Tins paper summarizes the design and results of a program of experiments conducted in a fully-automated, laboratory Caisson Element Test (CET) cell that has been developed to investigate fundamental aspects of caisson anchors under conditions of axial tensile loading. The CET cell comprises a miniature cylindrical caisson (outside diameter 5.lcm, wall thickness 0.15cm) which is installed within a consolidated, homogeneous, saturated element of resedimented clay. The model caisson cap and wall are separate units whose displacements are controlled independently; while additional instrumentation measures surface displacements as well as pore pressures beneath the cap and within the clay sample The experiments show that when the caisson is installed by underbase suction, almost the entire volume of soil displaced by the wall moves inside the sod plug. Very low effective stresses occur at the top of the sod plug, while the measured wall forces are consistent with skin friction data for driven piles. After installation, the equilibration phase involves a rapid redistribution of forces to the caisson wall with minimal settlement. The total load capacity and maximum wall resistance m undrained (short-term) monotonic axial load tests are mobilized concurrently at a cap displacement of approximately 0.2cm, comparable to the wall thickness. In contrast, there is a much stiffer response of the cap unit, which yields at a tensile force well below the expected cavitation condition and then deforms with negligible hardening, contributing 35% to the total resistance. The long term caisson capacity under sustained tensile loading is significantly less than the maximum wall resistance measured in short-term monotonic load tests. Limited data also suggest that the long term caisson capacity is affected by the method of caisson installation.
Although suction caissons have been successfully used as permanent anchors for prototype Tension Leg Platforms (TLP) m low permeability clay deposits, there are still many aspects of their behavior that are not well understood. In prototype applications, caisson installation is achieved through. 1) gravity penetration, due to buoyant self-weight; followed by 2) underbase suction obtained by pumping water from inside the caisson cell (i.e., differential vertical pressures acting across the lid) The relative contributions of each phase are controlled by the embedment to diameter ratio, undrained strength profile, etc The first phase of installation is analogous to the unplugged penetration of large Diameter open-ended piles, however, little is known about the disturbance effects caused by underbase suction Similarly, there is almost no information on the post-installation behavior including Dissipation of excess pore pressures, redistribution of stresses with the caisson cells and set-up of effective stresses m the surrounding soil. Tensile loads from the TLP superstructure include a) long-term mooring forces, and b) short-term surges and cyclic loads during storm conditions. Tensile loads applied to the top of the caisson are initially resisted by underbase suction pressures (i.e., negative increments of pore pressures acting between the lid of the caisson and the top of the soil plug) and shear resistance primarily along the outer caisson wall-soil interface.
