"

"

" ~."

,,,_' •• J'

2 PHIL~PS TECH~ICAL, REVIEW VOL. 8, NC).,1

SINTERED GLASS.':·

, , by E. G. DORGELO -, 666.189.4

."...In the manufacture of certain articles in which many metal, parts (e.g. leading-inwires) must be fused into glass close toeach other, it is sometimes impossible, due to thetoo, low fluidity of the molten glass,'. to }orcè the drop of glass between the metal parts,

'. liJ. such a case glass in powder form can be used, and this can be introduced between themetal parts before fusingv'I'he glass :obtafued after fusing; which is not completely clear,is called sintered glass, and 'contains manyvery smallnir bubbles. In this ar;icle variousproperties and possibilities of this glass are discussed.' "

l'···' ., .•.•

, ' .In the. manufacture of incandescent lamps, elec- sufficiently far' into them, perhaps .because of thetronie 'and gas-discharge tubes different special fact that cooling takes place' too rapidly ;as a con-kinds of glass are' used, which must satisfy certain sequence of the heat 'conduction through the metalrequirements for each type, of valve or lamp, and parts. This difficulty can bë overcome if, beforetherefore maydiffer very much from each, other. the, fusing in; the glass in, fin~ powder form 'isA kind of glass whose' properties and constructive introduced directly into the'space where it belongs,possibilities are particularly favourable for one the' whole then being heated tó a, temperaturetype' may be quite unsuitable for another type. at .which the glass melts, The strU<lture, ~f the

For' the development of '.à ,new lamp or valve, somewhat turHi~ sinte~e'd', glass which, is obtainedtherefore, the possibilities offered by' different upon fusing the', powdér, is' .not, homogeneous; it'kinds of glass should be subjected to an extensive cont~ins, nuÎri.~ro~s'~~;y s~all 'i~s'o~:'iair ,hubhlèé' '.'illvèstigation. If the existing kinds of glass are which more or less modify the different propertiesunsuitable for the application in view, an attempt of the glass. In general these changes. are-net of -,'is ~ade'to find a new and better kind of glass. In prime importance. Nevertheless" they may some-th,e,~as~;o.f dif!er~nt type!, of, vf:ll~esthe :progi~ss of times, make possible constructions which 'are .impos- , 'their development depended almost exclusively sihle with .normal glass. Examples, of such, casesö~'the inari.uia"cture of a:suit~blé, ri.e:wkind pi"gïa~s~ where preference is' given to powder glass will, be" ,~Oneexariipl~,is thè"~bdi~~ 'l~~p coveredon the given later in this paper. ' . ,inside' \vith.'~bo~àt;, gl;.'ss1); ~hi~h' is ~r~~isïant' to, ,Sintered glass offe~s great advantages in thesoai~~' v~p~ui; i~hè~ the, .hlgh-préss~r~' ~nd sripe; manufacture" of valves and lamps for experimentalhigh:prè~sU:re !~eré~y lamps, w:h~re'it '~as' 'neces- purposes, due to the rapid and simple manner insary to find types of glass for the covering of the which almost any desired lamp base can be made.'nietalleads through the q~~it~2)._IIi'the dèvelop~ Metal Jeads can be fused in at the same time that'ment of transmittirïg valves f~r' very. shott waves' the base is made, while the process can be used foruse was also, successfully ~;'de -of'sp~ci~i-n:e~'kind~ 'every kind of glass; including the kinds which areof glass, the elèctrolysis-fte{glgss(~s 3). I, ';' -:~/",:, " very difficult' to soften. 'Whilé in the examples riienti~ried he~é;~thè 'rie;v'

glasses .are distinguished fro~ thé, older" ones .:by"thei~ chemical eoinposition, in the case of "p'ç)wdér.,,' ,; . v·- '.glass" an attempt has been" made to create newpossibilities by a m~di6cation in the physical struc-ture. The stimulus for this attempt lay in a difficulty"which. occurs ill the manufacture of articles in'which' many metal parts (for example leading-in,wires) must be fused into glass close to each other.,The .liquid gÏass must then be forced betwê'eri. themetal parts, and the method fails when the spacesb~tween a~e so small that' a drop of molten glass,even under high external pressure, cannot penetrate. '. .

I"

The employment ~f sintered glass

,The raw material, powdered glass, is' obtained'by grinding up pieces of glass. This powder is cast'in .a mould in which the met~l parts 'tQ be fusedin are ;lrèady,present, ,Care'must be taken th.áithe~gl~ss P!lwder fills up- the spaces wellhetween tli.,e_"metal parts. After cov~~ing the-mould the' wliöle is' ;heated:to,~a 'temperature at which the glass {s' very -fluid, s~ "that ~nly slight pressure is 'enoiÎ~ghti>;fly: rev~n" .the"~mallest èävities.· ';,' ,~.,' :,', ,: ':.-'~, The cöeffi~ient of expansion 'óf. the niater~~róf:the ;rilöuld' mu~t be adapted in ~'.'certain', iv~y:to ~'that of the glass. The wall of the' mould must notclamp the solidified 'glass 'article;· the c~efficient

,1

ji

1) Philips techno Rey. 2, 87, 1937.2) Philips techno Rev. 3, 119; 1938.S) Philips techno Rev, 6, 255, 1941.

JANUARY 1946

(fig. 5). To prevent the exhaust tube from collap-sing during the fusion, it is previously sealed at thebottom and filled with fine sand, which is shakenout after the fusion.

SINTERED GLASS

of expansion of the materialof the wall of the mouldmust therefore be smaller than that of the glass.The bottom of the mould (which is separate fromthe wall) must, on the other hand, have by prefe-rence the same coefficient of expansion as the glass,since otherwise there is a danger that, upon cooling,any leads and the like which are fastened into thebottom will be bent. A mould which possesses theproperties mentioned is shown in fig. 1.

Besides lead-pins, differently shaped objectscan also be fused in, for instance metal strips,tubes, nuts; etc. (fig. 2). The number of pins, thedistances between them and their grouping issubject to practically no limitations when sintered

Fig. 1. Mould of metal with separate bottom and cover.

glass is employed. Fig. 3 illustrates the greatvariety possible.

Furthermore it may be pointed out that simul-taneously with the' fusing-in of leads in a lamp orvalve base the glass envelope can also be welded on.This envelope is then placed in the mould beforethe fusing of the powder glass base; the upper partof the mould must then be removed. In this waythe separate welding process is eliminated. Thisprocess is, of course, subject to the restriction thatit can only be applied in those cases where theelectrodes inside the envelope are resistant to theheat radiation of the glowing mould, The tubesand bases shown in the photographs fig. 4 aremade in this way. It is obvious that other glass parts,such as an exhaust tube, can be welded in simul-taneously with the fusing of the valve or lamp base

Fig. 2. Fused-in metal tubes and strips in powder-glassbases for valves.

Fig. 3, The leads can he fused-in through sintered glass basesin almost any number and arrangement.

45580

A special manner of fusing in, which is impossiblewith glass envelopes, can be applied to metal enve-lopes by strongly heating the edge of the metalenvelope and then pressing it into the likewisepreviously heated sintered glass base. The pro-jecting edge can be removed later. In this way it ispossible, for example, to fasten lead-glass discsinto 'an iron can. Because of the fact that the coef-ficient of expansion of the iron is larger than thatofthe glass, the iron upon cooling is clamped around

Fig. 4. Envelope andpinches, the powder-glassbases of which are weldedto the envelopes at thesametimetha t the powderedglass is fused.

the glass, glvmg a very reliable connection 4).Nor is it of importance in this method whether thecross section of the envelope is a true circle.

Properties of powder glass

1) Specific weight

The circumstance that there is a very largenumber of small gas bubbles in sintered glassaffects different properties; it is clear that forexample the specific weight is smaller than that of

4) In principle this method of fusing-in is also possible withdiscs of clear glass.

4 PHILIPS TECHNICAL REVIEW VOL. 8, No. 1

the original glass. The decrease depends upon thesize and the number of bubbles and is usually of theorder of magnitude of 5 to 10 percent; with veryfine powder the decrease is greater. The diameter ofthe bubbles usually lies between 10 and 50 [1..

The number of bubbles per mm" amounts to severalthousands.

2) Electrical properties

When a block of powder glass is situated in anelectric field the field strength in the glass will be

homogeneous glass (..1., e and tan 0). Thus if

volume of all air bubblesp=

total volume

we find that:

Î.

2-2p 3--ROl--p+ ...2+p 2

(1)

e' 2e+I-2p (e-l)e 2e+l+p (e-1)'

Fig. 5. Exhaust tubes can he welded in simurtaneously with the fusing of the lamp or valve base.

much smaller than in the gas or air bubbles, as aresult of the very different dielectric constants(glass: approx. 7, air: 1). The field is thus coneen-trated in the bubbles.If we consider a piece of clear glass with only a

few Ia r ge air bubbles, the field concentration inthese bubbles may be so high tbat the enclosed gasbecomes ionized, which may lead to completebreakdown. Glass for high-voltage apparatus (X-raytubes, for example) must thus satisfy the require-ment of being free of bubbles to a high degree.

The situation becomes quitc different when theair in the glass is divided into 'very many smallbubbles. The potential difference between theboundary surface is then uniformly distributedover the numerous intermediate gas bubbles, sotbat the potential difference per gas bubble is solow that danger of ionization is ·out ofthe question.With powder glass indeed values of the breakdownvoltage are found which are just as high as thosemeasured on glass which is free of bubbles.

The electrical field in a medium in which there areglobular gas bubbles can be calculated 5). Differentelectrical constants of powder glass, such as theelectrical conductivity).', the dielectric constant e'and the angle of loss, determined by tan 0', canbe calculated from the corresponding values for the

') K. W. Wagner, Archiv für Elektrotechnik, 2, 382,1914.

45586

which, e.g. when e =, 7, gives

e' 5--4p 6-=--~I--p+ ...e 5+2p 5

tan 0' 2 (5-3p) 3-- = -- R::3 1-- P + (3)tan 0 10-3 P 10

(2)

The formulae are valid only when p ~ 1. Inorder to obtain an impression of the influence whichthe air bubbles have on the constants mentionedwe substitute p = 0.1. Then ).'(). = 0.85, e']e =

0.88 and tg o'/tg 0 = 0.98.A further lowering of the values of e' and ).'

can thus be obtained by distributing much airamong many small bubbles. For this purpose it isnecessary to start with very fine powder and duringthe fusion the temperature must be raised veryrapidly. to prevent the escape or flowing togetherof the air bubbles. By the addition of substanceswhich give off gas the percentage can be very muchincreased, which, however, involves a lowering ofthe strength of the glass.

3) Thermal properties

The heat conduction in sintered glass showsalmost the same variation with p as the electricalconductivity (see formula 1). The heat conductivityof powder glass is thus somewhat less than that of

JANUARY 1946 SINTERED GLASS 5

clear glass of the same composition. It is sometimesnecessary to take special precautions in fusingbecause of this fact.

The coefficient of expansion of powder glass isthe same as that of normal glass; the coefficientof expansion is not changed by the presence of airbubbles, which is a result of the well-known factthat a hollow body expands as if it were massive.

4) Tensions

Objects built up of more than one differentmaterial with different coefficients of expansionare, not in general free of mechanical tensions atevery temperature, This holds also for the weldingtogether of two kinds of glass and for the intro-duetion of a metal lead through glass. The occur-rence of tensions in a glass object often leads tobreakage and therefore methods have been de-veloped for the checking of this. The tensions makethe glass optically anisotropic and give rise tophenomena of double refraction, which can forinstance be made visible with a polarizationapparatus.

In this investigation of tensions it has now beenfound to the advantage of sintered glass that; inobjects manufactured with the help of this glassfewer mechanical tensions occur than in ordinaryclear glass. This can be demonstrated by fusingtogether in pairs discs of different kinds of glass.When the discs fused together consist of powderglass smaller tensions appear after cooling thanwhen the two discs are composed of the corres-ponding kinds of clear glass.

The explanation of this phenomenon is probablyas follows. Since the solidification, is accompaniedby a decrease in volume, tensions will appear inthe glass which are smaller the better the stillsoft glass is able to accommodate itself.

Now this is more easily possible in sintered glassthan in clear glass of the same composition. Ifthere is a large number of gas bubbles in the glass,as long as the surrounding glas~ is still somewhatsoft they can compensate the volume decreaseof the solidifying glass by expanding. This con-ception is thus based upon the fact that some partsof the piece of work become solid before the rest. Thisphenomenon often plays a part when metal compo-nents are fused in.

Completely homogeneous objects can also bemade free of tension when normal glass is used bytaking care that they are cooled very slowly.One of the advantages of sintered glass is thatthere is no objection to the cooling taking placemore rapidly.

Applications

One of the chief applications of powder glass is inlamps and tubes for experimental purposes. How-ever, it also oilers a number of advantages forother applications whiclr in many cases are impor-tant. In the following we shall mention severalof them.

1) In tubes with complicated electrode systems

As already mentioned, the glass can 'be introducedinto the mould in the form of finely divided powderso that it can easily fill all the small cavities. Corn-plicated pieces of work, such as lamp bases withvery many leading-in wires or with leads veryclose to each other can be made of powder glasswithout much trouble (see fig. 3). The freedom ofchoice in distance hetween the leads makes itpossible to place them in such positions that thesimplest and most logical assembly of the electrode

Fig. 6. Electrodes of a transmitting valve for short wavesassembled directly on the powder-glass base.

system is achieved. Thus in fig. 6 may be seen thegrid-cathode part of a short-wave transmittingvalve, in which the two electrodes are weldeddirectly to the leads. These leads form two coneen-tric circles with diameters corresponding to thoseof the cathode and grid. Such a construction excelsin simplicity and sturdiness, while the slightness

6 PHILIPS TECHNICAL REVIEW VOL. 8, No. 1

of the self-induction of the lead ele~ents makespossible a satisfactory functioning on short waves.In extreme cases, in order to keep the selfinductionas low as possible, it is possible to lead the electrodeitself through the glass..Three examples are shownin fig. 7 (see also fig. 2).

In the left-hand valve there are two electredesbent in a U-shape which pass through the glassbase while retaining their cross section, The self-induction of the lead part is very small here. At thesame time good cooling is hereby ohtained. In thecase of the middle valve a shielding plate separatingthe two parts of the valve is led through the base.

the mould, and it can be heated to such a hightemperature that even a very hard glass stillbecomes sufficiently fluid. It has hereby hecomepossible to make lamp bases of kinds of glass, suchas the so-called electrolysis-free glass, which couldnot formerly be so used and this again has op~nedthe possibility of constructing new types of valves,especially in the field of transmitter valves fordecimetre wayes.

3) Connecting seals

It is possible to fill the mould with layers ofpowdered glass having different properties, and

Fig.7. Several transmitt.ing valves for very short waves in which the electrodes themselvesare fnsed into the base.

The right hand valve of fig. 7 contains two leads inthe form of strips which together form a Lechersystem. Experience has shown that with clearglass and the ordinary pressing technique suchconstructions are only made reliable with difficulty.It is difficult to obtain a satisfactory distributionof the drops of glass. At the edges of the metalalso the tensions are often too high, so that thevalve cracks at that spot.

ln the case of powder glass the first difficulty isnon-existent, while less hindrance is experiencedfrom the second.

2) Crowded constructions

Such constructions often result in high tempe-ratures ofthe glass wall. Itis then necessary to haverecourse to glass with a high softening point. Normalboro-silicate glasses with a softening point of500-600° C are often too soft or are too poorinsulators at the high operating temperatures,so that harder glasses must be used. As arule these glasses cannot be pressed into narrowinterstices in the ordinary way. Before thenecessary pressure is reached the drop of glasshas cooled off too far. The pressing in of metalparts directly is even more difficult. Due tothe fact that pressing is unnecessary in powdertechnique, high requirements need not be made of

then to fuse the whole. By choosing powderedglasses of gradually increasing coefficients ofexpansion graded seals can be made, i.e. tubularparts which at one end can be fused to a glasswith a high coefficient of expansion and at theother to a glass with a low coefficient of expansion.Another possibility offered is that the base of a

valve which is to contain the vapour of an alkalimetal can be protected by a thin layer of resistantglass. It is well known that most kinds of glass arevery severely attacked by such vapours. In thecase of glasses containing lead for example such astrong reduction takes place that the glass turnsblack, due to the liberated lead. Now by firstscattering powdered lead glass in the mould andover it a thin layer of borate glass, a protectinglayer is formed. Borate glass is not attacked; itcannot, however, be used· alone, since the t~mpe-rature interval in which softening takes place isparticularly short. A variant on the foregoingis the use of glasses of different colour, by whichmeans all kinds of indications can be introducedon the object.

4) Rapid fusing-in and cooling

Notwithstanding the fact that the conductionof heat in sintered glass is smaller than in thecorresponding clear glass, it appears in heating

~; ~.. , ..........~.~.~.... - ~, -

.....

_.f •

c-

JANUARY 1946 v ,

.SINTERJ):D GLASS 7

that with sintered glass without many' precaûtiort~ ,a more equal heatlngthrough the whole object isobtained. Probably the numerousgasbubhles'present'in sintered glass ensure 'that duririg the heating'the radiation of heat i~ strongly' dissipated. Oneresult of this is that the so-calledpre-heating whichprecedes the fusing of alamp base to the envelope

. ' -can . take place quite ' rapidly without _cràcking·· .the tube. It is clear that this is of great importancein màssproduction. Theanne~1fu.galso can illmost,cases proceed quite rapidly due to the previouslymentioned great elasticity of the still not completelysolidified powder glass, so that the occurrence oflarge tensions is combated.' ;,

.; ..

:. \. . ~.

..t,. • ~ ",

50 YEARS X.RAYS

';In November 1895 WiIhel m Con r a.d R ö 11: t g en, professor at theuniversity of Würzburg, made his fust observations on, Xi-rays; hispublication is dated December 27th 1895 and entitled: "On a new Kind 'of Rays", and it was a communication to the "Würzburger medizinîsch-"physikalischen GeselJ,schaft"'(1895"p. 137-141) .. ; . i

Now, 50 years afterwards, the whole world ,is góing to commemorate. this important contrihution to our present day scientific insight.- and to0l,lI medical and physical instruments. .T.olook backward PIJ, the' road achieved during these 50 years is 'indeed ' .

worth while.There the interaction is'reflectedwhich took placebetw~cn thevarious developing branches of science and technique, and it is possible~. .,'

to find there how an amazing.quantity ofWOJ k has given life to a series ofmost important practical applications of X-rays. 'These are too welllaÎown '.to be summed up here, but they engendered,.too, a great array of veryelaborate apparatus, The Philips' Fa~t~riès and LaboratorieS have also. had their part in this development. Here it may. suffice to refer to the··38 publications whicn appeared in the first seven years ~fthis periodicalon the subject of X-rays and their applicationar 9 'of these articles' dealt

. ' with Xvray tubes and X-ray apparatus; 21 on applications and 8 ont~e methods used i~ th~se applications. In this number, too, with whlch the.Philips Technic~l Review re-enters th~ world, after the forced interruptionduring the German occupation, the reader ~L1 find' a .description of anX-ray apparatus whi~h. in many respects is representative of the ideasand methods which hàve developed in the domain of X-rays ...in the 11946volume of this journal we' hope to find from our X-ray

Department still more contributions on the development of special tubes,'apparatus and ~èthods;" of': research. 'Thus the great discovery of.W. C. Röntgen 50 'yea~s ago will be given worthy commemoration.

, "

'I'