Nonlinear analysis for liquefaction simulation: implication of constitutive soil model

The nonlinear analysis for assessing the cyclic behavior of a specific location is widely applied to minimize geohazards, especially soil liquefaction. The Mandalika circuit project in the Lombok area is highly prone to earthquakes and is located on sand as the subgrade. The high soil liquefaction susceptibility leads to a comprehensive assessment by implementing the appropriate soil constitutive model. One-dimensional nonlinear analysis with two constitutive models, Modified Kondner–Zelasko (MKZ) and General Quadratic/Hyperbolic (GQ/H), provided by the DEEPSOIL program, is examined to generate the dynamic behavior aimed at by this research. Five boreholes from SPT data are conducted as part of the soil investigation data. Based on the pore water pressure, the MKZ and GQ/H in the maximum value (full liquefaction condition) are almost similar in output. However, in other situations, they show the opposite result. The GQ/H predicts a more realistic simulation in the low and medium liquefaction cases, which presents the most significant correlation between strain ratio and shear stress results in all sites. For an accurate evaluation of liquefaction behavior, the combination of dynamic features from an appropriate constitutive model should be considered to simulate the liquefaction behavior of sand.