Reconstructions of Inner and Outer Defects in Ferromagnetic Materials from Experimental Remanent Magnetic Measurements by using Neural Networks

摘要

The present paper presents the application of remanent field analysis for detecting defects in ferromagnetic materials. The remanent signal from outer and inner defects in magnetized ferromagnetic samples is computed using a FEM-BEM nonlinear code. Also, an experiment is set-up to measure the remanent field around defects. The parameterized defect shape is reconstructed using Neural Networks. Numerical results for the inversion are presented.