Magnetic Particle Inspection Simulation Model

Abstract

The Magnetic Particle Inspection (MPI) method is generally used to detect surface and near surface flaws. In a MPI test, a ferromagnetic specimen in unmagnetized state is sprayed with magnetic particles in an aerosol suspension. The particles are generally ferromagnetic oxides coated with fluorescent pigments and the suspension is a petroleum distillate of low viscosity. The specimen is then magnetized. In the presence of a flaw transverse to the direction of the applied magnetic field, leakage fields are established on the surface of the specimen. These fields exert a translational force on the ferromagnetic particles in addition to a rotational torque. These two in combination accelerate the particles toward the flaw increasing the density of particles in the vicinity of the flaw. When excited by ultraviolet light, the particles emit visible radiation indicating the location of the flaw. This paper attempts to model the physical principles underlying the MPI method including imaging techniques to recreate the dynamics of the particles prior to their reaching an equilibrium around the flaw. The proposed approach has the capability to predict the time to equilibrium of the magnetic particles and the efficiency of the method in terms of the fraction of the total number of particles in the MPI image.KeywordsMagnetic ParticleGray Level ImageFerromagnetic ParticleFlaw TransverseRotational TorqueThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.