This grant provides funding for an investigation of high performance magnetic levitation for manufacturing. The growth of manufacturing processes that require extreme cleanliness, very high precision positioning or non-contact part handling suggests an increasing role for magnetic levitation in production environments. As a technology, magnetic levitation is developing rapidly, spurred by dramatic advances in power electronics, precision position sensing, high speed digital signal processing, and control theory. Magnetic levitation has found application in many processes including metal conveyance, metal coating, silicon wafer transport, film stretching, and photolithography. Almost all work in high performance magnetic levitation assumes that the actuator can be constructed from laminations so as to minimize the production of eddy currents in the face of the rapidly varying magnetic fields caused by active control. However, in many current or anticipated applications, laminated construction conflicts with the basic process or would be highly costly. Without laminations, a rapid change in current applied to the electromagnet's coil results in a much slower change in the applied force due to the eddy currents induced within the actuator by the changing field. These currents interact with the produced field and greatly complicate the actuator's dynamics. If known, the actuator dynamics can be partially compensated by a well-designed control algorithm so as to achieve improved stiffness and servo bandwidth. However, they also impose fundamental limits upon levitation performance. Thus, the ability to predict, identify, and model non-laminated actuator dynamics is critical to attaining high stiffness/ bandwidth levitation and to understanding what can not be achieved. The objectives of the theoretical and experimental tasks of this research effort are to (1) determine the fundamental limits that eddy current effects impose upon the performance of active magnetic suspensions, and (2) develop robust control strategies that provide very high levels of performance with non-laminated actuators.
If successful, this research effort will enable non-laminated magnetic suspensions that provide highly accurate positioning in spite of disturbance forces and reposition the levitated object quickly. The effort will also provide a fundamental knowledge base that will connect an actuator's dynamic properties to its size, geometry, and composition. This knowledge will allow the prediction of magnetic suspension performance limits as well as the optimization of the stator design to maximize stiffness and servo bandwidth. Finally, it is anticipated that this research will help define new opportunities in manufacturing for high performance magnetic suspension.
