Thermal Modeling of Electrical Machines with Advanced Fluid Cooling

Advanced cooling methods, especially direct oil cooling, are used to maximize the power density of electrical traction machines. However, thermal monitoring with temperature sensors can only provide limited information about the thermal stress distribution. This work combines computational fluid dynamics and lumped parameter thermal networks to develop a thermal model of a prototype machine with direct oil cooling. The generated model represents the inhomogeneous spatial temperature distribution caused by the asymmetrical oil distribution in radial and axial direction of the machine. The proposed method allows a rapid study of the thermal behavior of electrical machines with direct oil cooling. Experimental measurements validate the accuracy of the thermal model. The temperature difference between simulations and measurements is within 5 °C.