Flat bottomed vs dished

[hydrosim]

A simple question from a purchasing bod ..... what are the design / specification constraints that determine whether a vessel should be flat bottomed vs dished?

 

[LSThill]

What is the services condition?

 

[hydrosim]

5 to 14.5 bar,g at a maximum discharge temperature of 120 degrees C

 

[codeeng]

Generally a dished head will give you the best in terms of thickness and additional volume. An example would be a 2:1 head with a calculated thickness of 0.2" versus the equivalent flat head at 2" thick in rough numbers. On the other hand flat heads for small diameters and low pressures may be more practical.

 

[JStephen]

If it's supported on legs or a skirt, and process considerations don't mandate one or the other, just specify either acceptable, give the minimum total capacity, and let fabricators choose the most economical.

For field erected tanks, you can use a flat bottom, supported by the foundation, for pressures up to 15 PSI, with the tank held down by anchor bolts (API-650 design). But I'm not aware that the pressure vessel codes have any provision for this kind of construction.

 

[JohnGP]

hydrosim,
For the pressures you have (14.5 bar, maybe more for design pressure), dished heads are going to be more economical. Diameter would need to be less than say 150mm before you might consider flat ends, and even then you could be using standard pipe and pipe caps.

There are all sorts of factors depending on how the flat end is attached that would affect the result, but as codeeng suggests, the flat end will need to be of the order of 10 times the thickness of say a 2:1 semi-ellipsoidal head. This using ASME VIII requirements, but other PV design codes won't be much different.

Cheers,
John

 

[panduru]

JStephen:
More often the process-side reckoning seems to be that where there is NO requirement of (a) a discharge nozzle at the bottom AND (b)a skirt support for any other reason, then the natural choice is a flat closure at the bottom. The logic seems to be that a flat plate resting its entire surface on a concrete bed can be made as thin as a formed head for the same loading since it is resting on the bed. But I have not seen such design approach favored in any technical literature.

In fact, it is not uncommon to find even flat-plate top closures with a "stiffening frame" or "ribbing" which serves two purposes, viz. to support attachments above the head and to provide rigidity to the span of the closure.

I was always apprehensive if such an approach is encouraged in the design codes, or if advisable at all.

Comments from both process and mechnical sides, most appreciated. Thanks.