Skip to Main Content
This paper addresses the difficulty of modeling and optimizing transverse flux machines (TFMs). 3D flux line patterns, complex leakage paths and saturation of the magnetic material significantly add to the complexity of building accurate magnetic models to optimize TFMs. Therefore, common TFMs design approaches usually rely on time-consuming finite element analyses, guided by the designer's knowledge. In this paper, a new design method is presented and applied to maximize the no-load flux of a Clawpole TFM. An error compensation mechanism combined to an analytical reluctance model is proposed as a solution to overcome inherent inaccuracies of TFM analytical models. It is shown how finite-element derived factors applied to selected reluctances of an analytical model can compensate for the model errors and validate the optimal solution found in a TFM design process, with a limited number of finite element simulations.