Influence of "long-term time" effects on soil stiffness in local seismic evaluation

Dynamic shear modulus determined at low shearing strain amplitude by field testing methods, G0,field, is generally higher than the value determined by laboratory testing methods on “undisturbed” soil specimens, G0,lab. Moreover the shape of the modulus ratio, G/G0, versus shearing strain curve, γ, in field differs from that determined by laboratory testing. The differences derive from many causes, including specimen disturbance, incorrect laboratory representation of grain-size distribution and field confinement, also for the same average confining pressure, and long-term time effects. Long-term time effects, both on the G0 value and on the shape of the G/G0 vs. γ curve can be evaluated by the Anderson and Stokoe II (1978) method. The method can be applied to evaluate the local seismic response of sites for which the results of dynamic laboratory tests from “undisturbed” soil specimens and a reliable geologic dating of the deposit are available, but VS profiles have not been determined in situ. The aim of the research is to check to what extent the difference between the values of the initial stiffness from laboratory and in situ testing can be predicted by applying the Anderson and Stokoe II method and analyse the influence of correction of the shear modulus curve on the results of local seismic response numerical analysis, depending on the shear strain and therefore on the design input motion. The study was performed on a site of Northern Italy where dynamic laboratory and in situ tests were performed to characterise soils and geologic age was known. The adopted methodology and the results are presented and discussed in this paper.