Yann Le Bihan

Study on the transformer equivalent circuit of eddy current nondestructive evaluation

The modeling of a practical configuration of nondestructive evaluation (NDE) by eddy currents (EC) generally requires extended analytical or numerical developments. This is a drawback to any understanding of the EC NDE principle and of its basic features. Nevertheless, it is possible to exploit a simple equivalent circuit based on an analogy with the operation of an electrical transformer. In this model the EC sensor constitutes the primary of the transformer whereas the evaluated conductive plate forms the transformer secondary and its electrical load. In this paper, thanks to the transformer analogy, the author obtains a simple analytical expression of the impedance of the EC sensor by using a distributed constant representation of the evaluated plate. This expression allows to exhibit some fundamental features of the EC NDE method.

Pebbles Tracking Thanks to RFID LF Multi-Loops Inductively Coupled Reader

The Radio Frequency IDentification (RFID) Low Frequency (LF) serial loops structure is proposed to improve TAGs detection when a TAG coil-antenna rotates by any angle, due to the tagged pebble moving. The detection zones of two types of TAGs (the token and glass TAG) and two types of reader coils, in function of the TAG size, TAG orientation and shape of the reader coils are tested. The effect of the proposed multi-coil inductively coupled is confirmed by measurement using a commercial LF RFID system.

Microwave Characterization Using Least-Square Support Vector Machines

This paper presents the use of the least-square support vector machines (LS-SVM) technique, combined with the finite element method (FEM), to evaluate the microwave properties of dielectric materials. The LS-SVM is a statistical learning method that has good generalization capability and learning performance. The FEM is used to create the data set required to train the LS-SVM. The performance of LS-SVM model depends on a careful setting of its associated hyper-parameters. Different tuning techniques for optimizing the LS-SVM hyper-parameters are studied: cross validation (CV), genetic algorithms (GA), heuristic approach, and Bayesian regularization (BR). Results show that BR provides a good compromise between accuracy and computational cost.

MICROWAVE CHARACTERIZATION OF DIELECTRIC MATERIALS USING BAYESIAN NEURAL NETWORKS

This paper shows the efficiency of neural networks (NN), coupled with the finite element method (FEM), to evaluate the broad- band properties of dielectric materials. A characterization protocol is built to characterize dielectric materials and NN are used in order to provide the estimated permittivity. The FEM is used to create the data set required to train the NN. A method based on Bayesian regularization ensures a good generalization capability of the NN. It is shown that NN can determine the permittivity of materials with a high accuracy and that the Bayesian regularization greatly simplifies their implementation.

Statistical Approach for Nondestructive Incipient Crack Detection and Characterization Using Kullback-Leibler Divergence

This paper is a contribution to the detection and characterisation of small cracks using Eddy Current Testing in the Non Destructive Evaluation framework. Small cracks are considered as incipient faults defined as gradual faults whose signature is weak and concealed by the noise. They are characterized by high signal to noise ratio and low fault to noise ratio. The detection and diagnosis of such faults is still an open challenge. For complex systems, model-based incipient fault detection and diagnosis (FDD) methods usually fail because of the inaccuracy of the model to describe all the phenomena and their interactions. Data-driven methods using statistical features are very promising as long as historical data are available. However in the case of incipient faults, there is not a significant variation of a single feature. The fault signature lies in the global variation of the signal properties. The proposed method relies on the Kullback-Leibler Divergence (KLD) as a nonparametric fault indicator. It measures the slight dissimilarities between the probability density functions of the current signal compared to the faultless or healthy one. Through experimental results, the KLD exhibits a higher sensitivity than the usual statistical features for the detection of small cracks (with dimensions in the order of 0.1 mm) realized in a nickel-based superalloy plate. Moreover, the detection is done with zero missed detection probability. Furthermore, the fault severity is assessed through the characteristics of the crack (surface, length, and depth). In the principal component analysis framework, the analysis of four statistical features (KLD, mean, variance, and maximum) dependency to the excitation frequency allows to discriminating among the cracks.

A model-based method for the characterisation of stress in magnetic materials using eddy current non-destructive evaluation

A precise knowledge of the distribution of internal stresses in materials is key to the prediction of magnetic and mechanical performance and lifetime of many industrial devices. This is the reason why many efforts have been made to develop and enhance the techniques for the non-destructive evaluation of stress. In the case of magnetic materials, the use of eddy current (EC) techniques is a promising pathway to stress evaluation. The principle is based on the significant changes in magnetic permeability of magnetic materials subjected to mechanical stress. These modifications of magnetic permeability affect in turn the signal obtained from an EC probe inspecting the material. From this principle, a numerical tool is proposed in this paper to predict the EC signal obtained from a material subjected to stress. This numerical tool is a combination of a 3D finite element approach with a magneto-mechanical constitutive law describing the effect of stress on the magnetic permeability. The model provides the variations of impedance of an EC probe as a function of stress. An experimental setup in which a magnetic material subjected to a tension stress is inspected using EC techniques is tailored in order to validate the model. A very good agreement is found between experimental and modelling results. For the Iron-Cobalt alloy tested in this study, it is shown that a uniaxial tensile stress can be detected with an error lower than 3 MPa in the range from 0 to 100 MPa.

Mesh Refinement in Eddy Current Testing With Separated T-R Probes

In this paper we are interested in FEM mesh refinement procedure in ECT problems with separated T-R (Transmitter and Receiver probes) probes. Local error estimators used for a posteriori h-type mesh refinement are presented. They are based on the complementarity of the E and H formulations. A new estimator is proposed combining the solutions obtained by feeding alternatively the transmitter and the receiver. This estimator proves to be very accurate compared to classical ones.

A mortar element approach on overlapping non-nested grids

The paper presents a finite element approach involving strongly coupled field approximations on moving non-matching overlapping grids. It generalizes a preliminary version of the mortar element method on overlapping grids (Maday et al., 2003) to the case where the field source can be in the moving domain and the physical parameters of the moving domain can differ from those of the surroundings. This generalization has been developed to face the need of some industrial sectors for an efficient investigation numerical method in eddy current non-destructive testing.

An Overlapping Nonmatching Grid Mortar Element Method for Maxwell's Equations

In this paper, a new finite element mortar approach with moving nonmatching overlapping grids is introduced. The bidirectional transfer of information between the fixed and moving subdomains is realized for each new position of the moving part. Two numerical examples are presented to support the theory: 1) an electrostatic problem with known solution, to state the optimality of the method and 2) an eddy current nondestructive testing configuration, to underline the flexibility and efficiency of the proposed approach.

Improvement of HF RFID Tag Detection With a Distributed Diameter Reader Coil

This letter focuses on 13.56 MHz high-frequency radio frequency identification (RFID) in the case of small tags detection, with an effective area below 1 cm2. In such an identification system, based on load modulation principle, the magnetic coupling coefficient k and quality factor of the RFID reader coil are the key parameters. The main goal of this letter is to improve the detection of small tags over a given surface of 10 × 10 cm2 by modifying the reader coil structure, and consequently the coupling coefficient k. Several coil designs are compared experimentally by distributing the diameters of their turns among three possible values. The design of the coils is based on empirical formulas that are in good agreement with experimental measurements. Electromagnetic simulations are performed to confirm the magnetic field distribution of the different designs. The results show that distributed diameter coil (DDC) as RFID reader coil is clearly efficient in this context of the RFID detection. The DDC structures determine the k factor, and, as k is low, the quality factor Q is a second parameter that can improve, in a second step, the RFID detection performances in function of the tag position and orientation.