Pressure vessel sidewall loading

[penpe]

I'm a structural engineering contractor on assignment at an oil refinery, working with piping designers and process engineers. The objective is to design and implement pressure-relief safety valves for heat exchangers. The valves are often located on or near existing platforms for accessibility. Often the available platforms are attached to, and supported by the sidewall of a (vertical steel cylinder) pressure vessel. I'm hoping to find a reliable and conservative, (and hopefully simple), way to analyze the vessel's sidewall. The platforms are typically supported by cantilevered steel beams. Therefore the moment would often the controlling factor for loading. The new loads introduced for the safety valve, block valve, and piping usually are located at the cantilevered end, so moments are increased dramatically. My boss typically refers this analysis, when needed, to an available mechanical engineer; but getting the results is like pulling teeth! Any tips, tricks, or rules of thumb for this type of analysis would be appreciated. Thanks.

Pete, P.E.

 

[BAretired]

Add columns below the cantilever to relieve the tank wall of any additional stress not contemplated in the original design.

BA

 

[fancypants]

I would suggest you get a copy of "Pressure Vessel Design Manual" by Dennis Moss and published by Elsevier. If you are a structural engineer, you should have no problem digesting and using the needed information in this book. There are other books etc...

 

[JStephen]

Bednar's pressure vessel handbook has an approximate method for checking stress due to line loads applied to a vessel wall, but the results tend to be overly conservative.
WRC 107/297 (or whatever the latest version is) will give stresses in a shell for certain types of applied loading. Simplified versions are included in some of the pressure vessel handbooks. And also implemented in some of the pressure vessel design programs. This would generally assume that you have a reinforcing pad to which you are welding.
You have to consider the combination of local stresses and pressure-induced stresses in the shell.
You're also causing some variation in the base moment, anchorage, vibration, and foundation calculations, which may or may not be negligible, depending on vessel size, etc.
If the original vessel and platform designs are available, it may not be too bad to go through and see what effect an added moment has. If existing designs and/or drawings don't exist, you'll pretty much be designing the vessel over again by the time you're done.

 

[BAretired]

I would be reluctant to mess with someone else's vessel design. If problems arise, he can wash his hands of the entire matter because someone changed his design. Much better (and probably cheaper) to add materials to carry the loads without reliance on the tank wall.

BA