Abstract
The evaluation of Negative Skin Friction (NSF) is a common problem in the design and construction of pile foundations in soft ground. However, to date most of the current design approaches are based on simplified methods and are not satisfactory. Furthermore, large group effects are predicted using current design guidelines, especially for central piles. However, relatively small group effects have been reported from previous experimental observations documented in the literature. Investigations involving both centrifuge tests and Finite Element Analyses (FEA) therefore have been performed to clarify aspects of the pile behaviour subjected to NSF.A single instrumented model pile, passing through clay and bearing in sand, was subjected to various changes of loading in centrifuge tests carried out at the Cambridge beam centrifuge. Loading events included sudden changes of centrifugal acceleration and rapid application of load at the pile head. Consolidation and swelling of the ground was simulated through changes of self-weight (i.e. gravity-turn-on-technique) during the centrifuge flight. Soil surface and pile movement and pore pressure changes inside the clay were monitored. Greater soil movement than pile movement was observed and hence the development of the NSF effect was clearly demonstrated. However, it was found that there was a strong thermal effect in the pile instrumentation system, thereby reducing the stability of the instrumentation system. This instability presented some difficulties in the interpretation of the test results. Various attempts were made to improve the pile performance and to allow better interpretation of the centrifuge test data.
2D and 3D FEA were conducted to investigate the development of NSF for a single pile and piles in a group. The development of dragload and the resulting group effect was found to be heavily dependent on slip at the pile-soil interface. This mechanism is governed by the interface friction, the pile configuration and the soil settlement. Various factors, such as the relative position of piles within a group, the number of piles, the pile spacing, the surface loading, the soil stiffness, as well as the pile type should therefore be considered in the determination of dragload and group effect. Even when the potential exists for the development of NSF on piles in soft ground, dragload (compressive force) is not normally a major design issue for the pile itself, since dragload rarely exceeds the strength of the pile material. However, downdrag could present some difficulties from a serviceability viewpoint if excessive settlement is induced. Piles should therefore be installed in a stiff layer to counteract downdrag, depending on driveability and the expected dragload. Friction piles in soft ground should therefore be designed with great care.
It was found that the estimation of dragload and group effects from current design practice was neither satisfactory, nor realistic. Dragload was normally over-estimated from conventional methods (i.e. elastic and continuum analyses). However, less dragload and less reduction of the dragload for piles inside the group were reported from field observations. The results from the FEA predicted a reduction in dragload due to group effects varying from 5-48% and 19-79% for 3x3 and 5x5 groups in 2.5D spacing respectively. These group effects are significantly smaller than current design guidelines. The results of analyses were in reasonable agreement with elastic solutions, recent theoretical studies and a number of field observations.
The behaviour of a pile in swelling ground was also studied. It was found that reinforcement should be provided for the full length of the pile when the possibility of uplift of the pile exists in order that the pile has sufficient tensile strength. Similar trends were obtained regarding development of tensile force and group effect as for piles in consolidating ground.