OPTIMUM DESIGN FOR IMPROVING MODULATING-EFFECT OF COAXIAL MAGNETIC GEAR USING RESPONSE SURFACE METHODOLOGY AND GENETIC ALGORITHM

Coaxial magnetic gear (CMG) is a non-contact device for torque transmission and speed variation which exhibits promising potential in several industrial applications, such as electric vehicles, wind power generation and vessel propulsion. CMG works lying on the modulating-effect aroused by the ferromagnetic segments. This paper investigates the optimum design for improving the modulating-effect. Firstly, the operating principle and the modulating-effect is analyzed by using 1-D field model, which demonstrates that the modulatingeffect is essential for the torque transmission capacity of CMGs, and the shape of the ferromagnetic segments have impact on the modulatingeffect. Secondly, the fitted model of the relationship between the maximum pull-out torque and the shape factors including radial height, outer-edge width-angle and inner-edge width-angle is built up by using surface response methodology. Moreover, FEM is engaged to evaluate its accuracy. Thirdly, the optimum shape of the ferromagnetic segment is obtained by using genetical algorithm.