Harnessing atomistic simulations to predict the rate at which dislocations overcome obstacles

Predicting the rate at which dislocations overcome obstacles is key to understanding the microscopic features that govern the plastic flow of modern alloys. In this spirit, the current manuscript examines the rate at which an edge dislocation overcomes an obstacle in aluminum. Predictions were made using different popular variants of Harmonic Transition State Theory (HTST) and compared to those of direct Molecular Dynamics (MD) simulations. The HTST predictions were found to be grossly inaccurate due to the large entropy barrier associated with the dislocation–obstacle interaction. Considering the importance of finite temperature effects, the utility of the Finite Temperature String (FTS) method was then explored. While this approach was found capable of identifying a prominent reaction tube, it was not capable of computing the free energy profile along the tube. Lastly, the utility of the Transition Interface Sampling (TIS) approach was explored, which does not need a free energy profile and is known to be less reliant on the choice of reaction coordinate. The TIS approach was found capable of accurately predicting the rate, relative to direct MD simulations. This finding was utilized to examine the temperature and load dependence of the dislocation–obstacle interaction in a simple periodic cell configuration. An attractive rate prediction approach combining TST and simple continuum models is identified, and the strain rate sensitivity of individual dislocation obstacle interactions is predicted.