Research On Dynamic Characteristics Of Welded Truss And Diagnosis Method Of Crack

Welded truss has been widely used in industrial and civil buildings as a kind of important mechanical bearing structure. Under the influence of long-term mechanical load, environmental temperature and various corrosion, it exists safety problems inevitably when the welded truss is used. Currently, the main potential threat of the safety of welded truss application is crack. The crack growth will lead to the structure break, which will cause damages to people’s lives and properties. Therefore, researched the dynamic characteristics of welding angle steel truss and diagnosis method of crack in the thesis.Solid model of the welded truss was made and the static and dynamic characteristics were analyzed by using finite element analysis software ANSYS. The contour of the deformation, the stress and the strain were obtained after the static analysis, and made a comparison of the static analysis results of the truss crack before and after. Made dynamic analysis of trusses with different cracks after the static analysis, and obtained the inherent frequencies and vibration modes in various circumstances. Analysed the influence of the crack parameters on these two kinds of modal parameters. Obtained the changing rules of these two parameters—the inherent frequency and vibration model—changing with the crack began and the changes of the location and the length of the crack.Made experiment modal analysis of truss without crack and cracked truss respectively. Researched each modal parameter changing with crack began, and made a comparison with the finite element analysis results. It indicated that two analysis methods had the same changing rule, and verified the reliability and accuracy of the finite element analysis results.Based on that the crack can produce weak friction vibration signals, made crack identification by using autocorrelation combined with wavelet packet transform. Results show that the delayed autocorrelation is able to get rid of the effects of Gaussian white noise, the wavelet packet transform can effectively extract the weak friction vibration signals, and the combination of delayed autocorrelation and wavelet packet transform is able to diagnose that there exist crack. The thesis provide the basis for optimization design of structure dynamic characteristics through the analysis of the dynamic characteristics of truss structure, and improve the efficiency of the structure design. At the same time the methods used in this subject—the finite element analysis and the experimental modal analysis—are also easily extended to other similar research and analysis. The crack diagnosis method used in the thesis is of some reference value to other structural crack diagnosis.