Experimental assessment of the performance of a bridge pier subjected to flood-induced foundation scour

Scouring around piers is recognised as one of the main causes of bridge failure. Local scour refers to the localised erosion of soil at piers and abutments caused by flow-induced vortices around their base. This paper studies experimentally the response of a bridge pier, supported on a cylindrical embedded foundation, subjected to flood-induced scour. A hybrid two-step methodology is developed to study the hydraulic and the mechanical part of the problem. In the first step, the hydraulic problem of local scour around a pier is experimentally modelled in 1g using a recently developed miniaturised tidal generator. The experimentally generated scour hole is then three-dimensionally scanned to produce a three-dimensional (3D)-printed mould. The latter is used in the second step to reproduce the scour hole in an Ng model, subsequently tested in a drum centrifuge to study the mechanical part of the problem under proper stress scaling. Foundation performance before and after local scour is studied through vertical, lateral monotonic and slow cyclic pushover tests. The effects of general (uniform) scour are also investigated by removing a soil layer of constant thickness. Local scour is shown to have a minor effect on vertical bearing capacity. In stark contrast, the lateral performance is significantly affected, with the foundation moment capacity M_ult being reduced by up to 38%. The effect of general scour is even more pronounced, leading to 48% reduction of M_ult. The rate of cyclic settlement accumulation is also much more severely affected by general scour as compared to local scour. Overall, the effects of local scour on foundation performance differ substantially to those of general scour, bringing into question the common simplification of ignoring the geometry of the scour hole, making no distinction between local and general scour.