Optimal Transducer Placement for Deep Learning-Based Non-Destructive Evaluation

Abstract

In this study, the Convolution Neural Network (CNN) algorithm is applied for non-destructive evaluation of aluminum panels. A method of classifying the locations of defects is proposed by exciting an aluminum panel to generate ultrasonic Lamb waves, measuring data with a sensor array, and then deep learning the characteristics of 2D imaged, reflected waves from defects. For the purpose of a better performance, the optimal excitation location and sensor locations are investigated. To ensure the robustness of the training model and extract the feature effectively, experimental data are collected by slightly changing the excitation frequency and shifting the location of the defect. The high classification accuracy for each defect location can be achieved. It is found that the proposed algorithm is also successfully applied even when a bar is attached to the panel.

Automated Debonding Characterization in Reinforced Structures Based on Finite Element Analysis and Convolutional Neural Networks

“…Subsequently, the scattering response of the Lamb wave passing through the structural defect is measured with the lead zirconate titanate (PZT) sensor array. Measured responses are converted into color bands by assigning different colors according to the amplitude of the response in the time domain and converted into a single image by laying them in the vertical direction [43]. This resultant image is stored in a folder labeled with the respective location of the structural defect, for instance, "debonding location#1", to match the debonding location and its response.…”

“…Virupakshappa and Oruklu [40] achieved high classification accuracy by employing Split Spectrum Processing (SSP) decomposition with A-scan data and a Support Vector Machine (SVM) classifier to detect the presence of holes in a steel block. When a plate structure with a defect is excited, the energy level difference according to the defect's location [41], the frequency domain response of the direct wave passing through the defect [42], the time domain response of the reflected wave from the defect [43], and the damage index [44] were imaged in 2D for training. Then Convolutional Neural Network (CNN) effectively classified the defect's location.…”

Understanding the Role of Sensor Optimisation in Complex Systems

“…In a recent study, Kim et al [ 42 ] discuss utilising the convolution neural network (CNN) algorithm for the NDE of aluminium panels. The objective is to classify the locations of defects by exciting the panel to generate ultrasonic Lamb waves, capturing the data through a sensor array, and then utilising deep learning to identify the features of 2D reflected waves from the defects.…”