Laboratory investigation into evaluation of sand liquefaction under transient loadings

The current laboratory procedures for evaluating liquefaction potential are still the same as 40 years ago, with minor updates. The complex seismic loading motions are simplified to a series of uniform harmonic sinusoidal loading cycles with amplitudes related to the maximum amplitude of a given ground motion; liquefaction resistance is then evaluated as the load generating liquefaction in a predefined number of harmonic loading cycles. The simplified methods of loading and resistance characterizations are a crude proxy and provide limited information in predicting the time of liquefaction triggering and therefore, the expected effects/damage of seismic events. Specific details of the time of soil liquefaction within a ground motion can be better understood from laboratory testing. Among the available element-level types of cyclic testing, cyclic simple shear (CSS) tests are the most popular and commonly used. The CSS tests provide a satisfactory simulation of seismic induced in-situ stresses. A testing program consisting of a series of multi-stage undrained direct simple shear tests was performed using the hydraulically-actuated GCTS cyclic simple shear apparatus. The apparatus had been modified and upgraded so that it is capable of applying user-specified, transient loading histories to Nevada Sand soil samples. Reconstituted specimens were prepared by the wet pluviation method at two different densities, 40% and 70%, followed with back-pressure saturation and K0 consolidation. The shearing phase was conducted in three distinctive stages: (1) Scaled transient stress histories, (2) modulated sinusoid with a taper-up shape stress histories, (3) static monotonic loads. All shearing stages were performed under continuous undrained conditions. This research program had two major motivations. The first motivation is to provide element-level tests subjected to transient loadings, so that the soil responses of excess pore pressure generation and shear strain along the time domain can be measured. The transient loading was selected from a suite of ground motions with different spectral and temporal characteristics to cover a wide range of possible ground motions. The second motivation is to investigate the performances of four Intensity Measures (IMs): CAV5, Arias Intensity, Normalized Energy Demand and PGA magnitude. These IMs were proven to be more efficient predictors of soil responses than peak acceleration. The experiments provide a database that can systematically illustrate the response of liquefiable materials subjected to transient ground motions before and after liquefaction; such a database was virtually non-existent prior to this study. Therefore, the data generated in this study supports the development of improved and more informative procedures for the evaluation of liquefaction potential, the effects of liquefaction, post-liquefaction responses, and more accurate constitutive models for liquefiable soils.