An optimized random structures generator governed by chemical short-range order for multi-component solid solutions

Multi-component random solid solutions have caused much interest for research because they often exhibit excellent mechanical properties. Many computational efforts have been made through various length scales, by first principles, molecular dynamics or finite elements. Here, a quasi-simulated-annealing (Markov Chain Monte Carlo algorithm) with prior burn-in procedure using chemical short-range order parameters as input values is developed for searching the global optimal distribution of species with regard to the phase stability of face-centered cubic CoCrNi alloys. The results show that this new approach could effectively generate the stable structures with a desired concentration and moreover, the stacking fault energies of these configurations have less uncertainties statistically. (paper)