Pinned base in a concrete wall - is it really achievable? 3

[Lion06]

I have a somewhat theoretical question regarding concrete walls as they are a reasonably popular topic of discussion on this forum.
I, as many others, often design concrete foundation walls as pinned-pinned when there is a first floor diaphragm to tie into. I understand people have been doing this forever and I am not questioning the practice. My question is this: Is a pinned base really able to be achieved in any typical detailing that anyone does?
I am thinking that any rebar tying the wall to the footing is providing some moment capacity. I understand the pinned-pinned case is the worst case for the wall, so that is being conservative is you get some end moment capacity out of the foundation to footing connection. Assuming a pinned base is unconservative for the footing since there WILL be some moment at the base. Even if the steel yields and a "plastic hinge" forms at the base it doesn't lose its moment capacity.
Also, there must be some steel at the interface of the footing/wall since they are usually seperate pours and the steel is need for shear.
Any opinions on this would be greatly appreciated.

 

[shin25]

The term "yielding" that you have used is associated with large rotation at that loaction, which dissipates the energy. On top of that the amount of reinforcement that are usually provided at the joints is quite small than the amount that will require the joint to act as fixed. In other words the small amount of moment that may generate due to the placement of the minimum reinforcements at the joint should not change the design assumptions (pinned-pinned).

 

[Lion06]

shin-
I understand that a large rotation is required. I also understand that this is not overly relevant for the wall since the most conservative case for the wall is pinned-pinned (assuming that you are placing the rebar in the center of the wall).
That being said, the rebar you provide at teh interface will provide some moment capacity. Why wouldn't you take that phiMn at ultimate and dump that into the footing? I realize that this phiMn will likely be smaller than that for the wall at midspan, but is a moment nonetheless and should probably be accounted for.
I have never put any numbers to this, so it is possible that it is very small and negligible, but given the way moments affect footing design, shouldn't it at least be checked?
Just thinking out loud here, but if you have #6@12" for the wall reinforcing (not unreasonably small) and #5@18" for the dowels (not unreasonably high), phiMn at the base is about half of that for the wall section. That is not insignificant in my estimation.
Also, given that the wall will act as fixed at the base until the reinforcement yields, it will attract moment. Is that a fair statement?

 

[271828]

I think you're missing a big source of flexibility at the wall base: the footing.

I've never modeled it, but I suspect that the footing will rotate to relieve most of the moment in most cases. It would be pretty easy to test this. Create a model of the wall and footing in your favorite program. Use the subgrade modulus to determine spring stiffnesses to model the soil below the footing. Run it and see how much moment is generated at the base.

Granted, there are exceptions such as pile supported grade beam as the base or a strip footing on solid rock.

 

[connect2]

211828
Or a foundation wall to a rigid mat. zero rotation of mat.

 

[271828]

connect2, I agree. I'm sure there are lots of examples.

Most walls are supported on narrow strip footings supported on plain ole soil. Those are the cases I was typing about.

 

[Lion06]

271828 and connect2-
Please excuse me if I am missing something that is staring me in the face.
If you have a narrow strip footing on plain soil and the footing rotates, how do you automatically assume pinned? The rotation of the footing may cause the bearing pressure to be exceeded, correct? The rotation is caused by the moment at the base of the wall, correct? I am not understanding why that amount of moment can be just neglected.
If you have a rigid mat this is less of an issue because the moment thrown in by one section of wall footing will likely be much less significant with respect to the capacity of the footing than for an isolated strip noted above. That being said, the lack of rotation of the mat doesn't mean zero moment being transferred from the footing to the mat, does it?
All I am saying is this. We obviously have our choice when we begin the design process of designing as a fixed base or not. If we arbitrarily decide to design as not fixed does not necessarily mean it will behave that way. I am trying to understand how you neglect the moment that will be transferred. As I stated in the second post, the base connection will have some moment capacity, that is not debateable in my opinion unless someone can impart something to me that I haven't learned yet (I know there is a ton I haven't and I am always looking to learn, so please pass the wisdom along!). That moment capacity does not go away when the rebar yields and the walls gains curvature (a plastic hinge forms). That plastic hinge maintains that moment capacity and I do not see how it does not get transferred into the footing.

 

[271828]

"If you have a narrow strip footing on plain soil and the footing rotates, how do you automatically assume pinned? "

Design simplification. Because the moment will probably be small and, like you typed earlier, it's on the safe side to assume it's zero.

"The rotation of the footing may cause the bearing pressure to be exceeded, correct?"

Seems unlikely to me because the rotation angle will be extremely small.

"The rotation is caused by the moment at the base of the wall, correct?"

Not really. Imagine the end of a simply supported beam. There's no moment there, but it certainly has some rotation due to curvature along the beam's length.

 

[Lion06]

""The rotation is caused by the moment at the base of the wall, correct?"

Not really. Imagine the end of a simply supported beam. There's no moment there, but it certainly has some rotation due to curvature along the beam's length."

The simply supported beam has no end restraint, a wall with reinforcement does have end restraint.



""If you have a narrow strip footing on plain soil and the footing rotates, how do you automatically assume pinned? "

Design simplification. Because the moment will probably be small and, like you typed earlier, it's on the safe side to assume it's zero."

It is only conservative for the design of the wall. It is UNconservative for the design of the footing.



- I understand there is a lot of design simplification involved in this. What I am trying to understand is whether someone has run any numbers at any point to determine if it is actually negligible. Maybe I should make the question a little more clear, also. While I am talking about the wall/footing assembly, I am more concerned with the moment at the base of the wall going into the footing, not the wall design.

 

[271828]

I've never heard of anybody running such a number. Maybe you can be the first and report back the result!

 

[Atomic25]

StructuralEIT

Need to really hone in on soil spring constants. If the soil deflects enough to create an active pressure on the main vertical span of the wall, it will certainly deflect enough below the basement slab where the passive pressure exists.

If your wall has enough moment capacity to span vertically 10' or 12' to resist the active pressure moment, it will likely have enough reinforcement carried through to the footing to resist the passive pressure moment.

BTW, footing failure is relative to the amount of deflection one can sustain. The building isn't going to crack in half if you exceed the bearing capacity on the toe of a strip footing.

 

[Lion06]

I appreciate everyone's responses with this.
Atomic-
I understand that the building is not going to crack in half, but if you take moment on the footing into account you will get a larger footing. The soil bearing pressure being exceeded may not cause the building to crack in half but it could cause serviceability issues, no?
Again, I am not concerned about the moment in the wall. I am concerned about the moment being transferred into the footing.

 

[DaveAtkins]

I think it gets back to deformation compatibility.

How much does the typical strip wall footing (not widened to take any assumed fixed base moment) need to rotate to create a pin condition at the base of the wall? (Answer--wL^3/24EI). Assuming this rotation, how much bearing pressure is generated under the footing?

DaveAtkins

 

[271828]

Think about what happens when load is shared by two springs, a stiff one and a flexible one. The stiff one gets most of the load.

If you made a model of this, you'd have a wall with a certain EI and L. In effect, you'd have a rotational spring at the bottom from the soil subgrade modulus.

I'm speculating that the spring at the bottom is so flexible for a soil-supported footing that it won't take much moment.

Here's an exercise that might help. Calc the rotation in degrees at the bottom of the wall considering it simply supported. Then calculate how much the corner of the footing dips downward due to this rotation. I think it'll be a tiny number, just guessing something like 0.1" or less. Then try to imagine a soil bearing failure due to the footing corner (only, the rest is less) pushing into the soil that tiny distance. Turning the problem around and thinking about it from a flexibility standpoint instead of a stiffness standpoint.

 

[271828]

Dave, you fiend! You beat me by 5 min.

 

[PMR06]

StructuralEIT correct me if I'm wrong, but you are talking about a typical concrete wall, on a continuous narrow footing, with a slab acting as the "pinned" support? If the slab is there, then whatever moment does or does not pass thru from the wall to the footing is irrelevant. If the dowels don't have the moment capacity, OK. The footing can still take gravity load, and the slab is still there providing the "pin".

If the slab is not there, then yes I would agree that there is a moment across the wall/footing interface to consider. But assuming pinned again, that moment will not be as large as the maximum in the wall, and the dowels will probably be sufficient.

I think the real answer to your question is that if you assume pinned, then there isn't much else you have to do. It's an easy detail, everybody understands it. If you assume fixed, then there is calculation and detailing rigor that must follow to assure you will get fixed behavior. I do not think it would be good engineering practice to try and "massage the numbers" to justify a narrow footing acting as fixed, when everyone else in the world sees as pinned.

 

[shin25]

This takes some amount of energy to put the connection into yielding and cause rotation in the inelastic range. This energy is contributed by the moment generated at the joint by virtue of little bit of steel at that joint.

Now, since the energy from the moment is used up to cause a rotation at the joint,the energy is lost, then theoretically the moment should not transfer into the footing.

 

[Lion06]

OK, let me restate my question with an exaggerated situation so that everyone can clearly understand my question and concern. It seems that many people are missing the point of my question.
The question is regarding the actual bearing pressure of the footing and has nothing to do with the wall except how it interacts with the footing.
Assume a 16" restrained retaining wall. It is restrained at the top by a floor diaphragm and at the bottom (either by a slab for sliding only or the ftg is acceptable against sliding on its own). Let's say we start by assuming the wall is pinned-pinned and the moment requires #9@12" each face. Now, we detail the wall/footing interface to have #5@12" each face for the dowels. There is no dead load on the wall other than self weight.
The footing is sized as gravity only and a very small footing width is required because there is little dead load, and say that the actual bearing pressure is about 95% of allowable (3.8 ksf vs. 4.0 ksf allowable).
Now, we assumed a pinned base, but those #5@12" have a moment capacity of 17 K-ft/ft of wall. This connection is going to act fixed until the base connection reaches it moment capacity at which time that connection will become a plastic hinge and the wall will act as pinned-pinned for any additional load, but that 17 K-ft/ft of wall moment is still there, it doesn't disappear. Once you add that 17K-ft/ft of wall moment to the footing that was just sized for gravity only it is going to exceed the allowable bearing pressure considerably.

PMR-
I understand that everyone in the world sees this as pinned. I am not trying to reinvent the wheel. As I stated in the OP, this is a somewhat theoretical question, but I would like to understand it better.

 

[Atomic25]

EIT, 271828's example is what I'm talking about.

No serviceability issues for something like this.

 

[271828]

"This connection is going to act fixed until the base connection reaches it moment capacity at which time that connection will become a plastic hinge and the wall will act as pinned-pinned for any additional load, but that 17 K-ft/ft of wall moment is still there, it doesn't disappear. Once you add that 17K-ft/ft of wall moment to the footing that was just sized for gravity only it is going to exceed the allowable bearing pressure considerably."

The footing won't have 17 kip-ft/ft unless it's stiff enough compared to the wall EI to attract that much moment. Again, go back to my analogy. You have a wall with a HUGE EI and fairly short L. You also have a teeny, tiny, strip footing supported on soil springs. The springs aren't very stiff and the strip footing is narrow. The bottom of the wall will rotate a little, but its own EI will limit the rotation to be very small.

This is just like having a steel beam with a simple shear connection. A rigorous model of that would have rotational springs at the ends. These springs would be very flexible, so won't pick up much moment for the tiny rotations at the end of the beam. We rightfully ignore moment going to these connections.