The dynamic driving and monotonic axial load-displacement behaviours of large driven open-ended piles often pose key design questions for major structures including bridges, port facilities and jacket foundations supporting offshore wind turbines. The latter structures, which are becoming increasingly common worldwide in water depths ranging from 30 to 60 m, are usually designed to carry their environmental wind, wave and current loads through a primarily axial push-pull action, as well as sustain associated lateral and pile head moment loads. Tension leg platforms, that may be adopted for deeper water sites, also rely primarily on their piles’ axial resistances.
The design of axially loaded piles for supporting such structures presents particular problems in the structured, brittle and sensitive, porous chalk, due to the lack of reliable predictions regarding piles’ driving resistances, long-term axial capacities and load-displacement response. Built on recently comprehensive research carried out in the low-to-medium density chalk, this Thesis contributes to the improving understanding of dynamic and monotonic axial pile responses. It applies a range of analytical techniques, including dynamic signal matching, machine learning, 1D load transfer and finite element techniques, to help analyse the extensive set of experimental data gathered from recent field and laboratory work at low-to-medium density chalk sites.
Noting the complexities of chalk behaviour, the Thesis first reports analyses made of the EURIPIDES Joint Industry Project’s (JIP) well-instrumented dynamic and static tests on large open-ended steel piles driven at a dense sand site. An assessment is made of how time affects pile compressive capacity through careful analyses of dynamic driving data and high-quality static compression tests conducted at different ages after installation. These preliminary analyses provide a good basis for moving to the problematic and challenging chalk, in order to add value to the recent Wind Support JIP; Innovate UK, ALPACA and ALPACA-Plus JIPs, which centred on pile behaviour in low-to-medium density chalk.
The Thesis moves from its extensive review of chalk properties to investigate the static Soil Resistance to Driving (SRD) in chalk, as mobilised during pile driving or restrikes. A modest database of high-quality dynamic force/velocity signals, pile characteristics and site investigations is presented before setting out a new Bayesian Optimisation-based Random Forest, Machine Learning, model for SRD analysis and ‘CASE-SRD’ method damping factor calibration. The Thesis’ next topic is the development of new non-linear (t-z and Q-z) Winkler-type axial load transfer model for piles driven in chalk. Noting that the ALPACA pile tests indicated a pronounced impact of displacement rate on axial loading behaviour, the Thesis shows next how maintained axial load tests can be re-interpreted within an isotach framework and suggests a load modifier approach to project how the effects of loading rate changes, or maintained load pauses, affect the piles’ nonlinear shaft stiffness and capacities in low-to-medium density chalk. Finally, the Thesis offers further insights into the mechanical response invoked around axially loaded piles through finite element (FE) simulations of representative ALPACA piles driven in low-to-medium density chalk. Alternative routes are set out for addressing the effects of chalk damage caused by pile driving and the effective stress states that apply under long-term loading conditions. The FE analyses also investigate the different ways in which shaft resistance, axial capacity and load-displacement behaviour develop in compression and tension tests.
Overall, the Thesis offers new understanding, insights and tools that should improve the ability to model the dynamic and monotonic response of piles driven in the low-to-medium density chalk, and so assist geotechnical engineers design safe and cost-effective foundations for many types of structure, including offshore wind turbines in this problematic geomaterial.