Evaluation of liquefaction potential by energy-based and stress-based methods and gene expressing programming (case study: Tabriz city)

Abstract

ABSTRACT Liquefaction in soil layers is a vital factor intensifying earthquake damages.This study compares the numerical evaluation process and the results of two methods based on stress and energy. It is inferred from the calculations that the stress-based method predicts a higher liquefaction potential with a lower safety factor as it promises liquefaction in deeper soil layers. In return, liquefaction tends to occur at a shallower depth with higher intensity in the energy-based method. Through applying the two approaches based on the data collected from different areas around Tabriz, a liquefaction-zoning map is presented. Despite being far from the fault, the central to the western and the southwestern parts of Tabriz has a high liquefaction potential. Eventually, based on the evaluated liquefaction potential using the stated methods and adopting the gene expression programming (GEP) approach, an equation is introduced to estimate the liquefaction potential for the case study. The predictions of the proposed models were consistent with the findings of experimental methods, demonstrating appropriate statistical measures and parametric analysis. It can also be concluded from the results of the parametric analysis, that the parameters of maximum acceleration, earthquake magnitude, and SPT number have more great impact on soil liquefaction.