Study, Design and Realization of GMR Probe for Eddy Current Non-Destructive Testing

Abstract

The aim of this thesis is to develop strategy of elaborating software and hardware tools for design and construction of high-sensitivity magnetic sensors (GMR probe) in eddy current non-destructive testing in its experimental and simulation aspects for the characterization of cracks in "critical" parts of conducting materials. In this context and to place the work in a more general framework. Two probes realize during this thesis simple and differential GMR probes. We propose a sample GMR probe and we show that the giant magnetoresistance (GMR) -based eddy currents probe is more sensitive than the inductive probe with a difference of 80 %. We propose a new design and implementation of a high-sensitivity eddy current (EC) sensor based on GMR to assess cracks in conductive materials. This approach's originality uses two symmetrical Giant magnetoresistance sensors in a differential configuration using commercial GMR elements inserted on a coil in a ferrite pot. The background signal measured by the sensor is infinitesimal if there is no crack in the sample. Therefore, the designed sensor demonstrates a high sensitivity to the presence of defects where the GMRs mounted in differential allow for reducing the background voltage's impact. On the other hand, The GMR-based EC probe with a ferrite pot core is more sensitive to the presence of cracks than the conventional EC sensor without a ferrite pot core. This work introduces the notion of the GMR sensor's effective area (EA) after being calculated and optimized using the inverse problem (Particle Swarm Optimization method). The operation of the differential GMR sensor is validated using a 3D Finite Element Model based on the (A, V–A) formulation and experimental measurements. Finally, the prototype of the differential GMR sensor is developed and tested. The experimental results are obtained to evaluate cracks machined on an aluminum standard. We validate (direct and reverse) by comparing the database comprising the tension of the GMR probe-crack (crack signature). We use the artificial neural network's method for the inversion of data in the NDT-EC differential GMR probe to detect and size the different defects in conductive materials while offering the advantages of high sensitivity and excellent reliability.