RECONSTRUCTION OF DEFECT SIZE AND SHAPE IN EDDY-CURRENT TESTING USING BENCHMARK PROBLEMS VALIDATION AND NEURAL NETWORK APPROACH

Authors

  • F. Barrarat Department of Materials Science, Physics of Materials Laboratory, University of Laghouat, Laghouat 03000
  • K. Rayane Department of Mechanical Engineering, Process Engineering Laboratory, University of Laghouat, Laghouat 03000
  • B. Helifa Department of Materials Science, Physics of Materials Laboratory, University of Laghouat, Laghouat 03000
  • B. Chettouh Department of Materials Science, Physics of Materials Laboratory, University of Laghouat, Laghouat 03000
  • I.K. Lefkaier Department of Materials Science, Physics of Materials Laboratory, University of Laghouat, Laghouat 03000

DOI:

https://doi.org/10.4314/jfas.v12i1S.7

Keywords:

Eddy-current testing; benchmark problems; flaw sizing; finite element method, inverse problem; neural network.

Abstract

The benchmark problems in eddy-current nondestructive evaluation (EC-NDE) are based on careful measures of the change in coil impedance as a function of circular air cored coil position which is scanned along the axis of machined slot by electrodischarge in aluminum plate. Tow benchmark problems (TEAM workshop n° 15-1, and JSAEM n° 2-5) are presented to validate and verify ANSYS Maxwell 3D-resolution of defect size and shape using electromagnetic formulation. In order to provide a challenge for current theoretical models, slots of rectangular, elliptical, slope and triangular profiles are considered. The final impedance data can be directly used to verify theoretical inversion algorithms.

Downloads

Download data is not yet available.

References

[1] Martinos J, Theodoulidis T, Poulakis N, Tamburrino A. IEEE Trans. Mag., 2014, 50(2), 1053-1056, doi : 10.1109/TMAG.2013.2282259.
[2] Ditchburn R.J, Burke S.K, Posada M. Eddy-current nondestructive inspection with thin spiral coils: Long cracks in steel. J. NDE., 2003, 22(2), 63-77
[3] Miorelli R, Reboud C, Theodoulidis T, Poulakis N, Lesselier D. IEEE Trans. Mag., 2013, 49(6), 2886-2892, doi : 10.1109/TMAG.2012.2236102
[4] Burke S.K. A benchmark problem for computation of in eddy-current nondestructive evaluation (NDE). J. Nondestruct. EvaL., 1988, 7, 35-42
[5] Norton S. J, Bowler J. R. J. Appl. Phys., 1993, 73, 501, doi: 10.1063/1.353359
[6] Burke S. K. J. Appl. Phys., 1994, 76, 3072-3080, doi: 10.1063/1.358454
[7] Harrison D. J, Jones L. D, Burke S. K. Benchmark Problems for Defect Size and Shape
Determination in Eddy-Current Nondestructive Evaluation. J. NDE., 1996, 15(1), 21-34
[8] Ida N, Palanismy R, Lord W. Eddy current probe design using finite element analysis. Mater. evalu., 1983, 41(12), 1389-1394
[9] Ida N. Alternative approaches to the numerical calculation of impedance. NDT & E Int., 1988, 21, 27-35, doi: 10.1016/0308-9126(88)90381-1
[10] Burrascano P, Fiori S, Mongiardo M. Int. J. RF and microwave CAE., 1999, 9(3), 158-174, doi: 10.1002/(SICI)1099-047X(199905)9:3<158::AID-MMCE3>3.0.CO;2-V
[11] Fiori S. Neural Networks., 2000, 13(6), 597-611, doi: 10.1016/S0893-6080(00)00044-7
[12] Fiori S. Neurocomputing., 2000, 34(1-4), 239-252, doi: 10.1016/S0925-2312(00)00161-2
[13] Fiori S, Burrascano P, Cardelli E, Faba A. A blind separation approach to electromagnetic source localization and assessment. Proc. of 7th international conference on engineering application of neural networks (EANN’2001), July 16-18, 2001, Cagliari (Italy)
[14] Rekanos I. T, Theodoulidis T. P, Panas S. M, Tsiboukis T. D. NDT&E Int., 1997, 30(2), 69-14, doi: 10.1016/S0963-8695(96)00047-3
[15] Hertz J, Krogh A, Palmer R. G. Introduction to the Theory of Neural Computation, Addison-Wesley, Reading, MA, 1991.
[16] Mann J. M, Schmerr L. W, Moulder J. C. NDT&E Int., 1991, 34-39, doi: 10.1016/0963-8695(92)90071-N
[17] Coccorese E, Martone R, Morabito F. C. IEEE Tran. Mag., 1994, 30(5), 2829-2839, doi: 10.1109/20.312527

Downloads

Published

2019-12-05

How to Cite

BARRARAT, F.; RAYANE, K.; HELIFA, B.; CHETTOUH, B.; LEFKAIER, I. RECONSTRUCTION OF DEFECT SIZE AND SHAPE IN EDDY-CURRENT TESTING USING BENCHMARK PROBLEMS VALIDATION AND NEURAL NETWORK APPROACH. Journal of Fundamental and Applied Sciences, [S. l.], v. 12, n. 1S, p. 78–88, 2019. DOI: 10.4314/jfas.v12i1S.7. Disponível em: https://www.jfas.info/index.php/JFAS/article/view/642. Acesso em: 24 mar. 2023.