Research On Damage Source Localization Method Of 3D Braided Composites Based On Acoustic Emission Time-frequency Analysis Technology

With the three-dimensional braided composite materials have been more widely used in the aviation,aerospace,automotive and construction engineering fields,the problems of damage and destruction have become the key to production and use.The technology of AE has the characteristics of dynamic,sensitive,real-time and strong anti-interference,so it has been gradually applied to the nondestructive testing of the three-dimensional braided composite materials.Because in the process of increasing load of the composite materials will generate a large number of AE signals which contain abundant damage source information.The analysis of these signals is helpful to obtain the damage mode of composite materials,and the obtain the damage location,and achieve the purpose of the damage source location.First of all,this paper will compare five typical methods of time frequency analysis,including STFT,WVD,Wavelet transform,S transform and Hilbert-Huang transform,and use the time-frequency analysis method to simulate the formation of five acoustic signals.The conclusion of the study shows that the STFT,WVD,Wavelet transform,S transform are affected by the problems of fixed window function,wavelet basis function selection and Heisenberg uncertainty principle,which cannot reveal the time-frequency characteristics of AE signals.However,the HHT based on the signal,without any artificial interference,is suitable for the analysis of nonlinear and non-stationary signals.Meanwhile,it has strong adaptability to the signal and can well reflect the local frequency characteristics of the signal.It is suitable for the time-frequency analysis of the three-dimensional braided composites AE signal.This thesis takes three-dimensional four-directional braided composite materials as the research object,in order to reveal the characteristics of the AE signal time domain or frequency domain of it’s stretch damage modes and establish the corresponding relation between damage mode and AE signal,through three groups of the three-dimensional braided composite materials experimental specimen of the same braiding angle and the different fiber volume content to conduct stretch experiment,and then to obtain the mechanical properties and AE signals at the different damage stages.In order to study the time-frequency characteristics of the AE signals,using the HHT time-frequency analysis method to isolate and extract the damage pattern information,it can obtain the relation between the damage mode of the three-dimensional braided composite materials and the AE signal,which provides a new signal analysis method for studying and analyzing the stretch fracture mechanism of the three-dimensional braided composite materials.Finally,according to the time-frequency characteristics of the AE signal HHT of the stretch damage model of the three-dimensional braided composite materials,on the basis of determining the damage stage and signal type,the two step method is used to locate the damage source.The first step combines the time frequency analysis results and the anisotropy of the experimental specimens to compensate the sound velocity,and uses the four point arc method to locate the initial position.The second uses the probabilistic neural network and the concept of crowding in chaos search to improve the algorithm,which will optimize the smoothing parameter of probabilistic neural network,reduce the number of iteration and improve the positioning accuracy.