International audienceThis paper describes a 3-D numerical hybrid method (NHM) of the permanent-magnet (PM) eddy-current losses in axial-flux PM synchronous machines (PMSMs). The PM magnetic flux density is determined using the multi-static 3-D finite-element method (FEM) at resistance-limited (i.e., without eddy-current reaction field). Based on the predicted flux density distribution, the eddy-currents induced in the PMs and the 3-D PM eddy-current losses are calculated by 3-D finite-difference method (FDM) considering a large mesh. Therefore, this 3-D NHM is based on a coupling between the multi-static 3-D FEM and the 3-D FDM. Two 24-slots/16-poles (i.e., fractional-slot number) axial-flux PMSMs having a non-overlapping winding (all teeth wound type) with stator double-sided structure are studied: 1) surface-PM (SPM) and 2) interior-PM (IPM) To evaluate the reliability of the proposed technique, the 3-D PM eddy-current losses are determined and compared with transient 3-D FEM (i.e., magneto-dynamical 3-D FEM). The same nonlinear properties of the laminations have been applied for multi-static/transient 3-D FEM. The computation time can be divided by 25 with a difference less than 12%
This paper describes a two-dimensional (2-D) nonlinear adaptive magnetic equivalent circuit (MEC) of radial-flux interior permanent-magnet (PM) synchronous machines (PMSMs) for automotive application, mainly for electric/hybrid/fuel cell vehicles (EVs/HEVs/FCVs). It includes the automatic mesh of static/moving zones, the saturation effect, and the connection of the zones for the rotor motion which is ensured by a new approach called "Air-gap sliding-line technic". The local/integral quantities at no-load/load (viz., the magnetic flux density, the magnetic flux linkage and the voltage) have been validated with the 2-D finiteelement analysis (FEA) in the case of radial-flux interior PMSM with 18-slots/16-poles having a double-layer concentrated winding (all teeth wound type). The semi-analytical results are in good agreement, considering both amplitude and waveform. The computation time is divided by 3/2 with an error less than 7 %.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.