Laboratory testing of large diameter open caisson sinking friction in sand

Abstract

The construction procedure for open-dug in-situ caisson shafts involves concurrent casting of reinforced concrete walls and excavation of soil from within the structure. This process causes the shaft to ‘sink’ into the ground, driven by the self-weight of the walls. A key design concern during construction is the build-up of frictional stresses (skin friction) on the exterior surface of the caisson walls by the surrounding soil. A lubricant filled annulus is typically used to minimise skin friction, however recent field monitoring results suggest this may not always be effective in sands. Ensuring minimisation of skin friction in sands is therefore a critical design priority. This paper presents results from novel experimental testing that simulates the sinking process at a reduced scale. Four tests were conducted focusing on two key factors that influence skin friction: annulus formation and lubrication. Interface stress sensors were used to measure the skin friction and normal stresses at the interface, whilst particle image velocimetry was used to track soil displacements. The results show that lubrication, either with or without an annulus, effectively reduces skin friction. However, lubricating without an annulus only leads to localised friction reduction, suggesting close spacing of lubricant ports can be used to ensure effective friction reduction even without effective annulus formation. The results also show annulus formation without effective lubrication can lead to an increase in skin friction, the opposite of friction fatigue for displacement piles, representing the first experimental measurements to validate caisson field monitoring results in the literature.