Modeling and optimization of permanent magnetic motors

Abstract

This thesis develops analytic models for the prediction and optimization of radial-flux permanent magnet motor torque and efficiency. It also facilitates the design optimization of electromagnetically-powered rotorcraft by characterizing optimal motor performance over a wide range of motor mass. The solution of the Poisson Equation, found as a function of the three spatial coordinates, is applied to the prediction of motor fluxes. Back EMF waveforms of prototype motors are measured in order to validate the analytical predictions. The solution of the magneto-quasi-static Maxwell's Equations are applied to the analysis of eddy currents and torque measurements are made to verify the theoretical predictions. Simplified motor models are discovered which yield symbolic solutions for optimal motor parameters as a function of mass. The Monte Carlo method is applied to the empirically-based motor model to compute optimal motor dimensions, number of magnet poles, and magnet height versus motor active mass for arbitrary material parameter values.