Tank/Retaining Wall Shear 2

[miecz]

Article 11.1.3.1 of ACI318 and ACI350 give critical section for concrete shear. This issue came up in a recent thread, thread507-244196, where JedClampett pointed out that, for tanks, the critical section for shear according to Article 11.1.3.1 can be taken at a point "d" from the support for exterior loads only.

Jeds remarks appear to be taken from R11.1.3.1, which clearly states that Figure R11.1.3.1(e) is an example of a condition the shear should be taken at the face of the support.

However, I notice that the Example 5.1 of PCA's "Rectangular Concrete Tanks" states that, for a full tank, shear could be checked at a distance "d" from the base.

So, I wonder if Figure R11.1.3.1(e) is not meant to apply to the base of a tank wall. A tank wall typically has a toe outside the wall, and so does not look exactly like Figure R11.1.3.1(e).

Further, this same load condition applies to any cantilever retaining wall. But my concrete text by Wang and Salmon allows shear to be taken at a distance "d" from the base of retaining walls. Can it be that the toe is considered sufficient to induce compression on the face opposite the load? Or can it be that the joint of a tank acts differently than the joint shown in Fig. R11.1.3.1(a)? Notice that Fig. R11.1.3.1(a) has the steel on the opposite face from the load in tension, while, for a wall/base connection, the steel on that face is in compression.

 

[JedClampett]

Well, I for one, think Jed is right.
Signed, Jed

 

[EQguys]

Miecz:

As I understand, the rational behind the "d" distance from the face of the support is that, if the structural element fails under shear (brittle failure), the failure plane is usually on an average (through experimental research) a distance of "d" from the face of the support (irrespective of the kind of support).

Hence, for a tank/retaining wall, the stem needs to be checked for shear at a distance of "d" from the top of the base slab. In addition, the base slab needs to be checked for shear at a distance of "d" from the inner edge of the stem AND a distance "d" from the outer edge of the stem, since the base slab overhangs from the stem.

In addition, if the tank/retaining wall is supported on piles, the base slab needs a check for single shear and double shear at the location of a pile with maximum reaction AND/OR least failure cone area. I hope this helps.

"Does the man make the journey or does the journey make the man" - Mark Twain

 

[miecz]

Jed,

Well, apparently, no one disagrees...or agrees...or cares. Let me say this. I've been reading your posts for years now, and they are always most informative. Anything you write carries a lot of weight with me. However, what you wrote here, while clearly following the code, appears to fly in the face of current practice. Aren't cantilever retaining walls normally designed for shear at "d" from the base? Isn't that the same condition as the base of a tank with internal pressure?

 

[miecz]

EQguys-

Hence, for a tank/retaining wall, the stem needs to be checked for shear at a distance of "d" from the top of the base slab

But doesn't that fly in the face of ACI Article R11.1.3.1?

 

[EQguys]

Miecz:

Consider the critical load for the stem. If you are looking at horizontal shear due to external earth pressure ONLY, then as per Article R11.1.3.1, the load case will be corresponding to figure(c) where the base slab is the supporting element for the stem. Now taking the stem horizontal shear due to internal water loads ONLY, its again figure(c).

However, depending on your construction condition, you need to decide between the above two critical load cases for the stem. Now, if you are considering both horizontal earth pressure and internal water pressure for the tank, obviously there is a negation of the above two forces, which tends to make this case not so critical.

Now, coming to figure(e) of article R11.1.3.1, the base slab is not a true representative of this figure, since the bearing pressure from the soil is not modeled. This bearing pressure results in a milder shear stress distribution across the section, which makes the critical section at a distance "d".

"Does the man make the journey or does the journey make the man" - Mark Twain

 

[miecz]

EQguys

Now, coming to figure(e) of article R11.1.3.1, the base slab is not a true representative of this figure, since the bearing pressure from the soil is not modeled.

We test the tank for leaks with no backfill, so the soil is not present for this load case.

Now taking the stem horizontal shear due to internal water loads ONLY, its again figure(c).

Ignoring the toe outside the wall, and turniing the book CCW 90[°], it look like figure (e) to me.

 

[EQguys]

I concur on the latter part. The critical section is top of base slab.

"Does the man make the journey or does the journey make the man" - Mark Twain

 

[JedClampett]

I'm not a big advocate of the PCA design of rectangular plates publication. I only use the old Bureau of Reclamation design tables. And like you miecz, we use a case of water without backfill in our designs.
As far as relating tank design to cantilever retaining walls, I think that's good to a point. But if I had a retaining wall with a very minimal or no toe, I wouldn't reduce the shear to "d" above the footer either.
I prefer to take the conservative design approach in this. I think from the discussions above there's really no clear right answer. Water loads are not like an active soil pressure or a floor live load, where you can count on inherent fat in the load itself. Liquid pressures are (unfortunately) extremely predictable. And problems are readily evident. I sleep better for it.

 

[miecz]

I agree there is no clear answer. I really was hoping somewone could clear this up, as I've always checked shear at "d", and now I see,that that violates the letter of the code. Very troubling, as that shear usually dictates the thickness of the wall, and can grow more than 20% over the distance "d". I wish ACI had drawn the failure plane for figure (e). It seems to me that the shear at the face of the wall is held by shear friction.

 

[EQguys]

Another twist to this problem is what if we had a shear key at the stem-base slab junction. Or what if we had an expansion joint for the stem close to the base slab ??

"Does the man make the journey or does the journey make the man" - Mark Twain

 

[JedClampett]

miecz, as far as designs that are done, I wouldn't worry about them. I've seen some scary thin wall designs that are overstressed for shear that have been working fine for many years. I guess the concrete doesn't know the code.

EQguys, as far as shear keys, I consider them either worthless or harmful. I delete them whenever I remember to. Yet there are good designers that swear by them. And if there is a EJ adjacent to a wall/slab joint, it's the same as a free edge and I would design it that way.

If you really want to get scared, check a tank corner design for shear after you take the reduction required by 11.3.2.3. But I digress....

 

[miecz]

EQguys-

The use of shear keys at the bottom of reinforced walls has been discussed in another thread. Opinions varied. I don't use 'em, unless I have to. DOTs like to step the footings, so the toe is thicker than the heel. Maybe this is why.

I can't picture an expansion joint in the stem close to the base slab...

 

[miecz]

JedClampett

I didn't want to bring up the corner, but, now that you have, the tank corner is a location that looks exactly like figure(e). Here, I design for shear right at the interface (not at "d"), and check 11.3.2.3 as well. In fact, I've taken to using additional diagonal bars at the corner.