A Systematic Method for the Calibration of FEA Model of Synchronous Reluctance Machines Considering Manufacturing Effects

Abstract

The recent advances in computational power and motor design software, supported by finite elements analysis (FEA), permit accurate prediction of the losses and electromagnetic performance of the machine at the design stage. Anyway, the simulation accuracy strictly depends on the real motor geometry and material properties, which are affected by the manufacturing process. This article offers a systematic method for accurately calibrating the electromagnetic FEA models, permitting to consider the production process within the motor design stage. A synchronous reluctance motor prototype is designed, manufactured, and experimentally characterized, and the measured flux and torque characteristics are exploited for ex-post calibrating the FEA model of the machine. This approach led to a reduction in torque estimation error from 18% to 4%, and an improvement of the flux maps evaluation, thus enhancing and validating the design procedure and permitting further optimization steps of machines with similar geometry and dimensions. The effects are particularly evident in the proposed motor due to the reduced size of the rotor flux barriers, stator teeth, and yoke. The mechanical tolerances are analyzed through a Monte Carlo analysis covering different areas of the machine. Then, the effects of uncertain airgap length and iron degradation due to the manufacturing process are considered, determining an equivalent degraded BH curve.