Structural Damage Detection Methods Based On Dynamic Analysis

Large structures may be damaged due to various loading applied in their lives. Damages in some important parts may even lead to catastrophic results. Therefore, it's important to find an efficient and effective damage detection method, to ensure the structure safety.Mechanical characteristics will change in damaged structures. The theory of damage detection is based on that damages of the structure could be reflected from its mechanical characteristic changes. As the dynamic analysis method, which could indicate all the nature of a structure, is better than traditional ones in overall detection, it's becoming the hotspot among both domestic and foreign researchers.Several damages of large structures are simulated in the paper. Firstly, typical modal analysis methods are applied in an offshore platform and in a cable-stayed bridge. Through these two examples, the validity and accuracy of each method in damage identification are discussed. Wavelet transform has found more and more applications in non-damage detections due to its ability of giving the details of the signal. Therefore, the wavelet transform is used in this paper to detect the damages appeared in the offshore platform and in the concrete bridge. Subsection method has been used for complex structures where many end points and turning points exist. And the modal expansion method is introduced for each subsection.From the numerical results, it's proved that the methods introduced in this paper could give good results if suitable form of modal and wavelet basic function are used. Furthermore, wavelet transform could eliminate high frequency noises appeared in the unsteady signal and keep its original sharp changes.Up to now, various methods have been successfully used in non-destructive damage detection, but none of them is universal. In order to get an accurate result with lower cost, it is suggested that several model analysis methods be used to locate the damage in one structure.