The influence of concentrated winding configurations and stator core structures on torque performance of fractional-slot PM machines are investigated in this paper. Analyzing the separated torque components of 12-slots/10-poles (12S/10P) PM machines by the frozen permeability method, it shows that the torque ripple is mainly due to on-load cogging torque in closed slot machines. Besides, the local saturation due to closed slot will jeopardize the merit of alternate teeth wound winding, such as resulting in reduced average torque. The analysis of stator tooth permeability variation over one electrical period demonstrates that the major reason for torque ripple difference among these machines with different stator topologies is the asymmetric saturation between the adjacent teeth. Both the subharmonic due to armature reaction (closely related to winding connection and electric loading) and the asymmetry of stator core structure contribute to such asymmetric saturation. Although 12S/10P PM machines were chosen as the example in this paper, the conclusion is effective for machines with other slot/pole number combinations as well. Experiments have been carried out to validate the predictions.
A modular linear doubly salient machine with dual-PM in the primary yoke and slot openings is proposed in this paper. First, the machine topology is introduced, where the PMs are employed in both the primary yoke and slot openings. The magnetic fluxes excited by the PMs in the slot openings pass through the air gap and link the secondary, not short-circuited by the primary teeth. Hence, the magnetic fluxes excited by both PMs can be concentrated and enhanced in the air gap. Then, the operation principle is investigated based on the permeance model, which shows that it obeys the flux modulation principle. The electromagnetic performances are analyzed in detail using FEM, which shows that the force density and overload capability can be significantly improved with the slot PMs. Finally, the 2-D finite-element (FE) predicted results are validated by the 3-D FE results that considering the transverse end effect.
25 May 2020 in IEEE Transactions on Industry Applications
In the past, various fundamental model based sensorless control methods have been proposed. However, most of them are developed under a healthy condition. Hence, in this paper, fundamental model based sensorless control of permanent magnet synchronous machines (PMSM) under rotor eccentricity is investigated which is a common issue in many drive systems. It is known that under rotor eccentricity, three-phase back-EMFs, flux-linkages, inductances and torque will be influenced. Mathematical derivation reveals that these effects will cause undesired harmonics in the position estimation. Since the frequencies of these low-order harmonics are much smaller than the bandwidth of position observer such as phase-locked loop, it is difficult to effectively eliminate them. Hence, the conventional back-EMF method is modified by adding an adaptive notch filter to suppress these undesired harmonics. Furthermore, the rotor eccentricity effects on PMSM characteristics are examined by measurements. The modified position estimator is implemented on a dSPACE platform and is verified by experiment results.
28 March 2019 in IEEE Transactions on Industry Applications
This paper proposes a double superposition method to fast and accurately predict the cogging torque and unbalanced magnetic force (UMF) of PM machines with axial-varying rotor eccentricities. The original axial-varying eccentric machine is cut into several uniform slices along the axial direction and each slice is considered to be axially uniform, i.e. axial-even. For each axial-even eccentric machine slice, the circumferential superposition method can be used to predict the cogging torque and UMF of this slice. Finally, the cogging torque and UMF of the original axial-varying eccentric machine can be obtained by using axial superposition technique which combines the cogging torque and UMF of each slice with different eccentricity magnitudes together. The 3D finite element analysis (FEA) is adopted to validate the efficacy of the proposed method and the merit of high accuracy and computation efficiency are also identified by comparing the results predicted from these two methods.