View article

Properly estimating strain-dependent shear stiffness of soils is necessary for accurate analysis of soil-structure interaction and seismic ground response problems during earthquake motions. In this research, an artificial neural network (ANN) model was developed for shear stiffness ratio of cohesionless soils. The input variables in this model are shear strain amplitude (γ), effective confining pressure (σ′ ₀ ), mean grain size (D ₅₀ ), and relative density (D _r ) and output is shear stiffness ratio (G/G _max ). A large experimental database was compiled from available published laboratory cyclic tests. Validation of model was carried out with using centrifuge tests results. Subsequently, sensitivity analysis and model accuracy was conducted. Finally, proposed model has been …