Demagnetization Fault Detection in Axial Flux PM Machines by Using Sensing Coils and an Analytical Model

The paper studies demagnetization detection in axial flux permanent magnet synchronous machines of the yokeless and segmented armature (YASA) type. In YASA machines, asymmetrical defects can occur: the demagnetization defects in the permanent magnets of both rotors differ from each other. Many techniques can indicate that there is a defect, even if it is an asymmetrical defect, but cannot detect which rotor has the defect, i.e., exactly locate the asymmetrical defect. In this paper, a method is proposed to detect and locate asymmetrical demagnetization defects in AFPMSMs. The method is useful for real-time condition monitoring of demagnetization defects, or as virtual temperature sensor for the magnets. A sensing coil is placed in each air gap. An analytical model is constructed, which uses the demagnetization defect and the three-phase current as input and calculates the sensing coil back-EMF. Sensing coils back-EMF data are combined with the data from the analytical model to solve an electromagnetic inverse problem. The results are the demagnetization amount and location of the demagnetization defect. The method is validated with the experimental data.