Nanoindentation Study on BCC Metals and FCC NixFe(1-x) Alloys: Unveiling Plastic Deformation through Machine-Learned MD Simulations
Deciphering the plasticity mechanisms in emerging alloys is paramount for optimizing their mechanical properties. This study employs an extensive series of machine-learned molecular dynamics (ML-MD) simulations to explore the behavior of single crystals in pure BCC tungsten, molybdenum, tantalum, and vanadium. This extends to their random solid solutions in W–Mo, W-V, and W-Ta alloys, as well as FCC Ni-based alloys. We analyze dynamic deformation processes, defect nucleation, and evolution, alongside concurrent stress–strain responses. Additionally, atomic shear strain mapping provides insights into surface morphology and plastic deformation. In BCC metals, the introduction of Mo, Ta, and V atoms into the W matrices induces lattice strain and distortion, heightening material resistance to deformation. This impedes dislocation mobility, especially for dislocation loops with a Burgers vector of 1/2⟨111⟩. Our comparative analysis reveals a remarkable suppression of the plastic zone size in the equiatomic W–V alloy. In contrast, the equiatomic W–Mo alloy lacks a similarly clear prediction for optimal hardening. Furthermore, we explore the influence of tantalum concentration in W matrices, focusing on twinning and anti-twinning mechanisms during nanoindentation, contributing to material hardening.
For FCC NixFe(1-x) alloys, we observe significant hardening effects due to Fe concentrations in Ni-based alloys. Experimental load–displacement data for the (001) crystal orientation show qualitative agreement with MD simulation results, providing strong evidence that the main strengthening factors are associated with sluggish dislocation diffusion, reduced defect sizes, and the nucleation of tetrahedral stacking faults. In this context, interstitial-type prismatic dislocation loops, mainly formed by ⅙⟨112⟩ Shockley dislocations, are nucleated during the loading process. Their interaction leads to the formation of pyramidal-shaped stacking faults, primarily created by ⅓⟨100⟩ Hirth dislocation lines.
The observation of both types of defects coexisting in the same plastic deformation zone is consistent with both approaches. Reported mechanical data, measured experimentally and interpreted numerically, also align with microstructural SEM and TEM investigations. Throughout this discussion, the highlights of the advantages and limitations of both conventional interatomic potentials and machine-learned models when simulating nanoindentation tests will be presented and discussed.
Go to the Events page.