In this work, classification of cracks in a non-ferromagnetic material (aluminum 1050) is presented using two supervised machine learning techniques called support vector machines (SVM) and decision tree. The classification performances between the two classifiers were evaluated using confusion matrices and miss-classification errors. In order to improve the classification performance, a multi-level SVM model is proposed wherein each level implements SVM using different features and kernels. The experimental tests were performed in two aluminum plates containing 18 defects with different depths and lengths. The 18 defects were inspected in the surface and sub-surface locations, resulting in a study of 36 different crack cases. The experimental data used to extract features for supervised machine learning were obtained using an eddy current testing (ECT) probe that includes a planar coil and a tunnel magnetoresistance (TMR) sensor. The probe was excited by a multi-frequency signal with four testing frequencies to evaluate the dimensions and the locations of the defects. Due to the experimental noise existing in the measurement signals, it is important to pre-process the signals before feature extraction. Signal processing techniques (such as moving average filter, spline interpolation and symmetry corrections) were used before extracting the features for classification.

Detection of corrosion and measurement of the thickness of stainless steel in industrial structures, such as pipelines, has been an important issue to avoid unwanted leakages. A recently proposed transient eddy-current oscillations (TECO) method has been used in this paper to measure thickness in non-magnetic stainless steel. The TECO method works based on the electromagnetic induction principle, as the conventional eddy-current testing method. An inductor (excitation coil) connected in parallel with a capacitor to generate damped oscillations is used to induce eddy currents into the test material beneath the coil. A Hall sensor, which is fixed at the bottom side and at the axial center of the excitation coil, detects the resultant magnetic field due to the induced eddy currents. Experiments were performed on the stainless steel specimens, having different thicknesses to test the potential of the TECO method. The results were analyzed using the fast Fourier transform. This paper proposes a lift-off point of intersection in the frequency domain that can be used to eliminate the false indications due to the lift-off variations during the thickness measurement of the test material. The results reveal that the proposed method can be applied to detect the thickness variations in metal structures covered with non-conductive paints or insulation coatings.

Steel Plate Cold-Rolled Commercial (SPCC) steel is known to have long-term durability. However, it still undergoes corrosion when exposed to corrosive environments. This paper proposes an evaluation method for assessing the corrosion level of SPCC steel samples using eddy current testing (ECT), along with two different machine learning approaches. The objective is to classify the corrosion of the samples into two states: a less corroded state (state-1) and a highly corroded state (state-2). Generative and discriminative models were implemented for classification. The generative classifier was based on the Gaussian mixture model (GMM), while the discriminative model was based on the logistic regression model. The features used in the classification models are the peaks of the perturbated magnetic fields at two different frequencies. The performance of the classifiers was evaluated using metrics such as absolute error, accuracy, precision, recall, and F1 score. The results indicate that the GMM model is more conducive to categorizing states with higher levels of corrosion, while the logistic regression model is helpful in estimating states with lower levels of corrosion. Meanwhile, high classification accuracy can be achieved based on both methods using eddy current testing.

Detecting deeply buried subsurface defect is a challenging task for eddy current testing (ECT). This study introduces a novel ECT probe designed for the detection of deeply buried defects in conductive materials. The key innovation of this probe lies in the utilization of two identical rectangular coils coupled with a high-sensitivity Hall sensor. The configuration of the coils generates maximum magnitude of eddy current while minimizing the background magnetic field perpendicular to the specimen surface, ideally achieving zero magnetic field in the center region of probe. This design allows the probe to effectively leverage magnetic sensors with high spatial resolution and sensitivity, enabling the detection of weak magnetic field disturbances caused by subsurface defects. To optimize the probe parameters, as well as maximizing the distribution of eddy currents in the plate, a theoretical model based on the second order vector potential (SOVP) formalism was derived. Moreover, FEM simulations and experiments were conducted to evaluate the detection capabilities of the optimized probe. The experimental results demonstrate that the ECT probe is capable of reliably detecting subsurface defects, when a defect with the size of 20×1×1 mm³ locates 9 mm below the surface of a 10 mm thick stainless steel specimen.

This work proposes a model based on the semianalytical boundary element method (BEM) to predict the perturbation of the voltage induced in an eddy current coil due to narrow opening flaws that are usually encountered in eddy current nondestructive tests. The modeling is performed for rectangular narrow opening flaws that are located on the surface of an aluminum plate of finite thickness. The excitation to the coil varies time harmonically and the BEM results are corroborated by experiments performed in a controlled environment in the laboratory.