Sealing a low-vacuum closed system

[dispersion123]

I've been researching this for quite a while and feel somewhat exasperated, so I thought I would ask more knowledgeable folk.

I need to seal a closed system for low vacuum, and I need to do it on a budget. My problem is that most of the information I have found deals with much higher vacuums (10^-6 torr) and the systems are continuously pumped. I just want an adequate solution, not overkill.

Problem Definition:
I am trying to make a small apparatus that demonstrates heat pipe operation using water. The system would have two chambers -- for evaporation and condensing -- connected by copper tubing. Both chambers will have glass windows so the boiling/condensing can be seen in action. The primary problem is sealing the glass windows to the chambers, but I will take advice on all other aspects as well (tube to chamber, valves to seal from vacuum pump etc...). I want the seal to last a while, so the demo unit can run without constant maintenance. The lowest pressure I will have to seal against should be 9 torr, the vapor pressure of water at 10° C. I expect the system to run well within 9-150 torr (i.e. H₂O vapor pressure 10-60° C).

Questions
How to seal the glass to the aluminum/copper chambers? (elastomer o-ring, epoxy, indium,...)
How to seal copper tubing to Al/Cu chambers? (compression fitting, solder, epoxy)
How to remove vacuum pump? (needle valve?)

I am thinking epoxy might work for joining/sealing the glass, due to the information I found in this http://www.fusor.net/board/view.php?bn=fusor_vacuum&key=1174261135. I can machine the parts I need (chambers), but for obvious reasons off-the-shelf solutions are preferred.

Thanks!

 

[Q_Goest]

The use of elastomers (o-rings) for sealing high vacuum is common. For lower vacuums such as you're suggesting, even RTV would work fine. I'd suggest a rubber gasket or o-ring and if you have any problems sealing with that, add a bit of RTV.

Copper to aluminum tubing - compression fittings will work fine. Probably the best and simplest method that's also very inexpensive.

When you talk about removing the vacuum pump, I'm assuming you want to valve off the chamber so you want to put a valve in. That's fine. Use a globe type valve. Note the valve will have one of the two sides that is also sealed by the bonnet seals. Don't put that side on the vacuum chamber side, you don't want to risk leaking through the bonnet seals and into your vacuum chamber. The best solution is typically a diaphragm valve though I've also seen weir valves used before, but these might be a bit more expensive and not really necessary in this case.

 

[dispersion123]

Already drafted a response and lost it because the system logged me off... urgh.

Thanks for the response, Q_Goest, I was beginning to get the impression that sealing wouldn't be difficult but I have a couple questions still.

I can get compression fittings and valves rated for 29.9" Hg vacuum, which sounds great but the big question I have is this: How is this pressure rated? Is this the pressure the fitting/valve can hold - and for how long? Most vacuum systems are constantly pumped, the question of sealing becomes, "Can my vacuum pumping system handle the cumulative gas loads of my seals at x pressure?". In a closed system, the question is how long will the seals hold a pressure below X? Which in my case is 9 torr. So I will rephrase my question:

How should I go about sealing a closed system, such that the partial pressure of air (O₂, N₂, CO₂) remains below 9 torr for about a year? That will ensure that the water will begin to boil as soon as heat as applied, rather than require the system to 'heat up'. I have a circuit that turns a MOSFET transistor into an effective little heater, which will work great for a "Push-Button to view in action" demonstration.

 

[Q_Goest]

My experience is with vacuum jacketed piping and cryogenic vessels. These are typically pumped down to around 10 microns (warm). Note 1000 microns = 1 Torr. The vacuum drops below 1 micron when the inner vessel or pipe is in contact with the cryogen - basically, any contaminants freeze out. To obtain that low of a vacuum level, the inner surfaces and all parts going into the vacuum space are meticulously cleaned and when put on a vacuum pump, the inner surface is heated to drive out any additional contaminants such as water. Once the vacuum space is sealed off, it won't be pumped down again for many years. It is expected to hold a vacuum for years despite having blow off plates and relief valves for example, that are sealed with nothing but an o-ring. So if the vacuum equipment you're familiar with needs a constant vacuum pump on it, I'd say there's something wrong with how the vacuum is being made. It sounds like there is either a leak or very poor cleaning of the vacuum space.

To get down to 9 Torr shouldn't be an issue. Conventional tube fittings will work. The rating provided by the manufacturer is an indication of the confidence they have that it can seal, it isn't a structural concern. Fittings have to be absolutely leak free for high vacuum applications but for a vacuum of only 9 Torr, I wouldn't think you'd have any problem with fittings or gaskets.

 

[dispersion123]

Thanks again Q_Goest, that is what I needed to hear. My experience with vacuum systems was a week long session required for my Engineering Physics degree. We operated a lot of vacuum equipment, all I remember was how easily leaks could botch pumping @ 10^-6 torr. I have read a number of articles about gas permeation through elastomer seals that seem to indicate pumping is required to maintain UHV. If you can seal sub-micron vacuum with elastomer seals on a cryo system for years, I'm sold.

Much appreciated!

 

[johnbbahm]

If you are attaching with with glass, with the system under vacuum, heat up the glass, and the negative pressure will seal itself. I have done laser tubes this way.
If copper you could coat the inside of the pipe with a layer of solder, after you reach the desired vacuum, heat and crimp the pipe

 

[sophiecentaur]

Welding / melting the tube to the pump would cause the tube to collapse and seal itself; just what you want. You would just need to make sure the glass isn't so soft that it collapses in on itself.

You could look at a commercial heat pipe and see how that's finished off.

 

[dispersion123]

Thanks for the suggestions! I like the idea of coating the inside of the copper tube with solder, heating and crimping it. Sounds like it will work great.

I think I may need to be clear about the glass, I will make the chambers out of copper/aluminum. The chamber will not be much more than a square pocket in the block of aluminum/steel, so I will use a pieces of square sheet glass (say 1/4" thick, 2" x 2" square) to seal off the chambers. I have a CNC mill available, so I can easily make o-ring grooves or seats for the glass.

If I made the chambers out of copper, I could use indium to join/seal the glass to the chamber, but that is an expensive solution. I don't think it is necessary.

 

[sophiecentaur]

When you've got it working, why not post a picture of it? Sounds a good project.

 

[sujeshy@gmail]

Sealing a high-vacuum system

Hi

Can anyone tell me where i can find system for high vacuum for filling oil into a capillary tube at 10-4 torr?

Regards,
Sujesh