ABSTRACT
In recent years, with the increase of the service life, the bridge scour has emerged as a growing concern for bridge engineers. Scour takes away the surrounding soil, resulting in the decrease of foundation depth, stiffness and bearing capacity, which influences the health status of bridge significantly. Existing research results show that the braking load can induce significant dynamic response along the bridge, which is helpful to detect the substructural damage. In this investigation, the acceleration responses of piers top under the vehicle braking were processed by wavelet packet decomposition, and the change of wavelet packet energy variance in selected frequency band was chosen as the damage index of scour state. Taking a concrete continuous beam bridge as an example, different scour positions are simulated by removing the corresponding soil spring elements to verify the proposed scour damage identification method, the results indicated that the proposed damage index had a high sensitivity for the single and multiple scour damage.
