Fusing metal and glass

[Rotate]

How is this done? I've seen it done when lightbulbs are made but doesn't the wire contract more than glass when it cools so that there's a gap?

Is there some special preparation that's done to the metal so that glass fuses to it better?

Albert

[Alistair Hosie]

Albert in a small town near me called Oban they make glass most of it is handblown or if not blown then all of it is hand made .It is a joy to go and watch the guys make paperweights for which they are particularly famous those guys can pick up a glob of soft molten glass and attach jut about anything which will give an effect into it .Old nails are pushed in and pulled back out to create bubbles small strands of metal and other pieces of glass are added of different colours all this adds to the overall effect.When the glass is soft it is like toffee/candy and it can stay that way for quite a time to be worked on before being reheated so it is not as difficult to do as you would think regards Albert from Alistair.

[wierdscience]

Well after all you have to remember that the metal is also heating and expanding too and its the first thing to cool off so it probibly wouldn't make to much difference.I noticed something similar when pouring some closed frame babbitt bearings with the shaft in place the babbitt heated and expanded the shaft enough so that when it all cooled the shaft had shrank to allow perfect oil clearence.I guess it boils down to the rate that a given material gives up heat.

[Tim#1]

Rotate,

A good example of the metal-glass seals is with tungsten and certain types of Pyrex glass. You heat the tungsten rod to white heat and it forms an oxide layer. You can then fuse Pyrex to it in the form of a short piece of Pyrex tubing. This is all done in a gas-oxygen flame (I use oxyacteylene).

The glass and oxide layer form a bond or seal. Additionally, for this to work, the metal and the glass most have similar coefficients of thermal expansion. If the difference is too great, the glass will crack. In lightbulbs, I believe they use dumet or kovar wire with soft glass; both form an oxide layer and the coefficients of thermal expansion for these alloys and the glass are similar.

So, these glass-to-metal seals are really a bond between a metal oxide layer and the glass. Hope this answers your question.

Regards,

TIM

[[email protected]]

Another way to get around the problem of sealing a metallic conductor though glass is:
Flatten the conductor very thin where it will pass through the glass. Once the thin metal is sealed in the glass the strenght of the glass is great enough to deform the metal as they heat and cool. This method was used for many years in physics and chem labs to make experimental items. Make sure the glass is plenty thick where the metal passes through.

[Samuel]

I think that you might be looking for the words: coefficent of expansion, in the glass industry it is used to determine the compatibility of different colors/chemical compostion to eachother. The farther the #'s difference the more likely internal stress will cause " defects " wich brings me to the point of annealing. its quite important, but the larger the work piece the more critical it becomes. I hope this helps

Regards,

Samuel

[yf]

I remember reading somewhere that the wires passing through the envelope of a light bulb are a special alloy that has the same coefficient of expansion as the glass used for the bulbs.

[Herb Helbig]

Rotate -
Chemical supply houses like Welch Scientific and Fisher sell kovar seals to solve this problem. These take the form of glass and kovar tubes sealed together so you can use glass blowing technique to seal the glass to a glass vacuum system, and solder the kovar to metal components. Kovar seals come in many diameters.

If you want to seal electrodes into a metal system, you can buy feedthroughs from places like Ceramaseal.

A trick I found useful is to use printed circuit boards with copper on both sides sealed with o-rings to a metal vacuum system. You can etch the copper away selectively to provide insulation where needed, and holes can be drilled through the board and filled with solder to provide a conducting path into the vacuum. The glass-epoxy circuit boards are very strong and have very low vapor pressure.

------------------

[Oso]

The old way to do this is to upset the wire to form a very thin fin all around it.

When fused into the glass, this fin maintains the seal even if the rest of the wire loses it.

If you thin out the wire, it will eventually break, sooner than the fin technique.

The correct alloy, which may be dumet, is better.

[Endmill]

Books on glassblowing or older books on high vacuum discuss different techniques. As indicated, the coefficents of expansion must be matched. (Soft glass has too high an expansion for reliable seals, which is why it must be annealed to remove strains, but
early day devices used platinum wire. Light bulbs and radio tubes used a less expansive glass (Nonex), and Dumet was developed to seal with it. In most cases, several layers of different glasses are used to make what is called a graded seal.

Laboratory equipment is usually made with Pyrex and Kovar seals are used. The Kovar is cleaned and treated in a hydrogen furnace, and coated with a sealing glass (Uranium or 7052) and then sealed to the pyrex. Tungsten can also be sealed in pyrex. It is first coated with an oxide layer by using sodium nitrite as a flux, coated with the sealing glass and then the pyrex. Copper oxide will also bond with glass in thin strips called a Housekeeper seal. Glass/metal electrical feed throughs are available commercially for buiding neon signs, and in a wide variety for rebuilding TV picture tubes etc.
Bob