O-Ring Between Flat Face Flanges? 3

[Slagathor]

I have a manufacturer telling my client to use a 7mm o-ring to seal between two 16" 125# cast flanges. There is no o-ring groove. The manufacturers sales rep is saying to just put the O ring between the flange faces, even though there is nothing to locate it. This seems absurd to me. Has anyone ever heard of this being proper piping practice?

The joint in question is between a pump suction nozzle, and a special pump suction elbow, both supplied by the mgf. I am concerned about cracking a flange in the short term, and the o-ring failing longer term as it creeps out. I am tempted to over rule the recommendation, and use a proper gasket (cloth inserted rubber, 1/16th thick). This is at a wastewater plant, so no ASME standards...all the standards are AWWA. Pressure is in the range of 20 psig, digester sludge service.

 

[C2it]

An O-ring seems the wrong solution, even at low pressure it is likely to blow out.

I would use a full-face soft gasket.

 

[LittleInch]

The phrase "manufacturers sales rep " is probably your problem.....

I've never heard of it and it is neither common nor good practice to use a o ring in such a situation.

Most flat face cast flanges require a full face gasket, normally made of neoprene (wet suit material) or similar rubber or other resilient material (EPDM) 3-4mm thick.

If you look hard enough, I'm sure AWWA has advise or a requirement for flat face flanges.

I didn't look very hard and found this - see 2nd page under gaskets

http://www.pacinst.org/wp-content/uploads/sites/2/Features-Flanged-DIP-and-Fittings.pdf

Refers you to the manufacturer of the flanges, not the equipment. The o ring idea still sounds very poor advice to me though.....

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[moltenmetal]

Dumb idea. Sheet rubber full face gaskets are the right way to do this.

 

[1gibson]

See how long you can get the sales rep to defend the o-ring idea, should be entertaining.

 

[Duwe6]

"See how long you can get the sales rep to defend the o-ring idea. . ." in writing. Wish I could be there to watch.

 

[btrueblood]

We use o-rings to seal on flange faces all the time in our shop testing. BUT...our shop tooling has o-ring grooves milled into the flange face. Without a groove, how would you even expect to keep the ring centered during makeup of the flange?

 

[1gibson]

Chewing gum, 4 locations equally spaced

 

[LittleInch]

Btrublood - Are they "cast #125 flanges"? Normally these flanges don't really like any bending forces imposed by the gasket or a raised face, hence why they are full face. The RF/FF issue has been discussed many times and from memory, most people recommend that FF flanges are mated with FF flanges and a full face gasket.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[btrueblood]

LI, re-read my post. We have an o-ring groove machined into the flat face (or raised face) of the test flange, this o-ring mates statically to the flat or raised faced flange of the device under test. I.e. the flanges end up mating face to face, with only a few 100'ths of an inch of o-ring compression to effect the seal. Much easier to get the raised face to lie flush during makeup as well. Much more reliable sealing for hydrotesting than the typical flat gasket, and you don't have to chisel/sand the gasket remnants off the device under test afterwards. The drawback is, you have to make the custom cut flange o-ring groove yourself.

 

[moltenmetal]

btrueblood's solution is a totally different matter. We've machined o-ring grooves into flanges for more than just testing and it works perfectly. But an o-ring just smushed between two flat faced cast flanged is a stupid replacement for a proper gasket.

 

[EnergyMix]

However btrueblood's solution sounds like a modification to the flange that you don't want to do yourself, whatever the sales rep reccomends. Have them come out and do it and supply the o-ring -- that way you're still under warranty.

Want to know the do's and don'ts of Eng-Tips? Read FAQ731-376.
English not your native language? Looking for some help in getting your question across to others or understanding their answers? Go to forum1529.

 

[LittleInch]

btrueblood. Apologies if it came across wrong, but the point was if you do this on cast iron flanges. If you have a raised face it isn't cast iron, but steel of some sort. I can appreciate you try to leave only a small gap, but even during tightening, you will then get a bending moment around the sealing point.

I think we're all agreed though that unless you put in a groove to hold the O ring, this idea is just nuts.

just for curiosity, how accurate do you need to machine the groove to get your clearance right. I presume you want only a very small gap to squash the o ring just enough and not leave it open to shear forces. what do the flange specs / codes say about O ring grooves?

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[rconner]

I see some interesting comments here but have to admit to me at least this whole thing sounds quite strange. Why is exactly this "sales rep" making this recommendation (is there some sort of problem with something else, or is this just a new green field installation, or what?)
While o-rings and o-ring like bulbs have long been used, with appropriate housing or support, and I think generally VERY good (and in many cases superior to at least some flat gasket) success in many sorts of gasketing means used in many sorts of "flanged joints", and going way back I think to at least to the early half of the 1900's, I'm just trying to understand why anyone would want to just throw a floppy o-ring itself in between a couple flat flanges (as I guess all other repsonders are assuming is meant by the OP)?

 

[btrueblood]

LI, I don't believe I can torque a set of flange bolts enough on raised face steel flanges to crack them, nor for typical ductile iron. But, I have not tried, either, and just torque the (typ. Gr. 5) fasteners to approx. 80% of their rated proof load. I could see a flange bending, maybe, but not cracking. We do this "all the time" for hydrotesting 300 lb. raised face flanges, and have never bent anything. We have seen bolt/stud stretching if we take the pressure all the way to "burst" (usually the o-ring extrudes when this happens, and it takes something over 1200 psi to do so - but the flanges don't bend. Is there something magical about a flat gasket that removes the possibility of bending a raised face flange? That rubber under the bolt area just squishes out, it doesn't offer any support. Pivoting around the seal point? Seriously? Rubber is about 3-4 orders of magnitude softer than steel, it just squashes until the flanges are in contact.

We machine the grooves per the Parker O-ring handbook, which means typical compression of about .025" on a .12" o-ring section.

As far as cast iron flat faced flanges, I'm not even sure where I could purchase a cast grey iron flange. Maybe I could find a cast ductile or nodular iron flange...but most likely what's in stock at the piping warehouse is plain old steel, and it's what we use.

I'm sure cutting an o-ring groove in a flange violates the piping code somewhere, for some condition. It certainly adds a stress concentration, though it's in compression if the flange is torqued down correctly. For factory test equipment at pressures well below the flange's capacity at room temp., it's just not a concern. We are not putting grooves in the stuff that goes out the door, only on the blind flanges and test connections (factory equipment).

Before we did this, we would routinely have to stop a hydrotest and repair leaks/replace gaskets. My opinion, is that O-rings are much more idiot proof in their ability to seal reliably.

 

[tbuelna]

I believe the step in the face of a 125# pipe flange is only around .06". So that would require the 7mm o-ring to be compressed around .155", which is around 50% of its free height. This much compression would likely result in the o-ring material taking a compression set and losing its sealing ability.

Installing an o-ring without any surface to back it up relative to the direction fluid pressure is applied is poor practice. Since the only thing keeping the o-ring in place is the friction provided by the opposing faces compressing it.

 

[MikeHalloran]

On a pump suction, an o-ring would be sucked in, not blown out.

It appears that AWWA 'hub flanges' would have room between the OD of the hub and the ID locus of the bolt ring for an o-ring to be fairly well trapped for suction service, but it would require a very fat o-ring, not 7mm cs. ... or non-standard flange dimensions, which is entirely possible between two parts from the same manufacturer, one of which is even described as 'special'.

We here have not seen photos or drawings of the affected parts, and it smells a little like Slagathor has not personally observed the details, either.

IMHO, it would be a good idea to dig deeper into the joint design before maligning the sales rep. Yeah, many of them are slimy BS artists, but every once in a while you run into one who really really knows his stuff.




Mike Halloran
Pembroke Pines, FL, USA

 

[C2it]

Mike,

That would be true if you ignore the pump needing an NPSH.

 

[LittleInch]

Mike H - Just because the inlet to the pump is commonly called the "suction" flange doesn't mean that the pressure is always less than atmospheric....

also pivoting - if you have low grade cast iron yes - it may be squashy, but force is located at a small point creating a high bending moment. full face gaskets spread the reaction load provided by the bolts over the whole face of the gasket. It is one of the problems of FF gaskets as the sealing force is generally low and hence leaks are difficult to eliminate just by tightening the bolts a bit. Ductile iron has got much better over time, but you still get some low grade cast flanges in this type of application.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[moltenmetal]

The flat face flange problem is one for gray cast iron flanges only, but there are plenty of components still made from brittle cast iron rather than ductile/malleable cast iron. Similarly there are many components (valves, cast flanged fittings, i.e. the entire Crane Resistoflex catalog) made from MI which have raised face flanges, and that's fine- a good grade of MI has an elongation on failure similar to that of ordinary carbon steel.

It's more a problem when you use a stiff gasket covering less than the full flange face than it is with a more compressible rubber gasket, for obvious reasons. A rubber or composition gasket covering the full face of the flange is the right solution when mating to another flat faced flange. If the mating flange is raised face, you need either a limiter ring of some stiff material to limit the bending when someone gets too aggressive with the bolt torque, or you need to change one of the flanges.