Investigating the Effects of Uncertainties in Equivalent Linear Site Response Models

Date: 4/26/2019

Time: 06:00 PM

Room: Grand Ballroom

There are several examples throughout the history of earthquakes that show how soil and geological features can amplify destructive effects of an earthquake. Performing site response analysis (SRA) is a good way to capture the behavior of local conditions of the soil in case of large earthquakes. There are two sources of uncertainty in performing SRA. First, it is the variability that exists in different seismic input records and second, is the uncertainties that exist in characterizing the represented soil profile at the study site. Parameters such as shear wave velocities, layer thickness, bedrock depth, shear modulus reduction G/G_max and damping curves have significant role in a site response analysis. In this study we are focusing to evaluate the uncertainties in these parameters of the soil. We evaluate the effects of uncertainties for each parameter separately and also evaluate them together to figure out how they affect the results of SRA.

To quantify uncertainties, for this study we consider actual recorded ground motions in a series of downhole arrays at the rock and the ground surface. By comparing the observed surface ground motion with predicted results from SRA, we can compute the uncertainties. Computer program SHAKE91 is used to perform 1-D equivalent SRA and we take advantage of KIK-Net vertical seismometer arrays, which provide variety of stations with numerous ground motions of different earthquakes.

Knowing the effects of uncertainties of the soil on SRA increase the precision of equivalent linear SRA and in many cases lowers the need of time consuming procedures like nonlinear site response analysis.

Presenting Author: Arash Yarahmadi

Authors