Study On Damage Identification Method Of Bridge Structures Based On Wavelet Analysis

Because of the bridge structures' complexity itself and the influence of factorssuch as natural environment, the recognition effect of most of the existing damageidentification methods is not very ideal in practical application. However, with the fastdevelopment of domestic traffic transportation and bridges construction, the problemof unsafe bridges and disease bridges is increasingly outstanding nowadays. Thereforeit is very necessary to further study the bridge damage detection method and explore asimple, easy and effective identification method. In addition, it has far-reachingresearch and engineering value.This paper mainly studies the bridge structural damage and the damage dynamicdetection method based on the wavelet analysis tool. The contents are as follows:(1) The vehicle-bridge coupling vibration experiment system, based on a doublepylon cable-stayed bridge model, is established by considering the vehicle bumpingcaused by the local larger roughness of the bridge deck. Through the experimentmethod, the effects of the vehicle bumping on bridge vibration are studied. Thecontinuous wavelet transform is adopted to analyze the time-frequency characteristicsof vibration signals of the girder. And the discrete wavelet transform is used to extractthe signal components caused by the vehicle bumping. At last, the fatigue damageproblem of the bridge deck caused by the vehicle bumping impact is briefly analyzed.(2) An approach of structural damage identification based on the change ofwavelet sub-band signal energy curvature is proposed. The structural responses underimpact excitation are decomposed into different sub-bands with separation of differentmodal responses. Then the damage index of relative energy curvature difference isdefined and used to identify the damage location. An example of a simply supportedbeam bridge model is used to show that the proposed method can examine the damagecorrectly and effectively, and the experiment results are well agreed with thenumerical simulation results.(3) A new method of structural modal parameter identification based on thewavelet analysis is proposed. On the basis of the discrete binary waveletdecomposition of the signals, each order modal response is extracted. For each ordermodal response, the method of vibration waveform is adopted to compute the frequency and damping ratio parameters. Each vibration mode of the structure is thencomputed by combining the same order modal responses of multiple points. Also asimply supported beam bridge model is used to verify the proposed method. Theresults are in good agreement with the modal analysis software.