The axial behaviour of displacement piles in chalk

Abstract

Current axial capacity design methods for piles driven in chalk are known to be unreliable, in particular where low-medium density material is encountered. The primary complexities associated with using driven pre-formed displacement piles to withstand predominantly axial loading include (i) determining their installation resistance (ii) assessing pile static load behaviour after allowing for set-up and (iii) the effects of axial cyclic loading on the behaviour. The development of improved design methods has been impeded by a lack of both instrumented pile test data and the understanding they can bring regarding the fundamental underlying effective stress mechanisms. The work described in this thesis was motivated by an Innovate-UK joint industry project related to the Wikinger offshore windfarm in the German Baltic Sea. The Author’s investigations included analysis of full-scale offshore static and dynamic pile tests on open-ended steel tubular piles founded in glacial till and chalk at Wikinger. The tests were supplemented by further experiments on reduced-scale open-ended driven piles and heavily instrumented closed-ended Imperial College Piles at an analogous onshore test site, in Kent, UK. The onshore test site facilitated a systematic investigation of how ageing and cyclic loading affects driven piles over an eight-month period. Shaft capacity was shown to increase significantly with time and the cyclic tests indicated that one-way axial cyclic loading is not as detrimental to aged capacity as previously feared. The understanding drawn from the Author’s work was used to map out a route towards better rules for predicting the axial capacity of open and closed-ended displacement piles in chalk. A new effective stress-based approach to predict radial effective stresses is proposed, based on the key phenomena identified, namely (i) the use of CPT cone resistance to allow for local variations in properties (ii) the marked effect of the relative distance from the pile tip below any given chalk horizon (iii) the interface effective stress shear failure characteristics and (iv) addressing the significant capacity gains which occur with time. The new approach is shown to lead to far better predictions of both soil resistance to driving and long-term capacity than the current industry design methods.