Advanced laboratory testing for offshore pile foundations under monotonic and cyclic loading

Abstract

Laboratory experimental research is described that supported the PISA Joint Industry monopile design project and a smaller scale international investigation into the ageing behaviour of micro-piles driven in sands. The study focused on the soils encountered at the PISA test sites, including the stiff high OCR, low plasticity, glacial till sampled at Cowden (Humberside, UK), and a silica-dominated fine marine sand retrieved from Dunkirk (Northern France). Tests were also conducted on sands from two other sites at Blessington (Ireland) and Larvik (Norway). Working in conjunction with parallel research by Ushev (2018), four important aspects of the PISA test sites’ soil conditions were studied: (i) Stress-strain-stiffness behaviour from the initial linear elastic range (Y1) over the full non-linear range and up to final critical states; (ii) Stiffness and shear strength anisotropy; (iii) Response to cyclic loading; and (iv) Soil-soil/steel interface shearing. Hollow Cylinder and fully-instrumented Bishop-Wesley triaxial apparatuses were employed in the Author’s tests that revealed the Cowden till’s shear strength and stiffness anisotropy, both of which impact on the interpretation of the lateral loading behaviour observed in the large scale PISA field tests. Extensive monotonic drained triaxial tests were conducted on reconstituted Dunkirk sand to characterise its highly non-linear small-strain stiffness characteristics, considering the effects of state as well as stress history (OCR), effective stress level and evolving anisotropy. The sand’s Y1 and Y2 kinematic surfaces were depicted and tracked in effective stress space. The effects of end restraint in triaxial testing were also examined carefully, before exploring the large strain behaviour of Dunkirk sand and interpreting the results with a critical state and state parameter approach. The monotonic testing was followed by an extensive programme of drained cyclic triaxial tests on Dunkirk sand that applied up to 104 cycles to demonstrate how the samples’ state, stress history, mean cyclic stress ratio and cyclic amplitude affected the sand’s response to repetitive loading. Cyclic strain accumulation flow and dilation characteristics were found to vary significantly with all of the above factors. The detailed response could not be depicted accurately by simple cyclic dilation models that consider only the mean cyclic stress ratio. Cyclic threshold conditions were characterised in terms of two kinematic yielding surfaces that correlated with those seen under monotonic loading conditions. Additional experiments on pre-cycled samples demonstrated that the repetitive loading enhanced the sand’s monotonic shear strength and stiffness characteristics. The associated study of sand-steel interface shearing employed Bishop ring shear tests to impose large shear displacements that revealed a notable dependence of the shear resistances and dilation angles on normal stress levels and ageing times, and so provide significant insights into the ageing mechanisms that boost the axial capacities of steel piles driven in sandy soils. Information from the Author’s test programmes contributed to the pile ageing JIP test interpretation and was central to the successful modelling of the PISA field tests undertaken by other members of the Imperial College team.