Large Concrete Tank Design 1

[ARLORD]

I am designing a concrete tank, 280ft x 100ft x 30ft deep, containing water. The b/a and c/a ratios are larger than tabulated PCA design guides. Therefore, I designed the walls as a retaining wall structure, no problem. 3ft thick at base for 12ft first of height, 2ft thick for next 12ft, 15" thick at top for 6ft.

My question is what to do at the corners. How should I design the horizontal reinforcing at the corners. Using the PCA tables for horzontal moments, the maximum moment is near the top, and require way too much reinforcing.

What is the best approach for the corner horizontal wall reinforcing design.

 

[cntw1953]

I would take unit plan x-section at a few feet interval, and analyze it as hollow rectangle structural frame with load intensity varies with depth. You can place uniform dummy (unit) load on it, then multiply depth to the results to obtain internal reactions at locations/elevations of interest.

 

[JedClampett]

First of all, some editions of the PCA Tables have major errors (corrected with an innocent looking errata), that might affect your results. Personally, I'd throw them away and get the Bureau of Reclamation Publication on the Design of Rectangular Plates. Then get the largest aspect ratio case in the publication, a/b=1.5, and design the corners for that. If the moments are large, they're large.

 

[ARLORD]

JD,

I have those tables also. You say use the largest ratio a/b=1.5. My a/b = 5. How can you justify using 1.5.

 

[cntw1953]

By the way, since the wall length is much greater than the height, if you have designed the wall as cantilever, then it would be excessive to place full load on the unit box frame again. I would use unit diplacement method (assume cantilever subjects to unit load) to estimate equivalent spring support at each frame level, then provide the spring supports along the box walls but at the corners. Any analysis program will be handy, however, the process can be lengthy and tedious. I think a 3D plate FEM maybe the best fit to take on this task.

 

[miecz]

If your tank is above ground, it will have tension in addition to bending at the corners. That tension will likely be highest near the top of the tank, as well.

In any case, the bending and shear are usually quite high at the corners. Consider keeping the walls 3 feet thick at the corners, and stepping the thickness as you did from the base.

 

[rowingengineer]

Chamfering the corners combined with angled bars is a possible way of handling the large moment.

Arguing with an engineer is like wrestling with a pig in mud. After a while you realize that they like it

 

[ARLORD]

miecz,

Yes, I think I will try going with 2ft thick wall at the corners. I initiailly thought of doing this but was not sure what others have done. I'm curious what has been done in the past to deal the the corner forces for tanks this large.

 

[miecz]

You know, one of the assumptions of the tables from the Bureau of Reclamation Publication is isotropy (See Method of analysis on sheet 3.) I think the PCA tables make the same assumption. I think that stepping the thickness of the walls violates that assumption. I would add an extra factor of safety if using the tables.

 

[JedClampett]

Back to your question about six posts ago; The reason that using the corner forces from a/b=1.5 is a good approximation is that with that aspect ratio, the interior of the wall is not affected very much by the corners (and vice versa), similar to your case. We commonly do this in our place.
I did a tank a while back that was sized 400 ft. by 88 ft. by 20 ft. deep. The walls were tapered from 32 inches to 21 inches. I used a cantilever approach plus analyzing the corners with the Bureau of Reclamation tables. I needed #9's at 6 inches horizontals at the corners, with reinforcing actually reducing toward the base.
I wouldn't be shy about trusting the numbers and providing the thicknesses and reinforcing as above. You'll only get one shot at this, and you don't want to lose sleep over it.

 

[ARLORD]

JD,

I see that you tapered your wall. I am stepping the wall because I think it is easier to construct. Do you see any disadvantages to stepping over tapering.

 

[JedClampett]

Technically neither one really meets the assumptions of the standard tools (PCA or Bureau of Reclamation), but I think if you're conservative with your approach you'll be fine. I've used stepped walls. If you want to be more precise, you can model the walls in STAAD or some other finite element program. But that's a lot of work.
One other point to consider; You're imposing construction joints at every step. You'll have to dowel in reinforcing to meet the new wall thickness and put in waterstop. With a tapered wall, the contactor can build gang forms, pour the wall full height and re-use the forms. And you have a lot of wall. Even though they complain about tapered walls, they do have advantages.

 

[cntw1953]

Is this an enclosed tank (with roof cover), or open top reservoir?

 

[ARLORD]

Open tank, no roof.

 

[cntw1953]

If that is possible, you may provide a ring beam on top of the tank to increase out of plan stifness of the wall (preferably the beam width is wider than the top width of wall), thus, reduce the upper corner forces. I am still thinking a 3D plate model is best choice for your situation.

 

[Splitrings]

Any consideration to a round tank? Much more efficient design without the stress concentrations. We have been building round tanks from 75' - 200' in diameter for a number of years. Only to heights of 12'. Contractors have also been using their conventional forms with the additon of a spacer on the outside every so often.

 

[RMunoz1361]

ARLORD,

You probably remember reading my post on a previous thread that you started, if not here it is with somemore info in it:

I designed an aeration basin matching these dimensions (146-ft long x 50-ft width x 27-ft height). Since the L/H was greater the 2, for simplicity the wall will perform as a cantilever member, unless I restrict the top of wall. The basin had an immediate wall running parallel with the 146-ft, so placed a 6-ft wide walkway (strut) cross all three walls, at the center of the structure (T'ing across the 50-ft dimension) to brace the wall at top. I also used a 6-ft wide (acting as a beam framing into it strut) walkway around all perimeter walls. If did not do this the required flexure strength would be like #10 at 6-inches on center, to much to close at the base. The end result was a 30-inch thick wall at the bass with #9 dowels at 6-inches on center with vertical steel at 12-inches on center. I stepped the wall down to 20-inches just past mid height. I sized the wall base (thickness) for the shear load without the additional environmental load factor (concrete only, I did not use any steel for shear capacity) and the flexure steel as required by ACI 530 with the environmental load factor. Be careful with the expansion and construction joints locations, I think I used 1-EJ and 3-CJs.

Make sure you follow the requirements in ACI 530 for temperature reinforcement. And double up on the temperature reinforcement as the base of the wall for at least 4D to 6D (height) to prevent vertical cracking as the wall cures. The foundation for the basin was a 3-ft thick mat with piles at 10-ft on center each way.

Just a reminder: The design of the crack control reinforcement in both directions is based on ACI 530 with unfactored loads.

What I worked on is not the same scenario as yours, but you could always create (3) walks crossing the 100-ft dimension, support them at (3) locations to provide yourself a modification of what I did and you must use STAAD to simplify analysis with the walkways. This should reduce you moment requirement at the base, but you will still probably be at a 32 to 34-inch wall.

Hope this helps.

RMunoz III

 

[ARLORD]

RMunoz,

I don't have control to add walkways. I have to live with the free top. I have #11@6 on the interior and #10@6 on the exterior, vert bars at the base.

I am planning to use 1 Expansion Joint and Construction Joints at 40ft oc. Did you continue the expansion Joint from the wall thru the mat foundation. Did you stop the reinforcing at your Expansion Joint. I am assuming CJ stand for Control Joint. Di you stop the reinforcing at that joint as well.

 

[ARLORD]

RMunoz,

I see CJ is construction Joint. I think you wanted to type ACI 350 and not 530 for the crack control design.

 

[ARLORD]

RMunoz,

Also, you mentioned crack control in both directions. Did you have bending in both directions. Did you use the actual calculated fs (aci 350 10.6.4.6), or did you use fs,max(aci 350 equation 10-4, section 10.6.4.1. aci 350-06.