Out of roundness in skirt between new and old stacked vessels

[tmurphy66]

Hopefully someone can point me in the right direction.

We have a set of two vertically stacked vessels and are replacing the top vessel only by cutting the skirt between the two. The problem we are facing is the out of roundness of the skirt, both old and new. The skirt is 5/8" thick and the OD of the skirt is 248 1/2". At the cut line, the new skirt is 1 7/8" out of round from a perfect circle. The difference between the new skirt and existing skirt at the worst place is 2 1/4".

Can anyone point me toward any information pertaining to this? How much out of round is too much? Any formulas for bending the skirt into place?

The party responsible for installation say they can dog it off and make it fit, they just need someone to give them the guidance on weld size, lug size, etc.

Thank you in advance.

 

[r6155]

If the two parts have de same outside perimeter -same OD- hence the party responsible can dog it off and make it fit.

Regards
r6155

 

[dhengr]

Tmurphy66:
That kind of out-of-roundness wouldn’t usually be a problem under normal fabricating and erection circumstances, but I’m not sure what the various ASME and client/plant specs. might say about the problem or the new welded joint. Does this joint have to be a full pen. weld? Are both the new and the old/existing skirts .625" thick? Are they both 248.5" O.D., do they have the same outer circumference length? How far is it from the girth cut line to the attachment on the vessels, on each, the new and existing vessels? What are your wind and EQ loadings, any other loadings on the new girth joint, and how do they come into play in that joint design? A couple inches out of round in a 20.7' dia. cylinder which has been flame cut would not be too unusual and should certainly be able to be brought back into line to make the new weld. The biggest thing might be coming up with a scheme to temporarily reconnect the two skirts in a simple efficient manner so the top tank doesn’t have to hang from a hook for days. You might want to have access to the inside of the skirt during the fit-up and girth rewelding process, and you would want to prep the weld joints before you brought the two tanks together. Depending upon how you did the skirt cutting, you may have to true-up the old skirt cut edge so that it really ends up being in one horiz. plane, rather than all over the place, above and below that plane. You might add a few 20.7' minus stiffeners to the inside of the shells to bring them more closely into a round configuration. Then add some temporary vertical ears to one of the skirts at about 30 or 45̊ or closer intervals so that the top tank/skirt can, more or less, be dropped right down on the lower skirt and engage fairly accurately, with some min. shimming and adjustments. These vert. ears would be welded to one skirt and bent in 5 or 10̊ so that they slip inside the other skirt. Then you can make holding welds at these ears so that the upper tank isn’t swaying in the breeze. Then the final gagging and dogging will involve a min. effort, and the final welds can be made. Take some care to sequence the girth welding so that you do not pull the upper tank in one direction, out of plumb.

 

[racookpe1978]

Well, if you're trying to mate two conic flame-cuts in mid-cone, each conic plate being 5/8 thick, you're only choice is to dog (force fit) the two conic plates together. Try stree-relieving them after welding, it "may" help relax the embedded strains and stresses, and "may" help prevent cracking.

I'd replace one or the other conic plates completely.

Out-of-round by 2.25 inch in 245 inches diameter is less than 1% difference.

 

[r6155]

@ racookpe1978. "Out-of-round by 2.25 inch in 245 inches diameter is less than 1% difference." is not applicable in this case.
See UW-31

Regards
r6155

 

[chaulklate]

I don't think Code out-of-roundness rules apply to non-pressure parts like skirts. I would dog it and full penn all around.

 

[ColinPearson]

Have you taken a number of diameter measurements and compared the average? How about measuring the outside perimeter of both sides? It can be hard to tell, because a 'wobbly' perimeter can be substantially larger than what the theoretical perimeter should be at that diameter, but by taking both kinds of measurement, you can start to get a feel for whether the parts will match up. If the old or new pieces are bevelled at this points, keep in mind that you are not measuring the perimeter at the same horizontal plane... you're 1/2 a bevel width above on the new and 1/2 width below on the old. Do they seem to be out of round in relatively small locations or wobbly all around? At 5/8", your fitters should be able to work quite a bit of slack around the vessel. But if they don't know how to weld U-bars or key plates (which I assume from you saying they needed help with weld sizing, lugs, etc), then you may want to look into what kind of boilermaker expertise they really have. It's no fun to leave a crane hung up for days while the field tries to make an impossible fit. Perhaps you meant lugs for wind, etc. Hope it fits great when you're done!