable for the computation of small-angle grain boundary. Aiming at this problem, a bi-region kMC method is proposed in this paper, which handles both situation of small-angle grain boundary and large-angle grain boundary. In the proposed method, the transition probability of orientation number is obtained using lattice energy based on potential functions, not using Read-Shockley formula. In order to solve the large computational problem caused by introducing potential functions, kMC single-region structure of traditional dynamics is divided into two regions: the one without grain boundary and the one with the grain boundary. Due to potential function computation only used in the region of grain boundary, the computing time is reduced. Experiment results show that the micro-structure evolution of bi-region kMC method is more consistent with the feature of grain growth than that of existing traditional and improved kMC, and the grain growth exponent is closer to the theoretical value than that of existing kMC.