Abstract. Experiments by Zhang and Karato [1995] have shown that in simple shear dislocation creep of olivine at low strains, an asymmetric texture develops with a [100] maximum rotated away from the shear direction against the sense of shear. At large strain where recrystallization is pervasive, the texture pattern is symmetrical, and [100] is parallel to the shear direction. The deformation texture can be adequately modeled with a viscoplastic self-consistent polycrystal plasticity theory. This model can be expanded to include recrystallization, treating the process as a balance of boundary migration (growth of relatively underformed grains at the expense of highly deformed grains) and nucleation (strain-free nuclei replacing highly deformed grains). If nucleation dominates over growth, the model predicts a change from the asymmetric to the symmetric texture as recrystallization proceeds and stabilization in the "easy slip" orientation for the dominant (010)[100] slip system. This result is in accordance with the experiments and suggests that the most highly deformed orientation components dominate the recrystallization texture. The empirical model will be useful to simulate more adequately the development of anisotropy in the mantle where olivine is largely recrystallized.
IntroductionSimple shear experiments of rock-forming minerals and analogs deformed by dislocation creep reveal some puzzling features. After moderate strains, the deformed original grains develop a characteristically asymmetric preferred orientation pattern with a monoclinic symmetry, including a mirror plane perpendicular to the shear plane and containing the shear direction. These experimental simple shear deformation textures (in this paper the term texture is used synonymous with lattice preferred orientation), such as those for quartz [ These orthorhombic textures cannot be predicted with polycrystal plasticity theory for deformation by dislocation glide such as Taylor, Sachs, or self-consistent models. Authors have referred to the textures as "easy slip" orientations [e.g., Schmid and Casey, 1986], implying that a microscopic slip plane of an active slip system is parallel to the macroscopic shear plane and a microscopic slip direction is parallel to the shear direction, and therefore deformation on such a slip system in simple shear is "easy." So far, there has been no explanation to how crystals rotate and remain within those orientations. Herwegh and Handy [1996] suggest a "rigid body rotation" into the easy slip orientation and an unexplained cessation of rotation once those special orientations are reached.Experimental