Kinetic Monte Carlo simulations of solute clustering during quenching and aging of Al–Mg–Zn alloys

Abstract

The physical mechanisms behind cluster formation during quenching and aging of age-hardening metallic alloys are poorly understood based on classical nucleation and growth theories, especially in multicomponent alloys with supersaturated vacancies and fast-diffusing solute atoms. Here, solute clustering in an Al–Mg–Zn alloy during quenching immediately after high-temperature (800 K) solution treatment is modeled with a newly developed kinetic Monte Carlo (kMC) framework. This includes accurate surrogate models trained using first-principles-calculated data to predict vacancy migration energetics on the fly as functions of local lattice occupations. First, kMC simulations at constant aging temperatures revealed that temporal evolution of the number, sizes, and compositions of solute clusters change with aging temperature. Such changes are consistent with our analyses based on classical nucleation theory (CNT) and Monte Carlo simulations. Second, the kMC algorithm was revised for simulating quenching processes by iteratively updating the kMC temperature based on simulated cooling rate profiles. These quenching simulations show that rapid solute clustering mainly occurs from ∼600 to ∼400 K during cooling. Below ∼400 K, the clustering is suppressed to a steady state because vacancies are trapped by stable clusters with sizes of ∼2 nm and high magnitudes of formation energies. The configurations of these steady-state clusters closely resemble the solute clusters we observed experimentally in 7XXX series Al–Mg–Zn-based alloy samples immediately after quenching. A two-stage vacancy trapping mechanism revealed in our simulations can provide physical guidance to tune precipitation kinetics during natural and artificial aging.