Superconducting Bearings for Space Telescope Applications

THE OBJECTIVE OF THIS PROJECT IS TO DESIGN AN equatorial mount for a telescope that will be deployed on the Moon for the purposes of observational astronomy. The Moon is an ideal location because it offers a stable platform with no atmosphere that would interfere with the observations, and it has a rotation period of 29 days, which allows 350 hours of continuous observation. However, the harsh environment on the lunar surface requires special consideration in the design of the more delicate components of the telescope. One of these components is the bearing around the polar axis of the equatorial mount of the telescope. This bearing must be smooth enough to allow the refined control of the rotation of the telescope about the polar axis at a steady rate of one revolution in 29 days in order to keep the telescope focused on any cosmic object under study.

Conventional mechanical bearings fail in two regards: (1) tight tolerances commonly found in these bearings are easily clogged by lunar dust that is present everywhere on the lunar surface, and (2) under the cold vacuum conditions on the Moon, solid lubricants must be used in these bearings, but the lubricants have the tendency to exhibit torque ripples that render the control of the steady rotation almost impossible.

Our approach is to use hybrid superconductor magnetic bearings. These bearings, almost frictionless, are designed with a wide tolerance of a millimeter or more. Such a wide tolerance will decrease the probability that the bearing will be clogged up by lunar dust. Furthermore, no lubricant is needed; hence, there will be no torque ripples. Combined with the frictionless character of the bearing, this mechanism allows very precise control of the rotation of the polar axis.

The main focus of this project is to design and construct a hybrid superconductor magnetic bearing system for the polar axis of an equatorial mounting system for a lunar telescope, complete with actuator and control for tracking stars from the Moon. What we have accomplished up to now are: (1) a preliminary design for the hybrid superconductor magnetic bearing to be used on the next prototype and (2) characterization of the first prototype superconductor magnetic bearing module that was constructed for an azimuth mount before the beginning of this project. This characterization provides us with valuable information for the design of the actuator and control system for our next prototype, which will utilize a new design of the hybrid superconductor magnetic bearing specifically adapted to support a fixed weight with the rotation axis tilted at an arbitrary angle to gravity.

Tilted Axis Bearing
By definition, the polar axis of an equatorial mounting system for a telescope is aligned parallel to the rotational axis of the planet or satellite that the telescope is stationed on. Except at the poles, the direction of gravity is not aligned along this same direction. Hence, the bearing on the polar axis has to support a weight that is not parallel to itself, but is tilted at an angle approximately given by (the complement of) the local geographical latitude. This is why we need to develop a tilted axis version of our hybrid superconductor magnet bearing. Even so, the equatorial mounting system is favored over the alternative altitude-azimuth mounting system for two reasons: (1) tracking with the altitude-azimuth mounting system requires refined coordinated control on both axes of rotation--the altitude and the azimuth axes, whereas in the equatorial mounting system, the declination axis is held fixed, while the telescope is rotated about the polar axis only and (2) additional steerable optical components are required to correct for a general background rotation of the field of stars about the center object of interest that occurs while following the object of interest with the altitude-azimuth mounting system, but not with the equatorial mounting system.

Methodology
We started our design of the tilted axis bearing from the hybrid superconductor magnetic bearing in the first prototype. The new element here is that the force of gravity is no longer aligned with the rotation axis. As a consequence, we cannot keep the axisymmetry of the earlier design. However, the interaction of high temperature superconductors with permanent magnets still demands that the rotor magnet be axisymmetric about the axis of rotation if we are to retain the near frictionless performance of the bearing. This feature we retained. We modified the supplementary magnets in the supports directly connected to the ground in such a way as to provide a vertical supporting force to the telescope that remains vertical and of constant magnitude, without regard to the tilt angle of the axis of rotation of the rotor magnet. We developed a preliminary design, but in order to simplify it, we gave up the capability to support a range of masses that we built into the previous prototype for the azimuth mount. With this change also went the capability of the bearing to provide the telescope with proper counterbalance. Thus, we have to develop some other means of fine-tuning the weight and the balance of the telescope. We do not expect any insurmountable difficulties associated with this requirement. Hence, the next logical step in the development of this new bearing will be refinement of design, fabrication, and performance testing.

Characterization of Bearing on Azimuth Mount
In previous work, we developed a prototype hybrid superconductor magnet bearing system for a lunar telescope azimuth mount that is based on passive magnetic levitation and the flux pinning effect of high-temperature superconductivity.^1,2 The rationale lies in the unique capability of high-temperature superconductors to adapt to the low temperature and vacuum environments in space or on the Moon, and to enhance system stability passively without power consumption.

In superconductor magnetic bearing systems, levitation forces are produced mostly by sets of two permanent magnets of the same or opposite polarities. There is, however, an inherent instability in the interaction of two magnets. High temperature superconductors are used to enhance system stability by pinning down the magnetic flux. The interaction between a high temperature superconductor and a magnet can also provide additional levitation force. A superconductor magnetic bearing system experiences nonlinear elastic restoring forces and hysteretic drag forces within some ranges of angular position as a consequence of the interaction between the magnet and high temperature superconductor.^3,4 Characterization experiments have been conducted to further the understanding of this dynamic behavior. This characterization provides information for the design of future superconductor magnetic bearing systems, such as the elastic restoring force, hysteretic drag force, position and velocity dependences, their stiffness, etc. The experiments shed light on the torque to be delivered by the driving motor for proper control of the desired rotational motion. For a generic superconductor magnetic bearing, we have pursued the characterization of our first prototype hybrid superconductor magnetic bearing system on an azimuth mount.

An experimental setup for the characterization has been assembled as follows. The superconductor magnetic bearing system is securely fixed on the Melles Griot solid aluminum optical breadboard. The platform of the telescope, which is attached on top of the bearing rotor, is coupled to a Himmelstein model MCRT 3L-08T, 10 oz-in (70 mN m) range, non-contact rotating torque transducer via a Thomas miniature flexible disc coupling, which provides relatively high torsional stiffness along the shaft axis and compliance along all five remaining degrees of freedom. The other shaft of the torquemeter is coupled to the MicroMo planetary gearhead 23/1 (gear ratio of 1526:1) via a Thomas miniature flexible disc coupling. The gearhead is attached to a MicroMo brushless D.C. servomotor 2444S024BK315, which is attached to a Hewlett Packard HEDS-5500 two-channel incremental optical encoder with a resolution of 512 counts-per-revolution with quadrature output.

The torquemeter is connected to a model 61201DL universal strain gauge amplifier, and the measured torque is displayed through a model 61201-OZ-IN digital, universal strain gauge amplifier-display. The brushless D.C. servomotor is connected to a low voltage pulse-width-modulation (PWM) amplifier, Western Servo Design BPW-S3-6/10.

For data acquisition and control, a motion control interface card for PC, Precision MicroDynamics model MFIO-3A-ISA-0, is inserted into a PC. This I/O card has a three channel encoder interface and DAC interface within one board and has an open architecture allowing hardware access at the register level. From Analog Devices a 12-bit and 600 ksps analog-to-digital converter (ADC) chip, AD7892AN-1, is integrated into the MFIO board for torque transducer analog input.

Mechanical supporting structures for the setup are designed and assembled. Computer codes were written for measurements. The integrated system for experiments is set up at a laboratory workspace at Johnson Space Center in Building 268 for carrying out the investigation of the Lunar Telescope.

To obtain constant velocity of the superconducting bearing for its characterization, the drive motor speed was regulated in a proportional plus integral (PI) velocity control loop. The positions of the bearing were measured using an encoder but differentiated to obtain velocity data. A finite-impulse-response (FIR) low-pass-filter was then applied using the MATLAB Signal Processing Toolbox to filter out the noise. The sampling rate was 1 kHz with cut-off frequency of 10 Hz. The pass band was 0.5 dB and the stop band attenuation was 40 dB. Drag forces were measured using the torquemeter with a resolution of 0.07 mN m. The same filter was applied to remove noise from the data.

To investigate the velocity dependence of the drag forces, the bearing was rotated with two different command velocities, 0.02rad/sec and 0.5 rad/sec.

Results and Discussion
Figure 1 shows the torques measured with angular velocities of 0.02 rad/sec and Fig. 2 at 0.5 rad/sec. Figure 1 spans data gathered in almost one complete revolution. It shows that the torque fluctuates between 0.5 mN m clockwise and 1 mN m counter-clockwise. Figure 2 spans data gathered in 15 complete revolutions. It shows almost the same behavior with the torque oscillating between 0.75 mN m clockwise and 1 mN m counter-clockwise, roughly, with a dominant periodicity at 0.26 hz. The resolution of the torque transducer was 0.07 mN m.

Fig. 1. Drag/torque vs. Time at 9.92 rad/sec

Averaged over complete cycles, the torque shows a positive bias, indicating that there is a drag which is dissipative. Then oscillations about this average could be identified with elastic restoring forces associated with slight uneven distribution of trapped magnetic flux around the ring of superconductor disks, caused by the fact that the superconducting ring is not one continuous piece around while the rotor magnets are not perfectly symmetric about the axis of rotation. The magnitude of the positive bias is about 0.4 mN m in Fig. 1, and 0.3 mN m in Fig. 2. Thus, there is essentially no change in the drag torque as one changes the rotational speed by more than an order of magnitude.

Fig. 2. Drag Torque vs. Time at 0.5 rad/sec

Conclusions and Future Work
Experimental characterization of the superconductor magnetic bearing did not reveal any basic flaws in the application of this bearing to achieve extremely precise control of the rotational motion for the purpose of tracking stars from the Moon. Extra work is still desirable, for the measurements were carried out at speeds that are still more than three orders of magnitude higher than that required for operation of the telescope on the Moon. This research would have to be accomplished with the construction of a direct drive to achieve very low rotation rates without depending on gear reduction. The design of this drive motor can now proceed with the data collected. This motor can be equipped to extend the measurements down to lower rotation speeds until we reach our goal of being able to control the rotation at 1 rev per 29 days. Before embarking on this endeavor, we should construct the next prototype with a tilted axis bearing and then design the motor for it.

Space programs and industry can both benefit by the development of a superconductor magnetic bearing using permanent magnets and high temperature superconductors. Such levitated bearings are free from friction, mechanical wear, mechanically induced torque ripple, and energy dissipation, thus resulting in long-life operation and less maintenance with less power consumption. With reduced system nonlinearity, such as friction including stiction and torque ripple, design of the control system becomes easier and simpler, rendering a more robust and high performance system. The velocity independence of bearing drag torque becomes very important in high speed applications. The rotational speed of the bearing was limited due to wear in conventional mechanical bearings. In active magnetic bearings, bearing loss is proportional to the square of speed, thus causing consumption of a lot of energy at high speed operation.

Superconductor magnetic bearing systems have many promising applications. A flywheel kinetic energy storage system incorporating superconductor magnetic bearings can be used for renewable energy storage from fluctuating sources such as the wind. Micro-electro mechanical systems (MEMS) may benefit from levitated bearings since wear is the most serious problem as a consequence of the high surface to volume ratio in the micro domain.

Most mechanical devices utilize bearings as important functional components. Compared to conventional mechanical bearings, the superconductor magnetic bearing is superior because of its extremely low bearing loss, but it lacks stiffness to support sizable loads and operates only under cryogenic conditions. Thus, it fits naturally in deep space which is a cold vacuum and away from any significant source of gravity.

Conditions on the Moon also correspond to a cold vacuum. The extensive presence of dust, which is a problem for conventional mechanical bearings, could also be mitigated by using wide gaps in superconductor magnetic bearings. The demand of extreme precision is met with the almost complete absence of friction in superconductor magnetic bearings. Moreover, the reduced gravity on the surface of the Moon makes the weight of the telescope a manageable load for the superconductor magnetic bearing, despite its weak stiffness. Besides, we are further enhancing the load carrying capacity of the superconductor magnetic bearings with additional magnets. Finally, the passive nature of the superconductor magnetic bearing means that no power source is required to run it, so long as it is adequately shaded to keep the superconductors cold enough. This makes it a winner over active magnetic bearings where power source is scarce. Putting all this together, we can see that the superconductor magnetic bearing is uniquely suited for telescope mounts on the Moon and could be advantageous in deep space missions.

References
¹K. Ma, M. Lamb, P. Chen, T. Wilson, R. Cooley, H. Xia, C. Fowler, Q. Chen, and W. K. Chu. "A Natural Application for High Temperature Superconductors: A Bearing for the Azimuth Mount of a Lunar Telescope," Applied Superconductivity 2.7/8 (1994): 479-86.
²M. Lamb, K. Ma, R. Cooley, D. Mackey, R. Meng, C.W. Chu, W. K. Chu, P. Chen, and T. Wilson. "High Temperature Superconducting Bearings for Lunar Telescope Mounts," IEEE Trans. on Applied Superconductivity 5.2 (1995): 638-42.
³F. C. Moon, K.-C. Weng, and P.-Z. Chang. "Dynamic Magnetic Forces in Superconducting Ceramics," J. Appl. Phys. 66.11 (1 Dec. 1989): 5643-45.
⁴F. C. Moon, M. M. Yanoviak, and R. Ware. "Hysteretic Levitation Forces in Superconducting Ceramics," Appl. Phys. Lett. 52.18 (2 May 1988): 1534-36.

Publications

TELESCOPE PROGRESS—Eunjeong Lee, post-doctoral aerospace fellow, points with pride to the development of mounting devices for the space telescope. The wall display provides insight into the history and progress of the project. Surrounded by superconducting magnets, the telescope will be able to withstand the harsh climate on the Moon better than conventional instruments.

Lee, E., K. B. Ma, T. Wilson, and W. K. Chu. "Characterization of Superconducting Bearings for Lunar Telescopes," IEEE Trans. on Applied Superconductivity. (To appear.)
Lee, E. "Superconducting Bearings for Space Applications," Preliminary Proc. of the National Science Foundation Civil and Mechanical Systems Workshop, Arlington, TX, Sept. 23-26, 1997.
Ma, K. B., J. H. Yu, E. Lee, and W. K. Chu. "High Temperature Superconductor-Magnet Momentum Wheel for Micro Satellite," Proc. of the 1998 Int'l Conf. on Integrated Micro/Nano Technology for Space Applications, Houston, TX, Nov. 1998.
Presentations
Lee, E. "Superconducting Bearing System for Lunar Telescope," Dept. of Mechanical Engineering, the Univ. of Nebraska, Lincoln, NE, May 4, 1998. Lee, E., K. B. Ma, T. Wilson, and W. K. Chu. "Characterization of Superconducting Bearings for Lunar Telescopes," Applied Superconductivity Conf., Palm Desert, Sept. 14-18, 1998.
Funding
"Basic Research Impacting Advanced Energy Generation and Storage in Space." Air Force Office of Scientific Research, 4.5 yrs, $800,000.
"An Advanced Momentum Wheel for Miniature Satellite." NASA/Langley Research Center, 2 yrs, $200,000.
Awards
The Workshop Material/Product Performance Award, 1997 Int'l Workshop on Superconductivity.

Contents
ISSO -- Institute for Space Systems Operations
1997-1998 Annual Report