Sparse random Fourier features based interatomic potentials for high entropy alloys

TL;DR

This paper introduces a sparse random Fourier features-based interatomic potential that enables efficient and accurate molecular dynamics simulations of high entropy alloys, capturing melting, defects, and high-temperature diffusion with reduced computational cost.

Contribution

The work develops a novel sparse random Fourier features-based interatomic potential for high entropy alloys, significantly reducing computational cost while maintaining high accuracy in modeling atomic interactions.

Findings

Atomic forces within 0.08 eV/Å of DFT

MD predictions within 7% of DFT values

94% reduction in computational cost

Abstract

Computational modeling of high entropy alloys (HEA) is challenging given the scalability issues of Density functional theory (DFT) and the non-availability of Interatomic potentials (IP) for molecular dynamics simulations (MD). This work presents a computationally efficient IP for modeling complex elemental interactions present in HEAs. The proposed random features-based IP can accurately model melting behaviour along with various process-related defects. The disordering of atoms during the melting process was simulated. Predicted atomic forces are within 0.08 eV/$\unicode{xC5}$ of corresponding DFT forces. MD simulations predictions of mechanical and thermal properties are within 7$\%$ of the DFT values. High-temperature self-diffusion in the alloy system was investigated using the IP. A novel sparse model is also proposed which reduces the computational cost by 94$\%$ without compromising on the force prediction accuracy.