Linear Collider Forum

Home » Detector Development » LC-TPC » Gas Sealing
Gas Sealing [message #1122] Thu, 20 September 2007 07:19 Go to previous message
Messages: 23
Registered: September 2007
Location: University of Rostock / D...
Dear all,
In yesterday's phone meeting there was a partly quite puzzled discussion going on about the shape of the O-ring grooves of the LP-TPC endplate. Since I put quite some effort and "research" into the design of the O-ring sealing of the electronics-testing prototype of the University of Rostock, I would like to share the results from this with you:

There are basic requirements for a groove that have to be fulfilled: First of all a very good surface quality is necessary. For the walls of the groove 0.4 m (16 in) Roughness Ra, for the top and bottom surface 1.6 m (63 in) Ra. The manufacturers told me that the shape of the groove doesn't really matter, as long as these surface qualities can be guaranteed.

Since it is easiest to produce a high quality rectangularly shaped groove this one is usually preferred. For high pressure applications triangular (nowadays trapezoidal with larger length open) grooves provide certain advantages. By using them, the effect of the O-ring being pressed into the tiny opening between the two contacting surfaces (so called extrusion) is moderated. However, there are other ways to achieve that goal nowadays (so called supporting or thrust rings). Also it is more laborious to produce these groove types with a sufficient surface quality, since you need specially shaped tools to cut precise radii and angles. We should also not forget that these solutions are only necessary for applications, where the O-Ring itself is deformed by the overpressure - seriously something like 70 bars.

For the purpose of holding the O-ring in place e.g. in a heavily moved lid, there is also a trapezoidal groove in use, where the smaller side is open. The O-ring may be inserted there once and will be safely kept in place. However disregarding the complicated turning of such a groove, involving several steps, it has other drawbacks: removing of the O-ring from such a groove will most likely result in its destruction. Much attention has to be paid to the edges of the groove, they have to be rounded with a certain radius in order not to damage the O-ring when inserted.

A special situation is given, if an O-ring sealing is used for gases or vacuum. Since gases can penetrate such a sealing easier than liquids (on which the dimensioning in industry standards usually rely on) special modifications have to be done. Some manufacturers have developed recommendations for such applications. These actually change a little bit the working mechanism of the O-ring seal. In a conventional O-Ring seal the medium that shell be sealed is allowed some space to have full contact to the O-Ring and deform it, press it onto the opposite side of the groove. Thus the O-ring groove has to provide some extra spacing in order to allow the medium to enter it. That's why the groove, you usually find in the standards (DIN, ISO, ...) are quite a bit wider than the O-ring diameter.

For applications involving gases, however, the concept changes such that one tries:
1. to cover as much surface as possible with the O-ring (in order to moderate possible scratches)
2. to block any possible way the gas could take through the groove
That leads to the famous four point contact, mentioned by Dan. Here the groove is dimensioned such, that in the compressed state the O-Ring contacts all four surfaces of the groove. This concept is widely used in vacuum applications, as the pressure of max. "-1 bar" directing inside the vessel is not enough to make the mechanism used for liquids work efficiently enough. Our application with 4 mbar overpressure ranges in the same category.

I used such a rectangular groove (3.03 mm x 4.6 mm) for a 4.0 mm thick O-ring with 215 mm inner diameter made of NBR 70 ("normal rubber") for my chamber. I tested the gas sealing of the chamber and found no measurable pressure loss at all, having it under pressure for two weeks. The grooves where machined into an aluminium alloy called here AlMgSi1. In order to protect the surface of the aluminium I decided to have it anodized. This process is, I think, very recommendable for aluminium parts since:
1. it increases the surface hardness to the level of corundum an such protects the groove very well against later damages caused by handling,
2. In the pre-process for this process minor scratches produced accidentally during machining are levelled
3. one can colour the surface in a very scratch resistive way (was an issue for me, since I plan to use it with laser tracks at a later stage)
4. it doesn't cost much
One drawback is, that the optical quality of the surface depends on the alloy. By chromatisation (?) one can keep aluminium surfaces permanently elctrically conductive (may be important for grounding).

In my case, were the groove was machined to fit exactly the inner diameter of the O-ring I observed a "fall out probability" of 50%. The manufacturers state, that the O-ring should be stretched to less than 3% to 5% of the original diameter, when build in. This little underdimensioning could be used to hold the O-ring inside the lid, which would avoid the complications introduced by a trapezoidal groove.

Something, that could interest Dan is that I also used the O-rings in a quadratically arranged groove (instead of the usual circular). Here the manufacturer states, that the bending radius in the corners shell be at least 4 times the O-ring thickness in order not to introduce additional material stress to the rubber, causing leaks. Specially shaped O-rings could be manufactured if this limit can not be reached, but they will be very expensive, since the tools have to be machined for manufacturing them.

Other things to consider are:
- Hardness: 70 Shore is the standard O-ring, softer ones (smaller numbers) should be even better for our application, since the are flowing better into small distortions of the surfaces.
- chemical resistivity: Viton could be good, since it resists most chemicals and comes in many versions. NBR (normal rubber) seems to be good enough for non aggressive Argon mixtures, like the ones used at DESY. (I'm thinking of CF4 as a problematic gas here). PTFE of course is also a good material, but it can not be greatly deformed and thus requires an even different groove-design.
- storage: O-rings shell be stored at room temperature, dry, light protected, oxygen protected, ozone protected. Depending on the material they can be stored such for 5 (NBR) to 20 (Viton) years before loosing functionality.

Oh yes: all the things stated above are mainly considering a static sealing, if you have to seal moving parts, like pistons, it's again different and other problems occur.

A very helpful source of information I found in the Technical manual ("O-ring book") of a company named ERIKS Also radiation influence is discussed in there. But also standards, and other more or less up to date books about O-rings where regarded.

Please consider also, that O-rings are only one way of a sealing design for our application. I have heard for example, that Aachen is thinking/designing a flat gasket, which, due to it's larger surface, could provide some advantages as well.

I hope these remarks will be helpful for the further discussion of this matter.

Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic:Ion Effects
Goto Forum:

Current Time: Thu Feb 21 16:46:27 Pacific Standard Time 2019
.:: Contact :: Home ::.

Powered by: FUDforum 3.0.1.
Copyright ©2001-2010 FUDforum Bulletin Board Software