Numerical-probabilistic modeling of the liquefaction-induced free fields settlement

Abstract

Liquefaction is a phenomenon through which saturated sandy soil loses its shear strength and turns into a liquefied state. One of the most detrimental consequences of liquefaction is the reconsolidation volumetric settlements after the earthquakes, which is due to the dissipation of excess pore pressure caused by earthquakes. Severe floods can follow these settlements in free fields such as grounds close to the sea or rivers. Several researchers studied this phenomenon using data obtained from experiments in the lab or observations in the fields. Previous works were mainly based on a limited number of experimental observations and considered loadings and boundary conditions that were different from reality in the field. In actual field cases, loading does not occur in fully undrained conditions, and earthquake excitations do not comply with load application limitations on the laboratory samples. Accurate simulation of liquefaction-induced settlement phenomenon requires using fully coupled hydro-mechanical numerical analysis. In the current study, a Numerical-Probabilistic framework is proposed to estimate the liquefaction-induced settlement of the fully saturated sandy layers in the free field subjected to earthquake excitations. A fully coupled (u-P) formulation is employed to analyze soil deformations and pore water pressure variations in order to assess the soil volumetric strains and settlement. The main advantage of the current work is considering the probabilistic nature of the phenomenon and providing insight into the prediction uncertainties. Additionally, it uses a comprehensive database obtained from extensive numerical analyses. Based on the obtained settlements from the numerical analyses, a new method is proposed that predicts the post-earthquake settlement using the Bayesian linear regression method. The proposed method predicts the free field liquefaction-induced settlement in a confidence interval with high accuracy, which is validated by comparing the current work results with the observed liquefaction-induced settlements in the field case studies.

Additional details