Cooling Fin Optimization on a TEFC Electrical Machine Housing Using a 2-D Conjugate Heat Transfer Model

The housing of totally enclosed fan-cooled electrical machines is usually finned to maximize the heat dissipation over its surface. Complex models involving heat and fluid flow simultaneously are necessary to evaluate a single design. An optimization over a wide parameter range is, therefore, not possible. In this paper, a novel approach on how to model this particular conjugate heat transfer problem in 2-D without the help of lumped circuits is presented. The proposed method is applied to an optimization process controlled by an evolutionary algorithm. Comparative, more detailed simulations in computational fluid dynamic (CFD) verify the tendencies obtained in the optimization. In another case, the method is used to evaluate the allowable degree of engineering tolerance for a given design. Although the method does not provide results as accurate as CFD, it is very helpful to study a large parameter set and determine the CFD candidates.