Undrained shear strength of clean sands to trigger flow liquefaction

This paper attempts to evaluate the undrained shear strength of sand during flow failures, based on both laboratory resting and field observations. In the laboratory, the minimum shear resistance during monotonic loading was taken as the undrained strength, based on the criterion of stability. Triaxial compression, triaxial extension, and simple shear test data on clean sand were examined and it was revealed that the undrained shear strength ratio could be related to the relative density of the material provided that the initial stress, p(i)', was less than 500 kPa. Three previous flow failures involving sand layers with relatively low fines contents and reliable cone penetration test (CPT) data were studied. Using existing calibration chamber test results, the Toyoura sand specimen densities in the laboratory tests were converted to equivalent values of CPT penetration resistance. The undrained sheer strengths measured in the laboratory for Toyoura sand were compared with those from the case studies. It was found that the behaviour of sand in simple shear in the laboratory was consistent with the field performance observations. Triaxial compression tests overestimated the undrained strengths, and triaxial extension tests underestimated the undrained strengths. From both the simple shear test result and the CPT field data, the threshold value of clean sand equivalent cone resistance for flow failure was detected. Based on these observations, a CPT-based guideline for evaluating the potential for flow failure of a clean sand deposit is proposed.