Parametric Fin Array Optimization for Radial Flux Machines Considering Time Varying Loads

This work utilizes a novel, generic, thermal model of an electric machine in conjunction with particle swarm optimization to optimize the electric machine’s fin array considering time varying loads. The maximum power rating on radial-flux electrical machines is typically based on the steady state temperature of the windings. This leads to over designs in applications in which only short periods of high power are required. The proposed optimization technique can be used in the design process to reduce the risk of over design therefore leading to reduced material costs for finned frames and increased power density in radial flux machines. Whilst many numerical optimization techniques exist, this paper will consider the application of particle swarm optimization techniques to optimize the fin array parameters. The parameter space to be investigated will consider the fin height (h_f), fin width (w_f), and fin spacing (s_f).