Page 1 of 4

Please let us know your experience with glass to metal seal

Posted: Tue Nov 04, 2014 7:32 am
by steve_rb
All those have some experience with making glass to metal seals please post your experiences here.

Re: Please let us know your experience with glass to metal s

Posted: Tue Nov 04, 2014 11:31 am
by Dennis P Brown
I have created glass to metal seals that can withstand cooling rates of over 1000C/sec and they have create strong bonds that are vacuum tight with copper wire even as the temperature ranges over a few hundred C. The single trick is to buy the special glasses that match the CTE of the metal. Sounds simple (and is) but these special glasses are generally costly (and some are custom orders only.) There are glasses for iron/steel, copper and even aluminum. Just google for the companies but Schott glass is your best bet. I should add - this should be considered only if you can melt the glass yourself (some are very low temperature melting - under 300 C.) They sell blocks of the glass so you need to make it into the size/shape you require - these are not tubes made of glass.

Re: Please let us know your experience with glass to metal s

Posted: Tue Nov 04, 2014 7:42 pm
by Carl Willis
As a very amateur glassblower with a self-sustaining hobby making decorative plasma tubes, I have made tungsten-borosilicate seals via uranium glass and Houskeeper ribbon seals from copper straight to borosilicate. Both types are challenging to perform reliably in my experience. The Dumet or platinum seal to soft soda-lime glass is easy by comparison and makes a very reliable seal, but you probably don't want to make apparatus out of soda-lime glass.

Given the experience, I strongly recommend hunting down some pre-made lead-ins rather than making them from scratch using materials of unreliable characteristics like tungsten wire. For wire seals, you are looking for a product known commercially as a "lead-in wire", preferably pre-beaded so you don't have to bother with the delicate challenge of prepping the wire and fusing a glass sleeve to it. The tricks when buying pre-beaded lead-ins are finding someone who will sell a small quantity (something less than thousands) and making sure you ask for one specifically for sealing COE33 borosilicate glass. They are all sourced in China or India. I haven't succeeded in finding a stable small-order supplier, but I have had a couple companies send me free "samples" of 50 or 100 leads. I figure these will last me essentially forever, although the sizes aren't appropriate for all jobs. You want to prepare an appropriately-thin-walled flange in your glass to accept the lead-in so that you don't risk overheating it (which will cause the joint to form bubbles and possibly leak). For sealing into larger, heavier walls, you make a maria in a piece of tubing that will slip over the lead-in bead and can be fused to it, and then you prepare a flange on the piece of glassware that will seal to the maria. Seals must be annealed in a kiln.

Borosilicate neon electrodes are another possibility. They are expensive, $5-10 apiece. (Contrast with Dumet-sealed lead glass neon electrodes, which are about $1 each.) They contain reactive gettering materials and are not suitable for some applications.

If you want a metal-tube to glass seal, you should look through the offerings from Larson and MDC / Insulator Seal. These seals are variants of the Houskeeper seal, but are strain-relieved through graded glass transitions. The cheapest and most common such seal is made from borosilicate glass to Kovar. The Kovar can in turn be soldered, brazed, or welded to other metals. It's worth noting that Kovar is embrittled by silver in brazing alloys.

Good luck!
Carl

Re: Please let us know your experience with glass to metal s

Posted: Wed Nov 05, 2014 12:18 am
by prestonbarrows
Unless you have access to scientific glassblowing equipment and fair skills (I'm guessing this is not the case or you wouldn't be asking), compression fittings are the simplest way to go for smaller diameters.

Example:
http://www.lesker.com/newweb/flanges/ad ... cfm?pgid=0

Just weld/braze onto a flange and you are done.

If you need diameters over an inch or two, you will likely want something pre-made. As mentioned earlier, this usually involves Kovar metal parts since the coeficient of thermal expansion is the same as many glasses/ceramics. Must it be glass or just an insulator? I would suggest going with brazed ceramics if you need large diameters. The ends of the ceramic are plated with metal than brazed directly to Kovar end caps. For large diameters, ceramic is much stronger and cheaper in general than a glass to metal seal.

Example:
http://mpfpi.com/stock-products/breaks/ ... reaks.aspx

Re: Please let us know your experience with glass to metal s

Posted: Wed Nov 05, 2014 5:46 am
by John Futter
Steve
You still have not replied to the intended use but I use Hysol 1C or torr seal works a treat to low ten to the minus 8 millibar and will handle 150 degrees C just

Re: Please let us know your experience with glass to metal s

Posted: Wed Nov 05, 2014 8:27 am
by steve_rb
I would like to join an about 70mm cylindrical glass to about the same diameter cylindrical metal to make a 60 KV insulation. Room temperature will be OK. I have heard usually some sort of wetting agent should be applied between the glass and metal interface for better sealing/temperature coefficient matching but couldn't find any info about glass to metal seal wetants.

Re: Please let us know your experience with glass to metal s

Posted: Wed Nov 05, 2014 8:36 pm
by Carl Willis
Hello Steve,

You have described a very challenging seal for a hobbyist to make by himself, and a costly one to obtain from a professional. I think there are preferred alternatives to the glass-metal seal in this situation as you have described it.

A master scientific glassblower would make this seal on a glass lathe by the Houskeeper method, and then anneal the assembly in a kiln. The diameter of 70mm is much too large to attempt without a glass lathe. You are correct that the Houskeeper seal requires a flux, but this is formed in situ in a thin layer by controlled oxidation of the metal (e.g. with boric acid in the case of copper). If your university in Tehran has a first-rate scientific glassblowing shop, you can see if they will do it. However, even a good glass shop would probably try to buy a suitable seal from a specialist manufacturer like those mentioned earlier in the thread. Either way, this will be a very expensive seal.

Ceramic vacuum breaks are available (as Preston mentioned) that are more durable and less expensive than a glass-metal seal in this size and that don't involve the sharp edge used in a Houskeeper seal, which can be a failure point if exposed to high electric fields.

I can't help but wonder if your real needs would be served by one of the standard feedthroughs used on hobby fusors all the time. One reliable approach to pass the higher voltages into a fusor is to use an inexpensive ($200 USD) 20-30 kV feedthrough insulated on the air side with oil.

Re: Please let us know your experience with glass to metal s

Posted: Thu Nov 06, 2014 5:39 am
by Andrew Robinson
I know absolutely nothing on this subject, but you guys peaked my curiosity. Discovered this video as a result. Pretty cool example of apparently what Carl was referring to: https://www.youtube.com/watch?v=OWw32BLodjY

Re: Please let us know your experience with glass to metal s

Posted: Thu Nov 06, 2014 6:48 am
by steve_rb
Interesting Video.
Also I found Varian TorrSeal resin epoxy bonder can stand up to 10^-9 torr. Anyone has used this bonder? How it behaves at 50 KV?
http://www.pchemlabs.com/product.asp?pid=2144

Re: Please let us know your experience with glass to metal s

Posted: Thu Nov 06, 2014 5:29 pm
by John Futter
steve
i'm not surprised you found torrseal as I had mentioned it to you further up the thread-- note Hysol 1C is exactly the same at a fraction of the price.
I use both up to 150kV with no apparent problems