Cavity nucleation at high temperatures involving pile-ups of grain boundary dislocations

Abstract

The current concept of sliding-induced cavitation is first reviewed, with due consideration to the respective role of the remote and internal stresses. The results show that the transient stresses produced by sliding play only a secondary role in cavity nucleation and that at high temperatures, the effect of sliding is eliminated in less than 1 millisecond. It is thus concluded that sliding cannot be the cause for cavity nucleation. Next, a model involving pile-ups of grain boundary dislocations (GBDs) is proposed. Unlike the sliding mechanism, the high strain energy ahead of the pile-up is a steady state phenomenon during secondary creep. It helps to compensate the large capillarity forces in the formation of sub-micron sized cavities, thereby rendering cavity nucleation barrierless. However, a threshold stress exists below which the cavities cannot grow to effect fracture. The present model suggests that cavity nucleation is feasible in single phase metals and alloys at the intersections of cell and grain boundaries. Predominant cavity formation after the onset of steady state creep and an intermediate temperature ductility trough during hot tensile tests are also features of the model. Good agreements are found between the model's predictions and the experiments.