Scouring at the base of bridge piers is the major cause of bridge collapses worldwide. Computing the scour risk of bridge foundations is therefore key for a correct management and allocation of resources for maintenance and scour mitigation works. Existing risk-assessment models compute the vulnerability of bridge foundations to scour by comparing the equilibrium scour depth associated with peak-flow discharges characterized by a given return period (usually of 100–200 years) with the critical foundation depth of the bridge. This approach neglects completely the history-dependent and time-dependent nature of scour. Yet, it is well known that bridge collapses can often be induced by the accumulation of scour during multiple flood events. This study aims at developing a novel probabilistic framework for the computation of bridge-pier vulnerability to scour using a Markovian approach to account for memory effects in scour development. The paper focuses on the case of local pier scour occurring in clear-water conditions whereby cumulative effects are significant, well understood and known to be the cause of recent reported bridge collapses. A simplified numerical example consisting of an idealised bridge pier in a canal is considered to clarify the application of the proposed framework and to shed light on the effects of some assumptions introduced to simplify the probabilistic scour assessment.
展开▼
