A modified grain growth Monte Carlo algorithm for increased calculation speed in the presence of Zener drag effect

Abstract

In its basic implementation, the grain growth Monte Carlo Potts algorithm is repeated millions of times in random matrix positions by using random number generator. The probability for possible site reorientation is bigger for a subset of lattice sites, and the remaining sites are rarely tested. However this method requires a long calculation time for large lattice systems. To increase the calculation speed for grain growth simulation, one proposes two modifications to the classical MC model. First, all lattice sites will be tested for reorientation at random with the same probability, i.e. without repetition for every Monte Carlo step. Second, the site energy will be divided by a parameter that represents the effective force acting on grain boundaries in the presence of particles. The second modification accelerates the growth of large grains to the detriment of neighboring small grains. The efficiency of these modifications has been tested for grain growth with the presence of Zener drag effect and its absence.