Publication

Experimental investigations of driven cast-in-situ piles

Flynn, Kevin
Citation
Abstract
The prediction of displacement pile capacity in sand is hampered by the extreme changes in stress which occur in the immediate vicinity of the pile during installation (Randolph 2003). High quality instrumented tests on steel model piles, such as those reported by Lehane (1992) and Chow (1997), have helped identify other factors that have a bearing on preformed displacement pile behaviour in sand. These factors include the extent of soil displacement during installation and loading, the reduction in shaft friction due to increasing load cycles during installation (referred to as friction fatigue), increases in radial stresses due to dilation at the pile-soil interface, differences in shaft resistance with loading direction (i.e. compressive and tensile loading) and increases in shaft capacity with time, i.e. pile ageing. The majority of these phenomena have now been incorporated in four new cone penetration test (CPT)-based design methods which give superior estimates of preformed displacement pile capacity in sand in comparison to traditional methods. The knowledge gained from the high-quality studies of displacement pile behaviour is now being applied to other pile types such as partial displacement piles (e.g. open-ended piles) and replacement piles (e.g. bored and screw piles). One category of pile which has received sparse attention is the driven cast-in-situ (DCIS) pile which is typically classified as a large displacement pile, despite sharing certain aspects of its construction with replacement pile types. Furthermore, there are relatively few case histories of load tests on DCIS piles in the literature to verify the assumption that they behave as full displacement piles. The behaviour of DCIS piles during installation, curing and maintained load testing was therefore investigated by constructing a total of seven instrumented DCIS piles in layered soils and sand at sites in the United Kingdom. The resistance of the steel installation tube during driving was derived using instrumentation fitted to the DCIS piling rigs. The variation in temperature and strain after casting was monitored continuously in three of the test piles to examine the development of residual loads during curing. After developing sufficient concrete strength, the test piles were subjected to maintained compression load tests to failure (i.e. a displacement in excess of 10 % of the pile diameter), with the shaft and base resistance during loading derived from strain measured within the test piles using vibrating wire strain gauges. The installation resistance derived by the rig instrumentation shows good agreement with the base resistance profile derived by the University of Western Australia UWA-05 method using the Dutch averaging technique. Residual loads developed during curing of the DCIS piles installed in layered soils due to consolidation settlement of soft soil layers as pore pressures induced by the driving process dissipated. On the other hand, residual loads for DCIS piles in uniform sand were negligible. The instrumented DCIS piles in sand exhibited a clear reduction in normalised local shear stresses and radial effective stresses at failure with distance from the pile base, i.e. friction fatigue, which is a well-known characteristic of preformed displacement piles and its existence for DCIS piles implies that radial stresses during driven installation of the steel tube are not erased upon concreting and tube withdrawal. The normalised base resistance at failure showed excellent agreement with the UWA-05 design method for driven closed-ended displacement piles, with the onset of degradation in base stiffness occurring at large base displacements. Design correlations have subsequently been developed based on the results of the instrumented DCIS pile tests. The main implication of the experimental data for DCIS pile design in sand is that the shaft, base and total capacities of a DCIS pile are similar to a preformed closed-ended displacement pile of equivalent dimensions. In keeping with this finding, an examination of the predictive performance of seven CPT-based displacement pile design methods using a database of 26 DCIS pile load tests with adjacent CPT qc profiles demonstrated that the recent methods provide improved estimates of DCIS shaft, base and total capacity in comparison to traditional simplified methods. However, a statistical study of DCIS pile load-displacement behaviour in sand using a database of 105 pile load tests showed that the total resistance of a DCIS pile tends to mobilise at a slower rate in comparison to preformed driven displacement piles, implying that DCIS piles may exhibit greater levels of displacement for a given applied load.
Publisher
Publisher DOI
Rights
Attribution-NonCommercial-NoDerivs 3.0 Ireland