PCA Circular Tank for Vortex Drop Structures

[kapow]

Hi,
I am working on a design for a Vortex Drop Structure and I am intending to use the "PCA circular Tank" document as a design reference. However, the parameters for my drop shaft are wayyyy beyond the "H^2/(D*t)" values provided in the appendix and the supplemental of the PCA circular tank document. My shaft is about 40ft dia. and extends nearly 200ft into the ground. I've been told that thicknesses of 2-3 ft are typical for these structures.
Since I am unable to use the PCA design tables, I decided to model the tank using Staad with fixed-pinned boundary conditions at the base and top of the shaft in order to determine the hoop stresses and moment demand.
My results are showing some pretty high moments at the base where I need bundled (2)-#11 bars at about 8" spacing. I thought this seemed reasonable given the 200ft depth, however, I have seen designs for similar shaft dimensions, but slightly shorter depths showing only single #8 bars at 12" spacing.

Does anyone have any experience designing these types of structures? Is the fixed pinned condition too conservative for design? or should I incorporate some soil springs?
Is there a reference that is more up to date or more appropriate to use than the "1993 PCA Circular Tank" article? I mean it is 2023 now.

Any advice would be greatly apprecated.

Thanks

 

[ALK2415]

are you refereeing to this type of structure ? with 200 ft depth ?
if so, you should apply soil-structure interaction and using a winkler springs model for the surrounding and bottom soil

 

[kapow]

Yes. The inner components will look different, but yes the picture is correct.

Thanks, I could try to input some winkler springs, however determining spring constants seems to be a bit of a rabbit hole. If I understand correctly, it is a function of the depth, FE plate sizes, and expected soil movement.

 

[JedClampett]

I've designed much shorter ones. PCA's Circular Tanks without Prestressing will likely work, but I would design the walls for the pure hoop tensions, reinforce the bleep out of it using a low allowable tension (14-16 ksi) and only using the PCA reference for wall moments (fixed case) into the slab.
One of the problems with the PCA reference (and it's still pretty good) is that the resolution of the hoop stresses into the slab is not well explained. You need to add circular bars to account for it. So I would:
[ul]
[li]Design the cylinder for the pure hoop stresses. No PCA. No finite element.[/li]
[li]Design the wall cylinder joint for the PCA generated moment. Are you taking it into account twice? Yes![/li]
[li]Add hoop bars in the base slab to match the bottom few feet of hoop tension.[/li]
[/ul]

 

[kapow]

This definitely helps. @JedClampett, I've seen some of your other posts mentioning the hoop stress in the bottom slab. I will definitely add those circular bars.
From taking a deeper look at the PCA tables, it make sense why I've seen design examples specifying smaller amounts of flexural reinforcement in the walls. It appears that the moment demand decreases as the as the H^2/(D*t) gets larger. It seems it would be conservative to use the maximum H^2/(D*t) values provided in the PCA tables in the case where the actual H^2/(D*t) exceeds the table limits.

 

[WesternJeb]

For what it's worth, I did a few of these a couple of years ago and also found there were no up to date documents. I would recommend doing hand calculations for this type of structure under those loads to verify your model; the calculations are awful but worth having the verification. Ours was 100' diameter and 100' deep and we came up with 8" walls being required but our senior engineer told us he usually specifies at least 15" as a serviceability claim. That being said, 2'-3' for a 200' shaft seems pretty reasonable to me and sizes likely aren't being controlled by the actual stresses in the structure but more of a "this feels right".

Look up Lamé's equations for thick walled cylinders.

 

[ALK2415]

check this Reference .. (if you have a general layout, will try to modelled in STAAD.PRO), I think it is a Vital FLOOD CONTROL structures, which every big city will require soon or later....
Link

 

[JedClampett]

Skimping on wall thickness is a big mistake. Concrete is cheap. If the walls are 2'-3" thick, the rodbusters can go in the gap and tie the bars. Plus, that wall thickness is much less likely to leak. You're not going to get good mud consolidation in an 8-inch wall no matter what.