O-ring groove design for non-circular groove 4

[RayJ2]

Hi,

I am designing a static seal with an o-ring.
The particularity is that the groove is non-circular. It has an oval-like shape (see picture in link below).
I am wondering how to calculate groove dimensions.

The medium is compressed air of 10 bar.
Pressure is on the inside, so the o-ring is pressed to the outermost groove wall.

I was thinking to calculate the groove the standard way as if the o-ring was placed in a circular groove, using some online calculator or the Parker Handbook.
Then, I would calculate my oval groove so that the outside wall ('outside circumference') of the oval groove has the same length as the outside wall of the circular groove.
Groove width and depth according to the standard calculations.
Would this be an acceptable calculation method?

Kind regards
Ray

 

[hydtools]

Yes.

Ted

 

[Eltron]

Sounds good except that I would use the centerline of the oval groove and the centerline of the circular groove instead of the outside wall.

Dan

http://www.eltronresearch.com

Dan's Blog

 

[jbeckhou]

In the past if I needed a custom made oring, I would allow the manufacture of the oring to design the groove, provide dims and tolerances, etc. Just my 2 cents.

Petrotrim Services

 

[c0rd0bes]

Hi just my opinion here, I have made numerious Oring grooves - albeit for ultrahigh vaccum applications.

Its a static application (flanged faced) so simple square / rectangular groove design.

I would design the groove and oring so that it hugs the inside circumfrence of the groove with a small % stretch compression. This is purely paractical for assembly disassemble and retention etc..

Height of the groove approx 2/3rds of the O ring O.D.
Width of the groove will be a function of the area left required to give the appropriate compression percentage & area for the oring to deform into.

Work in terms of oring compression percentage most manufacturers data can range from 18 - 22 % oring compression (personally i have used 25% for ultrahigh vaccumm systems using viton orings)

Hope this can help

 

[CorBlimeyLimey]

c0rd0bes ...

I use very similar for our static HV and UHV Viton/Kalrez O-ring grooves;
1/3 compression.
Cross-sectional area of groove 5%< or = cross-sectional area of O-ring, to ensure the O-ring completely fills the groove.

 

[MikeHalloran]

I would urge some caution in approaching 95..100 pct fill of the groove with the o-ring.

O-ring dimensional tolerances are rather large relative to typical groove tolerances.
Most o-rings arrive toward the low side of the tolerance range, because manufacturers still buy material by weight.
Once in a while, you'll get an o-ring that comes in near the large end of the tolerance range for dimensions, and hence volume. Put that in a groove designed for nominal size o-rings, and you can't close the joint, or you'll fracture something while trying.



Mike Halloran
Pembroke Pines, FL, USA

 

[CorBlimeyLimey]

Mike ... I always check the O-ring diameter before using.

And I forgot to mention that, where possible, the grooves inner perimeter is sized such that the O-ring is slightly stretched when inserted.

 

[hydtools]

Alternatively, it works if the outside of the groove is smaller than the OD of the o-ring requiring a squeeze to install the o-ring. Leave excess gland volume so that the o-ring does not completely fill the gland when the joint is closed.

Ted

 

[MikeHalloran]

ISTR there is a slight increase in service life of typical compounds with circumferential compression as hydtools describes.



Mike Halloran
Pembroke Pines, FL, USA

 

[RayJ2]

@cOrdObes,
"Height of the groove approx 2/3rds of the O ring O.D."
--- I assume you mean 2/3 of the o-ring cross section diameter?

@CorBlimeyLimey (whatever happened to simple usernames like Joe or José ..? ;-))
"Cross-sectional area of groove 5%< or = cross-sectional area of O-ring"
--- So you mean a negative groove fill ? (groove section smaller than o-ring section?)

@hydtools,
"it works if the outside of the groove is smaller than the OD of the o-ring requiring a squeeze to install the o-ring. Leave excess gland volume "
--- I was planning to do just that. What I haven't told you is that this particular application requires that the seal is often 'opened'.
The groove is in the 'ceiling' of the application, so I want that the o-ring is compressed into the groove to prevent it from falling out when the seal is opened.
The inside circumference needs to be sufficiently small to provide space for the o-ring to be squeezed. In other words, the groove needs to be sufficiently wide for the o-ring to expand.

 

[CorBlimeyLimey]

RayJ2 ... Yes, except the 5% was a typo ... should have been 2%.

 

[israelkk]

RayJ2

As a developer of high pressure systems for up to 20000 psi (~700 atm), for aerospace systems, for the last 30 years, I can assure you that your logic it correct. Since the pressure comes from the inside it tends to stretch the o-ring until it will touch the outer side of the oval. This stretch will degrade, crack or tear the o-ring with time. Therefore, Eltron's suggestion to use the center line is wrong and sc0rd0bes's suggestion to hug the inside is even worst (the maximum air pressure difference in vacuum systems is only 14.7 psi (~1 atm) which is quite small compared to your case. However, sc0rd0bes's is right in case of vacuum systems where the outer pressure is larger than the inner pressure. Regarding to assembly, if you make the outer length of the oval a little bit smaller than the outer length of the o-ring it will not fall when you push it into the grove. Regarding fill percentage, it should always be smaller than 100% to allow the o-sing to deflect sideways as a result of the squeeze. O-rings designed to seal only on three surfaces/sides. In your case: the flange, the bottom of the groove and the outer side of the groove. The inner side of the groove should not touch the squeezed o-ring to allow a free passage of the gas to apply pressure on the inner side of the o-ring thereby, to press it simultaneously to the flange, the bottom of the groove and the outer side of the oval.

Regarding the squeeze, for small pressure as in your case, there is no need to go to more than 10-15%. 25% and above are intended for really high pressure systems. However, in such high squeeze rates the o-ring needs to be replaced if it was disassembled, because such high squeeze may create too much permanent set of the o-ring thickness. See o-ring manufacturer data regarding the minimum squeeze, compressions set, etc, with respect to pressure, o-ring type of material and hardness.

 

[israelkk]

Sorry, 20000 psi is 1360 atm.

 

[c0rd0bes]

for rayj2

"Height of the groove approx 2/3rds of the O ring O.D."
--- I assume you mean 2/3 of the o-ring cross section diameter?

yes, Cross section diameter to use the correct nomenclature in this case.


Israelkk - great points thanks for correcting the obvious difference.

 

[tbuelna]

rayj2,

Your working pressure (10 bar) is relatively low so the amount of o-ring compression (ie. groove depth) is not critical. With low pressures, I prefer to reduce the amount of o-ring compression so that it requires less clamping force and also so that the o-ring elastomer is less likely to take a compression set.

With an oval shaped face seal like yours, I would also size the o-ring so that it fits snugly around the inner wall of the groove. This makes things easier during assembly because the o-ring won't fall out of the groove when turned upside down. If the o-ring sags out of the groove during assembly, it is easy to pinch or cut the o-ring when the cover is clamped tight.

Hope that helps.
Terry

 

[RayJ2]

Thanks everyone, for your contributions!
The compression % is definitely something I need to give some attention to. Especially because the seal is often opened and closed. If compression is too high, there might be some compression set which is not good if you need to open and close all the time.

Kind regards
Ray

 

[MikeHalloran]

Given the duty cycle, you might want to bias the seal toward the outer lip to get some circumferential compression, and give that lip a reverse taper to help retain the seal.



Mike Halloran
Pembroke Pines, FL, USA

 

[hydtools]

A little grease will also help retain the o-ring.

Ted

 

[RayJ2]

@Mike,
I milled a groove in a little test part where the o-ring was compressed against the outer circumference. It seemed to stay fairly well in place. Tapering the groove wall will ofcourse help a lot. But it will also increase production cost. The actual setup will have 80 of these o-ring seals.

@hydtools,
Yes, adding some grease will help. But in this particular application, the seals will undergo temperature cycles from -50°C to +150°C (-58°F to +302°F). So probably the grease will quickly dry out. Perhaps a few drops of cyanoacrylate glue ('superglue') might also help.