Axial capacity of piles in sand

Abstract

This dissertation presents axial compressive load tests where the pile base and shaft resistances were measured with distributed fibre optic sensors. Two test sites were established: one at Amaliahaven in the port of Rotterdam, and another in Delft. At Amaliahaven, the performance of three different pile types was compared in very dense sand: driven precast, driven cast-in-situ and a screw displacement pile type known as a screw injection pile. The screw injection piles were further investigated in the medium dense sand of Delft, comparing screw injection piles installed with a removable casing to screw injection piles installed with sacrificial casings. With these tests, the influence of installation method on the axial pile capacity was examined, and the results were considered in the context of design methods which use the cone penetration test to predict the axial pile capacity.
The piles at Amaliahaven reached very high base and shaft resistances, up to 30 MPa and 600 kPa respectively. These values are nearly three times greater than limiting resistances in design standards, suggesting that limiting resistances lead to excessive conservatism in dense to very dense silica sands. On the contrary, the screw injection piles at Amaliahaven mobilised much lower base capacities than anticipated. This became the focal point for the tests at Delft, and likewise, the piles at Delft also mobilised much lower base capacities than forecasted.
The influence of different installation methods on the pile base resistance was then examined in a review of other instrumented load tests—a review extending beyond screw injection piles and to screw displacement piles overall. The analysis confirmed that the installation of a screw displacement pile leads to little soil improvement around the pile base. In other words, a screw displacement pile tends to mobilise base resistances comparable to a soil-replacing (bored) pile rather than a soil-displacing (driven) pile.
A larger database of both instrumented and uninstrumented test records was then used to consider the implications of these findings. To do so, the base, shaft, and total capacity of each pile was compared to design methods from Belgium, France, the Netherlands and the USA. These design methods tended to overestimate the pile base contribution yet underestimate the shaft contribution, especially at cone resistances greater than the limiting resistances. A best-fit to the measured base and shaft resistances gave the best agreement on average to the measured total capacity. Nevertheless, there is room for improvement with this new formulation, particularly for the shaft resistance of screw displacement piles with an enlarged displacement body.
From the findings in this dissertation, a series of adjustments have been proposed for the Dutch pile design standard NEN 9997-1. The tests presented in the dissertation are also the first set of pile tests to be incorporated into the Dutch national pile test database, with the findings also being used to refine and optimise quay wall design across the port of Rotterdam.