Search

A 3D finite-element model is developed to study soil-structure interaction (SSI) effects at a Large-Scale Seismic Test (LSST) site in Lotung, Taiwan, during the earthquake of May 20, 1986. Analyses are carded out by direct method incorporating a 1/4-scale nuclear plant containment structure. The containment structure is modeled as a linearly elastic material, while the subsoil is modeled as an elastoplastic continuum material that deforms plastically according to a bounding surface plasticity theory with a vanishing elastic region. Eigenvalue analyses are performed to see how the presence of the structure affects the fundamental frequencies and modes of vibration of the system in the limit of elastic response. SSI effects are shown to be partly responsible for the reduced peak north-south ground surface acceleration recorded by a downhole array near the containment structure. Eigenvalue studies suggest that the local effect of the containment structure is to generate rocking and torsional vibration modes, in addition to the usual lateral and vertical modes. However, results of time domain studies indicate that the former modes (rocking and torsional) were not triggered by the 1986 earthquake.

A fully nonlinear finite-element (FE) model is developed to investigate the impact of hysteretic and viscous material behavior on the downhole motion recorded by an array at a large-scale seismic test site in Lotung, Taiwan, during the earthquake of May 20, 1986. A stick model with the same spatial interpolation accuracy as a three-dimensional FE model is used for vertical wave propagation analysis. The constitutive model is based on a three-dimensional bounding surface plasticity theory with a vanishing elastic region, and accounts for shear stiffness degradation right at the onset of loading. The model is cast in a time-domain nonlinear FE code SPECTRA and is used to analyze the 1986 earthquake data. It is shown that the recorded downhole motion of Lotung was dominated by nonlinear response. Results of the fully nonlinear analysis are compared with the predictions of the program SHAKE so that the performance of the nonlinear model may be assessed relative to that of an equivalent linear model.