Design and modelling of a passive bearingless motor with buried permanent magnets

Bearingless machines are developed for applications which require high rotation speed, power density, long lifetime and high purity environment. The most compact machines of this category are the combined winding machines which gather both suspension and driving functions in only one set of winding. During the last few years, an Electrodynamic Thrust Self-Bearing Machine (ETSBM) has been developed by Dr. Joachim Van Verdeghem and Prof. Bruno Dehez. This motor is an Axial Flux Permanent Magnet (AFPM) machine which drives a flat disc rotor that passively levitates at high speed due to a specific connection of the windings. However the rotor topology of this machine uses a basic structure of Surface Mounted PMs (SMPMs) without any ferromagnetic part present on the rotor. More generally, the current researches in the field of axial bearingless machine have not yet explored the use of rotors with Buried PMs (BPMs). This kind of rotor should improve the suspension performance at low speed by increasing the inductance coefficients. Moreover, these parts will add axial and angular variations of the same coefficients which will create reluctant force components. With this in mind, this Master’s thesis aims to study the influence of a BPM rotor on the ETSBM performances. First, a review of the literature dealing with the AFPM machines using a rotor with buried permanent magnets (BPMs) is done. Thanks to it, a topology is chosen and a first design proposal is given. After that, an electromechanical model suited to the selected topology is developed. It describes the axial and spin behaviours of the machine. This model also highlights the modulation capability of the suspension through the motor currents and the reluctant effects generated by the structure of the rotor. Then static FEM simulations determine the parameters of the model and validate the assumptions made in this one. Next a bi-objective optimisation process consisting of the minimisation of the losses and the maximisation of the axial stiffness is realised in order to size the AFPM BPM machine. Finally a case study based on the best candidate previously computed is done. It aims to compare SMPM and BPM rotor topologies through both a quasi-static and a dynamic analysis. These analyses show that the BPM rotor can significantly upgrade the suspension performance. However, the analyses point out that the motor currents do not have enough influence on the suspension to modulate it. Besides of that, a comparison with a modified BPM rotor design shows that better performances could still be achieved. This paves the way for future researches concerning bearingless motors with a BPM rotor.