Comparison of design methods for quay walls

Abstract

The aim of this research is to investigate which design method results in the best approximation of the deformations of a quay wall consisting of an anchored combined wall with relieving platform, looking at the elastoplastic method and the finite element method. The elastoplastic method is performed with the program D-Sheet Piling and the program Plaxis is used as a finite element method. Measurements on a quay wall in Eemshaven Groningen are used to analyze which design method results in the best approximation of the real deformation.

The D-Sheet model, after some adaptations in the calibration, gives a relative good approximation of the deformations, except for the deformations at the top. This deviation can be up to 113%.The calculation with Plaxis results in a better approximation of the deformation with a deviation from the measurement of only several mm. Comparing the results of both methods shows a large difference in top deformation. Two possible explanations for this large difference are investigated. The first explanation is that the modelling limitations of D-Sheet lead to different results. A simplified Plaxis model simulating these limitations results in a top displacement of 17 mm. Secondly, the difference in the input of the soil stiffness between both models is investigated by linking both stiffness parameters E and kh to each other using the relation of Ménard. A D-Sheet model with adjusted values for kh results in only a small change in the top deformation. The remaining differences in results are attributed to the difference in calculation method.

Several variations are performed in a sensitivity analysis to investigate to which extent different input parameters would have led to different results. This includes variations in phasing, modelling of the bearing piles, soil model, surcharge load, sea water level and soil parameters.