A Generalized Anisotropic Finite Element Formulation for the Characterization of Synchronous Reluctance Motor Drives

A generalized tensor two-dimensional (2D) finite element (FE) computational model, which fully accounts for material and structural anisotropy, is presented. The results presented in this dissertation show for the first time a method that takes into account the curvature in anisotropic and composite material flux barrier regions by introducing equivalent reluctivities used to evaluate the elements of the reluctivity tensor. In this work, the novel approach developed is used to study the effects of accounting for anisotropy on the energy conversion performance characteristics of a prototype axially laminated anisotropic (ALA) rotor synchronous reluctance (SynRM) motor drive system. Therefore, the characterization of the ALA rotor synchronous reluctance motor provides a means for verification and validation of the proposed reluctivity tensor 2D-FE computational model. In order to predict the performance of the ALA-rotor SynRM drive system, a computer aided model based on the use of the indirectly coupled magnetic-electric circuit algorithm was developed. The model indirectly couples a two dimensional nonlinear finite element algorithm, which accounts for material and structural anisotropy, to a time variant state space algorithm describing the SynRM drive system electrical and mechanical dynamics. Because of the necessity to accurately predict the SynRM drive system performance, and to rigorously account for space and time harmonics, the natural ABC flux linkage based frame of reference was chosen as the basis for the state space model formulation. Computer aided models based on the natural ABC frame of reference are successfully implemented with considerably less simplifying assumptions in comparison to d-q type based models. Therefore, the natural ABC frame of reference flux linkage based model fully accounts for the impact of space harmonics from the machine geometry, and time harmonics from the pulse width modulated power electronics inverter. This model, indirectly coupled through time varying inductance matrix coefficients and windings currents, provides a computational framework with improved accuracy over d-q type models for the characterization and performance prediction of the axially laminated anisotropic rotor synchronous reluctance motor drive system. In the implementation of the coupled magnetic-electric circuit algorithm, the inductance profiles, calculated in the 2D-FE algorithm, are used as inputs to the state space model in the form of time varying coefficients of the inductance matrix. The method of current-energy perturbation is therefore modified to include the anisotropic tensor formulation. In addition, post processing algorithms for the calculation of the machine parameters such as core losses, taking into account the axial rotor structure, and material and structural anisotropy are developed. To evaluate the accuracy of the computational model, its results, where possible, are compared to test data obtained on a 100 kW ALA-rotor synchronous reluctance motor drive system