Dynamic characterization of sand under low confinement stress via shear box testing on shaking table

An experimental study is presented herein, using an equivalent shear beam box placed on a shaking table to quantify the dynamic behavior of sand under low confinement stress levels, commonly encountered under 1-g testing conditions. The obtained experimental results are systematically compared with the predictions of well-established analytical models, which capture the shear modulus increase with relative density and mean effective stress while also accounting for shear strength degradation with increasing cyclic shear strain amplitude. A novel recalibration methodology is subsequently introduced to refine these models by evaluating the dynamic response of the entire soil column rather than a single soil element, which would be difficult to test under low-stress conditions. To this end, shaking table tests are conducted across a range of sand densities and shaking intensities, and transfer functions are computed at various depths. The analysis reveals that, under low confinement stresses, empirical models tend to overestimate small-strain sand stiffness, while also overestimating shear-strain-induced modulus degradation and thus underestimating large-strain stiffness. The recalibrated approach suggests a less nonlinear response, more accurately capturing the sand’s dynamic response across all strain levels.