Assessment of Anchor Forces for a Drilled Pile Wall - An evaluation of the correlation between measured anchor forces, analytical calculations and numerical simulations

Abstract

A large infrastructure project in central Gothenburg, Hamnbanan, involves the construction of a new railway that will run throughout the city. The project entails tunnelling through soil as well as rock and aims to increase the capacity of freight train transport headed to and from the port of Gothenburg. The construction of tunnel in soil requires an excavation where an adequate retaining system is essential. A drilled pile wall can be used for friction material where driving a retaining wall in steel can be complicated or not possible. Combined with back-striving anchors as lateral support, a workable excavation process is enabled and is the system used for the reference project in this thesis. The aim of this thesis was to measure the anchor forces for a specific segment at the project Hamnbanan and compare these results to analytical calculations and assumptions. The calculations where carried out in SLS where the wall was modelled with both a free end supported beam and as a free end supported beam with an additional forcing moment. Furthermore, it was analysed how a drilled pile wall can be modelled in a 2D numerical program to account for the embedded support in rock. This was accomplished by correlating numerical anchor forces with site-measured values. By pre-stressing the two walls to different lock-off loads, the anchor loads could be evaluated over time. The results shows that a reduced lock-off load causes a larger variation in anchor force during excavation although no anchor reached the originally designed lock-off load. It can be seen how blasting inside the excavation increases all measured values except the ones that are lowered to a greater extent. When a free end support is assumed the analytically calculated results are significantly higher than the measured results. For a free end support with a forced moment at the support, the analytic results correlate to a greater extent and it is considered more adequate. An evaluation is made on how to model a drilled wall into rock in a numerical geotechnical program by resemble the numerical anchor forces to measured values. It is stated that a lower strength material around the embedded section of the wall foot is the method with the most coincident results. To conclude the thesis, load-bearing structures can be optimised by utilising the moment capacity of a drilled pile wall when designing a retaining system. When modelling a geotechnical problem, a numerical model can be useful, and the embedded section in the rock for a drilled pile wall must be carefully examined.