An integrated design process for optimized high-performance electrical machines

Many high-performance applications require an optimized electrical machine to achieve desired operating characteristics within a given environment and minimum cost and weight. Traditional design methods that focus on single-point performance optimization (nominal load) often do not produce the best possible solution, because they are based on empirical design formulations and rules gathered from standard motor designs. To facilitate machine design for such applications, an integrated design process is introduced that utilizes multi-physics modeling to determine an electrical machine's performance in the electromagnetic, electrical, mechanical, and thermal domains, and leverages a genetic algorithm optimization process to enable multi-objective optimization. The integrated design process avoids empirical design rules and allows a machine to be optimized simultaneously for several performance requirements, which can be transient and/or steady-state in nature, by taking advantage of application-specific system information. The capabilities of the process are demonstrated in an example design of an aerospace generator that has to satisfy several steady-state and transient performance requirements under varying environmental conditions. The results indicate how such a process drives design criteria and solutions and considers performance limitations that were not necessarily obvious from the requirements.