A new method of alloy design using a genetic algorithm and molecular dynamics simulation and its application to nickel-based superalloys

Abstract

A new method of alloy design was proposed using a genetic algorithm and molecular dynamics simulations. In this method, evaluations of physical quantities using molecular dynamics simulations and an efficient global search using a genetic algorithm allow optimization of an equilibrium composition for a multi-component alloy satisfying a given design rule. The alloy design method was used to design nickel-based superalloys in the elemental systems simulating those of TMS-63 and CMSX-2. The obtained atom fractions in the γ phase and γ′ phase, and atom fractions in the Ni and Al sublattices of the γ′ phase were compared with the experimental values. The comparisons showed that the global features of TMS-63 and CMSX-2 were reproduced well. Furthermore negative lattice misfits were of about 1% and positive elastic misfits for c11 and c12 were of about 10% for both alloys. Finally sources of the inaccuracies for the obtained atom fractions and possible improvements for the design method were discussed.