Date of Award

1-2011

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Civil Engineering

Advisor

Andrus, Ronald D

Committee Member

Juang , C. Hsein

Committee Member

Moysey , Stephen

Committee Member

Ravichandran , Nadarajah

Abstract

Liquefaction potential of major Pleistocene deposits in the Greater Charleston area, South Carolina is investigated in this dissertation. The data considered to characterize liquefaction potential include field performance information the 1886 Charleston earthquake and the results of many seismic cone penetration tests with pore water measurements (SCPTu). The investigation begins with the Mount Pleasant area, and then expands to the entire Greater Charleston area.
A liquefaction potential map of Mount Pleasant is created through reviewing available first-hand accounts of ground behavior during the 1886 earthquake, analyzing cone penetration test and shear wave velocity data, and correlating the results with geology. Careful review of the first-hand accounts reveals that nearly all cases of surface effects of liquefaction can be associated with the younger sand deposits that lie adjacent to the harbor, rivers, and creeks. Only one documented case of minimal surface effect of liquefaction can be definitely associated with the older sand deposits of the 100,000-year-old Wando Formation. Ratios of measured to estimated shear wave velocity (MEVR) indicate that the younger sand deposits and the older sand deposits have measured velocities that are 9% and 38%, respectively, greater than 6-year-old sand deposits with the same cone tip resistances. Liquefaction potential is expressed in terms of the liquefaction potential index (LPI) proposed by Iwasaki and others. LPI values for the older sands computed from the SCPTu profiles are incorrectly high, if no age corrections are applied. If age corrections are applied, computed LPI values match well the observed field behavior in both the younger sands and the older sands. The results are combined with a 1:24,000 scale geologic map to produce a liquefaction potential map of Mount Pleasant. The findings of the Mount Pleasant study agree remarkably well with a previous liquefaction potential study of aged soil deposits on Charleston peninsula.
Liquefaction potential of Pleistocene sand deposits in the Greater Charleston area is characterized by reviewing cases of conspicuous craterlets and horizontal ground displacement that occurred during the 1886 earthquake, and analyzing eighty-two seismic cone soundings. Nearly half of the cases of ground failure in sand deposits are associated with the 200,000-year-old Ten Mile Hill beds located within 13 km of the Woodstock fault, the likely source of the earthquake. One quarter of the cases of ground failure are associated with the 100,000-year-old Wando Formation located within 17 km of the fault; and another quarter are associated with the younger deposits that lie adjacent to the harbor, rivers, and creeks located within 31 km of the fault. The influence of distance to the fault on LPI and MEVR is investigated. Computed LPIs are corrected for the influence of diagenetic processes using MEVR. The liquefaction probability curves developed for four major sand groups agree well with the 1886 field observations.
The influence of depth to top of the Cooper Marl and depth to the groundwater table on LPI values of the younger sand facies of Wando Formation (Qws) is also investigated. Liquefaction probability curves are developed considering the influence of depth to the groundwater table and depth to the non-liquefiable Cooper Marl.
Liquefaction potential of areas now covered by artificial fill (af) in the Charleston area are characterized through reviewing cases of conspicuous craterlets and horizontal ground displacement that occurred during the 1886 earthquake, and analyzing twenty-three seismic cone soundings. All cases of 1886 ground failure that plot in af areas on Charleston Peninsula and around Mount Pleasant are located where Qhes or younger sand deposits are believed to be in the subsurface. SCPTu sites mapped in af are grouped into three categories based on dominant geology in the top 10 m. Liquefaction probability curves are developed for the three categories considering the influence of depth to the groundwater table and depth to the non-liquefiable Cooper Marl.
The liquefaction potential of areas covered by surficial clayey deposits in the Greater Charleston area are characterized through reviewing liquefaction and ground failure cases that plot in these areas and analyzing thirty-two seismic cone soundings. Liquefaction probability curves developed for four major clay groups are compared with the liquefaction cases that plot in the surficial clayey deposits. The liquefaction probability curves developed for the surficial clayey deposits do not agree well with the high number of ground failures that occurred in these deposits during the 1886 earthquake. Conservative liquefaction probability curves are suggested for the surficial clayey deposits.
Laboratory tests conducted on samples collected from various Pleistocene deposits indicate little or no carbonate in the beach sand deposits in the Greater Charleston area. Thus, the higher shear wave velocity and MEVR values associated with Qws are not the result of carbonate cementation.
The probability curves can be used to develop geology-based liquefaction hazard maps of the Charleston area. Liquefaction hazard maps are useful tools for identifying areas with high likelihood of liquefaction-induced ground deformation, a major cause of damage in many earthquakes. Information about areas with high likelihood of ground deformation can be used for effective regional earthquake hazard planning and mitigation. Liquefaction hazard maps are also useful for identifying areas where specific investigations for liquefaction hazard are needed or should be required prior to project development, but in general these maps should not be used for site-specific engineering design.

Share

COinS