Approximating the distribution of flaws in magnetic materials using the generalized inverse

Non-destructive identification of voids in ferromagnetic materials is of great importance for industrial applications. Magnetic flux leakage technique is used here to examine the defected structure. To this end, an inverse problem should be solved in order to infer the location and depth of internal flaws from the measured leaked magnetic signals. Currently generalized inverse method and singular value decomposition are used for solving such inverse problem. Considering the cracks separation has significant effect on the absolute value of magnetic flux leakage signals, we study different distributions of cracks. In this paper, the magnetic dipole model is proposed to reconstruct the extent of the flaws. Comparing the reconstructed results with simulated data, the accuracy of proposed method for different distributions of cracks is shown.