Experimental study of axial behaviour of tapered piles: Discussion

Abstract

Discussions 1203 The authors present test data and analysis from testing tapered piles in a 1.5 m high and 1.5 m diameter cylindrical test chamber lined with an “air bladder” enabling the lateral stress to be increased. No similar arrangement is used to increase also the vertical stress. This creates a soil that is highly overconsolidated. The authors state that the test arrangement does not represent the state of stress along a prototype pile. The discussers agree. However, the authors also state that the purpose of the device (the testing chamber) “is to model the state of lateral stress [against the pile] along different “segments” of the pile...”. If this means claiming that the state of lateral stress against the model pile would be similar to the state of stress against a real (prototype) pile, the discussers disagree. Inflating the lining bladder compresses the soil and increases the stress acting against the pile (placed in the centre of the device). The authors appear to assume that the distribution of the so imposed stress is lateral and uniform. However, the soil will not move uniformly and it will not only move horizontally. There will also be an upward component near the pile head and a downward component near the pile toe. These movements will both cause stress rotation in the soil adjacent to the pile and impose load (residual load) in the pile. The stress against the pile will therefore vary along the pile. Moreover, the resulting increase of density cannot be uniform. When testing, the authors first installed and secured a model pile in the test chamber and then poured sand around the pile. As Hanna and Tan (1973) have shown in their experiments with instrumented model piles, the placing of the sand has a pronounced effect on the state of stress in the sand and in the pile. The authors’ arrangement will have had the effect of introducing additional load, i.e., residual load, in the pile. For information on residual load, see Nordlund (1963), Hunter and Davisson (1969), Altaee et al. (1993), and Fellenius and Altaee (1994). The authors report that all strain gauges were zeroed before the start of each test, but they do not explain why. However, zeroing the gauges does not remove the residual load in the pile; the pile and the measurements will remain under the influence of the residual load. If the gauges had been zeroed immediately on placing the test pile in the chamber and before the sand was poured around the pile, that is, when the load in the pile truly was zero, then the load in the test pile at the start of the test (the residual load) could have been ac counted for.