Horizontal tanks

[L17Aurora]

Being a nosey mechanical engineer I was wondering what calculated pressure is used for a round horizontal atmospheric tank.
I know pressure varies with depth, so on a vertical one the pressure is the same all round the tank depending on depth. For the horizontal one you would have "zero" at the top and max on the bottom of the shell. Are calcs usually based on the max?.
Regards
Mick

 

[europipe]

Yes, plus more because the operatig press. is lower then the design press.

 

[L17Aurora]

Cheers
Mick

 

[L17Aurora]

Another one on Horizontal tanks.
Most have 2 supoorts (best for everything) but is there a rule of thumb for when to add a central one to avoid buckling. ie if L/D gets over a certain figure?

 

[MJCronin]

L17...

Yes.... there is a rule of thumb... don't do it

Using three or more supports on a metalic horizontal tank or vessel is not a good idea and is generally discouraged in the industry. Buy more tanks if necessary

It creates an indeterminate structure that make a mystery of the question "how much load is on this support saddle"

My opinion only

-MJC

 

[L17Aurora]

That's what I thought
Cheers

 

[SnTMan]

L17Aurora, the "Zick analysis" can be useful for locating the saddle supports and analyzing resulting stresses in the shell, etc.

Regards,

Mike

 

[JohnGP]

I don't know of any rules of thumb apart from MJC's, but that doesn't mean to absolutely not do it. I've recently seen some process vessels (I know, I know, we started off talking about atmospheric tanks) that were about 4.5m diameter x 29m long, and designed for 70 Bar - these had a centre support, which was fixed. That allowed the two outside ones to be sliding type to cater for the 280C operating temperatures.

John

 

[SnTMan]

You see a lot of older stuff with more than two supports. We're looking at a replacement exchanger with four.

I don't believe support loads were calculated accurately however.

Regards,

Mike

 

[L17Aurora]

So it could be said then that as long as the tangential shear stress q is less than the 0.06 x E x t / r formulae. (or .8f), and all the other calculated stresses are below their allowables it won't buckle.
I'm not designing this and I'm not a structural eng so please excuse my ignorance.

 

[vesselguy]

L17Aurora,

Answer to your above question is yes. Just run thru the Zick's formulas and not exceed the allowable requirements and you're OK; for 2 saddle supports that is.

MJC is right on the "rule of thumb". More than 2 saddles is to be avoided, but can be done if required. Example, this is normal for extremly long bullets. For these extremly long vessels, no matter how you move the saddles, the mid-vessel bending stresses are so high that you must put in a third saddle to resolve the problem. You solve the reactions as a beam on elastic foundation with multiple supports and then you take the loads and apply to zick's formula. We have done this 16yrs ago and it was no problem. Search around and you'll find a few paper on this subject. Naturally, this kind of problem requires close cooperation with the civil/structural engineer to provide a solid foundation to reduce chances of significant differential foundation settlement. No-a-days with FEA, you can model the system and solve it.

 

[MJCronin]

I agree with vesseguy.....

I also would like to point out that rules of thumb are not absolute....

The key here is the phrase:

" Naturally, this kind of problem requires close cooperation with the civil/structural engineer to provide a solid foundation to reduce chances of significant differential foundation settlement"

If there is significant settlement in ANY of the multiple tank/vessel supports, the design can be in jeopardy... safety must be considered!!

A geotech buddy of mine one said "the whole world is built on clay"

Why take a chance ????

Support settlement is not such a big deal in two saddle designs.

Why not simply select multiple tanks ??

My opinion only

 

[L17Aurora]

Thanks Guys, excellent reponses & theory's.
I do hate it when one line answers come back in forums like "go to first pinciples or Hire a vessel engineer etc." where you don't get the expertise of the engineers.
Thanks again.