Modeling of quay wall systems subjected to seismic liquefaction

Abstract

The purpose of this research was to identify the seismic damage mechanism of caisson type gravity quay wall and cantilever sheet pile quay wall in liquefiable site. It also investigated the effectiveness of some available mitigation techniques for both the quay wall types under dynamic loading. For these purposes, numerical simulations and shaking table model tests were carried out. In the numerical simulations, finite element (FE) code GEFDYN was used. Newmark type implicit-explicit time integration scheme was considered in the dynamic analyses. Hujeux cyclic elastoplastic multimechanism model was used as constitutive model for the soil. The numerical analyses were divided into three parts; the case study, the parametric study and the mitigation study. The case study was undertaken for both the quay wall types to understand the applicability of the FE code and the constitutive model in the dynamic analysis of quay wall sites. The size of the domain and mesh were fixed up from the case study. The results obtained from the case study were found to be well simulated the field observations and experimental results of both the types of quay wall. In the parametric study, for both types of quay walls, nine models having different input motion parameters were applied to investigate the effect of input motion on non mitigated quay wall damage. The applied input motions were cyclic, and those motions have three frequencies (1 Hz, 3 Hz and 5 Hz) and three amplitudes (0.2g, 0.4g and 0.6g) for each frequency. From this parametric study, the displacement of gravity type quay wall and maximum bending moment in sheet pile quay wall were found to increase with decreasing frequencies of input motion. The displacement of both types of quay walls was found to increase with increasing amplitude of input motion. The maximum bending moment in sheet pile wall was also found to increase with increasing amplitude of input motion. For the gravity type quay wall, the stiffness of soil at toe and relative inertia force of the wall with respect to base motion was found to be the most important influencing factor for gravity type quay wall. Whereas, dynamic earth force on upper part of the sheet pile wall was found to be the most prominent influencing factor for increasing bending moment in the sheet pile quay wall at dynamic loading. In the numerical simulations of mitigation models, nine mitigation options were analysed for gravity type and three mitigation options were analysed for cantilever sheet pile quay wall. Mainly densification of different locations of the gravity quay wall site was taken as mitigation options with an exception of a mitigation model with stone wall as an option. On the other hand, densification of backfill soil up to different extent was taken as mitigation options for cantilever sheet pile quay wall. The 1995 Kobe earthquake was applied to these models for investigating the performances of these mitigation options during earthquake. The numerical analysis of mitigation models of gravity quay walls showed that densification of foundation could mitigate the damage at strong earthquake motion but densification of backfill could not do it effectively. Densification of foundation soil at sea side up to toe of the gravity type quay wall was found to be optimum. For mitigation of cantilever sheet pile quay walls, densification of backfill up to a lateral distance of double the retaining height was found to be effective. It was found that the gravity type quay wall damages or displacements were caused by accumulated shear deformation of foundation soil. The experimental study was basically mitigation study using shake table model tests. Two series of model tests were conducted; one for gravity type and another for sheet pile quay wall models. The first series consist of a benchmark (without mitigation) model and four models with different mitigation options of gravity type quay wall, and the second series consists of a benchmark (without mitigation) model and two models with different mitigation options of sheet pile quay wall. The results of these shake table model tests show similar effects of mitigation that was predicted from the numerical techniques.