Accelerated Multi-Criteria Design Optimization of Axial Flux Motors with 3D FEA
Electromagnetics
PRESENTER:
Dr. Hossain Mohammadi, SCHAEFFLER
EXPERTISE:
Electromagnetics
Software:
JMAG
Axial flux motors (AFMs) offer compelling advantages in power density and efficiency for next-generation electric drivetrains, but their design space is highly complex due to strong geometric, electromagnetic, thermal, and control-related couplings. This presentation introduces a computationally efficient, data-driven methodology for accelerated multi-criteria design optimization of axial flux motors using 3D finite element analysis (FEA) in JMAG Designer.
The proposed workflow integrates coarse and fine electromagnetic models, extensive parametric sampling, and automated scripting via MATLAB and Python to evaluate thousands of design candidates under realistic operating constraints. A multi-stage optimization process is employed, combining global design space exploration with localized refinement, surrogate modeling, and Pareto-front analysis to balance conflicting objectives such as torque and power density, efficiency, material cost, losses, and manufacturability.
A detailed implementation is demonstrated for a 3D AFM topology with fourteen independent design variables, leveraging Latin Hypercube Sampling and DQ-model-based post-processing to rapidly assess multiple operating points, including peak torque, maximum speed, and partial-load conditions representative of WLTC cycles. The results show close agreement between FEA and DQ-based models, enabling reliable performance prediction with significantly reduced computational effort. The presented approach substantially shortens engineering development cycles while maintaining high fidelity, and it has been successfully applied to both axial and radial flux machines, including IPMSM and PM-assisted synchronous reluctance motors. This scalable framework supports sustainable motor design strategies, including reduced reliance on rare-earth materials, and is open to collaboration for advanced R&D and industrial applications.
