Numerical modeling of soil liquefaction and lateral spreading using the SANISAND-Sf model in the LEAP experiments

Laboratory element and centrifuge tests from LEAP-UCD-2017 and LEAP-Asia-2019 were used for model calibration and evaluation in a dynamic coupled analysis of a saturated and gently sloped deposit of sand subjected to base excitation. The paper focuses on using a recently proposed novel constitutive ingredient for modeling the post-liquefaction large cyclic shear strains in sands. An existing critical state compatible, bounding surface plasticity reference model is used, with and without this new constitutive ingredient, to explore its improved modeling capabilities. The constitutive model was first calibrated based on available laboratory element tests on Ottawa-F65 sand, and compared to the reference model showed significantly improved performance in capturing the strain-based liquefaction strength curve of a series of undrained hollow cylinder cyclic torsional shear tests. The calibrated models were used in Class-C prediction of the slope surface deformation in five centrifuge tests on a mildly sloping liquefiable ground of the same soil subjected to dynamic loading, in the three-dimensional finite difference program FLAC(3D). The simulation results were compared with experiments in terms of acceleration history, spectral response, excess pore water pressure development, and horizontal displacement evolution at specified control points. The vectors and contours of displacements at the end of shaking also revealed a pattern of slope deformation consistent with that of a gently sloped liquefiable ground. Following the insights from the performance of the models in simulation of the slope response, the calibration was adjusted to more realistically account for the impact of initial static shear stress on the development of post-liquefaction shear strains. The models were again used for Class-C1 prediction of the slope deformation of the same centrifuge tests. The overall assessment revealed the capabilities and limitations of the models in simulating the soil liquefaction strength and its post-liquefaction response.

Automated summary

This study utilized laboratory and centrifuge tests to investigate the behavior of a sand slope subjected to base excitation through dynamic coupled analysis. By using a novel constitutive ingredient to model post-liquefaction large cyclic shear strains, the calibrated model showed improved performance compared to a reference model. The study revealed the capabilities and limitations of the models in simulating soil liquefaction strength and its post-liquefaction response.