LINEAR AND NONLINEAR MODELING OF ASPERITY SCALE FRICTIONAL MELTING IN BRITTLE FAULT ZONES

Study of pseudotachylytes (PT) (frictional melts) can provide information on the physical and chemical conditions at the earthquake source. This study examines the influence of asperity-scale fault dynamics on asperity temperature distribution, and therefore, the potential for frictional melting to occur. Frictional melting occurs adiabatically, and is initiated between opposing asperity tips during fault slip. Our model considers 2-D heat conduction in elastic, isotropic, hemispherical asperities, with temperature dependent thermal properties. The only heat source is a point heat flux pulse at the asperity tip. The non-linear problem was solved using the δ-form of Newton-Kantorovich procedure coupled with the δ-form of Douglas-Gunn two level finite difference scheme, while the linear problem required only the latter method. Results for quartz and feldspar indicate that peak temperatures can reach melting point values for typical asperity sizes (1-100 mm), provided that contact (frictional) shear stress is sufficiently high. For any asperity size, the temperature distribution peak becomes insignificant by the time it reaches the asperity center. These results imply that much of asperity scale melting is highly localized, which may explain why most PT veins in the field are usually very thin. However, in some cases, successive asperity encounters may generate temperature increases large enough t