Abstract:Currently, the small-strain damping of soil used in seismic site response analysis is generally determined based on laboratory tests. However, since laboratory tests can only represent material damping of soil, they cannot reflect the scattering effects caused by spatial variability of the site during seismic wave propagation. Therefore, directly using the small-strain damping obtained from laboratory tests often leads to a significant overestimation of the predicted site amplification. To address this issue, this study selected four vertical seismic arrays from the United States and Europe. First,the in-situ shear wave velocity structures of these sites were extracted from the seismic data using the seismic interferometry method. Then, the in-situ small-strain damping ratio structures were inverted using the simulated annealing algorithm, and the variation of the small-strain damping ratio of the soil with depth was discussed. Finally, the performance of the uniform damping assumption and the non-uniform damping assumption in reproducing and predicting seismic site response was compared. The results showed that the in-situ soil dynamic parameters inverted using seismic interferometry and the simulated annealing algorithm could well reproduce and predict the seismic site response under small-strain conditions. Meanwhile, the in-situ small-strain damping ratio profile did not exhibit a clear variation pattern with depth. Moreover, in most cases, whether the uniform damping assumption was adopted had almost no effect on the simulated values of surface acceleration response spectra. Therefore, the uniform damping assumption could be directly adopted in seismic site response analysis under normal circumstances, without the need to consider the variation of damping ratio with depth.
