抄録
Sand compaction pile (SCP) is a method of constructing large diameter sand columns in the ground. The process of SCP installation is known to affect the in-situ soil since it is compressed as the cavity expands to accommodate the sand pile. It is difficult to correctly capture these installation effects in 1g tests on the laboratory floor because of the scale effects. This paper compares the results of frozen pile method of installation at 1g to the in-flight method of SCP installation at high-g using the centrifuge. Pore pressure changes were recorded during the entire installation procedure. No setup of stresses was observed in 1g tests. However, in high-g tests, stress relaxation did not occur during the sand injection stage. The centrifuge test results were then compared to plane strain cavity expansion theory (CET). The findings seem to show that the CET gives a reasonably good estimate at large depth for the entire installation process but not for the residual stress after the casing jack-in during the first stage. Deviation from plane strain CET was also noted at shallow depths. These findings imply that in order to mobilize significant set-up of stress in the improved ground, there must be substantial further cavity expansion during the sand injection stage of SCP.
