Point Loads on Water Tank Shells 1

[scofie]

I am designing an antenna array support frame to be mounted on top of a water tank (on the shell's crown). Rather than a finite element modelling of the entire tank shell, does anyone have any references for hand calculations and approaches for analyzing the tank shell for local loads introduced by the antenna support frame?

The entire antenna support frame will be subject to overturning due to wind loads, so at each of the connections of support posts to the shell there will be either an uplift or downward load. Therefore, the load on the shell will not just be a uniform pressure or suction.

Thanks for the help.

 

[tincan]

It sounds like you are installing the tower independent of the tank roof with the supports resting on the tank shell. If so I would assume its a small diameter tank. For a preliminary I would assume the point loadings of the tower as anchor bolts in reverse to check the loading. Most tanks have a stiffining ring (angle or plate) at the junction of the shell and the roof. The ring may need to be reinforced for any lateral/tension ring effect by the tower support legs. This may give you a start, whats the tank diameter, weight,height, etc for the tower and its windloading?

 

[scofie]

Tincan, thanks for the reply. The antennas will be mounted on a 12 foot face width triangluar frame made up of pipe. It is not an antenna tower. The water tank is a 100,000 gallon water tank, 130 foot tall, 30 foot diameter, torospherical head, 1/4 inch thick plate. My point reactions at the antenna support posts will be in the ballpark of 3000 to 5000 lbs. The issue is basically point loading on a torospherically shaped plate... what is the bending stress induced and what is the buckling condition.

 

[corus]

Local loads on shells are assessed using BS 5500.

 

[warrenw]

Scofie:
The "pressure Vessel Design Handbook" by Henry Bednar
has a section on local stresses in shells due to attachments.
For your case;
radius (r) = 144 in
t = 1/4" - c.a.
corrosion allowance = 1/16"
force (F) = 5000 lbs
L = perimiter of attachment pad (foot print)
assume L = 72" ( 18" square pad)

Stress = [1.75*(F/L)*sqrt(r*t)]/t^2
= 17,962 psi

Also check out ;
Welding Research Council (WRC) bulletin 107

hope this helps
Warren

 

[VoyageofDiscovery]

Hi scofie,

I have been down this road before. The telecom industry is pushing hard. I think what you are attempting is not safe. Your point loads are too high. In all probablity the original design may not have treated the tank as a shell. The old guys designed these as membranes, large steel bladders. The only way I would attempt this is by keeping my reactions to less than 300 lbs.

Regards

VOD