Effects of System Compliance on Liquefaction Tests

A theoretical analysis of the errors arising from volumetric compliance in cyclic liquefaction tests on saturated sands is presented. For cyclic simple shear tests, compliance occurs with changes in membrane thickness, membrane stretch in corners, and expansion of the confining frame as pore-water pressures increase. For cyclic triaxial tests, compliance arises from reductions in membrane penetration. Compliance is shown to cause significant increases in stress ratios causing initial liquefaction. For triaxial tests, computations indicate that measured increases in stress ratios causing initial liquefaction with increase in grain size are primarily attributable to the effects of membrane penetration. The compliance effect arising from partial saturation is also examined, and significant increases in stress ratios causing initial liquefaction with only small reductions from full saturation are shown to occur. The difficulties of eliminating the effects of system compliance from undrained liquefaction tests suggest the use of constant volume simple shear liquefaction tests where drained samples are tested under constant volume conditions.