Verification and validation of dynamic fault rupture simulator using consistent fault stress induced by far-field loading
Please login to view abstract download link
One of the key aspects of fault rupture simulation is the determination of initial fault stress state, which plays a major role in rupture evolution. Although methods such as projection of assumed regional stress onto the fault is a commonly used approach, such empirical methods do not necessarily produce a stress field that is compatible with the surrounding material properties, far-field loading conditions, fault frictional behaviour, and fault geometry[1]. An alternate method is to use far-field loading conditions, through which it is possible to determine a mechanically consistent initial fault stress that satisfies equilibrium and is compatible with the assumed frictional and material behaviour. Such an approach allows for rupture simulation using consistent fault stress even for complex scenarios e.g. heterogeneous non-linear material with a geometrically complex fault system. We adopted Particle-Discretisation Scheme FEM (PDS-FEM) to develop a model capable of rupture simulations under far-field loading, thereby ensuring initial conditions are consistent with mechanics. Although we reproduced a past supershear event using the developed model[2], we had yet to formally perform verification and validation. Lack of verification problems for far-field loading was a major challenge in verifying our rupture simulations. In this work, we adopted our model to verify it using the standard SCEC/USGS test suite, where our results show good agreement with the benchmarks. Validation against laboratory scale earthquakes whose full-field data were measured using high-speed cameras and Digital Image Correlation methods[3] was also conducted. We observed good agreement in the rupture front and full-field displacements and velocity, indicating that observed ruptures can be reproduced according to the far-field loading approach.
