Wave Signal Demodulation and Artificial Neural Networks for Damage Identification

This presentation will discuss the Lamb wave-based damage detection using wave signal demodulation and artificial neural networks. The interaction between Lamb wave and damage will modify the response wave signal from which information related to damage can be extracted for automated damage detection. However, the interpretation of the response wave signal is not easy due to the complex nature of the wave-damage interaction. This paper discusses a damage detection algorithm based on wave signal demodulation and artificial neural networks(ANNs). The response wave signal is considered as a low-frequency signal modulated by a high-frequency carrier signal. After baseline subtraction, frequency domain convolution and filtering, the original signal is demodulated and transformed into a new simplified signal related to the energy change due to damage. Subsequently feature extraction is carried out by finding the local maxima in the new signal and the obtained peak values and locations are used as inputs into the ANNs for damage characterization. The validity of this damage detection algorithm is then verified using a finite element(FE) model of a composite laminate with notch defects. The response wave signals of different notch depths and locations are acquired from the simulations and used as the training and testing samples. Finally the assessment of the networks accuracy and generalization ability is performed and the result is satisfactory.