Numerical Simulations And Experiments Of Structural Damage Detection Based On Acoustic Emission

Acoustic Emission technique is a new dynamic NDT method based on the elastic waves from internal structure, from which the properties of acoustic emission source and therefore the damage degree of internal structures can be determined. However, transient elastic waves produced by acoustic emission sources can generate acoustic emission signals with various frequencies and modals in the propagation, which makes it difficult to locate the acoustic emission sources and make accurate determination about their properties. Thus it is not easy to apply acoustic emission technique in non-destructive testing of structures. Consequently there is a pressing need to investigate the propagation of acoustic emission in structures. Simulation of propagation of acoustic emission based on finite element method, and the corresponding experiment using both by a home-made and a commercial acoustic emission detection systems, were investigated; also the numerical and experimental results were compared.Aiming at investigating the propagation of acoustic emission signal, finite element method was used to simulate the propagation of acoustic emission signals in thin aluminum and the reflection waves from structural boundaries. Time-domain waveform of acoustic emission was obtained from the finite element model. Discrete wavelet transform (DWT) was used to decompose the signals in order to analyze its characteristics. Daubechies wavelet was selected to decompose the signals for multiple levels. It was demonstrated that low-frequency analog signal covered the reflected signal from the geometric boundaries. Moreover, reflected signal could clearly be seen from the detailed coefficients of DWT decomposition. FFT was used to analyze the detailed coefficients and the result showed its frequency range from which the filters could be designed to use in the following signal processing..An acoustic emission detection system based on virtual instrument was developed, including hardware and software systems. The hardware system include a Kistler piezoelectric acoustic emission sensor, Kistler acoustic emission coupler, and a high-speed National Instrument data acquisition card etc. The software system mainly included five parts, user login, data acquisition and storage, signal processing such as denoising, data display and playback and data printing. The home-made system could be used to acoustic emission data acquisition and signal processing.For the experimental investigation, two emission detection systems including the home-made system and SAMOSTm detection system, were built to collect the pencil-lead signals. These two experimental signals were both processed by band-pass filters. The two experimental results have an agreement each other and it meant that the high reliability of the home-made system. In the meanwhile, there were good agreement between the simulation and experimental results and therefore it validated the finite element model of acoustic emission and also finite element numerical method could be an effective method to analyze the propagation of acoustic emission. It could provide numerical models to detect structural damages based on acoustic emission and precise theoretical foundations to use acoustic emission in non-destructive testing.