Ricardo Taborda

Thanks to the continuous growth of computational capabilities and advances in modeling methodologies, seismologists and engineers are now able to simulate strong ground motion at a scale and frequency-level never envisioned before. Most of these simulations, however, continue to be done considering linear soil conditions. Although nonlinear soil behavior was largely debated for decades, its occurrence during moderate to large earthquakes has now been widely acknowledged, especially due to the evidence gathered after events like the Loma Prieta, 1989 and Northridge, 1994 earthquakes. Despite direct observations from earthquakes and in the laboratory, three-dimensional (3D) effects of soil nonlinearity on the spatial and temporal distribution of ground shaking are still far from being well-understood. In fact, its effects are still usually omitted or only treated approximately as part of ground shaking forecasting methods. Under this panorama, the main objective of our research was to develop the capability for modeling ground motion in basins due to moderate to strong earthquakes, taking nonlinear soil behavior into consideration, and to apply and validate this capability to realistic situations. Our plan was to first gain physical insight through simplified problems under idealized conditions regarding both nonlinear soil behavior and 3D basin shapes. This report summarizes the approach and first results of our work toward these objectives. Although we initially envisioned this being done in a hybrid two-step manner, combining our own software for large-scale anelastic earthquake simulations with other commercial software capable of modeling plasticity, we later realized that this approach, despite its advantages, had computational constraints in its efficiency and scalability for our own long-term objectives. This led us to reset the course of action and tackle the problem in a direct manner using and enhancing our own software. We have successfully implemented the necessary tools to represent nonlinear soil behavior by incorporating two basic material models for plasticity and have conducted experiments to test our implementation. In this report we describe our methodology and present results for the simulation of a realistic basin with nonlinear soil conditions subjected to two moderate earthquakes, one immediately below and the other far from the basin. We compare the elastic and inelastic response of the basin and discuss the di®erent contrasting results obtained from considering its nonlinear response.