Torque specs. for ductile iron backup flanges

[numberfive]

Experts,

We are currently installing a stainless steel cooling water system for a condenser consisting of 8" thru 20" sch. 10 pipe. All flanged joints are ductile iron backup flanges and gaskets are a compressed aramid fiber with nitrile binder.

At initial pressurization of the system (50psi.) several of the flanged joints leaked.

We have been asked to torque all of the flanged joints to the gasket manufacturers torque specs for ASME B16.5 150# class flanges.

I'm concerned that we may be over torquing these ductile iron backup flanges (especially the larger ones) due to the flexing that I'm seeing in them.

For example, the torque spec. for a 20" flange is 415 ft./lb.

I've never used a torque spec when installing ductile iron backup flanges before but my experience tells me that we may be over tightening these flanges.

Any input would be greatly appreciated.

numberfive

 

[gerhardl]

First the obvious : The purpose of a gasket is to prevent leakage. This is obtained by a proper and even pressure on the gasket. The form and material of gasket and flange must be suitable for this.

Secondly the cause of leakage: Missing or not satisfactory support or false unlinear mounting of the pipeline might give skew pressure on sealings. It is fairly common that leakages might occur at pressure tests. The normal procedure in 'good old days' was to visually inspect and hand-tighten bolts to 'firm' and then release a tiny bit, and test or check leakage once more. This will not necessarily give equal torqe for all bolts.

Inspection: Check also that none of the flanges are deformed or deflected. This is a clear indication of skew forces (false mounted pipeline). If the pipieline is correctly lined and supported, and the flanges still are deformed, or if sealings still are leaking: either the flangeform or sealing type (or combination) is not suitable for the purpose.

Note: The tightening force for the flangebolts sshould be far from any max bolt-torque load given. Anything else is again indication of something wrong.

 

[LittleInch]

I can't quite figure out why you're using ductile iron backing rings for stainless pipe, but if that's what you've got then it's what you have to work with.

you are totally correct to be concerned. There is no way bending or flexing of rings should be allowed and It would be good to clearly see or understand what the gasket manufacturer thinks he is quoting for - my guess is a standard weld neck #150 RF flange. There is something seriously wrong here.

what sort of flange faces are being used here? A dimensioned sketch of the joint or photo would be very useful.

Bolt torque is notoriously difficult to use to obtain bolt stress due to varying friction, whether lubricated or not and once the backing ring starts to flex then you cannot tighten it any more as the increase in bolt stress is very low.

My motto: Learn something new every day

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

 

[rconner]

fyi

http://www.eng-tips.com/viewthread.cfm?qid=146529

 

[rconner]

Also, it sounds like the OP is talking about some sort of "lap joint​" flange arrangement. If so, the stub ends of particularly very thin pipes like you are talking about are obviously not very "stiff", from the bending e.g. Bourdon thrust/bellows-type deformation standpoint. I believe some references say the pressure capabilities of such joints are less that those with welded flanges(while it may be hard to find quantified specifically, at least in standards?) , stress intensification factors are higher, and perhaps of more long term concern in many applications particularly with temperature and pressure etc. fluctuations is that references report the "fatigue strength" of such joints is only one tenth of others! {LI, references also indicate ductile backing flanges are used in stainless steel piping systems primarily for economy, and I don't doubt that is true when ductile castings are compared to a forging or flanges cookie(or donut?)-cut from steel plate, let alone stainless the latter process resulting inevitably in some drop waste.]
All this being said, 50 psi isn't much pressure and a lot of things can make flange joints leak. You have reportedly asked the "gasket manufacturer", and I think that is a good move. Another entity to ask may be the specific designer or design entity who married this particular pipe, these particular gaskets, as well as whatever specific type of bolting that has been specified for this application all together (i.e. who chose/specified -- while I'm not real familiar with the specific gaskets you describe; however, if they are rather hard compared e.g. compared to traditional rubber gaskets it is understandable that it might take more healthy torque to get them to flow into spaces in joint members or alignments that need sealing, as well as seal the elastomeric matrix around any fibers etc, whether the rest of the joint can take that and subsequent loadings or not!) All have a good weekend.

 

[chicopee]

It would be a good idea to provide us some pictures of those sections where the joints are leaking. Why do I get the feeling that schedule 10 piping may be a bit flexible and thereby cause leaks?

 

[Duwe6]

Torqueing Flanges:

If you do not see numbers written in 'crayon' on the flange, assume that no torque pattern was used. That can be the cause of most of those leaks. And I have no idea where the 420 ft-lbs comes from on a nitrile gasket. Try using half that, or less, applied PROPERLY.

Number out the torque pattern on the flange. Align the flanges, and drop the gasket in. Finger-tighten all the bolts. In the pattern, snug all the bolts with a crescent wrench or channel-locks, NOT the big wrench. In the pattern, apply 1/3 of the torque. Check the flange all around for evenness and alignment. In the pattern, apply 2/3 torque, and go around twice. In the pattern, apply full torque, and go around 3 times. Then make a 'check pass' in a circle, and keep going until no nut moves before the wrench clicks.

Takes a lot of time to get it right. But how much time will you spend fixing those leaks?