Finite element prediction of the effects of faulting below an existing building

A case history is presented here as a contribution to the panel discussion on the impediments to use advanced constitutive and numerical models in geotechnical practice. The problem concerns the evaluation of the effects induced in an existing building by a fault that could develop underneath it during a seismic event. The spreading of the fault and the consequent effects on the building are studied in static conditions through a series of elastic-plastic finite element analyses that account for the reduction of the shear strength and stiffness characteristics of the faulting zone with increasing irreversible strains. Even though the calculations require only “standard”, i.e. peak and residual, material parameters it is shown that these properties can hardly be obtained for the geotechnical medium at hand. It turns out that the analyses carried out by varying these characteristics within reasonable ranges lead to substantially different results. This example supports the observation that a non negligible impediment in using nonlinear numerical models in geotechnical practice derives from the difficulty in defining reliable values of the necessary material constants.