Pressure Vessel with Multiple Saddles

I havent heard about this method, at least with that name.

So you have a "superstatic" beam? Replace one reaction with unitary loads, solve the equation system and go back solving for that load. That's the way I have learned (as far as I remember...).

Before you go any further - are you ENTIRELY sure you need/want to go with three supports?

It is usually a BAD idea to go with more than two supports because any settling of foundations/supports on 3 or more will disrupt the stress distribution and could cause major problems.

The three saddle supports where specifically requested by our client, even with our advice to just use two. Unfortunately, three saddles will be used. The purposed of determining the load distribution on the saddles is so the structural engineers can design the pit that this monster will rest in.

...I love this job!

Yeah - gotta love clients.

You should still strongly suggest against more than two supports - it is preferable to install stiffener rings around the vessel in between the span.

Barring better judgement - if you have any say in the design of the foundations (if this is a new installation, or if possible) you should strongly suggest they design the foundations so that they are tied together and can not settle independently.

Lastly - if you put your email address in your user profile (it isn't recommended to post it in a thread because of spambots) I will send you some good information.

This is an intersting problem. During my vessel engineering days, the longest vessel I designed was about 60-70 feet in length and if my memory serves me right it was 10-12 feet in Dia. I do not exactly recall but with not very high pressure and temp as it was an storage vessel. After many meetings and discussions, we had only two 150 degrees saddles. Back then it was considered a bad idea as well to support on three saddles.

If you have some info regarding this, please do not deprive us and share with us. All of us, the readers, will appreciate it.

May be Mr. Abdel Halim Galala of Cairo, Egypt will shed some more light on this problem if he comes along in this discussions.

Have a great day.

Regards;

Nadeem

12012008

Good point - I shall share with all:

Please download:

http://ifile.it/lzb2a3r

You are correct in that the vessel is 50ft long and 12ft in diameter. We will be converting the ASME vessel into an autoclave that will be used for the aerospace industry. It will see a max working pressure of 100psi and about 450 deg F. My client insists on using three saddles because their previous vessels where out of round (not stiff enough) at the door section. Our normal design uses two 120 deg saddles and certain distances along the vessel. Using three saddles can be problematic (according to my research) but it has been successfully done.

Then the papers I posted will help you (note the questions and answers at the end of the singapore paper in reference to three supports)

I had a problem like this years ago (before computer design). I got the reactions to the supports by considering the vessel as a continuous beam. I sized the three independent foundations based on these loads. However, I later realized that due to the stiffness of the vessel in relation to the stiffness of the foundations, it would have been just as good to get the loads based on the simple beam reactions of the vessel between the supports. The stiffness of the vessel was a lot more than the stiffness of the supports.

If settlement is a problem, then by all means check with the soils people to check your settlement. Differential settlement can cause great damage sometimes. At the chlorine plant I worked at, a chlorine storage tank settled so much that it broke a connecting pipe and liquid chlorine spewed into the air. Fortunately, no one was injured. A continuous monitoring of the chlorine storage tanks was then initiated.

You could us a flexible medium between the saddle and the vessel or you could make the radius of the saddle larger than the radius of the pipe that is going to be resting on it.

I think we are all assuming that the vessel is cylindrical, and perhaps 3 times as long as its diameter, hence very stiff in bending. However, perhaps it is thin and long, in which case the client might not be wrong in suggesting more supports. If the vessel is stiff, then using three supports would be a very bad idea, because you can easily imagine cases in which the entire weight of the of the vessel is supported on just the center support, as the foundation shifts. Perhaps you could model the vessel and likely foundations in FEA and show the client some of the possible problems? Perhaps the supports could be compliant (e.g., springs)?

Give them what they want. But, don't give them problems. Two saddles can be supported by three footings, if the footings are laid out in a triangle and two of them support one saddle. This geometry is odd and probably not optimum but has advantages as three points determine a plain while two determine a line.

If the saddles are anything other than lineal there will be potential problems intrinsic to the design.

However, three footings is quite sensible and could be optimal if they are dealing with soil bearing capacity limits with no bedrock within reach. Please present this option as the alternative they are specifying is problematic as well as complex to calculate.

The ideal load distribution of a three point suspension is a well described pre-solved problem with applications from surface plates, to moon landers. Adapting this to a double saddle suspension for a tank should be relatively easy.

Otherwise, I would estimate that stress fractures of this tank are inevitable.

Sincerely,

Mr. Gee