Calculating stress and force when forcing a tubular round 2

[johg]

We are fabricating some heavy wall tubulars. Diameter up to 5 mtr and wall thickness up to 110 mm. Material grade S460. Due to some welding the tubulars are out of circularity. We cannot calibrate the tubular by rolling. We need to make use of a hydraulic pump to push the tubular round and weld in a beam or pipe to keep the roundness. Could someone advise me in calculating the force to be used to get the tubular round. We also need to calculate the stresses in the tubular.

All help or information is appreciated.

 

[desertfox]

Hi johg

I'm not convinced that hydraulic pumping will work, for a 5m diameter and 110mm wall thickness I estimate a pressure of 334BarG
just to bring the inside wall of the vessel to a yield stress of 380N/mm^2

 

[dhengr]

Johg:
You are using some kinda funny terminology for someone who should know what he is doing if he is fabricating this kind of stuff. Let me see if I’ve got this right. You are roll forming some pipes 5000mm o.d. x 110mm wall thickness x 10,000mm long (for my own visualization that’s about 16.4' o.d. x 4.3" t x 32.8' long cylinder); and you are running a full length longitudinal weld to join the two edges. Then, is this longitudinal weld a full penetration weld or only partial pen. weld, with the edges preped. in some way? This welding is causing the pipe to oval due to the welding? Off hand I don’t know what S460 is. You may be roll forming this in 10' lengths and doing circumferential welds also. The longitudinal welds are usually done first, on the 10' lengths. Then the circumferential welds are fit-up and welded to make the full length cylinder. In any case, is this the right picture and issue?

You need to involve an experience welding engineer or manufacturing/fabricating engineer with some experience with this kind of fabrication. You need some positioners/rolls for the pipe, and you need some automated welding equipment that can lay down that kind of weld. Welding process and procedures must be established for this kind of work. Basically, by sequencing your welding, inside and outside, a few passes at a time you can minimize the ovaling you are experiencing. You might need some internal braces during the welding and fit-up, as you’ve shown. That’s just brute force not something you really calculate. Generally, you are better off to under roll the pipe shape by a little bit, that is, not to quite close on itself. Then, from the outside you have to pull the two edges together by an inch or so for fit-up and to tack them together. An experienced fabricator will know how to do all of this, but fine tuning the exact process is also a bit of trial and error too.

 

[kingnero]

you might want to contact a company that makes large (pressure) vessels, like

http://www.ggi.be/

in Belgium, and some others (but there are not too many) in the world who deal on a daily base with equipment of this (and much larger) size.
A real-life visit shows a lot about the production.

 

[Cockroach]

I've seen this before, differential heating from the weld pulls it out of round. That's why per heats are used, albeit on much smaller sizes.

I know you can either roll the piece back to round or turn it. There is a local company here that will turn up to 60 feet ans machine off the high spots. Truly something to see, being that big, but it works and is very expensive.

I would bet that pressing between three rollers while turning, similar to tubular manufacturing in the steel industry, would work. You definitely would need to soften up the steel by heating, thus reducing the forces involved.

But let's not be so critical on the man, just some ideas or observations of mine.

Regards,
Cockroach

 

[berkshire]

Part of the problem can be a little flat spot at the edge of the plate where it leaves the roll.
It is common to bump form that final nip in a press brake.
This may be what is giving the poster a problem with ovality.
B.E.

The good engineer does not need to memorize every formula; he just needs to know where he can find them when he needs them. Old professor

 

[johg]

The tubulars has been made from several smaller tubulars (length 10") welding circumferential weld after welding the longitudinal welds. All welds are full penetration welds. The out of circularity is due to welded internals. We only need to adjust the outer end of the tubular, so we can weld another tubular to it.
Indeed it will be brute force. I was wondering if a local load calculation would do the trick.

Thanks already for the replies

jogh

 

[johg]

Sorry made a typo, length tubular is 10 mtr long instead off 10".

 

[prex]

You have a circular pipe loaded with two diametrically opposed forces. If these forces where continuously applied along the 10 meters of pipe length, the calculation is in the first site below, here, and is also found in the Roark.
However you want to displace only one end of the pipe, and I don't know of a closed form method to calculate this. It can only be estimated by assuming an equivalent length of actively reacting pipe : I guess that 10 times the thickness is a good guess.
If this is correct, the force required to overcome a 10 mm misalignment would be some 100 kN, a force that can easily be provided by a suitable jack.
However applying a single diametrical deformation is likely not what you need, you could experience non acceptable misalignements at other circumferential locations.

prex

http://www.xcalcs.com

: Online engineering calculations

http://www.megamag.it

: Magnetic brakes and launchers for fun rides

http://www.levitans.com

: Air bearing pads

 

[metman]

We only need to adjust the outer end of the tubular, so we can weld another tubular to it.

What if you wrap a steel cable around the circumference near the oval end and connect the two ends of the cable with a hydraulic cylinder then tighten the cable to attempt to force the tube closer to round. Or instead of a hyd cyl use a toggle action like used on a load binder with a huge long lever to pull on the cable ends.

Intuitively it seems like it might work but I could be all wet. If nobody here claims this idea is worthless, you could experiment with a really small scale model mock-up.

just to bring the inside wall of the vessel to a yield stress of 380N/mm^2

Fox: Are you sure he needs to go to yield? Would that not depend upon final out of roundness tolerance?

BTW Fox: the only reason I am putting in my two cents here was so I can give you an LPS which I could not do where you really deserved one on a closed thread. I needed to calc bending stress on a round tube "weld treated as a line." I don't have Shigley nor Blodget but I do have Scahums Machine Design. You saved my bacon once again. Kudos

Design for RELIABILITY, manufacturability, and maintainability

 

[SnTMan]

Commonly done with multiple jacks and pieces of pipe. Also screw-type dogs welded to one cylinder to force against the other. The thicker, the bigger. You may not get really, really round, just acceptable alignment. Yielding may not be necessary if the goal is to just align and join. It'll likely still move around after removing the jacks, dogs, etc.

Regards,

Mike

 

[desertfox]

Hi metman

If you correct the out of round of the pipe by whatever means ie:- hydraulic pumps,jacks etc and if the stress imparted to the pipe is only within the elastic region, the pipe will return to its out of round form when the pump or jack is removed, so yes to correct the form it needs to go beyond the yield stress.

desertfox

 

[prex]

desertfox, he needs to correct the out of roundness to properly line up the edges of two tubes for welding: they won't return back after welding.

prex

http://www.xcalcs.com

: Online engineering calculations

http://www.megamag.it

: Magnetic brakes and launchers for fun rides

http://www.levitans.com

: Air bearing pads

 

[desertfox]

hi prex

I read the OP's post as he needs to correct out of roundness post welding not prior too.

desertfox

 

[desertfox]

The quote from the OP

"Due to some welding the tubulars are out of circularity"

 

[johg]

Thanks all for your input. The out of roundness was due to welding longitudinal weld and some stiffeners at the other end of the tubular. We needed to fit up some other tubular and weld the circumferential weld. We calculated the needed pressure and beam to push out the out of roundness. The calculation was necessary because we had no clue what pressure was needed. We did this before but only on thinner tubulars and just tried.
The link to Roark was very helpfull. We had the book on our shelf but not thinking about it. Somebody just has to trigger you and put you in the right direction.
For safety reasons we calculated for a maximum of 30 metric tons. This was sufficient to push the out of roundness to an acceptable level.

I would like to close the thread by thanking you all again.

Johg