Earthquakes on the well-established and highly slipped fault zones which host major events seem to occur at overall levels of shearing stress which are notably lower than "static friction" stress levels required to initiate slow frictional sliding between the fault walls. If those static friction stresses prevailed during earthquake slip, they would produce perceptible localized heat outflows along faults and leave abundant signs of melting and re-solidification, even at shallow crustal depths. Neither are generally found. Also, recent field and lab observations show that the majority of deformation during rapid shear is generally localized to a remarkably thin principal shear zone along the fault, often less than a millimeter to a centimeter wide, with that feature forming within a much broader, say, one to a hundred meters wide, zone of granulated and damaged rock. Our aim in the planned study is to understand the materials and thermal physics responsible for those features of fault zone response, and to establish some of their consequences for the manner by which slip-ruptures propagate along faults in major earthquakes. It is hoped that such basic understanding of the physics of earthquakes may ultimately have payoffs in the improved predictability of seismic phenomena and effects.
We have developed the concept that thermal heating of groundwater-saturated fault gouge during shear leads to strong localization of strain into realistically narrow zones. That focuses further heating and temperature rise, but rather than leading directly to melting, weakening mechanisms are triggered that sufficiently limit strength, and hence continued heating, so as to make bulk melting of the fault zone rare, at least at shallow crustal depths. A relatively universal form of weakening is that groundwater thermally expands much more than its mineral host, causing the mineral constituents to push less strongly against one another, and hence to have low frictional strength. A variant of this process is that thermal decomposition of common fault constituents such as carbonates and hydrated clays occurs, at temperatures far below melting, and creates a highly pressurized volatile product phase (CO2 or H2O, respectively) which similarly reduces strength. Further weakening processes, of which the physical details are still unclear, relate to the nanometer size range of the solid decomposition and wear products. We will model how such weakening processes influence features of propagating earthquake ruptures (e.g., crack vs. slip pulse, rupture velocity, stress drop, total slip), how rupture relates to the fault mineralogy and depth, and how different dynamic weakening processes might be identified in seismic observations. Hypotheses to be tested are that thermal decomposition combined with variation in fault mineralogy could explain how rupture stops at the base of the seismogenic zone, and that thermal decomposition could provide a mechanism for occasional extreme earthquakes on faults that generally experience smaller events. We will model the material lying outside the narrow highly-deforming fault core as an elastic or an elastic-brittle-plastic solid, and use our analyses of the localized shearing processes within the deforming fault core as the basis for imposing boundary conditions along the fault surfaces in the larger analysis. The study should contribute towards a unified overall understanding of seismic processes. It will have inputs from fine scale materials physical/chemical theory, geologic fault core studies, rock mechanics lab friction experiments, spontaneous rupture simulations, seismic observations of the slip mode and extent of seismic ruptures, and large scale constraints, by heat flow, topography support and related studies, of the stress regimes under which major earthquakes occur.
