A dislocation-based model for creep recovery in ice

Creep recovery strain is significant in polycrystalline ice, and its stress dependence is strongly dependent on ice type and deformation history. Although it is generally recognized that creep recovery strain is largely attributable to dislocations, a dislocation-based model that rectifies the observed differences in two important ice types (freshwater and sea ice) has not previously emerged; and the development of such a model forms the goal of the present effort. The model considers basal dislocation distributions, employs a dislocation density–stress relationship from previous work, and uses an empirical expression for the decrease in slip-line spacing with increasing stress. The dislocation processes are taken to operate over a subgrain-sized domain, and the strain associated with the relative motion of neighbouring domains is considered. The model accounts for dislocation multiplication and the decrease in slip-line spacing if the applied stress is sufficiently high. The model explains the observed differences in creep recovery strain for freshwater and sea ice and adequately reproduces both the stress dependence and the limiting value of the experimentally observed behaviour. It also reproduces the shift from nonlinear to linear behaviour that has been observed in prestrain experiments on freshwater and sea ice cores.