Allowable local buckling compressive stress in tub ring 1

[Soln]

I'm trying to gain a better understanding, and form parallels, of allowable stress formulas listed in AWWA D100 (eq 3-11) and API 650 (E.6.2.2.3-2a/2b). How are these formulas derived? AWWA D100 eq 3-11 is similar to Boardmans's equation (S=Et/29r(2-200t/3r)), and these are both conservative compared to API 650 E.6.2.2.3-2a/2b. Can anyone steer me to literature, or help explain what's behind the curtain. I'm dealing with tanks supporting large platform loads (400,000lb) and need to ensure that a) my tub ring can take the stress, and b) that I transfer point loads at the top into the shell adequately. I'm looking at vertical stiffeners to transfer the loads down into the shell at the top (Boardman approach), but because of the proximity of the point loads to each other, I may need to count on a ring girder to act as a beam to help distribute the loads into the shell. Again, any help on this would be much appreciated- shell's purchased, so thickening the shell is not a friendly option.

 

[JStephen]

The equation in App. E of API-650 is specifically for seismic loading, and not intended for general compression-in-a-cyclinder situations. It should allow for the increased stability due to having fluid pressure inside the cylinder, may allow for an increase due to bending compression as opposed to axial compression, and should include a 1/3+/- stress increase normally added for seismic. The stresses in AWWA are intended for supporting sustained loads. And AWWA is generally more conserative than API-650 anyway.

Roark's Formulas For Stress and Strain includes load cases for loading in a cyclinder similar to AWWA, but also gives the references for them.

 

[Soln]

Thanks for the explanation JStephen. Can you direct me to where, in Roark's, are the similar load cases. I've been reviewing, and must have missed it.

 

[JStephen]

Actually, it was an external pressure load case I was thinking of when I wrote that. But refer to the chart "Formulas For Elastic Stability of Plates and Shells", which is Table 35 in the 5th Edition, specifically, cases 15 and 16.

 

[MJCronin]

When the load atop a thin-walled shell structure (like a tank) becomes massive, another method of structural support should be considered.

Is there some process reason why this massive load must be located on top of the tank, or did the CAD designer simply think that this was a good idea ? Would a structural steel frame that is spanning the tanks be acceptable ?....If not why not ?

It will be much more expensive to try to increase the tank wall thickness enough to support this load rather than use a supporting frame

 

[Soln]

Thanks for the clarification (JStephen)- I can see by working out the case in Roark's that it's in the same vein as AWWA D100 (eq 3-11) (1/sqrt(3)E/sqrt(1-v^2)=17.5E6)-- Roark's critical stress is greater by a factor of 10- could this be a safety factor?
MJCronin- Yeah, somebody overlooked the shell stress when they dreamt up this configuration- I'm close to taking up the loads, but I need to transfer point loads into the shell with vertical stiffeners (Boardman approach) and take up overlapping loads with a ring girder acting as a beam- any suggestions here? Experiences?