Earthquake hazard analysis for the city of Evansville, Indiana

Kamil Kayabali, Purdue University

Abstract

The proximity of Evansville to the Wabash Valley fault zone and the New Madrid seismic zone exposes the area to 2 potential earthquake-related hazards: soil liquefaction and site amplification. Engineering soil data were examined and various methodologies were employed to evaluate earthquake hazards. The input ground motion parameters for liquefaction and site response analyses were determined through seismic hazard analysis. The computed maximum-magnitude earthquakes are m$\sb{\rm b}$ = 6.9 and m$\sb{\rm b}$ = 7.4 for the Wabash Valley fault zone and the New Madrid seismic zone, respectively. The peak horizontal ground accelerations of 0.066g, 0.133g, and 0.178g were computed (for bedrock) for the return periods of 100, 500, and 1,000 years, respectively. Four liquefaction evaluation procedures, 2 SPT- and 2 CPT-based, developed based on Western America and Japanese data, were employed. Three of these methods indicated good agreement in the results. The fourth method based on q$\sb{\rm c}$/N$\sb{60}$ ratio was further analyzed by modifying the q$\sb{\rm c}$/N$\sb{60}$ ratios with respect to the Evansville data. About 50 to 60% of the layers studied were found liquefiable. New Madrid seismic zone had the greater influence on the liquefiability of the layers. Liquefaction Potential Index results show that 50% of the layers have very high risk of liquefaction (LPI $>$ 15). Threshold ground acceleration values for liquefaction were computed using the strain approach based on the shear wave velocity data. The results indicate that, when compared with the stress approach, the strain approach overestimates the threshold ground accelerations. Site response analysis was performed for 21 sites using the worst case and reasonable case scenario earthquakes in both seismic zones. The input time histories used were the records of the 1988 Saguenay, Canada earthquake because there was no earthquake data at bedrock for a sizable earthquake in Central North America. Results indicate that amplification ratios due to earthquakes generated in the New Madrid seismic zone are about twice as high as those caused by the earthquakes originating in the Wabash Valley fault zone. Regardless of the magnitude and the location of the input earthquake, the majority of clay sites showed higher amplification ratios than did the alluvial sand sites.

Degree

Ph.D.

Advisors

West, Purdue University.

Subject Area

Geology|Civil engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS