Abstract
A programme of centrifuge tests has been performed at the National Geotechnical Centrifuge
Facility (NGCF) at the University of Western Australia (UWA) to provide insight on the installation and capacity performance of suction caissons in different soil types.
Tests were completed at a centrifuge acceleration of 100g in various model seabeds prepared
by the NGCF including a silica sand, a sandy silt, and a clay-over-sand. The model caissons
investigated in this programme of work have an outer diameter of 80 mm – simulating 8 m at
prototype scale and skirt length to diameter ratios of either 0.5 or 1.5.
The test programme comprised 20 caisson tests with a further four half-model tests conducted adjacent to a viewing window to provide details on the installation response. Each caisson test typically involved installation through a combination of self-weight and pumping, followed by pre-shearing and vertical loading. In addition to the model caisson tests, over 40 in-flight penetrometer tests were completed to characterise the soil samples.
The self-weight installation phase was modelled by penetrating the caisson to a prescribed load before commencing a suction-assisted installation phase. The suction installation was achieved by continuing to hold the prescribed self-weight while extracting fluid from the interior of the caisson at a constant flow rate using an actuated syringe pump. After installation the caissons were subjected to a pre-shearing loading sequence, and half of all caisson tests involved cyclic loading sequences followed by compressive loading and then tensile loading.
For the silica sand sample, the cyclic loading sequences resulted in additional vertical
movements of up to 0.37 m for the short caisson and 0.7 m for the long caisson. For the sandy
silt samples, the cyclic loading sequences resulted in additional vertical movements of up to
1.67 m for the short caisson and 1 m for the long caisson.
Following final compression stresses of up to around 2000 kPa in the silica sand, the maximum sealed tensile resistances were approximately 240 kPa for the short caisson and 600 kPa for the long caisson. Following cyclic loading sequences and final compression stresses of up to around 1000 kPa in the sandy silt samples, the maximum sealed tensile resistances were approximately 400 kPa for the short caisson and 830 kPa for the long caisson.
One PIV test using the short half-model caisson (skirt length to diameter ratio of 0.5) was
successfully completed. However, a seal could not be maintained between the caisson and the Perspex window for the longer half-model caissons. A hybrid installation approach was adopted to install the long half-model caisson, which involved jacking and attempted pumping at different depths.
Facility (NGCF) at the University of Western Australia (UWA) to provide insight on the installation and capacity performance of suction caissons in different soil types.
Tests were completed at a centrifuge acceleration of 100g in various model seabeds prepared
by the NGCF including a silica sand, a sandy silt, and a clay-over-sand. The model caissons
investigated in this programme of work have an outer diameter of 80 mm – simulating 8 m at
prototype scale and skirt length to diameter ratios of either 0.5 or 1.5.
The test programme comprised 20 caisson tests with a further four half-model tests conducted adjacent to a viewing window to provide details on the installation response. Each caisson test typically involved installation through a combination of self-weight and pumping, followed by pre-shearing and vertical loading. In addition to the model caisson tests, over 40 in-flight penetrometer tests were completed to characterise the soil samples.
The self-weight installation phase was modelled by penetrating the caisson to a prescribed load before commencing a suction-assisted installation phase. The suction installation was achieved by continuing to hold the prescribed self-weight while extracting fluid from the interior of the caisson at a constant flow rate using an actuated syringe pump. After installation the caissons were subjected to a pre-shearing loading sequence, and half of all caisson tests involved cyclic loading sequences followed by compressive loading and then tensile loading.
For the silica sand sample, the cyclic loading sequences resulted in additional vertical
movements of up to 0.37 m for the short caisson and 0.7 m for the long caisson. For the sandy
silt samples, the cyclic loading sequences resulted in additional vertical movements of up to
1.67 m for the short caisson and 1 m for the long caisson.
Following final compression stresses of up to around 2000 kPa in the silica sand, the maximum sealed tensile resistances were approximately 240 kPa for the short caisson and 600 kPa for the long caisson. Following cyclic loading sequences and final compression stresses of up to around 1000 kPa in the sandy silt samples, the maximum sealed tensile resistances were approximately 400 kPa for the short caisson and 830 kPa for the long caisson.
One PIV test using the short half-model caisson (skirt length to diameter ratio of 0.5) was
successfully completed. However, a seal could not be maintained between the caisson and the Perspex window for the longer half-model caissons. A hybrid installation approach was adopted to install the long half-model caisson, which involved jacking and attempted pumping at different depths.
| Original language | English |
|---|---|
| Number of pages | 257 |
| Publication status | Published - 18 Jun 2024 |