Home | english  | Impressum | Datenschutz | Sitemap | KIT

Thermal Modeling of the Stator Slot in Electrical Machines Using an Extended Layer Approach

Thermal Modeling of the Stator Slot in Electrical Machines Using an Extended Layer Approach
Autor:

Felix HoffmannBenedict JuxMartin Doppelbauer

Links:
Quelle:

IECON 2019 - 45th Annual Conference of the IEEE Industrial Electronics Society

Datum: Oktober 2019

Abstract:

In the design process of efficient and high-power density machines, it is essential to couple the electromagnetic and thermal analysis. The thermal analysis can be done by numerical approaches or analytical lumped circuit analysis. The lumped circuit analysis has the advantage of a fast computation time, whereas the numerical approach has a high accuracy. In this paper an analytical approach for the stator slot is presented, which homogenizes the stator slot considering four directions in predefined layers. With the extension to model the layers in different directions, it is possible to predict a more precise temperature distribution inside the slot. Consequently, unlike the existing approaches, maximum and minimum temperatures can be predicted. We have implemented the extended layer approach, simulated and validated it with a Finite Element Analysis (FEA). The extended layer approach is simulated with six layers, which shows a good trade-off between accuracy and simulation time. The maximum temperatures of the hottest to the coolest layer differ by 21.8 K. The temperature difference between the different directions is up to 16.1 K, showing the importance of modeling the slot in all directions. The largest difference between the maximum temperatures of the FEA and the extended layer approach is 6.1 K in the transient case and 3.4 K in thermal steady state, whereas the average temperatures differ by a maximum of 3.5 K in the transient case and 0.2 K in thermal steady state. The validation with FEA results shows a good accordance to the analytical approach, while the simulation time of the analytical extended layer approach is about 150 times faster.