Nanoindentation induced plasticity in equiatomic MoTaW alloys by experimentally guided machine learning molecular dynamics simulations

TL;DR

This study combines experiments and machine learning-enhanced molecular dynamics simulations to understand nanoindentation-induced plasticity in MoTaW alloys, revealing orientation-dependent deformation mechanisms and dislocation behaviors.

Contribution

It introduces a combined experimental and atomistic modeling approach using machine learning potentials to accurately simulate and analyze plasticity in MoTaW alloys.

Findings

Excellent agreement between experiment and simulation in elastic response.

Orientation-dependent slip and dislocation activity observed.

Elevated fault energies suggest dislocation-mediated plasticity dominance.

Abstract

Refractory complex concentrated alloys (RCCA) exhibit exceptional strength and thermal stability, yet their plastic deformation mechanisms under complex contact loading remain insufficiently understood. Here, the nanoindentation response of an equiatomic MoTaW alloy is investigated through a combined experimental and atomistically resolved modeling approach. Spherical nanoindentation experiments are coupled with large scale molecular dynamics simulations employing a tabulated low dimensional Gaussian Approximation Potential (tabGAP), enabling near DFT accuracy. A physics based similarity criterion, implemented via PCA of load-displacement curves, is used to identify mechanically representative experimental responses for quantitative comparison with simulations. Indentation stress-strain curves are constructed yielding excellent agreement in the elastic regime between experiment and simulation, with reduced Young's moduli of approximately 270 GPa. Generalized stacking fault energy calculations reveal elevated unstable stacking- and twinning-fault energies in MoTaW relative to pure refractory elements, indicating suppressed localized shear and a preference for dislocation-mediated plasticity. Atomistic analyses demonstrate a strong crystallographic dependence of plastic deformation, with symmetric {110}<111> slip activation and four-fold rosette pile-ups for the [001] orientation, and anisotropic slip, strain localization, and enhanced junction formation for [011]. Local entropy and polyhedral template matching analyses further elucidate dislocation network evolution and deformation-induced local structural transformations. Overall, this study establishes a direct mechanistic link between fault energetics, orientation-dependent dislocation activity, and experimentally observed nanoindentation behavior in RCCA.