Observations and challenges in simulating post-liquefaction settlements from centrifuge and shake table tests

Abstract

Liquefaction-induced reconsolidation settlements occur as excess pore pressures generated during shaking dissipate and can lead to significant damage to overlying infrastructure. Designing resilient infrastructure in areas affected by liquefaction requires methods to predict these settlements for different soil types and boundary conditions. Simplified empirical models are commonly utilized to evaluate settlements but they exhibit several limitations that might hinder the accuracy of settlement predictions, including effects of partial drainage, thin layers, non-liquefiable crusts, or soil fabric. Numerical models can capture these effects but require proper calibration and validation. This study uses a numerical approach to simulate centrifuge and shake table experiments with free-field level, ground conditions and five different sands for evaluating reconsolidation settlements. The numerical platform FLAC and constitutive relationship PM4Sand are utilized. The necessity of soil-specific calibration of post-liquefaction stiffness and an excess pore pressure ratio dependent hydraulic conductivity to accurately model the observed pore pressures and settlements is analyzed. A new relationship between the increase in hydraulic conductivity due to liquefaction and grain size diameter is proposed. The numerical simulations are able to capture the general trends, but the bias in the results appears to be correlated with the grain size distribution of the tested soil.