Benchun Duan
Published: 2006
Total Pages: 468
Get eBook
Earthquake faults are geometrically complex, being segmented, bent and bifurcated. Understanding earthquake rupture processes on these fault systems is crucial to characterize source effects on resulting ground motion and to assess the possibility of rupture progressing across geometrical discontinuities to cascade into a large earthquake. However, most previous studies on this subject focus on a single earthquake with an ad hoc assumed initial stress on faults, which is one of most important components for dynamic faulting models. In this dissertation, I explore fault geometry effects on dynamic rupture processes and resulting ground motion in the context of multiple earthquake cycles. The earthquake cycle is modeled to consist of two phases: the coseismic dynamic rupture and the interseismic period. For coseismic processes, I use the finite element method to numerically simulate spontaneous rupture propagation on faults and wave propagation in the medium. I use approximate approaches to track fault stress evolution during interseismic periods. Thus, the initial stress on faults before an earthquake is a combined result of both tectonic loading and residual stresses from previous earthquakes. I examine dip-slip faults and strike-slip faults with bends, stepovers, or branches. I find that heterogeneous stresses develop on these faults over multiple earthquake cycles. These heterogeneous stresses have significant effects on the dynamic rupture process. A low normal stress developed from previous events near geometrical complexities facilitates rupture to initiate near these locations, and to jump across geometrical discontinuities. On the other hand, the high normal stress that can also develop near these locations can stop rupture. These heterogeneous stresses can allow rupture to jump larger offsets than has been previously proposed. They also allow rupture to propagate through complex paths that would be difficult to be understood in a uniform regional stress field. Fault systems with limited geometrical complexity evolve to a steady state after a number of earthquake cycles, with several typical patterns of initial stress distribution and earthquake rupture alternating in sequential earthquakes. Results from this dissertation advance our understanding of earthquake source processes on geometrically complex fault systems and may have important implications for seismic hazard analysis.