Help with O-Ring Groove Design (Static Radial) 1

[spazzyfry123]

Hello all.

I'm finishing up a design that will be utilizing an O-ring for sealing the OD of a spacer to the ID of a bearing support housing. This will be the final seal point for oil within the gearbox along with rotary seals at the ID of the spacer to seal to the spinning shaft. The last bit that I am unsure of is relative to the O-ring groove (gland) dimensions...specifically the diametrical clearance.

Unfortunately this is a retro-fit type job and with the space constraints within the design area, I am left with needing to have a press fit between the spacer and the bearing housing to prevent the spacer from potentially backing out of the housing.

With this known, is this diametrical clearance absolutely necessary with very low pressure? I would like to have a press fit across the face of the OD of the spacer. The clearances given within the Machinery's Handbook are valid up to 1500 psi. My specific application will see ~7 psi at the most as I have relief valves on the gearbox.

However, if this diametric clearance is necessary, I would assume it will be needed from the direction of oil flow (Shown marked in red) until some distance past the groove width for the O-ring. My question: How far past the groove width is this required? (Shown marked in blue)

I have attached a few snap-shots to help get a visual across.
The blueish, purple is representative of the spacer mentioned above.
The black is the O-ring - AS568A-438 A70 Durometer rating.
The component that the OD of the spacer is mating to is the bearing housing.

Thanks for your help!
-T

 

[MikeHalloran]

The clearance is of no consequence to an o-ring, except that if the clearance is large enough, the o-ring can extrude into it, which damages the o-ring enough to leak on subsequent pressure cycles even if the o-ring doesn't escape entirely.

In your case, it appears that the o-ring will be assembled axially, 'under' a chamfer that is short and small, and it will probably be torn on assembly, unless an assembly tool is used, in which case the chamfer can be trivial.

The one dimension that is important is the radial squeeze.

Download the Parker o-ring handbook, and do not deviate from their recommended dimensions.


Mike Halloran
Pembroke Pines, FL, USA

 

[ornerynorsk]

I'll strongly second Mikes suggestion.

Here ya go -

http://www.parker.com/literature/ORD 5700 Parker_O-Ring_Handbook.pdf

It is better to have enough ideas for some of them to be wrong, than to be always right by having no ideas at all.

 

[spazzyfry123]

Thanks for your responses. If you can't tell, this is a first-time learning experience with O-rings. The designs generally require seals; this specific design required custom-sized seals to meet the larger bearing size and ended up leaking. As a result, this design came into motion to use an off-the-shelf seal with minimal surface area. Here comes the "spacer."

I took a look at the dimensions given within the Parker handbook (sec. 4-9), and they match those that I've utilized out of the Machinery's Handbook. The amount that the O-ring overlaps into the bearing housing is typical. The reason such a large O-ring was chosen is (cross diameter = Ø0.275"), designed accordingly to the standards given, this is the only one that will match to the metric diameter (Ø170mm) of the bearing housing while not having a smaller groove diameter (6.2349...") for the O-ring than that of the O-ring's ID (6.225"). If the O-ring is still able to correctly seal with a slight gap between the groove diameter and the ID of the O-ring, I can step down in size (cross diameter = Ø0.139") to bring down the amount of interference between the O-ring and the bearing housing. I don't have those numbers here in front of me, but I believe the diametrical difference between the two was ~.02"

The chamfer that the O-ring will need to pass over is already incorporated to an existing piece and cannot be changed (1x30°).

You mention that the clearance is of no consequence to the O-ring. But is it beneficial (read: required) to the ability of the O-ring to seal properly? I am seeing some cut sheets that show the O-ring extruding (minimally of course) into the clearance. Surely the O-ring interference between the bearing housing and the spacer is not enough to keep this spacer in place...Hence my want for a press fit. But if at all possible, I would like the entire axial length of the spacer to be a press fit (aside from the grove diameter portion) as I have no other way to axially "fasten" the spacer to the bearing housing.

Sorry...I feel as the more I type this out the more confusing I am making this to understand. Let me know if there are any questions I can answer.

Thanks again for your help!
-T

 

[MikeHalloran]

From the o-ring's perspective, the clearance gap at the ends of the gland is ideally kept to zero, so a press fit of the spacer into the other part is not a problem.

Getting the o-ring past that 30 degree half-angle chamfer is a real problem. As a point of reference, seven degrees of half-angle makes it easy to assemble an o-ring axially. There is no functional difference between 30 degrees and 45 degrees; both require assistance and lubrication and still have a fair probability of cutting the o-ring.

You may have to make up a (possibly segmented) squeezer with a gently tapered bore, like a piston ring compressor tool, to guide the o-ring into place while you are pressing in the spacer.

That's the only downside of the press fit; you won't be able to feel the o-ring being cut during assembly, so you definitely need a squeezer tool of some kind. I have seen it done with rope in the field, but in the shop, a nicely machined part is a better idea.



Mike Halloran
Pembroke Pines, FL, USA

 

[Timelord]

Use the Parker Handbook!! Years and years of research went into it and if you follow their recommendations, it won't leak.

Timelord

 

[JJPellin]

I would recommend the smaller cross-section o-ring. You asked if it was acceptable to have the inside diameter of the o-ring larger than the outside diameter of the bottom of the groove. That would mean that the o-ring would be a sloppy fit within the groove and much more likely to cut on installation. It is better if the o-ring has to be slightly stretched and pulls tightly into the bottom of the groove. If a standard o-ring diameter does not work, you can splice the o-ring to a smaller diameter.

If you can shrink the spacer ring (dry ice) so that it drops right in, you can still “feel” the o-ring slip up into the bore to reduce the chance of cutting it.

Johnny Pellin

 

[tbuelna]

For your 170mm bore o-ring that is designed to provide an oil seal with minimal pressure delta, a 1/8" (.139") section o-ring should work well with a groove that provides around .020" radial compression. There is no benefit to using more compression of the o-ring than is necessary. Excessive compression of the o-ring requires greater effort to install the o-ring, and will result in greater possibility of the o-ring to take a compression set.

As others noted, you should make sure there are generous chamfers provided at the edges the o-ring must pass over during installation to prevent damage to the o-ring.

 

[spazzyfry123]

Thanks again for everyone's help. Sorry for the delayed response; I've been running all over with different projects.

I've changed the design to utilize a Ø0.139" cross-section O-ring (AS568A-260 specifically). I must say that I am confused how the OD of the groove does not meet/intersect with the ID of the O-ring, however. Even with removing the radial clearance dimension from the overall size of the gland depth to obtain a press fit across the spacer (thus creating a slightly larger OD for the groove), I am still left with a ~0.0055" radial clearance with all values at their maximum within the tolerance ranges. I would much rather use the design parameters that are given within the O-ring handbooks, but doing so leaves this small radial clearance between the groove and O-ring. Should the groove OD be slightly enlarged to create an intersection between the O-ring and the groove to have it "stretched"? If so, what guidelines do I have to ensure the correct amount of overlap is used? This is the sole reason I chose a Ø0.275" O-ring as it was the only one that fit the groove correctly. Something just doesn't seem right...

MikeHalloran,

Can you provide an example of the squeezer you are referring to? I've dealt with the piston ring compressors in the past when finishing up an old small block Chevy rebuild in my garage some odd years back, but that is my only experience with something like that. This is honestly the first time I have dealt with O-rings.

Thanks!

 

[ornerynorsk]

A squeezer is nothing more than a thimble or mandrel used to aid in assembling an o-ring. If you make your glands to the dimensions in the handbook, you will not need one! Assemble with appropriate lube on the o-ring, and using a little finesse and care, your assembly should go together smoothly without biting a chunk out of the ring. I did a short stint as a pump assembler 10 to 15 years ago for medical hydraulics, and out of approximately 75,000 O-rings and seals installed, I can count on one hand how many I damaged during assembly, and those instances were my own fault for hurrying.

It is better to have enough ideas for some of them to be wrong, than to be always right by having no ideas at all.

 

[spazzyfry123]

As I understand it from reading other replies in here, the assembly tool is needed as a result of the 30° chamfer on the housing that it is being pressed into.

 

[btrueblood]

Actually, it's the small size of the 30 deg. chamfer (smaller in its outside dimension than the outside dim. of the ring) that will potentially cut the oring. Either make the chamfer larger, or make the oring smaller in section, so that the chamfer outer edge is larger in diameter than the o-ring. T the op, yes, you should design the ring to have about 10-20% of its section compressed into the groove, and the groove should cause the o-ring to stretch by somewhere between 1-10%. Having the o-ring i.d. larger than the groove floor (resulting in a radial squish to seat the ring) will result in the ring buckling/squirming, increasing the likelihood of cutting on installation, or wrinkling the surface and causing loss of seal.

 

[tbuelna]

As the gland diameter of standard o-ring sizes get larger, the increase in diameter from one dash number to the next gets bigger. What you should do in this case is select the next smaller dash number o-ring so that it gets slightly stretched when installed. The typical limit for installed o-ring stretch is about 10%-15%. The groove dimensions can also be adjusted to provide a groove volume that is at least 1.25 times the o-ring volume.

 

[spazzyfry123]

Fair enough. We'll shoot to alter the housing for the larger diameter of the chamfer to be Ø172mm (45° x 1 chamfer as opposed to the original 30° x 1) to clear the OD of the O-ring.

AS568A-260:
Cross section = Ø0.139"
ID = 6.484"
OD = 6.762"

I've also changed the gland diameter to be 6.5118" to allow 10% stretch into the ID of the O-ring's cross section diameter. 6.5118 - 6.484 = 0.0278 | 0.0278/2 = 0.0139 = 0.139 * 0.1

This should be all that's needed, correct?

Thanks!

 

[JJPellin]

You may have misunderstood the recommendations above. We were not recommending increasing the outside diameter at the bottom of the groove to achieve a target pre-stretch of the o-ring. We were suggesting dropping down one dash number to the next size smaller o-ring. By increasing the OD at the bottom of the groove as you have described, you will be increasing the crush on the o-ring. The standard depth of this o-ring groove would be 0.123” to 0.125”. Your dimensions would correspond to a final groove depth of 0.091”. That is too much crush on a 0.139” o-ring. You would have great difficulty installing it without cutting it, even with the larger chamfer.

I would suggest dropping down from the dash-260 o-ring to a dash-259. If that it too tight, then I would custom cut the o-ring to fit your groove.

Johnny Pellin

 

[tbuelna]

You should remember that the standard radial gland dimensions listed in o-ring catalogs are designed for something like 1500psi pressures. Since the pressure delta in the example being discussed is less than 1% of that it would be prudent to adjust the gland dimension to reduce the radial compression of the o-ring. This will make the o-ring less susceptible to compression set and it will make the o-ring much easier to install.

It is also very important to pay attention to providing adequate chamfers at the edges the o-ring must pass over during installation. The diameter of the chamfer should always be sufficient to clear the uncompressed o-ring surface, and the chamfer should preferably have as shallow an angle as practical, ideally no greater than 30deg/side. The reason for paying close attention to housing/shaft chamfers for o-ring installations is because the o-rings are installed blind. Without adequate chamfers it is very easy to cut an o-ring at installation, and you would not likely know the o-ring was cut until it began to leak in service.

 

[MikeHalloran]

Yeah, but don't go crazy relaxing the compression on the o-ring because of the low pressure. 1500 psi is easy. Below 100 psi it gets harder, because the ring won't deform enough to seal well at low pressures.

Also, a 30 deg half-angle chamfer makes for difficult assembly, even greased. Its only virtue is that a 45 deg half-angle is worse by far. I like to see a max 15 deg half-angle, and I will go to 7 deg half-angle if I can find the axial space. The difference in ease of assembly (and in likelihood of cutting the o-ring) is huge.

It may be helpful to make up a spreadsheet to calculate every possible dimension associated with the seal and the gland, with multiple columns so you can evaluate alternatives against each other, and look at the range of tolerances too. Do not forget to to include the seal volume vs. the gland volume, and make sure you always have some margin there; rubber is not easily compressible.



Mike Halloran
Pembroke Pines, FL, USA

 

[spazzyfry123]

I hate to sound ignorant and almost like throwing darts and hope it sticks, but I'm really in the dark with these O-rings. Thanks to everyone for your help; this has been a great learning experience.

I'm afraid I'm a little confused as to why to drop down to a -259 O-ring that has an OD of 6.512" when the mating surface is Ø170mm (6.693"). Am I relying on that great deal of stretch?

The O-ring only needs pass over one edge: the chamfer in question. The remaining surface will be Ø170 H7 (3.2 surface finish) in which the spacer will be inserted (Ø170p6). I think I'll be able to get away with a 15° chamfer. This will nearly quadruple the axial depth of the chamfer but should leave ample surface for the spacer to mate to.

I am really trying to keep all components "off-the-shelf." The whole reason for the spacer is to avoid custom sized seals. I would like to keep it the same for the O-ring as well. Why couldn't we just be metric like the rest of the world....... Anyway, where am I left with the groove diameter (with a -260 unless I'm still misunderstood utilizing the -259?). I have to agree with MikeHalloran given the low pressure being built in the application. A "standard" gland depth certainly leaves a gap and a "modified" gland depth creates too much compression for the O-ring.

With the latest gland dimensions I have given with modified gland depth dimensions and using 0.005" radii at the "top" of the gland and 0.025" radii at the bottom...

Volume:
Gland - 5824.08mm³
O-ring - 5173.99mm³

 

[dgallup]

Dropping the o-ring down one dash size given the dimensions you cited will give you a stretch of about 2% which is ideal. I disagree with those above that said stretch can be as high as 15%. I have never seen a handbook recommend that much. I try to keep the stretch between 1% and 5%. I also would not reduce the squeeze. Low pressure sealing is the most difficult case as there is not enough pressure to "self energize (deform)" the o-ring. Installation considerations are paramount to avoid cut o-rings. You either must have an assembly tool that compresses the o-ring or provide a nice lead in. In addition to the chamfer, a nice radius where it intersects the bore can help a lot. I have had very good results with 20 degree chamfer and a 1 mm radius in high volume production assemblies (over 1 million a year).

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[spazzyfry123]

So it is safe to have...

A groove diameter of Ø6.471" (using standard gland depth of 0.111" off Ø170mm bore)
A O-ring ID of 6.234" (AS568A-259)

And this will mate to a Ø170 bore?