Title: A Tungsten Deep Neural-Network Potential for Simulating Mechanical Property Degradation Under Fusion Service Environment

Abstract: Tungsten is a promising candidate material in fusion energy facilities. Molecular dynamics (MD) simulations reveal the atomistic scale mechanisms, so they are crucial for the understanding of the macroscopic property deterioration of tungsten under harsh and complex service environments. The interatomic potential used in the MD simulations is required to accurately describe a wide spectrum of relevant defect properties, which is by far challenging to the existing interatomic potentials. In this paper, we propose a new three-body embedding descriptor and hybridize it into the Deep-Potential (DP) framework, an end-to-end deep learning interatomic potential model. Trained with the dataset generated by a concurrent learning method, the potential model for tungsten, named by DP-HYB, is able to accurately predict a wide range of properties including elastic constants, stacking fault energy, the formation energies of free surfaces and point defects, which are included in the training dataset, and formation energies of grain boundaries and prismatic loops, the core structure of screw dislocation, the Peierls barrier and the transition path of the screw dislocation migration, which are not explicitly included in the training dataset. The DP-HYB is a good candidate for the atomistic simulations of tungsten property deterioration, especially those involving the mechanical property degradation under the harsh fusion service environment.