Geometric design of electric motors using adjoint-based shape optimization

Abstract

Interest in electric aircraft has increased due to developments in electric propulsion technology and concerns regarding aircraft carbon emissions. The emerging urban air mobility (UAM) industry aims to provide convenient short-range air travel using electric aircraft. An important factor in the design of electric aircraft is the modeling and design of electric motors. The many degrees of freedom in electric motor design makes it a complex design problem. To mitigate this complexity, we have developed a novel adjoint-based electric motor design methodology using finite-element electromagnetic analysis and PDE-based mesh warping with exact derivatives. This paper highlights the approach and details of the proposed method and presents results from its application to a representative motor design problem. The optimization results in a 35% decrease in motor mass and a 3% increase in efficiency in comparison to the baseline design. These results demonstrate the efficacy of an adjoint-based optimization approach with exact analytical derivatives for electric motor design.