HR: 08:15h
AN: S51B-02    [PDF]
TI: Low-Heat and Low-Stress Fault Operation in Earthquake Models of Statically Strong but Dynamically Weak Faults
AU: * Lapusta, N
EM: lapusta@caltech.edu
AF: Division of Geological & Planetary Sciences and Division of Engineering & Applied Science, California Institute of Technology, 1200 E. California Boulevard, Pasadena, CA 91125 United States
AU: Rice, J R
EM: rice@esag.harvard.edu
AF: Department of Earth and Planetary Sciences and Division of Engineering and Applied Sciences, Harvard University, 29 Oxford Street, Cambridge, MA 02138 United States
AB: Observations suggest that San Andreas fault generates much less frictional heat that one would predict based on laboratory static friction coefficients of 0.6 to 0.8 for most rocks and effective normal stresses comparable to overburden minus hydrostatic pore pressure. Hence one concludes that earthquakes there happen under low shear stress. Two explanations are most commonly proposed: Either (1) effective normal stress is very low everywhere on the fault or (2) static friction coefficients are very low (less than 0.2) and so the laboratory values are not appropriate for real faults. There is another possibility, that of a statically strong but dynamically weak fault with small defect regions to nucleate ruptures. On such a fault, friction coefficients at low velocities and effective normal stress are high almost everywhere, except for small defect (weak) regions. However, frictional strength deteriorates significantly at high slip velocities or slips, making the fault dynamically weak. The idea is that the fault would then operate under low shear stress (and hence with low heat generation) as follows: Model earthquakes nucleate under low shear stress in a defect (weak) region and then propagate into strong regions due to significant dynamic weakening. We study this possibility in the framework of a 2D depth-averaged elastic model of a faulted crustal plate, slowly loaded via coupling to a steadily moving substrate. We use the classical Dieterich-Ruina rate and state friction law, but modified to permit much stronger weakening at high slip rates V than the logarithmic weakening of the standard rate and state formulation. In the new law with enhanced weakening, the steady-state strength varies, essentially, as 1/(1+V/constant) at high V. Such functional dependence is suggested by theoretical studies of flash heating of contact asperities at small slips and behavior of partially DRAINED, THERMALLY PRESSURIZED fault gouge, AND PERHAPS LIQUEFIED GOUGE, at larger slips. Note that other weakening processes of dynamic nature such as sliding between different materials or UNDRAINED THERMAL PRESSURIZATION may cause the fault to become dynamically weak, but their descriptions would be different. We find that this model can operate with much smaller heat generation than the one with the classical Dieterich-Ruina law. The static value of the friction coefficient is around 0.6 in our model. Even with no weak regions, the friction coefficient inferred from average shear heating stress can be as low as 0.2. This is due to the fact that most of the slip happens at high slip velocities when the friction coefficient is low, and hence there is little heat generated. However, with no defect regions, the average shear stress before model earthquakes is high, corresponding to the friction coefficient of about 0.5. When a defect region is introduced with much lower effective normal stress than for the rest of the fault, the average shear stress before model events reduces; the lowest we observed corresponded to the friction coefficient of about 0.3. At the same time, the heat generation further decreases, corresponding to the friction coefficient of less than 0.15. These numerical simulations are very challenging and we have not yet achieved full stability. We will report on our current attempts to find a set of parameters that allow for well-resolved simulations.
DE: 7209 Earthquake dynamics and mechanics
DE: 7215 Earthquake parameters
DE: 7260 Theory and modeling
SC: Seismology [S]
MN: 2003 Fall Meeting