Torque modeling of a novel magnetic coupler with multi-disk flux regulation via angular displacement: analytical FEM co-design

Abstract. This paper proposes a novel axially fixed multi-disk magnetic flux-regulatable magnetic coupler with modular topology, which addresses critical limitations such as the structural scalability of conventional magnetic couplers. The innovation integrates three key advancements: (1) a stationary axial architecture that eliminates limitations to torque capacity expansion through concentric modular disk stacking, (2) a flux-regulation mechanism enabled by tunable magnetic reluctance paths by rotational alignment featuring through-hole high-permeability regulator disks and (3) a hybrid analytical finite element method (FEM) framework incorporating three-dimensional edge magnetic effects and eddy-current-induced counter fields. An equivalent magnetic circuit model with variable reluctance components is developed to quantify flux modulation, enhanced by Russell–Norsworthy correction factors for edge field quantification. The derived torque equations explicitly resolve skin depth dynamics and transient eddy current distributions through correction terms, and these equations are further amended based on finite element. Comprehensive three-dimensional finite element analysis validates the model's accuracy while optimizing critical parameters: the relative permeability of the magnetic flux regulator disk (MFRD) (optimal μrs=7000), regulator thickness (0.8 mm) and speed differential threshold (70 rpm peak torque). Simulation results demonstrate 0–418 N m continuous torque regulation within the range of 7–15° angular displacement. The modular design achieves torque capacity improvement over conventional axial displacement couplers through scalable disk-group multiplication. This co-design methodology establishes a foundation for high-torque magnetic transmission systems in space-constrained industrial applications, providing significant reference for detailed design and also prototype manufacturing in next steps.