Dynamic O-ring seal design - floating O-rings

[Mike R]

Ok, Chris's difficulties with his battleship Ohio's engine spool valve o-rings had me nerding out and revisiting one of the bibles of the industry - the Parker O-ring handbook (ORD-5700  https://www.parker.com/content/dam/Parker-com/Literature/O-Ring-Division-Literature/ORD-5700.pdf )


In it I found something new to me - and I think relatively new in general (haven't checked how long ago it snuck into the handbook).
In the Dynamic Seals chapters, page 5-16, section 5.25 it talks about Floating Seals. 
This is actually very much a guide to how many people have been using o-rings as piston rings on our model engines, and it has the dimensional guidelines for using o-rings in this manner further down (page 5-35).


I'd be interested to hear form modelers who have used piston ring o-rings how close their implementation is to the handbooks recommendations. 
To me there are 2 big takeaways:
1.  Small amount of leakage on change of direction of the piston
2. Up to 60% reduction in breakout friction versus "normal" piston rod o-ring seal design

[crueby]

Just took a quick look - some interesting ideas there. I've never used more than one ring on a piston, though I have done what could be called a sorta-floating ring  seal, not with the float on the inside though, just with the groove wide enough for the ring to float side to side slightly. Not what they are describing really, more of just having cut the groove too wide!

I wonder if that would be worth trying on my current build, next three cylinders are much larger diameter, 2.5 to 3.18", than I am used to doing. Hmmm.
I'll have to do some more reading, that looks like a great resource. Thanks for the link!

[RReid]

That's really interesting. For my most recently completed engine, the Halfa, I used nominal 1/16" o-rings. My piston gland dimensions were 0.075"W x 0.064"D. The Parker table says I should have used 0.075"-0.079"W x 0.065"-0.067"D. I think I can live with that! So does the engine, as it seems to run quite well. The previous engine, the 2-Bits, also used the same size 0-rings, but with slightly tighter gland dimensions, 0.070W x 0.065"D, and also runs well.

[Jasonb]

Not been following all of Chris's thread or read the Parker stuff but I tend to us efigures From Model Engineers Handbook based on testing done by Arnold Throp ( Corless engine fame) and they generally compres sthe ring about 5% which is a lot less than many commercial table sshow and is enough to get a good seal but does not create excess friction. Depth taken from bore so take off any piston clearance to get actual ring groove depth.

That sounds like the Parker figures as 1/16" nominal ring is approx 0.070" so 5% off that would be 0.0665" depth

[derekwarner]

Technical catalogue Publications listings whether BS or ANSI Standards, will list recommended o-ring cavity sizes for industrial applications of 100 Bar or 250 Bar WP design...not for our lower pressure applications

So these Publications or Standards documents  are indicative only of where we should be considering

Derek

[internal_fire]

Technical catalogue Publications listings whether BS or ANSI Standards, will list recommended o-ring cavity sizes for industrial applications of 100 Bar or 250 Bar WP design...not for our lower pressure applications

The Parker description very clearly says this approach is for "low pressure", not over 200 PSI.

The concept is that the inner shape does not squeeze the o-ring at all; it just floats freely. The only (small) compression comes from stuffing into the bore.

Gene

[Mike R]

Ron,
Thanks for confirming that in your application you're close to what they are recommending.  I think all of these publications are great starting points and we can deviate as needed to achieve our desired results.  That's what its all about for me anyways. 

Derek,
This is why I posted this - the section specifically mentioned is a deviation from the "normal" high pressure practice and deals with low pressure (les than 200psi) pneumatic dynamic sealing - in a different way - i.e. low level of squish of the o-ring. In a sense its a validation / confirmation by a big company of the way modelers have been using O-rings on some small models for a while now, and gives another resource to reference when designing a piston to use an O-ring.