Application Of Two-Dimensional Structure Damage Identification Method Based On Wavelet Analysis

Beams and plates are relatively common structures in high-speed trains’car-body, and when trains are running, various damages will occur in the structural components. However, the damages reached to a certain extent, which will affect the safety of train’s operation, even in severe cases, which may lead to disastrous consequences and huge loss. Therefore, researches of damage identification on the beams and plates have become an important study topic, and there are important theoretical and engineering applications.This paper firstly introduced the structure damage identification based on Wavelet analysis, which have space for localized nature. The status of research on damage identification wavelet analysis is analyzed in detail, and the advantages and existing problems are summarized. And the basic theory of wavelet analysis is elaborated systematically. By means of finite element numerical simulation, and the damage identification theory based on wavelet analysis, the damage identification of the cantilever in one-dimensional and the plate in two-dimensional is studied.This paper established the cantilever beam damage model, by the numerical simulation in ANSYS, the shear stress of the cantilever beam was calculated, and the continuous wavelet transform was conducted. Finally, the local maxima of the wavelet coefficients can identify crack’s position. The results show that the crack identification by the Daubechies (db4) wavelet has good result, and the correspondence between the extent of damage and the maximum value of the wavelet coefficients was established.This paper established the two-dimensional plate’s damage model, by numerical simulation in ANSYS, the modal shapes of the damaged plate was calculated and obtained, and two-dimensional wavelet transform was conducted. Based on the wavelet transform’s high-frequency coefficients, the identified damage locations in plates was identified. Damage identification in thin plate having different shapes and different wavelet function was studied. The results show that higher modes are more sensitive to damage, and the Symlets (sym4) wavelet is more suitable for two-dimensional identification. Using the 1:8 ratio of high-speed trains the car-body as the application background, and designed a pilot program for the ratio car-body. Studied the roof of the ratio car-body’s damage identification problems of single-point or multi-point position, to prove that the proposed two-dimensional wavelet transform recognition validity of sheet crack damage detection.In this paper, take the damage identification of one-dimensional cantilever and two-dimensional plates for example. The parameter identification, wavelet selection, mode selection and other issues have been studied by wavelet transform theory. However, the theory that for any one-dimensional or two-dimensional structure has some applicability. It shows that the method is simple, easy to operate, with high precision, can be applied to the proportion of high-speed train car-body damage identification beams and board.