11:30 AM - 11:45 AM
[SSS10-14] Effects of bimaterial interface on rupture along strike-slip branch faults
Effects of bimaterial interface on rupture along strike-slip branch faults
Keywords:Earthquake, Branch Fault, Dynamic Rupture, Finite Element, Numerical Model, Seismic Hazard
Complex crustal faults can sometimes separate material with different elastic properties. Dissimilar media around faults, also referred to as bimaterial interface, has been shown to cause effects on the rupture process along vertical strike-slip faults. Some previous studies suggest that asymmetry in wave propagation across a bimaterial interface can introduce normal stress changes on the fault near the rupture front that can lead to asymmetric bilateral or unilateral propagation. Furthermore, a bimaterial interface can also lead to differences in strain release across a fault interface for a fixed stress drop. Considering the effects caused by bimaterial interface on rupture propagation, it is worth understanding whether these effects can impact through-going rupture across a geometric complexity such as a branch fault. In this work, we use 3D dynamic rupture simulations to investigate the effects of material property contrasts on rupture propagation along branch faults and the possible implications for seismic hazard. We generate a suite of finite element meshes with tetrahedral elements for 2 different planar fault geometries with: one fault geometry is a Y branch model which consists of a planar fault branching into two segments separated by an angle of 30° and the other consists of a planar fault intersected by a secondary fault at an angle of 50°. For the finite element models, we vary the location and magnitude of the material contrast (0-20% difference in velocity across the interface). We also vary the magnitude and sense of initial shear stress on the models. We force nucleation by increasing the shear stress in a circle patch of nodes and the frictional breakdown is governed by a linear slip weakening friction law.
Results indicate that the location and magnitude of the material contrast, as well as the initial shear stress conditions can affect the path of self-determination of rupture. Although there are seismically induced normal stress perturbations caused by the bimaterial interface, these do not appear to be the dominant factor in determining the path of rupture propagation. Rather we believe the differences arise from strain asymmetries caused by the varying shear modulus. The higher shear modulus in the material causes a reduction in slip for the same stress drop and also leads to a difference in the resolved shear and normal stress changes on the branch segments. Furthermore, in models where we vary the velocity contrast but hold the shear modulus constant, we find that the normal stress perturbations are still present, but we do not see a difference in the self-determined rupture path when compared to a homogeneous model.
When rupture is nucleated on the main fault for the 50° branch geometry it is less likely to rupture the secondary segment as the material contrast increases if the main fault and secondary faults have the same sense of initial shear stress. If the faults have opposite senses of shear stress, we find that a larger material contrast promotes rupture propagation on the secondary fault. For the 30° Y branch geometry, we see the rupture propagate onto both the extensional and compressional branch if the extensional branch is surrounded by stiffer material and the compressional branch is surrounded by softer material. This could have implications for the assessment of possible rupture scenarios in regions with varying material properties and helps to emphasize the importance of near fault rheology and seismic velocity studies.
Results indicate that the location and magnitude of the material contrast, as well as the initial shear stress conditions can affect the path of self-determination of rupture. Although there are seismically induced normal stress perturbations caused by the bimaterial interface, these do not appear to be the dominant factor in determining the path of rupture propagation. Rather we believe the differences arise from strain asymmetries caused by the varying shear modulus. The higher shear modulus in the material causes a reduction in slip for the same stress drop and also leads to a difference in the resolved shear and normal stress changes on the branch segments. Furthermore, in models where we vary the velocity contrast but hold the shear modulus constant, we find that the normal stress perturbations are still present, but we do not see a difference in the self-determined rupture path when compared to a homogeneous model.
When rupture is nucleated on the main fault for the 50° branch geometry it is less likely to rupture the secondary segment as the material contrast increases if the main fault and secondary faults have the same sense of initial shear stress. If the faults have opposite senses of shear stress, we find that a larger material contrast promotes rupture propagation on the secondary fault. For the 30° Y branch geometry, we see the rupture propagate onto both the extensional and compressional branch if the extensional branch is surrounded by stiffer material and the compressional branch is surrounded by softer material. This could have implications for the assessment of possible rupture scenarios in regions with varying material properties and helps to emphasize the importance of near fault rheology and seismic velocity studies.