Behaviour and efficiency of perimeter pile groups

Groups of piles are commonly used as high capacity foundations. It is recognised that the load distribution among piles in a group will vary and it is thought that the inner piles are likely to make a relatively small contribution to the total load carried. The essence of the research undertaken is to establish the relative effectiveness of pile groups with either no inner piles (perimeter group) or a single central pile (target group) when compared to the more commonly used grid group arrangement. Pile groups in which the central piles were omitted were used for the Cannon Place redevelopment in London and provided the impetus for the research project. The main research technique used is geotechnical centrifuge modelling. Samples of overconsolidated kaolin clay were prepared and tested on the centrifuge at City University London. This provided a firm clay into which pile groups could be installed in a wide variety of arrangements. Three or four different pile groups were located in each centrifuge model and loaded to failure using a strain-rate controlled load actuator. The individual model piles were made of 5 mm diameter aluminium rod placed in holes pre-drilled in the consolidated kaolin prior to the centrifuge test. All piles extended to a depth 250 mm in the clay giving an lid ratio of 50. The ranges of pile groups tested are linear, circular and square perimeter, circular and square target . and square grid. Single pile tests provided the reference pile capacities used to normalise the data from the 23 centrifuge models tested. The experimental work was complemented by a parametric numerical modelling study using the finite element programme Abaqus. This gave insight into the pile-soil interaction and permitted a more meticulous analysis of the soil stresses and displacements. In addition, the numerical modelling enabled extension to the original variables tested as part of the centrifuge experiments and the soil shear strength and lid ratio were varied. The pile groups failed in one of two ways: either as individual piles with the piles settling into the ground with no noticeable settlement of the soil surrounding a pile, or as a block with the soil contained within the outer ring of piles settling by the same or similar amount as the piles. The change from block failure to individual pile failure often occurred at a pile centre-to-centre spacing of about two pile diameters though variables such as number of piles, the presence of a target pile and the strength of the soil all had an effect. The efficiency of a pile group is defined as the load capacity of a pile group expressed as a ratio of the number of the piles in the group multiplied by the load capacity of a single isolated pile. It was demonstrated that a grid group arrangement was the least efficient of the groups tested, whereas a perimeter group arrangement could achieve higher efficiencies of greater than unity and the inclusion of a target pile could further enhance the group efficiency. It has been shown that a target group comprised of 17 piles (16 piles plus one central pile) has a significantly higher efficiency than a 5x5 grid group comprised of 25 piles, such that the capacity at lower settlements is the same for both groups.EThOS - Electronic Theses Online ServiceGBUnited Kingdo