This paper presents a calculation method to predict the installation process of a suction caisson in sand overlaid by clay. Calculation methods are first described for homogenous clay and sand profiles, with the approach for the latter case being based on the measured cone resistance profile and a simple linear reduction in resistance with increasing suction. Example calculations for a sand profile are compared with measurements to help validate the approach. Modifications to account for an overlaying clay layer are then discussed, and two alternative approaches are described, depending on the extent to which the clay layer remains intact within the caisson. A flow model is introduced that balances the water pumped from the caisson with that displaced by the caisson and seepage flow through the soil. By combining the installation resistance models with the flow model, predictions are made regarding the complete installation process over time for given suction caisson geometry, soil conditions and pumping speed. Theoretical predictions are compared with model test data and are provided in the paper.
Installation of a suction caisson consists of a self-weight penetration phase and a suction-assisted phase. During the latter phase, water is pumped out from inside the caisson creating a differential pressure between the inside and outside of the caisson. This differential pressure provides an extra driving force to help overcome the soil resistance during installation of the caisson. In the case of relatively permeable soils the reduced internal pressure will also induce seepage flow through the soil, leading to reduction in the effective stresses and hence penetration resistance.
The installation resistance of a suction caisson comprises internal and external friction along the caisson wall and bearing resistance at the tip. The unit resistance, either friction or bearing, must be estimated from some measure of the strength of the soil. In clay soils the obvious choice is the undrained shear strength, su, which is almost universally adopted in practice1. In sands, shaft friction and bearing resistance values for suction caissons and piles have traditionally been evaluated from fundamental properties such as the friction angle, F', and in situ effective stresses2, 3. However, the most direct measure of strength is the cone resistance, qc, and design methods based directly on qc have become increasingly popular over the last decade4–6. This approach will therefore be used here to predict the installation resistance in sand, following the basic methodology described in Senders and Randolph7.
For caisson installation in sands, it has been shown that dramatic reduction in installation resistance occurs due to suction-induced seepage8–10. This has been implemented in the proposed installation resistance calculation for caissons in sand, using a simple linear reduction in internal friction and tip bearing resistance with increasing suction7. In a soil profile comprising clay overlying sand, the question arises as to the extent that the clay will impact the seepage, and thus limit the decrease in tip resistance during penetration of the underlying sand. Of particular concern is the potential lifting of the clay plug