Abstract

Time of Flight Diffraction Technique is one of the NDE methods, used in weld inspection to identify the weld defects. The classi?cation of defects using the TOFD technique depends on the knowledge and experience of the operator. The classi?cation reliability of defects detected by this technique can be improved by applying the Artificial Neural Network. In this work, four austenitic stainless steel weldments with defects viz, Lack of Fusion, Lack of Penetration, Slag, Porosity and one with out any Defect were fabricated. TOFD experiment is conducted on these weldments. Discrete wavelet transform based denoising methods were applied to denoise the resultant A scan signals. Time scale features are extracted from the denoised signals. A multi layer feed forward network with Resilient Back Propagation algorithm has been applied for classification of the signals. The number of hidden layers in the network are increased from 0 to 6. Various performance functions are also employed to achieve a better classification efficiency. The results are promising to proceed the automatic defect classification by TOFD technique.

References

• Elineudo P. Moura,Romeu R. Silva, and Joao Marcos A. Rebello Marcio H. S. Siqueira, Pattern recognition of Weld Defects in Preprocessed TOFD Signals Using Linear Classi?ers, Journal of Nondestructive Evaluation, Vol. 23, No. 4, December 2004
• Shyamal Mondal,Dr. T. Sattar ,An Overview of ToFD method and its mathematical model, NDT. net-April 2000, Vol 5. No. 4
• S. Haykin:Neural Networks, A Comprehensive Foundation. 1994, Macmillian College Publishing
• S. Sambath , P. Nagaraj, N. Selvakumar: Automatic Defect Classification in Ultrasonic NDT Using Artificial Intelligence. Journal of Nondestrucive Evaluation (2011) 30: 20–28
• E P de Moura, M H S Siqueira, R R da Silva, J M A Rebello and L P Calôba: Welding defect pattern recognition in TOFD signal Part 1. Linear classifier. Insight Vol 47 No 12 December 2005
• J. L. B. C. Veiga, A. A. de Carvalho,I. C. da Silva,and J. M. A. Rebello:The Use of Artificial Neural Network in the Classification of Pulse-Echo and TOFD Ultra-Sonic Signals. October-December 2005 ABC394 M
• Shaun W. Lawson and Graham A. Parker: Automatic detection of defects in industrial ultrasound images using a neural network.
• C'Shekhar N Shitole, O Zahran and W Al-Nuaimy, Combining fuzzy logic and neural networks in classification of weld defects using ultrasonic time-of-flight diffraction, NDT 2006, the 45th Annual British Conference on NDT, Stratford-upon-Avon, UK, September 2006
• Sergio Roa and Fernando: Classification of Natural Language Sentences using Neural Networks, FLAIRS 2003
• YuwenCao, HaijiangZhu, PingYang: Ultrasonic Time of Flight Diffraction Defect Recognition Based on Edge Detection, Biomedical Engineering and Computer Science (ICBECS), 2010 International Conference, 23-25 April, 2010 Wuhan, China
• S. LalithaKumari, B. SheelaRani, B. Venkatraman, Wavelet Transform based De-noising of ToFD signals of austenitic stainless steel welds, CiiT International Journal of Digital Signal Processing, Oct 2011
• M. S. Obaidat,M. A. Suhail and B. Sadoun : An intelligent simulation methodology to characterize defects in materials. Information Sciences Volume 137, Issues 1-4, September 2001,pp33-41
• Amitava Roy, P. Barat and Swapan Kumar De :Material classification through neural networks. Ultrasonics ,Volume 33, Issue 3, May 1995, pp 175-180
• Karray and de Silva : Soft Computing and Intelligent Systems Design. Addison Wesley September 27, 2004
• Martin Riedmiller , Heinrich Braun:A Direct Adaptive Method for Faster Backpropagation Learning, The RPROP Algorithm. (1993) IEEE International Conference On Neural Networks