On the role of cross-slip and collinear annihilation in dynamic recovery

Abstract Dynamic recovery (DR) during single crystal stage III is believed to result from cross-slip or at least be triggered by this dislocation mechanism. However, the precise causal chain of events that induces a return to a low hardening value, similar to stage I, after the strong hardening of stage II, remains unclear. Unlike strain hardening, which has been the subject of numerous successful 3D Dislocation Dynamics simulations (3D-DD), DR has been sparsely studied. This is primarily because it requires achieving large strains to reach stage III, which is difficult to obtain in a sufficiently large volume to be statistically meaningful for dislocation modeling, especially under quasi-static loading. However, it is possible to conduct 3D-DD simulations with high dislocation density to approximate conditions close to stage III and attempt to identify the potential causes of recovery. This is what we aimed to do in the present study, with simulations at different dislocation densities and by testing specific dislocation mechanisms separately to determine their exact contributions. Our results, while not definitive, provide insight into the origin of DR. Firstly, we demonstrate that a previously neglected athermal mechanism, i.e. collinear annihilation, plays an important role in DR, being the only forest mechanism that induces virtually no storage. Secondly, we show that plastic instabilities favoring a reduction in the number of active systems are a highly effective way of reducing dislocation storage. Lastly, cross-slip is found to reduce dislocation storage under very specific slip conditions, suggesting that these conditions should exist during Stage III.