Atomistic simulation of creep deformation mechanisms in nickel-based single crystal superalloys

Abstract

Abstract In this paper, the creep deformation mechanisms are investigated in nickel-based single crystal superalloys. Two-dimensional molecular dynamics (MD) simulations are conducted to model various temperatures, stress conditions, and phase interface crystal orientations. Ni-based single-crystal superalloys are of great importance in the aircraft industry due to their excellent high temperature creep resistance. This characteristic mainly originates from two features considered in their structure; firstly, their two-phase micro-structure comprising gamma γ and gamma prime γ ′ , and secondly the nature of this superalloy itself, which is a single-crystal. MD is a powerful tool to gain insight into creep behavior at small scales, where dislocations and high-temperature diffusional phenomena are the most critical deformation agents. The parameters considered in the creep deformation are temperature, stress, and phase interface crystal orientation. The simulations are observed in various temperature conditions including 1100, 1200, 1400, 1600, and 1700 K. Stress levels are applied from 0.5 to 5.0 GPa with a sequence of 0.5 GPa, and phase interface crystal orientations are imposed on (001), (011) and (111). Various mechanisms are detected, including the γ ′ precipitate shearing, micro-twinning, the diffusion-mediated climb that allows the dislocations to bypass the γ ′ precipitates, and rafting that refers to directional γ ′ coarsening. The results are shown to be in good agreement with available experimental data. The steady-state creep phase is associated with a constant strain rate, which is calculated for each model. The power-law equation is employed to predict the steady-state strain rate as a function of stress, temperature, the required activation energy, and the stress exponent parameter. Finally, a deformation map is presented for different phase interface models based on the stress exponent parameter values.