Background ofthe Invention

This patent application is a continuation-in-part patent application of United States Patent Application Serial No. 08/025,950, filed March 3, 1993. The United States government has rights in this invention pursuant to Contract W-31-109-ENG-38 between the United States Department of Energy and the University of Chicago, representing Argonne National Laboratory.

The present invention is directed to superconducting bearings. More particularly, the present invention is directed to superconducting bearings offering increased stability and levitation force, as well as the structural integrity to rotate at very high speeds.

When superconducting bearings are used in flywheel energy storage devices, the efficiency ofthe flywheel can be very high and flywheels of this type become economic for diurnal energy storage and other applications where high energy efficiency is important. Diurnal storage of electricity is important to electric utilities in order to efficiently utilize base load generating plants and to meet the varying load demands of their customers. For example, the base load generating plants can charge storage units at night when demand is low. Peak electrical demands can then be met by discharging the storage units during the daytime peak hours. Energy storage can also play a substantial role in eliminating or postponing the installation of power lines with larger capacity. Power can be transmitted at night to a substation or user energy storage unit when demand is low, and then during peak power times, the energy storage units can be discharged. The placement ofthe energy storage units can occur in various parts ofthe electrical distribution system: utility parks where large amounts of energy are stored; in tandem with photovoltaic or wind energy generation facilities that are time dependent; substation units; individual companies and houses. The invention can also be used for energy storage on electric vehicles such as cars and buses, or as wayside energy storage for electric trains or other transit vehicles.

Flywheels are often considered for energy storage applications. Their primary advantages are modularity, mechanical simplicity (low cost), high energy storage density

(Wh/kg), and high efficiency input and output of electrical energy. The ability to produce high strength flywheel rotors and the ability to efficiently transfer energy in and out of a flywheel are well known and will not be discussed herein.

The primary disadvantage of conventional flywheels is inefficiency in standby storage mode. These losses occur because the bearings that support the flywheel structure produce high rotational losses. The present invention provides bearings having very low rotational losses and can enable standby losses in flywheels to be 0.1% hr or less. High temperature superconductor bearings designed in accordance with one form ofthe invention provide passive stability in all directions; e _^ , they provide a positive stiffness in all displacement directions. Further, they allow rotational motion with very low friction.

A preferred high temperature superconductor bearing embodiment ofthe present invention provides a number of improvements that decrease rotational losses and overcome reluctance instability. Other forms ofthe invention provide these advantages and incorporate a ferromagnetic part of a rotor directly into the composite winding ofthe flywheel, allowing even higher rotor speeds to be obtained. Another preferred embodiment enables greater levitation force to be produced while still providing other previously described advantages ofthe invention.

It is therefore an object ofthe present invention to provide a low-loss mixed-mu high temperature superconductor bearing that can enable a flywheel energy storage device to be highly efficient.

It is a further object ofthe present invention to provide novel method and apparatus that overcome reluctance instabilities in mixed-mu high temperature superconductor bearings.

It is a further object ofthe present invention to provide a mixed-mu high temperature superconductor bearing capable of rotating at high velocities without sacrificing the bearing's structural integrity.

It is a further object ofthe present invention to provide high levitation pressures in a mixed-mu high temperature superconductor bearing.

These objects and other advantages ofthe invention will be readily apparent from the following description ofthe preferred embodiments thereof, taken in conjunction with the accompanying drawings described hereinbelow.

Brief Description ofthe Drawings

FIG. 1 illustrates a mixed-mu bearing constructed in accordance with one form of the invention;

FIG. 2A shows a mixed-mu bearing without reluctance instability effects, and FIG. 2B shows a mixed-mu bearing experiencing reluctance instability;

FIG. 3 illustrates the coefficient of friction of a mixed-mu bearing as a function of rotor rim velocity; and

FIG. 4 shows a top view of a mixed-mu bearing rotor using ferrite or steel wedges to increase radial and circumferential reluctance;

FIG. 5 illustrates a top view of a mixed-mu bearing rotor including ferrite or steel pins;

FIG. 6 shows a top view of a mixed-mu bearing including staggered ferrite or steel pins;

FIG. 7 illustrates a side view of an alternative embodiment of a mixed-mu bearing with a rotor surface shaped to lessen the effects of reluctance instability;

FIG. 8 shows another alternative embodiment of a mixed-mu bearing design capable of achieving large levitation force; and

FIG. 9 shows an alternative embodiment ofthe bearing shown in FIG 8. Detailed Description of Preferred Embodiments

A basic mixed-mu bearing constructed in accordance with one form ofthe invention is indicated generally at 10 in FIG. 1. The basic mixed-mu bearing 10 is preferably comprised of a stationary permanent magnet system, stationary high-temperature superconductors, and a rotating ferromagnet (e.g., magnetic steel). The bearing 10 is referred to as "mixed-mu" because ofthe use of both ferromagnetic (mu > 1) and diamagnetic (mu < 1) phenomena in the levitation, where "mu" refers to the magnetic permeability of a component.

A first stationary permanent magnet 12 is fixed above and adjacent a stationary upper high temperature superconductor 14 and a second permanent magnet 16 is disposed below and adjacent a stationary lower high temperature superconductor 18. A rotor 20 is disposed for rotation vertically between the two stationary permanent magnel high temperature superconductor pairs as shown in FIG. 1. While the rotor 20 can comprise a variety of durable materials, preferably ferromagnetic material, such as steel, is used.

The entire bearing 10 is field-cooled (i.e., the high temperature superconductors 14 and 18 are immersed in liquid nitrogen until their temperature lowers sufficiently to enter the superconducting state) with the rotor 20 at approximately vertical equilibrium (with the upward magnetic force nearly balancing the downward gravitational force). The high temperature superconductors 14 and 18 are encased in a conventional liquid nitrogen cryogenic chamber, which is not shown for increased clarity. The entire bearing assembly is encased in a conventional vacuum chamber, which is also not shown.

The permanent magnets 12 and 16 are both polarized vertically in the same direction and induce magnetization in the rotor 20. Without the high temperature superconductors 14 and 18 in the superconducting state, the bearing 10 is radially stable but vertically unstable. Accordingly, the rotor 20 generally stays centered along its vertical axis of symmetry, but either falls toward the second permanent magnet 16 or rises toward the first permanent magnet 12. Because ofthe diamagnetic properties ofthe high temperature superconductors 14 and 18, magnetization ofthe steel in the rotor 20 induces shielding currents in the high temperature superconductors 14 and 18. This results in a repulsive force between the high temperature superconductors 14 and 18 and the steel in the rotor 20 in this embodiment. This repulsive force increases as the distance between the steel and one ofthe high temperature superconductors 14 and 18 decreases. This diamagnetic property tends to stabilize the bearing 10 in the vertical direction. When the steel is close enough to one ofthe high temperature superconductors 14 and 18, the stabilizing influence of one ofthe high temperature superconductors 14 and 18 is enough to overcome the destabilizing influence ofthe permanent magnet/rotor interaction.

Flux pinning in the high temperature superconductors 14 and 18 further acts to stabilize the bearing 10. This flux pinning is most evident with "Type II" superconductor materials that allow some magnetic flux to enter the materials. Once the high temperature superconductors 14 and 18 are field-cooled, one or more ofthe permanent magnets 12 and 16 can be moved vertically to adjust the vertical position ofthe rotor.

The present invention can also work with only a single high temperature superconductor or single permanent magnet as discussed in the parent application (United States Patent Apphcation Serial No. 08/025,950, filed March 3, 1993) ofthe present application. Alternatively, additional superconductors and magnets can be added to the embodiments disclosed herein without departing from the invention.

In another preferred embodiment ofthe invention, rotational losses are lessened by using permanent magnets 12 and 16 with a high degree of rotational symmetry. To a first approximation, the high temperature superconductors 14 and 18 experience a constant magnetic field and no loss. The only actual losses are the hysteresis and eddy currents in the rotating rotor 20 (which preferably comprises steel) which are extremely small. High mechanical strength ofthe rotor 20 allows the rotor 20 to spin very fast without structural failure or degradation.

During experiments on the mixed-mu high temperature superconductor bearing 10, it was noted that some geometries exhibit a reluctance instability. The instability is especially prominent for geometries ofthe rotor 20 in which the radius is larger than the height of a cylindrical shell or tube-shaped rotor 20.

The basic configuration ofthe reluctance instability is illustrated by FIGS. 2 A and 2B. The preferred levitational orientation ofthe rotor 20 is shown in FIG. 2A. The air gap in a magnetic circuit in the vicinity of the rotor 20 is composed of distances 22a plus 24a. FIG. 2B illustrates the position ofthe rotor 20 when the instability is active and the rotor 20 tilts. It is seen here that the effective air gap 22b plus 24b is diminished compared with FIG. 2A. Because the mu ofthe rotor is so much larger than that of air, the magnetic flux prefers to go directly through smaller air gaps 22b and 24b and diagonally through the

rotor 20. In other words, the total reluctance decreases when the rotor 20 tilts. This is the basis ofthe reluctance instability.

It has been shown that long narrow cylinders or shells can be stably levitated, whereas squat cylinders or rings undergo substantial reluctance instability. However, for many practical applications, it is desirable to construct the rotors 20 with short height. Preferred embodiments ofthe invention can overcome the reluctance instability in the mixed-mu high temperature superconductor bearing 10.

As an indication ofthe low friction potential ofthe mixed-mu high temperature superconductor bearing 10, the coefficient of friction (hereinafter "COF") as a function of the rim velocity ofthe rotor 20 is shown in FIG. 3. In this case, the rotor 20 comprised a Mn-ferrite sintered material. The advantage of such a rotor 20 over a steel form ofthe rotor 20 is that the electrical conductivity ofthe ferrite is much smaller, so that eddy current losses are small. No special attempt was made to use uniform permanent magnets in this experiment, and the losses could be made even smaller if a more homogeneous set ofthe permanent magnets 12 and 16 were used. A linear extrapolation ofthe COF shows that a rim velocity greater of 100 m/s would be reached before the COF reached 1.0 x 10"6. Such a low value (compared to prior art designs) verifies this structure can be useful for a low-loss flywheel energy storage system.

Another preferred embodiment ofthe present invention locates the rotor 20 such that the rotor 20 is adjacent an air gap in the radial or azimuthal direction. If the rotor 20 begins to tilt, the total reluctance ofthe magnetic system is increased. The ferrite rotor 20 that was used in the experiment (whose results are shown in FIG. 3) is preferably constructed in this manner. Preferably, a significant gap of nonmagnetic material is provided between each ofthe ferrite grains.

In all embodiments ofthe present invention where permanent magnets are not part ofthe rotor 20, the rotor 20 is preferably mechanically constrained before and during (but not after) the field cooling ofthe superconductors 14 and 18. It will be apparent that the rotor 20 can comprise any material with high magnetic permeability (e.g., ferrite, magnetic steels or nickel and its alloys) and is preferably of high mechanical strength. Further, the

rotor 20 can comprise a wide variety of shapes including, but not limited to, cylinders, rings and even non-circular shapes.

The rotor 20 can also comprise a variety of materials in various shapes and configurations. For example, electrically nonconductive magnetic ferrite can be sintered into the shape desired for the rotor 20. Alternatively, ferrite powder can be mixed or placed into an epoxy or other nonconducting, nonmagnetic matrix. Another preferred embodiment uses steel, nickel, or other conductive powder in a sintered form or in an epoxy or other nonmagnetic matrix. Yet another embodiment uses the ferrite powder as ballast in the winding of a fiber composite ring (or other desired shape). Composite rings can be formed by incorporating a powdered mass or "ballast" into the matrix material (usually epoxy). By changing the density ofthe composite as a function of radius, radial stresses in the final composite can be minimized or eliminated. Because the composite is mechanically weakest in the radial direction, this allows the composite to spin faster before breaking. By using ferrite (or steel) powder as the ballast, the mechanical strength ofthe bearing rotor 20 is maximized and there is a significant radial (and circumferential) reluctance to overcome any reluctance instability.

A further preferred embodiment of the invention embodiment is shown in a horizontal cross-sectional view in FIG. 4. The figure is shown for the rotor 20 having a ring geometry 30, but can also take the form of a cylinder or other similar shape. Alternate steel (or ferrite) wedges 32 are separated by nonmagnetic wedges 34 such as epoxy or composite-epoxy. The entire ring 30 can be optionally surrounded by a composite ring 36 to increase the overall mechanical integrity when the rotor 20 is spun to high speeds. The nonmagnetic wedges 34 provide the desired reluctance in the c cumferential direction.

A further embodiment based around the geometry shown in FIG. 4 uses the steel wedges 32 constructed from nonisotropic laminations so that there is a higher magnetic permeability mu value in the vertical direction than in the azimuthal direction. Another preferred embodiment of this type is shown in FIG. 5, in which a series of steel (or ferrite) pins 38 is located within a nonmagnetic matrix 40. The pins 38 can also be staggered in the azimuthal direction as shown in Fig 6.

A disadvantage ofthe geometries shown in Figs. 4-6 is that the magnetization ofthe rotor 20 is not homogeneous in the azimuthal direction. This inhomogeneity will induce hysteresis losses in the high temperature superconductor components as the rotor 20 rotates. To partially overcome this disadvantage, the top and bottom ofthe rings 30 can be covered by a ferromagnetic plate 42. The thickness ofthe ferromagnetic plate is preferably thin relative to the height ofthe steel wedges 32 or pins 38. The ferromagnetic plate 42 smooths out the inhomogeneities experienced by the high temperature superconductors 14 and 18.

Another preferred embodiment shapes the bottom ofthe rotor 20, as shown in FIG. 7, so that the reluctance increases if the rotor 20 is tipped. While this simplifies the construction ofthe rotor 20, it complicates the high temperature superconductors 14 and 18 and high temperature superconductor-cryostat construction. It is desirable to have the high temperature superconductors 14 and 18 as close to the rotor 20 as possible in order to maximize stability. The preferred embodiment of FIG. 7 can provide this advantageous result.

One ofthe disadvantages ofthe geometry shown in FIG. 1 is that the levitational force is not very large. If the magnetic field is large enough to saturate the magnetization ofthe rotor 20, it will have nearly constant magnetization M. Then the levitation force F per unit volume will be given by

F = M dH/dz, where H is the apphed magnetic field and z is the vertical coordinate. Ideally, one desires dH/dz to be uniform over the volume ofthe rotor 20. If dH/dz is not uniform, then the magnetic system produces a magnetic stiffness on the rotor 20. This stiffness is always destabilizing. To stabilize the rotor 20, dH/dz should be as large as possible, with d^H/dz^ as small as possible.

An embodiment to achieve large levitation force in a mixed-mu high temperature superconductor bearing is shown in FIG. 8 for a ring geometry. The bearing assembly 50 comprises an upper permanent magnet 52, a lower permanent magnet 54, an upper high temperature superconductor 56, a lower high temperature superconductor 58, an iron flux

return path 60, and a ring-shaped ferromagnetic rotor 62. The dimensions ofthe permanent magnets 52 and 54 are chosen so as to minimize d^H/dz^. In a variation on this embodiment additional permanent magnets and iron pole pieces 64 can be inserted into the system to further minimize d^H/dz^ as shown for example in FIG. 9.

The preferred embodiments ofthe present invention provide a low-loss mixed-mu high temperature superconductor bearings that can be used to achieve a high-efficiency flywheel energy storage device. The configurations also overcome reluctance instabilities in mixed-mu high temperature superconductor bearings, are able to obtain high speeds, and high levitation pressures.

While preferred embodiments ofthe invention have been shown and described, it will be clear to those skilled in the art that various changes and modifications can be made without departing from the invention in its broader aspects as set forth in the claims provided hereinafter.