Article Article
Detection and Classification of Discontinuities using Discrete Wavelet Transform and MFL Testing

Damage caused by the degradation of steam generator tubes (SGTs), used in sodium-cooled reactors, leads to leakage of water or steam into the sodium coolant, which may further lead to a high-temperature reaction. This paper explains how to detect and classify the discontinuities that occur in the outer surface of SGTs using the magnetic flux leakage (MFL) method. Generally, SGTs experience stress corrosion cracking, wear, and mechanical damage, which can lead to catastrophic failures. MFL is one of the nondestructive testing methods that can detect a localized discontinuity on the outer surface of a SGT. The leakage flux detected by the magnetic sensor is used to extract information about the discontinuity. A multiphysics software package is used to obtain simulated MFL images showing discontinuities such as cracks, flat-bottom holes, and rectangular notches. Statistical features are extracted to detect the discontinuities in the images, and discrete wavelet transform (DWT) features are extracted to classify the discontinuities. Feed-forward back propagation training and testing of neural networks is used to classify the discontinuities. Using these techniques, a 100% discontinuity detection rate is achieved, and an average classification accuracy of 94.62% is obtained through the proposed algorithm.


Carvalho, A.A., J.M.A. Rebello, L.V.S. Sagrilo, C.S. Camerini, and I.V.J. Miranda, 2006, “MFL Signals and Artificial Neural Networks Applied to Detection and Classification of Pipe Weld Defects,” NDT&E International, Vol. 39, No. 8, pp. 661–667.

de Mesquita, Roberto N., Daniel K.S. Ting, Eduardo L.L. Cabral, and Belle R. Upadhyaya, 2004, “Classification of Steam Generator Tube Defects for Real-Time Applications Using Eddy Current Test Data and Self-Organizing Maps,” Real-Time System, Vol. 27, No. 1, pp. 49–70.

de Mesquita, Roberto Navarro, Daniel Kao Sun Ting, Eduardo Lobo L. Cabral, Luis A. Negro M. Lopez, and Belle R. Upadhyaya, 2002, “Development of a System for Moni-toring and Diagnosis of Steam Generator Tubes using Artifi-cial Intelligence Techniques on Eddy Current Test Signals,” INAC 2002: International Nuclear Atlantic Conference; 13.

Brazilian National Meeting on Reactor Physics and Thermal Hydraulics; 6. Brazilian National Meeting on Nuclear Appli-cations, Rio de Janeiro, Brazil.

Dutta, Sushant M., Fathi H. Ghorbel, and Roderic K. Stanley, 2009, “Simulation and Analysis of 3-D Magnetic Flux Leakage,” IEEE Transactions on Magnetics, Vol. 45, No. 4, pp. 1966–1972.

Huda, A.S.N., S. Taib, K.H. Ghazali, and M.S. Jadin, 2014, “A New Thermographic NDT for Condition Monitoring of Electrical Components using ANN with Confidence Level Analysis,” ISA Transactions, Vol. 53, No. 3, pp. 717–724.

Hur, Do Haeng, Myung Sik Choi, Hee-Sang Shim, Deok Hyun Lee, and One Yoo, 2014, “Influence of Signal-to-Noise Ratio on Eddy Current Signals of Cracks in Steam Generator Tubes,” Nuclear Engineering and Technology, Vol. 46, No. 6, pp. 883–888.

Ji, Fengzhu, Changlong Wang, Shiyu Sun, and Weiguo Wang, 2009, “Application of 3-D FEM in the Simulation Analysis for MFL Signals,” Insight, Vol. 51, No. 1, pp. 32–35.

Kang, Yihua, Jianbo Wu, and Yanhua Sun, 2012, “The Use of Magnetic Flux Leakage Testing Method and Apparatus for Steel Pipe,” Materials Evaluation, Vol. 70, No. 7, pp. 821–827.

Khodayari-Rostamabad, Ahmad, James P. Reilly, Natalia K. Nikolova, James R. Hare, and Sabir Pasha, 2009, “Machine Learning Techniques for the Analysis of Magnetic Flux Leakage Images in Pipeline Inspection,” IEEE Transaction on Magnetics, Vol. 45, No. 8, pp. 3073–3084.

Liu, Shudong, Chunling Du, Jianqiang Mou, Landong Martu, Jingliang Zhang, and F.L. Lewis, 2013, “Diagnosis of Struc-tural Cracks using Wavelet Transform and Neural Networks,” NDT&E International, Vol. 54, pp. 9–18. 

Papaelias, Mayorkinos, Liang Cheng, Maria Kogia, Abbas Mohimi, Vassilios Kappatos, Cem Selcuk, Louis Constan-tinou, Carlos Quiterio Gómez Muñoz, Fausto Pedro Garcia Marquez, and Tat-Hean Gan, 2016, “Inspection and Struc-tural Health Monitoring Techniques for Concentrated Solar Power Plants,” Renewable Energy, Vol. 85, pp. 1178–1191.

Postolache, O., H. Geirinhas Ramos, and A. Lopes Ribeiro, 2011, “Detection and Characterization of Defects using GMR Probes and Artificial Neural Networks,” Computer Standards & Interfaces, Vol. 33, No. 2, pp. 191–200.

Qiu, Xuan Bing, Lu Lu Liu, Chuan Liang Li, Ji Lin Wei, Ying Fa Wu, and Xiao Chao Cui, 2014, “Defect Classification by Pulsed Eddy-Current Technique Based on Power Spectral Density Analysis Combined with Wavelet Transform,” IEEE Transactions on Magnetics, Vol. 50, No. 9, pp. 1–8.

Sambath, S., P. Nagaraj, and N. Selvakumar, 2010, “Auto-matic Defect Classification in Ultrasonic NDT Using Artificial Intelligence,” Journal of Nondestructive Evaluation, Vol. 30, No. 1, pp. 10–28. 

Shafeek, H.I., E.S. Gadelmawla, A.A. Abdel-Shafy, and I.M. Elewa, 2004, “Automatic Inspection of Gas Pipeline Welding Defects using an Expert Vision System,” NDT&E International, Vol. 37, No. 4, pp. 301–307.

Xu, Z.H., X.M. Zha, H.G. Chen, Y.H. Sun, and M.Z. Long, 2015, “A Simulation Study for Locating Defects in Tubes using the Weak MFL Signal based on the Multi-channel Correlation Technique,” Insight, Vol. 57, No. 9, pp. 518–527.

Yang, Peng, and Qiufeng Li, 2012, “Wavelet Transform-based Feature Extraction for Ultrasonic Flaw Signal Classifi-cation,” Neural Computing Applications, Vol. 24, Nos. 3–4, pp. 817–826.

Yoon, Byungsik, Yongsik Kim, and Seunghan Yang, 2014, “An Experimental Result on Flaw Sizing of Steam Generator Tube at Nuclear Power Plants using Ultrasonic Testing,” Journal of Nuclear Science and Technology, Vol. 49, No. 7, pp. 760–767.

Zhang, Donglai, Min Zhao, Zhihui Zhou, and Shimin Pan, 2013, “Characterization of Wire Rope Defects with Gray Level Co-Occurrence Matrix of Magnetic Flux Leakage Images,” Journal of Nondestructive Evaluation, Vol. 32, No. 1, pp. 37–43.

Zhang, Qi, Tian-lu Chen, Guang Yang, and Li Liu, 2012,“Time and Frequency Domain Feature Fusion for Defect Classification Based on Pulsed Eddy Current NDT,” Research in Nondestructive Evaluation, Vol. 23, No. 3, pp. 171–182.

Usage Shares
Total Views
308 Page Views
Total Shares
0 Tweets
0 PDF Downloads
0 Facebook Shares
Total Usage