Earthquakes represent a significant natural hazard in many inhabited areas of the world. Yet, despite much study, the conditions that control the stability and nature of the slip on faults are still poorly understood. An important control on fault-stability arises from the fact that faults zones are often filled with granulated rock and the interstices of these grains are filled with groundwater. Variations in the pressure of this fluid have been observed to trigger movement on faults. The proposed project will result in a better understanding of the basic physics behind slip on faults, in particular the coupling between solid stresses and fluid pressures in the fault zone. The insight gained from this work will contribute to the efforts to assess the seismic hazard associated with individual faults, and the risk of triggering earthquakes by natural or man-made changes in local groundwater levels. This work should also add to the understanding of saturated granular flows, such as landslides and offshore turbidity flows (both of which constitute significant natural hazards to lives and structures).
The saturated gouge-fault block system includes several coupled mechanisms that may control fault stability and determine the conditions under which faults creep, slip slowly (slow earthquakes), or accelerate into earthquakes. Increases in fluid pressure reduce the effective stress across a fault, which weakens the fault and promotes sliding, while the dilation that accompanies the onset of slip will lead to pressure reductions that can strengthen the fault. Fluid flow into and out of the fault will mute pressure fluctuations, but is highly dependent on permeability in both the granular fault gouge and in the confining wall rock, which in turn can vary greatly with gouge dilation and stress-induced damage. We will study this system using a grain-scale numerical model. This model couples together the discrete element method (DEM) for granular dynamics with a continuum finite-difference solution for fluid flow through permeable media. This grain-scale approach allows for heterogeneity and localization within the fault gouge, and accounts for effects of the pressure on the dilation of the fault. We will simulate sections of a gouge-filled fault zone confined by breakable wall rock (simulated with bonded cohesive granular material). The modeled fault will be subjected to various types of loading, including slow increases in tectonic shear stress and the transient passage of seismic waves and fluid pressure waves. The goal will be to map out the conditions for fault stability, and quantify the relationships between fault stresngth and the evolving permeability and fluid pressure in both the fault gouge and damaged wall rock.