Damage Identification Based On Continuous Wavelet Transformation For A Real Cable-stayed Bridge Structure Under Construction Condition

Structural damage in the construction period produced by the environment and the structural temporary loads will greatly reduce the construction quality and increase the safety risk. Therefore, the damage detection techniques for structures under construction have taken more and more attentions in the field of bridge engineering.Wavelet analysis is a new approach and prevails for processing the structural testing data with a great prospect of engineering applications. The damage identification problem will be systematically investigated for the bridge structures under the construction conditions by using the continuous wavelet transform analysis techniques. The structural modal analysis is performed to obtain the structural modal shapes by using the FEM software ANSYS and MIDAS. Then, the modal shapes are input into the software MATLAB as the input singles to perform the Morlet wavelet transformation which will be applied to the damage identification. First, a damage model of two-span continuous beam structure is established by using the software ANSYS. The structural damage identification results for different damage location and degree are analyzed and discussed in details. Different order modal shapes and different kinds of modal shapes (i.e. displacement, angle displacement and curvature modes) are respectively applied to the damage identification. The corresponding results are discussed and compared accordingly. Second, Lijiasha Bridge is taken as a real engineering example. It is a large-span cable-stayed bridge and is now under construction in Guangdong province. Four typical construction conditions are selected to perform the damage detection, i.e., the minimum bi-cantilever condition, the maximum bi-cantilever condition, the cross-border closure condition and the cross-closure condition. FEM analytical models are developed by using the software MIDAS. Then, the damage identifications for the above four construction conditions are carried out by using the wavelet analysis technique.The results show that the continuous wavelet transformation of the curvature modes is sensitive to the structural damage condition. The damage identifying results are satisfied. However, the damage identifying results would be affected by the structural boundary conditions. In general, the continuous wavelet transformation of the structural curvature modes has an ability to quickly and reliably identify the damage location and degree in a structure under construction condition.