large elbow in a vessel 2

[rlacombe]

I need a little help from the xperts. i have a 16" diam vessel and i am putting an elbow smack dab in the middle of the thing. making the vessel "L" shaped. The way i read it for ug44, it's stating that the part needs to be formed to b16.9 and i also have to do hoop stress calcs as if the elbow were a straight piece of pipe. I also want to put couplet and or maybe a pipe with a weld neck flange in it. Would this be acceptable to do with straight shell calcs?

 

[TGS4]

There are some unusual stresses that occur in elbows (owing to the fact that they are toroidal sections). I don't think that I would be able to justify the design-by-rules approach. If it were me, I would use a U-2(g) approach.

 

[jte]

It's unusual for me to disagree with TGS4, but in this case I will. If I have a vertical vessel with an outlet on the bottom head, some length of vertical nozzle, to an elbow, and some length of horizontal nozzle to bring the flange outside the skirt, I would determine the necessary pipe schedule and order the sticks of pipe and a matching elbow accordingly. [Having said, that, I'd put the elbow and horizontal pipe in B31.3 scope, but that's another thread...] I have strong doubts that most folks would do an FEA on every elbow they ever put into such service. Now, as far as VIII-1 is concerned, the nozzle neck (pipe) calc's are performed as for any other cylindrical shell. Thus, though quite unusual, I'd take the position that you calculate the required pipe schedule and use an elbow of the same schedule.

On a related note, how would anyone ever use pipe caps if they had to calculate their thickness as heads? Their shape is not mandated and thus, can vary... You can't look at 'em as though they are 2:1 heads...

jt

 

[TGS4]

Good points jte.

I didn't interpret this as an outlet nozzle with an elbow (which I would also strongly argue should be in the B31.3-scope), but rather an L-shaped vessel. I wouldn't necessarily perform an FEA on this, in fact I'd probably use B31.3 as my U-2(g) approach. (I've had enough of elbow FEAs for at least several months...)

Elbows are unique beasts. Under internal pressure (by the Bourdon effect) they tend to "open". Apply any external load to them, and a high stress shows up on the neutral axis. And, they are rather sensitive to out-of-roundness. as a first pass,I would recommend treating the vessel as a piping system, with the requisite SIFs for the elbow. If you're using a forged elbow, then that would be one matter, but if you are using a fabricated elbow, then that may be something entirely different (see my comment about sensitivity to out-of-roundness).

 

[jte]

TGS4-

I understand that the subject elbow is intended for use as the main shell on an L shaped vessel. I don't disagree with you 'bout the variation of stresses within an elbow, and the sensitivities involved.

But why go to B31.3 when UG-44 explicitly allows you to just use B16.9 fittings at full strength? My point above is that you wouldn't design the elbow on the outlet to B31.3 if you are extending the scope of VIII-1 to the face of flange, so if it's acceptable for me to put a say 24" elbow on a nozzle extension based on schedule calc'd for straight pipe, what makes it inappropriate for me to use a 16" elbow with a different name (i.e. "shell")?

rlacombe-

To respond to your other two issues: Yes, it is ok for you to use B16.5 weld neck flanges in your design, within their rating of course. Be careful to use the ratings in the edition of B16.5 which is referenced by VIII-1, not the latest edition. Some AI's will get picky 'bout this.

For nozzles which require reinforcement [see UG-36(c)(3)] you probably will not be able to use 'olet type fittings unless you get detailed dimensional drawings and can show that the area of reinforcement is sufficient. Fortunately for you, a couplet is a fairly simple geometry and you can probably just take your own measurements. I'd check with your AI first, though.

jt

 

[vesselfab]

Well if the opening for olet or pipe is small...it will be exempt from calcs anyway, so no problems

 

[TGS4]

jte-

Gave UG-44 a thorough read-through. I think that I now agree with you on the pressure-temperature rating issue. NOTE that UG-44 says to use the seamless pie equivalent, INCLUDING the thickness tolerance, which for seamless pipe would be 12.5%. (My emphasis added).

I'm still not clear if there are any rules (hence my reference to U-2(g)) about how to treat these fittings in reinforced opening calculation. If you have external loads (wind, seismic, nozzle loads, etc), I am certain that there are no rules in VIII-1 for how to us eB16.9 fittings.

rlacombe-

You may want to search through the interpretations to see if that issue has been resolved. As vesselfab said, I would try to work with a small opening that doesn't require reinforcement in the first place...

 

[rlacombe]

just talked to my ai. He says it's ok as long as the elbow is a seamless elbow. A welded or seamed elbow is however acceptable if (a) it's got a partial. or (B) you weld it yourself. (c) cut out the weld and weld it yourself.


Thanks for all your help guys.

 

[vesselfab]

HMMMM

will have to check on that
I was under the impression that standard manufactured parts did not require partial data as long as they were marked properly with spec and mfg's trade mark

 

[rlacombe]

Thats true but on non-shell parts or head parts.

If you buy a welded head, for instance, i do believe you should have a partial for it?


definately for a shell. Cant buy a welded shell without a partial.

 

[vesselfab]

but a head is not a standard manufactured part to another spec

apples are not oranges

 

[vesselfab]

i mispoke....old you know

yes....you are correct....not for head or shell

sorry

Now my halo is dented and damage

 

[rlacombe]

alright guys thanks for the help. Much appreciated.