Leg Design

[JStephen]

In designing a vertical vessel with leg supports on the outside of the shell, it is possible to approach it in two different ways.

One is to assume a pinned connection where the leg attaches to the shell and then design the leg for the resulting eccentric loading that results. This will give a larger leg and smaller connection to the shell.

The alternative is to assume a fixed connection at the shell, which eliminates the eccentric loading from the column, but requires the design of the shell for the resulting moment.

In looking at Bednar's and Moss's pressure vessel books, it looks like both use the first approach.

However, when you assume lateral loads on the vessel, that seems inconsistent with the use of the pinned connection assumption.

Any thoughts on the issue? Any reason to avoid the second approach?

 

[dhengr]

JStephen:
I would say that the first approach (pinned column with eccentricity) should normally be conservative for the leg design, but is not the real world condition, it’s the quick and easy/dirty way. The second approach with the shell inducing a moment into the column (or visa-versa) is the way it really acts, and you can’t ignore that whatever assumptions you make to do your design of various parts. And, it more accurately treats the shell, which after all is really the weaker link or the important part. The column load plus eccentricity and the induced moment really should be (come out to be) the same value. I would worry about what the induced moment does in the way of radial loads on the shell (loads perpendicular to the shell plate). They may require some distribution into the shell at the top of the col. and down at the joint btwn. the shell cylinder and the lower head or cone. And, in the extreme these gussets or stiffeners become a tension or compression ring. Rather than your triangular load distrib. to represent the induced moment, I would be more inclined to have (design for) fairly concentrated reactions at the col. top and lower joint. As for lateral loads on the whole tank you probably do need full t/c rings or bracing btwn. the legs to distribute the global lateral loads, not just to the shell for resistance.

 

[r6155]

See Process Industry Practices PIP VEFV 1111 to 1115 and British Standards BS EN 13445

Regards
r6155

 

[MJCronin]

This book contains a methodology and discussion to evaluate vertical pressure vessel legs AS A GROUP.

http://www.springer.com/gp/book/9783662124437

Also, I believe that some of the pressure vessel design software (COADE's PV ELITE ?) automatically considers these more subtle design issues for vessel legs.

Contact them

MJCronin
Sr. Process Engineer