Ageing and axial cyclic loading studies of displacement piles in sands

Abstract

An investigation is presented into the mechanisms governing both the ageing and axial cyclic loading characteristics of displacement piles in silica sands. The thesis considers first the state–of–knowledge regarding the axial capacity of displacement piles in silica sands. Three main areas of uncertainty are identified; the stress regime setup by installation, the mechanisms of ageing that lead to capacity increases with time (set-up), and the effects of axial cyclic loading. New laboratory experiments are then described that involved tests with extensively instrumented 36mm diameter Mini-Imperial College Piles (Mini-ICP) with roughened (Rcla ~3.5μm) stainless steel shafts and 60o conical tip bases, that could measure axial loads, and interface radial and shear stresses at multiple positions along their shafts. Less extensively instrumented piles with varying diameters were also tested. Ten installations were made in the 1.2m diameter, 1.5m deep Grenoble– INP calibration chamber. Fresh pluviated sand masses were formed for each installation, which were typically instrumented with multiple commercially sourced (Kyowa and TML) miniature sensors to measure radial, vertical and circumferential stresses in the sand mass during pile installation, ageing, and axial static and cyclic loading tests. Key parameters that might affect pile behaviour were then isolated and considered in turn. The interpretation links the model tests to instrumented field studies and the ageing trends established from a field database. The model piles’ axial cyclic loading responses are analysed by developing cyclic interaction diagrams which are linked to full scale tests and laboratory experiments to identify the key mechanisms governing field cyclic behaviour. The interaction diagrams provide a straight-forward screening tool for addressing axial cyclic loading in practice. Reference is made to more elaborate procedures and to the experiments’ scope for validating numerical models.