Removal torque

[zeusfaber]

I suspect too much torque has been used to assemble some pipe couplings.

If you know how much torque was required to undo a threaded connection, is there a rule of thumb you can use to get some idea of how much torque was used to do it up in the first place?

A.

 

[CoryPad]

To assemble joints, there are three torque contributions - overcoming thread friction, overcoming bearing friction, and generating preload. To disassemble joints, you don't have the preload component, so it gets difficult to compare "on" and "off" torques.

In industry, it has been found that the best procedure is to verify the torque in the "on" direction. A typical procedure is to take a previously tightened, at-rest fastener and advance it in the tightening direction less than 5 degrees and record the maximum torque. This should compare fairly well (+/- 20 %) to the original tightening torque. This assumes there aren't large changes in joint behavior (embedment, corrosion, etc.).

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[byrdj]

another method would be mark both sides of the assembled joint, measure the torque to break it free. then tighten joint back to match marks measuring torque required.

 

[unclesyd]

What type couplings are you worried about?

 

[zeusfaber]

Don't want to go into too much detail about the precise type of coupling at the moment - may become less sensitive later.

Usual problem is that this is diagnosis from the far end of a not very good telephone line, so I'm left (again) with trying to interpret the results of someone else's test.

If I interpret Cory's first paragraph correctly, it would be a bit suspicious to find an "undo" torque nearly three times the specified assembly torque (on a reasonably clean component).

A.

 

[IRstuff]

But, that may or may not be due to an initial torquing. After all, a rusted bolt's untorquing is due to the rust, and not the original installation.

Was there any pipe compound applied initially?

TTFN

FAQ731-376

 

[zeusfaber]

Almost certainly no compound. This is non-ferrous, carrying dry gas. No detectable corrosion.

A.

 

[IRstuff]

Are these like standard 316SS with Swageloks?

TTFN

FAQ731-376

 

[CoryPad]

Minus any deformation/interlocking/corrosion, there is no reason why an off torque would be 3 times the on torque. Assuming, of course, you are above a certain threshold, as 3 N m is 3 times larger than 1 N m, but neither of these torques are suitable for an M10 thread.

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[zeusfaber]

Thanks to everyone for advice - been very helpful.

A.

 

[MrMyers]

Just a thought - can you measure the bolt tension ultrasonically?

*awaits criticism*

 

[zeusfaber]

Never come across ultrasonic tension measurement before - quick google has just added to my education.

Too late now to do much about it (everything's in bits, even the stuff that didn't disassemble itself) - but thanks for teaching me something new.

A.

 

[KENAT]

I had a similar problem some time back. We knew the theoretical fastening torgue but the undo torque was much (X 10) higher and we were trying to find out why. We knew the real fastening torque couldn't be much higher than the theoretical because using the same tooling it broke when we tried to undo one at only slightly higher torque!

A very experienced guy told me at the time that very rough rule of thumb was undo torque = 1.5 X fastening torque.

Obviously we had something else going on.

KENAT, probably the least qualified checker you'll ever meet...