This work involves monitoring and non-destructive evaluation by inspecting two separate pieces (conductor and ferromagnetic). A database is established from the resolution of the direct problem. The inverse problem for the reconstitution of supposed single and multiple defects in the plates is then solved and analyzed. The study of the direct problem was made by exploiting the finite element method in resolution of the 2D magnetodynamic electromagnetic equation. Experimental Results are then performed and compared to simulation ones. The neural network technique was used successfully for the reconstruction of a single and multiple defects
Słowa kluczowe: Non destructive testing technique, Ferromagnetic materials, Eddy current, Finite element method, Artificial neural network.
W pracy zaprezentowano monitorowanie i nieniszczące badanie dwóch elementów – przewodzącego i ferromagnetycznego. Rozwiązanie problemu odwrotnego wykorzystano do rekonstrukcji defektu. Wykorzystano też metodę elementów skończonych do rozwiązania równań elektromagnetycznych 2D. Wykorzystano też sieć neuronową do rekonstrukcji kształtu pojedynczego i wielokrotnego defektu.
Keywords: badania nieniszczące, metoda prądów wirowych, metoda elementów skończonych,. Sieć neuronowa.
There is nowadays an increasing attention and request towards the study and development of systems and techniques for increasing the human safety and security in all aspects of the everyday life. In this framework, the possibility to perform a non-invasive inspection of structures or objects by means of non-destructive testing and evaluation (NDT-NDE) technologies is of particular interests in several applicative fields, ranging from civil engineering to biomedicine, up to aeronautic and nuclear industries . In non-destructive test methods, eddy current technology increasingly gains in importance. The reason for this is that it is a test method that is flexible in its application and that its probes, usually coils of copper wire, are very adaptable and can be produced fairly easily . Eddy current non-destructive evaluation (NDE) is a method based on the fact that when a coil powered by a variable energy source is brought near a conductive part, a change in impedance at the terminals of the coil is driven by the changing of magnetic field lines due to the existence and the distribution of induced currents in the conductive part . Besides the inspection of metallic samples for which ECT has been applied for some decades to detect cracks, voids, corrosion and other defect typologies structures [4, 5], this technique can also be used to distinguish between ferromagnetic and non-ferromagnetic materials. However, this technique does not allow access to accurate information that can characterize a defect in any form; this is why it is essential to dispose for a rapid tool for signal inversion as eddy current. For this, we are interested in an artificial neural networks. This technique is capable of solving complex problems using an artificial reasoning system constructed with basis on the human brain . A calculation tool developed under Matlab environment was used. Simulation of direct problem with FEM The relevant c [...]
 Grimberg R., Electromagnetic Nondestructive Evaluation: Present and Future, SV-JME, 57( 2011), No. 3, 204-217.  Janousek L., Capova K., Yusa N., Miya K., Multiprobe Inspection for Enhancing Sizing Ability in Eddy Current Nondestructive Testing, IEEE Trans. Magn., 44 (2008), No. 6, 1618-1621.  Hamel A., Mohellebi H., Féliachi M., Imperialist Competitive Algorithm and Particle Swarm Optimization Comparison for Eddy Current Non-destructive Evaluation, PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), R. 88 NR 9a/2012, ISSN 0033-2097.  Marco R., Giuseppe S., Luigi F., Marco L., Ibukun D-A., Gui YT., Evaluation of the Lift-off Robustness of Eddy Current Imaging Techniques, NDT & E International., http://dx.doi.org/10.1016/j.ndteint.2016.10.001  García-Martín J., Gómez-Gil J., Vázquez-Sánchez E., Nondestructive Techniques Based on Eddy Current Testing, Sensors, 11 (2011), 2525-2565  Lorenzi A., Silva Filho L-C-P-da., Artificial Neural Networks Methods to Analysis of Ultrasonic Testing in Concrete, the e- Journal of Nondestructive Testing. 20 (2015), No. 11, ISSN 1435-4934  Ayad A., Benhamida F., Bendaoud A., Le Bihan Y., Bensetti M., Solution of Inverse Problems in Electromagnetic NDT Using Neural Networks, PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), R. 87 NR 9a(2011), ISSN 0033-2097  Haoyu H., Toshiyuki T., Tetsuya U., Fast Numerical Calculation for Crack Modeling in Eddy Current Testing of ferromagnetic materials, Journal of Applied Physics, 94(2003)., No. 9, 5866- 5872  Mingye Y., Satish U., Shreekanth M., Yushi S., Paul S., William L., Solution of Inverse Problems in Electromagnetic NDE Using Finite Element Methods, IEEE Transactions on Magnetics, 34(1998)., No. 5, 2924-2927  Thomas J-L., Modeling Simplified by Control by Eddy Current of Type of Generator of Steam report of internship ESA IGELEC, 1998, University of Nantes, France  Francesca C., Alessandra F., Augusto M., Pietro T., Mariangela U., Artificial Neural Networks for Non-Destructive Evaluation with Ultrasonic Waves in not Accessible Pipes, IEEE, 6(2005), No. 05, 685-692  Martin T., Hagan., Mohammad B., Menhaj., Training Feedforward Networks with the Marquardt Algorithm, IEEE, Trans, Neural Networks, 5(1994), No. 6, 989-993  Lande P-S., Gadewar A-S., October 2012. Application of Artificial Neural Networks in Prediction of Compressive Strength of Concrete by Using Ultrasonic Pulse Velocities, IOSR Journal of Mechanical and Civil Engineering (IOSRJMCE), 1 (2012) No. 3 ISSN: 2278-1684, PP 34-42