Self-bearing motors (SBMs), or bearingless motors, are electric motors with a magnetically integrated bearing function. The result is a single actuator that provides levitation and rotation simultaneously. The main advantages are an increased power density and a reduced shaft length which softens the performance limitations imposed by rotordynamics. Furthermore, SBMs extend the capability of active magnetic bearings (AMBs) by an extra degree-of-freedom (DOF), thus allowing for six DOF contactless positioning of rotors. However, functional integration also increases the system complexity and poses new modeling and control challenges.
SBMs have a high torque-to-inertia ratio which makes them suitable for high performance servo applications. As a result, research in the ANCL is focused on the modeling and control of a permanent magnet synchronous machine for use as a self-bearing servomotor. A contactless servomotor has numerous potential applications where precision pointing and slewing requirements are extremely stringent. An example is a spaced-based optical tracking system, where a laser beam must be precisely positioned to establish an inter-satellite cross-link. The unique features of the self-bearing servomotor are its slotless stator construction and all-Lorentz operating principle. The all-Lorentz principle is particularly interesting, as it is an unusual basis for magnetic levitation (most magnetic levitation systems exploit the Maxwell force principle). These unique features yield smoother torque production and a simple design.
Initial efforts have been to establish a control-oriented model of the system. Emphasis has been placed on an analytical derivation of the bearing force and torque characteristics that capture the linear and nonlinear characteristics. Doing this provides new physical insight as well as the flexibility to choose the control approach that will yield the maximum performance benefit. In addition, the model must be amenable to standard parameter identification techniques to address the inevitable model uncertainty which arises in practice. This work is reported in . The proposed analytical model is validated by computational modeling through finite element analysis.
Current research efforts are focused on developing advanced control strategies beyond the typical "working" design: decentralized PID. There are a number of avenues to explore. First, and most common, is the potential for nonlinear compensation of rotor-angle dependent effects, i.e. bearing force and torque ripple. Perhaps more interesting, given the multiphase nature of this self-bearing servomotor (it has 12 phases in total!) is exploiting untapped degrees of freedom for control. For instance, the servomotor is current driven and the motor input could be either the magnitude of the phase currents, the phase angle offset, or possibly both. As well, the motor is over-actuated in the sense that it generates more distinct forces (four) than physical degrees of freedom (three). Below is a computer generated image of an actual self-bearing servomotor test rig in the ANCL that was built by Airex Corp. (Somersworth, NH). This rig will be used for experimental model validation and to assess new control strategies. It has six DOF in total, with conventional active magnetic bearings (AMBs) supporting the remaining three DOF.
Infrastructure funding to investigate the modelling and nonlinear control of a self-bearing motor is provided in part by Airex Corp. and the Canadian Foundation for Innovation.