Date of Graduation

8-2022

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Engineering (PhD)

Degree Level

Graduate

Department

Civil Engineering

Advisor/Mentor

Clinton M. Wood

Committee Member

Christopher L. Liner

Second Committee Member

Michelle L. Barry

Third Committee Member

Sarah V. Hernandez

Keywords

Earthquake engineering, Geological engineering, Geophysical engineering, Geophysics, Surface wave method

Abstract

This dissertation details the development of a surface wave method (SWM) technique to generate deep shear wave velocity profiles (VS profiles), applying this technique at 24 sites across the Mississippi embayment and developing an approach to generate pseudo site signature consistent VS profiles from velocity functions and fundamental frequency. In the presented SWM technique, active and passive source surface wave measurements are inverted along with fundamental frequency to develop a site signature consistent VS profile. Multiple transformation methods, including MSPAC, HRFK and FK are used to resolve experimental dispersion data from surface wave measurements. SWM VS profile at the Central US Seismic Observatory (CUSSO) site is validated by comparing with the downhole measurements at the site. Both a-priori and blind approach parameterizations are used for the surface wave inversion. While a comparable result is found between the blind and a-priori approach, the blind approach is incapable of identifying velocity reversals. Therefore, if available, a-priori information should be used to guide the parameterization process. In the absence of a-priori information, engineering judgement, experience of local geology and iterative parameterization should be used.

Deep VS profiles are developed at 24 sites across the Mississippi embayment utilizing the SWM technique. While the Central US Seismic Velocity Model (CUSVM) provides a continuous 3D Vs model of the embayment, significant differences in layer interfaces and shear wave velocities are observed between the measured and CUSVM modeled VS profiles, particularly for the Memphis sand layer. Lower shear wave velocity for the deeper layers is observed in comparison to the measured SWM VS profiles, indicating a spatial bias across the embayment. A parametric study varying the depth and velocity of the Memphis sand is conducted to observe the impact of the layer on site response. While the velocity of the Memphis sand has an impact on amplification, the depth to the Memphis sand influences both the amplification and frequency range of the amplification. Inaccurate site characterization of the Memphis sand layer could affect the long period public infrastructures with 1 – 2 second natural period, such as highway bridges and skyscrapers.

Utilizing the 24 deep VS profiles, velocity functions are developed for the geologic units found in the embayment. Pseudo site signature consistent deep VS profiles are generated utilizing a layer interface model, velocity functions and fundamental frequency. Modeled VS profiles generated using the developed approach are compared to the measured VS profiles. While differences in the near surface characterization is observed, overall modeled VS profiles have similar linear site response as the measured VS profiles, i.e., capturing the site signature.

Download

Share

COinS