Life of an o-ring

[dextermech]

How long should an o-ring last (silicone or PTFE) that will be around the OD of a 6" dia. 304 SS cylinder which will travel linearly 4" (bidirectional) once every 20 minutes? It is sealing 13 psi 250F clean air.

 

[Zapster]

Is there any lubricant?

 

[dextermech]

No lubricant.

 

[MikeHalloran]

You have pretty much described the worst case conditions for life of _any_ o-ring. The only thing that could make it worse is making its infrequent motion either very slow or very fast.

- O-rings generally require lubricant, even when used as static seals.
- PTFE and PTFE-coated o-rings won't seal against an imperfect surface, won't seal particularly well on a perfect surface, and don't wear very well.
- There's not enough pressure to 'energize' a PTFE o-ring.
- There's not enough pressure to energize a silicone o-ring, either, but that may be a blessing because silicone has poor abrasion resistance.

In my experience, one seal that has a chance of working in your application is the Bal-Seal.



Mike Halloran
Pembroke Pines, FL, USA

 

[dextermech]

Mike that BalSeal looks perfet for my application. Thanks a lot for the tip, I'll give them a call Monday.

 

[scastillo]

You may want to also look at Variseals, which I've used on a number of applications with good results

Here is the link:

http://www.busakshamban.us/busak.shamban.brands.htm?wlb=1

 

[dextermech]

The Variseal looks good too. I will have a dust issue as well...there could be fine plastic dust particles on the "piston" (actually a hollow tube with the OD machined) so a secondary dust seal would be in order. What approximately does a seal like this cost in a 6" ID?

 

[scastillo]

Depends on the final seal design and size of the order. No doubt it will be pricey, but at least you have the advantage of customizing the design to suit your application.

 

[strokersix]

Do you care if it leaks? If some leakage is acceptable then no o-ring at all might be a good solution.

 

[davefitz]

Sounds like the story of O.

 

[Cockroach]

This is an impossible question to answer. There are numerous issues you need to address, your application is basically a dynamic fit seal, reciprocating.

What is the stretch and squeeze on your elastomer?
What is the surface finish on the bore of the cylinder?
No lubricant; what is the frictional coefficient rubber/steel? Use the textbook average?
What is the average piston velocity?
Does the elastomer have a backup ring? Flat or concave?
What is the rubber durometer? 70/75 or 90/95?
Is the thermal operating level static or variant?

The average lifetime of an o-ring is a couple of years. You have very low pressure, but your thermal is elevated which greatly effects the lifetime on the rubber.

I would probably GUESS you're looking at a single year.


Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[plasgears]

You may want to develop your sealed product on the cheap, then upgrade as req'd. If rubber o-ring doesn't work, go to one of the specialty seals designed for translating shafts. The trick is to just barely get by with the least expensive.

 

[strokersix]

Perhaps I should rephrase my question. Is your objective to seal or is it to provide force and motion? A close fitting cylinder within a bore can provide a low friction seal with minimal leakage and long life if you can tolerate some nominal leakage rate.

 

[dextermech]

It is sealing 28" Hg vacuum, and yes I need it 100% sealed, or as close to 100% as I can get. Thanks for all the help so far.

 

[israelkk]

Did you consider rolling diaphragm instead of O-ring?

 

[dextermech]

Not sure what you mean.

There may also be plastic dust particles on the "piston" to deal with when sealing...

 

[israelkk]

The pioneers were Bellofram

http://www.marshbellofram.com/diaphragm.htm

 

[MikeHalloran]

Pay close attention to the design information that Bellofram provides. My experience says it's truthful and accurate.

For a stroke of 2/3 the minor diameter, and a non- reversing pressure difference, you'd certainly do better than 10,000 cycles at room temperature. At 250F, I'm not so sure. Ask Bellofram for help.








Mike Halloran
Pembroke Pines, FL, USA