Magnetic flux leakage testing based on the dual testing probes resisting the negative effects of lift-off perturbations

Abstract

The traditional defect evaluation method for magnetic leakage flux (MFL) testing is based on the calibration relationship of signal characteristics on a given lift-off value. The unavoidable lift-off perturbations during the testing process cause changes in the collected MFL signals, leading to challenges in the quantitative evaluation of defect sizes by the traditional method. To address this, we propose an MFL testing system with dual testing probes to test signals containing lift-off information, and develop related algorithms with the Regulargridinterpolator function and particle swarm optimization method to achieve joint high-precision evaluation for radius and depth of defect in the presence of lift-off perturbations. Unlike the single testing probe used in the traditional evaluation method, stacked dual testing probes with a given lift-off difference and a self-developed experimental system for the MFL testing are prepared and optimized, realizing the jointly automatic and continuous acquisition of two sets of signals with a given lift-off difference. Then, the testing experiment of Q235 steel plate containing surface round hole defects (defect radius and depth are 2 mm ∼ 5 mm and 1 mm ∼ 5 mm) is executed, and the experimental results combined with the magnetic dipole model show that both sets of signals are positively correlated with the defect size and closely related to the lift-off value. In addition, the proposed defect evaluation algorithm is used to jointly quantitatively evaluate the radius and depth of surface round hole defects based on the collected MFL testing signals, where the prediction error of defect size based on dual testing probes is controlled within 0.4 mm, indicating that the evaluation capability is significantly higher than that from the traditional evaluation method based on the single testing probe in the presence of lift-off perturbations. The proposed method is demonstrated to resist the negative effects of lift-off perturbations, providing a pathway for the quantitative evaluation of defects under complex testing conditions.