Seismic Hazard Curve Considering Nonlinear Site Effect

Seismic hazard analysis of the site addresses the estimation of the size of future ground motion (or intensity), the probabilities of different ground motion (or intensity), or the probability to exceed a given level of ground motion (or intensity). Probabilistic seismic hazard analysis (PSHA) is the most common method of application, the final results of the analysis include the annual probability of exceedance of ground motion parameters, namely the seismic hazard curve, and the intensities or peak accelerations of bedrock and response spectra at different exceedance probabilities, and the results provide input parameters for aseismic analysis of engineering design.The characteristics of future earthquakes, propagation path and local site condition are the three dominant factors that affect the size of the role of ground motion. For the soil site, effects of all three factors should be considered. In the field of geotechnical engineering, the equivalent linearization is generally used to evaluate the effect of the site condition on the soil response, and the amplification of a given bedrock motion due to the soil structure is studied with the transfer function.In this thesis, the effect of site condition including the uncertainty of the nonlinear effect is incorporated into the framework of PSHA, by a convolution in total-probability, and the hazard curve at the surface of the soil site is calculated by this method.Firstly, on the basis of summarizing the development of the method of PSHA, the hazard curve at the bedrock of a given site is calculated, and the acceleration time history at the site is synthesized. Then, the amplification of the soil structure is obtained using the equivalent linearization method and the concept of the transfer function. Finally, the surface seismic hazard curves are calculated on the basis of rock seismic hazard curve respectively, using the method of statistical hypothesis, transfer function method and the convolution method considering nonlinear effects, and the results are compared. The comparison shows that there are obvious differences existing in the surface seismic hazard curves using different methods; the seismic hazard is overestimated by statistical hypothesis, and is a little underestimated by the transfer function method, while the seismic hazard is the lowest by the convolution method.