Comparing calculation methods of deep excavations in soft clay - Comparison of analytical and numerical methods with field measurements for sheet pile wall excavations in soft clay

Abstract

Using computational tools has become standard practice in all engineering fields. In geotechnical engineering, there are a large number of calculation programs available to capture the complex conditions associated with underground construction. These advanced models can achieve efficient and optimised designs in a short amount of time. But can the complicated conditions of soil construction be too complicated to capture in advance models? This thesis will try to answer when the implementation of numerical tools is the most appropriate option. The thesis has studied three cases of sheet pile wall excavations, one of which is based on a real case with measured data from a single-strut excavation, and the other two are based on empirical values for typical Gothenburg clay. In the last two cases, one is a single-strut excavation, and the other has two levels of struts. The cases were assessed using four different calculation methods. Two of the four methods use a numerical approach, including the Mohr-Coulomb and NGI-ADP material models. Numerical results include a parametric study of certain stiffness parameters, such as the bending stiffness of the sheet pile wall, the soil stiffness, and the axial stiffness of the supporting strut. The results include active and passive earth pressures, horizontal displacements and waler reaction forces. Results show that, in general, it is very difficult to create a model that accounts for all parts of a construction process, and that an advanced model does not always yield more realistic results than a simpler model. However, complicated models are more flexible to work with and easier to implement for finding optimised designs. The parametric study showed that earth pressures are not significantly affected by changes in stiffness inputs, whereas deflections and waler forces are more sensitive. Further, the soil stiffness input has a greater influence than the stiffness inputs of the structural elements. The effect of the strut stiffness in particular could be considered negligible, and the bending stiffness of the sheet pile wall only showed a minor influence on the results.