Title: Automatic detection of defects in ultrasonic testing using artificial neural network

Authors: S. Sambath, P. Nagaraj, N. Selvakumar, S. Arunachalam, T. Page

Addresses: Department of Mechanical Engineering, Mepco Schlenk Engineering College, Sivakasi – 626005, Tamilnadu, India. ' Department of Mechanical Engineering, Mepco Schlenk Engineering College, Sivakasi – 626005, Tamilnadu, India. ' Department of Mechanical Engineering, Mepco Schlenk Engineering College, Sivakasi – 626005, Tamilnadu, India. ' School of Computing and Technology, University of East London, London E16 2RD, UK. ' Department of Design and Technology, Loughborough University, Loughborough, Leicestershire, LE11 3TU, UK

Abstract: An artificial neural network with signal processing technique is proposed to improve the sensibility of flaw detection and to classify defects in ultrasonic testing. Features for discrimination of detected echoes are extracted in discrete wavelet representation and are then classified using ANN. The inputs of the ANN are the features extracted from each ultrasonic oscillogram. Two different types of defect are initially considered namely crack and porosity. The training of the ANN uses supervised learning mechanism and therefore each input has the respective desired output. The available dataset is randomly split into a training subset (to update the weight values) and a validation subset. With the wavelet features and ANN, good classification at the rate of 96% is obtained. According to the results, the algorithms developed and applied to ultrasonic signals are highly reliable and precise for online quality monitoring.

Keywords: nondestructive testing; NDT; ultrasonic testing; defects classification; wavelet transform; artificial neural networks; ANNs; signal processing; flaw detection; defect detection; supervised learning; cracks; crack propagation; porosity; condition monitoring; quality assurance.

DOI: 10.1504/IJMMP.2010.038155

International Journal of Microstructure and Materials Properties, 2010 Vol.5 No.6, pp.561 - 574

Published online: 19 Jan 2011 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article