Evaluation of performance of analytical and numerical methods to account for liquefaction effects on the seismic response of anchored quay walls

Liquefaction induced by earthquakes has shown to have potential devastating influence on seismic performance of anchored quay walls. Therefore, measures to mitigate liquefaction are commonly part of the design of quay walls in seismically active regions. Such mitigation measures are costly. Moreover, these measures are difficult to implement for existing structures in operation. For these reasons, proper tools that can accurately predict the effects of liquefaction on anchored quay walls are valuable for engineering purposes. Numerical tools like finite element analysis can potentially replace simplified code based methods, such as the Mononobe-Okabe method. However, performance of numerical models that account for liquefaction and pore pressure accumulation is crucial towards the use of numerical tools for this purpose. Initial stress states influence both the liquefaction resistance of the soil as well as the performance of the constitutive model. This study proposed a new calibration procedure in order to deal with the influence of static shear and overburden stress in the model. Zones around the structure with specific corresponding stress states are defined for which the stress state dependent constitutive model behaviour is calibrated based on laboratory results and literature.This study evaluates the performance of finite element calculations with the UBC3D-PLM soil constitutive model based on a reported case study of two quay walls in Akita Port, Japan for the 1983 Nihonkai Chubu earthquake. It also evaluates to what extent Mononobe-Okabe calculations with code-based corrections for liquefaction effects could reproduce the observed performance of the Akita Port quay walls. The results shown by the analysis employing the new developed calibration procedure indicate good correspondence with observations in the field. On the other hand, Mononobe-Okabe methods including corrections for liquefaction effects give a poor fit to the observed behaviour. The response indicates that dynamic analysis with the UBC3D-PLM model using the proposed calibration procedure is capable to give insight in effects of excess pore pressures on the seismic performance of an anchored quay wall. This study mainly only focussed on liquefaction triggering as a function of stress state and the post-liquefaction stress-strain behaviour predicted by UBC3D-PLM was only evaluated at a basic level.