Vacuum-type circuit interrupters having an axial magnetic field produced by condensing shield coils

[57] ABSTRACT

A vacuum-type circuit interrupter is provided having an axial magnetic field provided therein through the use of currents transmitted to the surrounding shield structure, the latter being electrically connected by means of a field coil to one of the separable contacts.

In one embodiment, the surrounding condensing shield is connected by an internal coil strap to one of the separable contacts. In another embodiment, the coil, providing the axial magnetic field, is disposed externally of the evacuated envelope, and is connected by a connection extending through the envelope to the condensing shield, which may be electrically isolated internally from the electrodes or contacts. In another embodiment, the axial magnetic field-producing coil is disposed externally of the envelope, but has a connection to an end plate, which, additionally, is electrically connected to the internal condensing shield.

Another embodiment has the condensing shield provided in the form of a helical conducting strap. In another embodiment, a split condensing shield is connected by strapcoil connecters to both of the separable contacts of the vacuum interrupter for creating an axial magnetic field. In another embodiment, the “floating” shield has a generally spiral configuration adjacent to the arcing gap. The “floating” shield will draw net electron current on the cathode side of the shield, and a compensating ion current on the anode side. The resulting circulating current will cause an axial magnetic field to be set up interiorly of the vacuum interrupter.

In yet another embodiment of the invention the condensing shield is connected by a field coil to one of the fixed electrodes of a fixed-gap vacuum device to assist in extinguishing the arc established between the two fixed electrodes.

Claims, 29 Drawing Figures

<img file="US3889079A_D0001.tif" />

<img file="US3889079A_D0002.tif" />

<img file="US3889079A_D0003.tif" />

<img file="US3889079A_D0004.tif" />

<img file="US3889079A_D0005.tif" />

PATENTED JUMIΟ 1975

SHEET 4

3,889'079

<img file="US3889079A_D0006.tif" />

<img file="US3889079A_D0007.tif" />

FIG. 9

P/Wwimo 1975

SHEET 5

3,889,079

<img file="US3889079A_D0008.tif" />

<img file="US3889079A_D0009.tif" />

PATFMTFPJU:,: I Ο 1973

SHEET 6

3.889,079

FIG.8A

<img file="US3889079A_D0010.tif" />

<img file="US3889079A_D0011.tif" />

B MAGNETIC FIELD (GAUSS)

INCREASING NUMBER OF TURNS

<img file="US3889079A_D0012.tif" />

ΡΛΤΓΠϊ^η JUK10 1975

3,889,079

SHEET

<img file="US3889079A_D0013.tif" />

FIG.I2A

------------—r

MOVABLE ELECTRODE — (CATHODE) ^L

CLOSING

<img file="US3889079A_D0014.tif" />

^CAPACITOR BANK

<img file="US3889079A_D0015.tif" />

ARC VOLTAGEFIG. 12

<img file="US3889079A_D0016.tif" />

C

<img file="US3889079A_D0017.tif" />

R4

TEST INTERRUPTER (FIG. 12 A) _L-0.00l OHM-ION CURRENT

VOLTAGE DIVIDER

<img file="US3889079A_D0018.tif" />

<img file="US3889079A_D0019.tif" />

SHIELD CURRENT MEASURING SHUNT

ARC CURRENT MEASURING SHUNT 2.5XI0^-4 OHM

ARC VOLTAGE MEASURED VIA VOLTAGE DIVIDERS

ΡΑΤΠ.ΐΤΓΠ n .

SHEET 8

3,889,079

<img file="US3889079A_D0020.tif" />

Arc Vol

FIG. 20

<img file="US3889079A_D0021.tif" />

3.889^079

<img file="US3889079A_D0022.tif" />

FIG.I4

PATENTFO J.UH 1 0 1975.

7889,079

Peak Shield Ion Current, kA

<img file="US3889079A_D0023.tif" />

FIG. 17

<img file="US3889079A_D0024.tif" />

FIELD COIL

FIG.I8

<img file="US3889079A_D0025.tif" />

6-TURNS IN

FIELD COIL

<img file="US3889079A_D0026.tif" />

i o 1975 3,889,079

SHEET 11

Arc Power, MW RATE OF RISE OF RESTORED VOLTAGE, KV/XS

<img file="US3889079A_D0027.tif" />

Arc Voltage,

FIG.2I

PATENTEilJUNlO 1975

SHEET

ACTS

LIKE

AUX.

CATHODE

NET ION CURRENT APPROX. 10% OFI

7ZZZZZZL

BIASED NEGATIVE RELATIVE TO ARC PLASMA

NET ION CURRENT

BIASED POSITIVE RELATIVE TO ARC PLASMA

<img file="US3889079A_D0028.tif" />

c

ANODE

X77777.

ARC VOLTAGE REDUCED

-ACTS LIKE AUX.

ANODE

-2.

NET ION CURRENT

FIG.23

NET ELECTRON CURRENT

CATHODE

ANODE ©-*

ARC VOLTAGE REDUCED

CATHODE

NET ELECTRON CURRENT

FIG24

PATENTED JUM 0 1975

3,889,079

SHEET

ANODE

<img file="US3889079A_D0029.tif" />

(CATHODE

<img file="US3889079A_D0030.tif" />

FIG.25

AXIAL MAGNETIC FIELD

ANODE

DIRECT CURRENT ARC (1,000°) (CATHODE

<img file="US3889079A_D0031.tif" />

FIG.26

NO AXIAL MAGNETIC FIELD

DIRECT CURRENT ARC (1,000°)

3,889,079

VACUUM-TYPE CIRCUIT INTERRUPTERS HAVING AN AXIAL MAGNETIC FIELD PRODUCED BY CONDENSING SHIELD COILS

CROSS-REFERENCES TO RELATED APPLICATIONS

Applicants are not aware of any related patent applications pertinent to the present invention.

BACKGROUND OF THE INVENTION

It has been known by those skilled in the art that certain benefits are to be achieved by utilizing an axial magnetic field interiorly of a vacuum “bottle,” or vacuum-type circuit interrupter. For example, reference may be made to U.S. Pat.: Lee 3,321,599; British Pat. No. 1,258,015; French Pat. No. 1,415,441; Polinko, Jr. et al. No. 3,345,484; Lucek et al. U.S. Pat. No. 3,263,162. Generally, the coils have been in series with the separable contacts of the vacuum interrupter.

Several investigators have disclosed that axial magnetic fields are effective in increasing the current interruption rating of vacuum interrupters. This axial magnetic field is applied parallel to the axes of the electrodes, and the field-producing coils are always placed in series with the current through the electrodes. In patents by Lee U.S. Pat. Nos. 3,321,599 and 3,372,259, 3,372,258 by Porter, the axial magnetic field is produced by a solenoidal coil wound external to the interrupter. In a third U.S. Pat. No. 3,158,722 by Porter, the electrode stems are constructed from materials of widely different electrical conductivities, and an axial magnetic field is produced from these “stem coils.”

The axial magnetic field is effective in increasing the current interruption ability of a vacuum interrupter for two reasons. First, the arc voltage is reduced during the critical half cycle of arcing prior to the natural current zero. This arc voltage reduction minimizes the energy released within the interrupter, and consequently minimizes the temperature rise of the internal interrupter components. In particular, recent experiments indicate that axial fields produce a significant increase in the threshold current associated with anode-spot formation. Thus, in low-current de experiments, where a small diameter anode is drawn to long electrode spacings from the cathode, the threshold current for anodespot formation in the absence of an axial magnetic field is approximately 500 amperes. When the experiment is repeated, with the axial field of approximately 300 gauss, however, the threshold current for anode-spot formation is raised to approximately 1500 amperes. Recent literature reveals that low-current anode-spot experiments are relevant to high-current phenomena in vacuum interrupters, and we therefore feel confident that axial magnetic fields will delay the onset of anodespot formation in a practical alternating-current interrupter.

A second benefit derives from the confining effect of an axial magnetic field. This field reduces the plasma contact with the vapor-condensation shield, and therefore reduces the probability of the arc striking to the condensing shield. This is a particularly important effect since many cost-reduced vacuum interrupters are being manufactured with the vapor-condensation shield connected directly to one of the electrodes. We have shown that the rate of rise of dielectric strength is only decreased slightly in these integral shield designs, but problems may be encountered at high currents (approximately 20 kA) due to the tendency for the arc to strike over to the shield. The strike-over tendency will be small when the shield is connected to the 5 anode electrode during arcing. In fact, this is a beneficial connection since the shield acts as an auxiliary anode and reduces the arc voltage. However, the shield will be biased negative relative to the arc plasma during those arcing half-cycles in which the shield is at cath10 ode potential. This connection does not reduce the arc voltage. Rather, under these circumstances, the arc plasma may initiate cathode spots on the shield with resulting catastrophic failure of the interrupter. We can conclude that axial magnetic fields will definitely prove 15 of value in cost-reduced interrupters, and the field will prove of particular value when the shield is at cathode potential during the arcing half-cycle.

SUMMARY OF THE INVENTION

In the present invention, we are now disclosing a method for applying axial magnetic fields to highcurrent vacuum arcs without the use of a seriesconnected coil. In our system, the “coil” is essentially connected in parallel with the arc, and only conducts current during the arcing half-cycles. The coil is effectively short-circuited during normal alternating-current operation with the electrodes or contacts closed, and thus power losses, due to eddy-current heating of the interrupter end plates, are non-existent. Our embodi³θ ments rely on the fact that currents flow from the arc plasma to the shield when the latter is connected to either anode or cathode potential, this connection being made through a coil. When the shield is connected to anode potential, the shield draws electron current from ³³ the arc plasma as it is this appreciably large electron current flowing through the coils that creates the axial magnetic field. This axial magnetic field lowers the arc voltage, and also reduces, to some extent, the plasma contact with the shield. More importantly, we have dis⁴⁰ covered that large ion currents also flow from the plasma to the shield when the shield is connected to cathode potential. This ion current can be of the order of 10% of the arc current. Thus, for those AC current cycles when the shield is connected via its coil to cathode potential, appreciable axial magnetic fields are also created.

In accordance with preferred embodiments of the present invention, there are provided various condens₅θ ing-shield configurations in which the ion or electron ^υ current, collected by the condensing shields, is utilized in an advantageous manner to result in an axial magnetic field provided interiorly of the vacuum envelope to result thereby in a confinement of the arc without serious anode-spot formation and a rapid extinction thereof following low arc voltage.

In one embodiment of the invention there is provided an internal coil, which is electrically connected between the stationaryy contact, or electrode and the ₆θ internally-disposed condensing shield. In another embodiment of the invention, an externally-located coil is provided being connected electrically between the stationary-contact stem and, by means of a probe connection, to the internally disposed condensing shield.

In still another embodiment of the invention an externally-located coil is electrically connected between the stationary contact stem and an end plate of the interrupter with, preferably, the internally-disposed con3,889,079 densing shield affixed to said end plate, so that the externally located coil is connected between the stationary contact and the internally-located condensing shield.

Yet another embodiment of the present invention provides a generally spiral coil disposed internally of the envelope, and supported from the end plate associated with the stationary-contact end of the vacuum circuit-interrupter. This coil constitutes the internallylocated condensing shield for the interrupter.

In yet another embodiment of the invention the internally-located condensing shield is of split construction, one half of the split condensing shield being electrically connected by a coil to the stationary contact of the interrupter. The other half of the split condensing shield is electrically connected by a second coil to the movable contact of the interrupter. The two coils act like a Helmholz coil in a cumulative fashion to provide an internal axial magnetic field.

In yet another embodiment of the present invention there is provided an isolated shield in which the shield is electrically floating, and is not connected to either electrode. This shield has a generally spiral configuration. It is anticipated that parts of the shield adjacent to the cathode electrode will collect predominantly electron current. Parts of this shield adjacent to the anode electrode will collect predominantly ion current. These currents will be of equal magnitude since a floating shield collects zero net current. Due to the collection of electrons at the cathode end of the shield, and ion curent at the anode end of the shield, there will be a circulating current through the coils which make up the shield.

It is not necessary that the coil, which produces the axial magnetic field, be connected only to the stationary contact of the interrupter. In yet an alternative embodiment of the invention, the coil is electrically connected between the movable contact of the interrupter and the internally-located condensing shield.

In a still further embodiment of the invention, a condensing shield of a fixed-gap vacuum device is connected via a coil strap to one of the two fixed electrodes.

Further objects and advantages will readily become apparent upon reading the following specification, taken in conjunction with the drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a vertical sectional view taken through a vacuum-type circuit interrupter embodying the principles of the present invention, the contact structure being illustrated in the open-circuit position;

FIG. 1A shows a modification of FIG. 1 wherein the condensing shield is solely supported by the magnetic field coil strap;

FIG. 2 is a sectional view taken substantially along the lone II—II of FIG. 1 looking in the direction of the arrows;

FIG. 3 illustrates a modified-type circuit-interrupter construction in which the axial magnetic fieldproducing coil is disposed externally of the envelope, and is connected to the internal condensing shield by means of a probe extending through the side wall of the envelope;

FIG. 4 is another embodiment of the invention utilizing an external axial magnetic field-producing coil with the coil connected to the contact stem, and also to the end plate of the circuit-interrupter, with the condensing shield likewise affixed to the aforesaid end plate;

FIG. 5 illustrates a vertical sectional view taken through a modified-type of circuit-interrupter construction in which the condensing shield assumes a spiral-strap configuration, and is electrically connected to an end plate of the circuit-interrupter;

FIG. 6 illustrates still another embodiment of the invention in which a split condensing shield of the circuitinterrupter has two coil-strap connectors electrically connected to the two separable contacts of the interrupter to result in an axial magnetic field interiorly of the vacuum “bottle”;

FIG. 7 illustrates still another embodiment of the invention utilizing a floating condensing shield of a generally spiral configuration in which circulating currents create the axial magnetic field within the interrupter envelope;

FIG. 8A is a diagrammatic view illustrating the principles involved in the interrupter of FIG. 7 for creating axial magnetic fields via circulating currents in a “floating’’condensing shield;

FIG. 8B is, a diagrammatic view, similar to that of FIG. 8A, further illustrating the circulating currents involved in the circuit-interrupter of FIG. 7;

FIG. 9 is a graph illustrating the principles of arcvoltage reduction by means of axial magnetic fields created by ion currents to the condensing shield;

FIG. 10 is a graph illustrating the relationship between the magnetic field created by current flow through coils of different geometries;

FIG. 11 is a graph which shows how ion-current flow through the coils produces stable operating points, which are a function of the coil geometry;

FIG. 12 is a diagrammatic view of the test circuit, which ws used to verify that the burning voltage of high-current vacuum-arcs is significantly reduced by connecting the condensing shield via a coil to one of the electrodes. These tests were performed on the interrupter of FIG. 12A which has a geometry similar to FIG. 3;

FIG. 13 is a graph of experimental data taken from the test circuit of FIG. 12. The arc voltage for a 36 Hertz a.c. arc is plotted as a function of arc current to 20 K.A. r.m.s. The arc voltage is plotted first with the condensing shield connected directly to the cathode, secondly with the condensing shield connected to cathode potential via an external coil of six turns, and finally with the condensing shield connected to the cathode via a coil of 12 turns;

FIG. 14 is one datum point taken from the curve of FIG. 13. In the graph of FIG. 14 we have shown the arc voltage during a half-cycle of arcing at a current level of 15 K.A. r.m.s. This curve was obtained for a vacuum-interrupter configuration, where the condensing shield was connected to the cathode via a six-turn field coil. We have also plotted the power expended within the interrupter;

FIG. 15 is a graph in which arc-voltage and power are again plotted for this 15 K. amp arc. Here the condensing shield is connected directly to cathode potential;

FIG. 16 is a graph which shows the peak shield-ion current plotted as a function of the 36 Hertz arc current. Three curves are shown. The first curve shows the ion current collected to the condensing shield when the latter is connected directly to the cathode. The second curve shows the ion current to the condensing shield

3,889,079 interrupter embodying applicants’ invention, the contacts, when closed, completely bypass the magnetic field structure, which only during the time of arcing, sets up the axial magnetic field internally of the inter5 rupter to benefit the interruption process.

This has the important advantage that heating effects are thus eliminated in the closed-circuit position of the interrupting device, whereas the axial magnetic field structures of the prior art, by having the coils energized 10 at all times, must put up with the heating problem, and also, with the eddy-current heating problem, both of which are obviously undesirable.

Referring to the drawings, and more particularly to FIGS. 1 and 2 thereof, the reference numeral 1 gener15 ally designates a vacuum-type circuit interrupter comprising an evacuated envelope 2 having end plates 3 and 4. Affixed to the upper end plate 3 of the interrupter 1 is a stationary contact stem 5, to the lower end of which is secured the stationary contact, or electrode 20 6 of the circuit-interrupter 1.

Making separable cooperating engagement with the stationary electrode 6, is the movable contact or electrode 7, actuated by any suitable external operating means, not shown, and being affixed to the upper end 25 of a flexible metallic bellows 8. The lower end 8α of the bellows 8 is secured within an opening 9 of the lower end plate 4 of the interrupter 1. As well known by those skilled in the art, the bellows 8 provides a suitable vacuum-tight seal for the operation of the movable ³⁰ stem 10 of the movable contact structure 7. The closed-circuit position of the device is illustrated by the dotted lines 11, whereas the full lines as shown in FIG. 1 indicate the fully open-circuit position of the device.

³³ Disposed within the insulating casing 12, which may be of a suitable ceramic material, for example, is a metallic condensing shield 13. As well known by those skilled in the art, the condensing shield 13 insures that no metallic vapor, sputtered from the contacts 6,7 dur⁴⁰ ing the arcing conditions, will be deposited interiorly on the inner casing wall 12α, which might, if not prevented, result in internal flashover due to the voltage existing between the contacts 6 and 7 in the opencircuit position of the device 1. For example, the device ⁴⁵ may operate at a voltage of 15 K.V., and the inner surfaces 12α of the ceramic casing 12 must withstand B.I.L. levels far in excess of 15 K.V. in the open-circuit position of the device 1.

It will be noted that the condensing shield 13 is par⁵⁰ tially supported from, and is in good electrical contact with, the upper stationary contact stem 5 by means of a coiled metallic strap 19 constituting a magnetic field coil. Thus, the condensing shield 13 is at the same voltage potential as the upper stationary contact 6. Addi⁵ tional support for the condensing shield 13 may be provided by an annular flange portion 15, which extends within an annular recess 17 provided on the inner wall 12α of the casing 12. The condensing shield 13 is sup„ ported by the coiled strip of metal 19, and it will be ^U noted that, consequently, the one turn of the magnetic field coil 19 produces an axial magnetic field within the interruper 1 during arcing.

The operation of the shield coil 19 will be described ₆₅ first relative to the important condition, where the ³ shield 13 is connected to cathode potential during arcing. In the absence of a confining magentic field, the probability of arcing to the shield is high. The plasms is in good electrical contact with the biased shield, and when this shield is connected to the cathode via a coil of six turns. Finally, we also show an experimental curve where the ion current was observed with the condensing shield connected to the cathode via 12 turns;

FIGS. 17-19 show typical oscilligrams of the ion current collected to the condensing shield during a.c. arcing at 15 K.A. r.m.s. In FIG. 17 the peak ion current is 3.2 K.A. and the condensing shield is connected directly to the cathode. In FIG. 18 the peak ion current is 1 K.A., and the condensing shield is connected to the cathode via a six-turn coil. In FIG. 19, the peak ion current is 0.42 K.A., and the condensing shield is connected to the cathode via a 12-turn coil;

FIG. 20 is a graph showing instantaneous values of arc-voltage and arc-power during 60 Hertz arcing at 14.5 K.A. r.m.s. This figure was determined with the condensing shield connected to cathode potential by way of a six-turn coil;

FIG· 21 is a graph which shows instantaneous values of arc-voltage and arc-power during 60 Hertz arcing at 14.5 K.A. r.m.s. These curves were determined for a condensing shield configuration in which the condensing shield was connected directly to cathode potential;

FIG. 22 is a graph which demonstrates that axial magnetic fields improve high-current interruption ability. Although the curves in FIG. 22 were determined by applying axial magnetic fields created by field coils in series with the arc current, they serve to demonstrate the superior interruption ability. Open points are interruption points observed with no axial magnetic fields applied, and the solid points are interruption tests observed with axial magnetic field applied;

FIG. 23 is a somewhat diagrammatic view indicating the instantaneous-current conditions in which the stationary contact acts as a cathode and the moving cooperable contact acts as a anode;

FIG. 24 is a view, somewhat similar to that of FIG. 23, indicating the instantaneous current conditions at a period of time in which the stationary contact acts as an anode and the cooperable moving contact acts as a cathode;

FIG. 25 is a photograph of a 1,000 amp. arc burning between a large-area cathode and a small-area anode. Here, there is an applied axial magnetic field;

FIG. 26 is a photograph of a vacuum arc of the same arc current burning between the same electrode geometries as in FIG. 25. Here there is no axial magnetic field; and,

FIG. 27 shows an application of the invention to a triggerable fixed-gap arc device.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

It is well known by those skilled in the art to utilize an axial magnetic field within a vacuum-type circuit interrupter, as evidenced by the following patents: U.S. Pat. No. 3,283,103-Greenwood et al; U.S. Pat. No. 3,345,484—Polinko, Jr. et al; U.S. Pat. No. 3,321,599Lee; U.S. Pat. No. 3,263,162-Lucek et al.; French Pat. No. 1,415,441; and British Pat. No. 1,258,015-Ookura et al. In all of the aforesaid patents, it will be noted that the magnetic field coil structure is in series with the separable contact structure, so that the magnetic field coils carry the series, current at all times in the closedcircuit position of the devices.

It is, however, an important feature of the present invention that in the closed-circuit position of the circuit

3,889,079 experiments to 20 K.A. have shown that the shield can receive an ion current i of approximately 10% of the arc current I. However, with a shield design, as shown in FIGS. 1 and 2, the shield current from the plasma must flow to the cathode stem via the coil, and will consequently produce a confining axial magentic field. The arc voltage and the probability of arcing to the shield will be reduced.

In the prior art, several investigators have disclosed that axial magnetic fields are effective in increasing the current interruption rating of vacuum interrupters. This axial magnetic field is applied parallel to the axes of the electrodes, and the field-producing coils are always placed in series with the current through the electrodes. In patents by Lee and Porter, the axial magnetic field is produced by a solenoidal coil wound external to the interrupter. In a third patent by Porter, the electrode stems are constructed from materials of widelydifferent electrical conductivities, and an axial magnetic field is produced from these “stem coils.”

We find that the axial magnetic field is effective in increasing the current-interruption ability of a vacuum interrupter for two reasons. First, the arc voltage is reduced during the critical half-cycle of arcing prior to the natural current zero. This arc-voltage reduction minimizes the energy released within the interrupter, and consequently minimizes the temperature rise of the internal interrupter components. In particular, recent experiments indicate that axial fields produce a significant increase in the threshold current associated with anode-spot formation. Thus, in low-current d.c. experiments, where a small-diameter anode is drawn to long electrode spacings from the cathode, the threshold current for anode-spot formation in the absence of an axial magnetic field is 500 A. When the experiment is repeated with an axial field of 300 G, however, the threshold current for anode-spot formation is raised to 1,500 A. Recent literature reveals that low-current anode spot experiments are relevant to high-current phenomena in vacuum interrupters, and we are therefore confident that axial magnetic fields will delay the onset of anode-spot formation in a practical a.c. interrupter. Comparative arc structures observed at 1,100 A. with and without an axial magnetic field are shown in the photographs of FIGS. 25 and 26.

A second benefit derives from the confining effect of the axial magnetic field. This field reduces the plasma contact with the vapor condensation shield, and therefore reduces the probability of the arc striking to the shield. This is a particularly important effect since costreduced vacuum interrupters are being manufactured with the vapor-condensation shield connected directly to one of the electrodes. We have shown that the rate of rise of dielectric strength is only decreased slightly in these integral shield designs, but problems may be encountered at high currents (approximately 20 K.A.) due to the tendency for the arc to strike over to the shield. The strikeover tendency will be small when the shield is connected to the anode electrode during arcing. However, the shield will be biased negative relative to the arc plasma during those arcing half-cycles in which the shield is at cathode potential. Under these circumstances, the arc plasma may initiate cathode spots on the shield with resulting catastrophic failure of the interrupter. We can conclude that axial magnetic fields will definitely prove of value in cost-reduced interrupters, and the field will prove of particular value when the shield is at cathode potential during the arcing half-cycle.

The operation of the shield coil will be described first relative to the important condition where the shield of FIG. 1 is connected to cathode potential during arcing. In the absence of a confining magnetic field, the probability of arcing to the shield is high, as shown in FIG. 23. The plasma is in good electrical contact with the biased shield, and our experiments have shown that the shield can receive an ion current i of approximately 10% of the arc current I. However, with a shield design, as shown in FIGS. 1 and 23, the shield current from the plasma must flow to the cathode stel via the coil 19, and will consequently produce a confining axial magnetic field. The arc voltage and probability of arcing to the shield are reduced. It must be appreciated that the coil 19 of FIG. 23, connecting the shield to the electrode stem, may be placed external to the interrupter, rather than internal, as shown in FIG. 3. Under these circumstances the shield is electrically isolated from both electrodes internally, but is biased to one of the electrodes via the coil and an electrical connection to the shield-support flange.

Concentrating our attention on the shield design of FIG. 1, let us consider the possible magnitudes of the magnetic field produced by the metallic coil 19. If we approximate the metallic coil 19 of FIGS. 1 and 2 by a single loop of radius R equal to the shield radius, then the magnetic field in the interelectrode gap is given by the expression:

i R²

Haxial ⁼ ~2 ( /? ² + X² ) ^3/2 ’ ( I I where λ is the distance from the coil plane to a point in the interelectrode gap. Since the coil can be located in the electrode plane, let us neglect x compared to R. Then equation (1) reduces to the form:

(amperes per meter).

But

Haxial ~ ,, K-o where v₀ is the permeability of free space (4 π X 10—7 henry metre “’).

_ i 4ir 10⁴ i

Whence B (gauss) = ⁼ ΊΓ^{36 X 10} ’·

In present interrupter designs R has a value of 3 to 4 cm. Allowing i, the ion current, to be approximately 10% I, the arc current; then:

B = I . 2 X 10^-2 gauss

For this example, the field will be 200 gauss when the arc current is 10 K.A. Naturally, the field at a given arc current can be increased by adding more than one turn to the “field spiral” 19 connecting the shield to the electrode stem 5. With two turns, for example, a field of 200 G. will be obtained at arc currents of 5 K.A. This is a significant and useful field. The arc voltage in a 600 A. interrupter at an electrode spacing of 1.9 cm, for ex3,889,079 ample, is of the order of 38 V. at 4.2 K.A. in the absence of a magnetic field. When a field of 200 G. is applied via external Helmholtz coils, the arc voltage reduces to 30 V., and the ion current i still has a value of 220 A. This ion current could have produced the required magnetic field by flowing through four turns of the described loop.

An alternative shield design is shown in FIG. 5. Here the shield is formed from a spiral strip of metal, and these strips overlap in order to protect the insulating envelope from vapor deposition. In FIG. 5 the shield is shown connnected to the stationary electrode, although either electrode support would prove satisfactory.

Let us now consider the less important situation of the shield connected to the arcing anode, as shown in FIG. 24. This connection is less important since the shield is now biased positive with respect to the plasma, and cathode spots are unable to form on the shield surface. Thus, the probability of the arc striking to the shield is low, even without a confining magnetic field.

In the absence of a field, a shield conhected to the anode acts as an auxiliary anode and draws a net electron current from the plasma. This is beneficial since the threshold current for anode-spot formation increases with the effective anode area. With coils connected to the shield, as shown in FIGS. 1 to 5, this electron current will be decreased to some extent by the resulting axial magnetic field. However, we can expect that the resulting decrease in the anode-spot threshold current will be offset by the confining effect of the magnetic field. In particular, we have observed experimentally that the arc voltage of high-current arcs to 25 K.A. is significantly reduced when the shield is connected to anode potential via approximately six turns of coil.

If necessary, a split-shield device can be designed, in which one-half of the shield is connected via a coil to each of the electrodes. A schematic of such a device appears in FIG. 6. Here the shield is split by a cut parallel to the axes of the electrodes. However, the shields could be split by a cut parallel to the electrode surfaces.

A final concept concerns the application of a shieldgenerated axial magnetic field during arcing, even when the shield is electrically insulated from both electrodes. Let us suppose that the “floating shield” of FIG. 7 has a generally spiral configuration adjacent to the arcing gap.

Since the floating shield is isolated from both electrodes, the shield, during arcing, collects zero net current from the adjacent arc plasma. However, we can confidently expect that the shield will collect electron current at regions adjacent to the cathode electrode, and an equal ion current to regions adjacent to the anode electrode. As a consequence of this distribution of current collection, a circulating current will pass through the spirals of the shield, and will produce an axial magnetic field. We will clarify the mechanism of FIG. 7 by referring to FIGS. 8A and 8B. In FIG. 8A we show an arc of current I burning between an anode and a cathode. We are also showing that the arc is burning through the hole bored in two metallic plates, plate 1 and plate 2. Let us consider that the plates are isolated from each other, and that there is an appreciable voltage drop in the arc column. Let us consider that the total arc voltage is perhaps 150 volts, and that the plasma potential adjacent to plate 1 is approximately 100 volts positive relative to cathode potential, and that the plasma potential in the arc column adjacent to plate 2 is 50 volts positive relative to cathode potential. From Langmuir probe considerations, we know that the potential of plate 1 will be approximately 100 volts positive relative to the cathode, and the isolated potential of plate 2 will be approximately 50 volts positive relative to cathode potential, that is, the plates assume the potential of the adjacent plasma. Now consider the effect of connecting plates 1 and 2 via an ammeter. This is shown by the dashed lines in FIG. 8A. The potential of plate 1 will drop from 100 volts positive relative to cathode to approximately the potential of plate 2, that is, the potential of plate 1 will drop from 100 volts to 50 volts. As a consequence, plate 1 is effectively biased minus 50 volts negative relative to the potential of the adjacent arc plasma. It will consequently draw a net ion current I] from the arc plasma. Due to the high electron mobility, plate 2 will rise only several volts positive relative to the adjacent arc plasma, that is, it might rise from initially 50 volts positive relative to cathode to 52 or 53 volts positive relative to cathode. Thus, plate 2 is effectively biased positive relative to the adjacent arc plasma, and will draw a net electron current L from the arc plasma. Thus, the two metallic plates 1 and 2 draw zero net current from the arc plasma, but the ammeter connecting the two plates together will register a total circulating current Ip It is this circulating current which we will use to generate our axial magnetic field.

In FIG. 8B, we are depicting that the two ends of the metallic shield, which we can relate to plate 1 and plate 2 of FIG. 8A, are connected together via a strip of metal in the shape of a coil. Circulating currents through this coil will generate the desired axial magnetic field.

In summary, the shield configuration of FIG. 7 will create an axial magnetic field due to the fact that there will be a significant voltage drop along the high-current arc-column adjacent to a shield of considerable overall length. This voltage drop will drive a circulating current through the shield coils.

In clarifying the mechanism whereby shield currents create axial magnetic fields as, for example, in FIG. 1, we will concentrate on describing the situation of shield biased to cathode potential. This connection is of particular importance. When a shield is connected to anode potential, even without shield coils, the shield acts like an auxiliary anode, and markedly reduces the arc voltage. However, if a shield is connected directly to cathode potential there is no reduction in the arc voltage. A shield connected via coils to either electrode does produce a reduction in the arc voltage, and we will now explain, in detail, the mechanism of voltage reduction with shields connected to the cathode. Let us first consider FIG. 10. If we take any coil of a given geometry, the magnetic field strength at the center of the coil will increase linearly with the current through the coil. For a given current in the coil, the field strength will be larger, the greater the number of turns in the coil. For the various embodiments of our invention, the coil is, of course, the metal coil whether external or internal connecting the shield to one of the electrodes. This current is an ion current when the shield is biased to the cathode. In FIG. 10, it is shown that as the ion current through a given coil increases, the axial magnetic field will also increase. This ion current is drawn to the shield from the arc plasma. Naturally, an increasing

3,889,079 axial magnetic field, generated by the coil, will decrease the arc plasma contact with the shield. Consequently, there are limited operating points determined by the ion current dependence on the axial magnetic field and the shield geometry.

Consider FIG. 11. At a given arc current, and no magnetic field, a large ion current will be collected to a negatively-biased shield. The intercept on the Y-axis shows the ion current to the shield when the shield contains no turns. From experiment we know that the ion current to a shield with on turns decreases with an externally-applied axial magnetic field. This decreasing ion current is shown by the dashed line of FIG. 11. Consider now a shield with a number of turns between shield and cathode. By analogy with FIG. 10, the created axial magnetic field increases linearly with the ion current to the shield. For a given arc current, the shield will operate at the stable operating point given by the intersection of the shield load-line and the ion current curve. Consider further the practical experimental data of FIG. 9. This figure was generated from data using a practical vacuum interrupter with an externally applied axial magnetic field. The data all refers to an arc current of 4.2 K.A. Let us first consider the arc-voltage curve. When there is zero axial magnetic field the arc voltage is 38 volts. With an applied magnetic field of slightly less than 200 gauss the arc voltage has reduced to 30 volts. The arc voltage then increases slowly with further increase in magnetic field. Consider now the ion current curve. This curve was determined with the shield connected directly to cathode potential, but for various values of externally-applied axial magnetic field. In the absence of an axial field, the ion current from the 4.2 K.A. arc plasma has a value of about 280 amps. As the shield applied to the arc plasma increases, this ion current reduces and reaches a value of about 120 amps when the field is 800 gauss.

Now, consider the arcing situation where, instead of applying external axial magnetic fields, the shield is connected to cathode potential via several turns. By analogy with FIGS. 10 and 11, the shield will have a coil-load line. This coil-load line is shown in FIG. 9 and intersects the ion current curve at point A. Let us consider the possibilities. It will be obvious that the arc will not burn with zero magnetic field. For that particular hypothetical situation, an ion current of 280 amperes would flow from the arc plasma to the shield and the arc voltage would be 40 volts. However, a shield ion current of 280 amps would create a large magnetic field dictated by the coil-load line. Thus, an operating point at the zero magnetic field is unstable. The only stable operating point is point A. It will be noted in FIG. 9 that the arc-voltage minimum is attained at relatively low magnetic fields. Fortunately, these magnetic fields can be created using coils of only several turns. As direct evidence of the shield-coils efficiency in reducing the arc voltage, let us consider the experimental data of FIGS. 13,16,17, 18 and 19. We experimented with a vacuum interrupter, as shown in FIG. 12A. Here the shield coil is connected externally to the cathode via a six-turn coil. This is similar to the embodiment shown in FIG. 3. We connected the shield either directly via a straight conductor to the cathode, or via a six-turn coil, as shown in FIG. 12A, or via a twelve-turn coil; and monitored at high currents the arc voltage and also the ion current, which flowed from th arc plasma to the shield. FIG. 13 shows the peak arc voltage during a half cycle of arcing at currents to 20 K.A. The arc voltage at 15 K.A., for example, with the shield connected directly to the cathode is approximately 150 volts. With the shield connected to cathode potential, via a six-turn coil, the arc voltage is approximately 85 volts. With the shield connected to cathode potential via a twelve-turn coil, the arc voltage is 55 volts. The curve of FIG. 16 shows the current which flows to the shield from the arc plasma for the experimental conditions of FIG. 13. At 15 K.A. an ion current of approximately 3 K.A. flows to the shield when the latter is connected directly to the cathode. This connection corresponds to the zero-field situation of FIG. 9. With the shield connected to the cathode via six turns, the peak ion current, during the arcing half cycle, reduces to approximately 1 kiloampere. Thus, in this situation, the operating point A of FIG. 9 has resulted in a one-third reduction in the ion current. However an ion current of 1 kiloampere flowing through the six-turns produces a marked effect on the arc voltge, as already discussed in FIG. 13. Again, with reference to FIG. 16, with the shield connected to cathode via 12 turns, the ion current through the coil turns is reduced to approximately 0.6 K.A. Typical data for FIG. 16, obtained using the configurations of FIG. 12A, appear in FIGS. 17,18 and 19. These figures are taken from single oscillograms of shield currents during a half cycle of arcing at 15 K.A. R.M.S. FIG. 17 shows the ion current to the shield when the latter is connected directly to cathode potential. This is the high arcing-voltage condition.

FIG. 18 shows that the ion current has been markedly reduced by connecting the shield via six turns to cathode potential. The arc voltage is now significantly reduced. FIG. 19 shows that the ion current is even more reduced when the shield is connected to cathode potential via 12 turns. This is the lowest voltlage condition. Again, the reasons for the ion current reductions observed in FIGS. 17, 18 and 19, for a given 15 K.A. arc but with different shield geometries, can be understood relative to FIG. 11. The test data, which we have discussed, were obtained in the interrupter of FIG. 12A using the test circuit shown in FIG. 12. This test circuit comprises a two megajoule capacitor bank which discharges through a reactor. In order to obtain arcvoltage data at comparative electrode separations, we deliberately obtained our data using a circuit tuned to 36 cycles per second. However, we obtained similar data to FIGS. 13 and 16 when we performed experiments at 60 cycles per second. Examples of data at 60 cycles appear in FIGS. 20 and 21. Data in these two curves was obtained during a half cycle of arcing for an arc current of 14.5 K.A. R.M.S. Consider FIG. 21. The shield is connected directly to cathode potential, and the arc voltage during the 8 milliseconds of arcing is high. The peak arc voltage during the arcing half cycle is approximately 120 volts. In FIG. 21 we have also plotted the instantaneous power dissipated in the interrupter. From this figure we estimate that an arc energy of 8.9 kilojoules was expended during the arcing half cycle. Consider now FIG. 20. Here the peak arc voltage is markedly reduced by using a six-turn coil between the shield and the cathode. In particular, the arc energy, expended during the off arcing half cycle, is only 4.9 kilojoules. Thus, the effect of the shield coils is to reduce not only the peak arc voltage, but also, and more importantly, the total energy expended in the interrupter during the arcing half cycle. Since the data of

3,889,079

FIG. 13 shows only the effect of shield-coil connections on the peak arc voltage, we include FIGS. 14 and 15 to demonstrate that the energy expended within the interrupter was markedly reduced by field coils. The data in FIG. 13 were obtained during a half cycle of arcing at 36 cycles per second, that is, the arc duration was approximately 14 milliseconds.

Consider FIG. 15 where the shield was connected directly to the cathode. During the arcing half cycle at 15 K.A. R.M.S., the peak arc voltage was approximately 150 volts, and the total arc energy was of the order of 16.7 kilojoules. The data in FIG. 14 was also obtained at 15 K.A. R.M.S., but with the shield connected to the cathode via a six-turn coil. Here the peak arc voltage is only 60 volts, and the total arc energy dissipated within the interrupter during the arcing half cycle was reduced from 16.7 kilojoules to only 8.3 kilojoules.

In summary, our experimental evidence indicates that the shield coils work as shown diagrammatically in FIGS. 23 and 24. In FIG. 23 the shield is connected externally, via external coils, to the stationary electrode, which, for this particular arcing half cycle, is the cathode. It will be appreciated that the cathode could also be the movable bellows electrode, as shown for example in FIG. 12A. With the connection, as shown in FIG. 23, the shield is biased negative relative to the adjacent arc plasma. With this connection the shield collects a net ion current from the arc plasma, and this current, flowing through the shield coils, produces the desired axial magnetic field. The axial magnetic field reduces the arc plasma contact with the shield, as exemplified by FIG. 9 and as shown in FIG. 16.

The tendency for the high-current arc to strike over to the shield is reduced, and in particular the arc voltage is markedly reduced, as shown in FIG. 13. For arcing half cycles in whidh the shield is connected to anode potential, the shield acts like an auxiliary anode. As shown in FIG. 24, the shield is biased positive relative to the arc plasma, and draws a net electron current from that arc plasma. Again, the circulating current results in an axial magnetic field which reduces the arc voltage. Experiments show that the arc-voltage reduction is similar to that depicted in FIG. 13. Additional experimental data, obtained by applying axial magnetic fields to vacuum arcs, appear in FIGS. 25, 26 and 22. FIGS. 25 and 26 show vacuum arcs burning between a large-diameter cathode, which is the upper electrode, and a small 1.3 centimeter diameter anode, which is the lower electrode. The photographs in both FIGS. 25 and 26 were obtained for a direct current arc of approximately 1000 amps. In the presence of an axial magnetic field, FIG. 25, luminous streamers propagate from each of the individual cathode spots, and these streamers terminate on the anode. The arc voltage is low, the plasma is confined predominantly to the interelectrode region, and this confinement is evidenced by the presence of the luminous streamers. In particular, there is no gross melting of the anode, since low confinement to the interelectrode region has reduced the anodevoltage drop. There is no anode spot activity. FIG. 26 shows the same 1000 amp arc in the absence of an externally applied axial magnetic field. We have the characteristic multiplicity of cathode spots. However, the interelectrode arc plasma is diffuse. The arc voltage is high due to the presence of an appreciably-high anode voltage drop. There is vigorous evaporation at the anode, with gross melting, due to the presence of a vigorous anode spot. FIG. 25 shows a more desirable arcing condition for a vacuum-interrupter. The arc plasma is confined away from the walls of the arcing chamber, the arc voltage is low and, consequently, the heating of 5 the electrodes and of the shield is reduced. The gross melting and deformation of the anode does not exist. Although the comparative photographs of FIGS. 25 and 26 were obtained at an arc level of 1000 amperes, experimental data at current levels of 15,000 and 10 20,000 amperes suggests that the same physical phenomena occur at high arc currents with and without magnetic fields. The beneficial effects of axial magnetic fields are also shown in FIG. 22. This figure shows interruption data following arcing to 15 K.A. at 60 cycles 15 with and without the presence of an external axial magnetic field. These particular data was obtained with the axial field generated by Helmholz coils. The open points show data obtained without an axial field. The solid points show data obtained with an axial field. It 20 will be noted that as the circuit current increases, and as the circuit voltage increases, the probability for arc reignition without with magnetic field is high. This beneficial effect of axial magnetic fields on the interruption ability, has been observed by many investigators.

To summarize the advantages of our disclosure, we can list these advantages as follows:

We have a method for applying an axial magnetic field whereby the field coils are in parallel rather than in series with the arc current. With this situation, the ³⁰ field coils produce no energy losses with the electrodes in the contact-closed position. The field coils are only inserted into the circuit when the arc is struck within the interrupter. The field coils produce an axial magnetic field, which (a) postpones the onset of anode spot formation, (b) reduces the energy expended within the interrupter during the arcing half cycle. This energy reduction minimizes the heating of the shield and minimizes the heating of the electrodes, (c) the probability of the arc striking over to the shield is reduced, (d) the interruption ability is increased. It will be noted that the shield configuration should lead to significant cost reductions for the manufacture of vacuum-interrupters. For a given interrupter size, the rating of the interrupter will be increased over the situation with a shield supported and electrically connected from one electrode having no magnetic field-coil arrangement. Further, arrangements for minimizing anode-spot activity can be expected to lead to vacuum-interrupters, which can withstand high voltages. Anode-spot activity leads to deformation of the surface and splashing of anode material onto the shield and the cathode, and this deformation and splashing can markedly reduce the basic impulse level and transient recovery of the interrupter FIG. 27 illustrates a further embodiment of the present invention, which is applicable to a triggerable fixedgap vacuum-device. It will be noted that the various embodiments of field-coils discussed hereinbefore wil also prove advantageous in arcing devices other than o: 0Q the separable-contact type. Consider, for example, the triggered fixed-gap vacuum device. Here, the elec trodes are permanently separated, and arcing is initi ated by spark formation of both a cathode spot and i tenuous inter-electrode plasma. The subsequent arcin; 0₅ and interruption phenomena are practically identica with those observed in the vacuum-switch. The presen invention, therefore, is also advantageously utilized ii the use of field-coils associated with triggered vacuun

3,889,079 gaps, as well as with vacuum-type switches and interrupters, as hereinbefore described.

The gap of U.S. Pat. No. 3,087,092 comprises a pair of spaced-apart main electrodes 40 and 42 disposed in a highly evacuated chamber 44 and defining a main gap 45 therebetween. Disposed adjacent one main electrode 42 is a trigger electrode 46 defined by a hydrogen-impregnated titanium film on a ceramic supporting rod 48. This ceramic supporting rod 48 is disposed coaxially of the main electrode 42 and is suitably sealed to the main electrode 42 about its outer periphery. A portion of the ceramic supporting rod 48 is uncoated and defines a trigger gap 51 along this uncoated surface that electrically isolates the trigger electrode 46 from the main electrode 42 under normal conditions. A conductive connection 49 extends through the ceramic rod 48 and across its upper end surface to the trigger electrode 46.

When an electric pulse is applied between the trigger electrode 46 and the main electrode 42, the trigger gap 51 breaks down, and the resultant spark liberates a small quantity of hydrogen from the hydrogenimpregnated trigger electrode 46. This hydrogen is quickly ionized and projected into the main gap 45, thus lowering its dielectric strength and initiating a breakdown of the main gap.

Additional theory regarding the operation of fixedgap triggerable vacuum-devices is set forth in U.S. Pat. No. 3,489,951, issued Jan. 13, 1970 to A. N. Greenwood et al; U.S. Pat. No. 3,087,092, issued Apr. 23, 1963 to J. M. Lafferty; and U.S. Pat. No. 3,465,192, issued Sept. 2, 1969 to J. M. Lafferty. Also additional information is available in a IEEE 1965 paper entitled, “Triggered Vacuum Gaps” by J. M. Lafferty, set forth in Volume 54, No. 1 of Proceedings of The IEEE.

The foregoing references set forth that in such a device the ability of a vacuum-gap to hold off high voltage and then to recover its dielectric strength rapidly after arcing, has made it pontentially attractive as an overvoltage protection device and current switch. For example, when such a device is utilized as a lightningarrester on power transmission lines, in such an application the gap was connected between line and ground and designed to hold off maximum system voltage, but to break down on impulse voltage surges associated with lightning strokes or switching transients. Following impulse breakdown, the gap would interrupt the flow of power-frequency current at its first current zero. In such an application it is desirable to interrupt arcing within the device following breakdown, and consequently the magnetic field coil 19 of the present application may advantageously be utilized in connection with such a fixed-gap device, as illustrated in FIG. 27 of the drawings.

From the foregoing description, it will be apparent that we have provided an improved vacuum-type circuit interrupter in which an axial magnetic field is utilized to reduce arc voltage only during the arcing period, and in the normal closed-position of the device, the heating problem and the eddy-current heating problem is eliminated, by virtue of the fact that the contacts, when closed, completely electrically bypass the field-coil structure.

Although there has been illustrated and described specific structure, it is to be clearly understood that the same were merely for the purpose of illustration, and that changes and modifications may readily be made therein by those skilled in the art, without departing from the spirit and scope of the invention.