Probabilistic Estimation of Elastic and Volumetric Cyclic Dynamic Shear Strain Thresholds for Italian Clays

The dynamic cyclic behavior of soil is remarkably complex, ranging from quasi-linear-elastic to highly non-linear and dissipative. Elastic and volumetric shear strain thresholds define relevant boundaries between different dynamic cyclic soil behavior domains. The estimation of threshold strains is an important step in the dynamic characterization of soil behavior for engineering applications such as ground response analysis and geotechnical design in the dynamic domain. The estimation of threshold strains relies on the results of cyclic dynamic laboratory testing on undisturbed soil samples. Past research has shown the existence of a direct correlation between dynamic behavior strain thresholds and soil index properties. This paper describes the probabilistic estimation of cyclic dynamic shear strain thresholds for a wide ensemble of Italian clays. Analytical relationships between plasticity index and strain thresholds are developed using quantile regression relying on an extensive database of resonant column and cyclic torsional shear testing results. These formulations refer to specific probabilities of exceedance, providing a quantitative measure of uncertainty which can be accounted for in best-practice dynamic geotechnical analysis.