Comparison Analysis of Iron Nitride‐Assisted Synchronous Reluctance Motor With Different Magnet Grades for High‐Speed Traction Applications

ABSTRACT

This paper analyses the performance of high‐speed permanent magnet assisted synchronous reluctance motor (PMaSynRM) with blended magnets for electric traction applications. The blended magnet approach utilises two magnet combinations of NdFeB, dysprosium‐free (Dy‐free) NdFeB and iron nitride (FeN) permanent magnets which reduce the contribution of rare‐earth permanent magnets in the double layer and triple layer rotor topologies. The iron nitride magnets with three different energy products (10, 24 and 36 MGOe) are employed for a systematic comparative design study of the high‐speed traction motor. Multiphysics multiobjective optimisation algorithm (MP‐MODE) based on differential evolution identifies the optimal designs satisfying the defined objectives and constraints based on a baseline traction motor with rare‐earth permanent magnets. The critical design objective of demagnetisation risk is evaluated at extreme temperatures of −20°C and 150°C for both magnet combinations. The lower energy product variants of the rare earth free magnets increase the overall magnet volume which poses challenges to the rotor design at high speed. Hence, ensuring the mechanical integrity of the rotor is necessary, and the centrifugal mechanical stress is treated as a true co‐optimisation objective within the MP‐MODE framework, ensuring all Pareto‐optimal designs are simultaneously electromagnetically and mechanically feasible. The electromagnetic and mechanical performance of optimal double layer and triple layer iron nitride‐assisted designs for both magnet combinations are compared to the baseline traction motor for validation. Results demonstrate rare‐earth permanent magnet reductions exceeding 65% with 36 MGOe FeN and up to 24% reduction even with the currently available 10 MGOe grade, with demagnetisation risk below 5% and mechanical stress within the lamination yield strength limit across all evaluated designs.