HR: 0800h
AN: T21B-0475    [Abstracts]
TI: Details of Fault Loading, Slip and Recovery From a Bonded Particle Model of the Hayward Fault
AU: * Strayer, L M
EM: luther.strayer@csueastbay.edu
AF: California State University, East Bay, 25800 Carlos Bee Blvd, Hayward, CA 94542 United States
AB: We are examining the details of cyclic fault loading, failure and subsequent redistribution of stress within a 3D bonded-particle model (BPM) of the Hayward fault in the Oakland Metropolitan area. The BPM is an implementation of the distinct-element method, where rock is simulated by an array of elastic, frictional spheres that may be bonded in shear and tension. In the BPM, deformation occurs by progressive failure of bonds between particles, the resultant `micro-fractures' link up, ultimately forming discreet faults within the model. Because the BPM is fully dynamic, `earthquakes' occur in the model, and we are able to calculate their magnitudes and monitor strong ground motion anywhere in the model. By fitting a curviplanar surface to double-difference relocated hypocenters and the surface trace, we have constructed a model of the Hayward fault that ideally captures the most robust asperities upon the fault surface. It is these asperities that produce much of the model seismicity, with a good correlation between model and historical seismicity in terms of earthquake locations and magnitudes. The goal of this work is to observe the details of how earthquakes evolve in the BPM and to derive rules-of-thumb about how slip is triggered and accommodated in these types of models. Specifically we want to determine the nature and geometry of the volume of rock that is involved (based on stress magnitude, velocity, etc.), and also monitor how this volume evolves during the loading, failure and recovery cycle. The discreet, particle nature of the BPM allows us to monitor the particle stresses throughout the model and also the arrangement of `stress bridges' - networks of loaded particles. The distribution and arrangement of these stress bridges appears to be a primary control of the location of future faulting. Initial results show that slip events may involve significant rockmass on either side of the modeled fault, and over the short slip interval the size and shape of the `involved volume' can vary greatly, sometimes involving the full depth of the model (12 km).
DE: 7223 Earthquake interaction, forecasting, and prediction (1217, 1242)
DE: 7290 Computational seismology
DE: 8010 Fractures and faults
DE: 8020 Mechanics, theory, and modeling
DE: 8111 Continental tectonics: strike-slip and transform
SC: Tectonophysics [T]
MN: Fall Meeting 2005