A 3-D Forward Model of Metal Magnetic Memory Method for Quantitative Defect Detection

Abstract

Metal magnetic memory (MMM) has gained attention in both academia and industry as a nondestructive testing (NDT) method capable of achieving early diagnosis of ferromagnetic materials and structures under various stress states and damage forms. The quantitative MMM detection, however, has been limited in engineering applications due to the imprecision of the quantitative relationship between stress-magnetization-defect size, the poor similarity between the simplified geometries of the characterization model and the actual defects, as well as the inability of the characterization signals to accurately reflect the spatial distribution of the magnetic field at the defects. In this article, in order to be consistent with engineering defects, the ellipsoid defect is adopted as the research object. Based on the combination of hysteresis characteristics, demagnetization energy, and finite element method, the quantitative relationship between stress vector, magnetic charge density, and defect size is derived, and an MMM detection forward model for defects containing an inhomogeneous distribution of stresses is established. The results of the experiments show that the predictions of the model agree with the experimental data both qualitatively and quantitatively. On this basis, the spatial distribution characteristics of the 3-D signal components at the defect were analyzed, and the advantages of the previously unnoticed X-directional component in determining the shape and size of the defect were revealed. Meanwhile, the influence laws of dimensional parameters, lift-off height, stress, and burial depth on the MMM signals are discussed in detail, which provides a theoretical basis for the quantitative identification of defects in practical engineering.