Electronic Theses and Dissertations

Date

2019

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Civil Engineering

Committee Chair

Shahram Pezeshk

Committee Member

Roger Meier

Committee Member

Chris Cramer

Committee Member

Charles Camp

Abstract

The local site conditions have paramount influence on the characteristics of seismic ground shaking. There are several methods of investigating seismic site effects which can be classified into two major categories; common site-specific response analysis, and procedures that use solely earthquake time series like the Standard Spectral Ratio (SSR), and Horizontal to Vertical Ratio (HVSR) methods. In the SSR method, the spectral ratios of earthquake records at a certain station can be evaluated with respect to a reference station located on a nearby outcrop rock. In the first part of the dissertation, we use earthquake records to find reference stations and use that reference station to find the site amplifications at the target stations inside the Mississippi Embayment (ME). We found WHAR, W41B, and UALR seismic stations, which have reference stations characteristics, close to the ME using horizontal-to-vertical spectrum ratio (HVSR) method. Amplification factors are obtained for six target stations inside the ME using an inversion method with WHAR seismic station as the reference station. Thirty-five local and regional earthquakes in the New Madrid and Arkansas seismic zones are used to perform the inversion. The results show the fundamental frequencies of these stations are as low as 0.2 HZ to 1 HZ, which corresponds to the soft features of their local conditions. In the second part of this dissertation, we focus on finding uncertainties in the 1-D equivalent linear site response analysis (SRA). In SRA these uncertainties are accounted for by generating random cases of soil parameters. Choosing suitable randomization bounds can decrease the effects of the uncertainties in soil parameters on SRA results like predicted spectral accelerations. These bounds quantified by the coefficient of variation (COV), which can be defined for different soil parameters. Using Vertical seismometer arrays, we can compare predicted and observed surface ground motions. In this study we evaluate COVs for different parameters of the soil that are the main input parameters of equivalent linear SRA. Coefficients of variation which generate minimum root means square errors of the observed and predicted response spectra are obtained for different site classes.

Comments

Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to ProQuest

Share

COinS