Time-varying Modal Parameters Identification And Damage Detection Of Steel-timber Composite Footbridge

The steel-timber composite(STC)structure is frequently used for footbridges because of its good bearing capacity and environmental harmony.During the service life,the STC bridges are inevitable to be damaged.The resultant response signals are time-varying and nonstationary.Therefore,it is urgent to perform applicability assessment,safety evaluation,maintenance and reinforcement on the STC bridges.However,most parameter identification methods of time-varying structures have drawbacks such as low time-frequency resolution and restrictions of signal types.To address these problems,a combined method is used to identify the instantaneous frequency of the pedestrian-bridge interaction system which aiming at the time-varying and non-stationary of the footbridge response signal under pedestrian excitation.After that,the frequency change equation of the pedestrian-bridge interaction system under single pedestrian and crowd load are proposed.In addition,a timevarying damping identification method is proposed here on the basis of instantaneous frequency identification.In view of the problem that composite structure footbridges are prone to bolt loosening damage,a bolt loosening localization method based on normalized curvature difference of frequency response functions is presented.The validity and accuracy of the proposed methods are verified by numerical examples and footbridge walking tests.The main research work and innovations of this dissertation are as follows.1.At first a model of STC footbridge interacted with human body is established and its vibration response is obtained.Then,a combined method(CM),which is a combination of extended analytical modal decomposition,recursive Hilbert transform and zoom synchrosqueezing wavelet transform,are introduced to identify time-varying modal characteristics and the instantaneous frequency curves are extracted with high time-frequency resolution.To investigate the impact of human parameters on the natural frequency of the human-bridge interaction systems,extensive simulations on STC bridges with different spans are carried out.After that,the relationship equation about the instantaneous frequency of human-bridge interaction system and the pedestrian movement is obtained by curve fitting.The results demonstrate that the proposed model is capable of simulating the interaction between human and STC bridge well.The identification of frequency change of the interaction system under single person and crowd load by CM is accurate and robust.In addition,the empirical equation provides a reference for the design and maintenance of pedestrian bridges.2.Since the identification of time-varying damping ratio based on continuous wavelet transform and standard synchrosqueezing wavelet transform is inaccurate and low recognition resolution,a damping identification method based on improved synchrosqueezing wavelet transform is proposed.At first,the improved synchrosqueezing wavelet transform is used to obtain the instantaneous frequency curve with high time-frequency resolution,and then the logarithmic amplitude slope of wavelet coefficients is calculated,leading to the instantaneous damping ratio curve with high time-frequency resolution.At last,a numerical example of STC footbridge and a walking test of composite structure footbridge are employed to verify the effectiveness and the accuracy of this method.3.A new bolt loosening localization method is proposed for STC structures by establishing a damage index called normalized curvature difference of frequency response functions(NCDFRF).In this method,fast Fourier transform is first performed on the excitation and acceleration response signals and then frequency response functions(FRFs)between them are calculated.Based on this,the index of NCDFRF is calculated to localize the position of bolt,aiming at avoiding damage position misjudgment caused by bolt drilling and preload.