A Novel H-Element Magnetic Network Model for Yokeless-Stator Axial Flux Machines Considering Magnetic Circuit Saturation and Convergence

The equivalent magnetic network (EMN) model is an efficient method for rapid evaluation of the electromagnetic performance of axial flux permanent magnet (AFPM) machines. However, traditional EMN models often exhibit large errors in core flux density due to magnetic circuit (MC) saturation and nonconvergence, resulting in reduced calculation accuracy and slower computation of the EMN model. To address this, a novel H-element EMN (H-EMN) model considering core MC saturation is proposed, with refined permeance along saturated MC directions and reduced permeance along the nonsaturated MC directions. Furthermore, a new adaptive dynamic convergence strategy is introduced to rapidly reduce iterative errors and improve convergence. To verify the accuracy and effectiveness of the proposed H-EMN model, comprehensive analyses are conducted on the machine's electromagnetic performance and iteration performance. The results show that compared with the traditional EMN model, the proposed H-EMN model improves core flux density accuracy by up to 21.9% while reducing calculation time by 16.6%. Moreover, it enhances the accuracy of air-gap flux density, back EMF, and average torque by 4.54%, 2.4%, and 1.1%, respectively, and provides more accurate predictions of back EMF harmonics and torque ripple. The H-EMN results also closely agree with finite element analysis (FEA) and experimental tests.