Abstract

This study presents the numerical results of a series of laboratory and dynamic centrifuge tests conducted by the team at Universidad del Norte, as part of the LEAP-2022 project. The soil's mechanical behavior was simulated using a pressure-dependent multi-surface plasticity constitutive model, which was carefully calibrated based on cyclic soil tests performed on Ottawa F-65 sand. These tests covered a wide range of initial densities, initial effective stresses, and cyclic stress ratios. The comparison between laboratory and numerical element tests revealed that the adopted constitutive model reasonably replicated most features of the material's undrained cyclic response, including liquefaction occurrence and the progressive development of double-amplitude permanent shear strains. The calibrated constitutive model was then used to blindly predict the dynamic behavior of centrifuge experiments composed of a sheet pile-soil system using the OpenSees finite elements software framework; these simulations are referred to as Type-B predictions. The numerical simulation showed that the model provided reasonable representation of soil responses in terms of accelerations and pore water pressure build-up; however, the simulations consistently overpredicted the displacements of the sheet piles. Therefore, based on the centrifuge experimental results, minor adjustments of the material parameters were performed, and the centrifuge tests were re-simulated; these simulations are referred to as Type-C predictions. The comprehensive evaluation exposed both the strengths and weaknesses of the modeling approach for the simulation of liquefiable deposits. Despite the discrepancies in sheet pile displacement, the study instills confidence in the model's applicability to liquefaction-related projects with similar conditions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call