Study On Refined Damage Identification Of Long-Span Bridges

Because the real-time features and complexity in bridge structural health monitoring system, the damage identification methods currently used have limitations, and the external excitation characteristics of the specific structural system wasn't considered. In the thesis, based on wavelet analysis, using the good time-frequency analysis capability of wavelet, the study on damage identification of long-span cable-stayed bridge under the wind load is conducted. The main subjects are summarized as follows:(1) The damage numerical simulation under random loads and the bearing damage test under shock load on continuous beams are made, the load-dependent and damage sensitivity when using the acceleration, the autocorrelation function, the crosscorrelation function and the impuls response function respectively to do wavelet packet energy analysis is analysed.(2) Three different methods of finite element modelling of long-span cable-stayed bridge are discussed, the static and dynamic results based on different finite element modal are compared. The wind field of long-span cable-stayed bridge is generated, and a simplified calculation method to transform wind field into wind load is adopted.(3) Dynamic time history analysis of long-span cable-stayed bridge under the wind load in different damage case is performed, the changing regularity of the damage criterion got from different scales and different directions in different damage degree is studied. The wavelet packet energy analysis of the impuls response function under different wind loads is made, the load-dependent and damage sensitivity is discussed.