Research on site response and soil-structure interaction of critical infrastructures such as nuclear power plants and transportation hubs has assumed that the underlying strata are horizontally distributed. Subsequent site response simulations have inherited this assumption, regardless of whether the analysis is conducted in the frequency or time domain. In reality, the natural status of soil layers rarely follows 1D assumptions. Hence, conventional site response analysis typically accounts for the variation of soil material for evaluating the influence of different soil layers and geological conditions. In this study, a 3D model based on an actual site with comprehensive geotechnical data was constructed to conduct 3D site response analysis (SRA). Simulations were performed in LS-DYNA using several earthquake recordings as input. The results were compared with those of conventional DEEPSOIL analysis. The results of 3D SRA are significantly lower than those obtained from DEEPSOIL analysis at corresponding borehole locations, indicating that non-horizontal subsurface conditions produce noticeable differences in surface response. Within the scope of the two verifiable events analyzed in this case study, the 1D approach was observed to predict a predominant frequency up to 104% higher than the recorded data, while peak PSa values were overestimated by as much as 115%. These results highlight the potential discrepancy of 1D analysis in complex geological settings. In contrast, the 3D approach exhibited significantly lower deviation and, in several instances, achieved a near-perfect prediction. Comparison with the actual seismic response recorded at the site's seismic monitoring stations reveals that the results from 3D SRA are more consistent with the observed data. Thus, 3D SRA conducted using LS-DYNA provides results that are closer to real-world seismic behavior. Comparison of different seismic events reveals that the surface response across the analysis domain in 3D ground response analysis varies significantly depending on the location. This highlights the advantages and unique features of 3D ground response analysis, indicating the need for further in-depth research. • A real site in Taiwan with a non-layered 3D shear wave velocity model • Comparing the nonlinear site response analysis in 1D and 3D Models. • Clarify the different between 1D and 3D site response analysis.
Yang et al. (Sat,) studied this question.