Investigation of Primary Radiation Damage in Nanocrystalline Tantalum Using Machine-Learning Interatomic Potential: An Atomistic Simulation Study

Tantalum is a plasma-facing material possessing superior thermal and mechanical strength, high melting point, chemically resistant refractory element and efficiently arrests radiation-induced material degradation. Under the impact of radiation damage, the metallic structural components of the nuclear power plants, like fuel cladding components, Reactor Pressure Vessels (RPV), etc., suffer deterioration in their microstructures and transmutation in the thermodynamical and mechanical behavior. In the present Molecular Dynamics (MD) based simulations study, bcc nanocrystalline (NC) tantalum system was irradiated at varying magnitudes of primary knock-on atom (PKA) recoil energies. The material response towards irradiation is stochastic for both length and time scales. The MD tool is apt for encountering and quantifying the primary radiation damage (or displacement cascade) in the atomistic scale. The reliability of the MD results depends on the critical choice of the interatomic potential files. To mimic the fast collisions of bombarding atoms and Ta–Ta atomic interactions, machine learning (ML) based interatomic potential, Spectral Neighbor Analysis Potential (SNAP) and universal repulsive ZBL potentials were splined at suitable cut-off distances respectively. The dynamic evolution of radiation-induced point defects, dislocations, and defect clusters was investigated. The grain boundaries acted as the sink sites of the interstitials and vacancies generated during the cascade formation, and the magnitude of the point defects increased with the increase of the PKA energy magnitude, respectively.