COMPUTATIONAL ASSESSMENT OF KEY PROPERTIES IN CoNiRu ALLOYS USING ATOMISTIC SIMULATIONS

Abstract

Reliable prediction of material behavior at the atomic scale is critical for the design, improvement, and development of high-performance structural alloys. Among them, the CoNiRu-based alloy has emerged as a promising candidate to many applications in aerospace, nuclear energy, and high-temperature manufacturing due to its high tolerance to mechanical loading, thermal stability, and potential for use in extreme environments. In this study, the CoNiRu ternary alloy system is systematically evaluated using atomistic simulations implemented in Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS), where a universal machine learning interatomic potential (uMLIP) model is benchmarked against density functional theory (DFT) results. The uMLIP is assessed for its ability to predict lattice parameters and elastic constants in face-centered cubic (FCC) and hexagonal close-packed (HCP) structures at 0 K and 300 K, as well as generalized stacking fault energies (GSFE) in FCC and HCP structures at 0 K. The findings of this work offer critical insight into whether uMLIP can accurately approximate key mechanical and structural properties of CoNiRu alloys when compared to DFT results. The capabilities and limitations of uMLIP for modeling multi-component alloy systems will be tested, with implications for accelerated alloy design and deployment in advanced engineering applications. Gathering atomic simulation data using uMLIP proves to be promising and resource-efficient when juxtaposed with DFT method of data collection.