Based On The Improved Damage Model Of Composite Container Structural Damage Detection

Online damage detection of initial and tiny crack in a cylindrical composite vessel is studied in this paper. The in-situ active damage detection of composite structure can be implemented using surface-bonded piezoelectric patches with signal processing based on vibration analysis.To meet the actual necessity of the vibration analysis of structural damage detection, an improved modeling method is proposed for small crack modeling in ANSYS dynamic analysis. This method establishes a specific, numerical relationship for crack status between the modification coefficients of element stiffness matrix. The first step of this improved modeling method is to determine modification coefficient of element stiffness matrix based on the coherence of natural frequencies for direct crack modeling and modification coefficient modeling. Next step is to verify the coherence of the frequency-response functions.For the integrity and damage status, the sub-signal energy spectrum variations are simulated on the base of response signal decomposition in various frequency bands by wavelet packet analysis.Taking the energy spectrum variations of the decomposed wavelet signals as the inputs of the artificial neural network, neural network is training for crack size and position identification.The result shows that the improved crack modeling method has goodcoherence with direct crack modeling. The energy spectrum variations as the damage dynamics feature have good sensitivity. Therefore it can detect very tiny cracks in cylindrical composite vessel.