|Publication number||US2217421 A|
|Publication date||Oct 8, 1940|
|Filing date||Jul 19, 1934|
|Priority date||Jul 19, 1934|
|Publication number||US 2217421 A, US 2217421A, US-A-2217421, US2217421 A, US2217421A|
|Original Assignee||Westinghouse Electric & Mfg Co|
|Export Citation||BiBTeX, EndNote, RefMan|
|Referenced by (13), Classifications (26)|
|External Links: USPTO, USPTO Assignment, Espacenet|
06f. 8, 1940. H SCOTT 2,217,421
CASNG FOR METALLIC VAPOR DISCHARGE DEVICES Filed Ju1y.19. 1934 Patented Oct. 8, 1940 Howard Scott, Forest Hills, Pa., allignor to Weltinghouse Electric Manufacturing Company, mEastPittsburgh, Pa., a corporation of Pennsyl- Appumion .my 19, 1934, serial No. vac-,ooo
zo mm. (c1. ssc-zas) My invention relates to a casing for metallic vapor discharge devices and especially such casings in which a metalportion forms a portion thereof in vacuum tight relation with a glass insulating portion of the casing.
This application' is a vcontinuation in part of my copending application Serial No. 376,291, for glass metal seals, led July 5, 1929.
One object of my invention is to provide a casing for a mercury vapor discharge device having a metal portion that will seal directly into a hard glass, that is, a glass of low expansivity and high silica content. f
Another object of my invention is to provide an exterior electrode for a metallic vapor discharge device which can be easily fabricated, rolled, machined or otherwise formed into convenient shapes.
Another object of my invention is to provide a casing of metal and glass portions which will be vacuum tight and yet withstand the action of certain metallic vapors.
Other objects of my invention will become apparent from the following specification and the drawing, in which Figure 1 illustrates a cross-section through a container illustrating the application ofmy invention in a preferred embodiment; and.
Fig. 2 is a curve disclosing the unit expansion with the temperature of my preferred metal with that of two hard glasses.
Heretofore the containers for metallic vapor discharge devices, such as those for mercury arc rectitiers, have had to have an all glass casing except for the leads therethrough or have had to have connected thereto a pump to maintain a suitable vacuum therein. 'I'he reason for this is that copper, which is the only convenient metal that can be easily machined or fabricated and .is 40 also non-porous for the desired vacuum, is attacked by mercury so that the use of copper in the prior art has been limited to vacuum tube devices. Platinum is too expensive to use except for small lead-in wires of expensive apparatus. Timgsten and molybdenum are very dilllcult, if not impossible, to fabricate, and their use is limited to small lead wires. Even these wiresare not very satisfactory with the hard glasses, such as a boro-silicate glass, and in fact their use is rather expensive due to the involved careful manufacturing processes required.
In searching for an alloy that would form a portion of the exterior casing of a mercury arc discharge device, these metals generally used for vacuum tube devices are accordingly eliminated.
The use of boro-silicate glass is desired as a portion of the container due to its her melting point than ordinary lead glass. The ro-silicate glass, however, presents the dimculty that it has a` materially lower coeilicient of expansion than 5 lead glass, and consequently the alloy to be secured thereto must also have a -vey low coeiilcient of expansion. Platinum could not be used with boro-silicate glass on account of its higher coefllcient of expansion. It is apparent, 10
of course, that if there is a. wide difference between the coeillcient of expansion of the metal and the glass that when the seal cools, cracks and leaks will develop due to the 'resulting strains in the seal. It is necessary that the metal and the glass shall have substantially the same ex pansivity with temperature over the range from the annealing temperature of glass down to room temperature.
Another dimculty is that a successful seal re- 2o quires that the metal be one which the glass will wet." Due to al1 of these diillculties, a satisfactory alloy suitable for sealing into hard glasses that would withstand mercury and yet be easily fabricated or machined has not hitherto been 25 found.
Nickel steels of low expansivity were tried out, but it was found that, while thislow expansivity y existed up to a certain critical temperature above which the expansivity increased rapidly, this tem- 30 perature, called the inection temperature, was
considerably lower than -that suitable' for annealing hard glass, with the result that when the temperature was raised tothe proper annealing temperature for hard glassthe expansivity 35 was so great that vacuum tight 4seals couldnot be produced.v The best of these nickel steels has an inflection temperature of around 325 C., while 400 C. is the preferred annealing temperature with the high silicate or hard glasses. At 400 C., 4,0 the expansion of theseprior nickel steel alloys would be 6.5)( 10-6 centimeters per centimeter per degree centigrade. which makes them undesirable for use with boro-silicate glass.
In accordance with the present invention I 45 have discovered that, -by using an alloy consisting essentially of iron, nickel and cobalt. the alloy has a sufficiently low coeilicient of expansion throughout the entire range' below the necessary annealing temperature of the boro-silicate glass;
that such alloys are wet by the boro-silicate glasses; and that the combination produces satisfactory vacuum tight seals.
I have especially found that alloys containing from 23% to 34% mm1, 5% to 25% comme than 1% manganese andthe remainder iron maintain an expansivity over'vthe temperature range up to the annealing temperature of borosilicate glass, that is to say, up to about 400 C., which is a suillciently near approximation to the annealing point that of glass to insure a successful vacuum tight seal. 'I'hese boro-silicate glasses have expansivities between 3.0 and 6.5 times 10 centimeters per centimeter per degree centigrade. Ordinary lead glass on the other hand has an expansivity of 9X 10-6 centimeters per centimeter per degree centigrade. In particular, I have found that an alloy comprising 32% nickel, 16% cobalt, .08% manganese and carbon up to 0.10% and the remainder iron maybe sealed successfully through boro-silicate glass having a coeilicient of expansion of 6.2 10-8 centimeters per centimeter per degree centigrade. As examples of such boro-silicate glasses, the following may be given:
1. 67% S102; 22% B203; 6.5% Naz, KzO 2% 10% Naz, Kao; 10% B203; 5%
2. '12% SiOz;
3.73% S102; 1l/2% Naz, KnO; ll/2% B203: 6% PbO. I have found that successful seals result from the employment of alloys within the ranges of the composition heretofore described in conjunction with either of the aforesaid boro-silicate glasses.
I have further found that a successful seal may be made between boro-silicate glass having an effective expansivity between 6.0 and 6.5 x10-6 cm. per cm. per degree C. and an alloy containing substantially 27% nickel, 22% cobalt, 0.5% manganese and the remainder iron.
Likewise, I have found that iron alloys containing from 23% to 28% nickel, 17% to 23% cobalt, to 0.5% manganese and 0 to 0.3% carbon may be vsealed successfully in boro-silicate glasses having effective expansivities between 3.0 and 5.5)(10-6 cm. per cm. per degree C. In particular, I have found particularly good seals to be produced between boro-silicate glasses having expansivites between 3.0 and 5.5 X 10-6 cm. per cm. per degree C. and iron alloys containing substantially 29.8% nickel, 15.5% cobalt, 0.22% manganesefand 0.3% carbon. Boro-silicate glasses having expansivities of substantially 3.6 x 10*6 cm. per cm. per degree C. apparently produce the best seals with the alloy last mentioned.
I have also found that a seal between the boro-silicate glasses known to the trade as G702P and G705AJ and manufactured by the Corning Glass Company, seals well with an alloy comprising 29.4% nickel, 16.3% cobalt, 0.2% manganese and theremainder iron.
By the expression "G705AJ, the Corning Glass Company designates a borosilicate glass which is kept as closely as possible at all times to a standard chemical composition. When new batches of this glass are manufactured, they are subjected to physical tests for expansivity between zero-C. and'320 degrees C. and for their softening pointl If the batch departs by morev than 2% from an expansivity of 46x10-rl centimeters per centimeter length per degree C. or
the softening point departs by more than 5 degrees from a value of 710 degrees C., the batch is discarded and not sold under the designation G705AJ. 'I'he softening point of the glass is determined by suspending a glass rod by one end in a furnace and varying the furnace temperature until the end of the glass rod moves at the rate of one centimeter per second under the vforce of gravity. Methods of measuring the 'expansivity of glass are well known and are described in the technical literature.
In general, I have found that, in the case of alloys of low expansivity containing nickel and iron, whether or not they also contain cobalt, it is desirable that the relationship expressed by the following formula should hold: Z1 nickel+2.5 manganese)+18 carbon) %iron the best value for the fraction being about 0.55. It also appears desirable, for most purposes, to minimize the amount of manganese, the maximum percentage of this element being preferably below 0.2%. Carbon, however, appears, in most instances, to be desirable up to the amount of 0.3%, since it permits a higher cobalt content than would otherwise be desirable and does not seriously aiect expansivity. It can be used. however, only if the metal is coated in such a manner as to prevent gas release from the metal surface during heating for sealing, as explained in copending case 17,855.
The temperature at which strains are readily relieved in glass is, of course, not an absolutely fixed quantity, even for glass of a given chemical composition. At very high temperatures, the glass is relatively soft and, as the temperature decreases, it hardens. The lower the temperature. and, consequently, the harder the glass, the less rapidly it is able to ex and relieve mechanical strains existing within it. HoweverLsince the inflection temperature, that is to say, the maximum temperature, at which the coeillcient of expansion of the above-described alloys remain low, decreases as'the coefficient of expansion itself is decreased, it is desirable to form seals with a glass having as low an annealing tem-- seals if the completed seal is maintained at a temperature of about 50 C. above the inflection temperature for a considerable period of time. Thus, while the strains are relieved slowly, they do so progressively and with certainty, .and a strong vacuum tight seal results in the end.
In order to make a successful seal, the expansion of the metal and glass must be substantially the same over the temperature range within which the glass is elastic. A large difference in expansion between metal and glass bonded by sealing produces stresses which may cause the glass to crack when cooled to room temperature. Some degree of differential expansion, however, is tolerable and sometimes desirable. The relaxation characteristics of the glass determine the upper temperature limit to which matched expan'sion is essential. Obviously, at temperatures where the glass is substantially' plastic, equal expansion is no longer necessary.
The relaxation of stress in glass under a given initial strain proceeds continuously and at a rate which rapidly increases with rising temperature. For purpose of exposition, it is necessary t6 select arbitrarily a temperature characteristic oi' the relaxation range of the glass. In glass technology, two such temperatures, the strain point and annealing point, are used and define well the cate or hard glasses previously described.
Expansivit Strain Annesling (25 to 325 point point Il'he values of expansivity given in the above table represent heating from 25 to 325 C. over which range the expansion is practically linear. In connection with sealing problems, however, the expansion during cooling and over the temperature range from the annealing point to room temperature is of chief interest. .The following gives the expansivity of glasses #l (called clear sealing glass), #3 (called by the trade name "Nonex), together with the expansivity of my preferred alloy (called by the trade name Kovar) in comparison with tungsten and molybdenum. My preferred alloy as previously described comprises substantially 29.4% nickel, 16.3% cobalt, 0.2% manganese and the remainder iron` It is to be noted that the expansivity of the glasses is much higher during cooling than on heating and also higher for the longer temperature range. Since the lowest'feasible annealing temperature of a glass is its strain point, the most useful single expansion characteristic of glass in this connection is the mean expansivity on cooling from the strain point to room temperature.
The expansion curve of my preferred alloy called by the trade name "Kovar is given in Fig. 2, together with that of two hard glasses previously described for comparison. The expansion of this alloy is reversible, that is, the same A on heating as on cooling. Values of its expansivity is lower than that of tungsten up to 450 C. and than that of molybdenum up to 500 C.
It is apparent from Fig. 2 that the match in expansion between my prefered alloy and either glass is not perfect, but that the match is much better in the case of Clear Sealing Glass than that of Nonex. Though the "Clear Sealing glass has a lower expansivity than Kovar at temperatures above 435 C., the discrepancy is small so that seals practically free from stress can be made with this combination. In fact, serviceable seals can be made with Kovar in Nonex also.
With so close a match in expansivity between Kovar and the Clear Sealing glass, the anhealing procedure for least development of stress is clearly indicated. The seal while hot from the blowing operations should be cooled to the annealing point, held there fifteen minutes and then cooled to the strain point at a rate determined by the permissible residual stresses. Cooling from the strain point to room temperature can be quite rapid, the only limitation being danger of cracking from temporary stresses due to temperature gradients necessary to maintain cooling. From Fig. 2 it will be seen that the expansivity of Kovar" and Nonex matches most clearly near 400 C., while the strain point is 467 C. If the seal be annealed over night at 467 C., no stresses will remain at that temperature.
An alternative method for annealing this seal is to cool to .390 C., and hold at this temperature. Complete stress release can be achieved by holding at' this temperature for some days. Sufilcient relaxation may occur and certainly the major part does occur during an overnight anneal. If the composition of the Kovar is modified to give an inection temperature of 390 C., no additional stress will develop on cooling thereafter to room temperature.
Figure 1 discloses an application of the invention to a mercury arc rectier. although the application thereof, is, of course, applicable to other types of electric discharge devices in which it is desired to have a metal that is not attacked by mercury or in which a vacuum tight seal with hard glass is desired or in which an exterior easily machined or fabricated alloy is desired. The casing of this rectifier, as shown, preferably consists of several metalportions insulated from each other by borosilicateglass. These metal portions are of thevalloyasprviously described.
The cathode container` Il tonfthe drawing is disclosed as a cupshapedlmember containing a pool of mercury II or other metal'such as gallium .or amalgamvaporizable into a metallic vapor. The alloy that I have-described is easily machined into this desired/shape.' -It will be noted that the mercury pool not be lapt to damage the glass portions ofthe tube incase of a jar to the device, because it is supported entirely within the strong metalnalloy portion. The cup-shaped member I0 may have suitable legs or extension 34 thereon to form a supporting means, and also a cover I2 may be attached to the extension with openings I3 therethrough to provide cooling or heating means as desired to the chamber I4 provided by the extension 34 and the cover I2. Fastening means I5 for the cover portion I2 may also act as a connection for th'e cathode lead I6. A boro-silicate glass portion I1 is sealed to the upper end of the cup-shaped electrode l0, and the seal between these two is a vacuum tight arrangement. At the upper end of the casing is an anode having an exterior p0rtion I8 and an interior portion I9 to receive the impact of the molecules of mercury vapor or other metallic vapor filling the interior of the casing. The shape of this anode may be of any desired configuration, and as shown may have a central hollow portion 20 that is closed by a cover 2I either separate or integral therewith. Tube means 22 and 23 lead into the chamber 20 to provide a means of cooling the anode if desired. A boro-silicate glass portion' 24 is sealed to the outer rim or skirt I8 of the anode assembly. f
In order to start the mercury rectier, a keepalive or make-alive is generally provided and in accordance with this custom, I have disclosed rather diagrammatically an auxiliary electrode 25 connected through a bar 26 to a central metallic portion 21 of the casing. While this auxiliary electrode may be of any desired construction, I prefer to make it of the make-alive type disclosed in the application of Joseph Slepian et al., Serial No. 626,866, filed July 30, 1932. A connection 28 may be made to the central metallic portion 7 y ro.
lsembled can be sealed od through any of the glassA portions, such as that illustrated by the portion i1 sealed oil.' at 32.
By the use of the term high silicate glass" in the following claims, I mean a glass havingl overl 60% silica and an expansivity of or below 6.5X"/ C.
While I have disclosed my invention as applied to one particular embodiment, it is obvious that the invention can be applied to many other types of embodiments and that various changes may be made in the shape, number and arrangement of the elements disclosed. Accordingly, I desire that the terms in the following claims to have the broadest interpretation permissible in view of the prior art.
'I claim as my invention:
1. A container having aportion comprising an alloy of 23% to 34% nickel, 9% to 25% cobalt, less than 1% manganese and the remainder iron and another portion comprising a high silicate glass, said portions being sealed together in vacuumtight arrangement and a metallic vapor within said container.
2. A container having a portion comprising an alloy of 23% to 34% nickel, 9% to 25% cobalt, less than 1% manganese and the remainder iron and another portion comprising a` high silicate glass having an expansivity of 3.0 to 6.5 10s centimeters per centimeter per degree centigrade, said portions being sealed together in vacuumtight arrangement and a metallic vapor within said container.
3. A container having a portion comprising an alloy of 23% to 34% nickel, 9% to 25% cobalt, less than 1% manganese and the remainder iron, and a. glass portion having a composition of substantially 67% Si02; 22% B203; 2% A1202; 4/2% Na2, K20, said portions being sealed together in vacuum-tight arrangement and a metallic vapor within said container.
4. A container comprising an alloy portion, said alloy containing substantially 29.8% nickel, 15.5% cobalt, 0.22% manganese and less than 0.1% carbon, the remainder 'of said alloy being iron, saidA container also comprising a glass portion, sealed .to said alloy portion, containing substantially 67% Si02; ll/1% B202; il/2% Na2, K20; 2% A1202 and mercury vapor in said container.
5. A container having a metal portion comprising an alloy containing substantially 24% to 34% nickel, 10% to 25% cobalt, less than 1% manganese and the remainder iron and a portion of high silicate glass sealed to said metal portion, said glass containing 65% to 75% S102, 10% to less than 10% PbO and less than 6% A1203 and mercury vapor in said container.
6. A container having a metal portion comprising an alloy containing substantially 29.8% nickel, 15.5% cobalt, 0.22% manganese, less than 0.1% carbon, the remainder oi' said alloy being iron, said container also comprising a high silicate glass portion having an expansivity of 3.0 to 5.5 10-6 centimeters per centimeter per degree centigrade sealed to said metal portion and mercury vapor in said container.
7. A cathode assembly for an electric discharge device comprising a cup-shaped metal portion 23% to 34% nickel, 9%
forming a portion of the exterior wall of said device and a pool of vaporiz'able metallic substance in saidLcup-shaped metal portion, saidv metal portion comprising 23% to 25% cobalt, less than 1% remainder iron. f 8. A metal cup-shaped electrode container for electric discharge `devices comprising an alloy of 24% to 34% nickel, 5% to 25% cobalt, less than 1% manganese and the remainder iron.
9. A vacuum-tight container for electric discharge devices having a portion comprising an alloy of 24% to 34%' nickel, 9% to 25% cobalt, less than 1% and a high silicate glass portion sealed thereto and a mercury vapor within said container.
10. An electrical device comprising a container, a vapor conned within said container, said container being mainly of metal, a portion of the container wall beingglass, and a metal portion adjacent said glass comprising an alloy of to 25% cobalt, less than 1% manganese and the remainder iron, said alloy sealed vacuum-tight to said glass.
11. An essentially strain-free seal comprising a glass in fused combination with a metal having a transformation zone below the v annealing temperature of said glass, the thermar expansion characteristics of said glass and said metal being substantially Acoincident for all temperatures at, as well as above and below, saidtransformation zone, at least up to the softening temperature of the glass.
12. An essentially strain-free seal comprising a glass in fused combination with/a metal having a transformation zone below the annealing temperature of said glass and having .a thermal expansion characteristic which is substantially straight at temperatures below the transformation zonebut curves upwardlyattemperatures above the transformation zone, said glass having a transformation zone which substantially coincides with the transformation zone of the metal and a thermal 'characteristic the slope of which substantially matches theslope of the metal characteristic at all temperatures above to 34% nickel. 9% manganese and the and below the transformation zone, atleast up to the softening temperature of the glass.
13. An essentially strain-free seal comprising glass welded to an alloy ,containing iron. nickel and cobalt, said alloy having a transformation zone which substantially coincides with that .of said glass and a thermal expansion characteristicwhich substantially corresponds in value and slope with the characteristic of said glass at temperatures at, above,and below said transformation zone. o
14. A strain-free seal comprising a glass in fused combination with a metal having a transformation zone below the annealing point of said glass, the thermal expansion characteristics of said glass and said metal being substantially coincident at temperatures at, below and appreciably above the transformation zone and continuing along lines of generally similar curvature up to the softening temperature of the glass.
' l5. An article comprising borosilicate glass welded to an alloy containing approximately 18 per cent cobalt, approximately 28 per cent nickel, and approximately 54% iron, said glass having a transformation zone which is substantially coincident with that of the alloy, and athermal expansion characteristic which substantially matches that of the alloy at temperatures at, below and appreciably above the transformation manganese and the remainder iron o zone at least up to the softening point of the glass.
16. An article comprising a glass having a thermal expansion curve which has at least one inection point and a metal having a thermal expansion curve which has at least one inflection point, the lowest temperature inflection points of said glass and metal curves being close together and the slopes of said curves being substantially alike through an appreciable temperature range above said last-named inection points.
17. An article comprising a glass having a thermal expansion curve which has at least one inflection point and a metal having a thermal expansion curve which has at least one inflection point, the lowest temperature inflection points of said glass and metal'curves being close together and the slopes of said curves being substantially alike for a temperature range above the inection temperatures extending to a temperature at which strain is relieved in the glass when the glass is maintained at said last-named temperature for a predetermined time interval.
18. An article comprising a glass having a thermal expansion curve, which has at least one inectionv point and a metal having a thermal expansion curve, which has at least one infiection point, said curves being close together up to the inilection points which occur at the lowest temperatures for the curves and the slopes of said curves being substantially alike for a temperature range above the temperatures of said last-named inflection points which extends to a temperature at which strain is relieved in said glass if said glass is maintained at said lastnamed temperature for a predetermined time interval.
19. An article comprising borosilicate glass welded to an alloy containing 15.5% to 18% cobalt, 28% to 30% nickel and approximately 54% iron, said glass having a transformation zone which is substantially coincident with that of the alloy, and a thermal expansion characteristic which substantially matches that of the alloy at temperatures at, below and appreciably above the transformation zone at least up to the softening point of the glass.
20. An article comprising a seal between Corning glass G-705-AJ and an alloy comprising about 29.4% nickel, 16.3% cobalt, 0.2% manganese and the remainder iron. HOWARD SCOTT.
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US2431152 *||Mar 1, 1945||Nov 18, 1947||Westinghouse Electric Corp||Mercury vapor tube|
|US2447719 *||Jan 22, 1945||Aug 24, 1948||Eitel Mccullough Inc||Electron tube|
|US2479201 *||Jan 31, 1944||Aug 16, 1949||Philips Lab Inc||Geiger-muller counter|
|US2504303 *||May 23, 1947||Apr 18, 1950||Bell Telephone Labor Inc||Metal to glass sealing method|
|US2525468 *||Mar 29, 1943||Oct 10, 1950||Westinghouse Electric Corp||Ultra high frequency tube|
|US2724057 *||Jan 21, 1944||Nov 15, 1955||Westinghouse Electric Corp||Ionic centrifuge|
|US2773211 *||Jan 14, 1955||Dec 4, 1956||Westinghouse Electric Corp||Large sealed ignitron-tube|
|US3384399 *||Oct 19, 1965||May 21, 1968||Int Nickel Co||Metal-to-glass seal|
|US3780328 *||Dec 30, 1971||Dec 18, 1973||Electronic Control & Surveilla||Lighting unit for use underwater and in wet environment|
|US4310773 *||May 16, 1979||Jan 12, 1982||General Electric Company||Glass flash tube|
|US4499774 *||Jul 17, 1981||Feb 19, 1985||Hitachi, Ltd.||Semiconductor pressure transducer|
|DE1259103B *||Oct 20, 1965||Jan 18, 1968||Wiggin & Co Ltd Henry||Verwendung einer Nickel und Kobalt enthaltenden Eisenlegierung als Werkstoff fuer mit Bor-Silikatglas verbundene Teile|
|WO2001074285A2||Mar 30, 2001||Oct 11, 2001||Bhm Medical Inc.||Support structures|
|U.S. Classification||313/567, 313/32, 420/95, 220/2.30R, 313/171, 501/15, 313/328, 220/2.2, 313/634, 501/66, 313/29, 420/581|
|International Classification||H01J61/30, H01J13/10, H01J5/26, C03C27/02|
|Cooperative Classification||C03C27/02, H01J61/302, H01J2893/0043, H01J5/26, H01J2893/0077, H01J13/10|
|European Classification||H01J5/26, H01J13/10, C03C27/02, H01J61/30A|