Key points are not available for this paper at this time.
The present work aims at providing a predictive numerical methodology for the thermal characterization of electric motors. The methodology relies on a 2D-FE simulation for the estimation of the electromagnetic (iron and joule) losses. The latter are then exploited in a 3D-CFD Conjugate Heat Transfer analysis for the evaluation of the thermal field. The CFD model includes both the solid components and the fluid domains. The main novelty of the paper is represented by the copper coil modelling. In fact, copper, air, epoxy resin and enamel are synthetized in a single homogeneous body able to reproduce the thermal behaviour without including the single components, to reduce the computational cost. The methodology is validated against experimental data on a three-phase squirrel-cage induction motor. As for the experimental data (available at three different operating conditions), temperature distributions are measured by thermocouples at the test bench for the validation of the 3D-CFD CHT model. In addition, experimental estimations of the losses are available for the validation of the 2D electromagnetic simulations. The numerical results in terms of motor performance, electromagnetic losses and thermal field are discussed and are proved to be close to the experimental counterparts, for all the investigated conditions.
Iacovano et al. (Sat,) studied this question.