A magnetic-bearing-supported shaft may have a number of concentricity and alignment problems. One of these involves the relationship of the position sensors, the centerline of the backup bearings, and the magnetic center of the magnetic bearings. For magnetic bearings with permanent magnet biasing, the average control current for a given control axis that is not bearing the shaft weight will be minimized if the shaft is centered, on average over a revolution, at the magnetic center of the bearings. That position may not yield zero sensor output or center the shaft in the backup bearing clearance.
The desired shaft position that gives zero average current can be achieved if a simple additional term is added to the control law. Suppose that the instantaneous control currents from each bearing are available from measurements and can be input into the control computer. If each control current is integrated with a very small rate of accumulation and the result is added to the control output, the shaft will gradually move to a position where the control current averages to zero over many revolutions. This will occur regardless of any offsets of the position sensor inputs. At that position, the average control effort is minimized in comparison to other possible locations of the shaft. Nonlinearities of the magnetic bearing are minimized at that location as well.
A satisfactory simplification of the method is possible in most situations. For frequencies where power amplifier phase lags are negligible, the actuator control current is proportional to the control output command for transconductance mode power amplifiers. Thus, an integral of the output command to each amplifier, added to that output, will achieve the same effect. Even at frequencies where there is appreciable amplifier phase lag, the slowly building integral should still be effective.
The effectiveness of the simplified method (integration of control output) was demonstrated on a magnetic-bearing-supported energy-storage flywheel (DEV1) at the NASA Glenn Research Center. Previous tests of the flywheel had required repeated sensor offset adjustments as rotor speed changed to keep the average current for each axis near zero. (The presumed cause of the speed dependence was severe sensor runout interacting with various dynamic effects and system frequency response.) The present method maintained zero average control current throughout the speed range without any operator attention. The integral gain was chosen so that the time constant for approach to the final shaft position was a few seconds. An unexpected benefit was that the average motor current required at any rotor speed was reduced as well, presumably because of better centering of the motor rotor within its stator.
The method was also applied to the thrust bearing of the rig, which has a vertically oriented axis. In this situation with permanent magnet bias, the control effort is minimized with the axial bearing thrust disk higher than midway between the two stator disks. The axial control current becomes zero there, and the shaft weight is carried by the bias field because the magnetic gap is smaller above the thrust disk than below.
Glenn contact: Dr. Gerald V. Brown, 216-433-6047, Gerald.V.Brown@grc.nasa.gov
Author: Dr. Gerald V. Brown
Headquarters program office: OAT
Programs/Projects: Flywheel Technology, FESS
Last updated June 6, 2001, by Nancy.L.Obryan@nasa.gov
Responsible NASA Official:
Gynelle.C.Steele@nasa.gov
216-433-8258
Point of contact for NASA Glenn's Research & Technology reports:
Cynthia.L.Dreibelbis@nasa.gov
216-433-2912
SGT, Inc.
Web page curator:
Nancy.L.Obryan@nasa.gov
216-433-5793
Wyle Information Systems, LLC
