On evaluation of shear modulus parameter (G<inf>0</inf>) in SANISAND constitutive model: Fid-Bau Portal

On evaluation of shear modulus parameter (G0) in SANISAND constitutive model

Darvishi, Reza / Jafarian, Yaser / Lashgari, Ali / Askari, Faradjollah

Abstract The SANISAND constitutive model has been widely employed for the simulation of sands cyclic behavior. The shear modulus constant (G0) in SANISAND should be theoretically estimated from maximum shear modulus of soil ( $G_{\max}$ ). To estimate $G_{0}$ , however, majority of calibration studies have used triaxial tests data in the shear strain levels larger than those associated with Gmax. This technical note aims to demonstrate how the selection criteria for $G_{0}$ parameter affects the SANISAND model predictions , particularly in terms of G-reduction (shear modulus -shear strain) curve of sands. First, experimental G-reduction curves for four sands are compared with SANISAND predictions based on the values of $G_{0}$ parameter estimated from triaxial tests. Then, a new calibration set of SANISAND for a single sand is proposed based on the $G_{0}$ value estimated from Gmax, obtained from resonant column tests. The results confirm that the numerical shear stiffness values fall significantly lower than the experimental values in small strain ranges when the $G_{0}$ constant is estimated from triaxial tests. However, the $G_{0}$ constant estimated from $G_{\max}$ appropriately predicts the experimental G-reduction curve of the sand in a wide shear strain range.

Highlights • The selection criteria for G0 parameter significantly affects SANISAND model predictions. • SANISAND calibration set is proposed for a sand based on resonant column tests. • Comparison is made for predicted G-reduction curves based on the G0 estimated from triaxial and resonant column tests. • Calibrations based on G0 parameter from triaxial tests in large strains can produce biased predictions.