Quantitative Nondestructive Evaluation with Ultrasonic Method Using Neural Networks and Computational Mechanics. Identification of Inclined Defect and Verification of Robustness.

This paper describes an application of a hierarchical neural network to defect identification with the ultrasonic method. The present method consists of three subprocesses. First, sample data of identification parameters vs. dynamic responses of displacements at several monitoring points on the surface are calculated using the dynamic finite element method. Second, a back-propagation neural network is trained using the sample data. Finally, the well-trained network is utilized for defect identification. This method is applied to the identification of location, length and inclination of a defect hidden in solid. Its performance and robustness are quantitatively discussed in detail.