Uncertainty in Dynamic Anisotropic Strength of Sand

A statistical model is proposed here to estimate the in situ anistropic cyclic shear strength of a saturated sand deposit. Triaxial test results are used to develop the basic model; it is then modified for in situ conditions considering all the factors affecting laboratory testing conditions as well as in situ conditions, i.e., mean grain size, overconsolidation ratio, multidirectional shaking, uniform cyclic loads, and secondary factors. A comparison is made between the anistropic and isotropic shear strengths. Using these strengths, the risks of liquefaction are estimated. It is observed that neglecting the presence of structures will always give a higher probability of liquefaction. The difference between the two cases is significant for shallower depths. The difference increases as the anistropic stress ratio, K_c, increases. A considerable amount of uncertainty is expected in the estimation of the in situ anistropic cyclic shear strength. The uncertainty associated with the in situ relative density contributes significantly to the overall uncertainty. The effect of the anistropic stress ratio as well as the uncertainty associated with this estimation need to be considered in any rational method for evaluating the susceptibility of a saturated sand deposit subjected to earthquake loadings.