Sealing vacuum chamber holes air tight

[zaillian]

TL;DR Summary

Question regarding how to best seal inlet holes in a vacuum chamber.

There is a very low pressure vacuum chamber with two small round holes on it, around an inch or so for inlets. What are the options for
1) Sealing the holes air tight
2) Sealing the holes air tight with inlet for low voltage electricity into the chamber.

I have been trying to search for ready made solutions which are preferred, but have not found any. Lacking those I have access to standard milling equipment.

Previously a plate was screwed on the hole with two o-rings, one between the hole and the plate and a other between the screw and the plate. The solution leaked air and did not work.

 

[Lnewqban]

Welcome, @zaillian !

What are the shape and dimensions of the low-pressure vacuum chamber?

 

[phyzguy]

How low pressure are you talking about? O-rings work OK down to about 10^-7 Torr. Lower than that you want metal-metal seals. A picture or drawing of the holes would help us give better advice.

 

[Vanadium 50]

How low pressure are you talking about?

I am quite concerned that this was not described in the OP, as it is critical to the answer.

 

[zaillian]

Welcome, @zaillian !

What are the shape and dimensions of the low-pressure vacuum chamber?

Thank you!

The chamber is small, maybe around 20 litres made from stainless steel. The holes are circle shaped on a flat surface. Previously some o-ring based solution was used which worked.

The ultimate pressure should be in the range of 10^-7 mbar, achieved with a diffusion pump.

In all likelihood the solution is trivial to the commoners here. However, I was not able to find specifications or parts and hence the question.

 

[phyzguy]

Well, O-rings should work. Again, a picture would help a lot. You say the old solution, "leaked air and did not work". Maybe the O-rings just need to be replaced and re-greased with vacuum grease. As for passing in low voltage, what you want is a feedthrough. You can buy one, or you can drill a hole in the plate and seal it up with TorrSeal.

 

[Paul Colby]

https://www.mks.com/p/100311605

Something along these lines should work if the flat surface surrounding the hole is smooth enough.

 

[gleem]

Previously a plate was screwed on the hole with two o-rings, one between the hole and the plate and a other between the screw and the plate. The solution leaked air and did not work.

I'm not sure what was done but this is how I think it should have been done.

 

[Baluncore]

I'm not sure what was done but this is how I think it should have been done.

O-rings seal by being pushed against the joint by differential pressure.

For a vacuum, the cap offset should be the negative of what you show. The system shown would seal against high internal pressure as the o-ring would increase in diameter to make the seal.

As shown, the o-ring will be pushed against the opening on the vacuum side, which will damage the surface of the o-ring as it is extruded into the gap, against the sharp edge. That inner joint should have a very small gap, mostly metal-on-metal inside with the vacuum. There should be a bigger gap, maybe 0.1 mm, where the screws attach the cap, so atmospheric pressure can push the o-ring against the joint.

 

[DaveC426913]

That inner joint should have a very small gap, mostly metal-on-metal inside with the vacuum. There should be a bigger gap, maybe 0.1 mm, where the screws attach the cap, so atmospheric pressure can push the o-ring against the joint.

Like this?

 

[marcusl]

A zillion companies make vacuum feedthroughs. (Ok I exaggerate, it’s not quite a zillion.) here are a couple
https://apexvacuum.com/electrical-vacuum-feedthroughs/
https://www.pavetechnologyco.com/
You can do your own search for more.

 

[Lnewqban]

The holes are circle shaped on a flat surface.

It seems important to verify the flatness and rigidity of that flat surface.
The O-ring, or any recommended metal seal, should work between two finely machined and perfectly parallel surfaces.

 

[Vanadium 50]

It seems important to verify the flatness and rigidity of that flat surface.

Why?

Why not simply buy something that will work out of the box?

 

[Lnewqban]

Because ...

Previously a plate was screwed on the hole with two o-rings, one between the hole and the plate and another between the screw and the plate. The solution leaked air and did not work.

The chamber is... made from stainless steel. The holes are circle shaped on a flat surface. Previously some o-ring based solution was used which worked.

 

[Baluncore]

The holes are circle shaped on a flat surface. Previously some o-ring based solution was used which worked.

But it stopped working later, probably due to lack of grease and O-ring damage.

 

[Dullard]

People who do this for a living use any of a number of different types of standard vacuum flanges. They would weld (via tube nipple) to your existing opening and provide a 'standard' interface for whatever you wanted to connect (plug, feed-thru...). Search QF, KF, and/or ISO flanges for info.

 

[Paul Colby]

The OP does specify “the best” solution. Best is subjective. In a pinch I would just hold the blank off in place by hand while roughing out the system. If I didn’t hear hissing and it roughed out in a timely fashion I’d start the diffusion pumps. I had data to take.

 

[gleem]

I exaggerated the gap in my diagram. I do not think the O-ring's compression should be determined by the fastening screws but by the depth of the groove so the metal-metal contact should be on the atmosphere side of the O-ring. This also helps outgassing issues with metal-to-metal contact surfaces within the chamber.

 

[Baluncore]

I do not think the O-ring's compression should be determined by the fastening screws but by the depth of the groove ...

That is correct. An O-ring is not a gasket.

... so the metal-metal contact should be on the atmosphere side of the O-ring.

That is where we differ. What you write is true for pressure vessels, but not for vacuum chambers.

The O-ring in the groove, must close the gap above and below, yet still be free to slide radially between the two parts within the groove. Without that side closure, differential fluid pressure will not push the O-ring against the junction of the parts.

There should be a minimum gap between parts on the low-pressure side of the O-ring groove. That is where the pressure seal will occur, and where the O-ring will be extruded or damaged if there is any significant gap.

The differential pressure should increase the force that pushes the O-ring against the junction of the parts. Any gap on that sealing junction, will reduce the force multiplication by the differential hydrostatic pressure.

 

[gleem]

There should be a minimum gap between parts on the low-pressure side of the O-ring groove. That is where the pressure seal will occur, and where the O-ring will be extruded or damaged if there is any significant gap

At least for O-ring materials and vacuum applications that I am familiar with it takes a lot of force to compress the O-ring sufficiently on the flat surfaces (the main sealing area) to ensure a seal compared to what atmospheric pressure can apply to push it in. The atmospheric pressure is 14.7lbs/in² but the compression force per inch on 1/8 in. diam O-ring is a bit more than a pound/in, vs the fastener force of up to 15 lbs/ in. to compress it to the manufacturer's specs.

 

[Baluncore]

If you use the O-ring as a gasket it will be more difficult to seal than if the O-ring is in a groove. The O-ring only needs sufficient compression to prevent the high pressure moving above or below. Once the pressure is applied, the pressure will seal the O-ring in a groove, but not between two flat faces.

There are captive ring seals that are rubber bonded to a metal surround. They do not require a groove to retain the O-ring and prevent extrusion. When they are placed between flat faces, the rubber ring is forced against the metal outer, sealing against the flat surfaces. For a vacuum, the metal ring would need to be inside the rubber ring.

 

[bobob]

If you are going to use o-rings, just get a copy of the parker o-ring handbook. The handbook will give you all the design dimensions for o-ring grooves, tolerances, etc for whatever size o-ring you want to use. Yes, it's easy to seal against a flat surface as long as the surface is free of scratches and clean, even without vacuum grease. If you need vacuum grease, check your surfaces and clean those up. I have not had a problem sealing down to mid 10^-7 without using grease.