Award Date

8-1-2014

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Civil and Environmental Engineering

First Committee Member

Barbara Luke

Second Committee Member

Ying Tian

Third Committee Member

William Savage

Fourth Committee Member

Edward Kavazanjian

Fifth Committee Member

Wanda Taylor

Number of Pages

297

Abstract

Abstract from Manuscript 1, "An alternative analysis of the probabilistic seismic hazard for Las Vegas Valley, Nevada": Probabilistic seismic hazard calculations relevant for rock-site conditions in the Las Vegas Valley, Nevada (LVV) have been computed that account for seismic sources that are not included in the current (2008) USGS national seismic hazard model (NSHM) because of insufficient knowledge or documentation, using the commercial software package EZ-FRISK. The LVV is underlain by a system of mapped, active normal faults that comprise the Las Vegas Valley Fault System (LVVFS), with maximum potential earthquakes to M6.8. The 2008 NSHM explicitly includes only one fault of the LVVFS. This analysis includes four more faults of the LVVFS plus four regional faults in addition to those included in the 2008 NSHM, and modifies parameters of two others. As such, this study demonstrates the effect of a significant modification to expectations for seismic sources. These results can be considered as a "what if" scenario for seismic hazard of the LVV. Nominal peak ground acceleration (PGA) and 5% damped spectral acceleration (Sa) were calculated for 0.2-s and 1.0-s periods using five ground motion prediction equations (GMPEs). A logic-tree formulation accounted for uncertainty in the GMPEs and source parameters. Hazard values for PGA and Sa, computed at -2.8-km spacing, are comparable to those of the 2008 NSHM on the margins of the LVV, but are considerably higher in the north-central parts of the LVV near downtown Las Vegas and the city of North Las Vegas. The study provides a rationale for the urgency to conduct deeper investigations of faults in and around the LVV.

Abstract of Manuscript 2, "Evaluating approaches for developing design ground motions and their effects on different sediment columns: A study in Las Vegas Valley, Nevada": This research investigates the effects of different approaches to generate design earthquake ground motions (GMs; time histories and spectra), including unscaled real GMs, scaled real GMs, and spectrum-matched GMs, on earthquake site response computations; design ground motion can be defined as the ground motion specific to a site predicted at its surface, ready for use in structural design calculations. The GMs are matched to the same target spectrum (uniform hazard spectrum for a 2500-year return period) for a bridge site in Las Vegas Valley (LVV), Nevada, and one-dimensional site response analyses (equivalent-linear and non-linear) are performed. Three soil profile models are tested. The first model (profile 1) is representative of the bridge site and is deep (-400 m depth), and has a simple profile with gradually increasing shear wave velocity (VS) with depth with a VS30 of -365 m/s. The second (profile 2) is a deep and complex profile, adding a 6-m thick high-velocity layer at -30 m depth to profile 1. The third (profile 3) is a shallow and simple profile to a depth of -30 m, which treats the high velocity layer of profile 2 as the model halfspace. Because the modifications to the profiles were made below 30 m, the VS30 (shear wave velocity over the top 30 m) of all three profiles remained constant. Results of the analyses using the three approaches to generate design GMs for the three different models are comparable, although some differences are notable. The deep profiles (profiles 1 and 2) deamplified the short-period motions while amplifying long-period motions. The shallow profile (profile 3) predicts the highest amplification for all cases, mainly at shorter periods. This outcome indicates that considering only the top 30 m of the sediment can significantly over-predict the response, mainly at shorter periods. In general, amplifications are greater and differences among the three approaches to generate design GMs are greater for the equivalent-linear than for the nonlinear approach.

Abstract of Manuscript 3, "How would the Las Vegas Valley, Nevada sediments respond to strong earthquake shaking?": One-dimensional site response analysis, equivalent-linear and nonlinear, has been performed for the Las Vegas Valley (LVV), Nevada for two earthquake scenarios - a close-in earthquake on the Eglington fault and a distant earthquake on the Garlock fault. These scenarios were chosen based on deaggregation of seismic hazard from a probabilistic seismic hazard analysis for a hypothetical bedrock outcrop. Site response calculations were performed for 45 target site response grid points (TSRGPs) across the LVV at a spacing of -5 km. Sediment columns for the TSRGPs were derived from existing 3-D shear wave velocity (VS) and lithology models for the LVV. Spectral ratios at 0.01, 0.2, 0.5, and 1.0 s periods were calculated and contour maps of amplification factors were produced. Results show that there is significant variability in seismic response of the sediments across the LVV, even without considering basin reverberations and near-fault effects. The results show that seismic waves would be amplified in most portions of the LVV basin as they pass through it. Amplifications increase on average with increasing period for both scenarios. Overall, amplifications are highest in the southern and western part of the LVV. For the Eglington scenario, the patterns of ground motion are similar to the input motion, indicating that the responses of sites to this scenario are overwhelmed by the strong, near-field input motion. For the Garlock earthquake scenario, higher ground motions are observed along the western and southern margins of the Valley, which are dominated by shallow coarse-grained sediments. Amplifications are lower for places dominated by fine-grained sediments with thick sediment columns and higher for places dominated by coarse-grained sediments with relatively thin sediment columns. This result does not correlate well with the pattern of weak ground motions that have been recorded in the LVV during distant earthquakes. This mismatch implies that site response in the LVV is not only a function of sediment properties that can be modeled by 1-D analyses, it is also significantly affected by three-dimensional reverberation of energy and basin-edge effects, which can be expected to further amplify ground motions. There are a number of uncertainties in the analyses presented; perhaps the most significant pertain to the depth to the halfspace and its VS. Still, the research demonstrates that despite the mismatch and the uncertainties, sediment response plays an important role in earthquake site response across the Las Vegas Valley.

Keywords

Design; Earthquakes; Earthquake hazard analysis; Ground motion; Hazard; Nevada – Las Vegas Valley; Seismic; Site response

Disciplines

Civil Engineering | Geophysics and Seismology

Language

English


Share

COinS