Thickness of vessel skirt?

[shahdadf]

Dear Friends,
is it possible that for a long length vessel skirt(about 7500mm) we consider 2 thickness (for example 3000mm with thickness of 13mm and for 4500mm thickness of 10mm)?
regards

 

[TomBarsh]

As 'devil's advocate' here, what advantage would that bring?

Most designers would use ASME method (UG-23(b)) (or similar method) to determine the allowable compressive stress in the skirt; the allowable stress is a function of the diameter and nominal thickness (but not, surprisingly, length). In your case the load is essentially the same in both sections of the skirt (slight difference due to the additional weight of the skirt acting at the bottom). So what gain is there to make the different thicknesses? There may be some for functional or other reasons, but there appears to be no structural benefit.

However, sometimes the skirt section may be made of different material at the top of the skirt where the higher vessel temperature exists or if vessel is expensive material, eg: stainless, etc., then use a carbon steel skirt.

This latter point does give some credence to changing the thickness: Where the skirt temperature is high, you might use thicker material in order to keep the allowable stress higher. Where temperature is lower you could use thinner material but maintain similar allowable stress. But I doubt this is done very often in actual practice.

 

[JStephen]

If wind or seismic overturning is governing the design, that would be a reasonable way to approach it.

 

[jte]

Tom-

It is done on occasion. As you pointed out there are times when it makes sense such as high head/shell temp and as shahdadf's example of a skirt which is over 20' high. It's not unusual for wind/seismic at the top of the skirt to govern skirt thickness since the allowable stress is reduced due to being at the vessel's design temp.

Another potential issue would be in lifting/rigging when erecting the vessel as a vessel with a relatively long skirt and short shell may see high bending stresses in the skirt at the initial pick.

jt

 

[innovation2]

A good rule of thumb is to design your shell length, including your skirt at 20% of the vessel height. You will get better results in your tower analysis. I have changed thickness due to static pressure with out any problems, so I would see any difference with skirt design. Just remember to add a taper.