Liquefaction Evaluations of Finely Interlayered Sands, Silts and Clays

University Of California-Davis, Davis CA

Investigators

Abstract

Recent earthquakes, including the 2010-2011 Canterbury Earthquake Sequence (CES) in New Zealand (NZ), have shown that current engineering procedures have a strong tendency to over-predict liquefaction effects in thinly interlayered sand, silt, and clay deposits. This was observed in the Riccarton area in Christchurch, NZ where minimal damage occurred, but various engineering analyses predicted that at least one of the 2010-11 CES events should have caused significant ground surface damage due to liquefaction of the soils between depths of 1 and 10 m. The over-prediction of liquefaction effects in these types of deposits is a major concern in engineering practice worldwide because it leads to potentially unnecessary and expensive ground improvement or structural strengthening efforts. This research will contribute to eliminating this bias in current liquefaction evaluation procedures by addressing fundamental limitations associated with: (1) insufficient stratigraphic resolution with cone penetrometer measurements, (2) biases in the liquefaction triggering models for intermediate soil types, and (3) insufficient accounting of spatial variability in the analysis models used to estimate ground deformations. The development of improved methods for interpreting cone penetration measurements and accounting for geologic variability has broad applicability and benefit to the geotechnical profession. The correction of over-prediction biases in liquefaction procedures will contribute to more realistic risk assessments and improved land use planning for communities. The collaboration with NZ researchers will enable direct implementation of results in the Christchurch area and thereby contribute to more efficient rebuilding efforts. The student training and module development will increase the exposure of civil engineering to girls and the enhanced mentoring of our graduate student group will foster more diverse and engaged leaders in industry and academia. This research will address a major limitation in current liquefaction evaluation procedures by seeking to eliminate the strong bias toward over-predicting liquefaction effects in finely interlayered sand, silt and clay deposits. The first task will use physical modeling on a geotechnical centrifuge and axisymmetric numerical simulations of the cone penetration process to refine procedures for using cone penetration test (CPT) data to delineate stratigraphic layering and material types. The second task will update a CPT-based liquefaction triggering model using recent field data from New Zealand with improved mechanics-based functional forms for intermediate soil types. The third task will use nonlinear dynamic analyses with spatially correlated geostatistical realizations of subsurface stratigraphy to examine how one-dimensional liquefaction vulnerability indices, with their inherent assumption of horizontal layering, are biased toward over-predicting earthquake-induced ground deformations in these types of deposits. The fourth task will reevaluate the performance of Riccarton sites in the 2010-11 CES using findings from tasks one through three, and evaluate the extent to which those findings resolve the differences between the predicted and observed performances of the Riccarton area in these different earthquake events. This research will thereby integrate experimental, theoretical, and case history findings toward resolving fundamental issues affecting liquefaction evaluations, with the broader findings regarding the interpretation of cone penetration test data in finely interlayered soils and the representation of spatial variability in ground deformation models also having potentially major impacts on general practice.