On the inter-relationship between grain size sensitive creep and dynamic recrystallization of olivine

Abstract

The recent theoretical results of Twiss (1976) and Goetze (1978) suggest that at high temperatures and sufficiently high stresses the creep behavior of dry olivine should be dominated by either nonlinear diffusion accommodated grain-boundary sliding or nonlinear Coble creep mechanisms. This would result following the production of a fine grain-size by dynamic recrystallization. For the high-temperature experimental work performed by Karato et al. (1982) dry single crystals of olivine were almost totally recrystallized during creep, and temperature changing experiments were performed on the resulting dynamically recrystallizing polycrystalline aggregates. However, the activation energy for creep determined by Karato et al. (1982) was far higher than that predicted by the models of Twiss (1976) or Goetze (1978), although the conditions required for operation of at least the model of Twiss (1976) apparently were satisfied. The data for the highly recrystallized specimens from the higher stress, lower temperature experiments of Zeuch and Green (1979) and Zeuch (1980) are in good agreement with the results of Karato et al. (1982). These latter experiments were conducted under conditions where either the model of Twiss (1976) or Goetze (1978) should have been applicable. I tentatively conclude that although fine grain sizes were produced in the experiments of Karato et al. (1982), Zeuch and Green (1979) and Zeuch (1980) by dynamic recrystallization, there is no evidence for a transition to grain-boundary diffusional creep mechanisms at high or low stresses despite the predictions of Twiss (1976) or Goetze (1978). Instead, deformation is dominated by dislocation movement with recovery by dynamic recrystallization.