U.S. patent number 3,768,054 [Application Number 05/240,350] was granted by the patent office on 1973-10-23 for low flux leakage magnet construction.
This patent grant is currently assigned to General Electric Company. Invention is credited to Wendell Neugebauer.
United States Patent |
3,768,054 |
Neugebauer |
October 23, 1973 |
LOW FLUX LEAKAGE MAGNET CONSTRUCTION
Abstract
A magnet construction is disclosed wherein the flux of a first,
principal magnet is conserved by placing a second magnet adjacent
the first magnet with the magnetic axes of the two magnets
perpendicular to each other. The second magnet is constructed of a
highly anisotropic material having low permeability perpendicular
to its magnetic axis and preferably having a high coercive force
and good magnetization retention. Preferably the second magnet
surrounds the first magnet to minimize the flux leakage of the
first magnet.
Inventors: |
Neugebauer; Wendell (Ballston
Spa, NY) |
Assignee: |
General Electric Company
(Owensboro, KY)
|
Family
ID: |
22906178 |
Appl.
No.: |
05/240,350 |
Filed: |
April 3, 1972 |
Current U.S.
Class: |
335/304;
310/254.1; 335/306 |
Current CPC
Class: |
H02K
23/04 (20130101); H02K 1/17 (20130101); H02K
21/14 (20130101); H01F 7/0278 (20130101) |
Current International
Class: |
H02K
23/04 (20060101); H02K 1/12 (20060101); H02K
1/17 (20060101); H02K 21/14 (20060101); H02K
23/02 (20060101); H01F 7/02 (20060101); H01f
007/02 () |
Field of
Search: |
;335/210,302,304,306
;250/49.5D ;310/254,261 |
References Cited
[Referenced By]
U.S. Patent Documents
Other References
W Hofman, Strayfield Neutralizing, IEEE Transaction on Magnetics,
Vol. 2, No. 2, pp. 279, 280, June, 1970..
|
Primary Examiner: Harris; George
Claims
What is claimed is:
1. A permanent magnet construction having a minimum leakage flux
comprising a first permanent magnet having a predetermined lineal
extent and a magnetic axis therealong and a second permanent magnet
forming a cladding completely about said first magnet and
coextensive therewith, said second magnet comprising a highly
anisotropic material having its magnetic axis substantially
perpendicular to the magnetic axis of said first magnet and having
low permeability in the direction normal to its magnetic axis to
inhibit leakage of magnetic flux from said first magnet.
2. The magnet construction of claim 1 wherein said second magnet
comprises a rare earth-cobalt alloy.
3. The magnet construction of claim 1 wherein said second magnet
comprises a material exhibiting good magnetization retention
properties.
4. A magnet construction comprising a first magnet having a
predetermined lineal extent and a magnetic axis therealong and a
second magnet comprising highly anisotropic material having a low
permeability normal to its magnetic axis, said second magnet
adjoining substantially the entire extent of said first magnet and
substantially completely surrounding the magnetic axis of said
first magnet, the magnetic axis of said second magnet being
substantially perpendicular to the magnetic axis of said first
magnet.
5. The magnet construction of claim 4 wherein the magnetic
potential of the second magnet is not uniform along the magnetic
axis of the first magnet.
6. The magnet construction of claim 5 wherein the magnetic
potential of the second magnetic is strongest adjacent at least one
end of the magnetic axis of the first magnet.
7. The magnet construction of claim 5 wherein the thickness of the
second magnet perpendicular to the magnetic axis of the first
magnet is tapered to provide a non-uniform magnetic potential along
the magnetic axis of the first magnet.
8. The magnet construction of claim 5 wherein said second magnet
comprises a rare earth-cobalt alloy.
9. The magnet construction of claim 4 wherein said first magnet
comprises a plurality of magnets aligned along a common magnetic
axis.
10. The magnet construction of claim 4 wherein said second magnet
comprises a plurality of magnets, each having its magnetic axis
normal to the magnetic axis of said first magnet.
11. The magnet construction of claim 4 wherein said first magnet
comprises a pair of coaxially aligned hollow cylinders disposed in
mutually axially spaced relation, and said second magnet is in the
form of a double cone having a wide-thickness portion surrounding
the adjacent ends of said cylinders, said wide-thickness portion
being of substantially uniform thickness over an axial extent
thereof corresponding to the spacing of said cylinders.
12. The magnet construction of claim 11 wherein said cylinders of
said first magnet are magnetized in opposite directions and said
construction further comprises a radially extending reversal pole
piece disposed within the space between said cylinders.
13. The magnet construction of claim 12 wherein the cylinders are
adapted to surround a linear-beam electron discharge device.
14. The magnet construction of claim 4 wherein said first magnet
comprises a pair of spaced aligned solid bars of uniform diameter
and having a common direction of magnetization, and said second
magnet comprises a pair of cylinders having a common axis and
uniform internal diameters conforming with the diameters of said
bars, the magnetic axes of said cylinders being in mutually opposed
directions and normal to the magnetic axis of said bars, and a
magnetic shell encasing said first and said second magnets to form
a complete magnetic circuit.
15. The magnet construction of claim 14 wherein the space between
said bar magnets is adapted to be occupied by a crossed-field
electron discharge device.
Description
BACKGROUND OF THE INVENTION
Permanent magnets are normally characterized as materials having
magnetic flux producing properties generally along an axis.
However, in actual practice, some of the flux is generated in
directions skewed from the main axis. This may be regarded as
leakage or lost flux since the flux so generated contributes
little, if any, to the desired magnetic intensity. The extent of
this loss varies with the type of material as well as the geometric
design of the magnet. In any event the usual remedy is to increase
the total size of the magnet to achieve the desired magnetic
intensity. This, however, has the undesirable effect of adding
excessive weight to the magnetic system.
SUMMARY OF THE INVENTION
It is therefore an object of this invention to provide an improved
magnet construction wherein the leakage flux is minimized.
It is another object of the invention to provide a magnet
construction having an improved magnetic intensity per unit weight
ratio.
These and other objects of the invention will be apparent from the
specification and accompanying drawings.
In accordance with the invention the magnetic flux of a first
magnet is conserved by placing a second permanent magnet adjacent
the first magnet with the magnetic axis of the second magnet
perpendicular to the axis of the first magnet to prevent leakage of
flux from the first magnet. The second magnet comprises a highly
anisotropic material having a low magnetic permeability in a
direction perpendicular to its magnetic axis. Preferably, the
second magnet comprises a high coercive force material with good
magnetization retention properties.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a cross-sectional view of a bar magnet constructed in
accordance with the invention.
FIG. 2 is a cross-sectional view of a magnet for a klystron
constructed in accordance with the invention.
FIG. 3 is a cross-sectional view of a magnet for a voltage tunable
magnetron constructed according to the invention.
FIG. 4 is a cross-sectional view of a modified version of the
magnet construction of FIG. 3.
FIG. 5 is a cross-sectional view of a loud speaker magnet
constructed according to the invention.
FIG. 6 is a cross-sectional view of a portion of a motor having a
rotor magnet construction in accordance with the invention.
FIG. 7 is a cross-sectional view of a portion of a motor having a
stator magnet construction in accordance with the invention.
FIG. 8 is a graph showing B-H curves for magnetic materials.
DESCRIPTION OF THE INVENTION
Turning now to FIG. 1, the invention is illustrated as comprising a
central bar magnet 2 and a second magnet 10.
Bar magnet 2, in the illustrated embodiment, comprises a solid rod
with the magnetic axis generally along the mechanical axis of the
rod. Magnet 2 can be made of any conventional magnet material such
as, for example, steel or Alnico.
Magnet 10 is of generally conical configuration with a central bore
therein conforming to the diameter of magnet 2 so that magnet 10
may be slipped over magnet 2. Magnet 10 is preferably conically
shaped or tapered for a reason which will be described below.
Magnet 10 can be permanently attached to magnet 2, for example, by
bonding the magnets together with generally magnetically
transparent means such as epoxy cement or the like. In accordance
with the invention magnet 10 is constructed of a highly aniostropic
material having a low permeability normal to its magnetic axis.
Preferably, the material also should exhibit a high coercive force
and good magnetic retention properties. The latter property is
particularly desirable to prevent demagnetization or alteration of
the magnetic properties of magnet 10 by magnet 2. Materials which
possess these properties and which have been found to be useful in
the practice of this invention include the rare earth-cobalt alloys
described and claimed in U.S. Pat. Nos. 3,655,463 of Apr. 11, 1972;
3,655,464 of Apr. 11, 1972; 3,684,593 of Aug. 15, 1972; and
3,695,945 of Oct. 3, 1972; and all assigned to the assignee of this
invention.
The use of the term "rare earth-cobalt alloys" is intended to
include one or more of the rare earth elements alloyed with cobalt.
The term "rare earth" is intended to include the 15 elements of the
lanthanide series having atomic numbers 57-71 inclusive. The
element yttrium (atomic number 39) is commonly included in this
group of metals and is therefore to be considered, in this
specification, as also included in the term "rare earth".
As shown by the arrows in FIG. 1, magnet 10 is in accordance with
the invention, radially magnetized so that its direction of
magnetization is generally perpendicular to the magnetic axis of
magnet 2. Any skewing of the flux lines generated by magnet 2 are
thus corrected or prevented from "leaking" out of the sides of
magnet 2. This conservation or increase of the available flux in
magnet 2 enables one to reduce the overall size, and thus the
weight, of magnet 2 to a size commensurate with the desired amount
of flux without the previous necessity of additionally compensating
for flux losses by "leakage" of stray flux.
The magnetic material used in accordance with the invention for
magnet 10 must have low permeability in a direction normal to the
magnetic axis, i.e. parallel to the magnetic axis of the main
magnet 2 to prevent shorting the field of the main magnet.
The term low permeability is intended to define a relative
permeability of the cladding magnet material in a direction normal
to its magnetic axis more like that of vacuum than that of
ferromagnetic materials.
The cladding magnet material must also, in accordance with the
invention, have a high anisotropy, i.e., the material should be
highly magnetizable only along one axis. The degree of anisotropy
of a material can be found by determining the alignment factor
which is the residual induction (B.sub.r) of the material divided
by the saturation magnetization (4.pi.J.sub.s). If a theoretically
perfect alignment factor was assigned a value of 1 and an alignment
factor representing completely random alignment assigned a value of
0.5, the alignment factor representing highly anisotropic material
would have a value of at least 0.80 and preferably 0.95 or
higher.
As previously mentioned, the magnetic material used in constructing
cladding magnet 10 preferably should have a relatively high
coercive force to provide the required magnetic potential necessary
for operation of the invention. Coercive force is a material
property defined as the field strength (H.sub.c) at which the
magnetic induction (B) becomes zero. High coercive force material,
then, is defined in this specification as referring to a strongly
magnetizable material requiring a coercive force of at least 2 Kilo
oersteds, and preferably 4 Kilo oersteds, to reduce the magnetic
induction to zero.
Referring to FIG. 8, B-H curves in the second quadrant of the
hysteresis loops for various magnetic materials are plotted using
the EMU system of units. It will be seen that the slopes of the
curves vary, with the cobalt-rare earth curves approaching a
45.degree. slope, the theoretically ideal condition. The ratio of
the value of H.sub.c to the residual induction B.sub.r the magnetic
induction B corresponding to zero magnetizing force H is an
indication of the magnetization retention properties of the
material. It can be seen that the value of H for the Alnico
materials with respect to B.sub.r is much lower than that of the
CoPt, ferrite and cobalt-rare earth materials and, in fact, such
Alnico materials are unsuitable for use as cladding magnets in
accordance with the invention. The term good magnetization
retention properties, then, is defined in this specification as
referring to a material having a value of H.sub.c at least 60% of
B.sub.r (when expressed in the EMU system in Oersteds and gauss
respectively) and preferably 80-90% or higher.
Referring now to FIG. 2, a specific application of the teachings of
this invention is illustrated. A magnet comprising a single
reversal permanent magnet circuit for a linear beam tube is shown.
Cylinders 13 and 14 may be constructed of conventional magnetic
material and are magnetized axially and placed in opposition as
shown. Cylinders 13 and 14 produce the flux that passes through
pole pieces 15 and 16 and the open space 19 to be occupied by a
linear beam tube such as a klystron or traveling wave tube.
Reversal pole piece 17 carries twice the flux as either of the end
pole pieces.
In accordance with the invention, magnetic cylinders 13 and 14 are
jacketed by a magnetic member 20 generally shaped as a double
cone.
Magnetic member 20 is constructed of magnetic material having the
properties described above for magnet 10. Member 20 is radially
magnetized, as indicated in the figure, substantially perpendicular
to the flux lines of cylindrical magnets 13 and 14. In this
embodiment, the second, flux conserving, magnet 20 is not
constructed of uniform thickness. Since the magnetic potential in
cylinders 13 and 14 varies linearly with distance from the
respective end pole pieces 15 and 16, the variation in thickness of
the cladding 20 must be linear along the axis of magnets 13 and 14.
Stated another way, inasmuch as the magnetic potential increases
along cylinders 13 and 14, the magnetic counter potential necessary
to prevent leakage of flux from cylinders 13 and 14 must be
increased by increasing the size or thickness of cladding 20. The
magnetic potential of the reversing pole 17 is constant and
therefore the thickness of member 20 is constant in this region.
Iron shell 18 carries no flux and is merely added to keep external
fields from penetrating into the structure.
The thickness of member 20 at any point is dtermined by the
magnetic potential on the principal magnet 13 or 14 as well as the
coercive force of the material used in constructing member 20. The
shape and configuration of the cladding magnet or magnets can thus
be determined either empirically or by calculations using the
magnetic potential of the principal magnet and the magnetic
properties of the cladding magnet material. These parameters can,
in turn, be determined by referring to the B-H curves for the
particular materials used.
Thus, while I do not wish to be bound by any mathematical theories
concerning the magnetic forces, it is believed that the shape of
member 20 can be calculated by first determining the magnetic
potential along the main magnets 13 or 14 by applying computational
methods well known to those skilled in the art and then applying
the results to the following formula:
T = U/H.sub.c
where T is the thickness of the cladding at any given point; U is
the magnetic potential (in gilberts in the EMU system of units) at
that point on the surface of the main magnet; and H.sub.c is the
coercive force of the cladding material.
Since the addition of the cladding material to the original
magnetic circuit alters the flux pattern previously computed, it is
essential that the computation be repeated with the cladding in
place. The new solution thus obtained will lead to a slightly
different magnetic potential distribution U on the surface of the
main magnets. The cladding thickness is then slightly adjusted
according to the above formulation. This process is repeated until
a solution is obtained that is self-consistent within the desired
degree of accuracy.
To demonstrate the method of computing the cladding thickness, the
circuit of FIG. 2 will be analysed. Still referring to FIG. 2, the
flux density B is specified to be axially directed with a complete
reversal of direction in the center of the structure. The magnetic
field vector H is therefore also axially directed and has a
magnitude equal to B in the EMU system of units in the open space
19. Since no surface current is flowing at the boundary between the
open space 19 and the cylinders 13 and 14, the field vector H is
the same in the material of cylinders 13 and 14 as in the space 19
by the law of continuity. The flux density of the material of
cylinders 13 and 14 is now established by referring to the
appropriate B-H characteristic of the material with H being known.
Designating B.sub.m as the magnitude of flux density in the
material of the cylinders 13 and 14 and B the magnitude of flux
density in the open space 19, the rule that the divergence of the
flux density vanish implies that:
A.sub.m B.sub.m = AB
where A.sub.m is the transverse area of the material comprising
cylinders 13 and 14 measured in a plane perpendicular to the axis
of these cylinders, and A is the transverse area of the open space
19 measured in the same plane. This relation yields the value of
A.sub.m because all the other factors are known or specified. From
this value of A.sub.m the thickness of the main flux producing
cylinders is determined.
The thickness of the cladding at any point X on the surface of the
main flux producing cylinders is established by first determining
the magnetic potential U along the outer surface of the cylinders
13 and 14 and then referring to the appropriate B-H curve of the
cladding material to obtain the coercive force H.sub.c which is the
value of H when B is equal to zero. The values of U and H.sub.c are
applied to formula (1) to determine the thickness of the cladding
at the point X. A similar procedure yields an equation for the
cladding thickness along cylinder 14. Since the flux patterns were
known beforehand, no repetition of the calculation is necessary in
this example, the given solution already being self-consistent.
Although the method of computation outlined above yields a definite
value of cladding thickness as a function of geometrical location,
a thickness different from the one computed may be used. In
particular, changes in the thickness lead to changes in the flux
density in the open space to be occupied by a specific device, and
thus afford a valuable means of adjusting the flux density for
optimum device performance. In general, a lesser thickness of
cladding yields a lower flux density and vice versa. The use of
cladding thicknesses other than computed may result in some leakage
flux which, however, is still far below the leakage flux normally
associated with unclad magnetic circuits. It should also be noted
here that non-tapered cladding, in some instances, may be used for
manufacturing efficiencies.
Referring now to FIG. 3, another magnetic construction, in
accordance with the invention, is illustrated in a structural form
most useful for crossed-field devices such as magnetrons,
voltage-tunable magnetrons (VTMs), and crossed-field amplifiers
(CFAs).
In this construction the useful flux is produced by cylindrical
bars 34 and 35 magnetized in the same direction. This flux passes
through the space 36 to be occupied by the crossed-field device and
completes the circuit through iron shell 31.
In accordance with the invention, cladding magnets 32 and 33
comprising radally magnetized cylinders sruuounding principal
magnets 34 and 35. Again the thickness of magnets 32 and 33 is not
uniform but is rather directly proportional to the magnetic
potential along the flux producing bars 34 and 35 or in the air gap
36. The cladding thickness, therefore, is zero at the very center
of air gap 36 because the magnetic potential is zero there. This
proper adjustment of the thickness of the cladding magnet 32 and 33
results in no flux being carried by these magnets. Magnets 32 and
33 are constructed of magnetic materials having the properties
previously described with respect to cladding magnet 10. FIG. 4
illustrates an alternate construction to FIG. 3 and useful in the
crossed-field devices previously described. In this embodiment
magnets 34', 35', 37 and 38 are each shaped as frustums of a cone.
Magnets 34' and 37 together comprise one main magnet while magnets
35' and 38 form the other main magnet.
Cladding magnets 32' and 33' have an outer cylindrical shape but
are each provided with a tapered center bore conforming to the
frustum shapes of the respective main magnets. An iron shell 31'
surrounds the magnet construction to serve as a return path and to
shield the magnets from penetration by external fields.
To further illustrate the practice of the invention, a magnet was
constructed as shown in FIG. 4. Frustums 34' and 35' were
constructed of Alnico 9 material with a base diameter of 1.25
inches, a diameter of 0.875 inch at the smaller end, and 1.013
inches thickness. Magnets 37 and 38 were constructed of Co.sub.5 Sm
alloy having a base diameter of 1.390 inches, a diameter at the
smaller end of 1.25 inches, and a thickness of 0.375 inch. Magnets
34', 35', 37 and 38 were all magnetized axially with the North
poles of magnets 35' and 38 facing the South poles of magnets 34'
and 37 as indicated by the arrows in FIG. 4.
Cladding magnet 32' was constructed of Co.sub.5 Sm alloy segments
forming a cylinder having an outer diameter of 1.65 inches and a
length or thickness of 1.388 inches. Magnet 32' was magnetized
normal to its cylindrical axis.
The segments forming magnet 32' were provided with an internal
taper to provide a conical bore to snugly receive frustum main
magnets 34' and 37. The segments of magnet 32' were retained to
magnets 34' and 37 using Eastman 910 cement.
The segments forming magnet 33' were similarly assembled about main
magnets 35' and 38 and the clad magnet assemblies were mounted in
iron shell 31' with the small ends of magnets 34' and 35' facing
one another in coaxial alignment with a spacing therebetween of
0.550 inch.
Measurements of the flux density were made and found to be about
7,400 gauss as compared to a theoretical value of about 8,000
gauss. While the flux densities of main magnets 34', 35', 37 and 38
were not actually measured without cladding, standard magnets of
this type, size, and gap are normally found to have flux densities
below 5,000 gauss.
In FIG. 5 a loudspeaker magnet constructed in accordance with the
invention is illustrated. The main flux producing member is a
radially magnetized disk 45 which is clad with axially magnetized
shaped disks 42, 43, and 44. The pole pieces 46 and 47 serve to
concentrate the flux to a density that is higher than that capable
of being produced by the permanent magnet material comprising disk
45. The large, generally circular, block 41 comprises an iron
return path for the flux from the main magnet 45.
The cladding or insulating magnets 42, 43, and 44 are, in
accordance with the invention, constructed of magnetic materials
having the properties previously described with respect to magnet
10. The axial magnetization of magnets 42, 43, and 44 produces a
magnetic counter potential perpendicular to that of radially
magnetized disk 45. The thickness of the cladding magnets 42, 43,
and 44 is, again, proportional to the magnetic potential on the
surface of disk 45. Since this is not linear along the disk, the
thickness of magnets 42, 43, and 44 is not a linear function but is
rather determined by the B-H characteristics of disk 45 and the
remainder of the magnetic circuit geometry.
It should also be noted here that in actual practice voice coil 48
must have some means of external communication and thus magnet 44
must be either modified or eliminated. This will result in a minor
amount of leakage flux. The amount of flux lost, however, with
respect to the amount conserved is small and thus the object of the
invention to minimize leakage flux is still carried out by magnets
42 and 43.
In FIG. 6, a motor is generally shown having a permanent magnet
type rotor comprising radially magnetized spokes 52 having
alternate directions of magnetization. In accordance with the
invention flux leakage is minimized by the insertion of
circumferentially magnetized sections 54 between spokes 52
comprising materials having the magnetic characteristics previously
described with respect to magnet 10. It should be pointed out
again, however, that the principal magnets, in this case spoke 52,
can be constructed of any magnetic materials depending upon the
flux density desired. The direction of magnetization alternates for
sections 54 to provide the desired flux leakage bucking system in
accordance with the invention.
Alternatively, as shown in FIG. 7, the stator may be constructed as
a permanent magnet structure comprising radially magnetized spokes
62 clad in circumferentially magnetized sections 64. As discussed
in previous embodiments each section 64 is not of uniform thickness
but is rather tapered to provide a profile of strongest magnetic
potential adjacent the end of each spoke 62 facing the rotor. The
amount of taper and the thickness is again determined as previously
described.
For the correct choice of thickness of cladding magnets 64 the
space 65 between spokes 62 will be a magnetic field free region and
may therefore be filled with a highly permeable material. This
space may, alternatively, be used by contouring the stator return
shell 66 to fill this region.
The cladding magnets used to provide magnetic potential in
accordance with the invention to inhibit leakage of flux from the
main magnet are dimensioned to provide the correct amount of
counter potential in accordance with the potential along the main
magnet as well as the coercive force of the particular material
used for the cladding magnet. The materials used for the cladding
magnets must have the particular magnetic properties previously
described with respect to magnet 10.
The cladding magnet may be shaped to the desired contour by
pressing the magnetic material in particulate form to the desired
shape followed by sintering of the shaped magnet. The magnetic
alignment of the particles before sintering and the magnetization
of the sintered product are carried out as described and claimed in
the aforementioned pending patent applications.
The material can also be ground or machined to the desired shape.
While this would normally be done before magnetization because of
the practical problems experienced when processing magnetized
materials, the desired high anisotropy of the material may make it
desirable to magnetize the material before fabrication. When, as in
the preferred embodiment, a magnetic material having high
magnetization retention properties is used, processing can be done
after magnetization with minimal risk of demagnetizing the cladding
material.
The resultant cladding magnets are joined to the main magnet using,
for example, bonding means such as epoxy cement. Other bonding or
mechanical retention means can be used provided they do not
interfere with the magnetization of either the main magnet or the
cladding magnets. In certain circumstances it may be desirable to
remagnetize the main magnet in situ, i.e. after being clad when a
material havin a low intrinsic coercive force is used for the main
magnet. Since, as previously described, the cladding magnet
material is preferably characterized by a high magnetization
retention, the choice of bonding methods or materials will have
little, if any, effect on its magnetization nor will
remagnetization of the main magnet effect the cladding magnet.
Thus, the invention provides an improved magnet structure wherein
the flux of a principal magnet is conserved by cladding the magnet
with a second magnet having its magnetic axis perpendicular to the
magnetic axis of the first magnet and further characterized as
preferably having a high coercive force, with low permeability
perpendicular to its magnetic axis, and highly anisotropic.
Preferably, the aterial is also characterized as highly resistant
to alteration of its magnetic properties. The cladding magnet is
preferably contoured as previously described to provide a
perpendicular, bucking, magnetic force proportionate to the
potential along the principal magnet.
* * * * *