API650 v11 - Appendix E - extra thickness of the shell

[bourguideche]

Hi gentlemen,

Please excuse me in advance for these beginner's questions, but I need to understand the API650 (version 11) code which is used by our customers.
I think of having succeeds in understanding the formulas and the typo mistakes but there remain to me these questions:

- When we calculated the values of compressive stress at shell bottom and maximum allowable compressive stress, how can we determine the extra thickness of the shell to apply for seismic risk? And this, for each shell course.

- How, in the current version of the code (v11), can we use the old value of Z (Zone coefficient) used in the previous version?

Thanks a lot for your help.

 

[JStephen]

If the calculated compressive stress is too high in the shell, then the additional thickness is a matter of trial and error. It may help to notice that there will be very little change in the moment calculation from the additional shell weight. For unanchored tanks with marginal stability, it may be better to thicken the annular ring to reduce the stress rather than thickening the shell.

You can't really use the Z factor in the current standard. You can check in prior versions, and see what the equivalent acceleration was, and apply that acceleration to your tank. But it won't be an exact match. The ratio of convective to impulsive acceleration may be different between the older and newer standards. The older standards used direct sum for calculating moment, while the newer one uses the square root of the sum of the squares, which will change the moment as well.

 

[bourguideche]

Thanks a lot for your answer.
The use of Z factor is now clear for me ^^

But on the thickness of the shell, I'm not really sure to understand. If the better method is to thicken the annular bottom ring, why my customer apply an extra thickness on each shell course? For example I have calculate the thickness of a little tank with the classical method (section 5 of the code) and the result is five courses with thickness from 6 (bottom) to 5mm (up). Our customer, with seismic risk, find five courses from 13 to 6mm...
I thought the seismic risk generate an (virtual) extra load which be use in the calculation of thickness of each shell course. Isn't it?

Thanks in advance

 

[JStephen]

If the shell thickness calculates out as 6mm and is then increased to 13mm due to seismic, I would assume there is either very very high seismic loading or some kind of mistake. The mistake could be numerical, could be in using the wrong equations or wrong approach, or could be a typo in API-650 itself. It could be a matter of adjusting the wrong variables to make the design work out. If you don't actually know how the calculations are being done, it's hard to guess where that is coming from.

The issue with the annular plate comes from Eq. E.6.2.2.1-2(a) and (b). When J is near 1.54, the denominator is small, and stresses are large. If wa is not limited by the 1.96HDG limitation, then thickening the annular ring (or adding an annular ring if there's not one) will help. If the tank is anchored anyway, or J is not near 1.54, or the wa is limited by the 1.96HDG factor, then thickening the annular ring won't help.

 

[bourguideche]

Ok, thanks to you...

I will try to obtain more informations from the customer and work on your ideas...

Have a nice day