On the In Situ Cyclic Resistance of Natural Sand and Silt Deposits

Abstract

This study presents cyclic resistances of an instrumented medium dense sand (i.e., the Sand Array) and medium plastic silt (i.e., the Silt Array) deposit deduced from in situ dynamic testing using the controlled blasting test method. Particle velocity records were used to calculate the cyclic resistance ratios, CRRs, and convert the transient blast-induced ground motions into their equivalent number of shear stress cycles, Neq, through consideration of the cyclic resistance observed from stress-controlled, constant–volume, cyclic direct simple shear (DSS) tests. The CRR-Neq relationship developed for the medium dense sand deposit demonstrated that the in situ cyclic resistance is larger than that (1) expected from cyclic DSS test specimens reconstituted to the in situ vertical effective stress, relative density, and shear wave velocity, Vs; and (2) calculated using case history-based, penetration-, and Vs-based deterministic formulations of liquefaction triggering models. Differences between the in situ cyclic resistance and that computed using probabilistic liquefaction triggering models reduced somewhat when considering probabilities of liquefaction exceeding 50% and 85%, depending on the model. Partial drainage during dynamic loading of the Sand Array appears to have contributed to the cyclic resistance of the sand deposit, with an increase of 6% to 27% compared to that estimated for fully undrained conditions. Differences between the cyclic failure criteria used to interpret the cyclic resistance of intact laboratory specimens of silt result in significantly different interpretations of the in situ CRR; the use of maximum excess pore pressure-consistent criteria appears to provide the best representation of the in situ, stress-based cyclic resistance when high quality, intact silt specimens form the basis for conversion of transient seismic waveforms to uniform shear stress loading cycles. The investigation described herein suggests that the reduction of cyclic resistance for plastic soils to account for multidirectional shaking ranges from 0% to 7% over Neq of 1 to 100.