Importance of Flux Fringing for 2-D Modeling of Permanent Magnet Axial Flux Motors

Modeling of axial flux permanent magnet synchronous motors (AFPMs) by means of 2-D finite element analysis (FEA) is well-established. Torque and induced voltage can be described with good accuracy as they are dominated by permanent magnet (PM) fluxes which behave mostly 2-D. However, due to 3-D effects, inductances are underestimated by 2-D FEA by 25% to 30% compared to 3-D FEA for the AFPMs considered in this work. This is most notable when dealing with d-currents for field weakening. Here, the authors of this paper find that flux fringing at the inner and outer diameter of the motor is the most relevant contributor to error. With the help of 3-D FEA, an approach based on 2-D FEA is presented for correcting inductances with a flux fringing factor. A flux map model deploying the adjusted inductances is then validated against actual test stand data. This reduces errors in d-current estimations from 40% to 5% for field weakening over the whole field of operation. While the above is based on central rotor AFPMs, the effect of flux fringing is also examined for central stator AFPMs using FEA.