Excessive building settlement and tilt on liquefiable soils has led to significant damage in previous earthquakes. The state-of-practice for evaluating liquefaction-induced building settlement still primarily relies on semi-empirical free-field relationships that have repeatedly been shown as unreliable and inaccurate during field and physical model studies. This is because these methods ignore the presence of the building, soil-foundation-structure interaction, and some of the dominant mechanisms of deformation near buildings. In a comprehensive numerical parametric study, the dynamic response of the soil-foundation-structure (SFS) system was assessed with a wide range of soil, structure, and ground motion characteristics. The primary objectives were: first, to identify the key predictors of foundation settlement and study their relative importance and interdependence; and second, to provide a comprehensive and mechanistically-sound dataset for the future development of a probabilistic predictive model of building settlement. The numerical simulations involved fully-coupled, 3-dimensional, nonlinear dynamic analyses of the SFS system, previously validated using centrifuge experimental results. For the conditions considered, the key predictors of building settlement were identified as the cumulative absolute velocity (CAV) of the outcropping rock motion, the relative density of, thickness of, and depth to the liquefiable layer(s), presence of a low-permeability cap, followed by foundation length-to-width ratio, embedment depth, contact area, and bearing pressure. The structure’s inertial mass and height/width ratio as well as the initial fundamental period of the structure and site were comparatively less influential. The relative importance and influence of most input parameters were shown to depend on ground motion intensity (e.g., CAV) and soil relative density.
展开▼
