Void Defect Identification Of Concrete Filled Steel Tube Based On Vibration Characteristics

Concrete-filled steel tube structures in service inevitably have a different type of voiding when they are subjected to human activities and external loads.With the void defects develop to an extent,structure failure or instability may happen,and it will greatly affect the bearing capacity and stability of the concrete-filled steel tube structure.Therefore,it is quite necessary to take effective measures for defect detection and safety assessment to ensure the safety,applicability and durability of concretefilled steel tube structures in service.With a consideration of the dynamic characteristics of concrete-filled steel tube structure and their service environment,the vibration signal is processed as a target signal.By detecting the change of structural vibration characteristics,reliable indexes of defect detection is established.After that,the accuracy and effectiveness of the indexes are verified via concrete-filled steel tube finite element models and concrete-filled steel tube dynamic tests.The main contents of this paper are as follows.(1)The indexes of the relative wavelet entropy(RWE),the energy change rate of wavelet packet(WPECR),the curvature difference of wavelet packet energy(WPECD)and the energy difference curvature of relative wavelet packet(RWEDC)are summarized.In order to highlight the information of the wavelet packet frequency band signal,the analytic modal decomposition theorem is first introduced to extract the low-order components of the structural vibration response signal.Then wavelet packet decomposition is performed and the component signal whose frequency band owns larger energy proportion is selected for reconstruction.At last,the index of the first order intrinsic mode function relative to the wavelet energy change rate(CRIMT)is defined,and the validity and applicability of the index are verified by finite element models and dynamic tests.(2)For most indexes,it is too difficult to extract structural defect features from non-stationary signals.In addition,the information under undamaged cases should be known beforehand.Thus,a new defect detection method by a combination of modal signal decomposition and kurtosis is proposed.Firstly,the signal is separated by a modal decomposition method.Then,weighted kurtosis(WK)is used to process each component and the component whose WK is greater than its mean value is selected as the optimal component for reconstruction.After that,the Teager energy operator(TEO)is used to calculate the reconstructed signal and then fast Fourier transform(FFT)is introduced to analyze the TEO of the reconstructed signal.Finally,the kurtosis of the new signal processed by FFT is solved and used for detecting structural defects,with the final results are summarized and analyzed.