Welding detail for pipe anchors

[KevinNZ]

Point for discussion on how best to weld a pipe anchor/trunnion to pipe. As below this could be
1) a set on weld with reinforcing ring on the outside (much like a reinforced tee)
2) full pad welded to the pipe and the anchor welded to the pad.

Fabricators like 2) much easier to make.

As a designer I prefer 1). I can work out an SIF for the stress model and know the pad is not pulling away from the pipe under load. Is there a way to work an SIF for 2)?

 

[1503-44]

US gas codes do not allow welding directly to pipe in any sense on the word. A full encirclement clamp must be used. That's sometimes inconsistent with full anchor forces, in which case we would weld to an anchor flange fitting. Just saying.

I'd use #2. Stronger, due to lack of a hole and far easier to fab.

Full encirclement would pretty much double the pipe stiffness and better distribute the eccentric bending moment the anchor force is likely to introduce into the pipe.

 

[LittleInch]

I agree.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[FacEngrPE]

The detail has a risk of stress from an overly constrained connection.
I once investigated a elbow anchor in a steam pipe that failed, pipe welded to an elbow. FEA determined that the thermal stress between the hot pipe, and the cold pipe anchor was high enough in the overly constrained weld joint that stress exceeded yield. The failure looked like a fatigue failure in the weld heat affected zone.

When attaching a pipe anchor in the manner shown, if there is signification temperature difference, you must consider the thermal stress from the temperature differences.

If your application is at ambient temperature, either approach should work.

 

[1503-44]

KevinNZ didn't specifically say, but I'm sure that is for a fully restrained anchor for relatively warm pipe. He knows the difference and wouldn't be thinking about that detail, if it was for anything less.

 

[FacEngrPE]

Our solution was to still to use the elbow, and replace the pipe between the elbow and the base plate with flatbar, The flatbar winds up with an X cross section. At the center of the X, where it joined the elbow, we added a notch out so the weld in one direction could be made to run through without interruption.

We did not subject the replacement elbow to FEA. In he early 1980's our access at work to that tool was rather limited.

 

[MJCronin]

https://whatispiping.com/supporting-of-piping-systems/

Trunnion anchors are tried and true designs and have been in use in piping systems for a hundred years. Anchors are a bad idea for hot steam systems and should be avoided if possible but NEVER be placed on elbows on steam systems.

https://www.processpiper.com/wp-content/uploads/2017/04/PSS-BS01-2-to-24-BASE-SUPPORT-.pdf

https://whatispiping.com/trunnion-checking/

Yes, steam and other piping systems that are operating above say - 750F should be supported WITHOUT INTEGRALLY WELDED ATTACHMENTS

Where they are necessary on hot piping, all integrally welded pipe attachments (including anchors, trunnions, lugs, batwings etc) should be insulated with the piping to prevent thermally induced differential stresses.

Well, 1503-44 states: "US gas codes do not allow welding directly to pipe in any sense on the word" ..... Perhaps,.... but much of the world is governed by ASME B31.1 and B31.3 and not "the gas codes". Integrally welded attachments are allowed by both of these Piping Codes as well as ASME Section III - Subsection NF (Nuclear Power Piping)



MJCronin
Sr. Process Engineer

 

[MJCronin]

Other rules apply for welded attachments to piping systems.

This topic has been discussed in the past on this forum

https://www.eng-tips.com/viewthread.cfm?qid=477802

MJCronin
Sr. Process Engineer

 

[1503-44]

I mentioned the US gas code because I know Kevin works with gas piping a lot and he is not in the US.
What's it to you?

Much of the world does not and much of the world does use B31.4 and 8, or is so close you would not know the difference. Not that it matters.

 

[KevinNZ]

Hi all,
I do not work with gas piping. Usually large long run hot steam/water piping to about 200C. For last 60 years all pipe shoes have been welded to the pipe. Biggest design loads are seismic and need extra strong anchor connections.

Thanks for the heads up on the differential temperatures between the pipe and the anchors. At least detail 2) would move the high stress weld to the non pressured plate. May need to find a better tool to work out the SIf for the non integral pad.

 

[1503-44]

For some reason you had me convinced you were in gas. Some photos a long time ago of an arched (gas?) pipeline crossing a river I think it was. Anyway, no matter. I wasn't saying that you should follow US gas codes by any means. Of the choices you presented, I voted for #2. I assumed that both would be acceptable to whatever code you might be using.

So BTW, what code are you using, just for satisfying curiosity purposes?

 

[LittleInch]

I'm assuming here that this is a one sided anchor / trunnion.

I'm for option 2 as well as I think this should distribute all those out of plane bending forces more evenly and allows the main pipe structural weld to be a good full penetration weld into the pad and not compromise the internal pipe to the same extent as all those fillet welds on the larger pad. That pad in my mind is a full encirclement split collar and I can't see it being ripped off anything.

Ultimately you still have a bending moment resulting in stress concentrations somewhere.

I struggle to see how the reinforcing ring really helps as the forces are all on the connecting pipe to pipe weld? That weld detail in the pipe to pipe to repad joint looks a bit dodgy as well. Lots of heat there is nothing else.

I'm sure some FEA jockey can tell us which is better....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[KevinNZ]

1503-44
The line over the river is hot water. (Geothermal reinjection line)
Our work is to B31.1.

 

[1503-44]

Not two minutes ago I was thinking maybe it was a geothermal steam line in the picture.

OK. Great. Weld away.
Thanks

 

[MJCronin]

Getting back to the original question, there is some worthwhile information in this experience based document.

https://s3.amazonaws.com/suncam/docs/167.pdf

Gerald May states positively:

GUIDELINE 10:

WELDED PIPE SUPPORT ATTACHMENTS SHOULD BE AVOIDED DUE TO POTENTIAL FOR FAILED WELDS.

 FOR RISER SUPPORTS USE SHEAR LUGS WITH ARISER CLAMP TO AVOID PUTTING THE ATTACHMENT
INTO BENDING AT THE PIPE WALL.
 ON HORIZONTAL PIPES USE CLAMPS INSTEAD OF WELDED ON PLATES.
 ON BENDS, ATTEMPT TO RELOCATE THE SUPPORT TO STRAIGHT PIPE INSTEAD OF WELDING TO THE BEND

(I do not agree with all of the opinions expressed by Gerald May, the information contained in his pipe support document --but his experience noted on pages 34 to the the end of the document is very useful and applies to the topic of discussion.)

May also states:

Some Piping Engineers have a design rule NEVER to weld to a bend or elbow because:

 An elbow is counted on for thermal flexibility and the attachment stiffens it
 A welded attachment creates an undefined stress intensification factor

I agree !!

MJCronin
Sr. Process Engineer

 

[1503-44]

Me too. Just because you can do something, does not mean its a good idea.

 

[KevinNZ]

Good find

https://s3.amazonaws.com/suncam/docs/167.pdf

Clamps might all good until they slip! Rather weld pipe shoes and know from experience they are good for 50+ years outside under insulation.

"Some Piping Engineers have a design rule NEVER to weld to a bend or elbow because:

 An elbow is counted on for thermal flexibility and the attachment stiffens it
 A welded attachment creates an undefined stress intensification factor"​

I agree too, but will do it occasionally and will stiffen the elbow in the stress model and work out an SIF with FEA.

 

[KevinNZ]

Out our curiosity I looked at FEA for rings and pads. Many geometry variables to consider. This did show;
1. high stress in the trunnion to pad weld
2, if the pad is thin or wide it is not stiff and does not support the trunnion weld. The pad/ring pulls away from the pipe wall.
3. Not a lot of difference in max stress between pad or ring if the pad/ring is stiff.

 

[LittleInch]

Would a set on "weldolet" work better?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Exactly why I mentioned the welded full encirclement option.
At least a half, "saddle-like" as LI suggests.