Numerical analysis of surface foundation subjected to strike-slip faulting: model boundaries, pre-softening volumetric response, parametric study

The paper investigates strike-slip fault rupture propagation through dense sand and its interaction with surface foundations, employing 3D finite element (FE) modelling. Two soil constitutive models (implemented in Abaqus through user subroutines) are employed for this purpose: (i) a recently developed simple yet efficient model with a Mohr-Coulomb yield criterion, which incorporates post-yield isotropic frictional hardening and softening (MC-HS model); and (ii) the basic version of the more sophisticated hypoplastic model for sand. Both models are calibrated on the basis of monotonic triaxial compression tests (and an additional oedometer test for the hypoplastic model), conducted as part of this study. The numerical predictions are comparatively assessed against centrifuge model test results. In accord with the centrifuge model tests, the free-field analyses with the MC-HS model reveal a complex fault pattern at the ground surface, consisting of Riedel (R) shears. These are the surficial manifestation of complex 3D structures of helicoidal geometry, attributed to the spatial variation of shear stresses within the overburden soil due to the imposed bedrock offset. The development of R shears is primarily controlled by the pre-softening volumetric soil response in monotonic compression and soil dilation. The latter is underestimated by the basic hypoplastic model, thus predicting the formation of a single straight shear band instead of R shears. A parametric study is conducted employing the validated MC-HS model, exploring fault rupture-soil-foundation interaction. Foundation response is shown to be sensitive to the surficial fault pattern (R shears vs. single fault trace), but the mechanism (rotational vs. translational) and foundation distress are not affected to the same extent. A two-step design strategy is outlined, requiring a free-field analysis to capture the surficial fault pattern, followed by a minimum of four interaction analyses, varying the fault-normal and fault parallel foundation location.