THE IMPACT OF TRANSIENT BRITTLE FAILURE ON DEFORMATION AND METAMORPHISM AT MIDDLE AND LOWER CRUSTAL LEVELS

Brittle deformation of metamorphic rocks is significant not only because high strain rate events accommodate relatively large displacements within narrow zones, but also because they modify the rock's microstructure, affecting subsequent deformation and metamorphism. Specifically, cataclasis results in grain-size reduction (increase in surface area / volume ratio of grains) and grain-scale mixing of phases that are commonly segregated into layers prior to brittle failure. Cataclasite is therefore both finer grained and more homogeneous than its protolith, facilititating development of an equilibrium mineral assemblage and changing the mechanical behavior of the rock.

Deformation of cataclasite and adjacent compositionally banded rock will proceed in the most mechanically and thermodynamically efficient way, depending on rock microstructure. The random distribution of phases in a cataclasite will help maintain a fine grain size during subsequent ductile flow by inhibiting grain growth. Because diffusivity is higher between unlike grains than between like grains, diffusivity will be enhanced in cataclasite relative to protolith (the grain-boundary area between unlike phases is greater in cataclasite). Both fine grain size and enhanced diffusivity favor grain-size-sensitive deformation mechanisms such as diffusion-accommodated grain-boundary sliding, which can accomplish superplastic flow. Thus, zones of cataclasite are likely to continue to accumulate higher strains than surrounding materials subsequent to brittle failure.

In contrast, phases are typically segregated into foliation-parallel mineralogical domains in polyphase mylonites. Diffusivity will be greatest at the boundaries between these compositionally and mechanically distinct layers or domains. Deformation mechanisms that are dependent on diffusion, such as diffusion-accommodated grain-boundary sliding, will thus be suppressed within mineralogical domains, whereas sliding between domains (cooperative grain-boundary sliding) may be facilitated. Evidence for cooperative grain-boundary sliding includes ridge-in-groove slickenlines and sharp boundaries between C-surfaces and surrounding domains.