Numerical simulation of spontaneous rupture processes on two non-coplanar faults: the effect of geometry on fault interaction

Analyses of earthquake sources have revealed that the earthquake rupture process is complex and that the rupture does not occur on a single plane. Earthquake faults are often composed of several subfaults, and rupture propagation tends to decelerate or terminate at places where the fault strike changes. These observations imply that fault geometry, including fault steps and fault strike change, plays an important role in earthquake rupture complexity. In this paper, we calculate the spontaneous rupture processes of two non-coplanar faults in 2-D in-plane problems, attempting to clarify the effect of fault geometry. We consider two simple models-models in which two faults are either parallel or perpendicular to each other. We calculate spontaneous rupture propagation on the faults by a finite difference method, and we then compare the results. In our simulations, rupture initially grows on the main fault, and stress perturbation from the main rupture then triggers rupture on the secondary fault. Propagation of the main-fault rupture controls a spatio-temporal pattern of stress difference in the uniform elastic medium, which determines the rupture process of the secondary fault. The rupture propagation and termination of the secondary fault are significantly different between the two models. The difference is obvious when rupture of the main fault is arrested and the secondary fault is located near the arrested end of the main fault. When the secondary fault is parallel to the main fault, rupture can propagate ahead on the secondary fault. However, when the secondary fault is perpendicular to the main fault, rupture is either not triggered on the secondary fault, or soon terminates if triggered. This variation of the rupture process implies that: fault interaction, depending on geometry, can explain the termination and change of rupture at places where the fault strike varies. This shows :the importance of the fault geometry in studying spontaneous dynamic rupture processes.