US2217423A

US2217423A - Glass-to-metal seal

Info

Publication number: US2217423A
Application number: US77221A
Authority: US
Inventors: Scott Howard
Original assignee: Westinghouse Electric and Manufacturing Co
Current assignee: CBS Corp
Priority date: 1936-04-30
Filing date: 1936-04-30
Publication date: 1940-10-08
Anticipated expiration: 1957-10-08

Description

Gd, 8, 1940. H. SCOTT 2,217,423

GLASS-TO-METAL SEAL Original Filed April '50, 1936 Fig. 1.-

' X3 Alloy 1784 -Ll I L o 100 200 300 400 500C.

Empemi'ure LL! I l l 1 l 0 I00 200 300 400 50070.

WITNESSES:

wig

Patented Oct. 8, 1940 UNITED STATES PATENT OFFICE GLASS-TO-METAL SEAL vania Application April 30, 1936, Serial No. 77,221 Renewed July 1'7, 1940 2 Claims.

This application is a continuation-in-part of my application Serial No. 376,291, filed July 5, 1929, now Patent 2,062,335 granted Dec. 1', 1936, and relating to seals between metal and glass.

My invention relates to vacuum-tight seals and it has particular relation to seals between electrical insulating materials of the class typified by glass and metallic compositions of the class typified by ferrous base alloys.

For many applications of vacuum-tight joints of the character named, the use of a hard or borosilica glass is of advantage. The high strain point or temperature at which such a glass will first begin to soften permits the assembled apparatus to be treated out under high temperature vacuum conditions without danger of collapse. In cfiect, the strain point represents the lower limit of the customary annealing range of the glass.

Iron-base alloys are of especial advantage in glass-sealing applications because of their low cost, easy fabrication, immunity from attack by mercury, and ability to wet glass. Such alloys may be compounded from iron, nickel and cobalt to have mean expansivities very closely matching those of the described hard glass from a static or room temperature up to the inflection temperature of the alloy.

As above used, the term static designates the lower limit of the temperature range to which the seal is in service to be subjected and the term inflection designates the temperature above which the expansivity of the alloy increases rapidly. The point of this abrupt rise in coefiicient of expansion of the metal may, in the case of the ferrous base composition above mentioned, be made to approach the strain point of commercially available grades of hard glass. It is only above the strain point that stresses from the differential expansion between the metal and glass can be relieved substantially by plastic flow.

These iron-base alloys, however, are further characterized by having expansion-temperature curves which decrease in slope as the temperature rises above room or static value and then start again to rise more rapidly as the infiection temperature is approached. The expansion curve of hard glass, on the other hand, is substantially linear in form within the range of temperatures below the strain point.

Typically, therefore, the curve of an iron-base alloy of comparable mean expansivity will cross and recross that of hard glass. Such a condition is contrary to the past accepted theory of glassto-metal sealsthat both the glass and the metal should have the same coemcient of expansion for each degree up to the annealing range of the glass. In the light of this theory, others have predicted many practical difliculties in sealing ferrous-base alloys into hard glass.

I have discovered that these difliculties are by no means insurmountable and that exact identity of expansion per degree up to the annealing range of the glass is not necessary between the glass and the metal if the overall expansions of the materials between static or room temperature and the strain point of the glass are substantially the same. In accordance with my discovery I am able to make satisfactory seals between the two materials named.

My invention will best be understood through the following description of two specific embodiments represented by Figs. 1 and 2 of the accompanying drawing.

In Fig. 1, I have represented at 10 the expan- 2O sion-temperature curve of one composition of hard glass known as Meth and at 12 the corresponding curve of a ferrous-base alloy which I have found to produce a satisfactory seal with this glass. This alloy is composed of 31.9% of nickel, 9.8% of cobalt, 0.79% of manganese, 0.01% of carbon and the remainder of 57.4% of iron.

It will be noted that whereas curve I0 is substantially a straight line, curve I2 is sumciently irregular in shape to cross and recross at points M and i5, respectively. Between the strain point l8 of the glass and the static or room temperature, however, the overall expansivities of the two materials are substantially the same. The differences at intermediate temperatures are, I have discovered, insuflicient to produce strains of an objectionable or disruptive nature.

In Fig. 2, I have represented a diiierent combination of sealable materials. Curve 20 applies to G-BO glass composed of 72.4% S102; 10.2% B203; 9.8% NazC); 5.10% A1203; 1.75% PhD; and 0.40% K20.

Curve 22 applies to a ferrous-base alloy suitable for sealing into this glass, which alloy is 5 made up 31.8% of nickel, 16.0% of cobalt, 0.65% of manganese, 0.01% carbon and the remainder of iron.

In the combination represented in Fig. 2, curve 1 22 also crosses curve 20 of the glass at two points, 24 and 2B. The overall expansion, however, between the strain point 28 of the glass and the static or room temperature is substantially the same for the two materials and no damaging stresses are set up at intermediate temperatures.

It will be obvious that other equivalent coinbinations 01' glass and ferrous-base alloys complying with this condition may be successfully Joined. My invention, therefore, is not to be re- 5 stricted except insofar as is necessitated by the prior art and by the scope -01 the appended claims.

I claim as my invention:

1. A glass-to-metal thermal joint in which the 10 glass has an overall expansion between the static temperature of the joint and the strain point of the glass substantially the same as that 01' the metal between these temperatures and a coei'ficient of expansion substantially difl'erent from the metal at an intermediate temperature.

2. A vacuum-tight seal between a ferrous-base alloy and a glass having an overall expansion between the static temperature of the seal and the strain point of the glass substantially the same as that 01' the alloy for the same range of temperatures and a thermal coeflicient of expansion diii'erent from that of the alloy at an inter mediate temperature.

i HOWARD SCOTT.