US2898395A

US2898395A - Spark plug seal

Info

Publication number: US2898395A
Application number: US447792A
Authority: US
Inventors: Harry G Schurecht
Original assignee: Champion Spark Plug Co
Current assignee: Federal Mogul Ignition LLC
Priority date: 1954-08-04
Filing date: 1954-08-04
Publication date: 1959-08-04
Anticipated expiration: 1976-08-04

Description

Aug. 4, 1959 H. e. SCHURECHT SPARK PLUG SEAL 5 Sheets-Sheet 1 Filed Aug. 4, 1954 w m E mw M WM m m5 mu V r a Aug. 4, 1959 H. cs. SCHURECHT SPARK PLUG SEAL Filed Aug. 4, 1954 s Sheets-Shet 2 G R L m m m.

wmw 6 M 0 n7 L M 5 5 0 0 CH C6 CONDUCTING HOLDING POWDER GLASS SEAL IN VEN TOR.

Harry 6. Sc/rurecbf 1959 H. e. SCHURECHT 2,898,395

' SPARK PLUG SEAL Filed s- 1954 5 Sheets-Sheet 3 FIRING TEMPERATURE F 0.1. emmu HlLdV .L3T13d d0 unswvm INVENTOR- Harry 6. SC/wrechf BY ATTORNEYS Aug. 4,

1959 H. G. SCHURECHT SPARK PLUG SEAL Filed Au 4, 1954 5 Sheets-Sheet 4 INVENTOR.

LATTORNEYS g- 4, 1959 H. G. SCHURECHT 2,898,395

SPARK PLUG SEAL File A g. 1 54 5 Sheets-Sheet 5 IN VEN TOR.

Harry 6. Jc/rurec/fl BYKQLLQQuQ.

'HTTCRNE Y5 United States Patent SPARK PLUG SEAL Harry G. Schurecht, Detroit, Mich., assignor to Champion Spark Plug Company, Toledo, Ohio, a corporation of Delaware Application August 4, 1954, Serial No. 447,792

8 Claims. (Cl. 174-452) This invention relates to a method of producing gas tight seals for spark plugs and is particularly directed to the improvement of seals of the type commonly referred to as glass seals.

It has heretofore been proposed to seal the center bore in which the center electrode of the spark plug is disposed by fusing a layer of pulverized glass in the annular space between the electrode and core or, more commonly, above a lower section of the center electrode and making the seal of partially conductive material. The composition of the pulverized layer has been altered in various ways, in some instances to obtain a desired degree of electric conductivity, and in other instances for the purpose of reducing the tendency of the pulverized glass to boil during the heating step in the formation of the seal. One of the continuing difiiculties which leads to an occasional unsatisfactory seal is the inability of the glass to adhere to the center electrode, if the electrode is made from any of the usual metals. To overcome this, it has been suggested to use thin connecting wires of special metals for that portion of the electrode within the seal, the wire connecting a lower center electrode section and an upper terminal of the same metal as the electrode since the connecting wire has been thought to be unsuitable for the protruding terminal part. Such construction results in three piece electrodes that are expensive to manufacture and require individual assembly.

In known constructions the parts to be sealed are usually assembled with the sealing material either in the form of a layer or mass of powdered glass with metal or other powders added, or with the composite sealing material in the form of a pellet. The material is then softened by heat and the parts brought to their final position by forcing an upper terminal section into place against or through the molten glassy material. It is apparent that such a method requires individual heating of each spark plug and makes impossible the production of spark plugs on a mass scale in which the formation of the seal by melting of the glassy material takes place during the passage of an assembled unit through a furnace. The seal produced by a method in which the glass is pressed while molten is a dense, inflexible mass which functions well as a seal only if no significant difference exists in the absolute thermal expansions of the sealed parts.

2,898,395 Patented Aug. 4,1959

The glassy material has a tendency to boil out during the heating step and it has been proposed to reduce this tendency by adding to the glass a portion of the material of which the spark plug insulator is made, such material being finely powdered prior to its addition. Other attempts have been made to confine the glass sealing material during the melting step with all of the parts disposed in their final position but these attempts have required the formation of screw threads within the insulator bore and have been otherwise cumbersome and expensive.

In still another known structure soapstone is used as the sealing material, and the seal is effected merely by pressing the soapstone in place. If desired, after such a soapstone seal has been effected, the parts between which the seal has been effected can be further supported relative to one another by sintering a mass of aluminum metal powder in place between the parts, and above the seal. Such sintering can be effected by passing an electric current through the aluminum powder. It is to be noted, however, that the aluminum powder, in such structure, has no relationship to the formation of the seal, being placed only after the seal has been completed by pressing of the soapstone in place.

The primary object of the present invention is to provide a novel method of forming and retaining a glass seal between metal and ceramic parts of a spark plug in such a manner that the material cannot be displaced during the formation of the seal, which formation takes place with the glassy material under a pressure produced only by its confinement. The seal produced is vesicular and flexible and thus capable of functioning where great differences of thermal expansion exist between the sealed parts. The pores of the seal are, of course, separated by walls of the sealing material so that gas leakage from one end of the seal to the other is prevented.

Another object of the invention is to provide for improved adherence between the material of the seal and any metallic part with which it is in contact so that a gas tight assembly may be made around a one piece electrode without resorting to special metals for the electrode parts.

Another object of the invention is to provide for the production by means of a holding material of a satisfactory glass seal which will mature at a lower temperature than the same glass used alone, without the holding material.

Other objects and advantages of the invention will become apparent from the following specification, reference being had to the accompanying drawings in which:

Fig. 1 discloses a central vertical sectional View of a spark plug having a seal formed in accordance with the present invention;

Fig. 2 is a fragmentary sectional view of a modified form of electrode;

Fig. 3 is a central vertical sectional view of an insulator and a simplified central electrode assembly including a seal according to the invention;

Fig. 4 is a central vertical view in cross section similar to Fig. 3, but showing a central electrode assembly incorporating an electrically conducting glass seal;

Fig. 5 is a diagram showing expansion characteristics when fired as pellets, of certain glass-mineral compositions;

Fig. 6 is a vertical sectional View of an unfired assembly for producing a seal according to the invention;

Fig. 7 is a fragmentary vertical sectional view, after firing, of the assembly of Fig. 6;

Fig. 8 is a perspective view of several components of the assemblies of Figs. 6 and 7; and

Fig. 9 is a vertical view in cross-section, similar to Figs. 3 and 4, except that an electrode part extends through both the seal and the holding powder.

A typical spark plug to which the present invention may be applied comprises (Fig. 1) a shell 10 threaded as at 11 for engagement with the cylinder head of an internal combustion engine and carrying a shell electrode 12. The shell is extended into a shield barrel 14 which is lined with an insulating material 15. An insulator 16 is seated and fixed in the shell by a metallic sleeve 17 and supports a center electrode having a firing tip 18 in the usual co operating relation with the shell electrode.

The center electrode of Fig. 1, above its firing tip, comprises a stem part 20 received in a bore 22 of the insulator. Since there is a clearance between these two parts the space must be sealed off if a gas-tight spark plug is to result. The glass seal, according to the present invention, is thus placed at the upper end of the center electrode as is known in the art. The center electrode 20 has a reduced portion 24 connecting the enlarged stern and a head or terminal 26 which is also enlarged. The center electrode may be made of nickel or the commonly known 42 percent nickel-iron alloy and the reduced portion 24 may be made integral with the stem 20 and terminal head 26. It has been found that satisfactory and reliable seals capable of withstanding an arbitrary test pressure of 1000 to 1500 pounds per square inch may be formed around the nickel or nickel alloy wire by the present invention.

Nickel and nickel alloys tend to oxidize quite easily and the oxide coating may on occasion become so thick that the glass sealing material is not in contact with the wire but with its oxide coating. The interface of pure oxide leaks under pressure so that it has been previously considered difficult or impossible to seal satisfactorily to wires of these metals. I have found, however, that by using a material that penetrates the oxide coating, such as a lead borate, as a first coating, or by plating, for example with copper, to prevent oxidation, particularly around the reduced portion 24 of the wire, the glass seal together with the lead borate or copper produces a dependable seal. It has been found that the most practical size of nickel alloy wire, e.g., a 42 percent nickel-iron alloy wire, through the seal has a diameter in the range of .030 inch, although the size may vary upward from this limit if the lead borate coating is used, and the sealing material is compounded as hereinafter described.

Under some circumstances it may be desired to utilize a metal such as molybdenum or tungsten for the reduced portion 24 of the center electrode and since molybdenum forms a volatile oxide so that only a thin oxide coating is present at any time, the lead borate coating is not necessary for the production of a tight seal. Further, thicker molybdenum or tungsten wire may be used through the seals than wires of nickel or nickel alloys. Good results have been obtained with molybedenum or tungsten wire up to 0.094 in diameter.

The use of a thin molybdenum connecting wire extending through a glass seal and electrically connected to lower and upper electrode parts has been proposed in the prior art. Molybdenum oxidizes so rapidly above about 1500 F. that protrusions of the wire above the seal are nearly lost during the formation of the seal. I have found that the molybdenum wire may be adequately protected by plating with a more slowly oxidable metal, such as nickel, over those areas that extend above the glass during heating. The plated part may then be used ,as an electrical contact with the full expectation that suflicient contact area will be available to carry the needed current.

The glassy sealing material, which is indicated at 28, may comprise powdered glass hereinafter for convenience referred to as Glass A, subsequently identified in more detail, in admixture with an infusible substance insoluble in Glass A. The admixture of Glass A and an infusible substance must be one which, when packed into the space to be sealed and heated to a temperature at least as high as the softening temperature of Glass A, expands into a vesicular structure and occupies, upon cooling, a volume greater than before such heating. It has been found that Bayer process aluminum hydroxide, after dehydration and calcining at about 1200 F., yields an aluminum oxide that, when ground and mixed with Glass A, and with other glasses, produces an admixture having the requisite expansive characteristics upon such heating. It has also been found that incorporation of bentonite in such an admixture lowers by several hundred degrees F., the temperature at which the requisite expansion occurs. Optimum results have been achieved with mixtures of 60 parts by weight of Glass A, 40 parts by weight of alumina, and two parts by weight of bentonite.

Other substances can be substituted for calcined alumina in admixture with Glass A, for example, satisfactory results have been produced with admixtures of this glass and petalite, fused quartz, cordierite, and even powdered tungsten metal, zinc oxide and tin oxide. In fact decomposing materials such as sodium carbonate, lithium carbonate, and others may be used, and the requisite expansion caused to occur at a desired temperature. Whiting and talc, as well as bentonite, when admixed with a sealing composition have been found to be highly advantageous in causing the expansion to occur at a relatively low temperature. Further, glasses other than Glass A can also be used, as will hereinafter be discussed in more detail. The production of seals according to the invention will now be discussed in detail with reference to the use of admixtures of Glass A, bentonite and calcined alumina, for convenience called Admixture I.

In general, Admixture I and other admixtures used for producing seals in accordance with the invention have a tendency to boil and foam during the time that they are heated in the formation of the seal, and, unless confined, leave a seal of such high porosity that it may leak if a continuous path of bubbles is formed when the seal hardens. The unconfined seal shows in cross section a very porous structure sometimes containing voids that are nearly equal in diameter to the thickness of the seal. This leaves only a thin wall available to prevent the passage of gases. In other seals the porosity may be such that adjacent voids may be readily connected by failure of the separating walls so that the seal ultimately leaks even though it appears to be gas tight at its first test. The most desirable seal, of course, exhibits a fine grained but porous or vesicular appearance with all voids surrounded completely by an adequate layer of homogeneous, strong material. The vesicular porosity of the seal resulting from the present invention is such that it is flexible enough to withstand the stress set up by unequal thermal expansion of the metallic and ceramic parts between which it is disposed.

Such a seal can be satisfactory only if excessive foaming or boiling is prevented. The present invention overcomes the tendency to boil and confines the seal by means of a packing, designated 30, which may be placed as powdered or finely divided refractory materials, including powdered metal particles such as chromium, nickel, copper, iron, silicon or aluminum, mixtures thereof, powdered refractory oxides, or mixtures of s powdered metals with refractory oxides. It is usually preferred that the packing comprises powdered refractory oxide material, or that it comprises aluminum, nickel, or silicon, most desirably at least about 25 percent by' weight of powdered aluminum in the latter instance, because exothermic oxidation of the metallic aluminum in the course of firing reduces the minimum kiln temperature required to form a tight seal. It has been found that a glass seal made either in a conventional manner by mechanical pressing, or under the confinement of a refractory oxide holding material matures at a much higher temperature than one made of the same glass and confined by a material containing powdered aluminum. The fine metal powder-containing packing is placed after the glass-containing seal mixture has been tamped in place around the center electrode. The packing is carried down to a point below the terminal head 26 and is engaged under the head in such a manner that pressure from beneath the packing layer cannot dislodge it. It has also been found practical to utilize one or more anchoring grooves 27 in the terminal head 26 for the purpose of anchoring the refractory packing, in which construction the lower face of the packing material can be raised to coincide with the cylindrical part of the terminal head.

The simplified central electrode assembly shown in Fig. 3 comprises an insulator 45 supporting a center electrode 46. Attached to the upper portion of the center electrode 46 is a metallic extension 47 which is advantageously of relatively small diameter, and preferably of molybdenum or tungsten, although other metals can be employed. The glassy sealing material is indicated at '50, and the packing at 51. No anchoring grooves are required in this structure, it being only necessary that the metallic extension 47 protrude into the packing 51. This structure contemplates that the packing 51 is electrically conducting after firing in an oxidizing atmosphere, as hereinafter described. As a consequence, the electrical connection to the center electrode can be completed merely by making suitable electrical contact, not illustrated, with the packing 51, and attaching this contact to the ignition system of an internal combustion engine.

It has been found that the packing 51 is electrically conducting when it is produced by firing 100 percent of a metal powder, for example as described above, or a mixture of metal powders, or a mixture of metal powders and refractory oxides in certain proportions. When a mixture of metal powders and refractory oxides is used to form the packing 51, such mixture can contain as much as about 75 percent by weight of refractory oxides so long as such mixture contains at least about 25 percent by weight of aluminum in every mixture containing more than about 50 percent by weight of refractory oxides. Most desirably, such mixtures contain not more than about 95 percent by weight of aluminum. Optimum results have been obtained when the packing contains from 70 percent by weight to 90 percent by weight of aluminum powder, from percent by weight to 15 percent by weight of one of the other above-listed metals, and from 5 percent by weight to 15 percent by weight of glass or ball clay.

The assembly shown in Fig. 4 comprises an insulator 54 carrying a center electrode 55 having an upwardly extending metallic portion 56. An electrically conducting glass seal, which can contain, for example, 35 percent of tungsten metal powder, 60 percent of glass, and 5 percent of graphite (see Table V), is packed in place around the upwardly extending portion 56 of the center electrode, and is confined by a conducting packing 61 thereabove. The electrical circuit to the center electrode may be completed in the same manner as described in connection with Fig. 3.

.It ,is sometimes advantageous to use a powdered refractory material below the sealing admixture as well 6 as confinement thereabove. For example, a metal or metal-containing material, so proportioned, can assist substantially in heat conduction from the firing end of the insulator.

The structures represented in Figs. 3 and 4 can readily be produced by inserting the

center electrode

46 or 55 into a central bore of the insulating

core

45 or 54, inserting a pellet of the desired sealing composition, upsetting the pellet by pressing, inserting a second pellet of the desired holding composition, and upsetting that pellet by pressing. At least some structures can be produced by inserting both pellets, and then pressing and crushing the two pellets simultaneously to fill the space around the upwardly extending portion 47 or 56. The assembly is then fired without the application of additional pressure.

It has been found that the metal powder in electrically conducting holding compositions surrounding the upwardly extending portion of the electrode in the structure shown in Figs. 3 and 4 is effective to minimize or almost completely eliminate oxidation of the electrode portion during firing. Therefore, the previously discussed need for plating a molybdenum electrode is eliminated by this preferred structure.

When the parts have been assembled, the glassy sealing material tamped in place with the powdered packing tamped over it, the entire plug may be heated to a temperature above the softening point of the glass, for example 1300 F.*l600 R, or even 1700 F. in some cases, preferably not higher than about 1650 F., held at this temperature for a period of approximately five minutes, and then cooled. Such heating is preferably in an oxidizing atmosphere for most reliable and consistent results. Since the sealing admixture has been held firmly in place, the tendency to foam and boil is overcome.

As has been previously stated, an admixture of a fusible, vitreous sealing material and a reactive substance used to produce a seal according to the invention must be one which, when packed in the space to be sealed and heated to a temperature at least as high as the softening temperature of the sealing material, expands into a vesicular structure and occupies, upon cooling, a volume greater than before such heating. Such expansion of this admixture, during heating, sets up an internal pressure which facilitates wetting of the metal of the center electrode by the sealing material. The confinement by reason of the powdered packing thus has two distinct advantages, namely, of overcoming the tendency to foam and boil, and increasing the capability of a sealing material to wet the metal of the center electrode, but the pressure on the glass is not sufiicient to produce an undesirably dense, inflexible seal such as results from the application of high mechanical pressures.

In some instances an admixture used for producing glass seals according to the invention is one, for example Admixture I, which, when pelletized and fired to a temperature at least as high as the softening temperature of the glass, expands and, upon cooling, occupies a volume greater than before such heating. This characteristic of several admixtures is illustrated in Fig. 5 of the drawings. The data constituting the basis for Fig. 5 was collected by preparing various admixtures of a glass and a finely divided substance, pressing a measured charge of the admixture in a suitable mold, under a pressure of about 14,000 p.s.i., into a cylindrical pellet having a diameter of about 0.196 inch, a height of about 0.25 inch, and a central cylindrical opening of about 0.060 inch diameter, firing the pellet to a predetermined temperature, allowing the pellet to cool, and measuring the diameter of the fired pellet. The procedure was repeated using different firing temperatures. The curves of Fig. 5 of the drawings constitute plots of diameter of the pellet in inches, after cooling, against firing temperature in degrees F.

From Fig. 5 it will be apparent that Admixture I is not appreciably affected upon firing to temperatures below about 1050" F., that, upon firing to temperatures be tween about 1050 F. and about 1250 F., shrinkage of Admixture I occurs, while upon firing to temperatures above about 1325 F. the pellet expands to an extent such that it is larger, after cooling, than before firing. It has been found experimentally that consistently tight glass seals can be produced according to the invention using Admixture I, and firing to temperatures of about 1350 to 1400 F., or higher, but that such seals fired to temperatures below about 1300 F. are seldom, if ever, gas-tight.

A similar series of tests has been carried out using what is hereinafter for convenience designated Admixture II, which is an admixture of 60 percent of Glass A and 40 percent of alundum fused alumina. It will be apparent from a comparison of the curve for Admixture I with the curve for Admixture II that a much higher firing temperature is required to cause pellets of the latter to expand than is the case with the former. It has been observed experimentally that a substantially higher firing temperature is required to produce consistently tight seals according to the invention using Admixture II than is required using Admixture 1. Thus, in the cases of Admixtures I and II there is a correlation between the temperature at which a pellet expands into a vesicular structure and occupies, upon cooling, a volume greater than before firing, and the firing temperature required to produce consistently tight seals. In fact, in this instance, the correlation is virtually complete, since the minimum firing temperature for producing tight seals using Admixture II coincides closely with the temperature at which a pellet of such admixture expands to a diameter of about 0.204 inch, or approximately the diameter to which a pellet of Admixture I expands upon firing to about 1350 F.

Admixture III requires an even higher firing temperature than Admixture II to produce a pellet which has expanded into a vesicular form and occupies, upon cooling, a volume greater than before firing, or to produce consistently tight seals according to the invention. Admixture III is made up of 60 percent of Glass A and 40 percent of a ground ceramic material produced by firing a composition of about 96 percent alumina, small amounts of whiting, talc and bentonite, and various impurities, to about cone 31.

Admixture IV is composed of 60 percent of Glass A and 40 percent of fused quartz. It will be observed that pellets of Admixture IV did not expand even when fired to 1700 F. It has also been observed that leakage occurred at less than 100 pounds per square inch gas pressure when it was attempted to produce seals according to the invention using Admixture IV and firing at 1400" F., 1600 F., or 1800 F.

It will be apparent from the foregoing discussion of the characteristics of Admixtures I through IV that there is a definite correlation between the capability of a particular admixture to produce tight seals according to the method of the invention and the capability of that admixture to expand upon firing to a predetermined temperature into a vesicular structure having an increased volume after cooling. However, in some instances, it has been found that consistently tight seals can be produced upon firing to a given temperature using admixtures which, when pelletized and fired to that temperature, do not show such expansion. For example, in the case of Admixture V, which is composed of 80 percent of a glass which is hereinafter for convenience identified as Glass B, and subsequently identified in more detail, and 20 percent of natural or raw petalite, pellets showed no expansion on firing at temperatures up to 1200 F. However, when incorporated in a holding powder glass seal according to the invention, Admixture V produced, upon firing to 950 F.,

a seal which withstood a 150 p.s.i. leakage test, and, upon firing to 1050 F., a seal which withstood a 1000+ p.s.i.

. 8 leakage test. As will subsequently be discussed in more detail in connection with different experimental work with Admixture V, it has been determined that such admixture, when fired in place in an annular space between a ceramic insulator and a metal part of a spark plug, does show the permanent type of expansion into a vesicular structure which has previously been discussed. Although the invention is not limited to the following theoretical explanation, it is believed that these experimentally observed facts indicate that, for example, in the case of admixtures of Glass B witlrpetalite, some reaction between the admixture and the insulator, or the metal component, when the seal is fired in situ, is responsible for such expansion.

In the case of Admixture VI, percent of Glass B and 20 percent of fused quartz, substantial expansion of the pellets was observed on firing to temperatures below 1200 F. Tight seals according to the invention were produced, however, upon firing to 1050 F., while pellets fired at such temperature showed contraction after cooling. Reaction similar to that discussed in connection with Admixture V is believed also to occur in the instance of Admixture VI.

As a result of extensive work with glass seals, it has been determined that, when sealing is attempted around a center electrode, leakage is most likely to occur along that electrode. This experimentally observed fact is believed to be explained by several factors. One significant factor is thought to be that the electrode material almost invariably has a higher coefficient of thermal expansion than the sealing material itself. As a result of this difference, the electrode is likely to contract away from the sealing material as an assembly cools after firing, tending to leave a void directly around the electrode through which gas can escape. This likelihood can be minimized by using an extremely fine wire through the seal, so that the total volume that the wire contracts during such cooling is minimized. The difficulty can also be alleviated by using, as a center electrode through the seal, a material having a relatively low coefficient of thermal expansion, such as tungsten or molybdenum. These expedients have previously been suggested and constitute no part of the instant invention. They are useful, but have not been found to constitute a practical solution to the problem. It is believed that the instant invention further assists in solving this specific problem by virtue of the expansion, under confinement, of the sealing admixture (during firing) to a vesicular structure which, even after cooling, occupies a larger volume than before such firing. By virtue of the expansion, under confinement, the sealing material is compressed tightly around the center electrode. By virtue of its vesicular structure, the sealing admixture is capable of at least limited expansion to fill the void which would otherwise be left by contraction of the electrode during cooling.

It has further been found that an additional characteristic of a sealing admixture according to the invention can be of assistance in solving the problem of electrode contraction during cooling after firing. This characteristic is the ability of the softened sealing admixture to wet the particular electrode with which it is in contact, and to adhere thereto after cooling. In general, other factors being equal, pure glass is more effective, when softened, at wetting metal, and adheres thereto better after cooling, than is an admixture of that glass with, for example, a mineral material such as alumina. However, as has been previously discussed, in order to produce a tight seal according to the invention it is essential that an admixture of a fusible vitreous sealing material, for example a glass, and some substance which, under the actual firing conditions employed to produce the seal, is reactive to cause expansion of the admixture into a vesicular structure having a larger volume after cooling than before firing. The maximum amount of any such reactive substance which can be used in a sealing admixture is the maximum at which the softened sealing material still retains its ability to wet the electrode or other metal component, and to adhere to the electrode after cooling.

This limitation on admixtt: res useful for producing seals according to the invention will be apparent from a consideration ofthe curve showing the characteristics of Admixture VII (Fig. Admixture VII contains 60 percent of Glass B and 40 percent of natural or raw petalite. By comparison between the curves for Admixture VII and for Admixture V it will be seen that, solely on the basis of expansion when in pellet form, Admixture VII would be expected to be more effective in producing seals than Admixture V. Quite the opposite has been found, however, to be true. For example, when Admixture VII was used as the sealing composition, and assemblies were fired to 950 F. and 1050 F., it was found that the seals were tight at gas pressures of 100 and 200 p.s.i., respectively, but leaked under higher gas pressures. Seals made with Admixture V, but otherwise identical, when fired to 1050. F., withstood 1000+ p.s.i. pressure tests. This difference between Admixture VII and Admixture V is believed to be attributable to the decreased capability of Admixture VII, when softened, to Wet the electrode, and, after cool ing to adhere thereto, by virtue of its containing a large petalite addition.

The inability of Admixture VII, when softened, to wet an electrode and, after cooling, to adhere thereto, has been demonstrated by a simple test, as follows: Admixture VII was mixed with water to form a paste, which paste was then applied to a ceramic disc. Lengths of 0.050 wire, iron, stainless steel, nickel, and even molybdenum, were then approximately half imbedded, axially in the paste on different discs, and the resulting article fired to 1150 F. It was found that the wires fell from Admixture VII, after firing, either upon inversion, or upon application of slight pressure thereto. When the same procedures were repeated, using admixture V in place of Admixture VII, the wires were all found to be tightly adhered, after cooling.

It has also been found, other factors being equal, that a sealing composition having a low coeificient of thermal expansion, is more effective at producing tight seals than is another admixture having a higher coefiicient. In this connection, the use of a substantially non-expanding mineral as the reactive substance in a sealing admixture is further advantageous as such use lowers the coefiicient of thermal expansion of the admixture. This effect is especially significant when the fusible material is a low softening temperature glass having, as do most such glasses, a high coefficient of thermal expansion.

A slightly modified way for forming a seal according to the invention is illustrated by Figs. 6 through 8 of the attached drawings. The assembly, shown in Fig. 6 before firing, and in Fig. 7 after firing, comprises an insulator 65 carrying a center electrode 66 having an upwardly extending metallic portion 67. A pellet of a seal ing admixture 68 is shown in Fig. 6 disposed in the annular space between the metallic portion 67 of the elec trode and a bore 69 of the insulator. The pellet 68 can be composed of Admixture I, and produced by pressing as described above and firing at about 1150 F. A lead borate pellet 70 is positioned around the portion 67 of the electrode, above the pellet 68, and inside an enlarged axial opening 71 of a cup-shaped metal member 72 having a reduced axial opening 73 in its upper portion surrounding the electrode portion 67. The assembly of sealing material pellet 68, lead borate pellet 70, and cupshaped washer 72 is supported against upward movement from the position represented in Fig. 6 a tube 74 threaded into the upper portion of the insulator 65. The tube 74 can be tightened finger-tight against the upper portion of the cup-shaped member 72 while the assembly is cold. As the assembly is heated during firing it is believed that the lead borate pellet first melts and flows down along the upwardly extending electrode portion 67 to provide, in situ, a coating thereon, and that this coating is at least partially absorbed by the pellet 68. Asthe temperature of the assembly is increased above that at which the lead borate pellet melts, a temperature is ultimately reached at which expansion of the sealing admixture in the pellet 68 into a vesicular structure begins, as has been previously discussed in detail. Since the cup-shaped member 72 confines the Admixture I of the pellet as it expands into a seal 75 indicated in Fig. 7, the same conditions prevail as the seal cools as when a holding powder is employed, namely, the vesicular sealing material is compressed against the electrode extending therethrough. It has been found that seals which will consistently withstand the 1000 p.s.i. leakage test can be produced from Admixture I, using a lead borate pellet, even around an ordinary iron wire having a diameter as large as 0.090 inch by this technique, and using a firing temperature of about 1400 F. Satisfactory seals have been produced by this method without using the lead borate pellet where electrode portions corresponding to the portion '67 through the seal of a smaller diameter and other metals to which sealing is less ditficult have been employed. Satisfactory seals even around large iron wires, e.g., of 0.050 inch diameter, have also been produced without a lead borate pellet, by mixing 10 parts by weight of lead borate with parts by Weight of Admixture I, pressing into pellets, firing the pellets to 1130 F., assemblying as shown in Fig. 6, but without a lead borate pellet, and firing the assembly to 1500 F.

It will be apparent that the member 72 could be supported in various ways other than as shown in Figs. 6 and 7, for example by a permanent stud either threaded or press fit into the insulator, or by a weight which prevents upward movement thereof. Alternatively, the member 72 can be allowed to move upwardly as the admixture expands, and then forced back substantially to its original position. Where the member 72 is forced downwardly into the sealing admixture, it is not necessary that the admixture be one which expands upon heating, as described herein. This method for forming a seal where a member similar to the member 72 is forced downwardly into a sealing composition is claimed in a copending application.

The cylindrical lead borate pellet 70 is shown enlarged and in perspective in Fig. 8, as is the cup-shaped member 72. An alternative form of confining member is indicated at 76 in Fig. 8 as comprising merely a short length of a cylindrical member having an axial cylindrical opening 77 central thereof. More effective sealing has been accomplished with a confining member of the shape of the member 72, but satisfactory results have been achieved with the member 76.

As a result of the visual examination of sectioned seals produced according to the method discussed with'reference to Figs. 6 through 8 of the drawings, it has been positively established that Admixtures V and VI, discussed above, when fired-in situ, exhibit the requisite expansion, as discussed previously, and at temperatures which explain the production of satisfactory seals by the holding powder technique at the temperatures previously reported. It has also been observed that the presence of a lead borate pellet as discussed above and as shown in Fig. 6 increases the extent to which a sealing admixture expands upon firing to a given temperature.

During the heating to form the seal where a holding powder, e.g., a powdered refractory oxide, a powdered metal or mixture of powdered metal with refractory oxide or oxides is used to confine the sealing admixture, this holding powder is not completely melted in the usual sense, as it contains sufficient refractory oxide or metal to cause it to retain its shape. This is true even in the lcase where 100 percent powdered aluminum metal is used as the holding powder, of reaction which proceeds during firing, such reaction being either partial conversion of the aluminum to alumina, or reaction between aluminum and I the sealing admixture. The powdered.

metal or mixture thereof with refractory oxides is not a seal, but a holding powder; the holding powder itself, as distinguished from the major constituent thereof, must be more refractory, which means that it must have less plasticity or melt at a higher temperature, than the sealing composition it confines. In this connection, it has been found that a pressed body consisting essentially of powdered aluminum metal can be fired in an oxidizing atmosphere to about 1800 F., or almost 600 F. above the melting point of pure aluminum, without losing its shape. Similarly, where a solid confining means is used, e.g. the

member

72 or 76, such means must be infusible, i.e., must not melt under the actual firing conditions.

It will thus be seen that the present invention provides a glass seal for a spark plug in which the sealing material is prevented from foaming or boiling out while the seal is being fired and in which the confinement of the sealing material creates an internal pressure which promotes wetting of the metal electrode parts and of the ceramic insulator by the glassy sealing material. Further, it will be seen that the invention provides for a glass seal around a wire of nickel or nickel alloy, or even around a thick iron wire, with complete dependability so that it is unnecessary to take into consideration the possible formation of oxide coatings on the wire which might occasionally prevent the formation of a completely reliable seal. Since it is now possible to seal reliably around wires which are of material commonly used in spark plug electrodes, a one-piece electrode structure can be utilized with the advantageous ease of assembly and constructional results, so that it is possible to provide a completely reliable and permanent electrical path through the seal which is not likely to become oxidized in use by the formation of ozone at reduced operating pressures or broken by mechanical impact on the electrode. If it is desired to use thick molybdenum wires through the seal of Figs. 1 and 2 it is only necessary to form a single-piece electrode by fastening the molybdenum section to the lower electrode section 20, using a piece of molybdenum that has been plated at its upper end at 21 (Fig. 2) where it will act as an exposed electrical contact in the finished spark plug.

To produce a spark plug that is gas-tight, it is of course, necessary to seal the space between the core 16 and the shell (Fig. 1). This may be done by utilizing a glass seal placed in accordance with the present invention since the flexible seal produced is capable of withstanding without rupture any tensile stress set up by unequal expansion of the metallic and ceramic parts. Such a seal is indicated at 40 and the confining body of powdered refractory oxide or oxides, powdered metal or metal-refractory oxide mixture is indicated at 42. The sealing material may be introduced initially as a preformed collar held together by a suitable binder and the powdered confining material layer may, if desired, be similarly placed. The layers are tamped separately prior to melting of the glass, the confining material remaining relatively implastic during the formation of the seal.

As the steel shell 10 oxidizes quite readily there is some possibility that the glass sealing material will not adhere to the metal as completely as is desired. The adherence can be improved and uniformly good bonds produced by coating the interior of the shell with either lead borate or an enamel ground coat. The coating is, of course, restricted substantially to that area where the final seal is made. A typical ground coat that has been found suitable is used in the production of sheet steel enamelware and comprises:

Fluorspar 9.0

Percent Cobalt oxide 0.5 Nickel oxide 0.4 Manganese oxide 0.9

This mixture is fritted and mixed with about 6.75 percent of clay, Vallender clay being suitable. Any desired method of application to the shell may be used, and the coat is, of course, fired to form a glassy coating on the shell prior to the introduction of the sealing material as set forth above. It has been found however, that such coating of metallic parts is not essential to the production of fully reliable glass seals, as only reasonable care is required to minimize oxide coatings to an extent such that completely reliable seals can be produced without this extra step.

Glass seals have not, in the past, been used between the core and shell of spark plugs, largely, I believe, because the wide dilference in thermal expansion between the parts tore the sealing material away from one or the other. The seals thus were unable to withstand repeated heating and cooling cycles such as are encountered in engine use. By applying the seal with a coating tending to improve its adherence to the shell and by forming the seal under the pressure of its confinement by packing 42, a sufiiciently flexible body of glass mixture is formed that will withstand repeated temperaturechange cycles. Seals so formed have gone through an arbitrary acceptance test of 100 heating and cooling cycles Without leakage at 1500 p.s.i. test pressure.

While satisfactory seals have been produced in an oxidizing kiln atmosphere, they may also be produced in a controlled atmosphere furnace, such as a hydrogen furnace, with the advantage of preventing oxidation of the metal parts.

Fig. 9 of the attached drawings shows a complete spark plug including an insulator-electrode assembly similar to that of Figs. 1 and 2, but substantially simplified. The spark plug of Pig. 9 comprises a ceramic insulator carrying a center electrode 81 having an upper electrode part 82. A glass seal produced according to the in vention is indicated at 83, confined by a holding powder 84. The ceramic insulator is carried by a metal shell 85 threaded at 86 for insertion in the firing chamber of an internal combustion engine. The electrode-ceramic insulator assembly can be produced as described above by packing Admixture I, for example, in place in the ceramic insulator-electrode assembly, packing the holding powder above the sealing material, and then firing the assembly to fuse the admixture and to complete the seal.

Using Admixture V, and alumina produced by dehydrating Bayer process aluminum hydroxide and calcining at about 1200 F. as a holding powder, seals which withstand a 1000 p.s.i. gas leakage test have been produced upon firing to 1050" F. with the assembly of Fig. 9. Using Glass B and 25 parts of raw petalite around a center wire of a 42 percent nickel-iron alloy having a diameter of 0.020 inch, and the same alumina holding powder, tight seals (1000+p.s.i.) have been produced upon firing to 950 F. or to 1050 F. Identical results have. also been achieved using compositions comprising 85 percent of Glass B and 15 percent of petalite, 80 percent of Glass B and 20 percent of petalite, and 70 percent of Glass B with 30 percent of petalite, upon firing to 1050 F.

If, for purposes of comparison, but not in accordance with the invention, petalite which has been previously calcined at a temperature of either 2400 F. or 2600 F. is mixed with Glass B in place of raw petalite in any of the above proportions, it has not been found possible to produce a tight seal upon firing even as high as 1400 F.

Consistently tight seals which have withstood the 1000 p.s.i. leakage test have also been produced by the method discussed with reference to Figs. 6 and 7 of the Some TABLE IV ing admixtures, the type of confinement used, and firing temperatures at which seals withstanding the 1000+ p.s.i.

kage test have been produced are presented in Table IV below.

of the available experimental data concerning other seal- V, but it has been observed rial in producing seals according to the invention.

admixture except that petalite F. or 2600 F. is substituted 5 lea previously n Tables I quently Glasses A and B, to which reference has be f m u m w .w m m u u m F 6 e 3W mm o 2 2 1 2 2 1 2 .2 Vh m r G 1 v 1 1 y y v 1 a t m n+ Sh m bm fi 1 u l 1 1 1 1 n 1 1 h 0 u t m OOOOOO O WWW MEWJWDMM I o h n t f mm F mmmmmm122212244 me nmm n n n u n a g mmm us LLLLLLLLLLLLLLL .Wit t W u u u m n e .1 8 g S t a mm m a h m m u n w m m m n m m m r i w ur 3 2 4. 3 3 4 3 6 u w 9 ee w h wmmw H 1 1 u n 1 1 L 1 n 1 1 d c n m ae w k o o u u n n e a h HMMWAatOH C S P 55 It 21 m m am m P e mo eh m m m m w m m m m o o 1 8d MWW 2 2 2 2 2 2 1 2 2 5 m 61D m w 0 1 L 1 1 L 1 1 1 L L L n V g n m d ww 0 m@ e a en Mm 5% 0% man I E6 4 .l. V 5a mmm mm m m w m H. mm mummn enema n d e r f. b mmm 0 0 O O O 0 0 0 0 0 0 p T x S Ha mu m m w a m u u n a n w m m m m m .mmm MW 0 1 L 1 L 1 l 1 1 L 1 1 mww m P u b T S 6 nu H ..L C m M I S n g m e u m 0929 7698354833598228 02676979602829368886 m E M m A mmmm WHB QRw MMB HB WMH wmm wm% mm% ww %%E a m P L w s d MUHWm a M m B I! W m 0 0O 00 m i .1 m v m m fi a A O m Mk 3%22n 4mm11 h oembe T TUTHTH I" e n Tn a A 1 r. d d h 0 0 e C mm m "m u m n .m u .m u n m wfi n R h e .A A A AO H .1 cu b L L L S eem oem f. F m w 0 m o D b0. bowhb omhbow mm mm mm c n P a PB PBSAPBSAPBS Bs BsmGG i m w n 0 5 0 5 0 0 C n 1 1 2 2 3 5 T t 25953536 8 77 6&265 6590 S. m m nnna ssm 1 m wmmnm @6117 m M n C 11 4 4300 .1 .19 01 C62 5 S 550005555 .07 M m IN 9 J7 TT N n v. m 1 443331111 n. 67 a P P P r g .w n u n u u u u n n .m u m u u u u n mla. MW 0 n u n n u n e 0 6 m .0 .00 000 .000 h S a .2 24 .246 .246 C u u S S 2b 1 n u n n Y n n w H m G b n u u u u u u u n h n u a w e u n n n u mm Sp wrafiuxsfih %5%: o n u n M n u n n u v m 8 n7 & 7 0a "seas u d n n I n i m 1 n "7531 I I e hm m A u n n I B u n u H C g 1 I. E d S H .u. L H L M M n m m m Hm W a B d a a 3 .l .l. B l 6 w y T G F A G u n A G Ah b T t. S T S T e S a g 2 S n a 2 a .1 t n u u I u u m w m u u u u u u n u u u G m fi m C u n w vu 6 C m n P o S .0 S M e 6 5 S S n 5 lb 1 mm 0.w H M n n u n u u n u d w l t n t n d1 0 a 0 a O a a 3 3 3 3 3 mm pdwuwwwpddfi pwwdpm wo m m 1 1 1 2 1 2 2 m SPMTFCMNKLBHR P&AFMBHh PSAB h drawings using Admixture that virtually no expansion, and, therefore, no sealing, occurs using an identical calcined at either 2400 for the raw petalite.

been made, are more completely identified i and II and Glass C to which reference is subse It will be apparent that various changes and modifications can be made from the specific details described and represented in the attached drawings without departing from the spirit of the attached claims, and that the method of the invention, in its essential aspects, provides a method for forming a gas-tight seal in an annular space between a ceramic insulator and a metal component of an assembly. The method of the invention, in its essential details, consists in packing between the parts an admixture comprising a fusible, vitreous sealing material and a substance which, upon firing to a temperature of T", at least as high as the softening point of the sealing material, when the admixture is so packed, is reactive to cause the admixture to expand into a vesicular structure and occupy, upon cooling, a volume greater than before such heating, the amount of such substance in the admixture being sufiicient to cause expansion as aforesaid, but insufiicient to prevent wetting of the metal component by the sealing material, and confining the admixture in the annular space while heating at least the admixture to a temperature between T and about 1650 F., whereby the admixture is confined and compressed against the metal component by virtue of its expansion. The admixture used as above stated is ordinarily one which is capable of producing a tight seal if there is not an electrode or other part extending therethrough and also having the additional characteristics recited in this paragraph, and hereinbefore discussed in more detail.

This application is a continuation-in-part of the following copending applications: Serial No. 75,529, filed February 10, 1949, and Serial No. 392,395, filed November 16, 1953, both now abandoned.

What I claim is:

1. In a spark plug assembly comprising an insulator having a central bore and an electrode part supported in the insulator bore, the improvement comprising a substantially gas-impervious, vitreous sealing means disposed in at least a part of the insulator bore, and in contact with the electrode part, and restraining means at least partially confining said vitreous sealing means within the bore, said vitreous sealing means being expanded into a vesicular structure against said restraining means and pressed by its expansion into sealing relationship within the insulator bore, and acting to prevent the How of a gas therethrough.

2. In a spark. plug assembly comprising an insulator having a central bore and an electrode part seated in the insulator bore, the improvement comprising a substantially gas-impervious, vitreous sealing means disposed in at least a part of the annular space between the electrode part and the walls of the insulator which define the bore, and restraining means at least partially confining said vitreous sealing means within the bore, s'aid vitreous sealing means being expanded into a vesicular structure against said restraining means and pressed by its expansion into sealing relationship both with the walls of the electrode part and with the walls of the insulator which define the bore.

3. In a spark plug assembly comprising an insulator having a central bore and an electrode part seated in the insulator bore, the improvement comprising a substantially gas-impervious, vitreous sealing means disposed in at least a part of the annular space between the electrode part and the walls of the insulator which define the bore, and restraining means at least partially confining said vitreous sealing means within the bore, said vitreous sealing means comprising glass and calcined alumina, and being expanded into a vesicular structure against said restraining means and pressed by its expansion into sealing relationship both with the walls of the electrode part and with the walls of the insulator which define the bore.

4. In a spark plug assembly comprising an insulator having a central bore and an electrode part seated in the insulator bore, the improvement comprising a substantially gas-impervious, vitreous sealing means disposed in at least a part of the annular space between the electrode part and the walls of the insulator which define the bore, and restraining means at least partially confining said vitreous sealing means within the bore, said vitreous sealing means comprising glass and petalite, and being expanded into a vesicular structure against said restraining means and pressed by its expansion into sealing relationship both with the walls of the electrode part and with the walls of the insulator which define the bore.

5. In a spark plug assembly comprising an insulator having a central bore and an electrode part seated in the insulator bore, the improvement comprising a substantially gas-impervious, vitreous sealing means disposed in at least a part of the annular space between the electrode part and the walls of the insulator which define the bore, and restraining means at least partially confining said vitreous sealing means Within the bore, said vitreous sealing means comprising glass and fused quartz, and being expanded into a vesicular structure against said restraining means and pressed by its expansion into sealing relationship both with the walls of the electrode part and with the walls of the insulator which define the bore.

6. In a spark plug assembly comprising an insulator having a central bore and an electrode part seated in the insulator bore, the improvement comprising a substantially gas-impervious, vitreous sealing means disposed in at least a part of the annular space between the electrode part and the walls of the insulator which define the bore, and restraining means at least partially confining said vitreous sealing means within the bore, said vitreous sealing means comprising glass and cordierite, and being expanded into a vesicular structure against said restraining means and pressed by its expansion into sealing relationship both with the walls of the electrode part and with the walls of the insulator which define the bore.

7. In a spark plug assembly comprising an insulator having a central bore and an electrode part seated in the insulator bore, the improvement comprising a substantially gas-impervious, vitreous sealing means disposed in at least a part of the annular space between the electrode part and the walls of the insulator which define the bore, and restraining means at least partially confining said vitreous sealing means within the bore, said vitreous sealing means comprising glass and tungsten metal powder, and being expanded into a vesicular structure against said restraining means and pressed by its expansion into sealing relationship both with the Walls of the electrode part and with the walls of the insulator which define the bore.

8. In a spark plug assembly comprising an insulator, a metal part, and a sealing material disposed in an annular space between the insulator and the metal part, the improvement comprising a substantially gas impervious, vitreous sealing means disposed in at least a part of the annular space, and in contact with the insulator and with the metal part, and restraining means at least partially confining said vitreous sealing means within the annular space, said vitreous sealing means being expanded into a vesicular structure against said restraining means and pressed by its expansion into sealing relationship with the insulator, and acting to prevent the flow of a gas through the annular space.

References Cited in the file of this patent UNITED STATES PATENTS 1,999,785 Rohde Apr. 30, 1935 2,248,415 Schwartzwalder et a1. July 8, 1941 2,311,647 Doran Feb. 23, 1943 2,317,305 Schwartzwalder et al. Apr. 20, 1943 2,597,978 Doran May 27, 1952