A parametric, self-segmenting steady-state thermal estimation for switched reluctance machines

This work focuses on the creation of a thermal model used for the estimation of the temperature distribution within ideal switched reluctance machine (SRM) design candidates. The estimation is performed using a hybrid finite difference (FD) and thermal circuit approach. Separate frame and shaft thermal transport models using thermal circuit account for axial heat transfer. Empirical Nusselt correlations for laminar shear flow, laminar flow with vortices, and turbulent flow (based on rotor speed) are used to estimate the convective heat transfer coefficient in the air gap. The results from the shaft thermal transport validation were a maximum temperature error of 6.86°C and 4.10%. A 3D comparison between the FD and finite element analysis (FEA) temperature distribution in the frame, stator, and rotor of the machine was conducted. The FD thermal model slightly overestimated the temperature when compared to a 3D FEA with an error of 3.8°C and 3.78% respectively. The FD thermal model was able to predict the maximum temperature to within 10.80°C when compared to data collected from an experimental SRM.