Foundation Wall Keyways 1

[EngWade]

Few quick questions for you guys.

Why are keyways typically incorporated into the junction of foundation footings and foundation walls? The only reason I can come up with is for simpler construction. But my intuition tells me that when doing this, you significantly weaken the foundation wall/footing against bending/rotation. Meaning - any moments about the foundation wall/footing will only be transferred by the rebar (if it is specifed) that connects the foundation wall from the foundation footing. If what I am saying is correct, then you should never have keyways in concrete, cast-in-place retaining walls, right?

Thanks in advance.

 

[hokie66]

Michael,
Not sure if I understand. When I use a full width key, the entire footing or slab outside the inside face of the wall is at the same level. So if the footing is 12" thick inside, it is 10" thick from the inside face of the wall outward.

 

[paddingtongreen]

hokie, I meant the footing outside of the wall, using your numbers, 12" thick inside the wall, 10" under the wall, 12" outside the wall. It's the width of the outside part that interests me, it obviously needs to be the same width as the wall thickness to have the same shear area but what intrigues me is that if I rotate the detail 90 degrees, so that the wall is trying to move downward, that outside looks like a corbel off the bottom of a beam and I wonder why it is not subject to similar requirements.

For clarity, I should say that I have a water tank in mind through these musings.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[Ltwine]

pad:

I think you better to provide a sketch, I am curios to know your concerns, but getting lost here.

 

[theonlynamenottaken]

We could discuss the theoretical musings of keyways ad nauseum. When it comes down to it testing would be the only basis for a concrete (pun intended) judgement.

From ACI224.3-08 "Joints in Concrete Construction" (which has just been re-reapproved from 1995's version)
"Shear transfer is provided by shear friction between the old and new concrete, or dowel action in the reinforcement through the joint. Shear keys are usually undesirable (Fintel 1974), since keyways are possible locations for spalling of the concrete. The bond between the old and new concrete, and the reinforcement crossing the joint, are adequate to provide the necessary shear transfer if proper concreting procedures are followed."

I checked the Fintel reference. It's a book, not a test report. Anyone know if there's any testing data out there or have I just found my PhD thesis?

 

[Ltwine]

I am questioning ACI on this blanket statement.
A member build on the other with rotation caused by large bending moment, there is possibility losing partial contact due to tension at hell, though the compression toe will gain extra force to deliver required shear friction, but in a very limit area (C small). I don't know how reliable the shear friction carried under such condition.

Further more, by any chance once debris get into the joint, there may not have any shear friction capability.

I wondering if there is test done on member subjects to both shear, moment, and stressed to near/at yield.

 

[miecz]

with relatively smooth contact surfaces, the shear friction becomes mostly rebar dowel action. I worry that these bars placed near the surface of the concrete will cause spalling. Anybody else worry about this?

I worry about this too. My solution is somewhat convoluted. First, I meet the code by checking the shear friction provisions for the tension steel. This leaves me feeling uneasy, so I provide dowels across the joint at the compression face that are able to take the shear. (I would have provided dowels there anyway, so this is usually just a check) On the compression face, the shear force is away from the face of the wall, so the concrete won't tend to spall.

 

[Ltwine]

The only objective I can accept is to be construction friendly, to minimize mistake/potential weakness created by varies reasons, and in application there is no/little chance to loss contact area, dowel alone can not prevent slip accross gap.

 

[KootK]

JAE: Agreed, a full depth key is better. I always seem to encounter this as ramps (often curving in plan) that need to be keyed into walls. Contractors complain that, using a full depth key, the key is tough to place accurately and you end up with some of the key beyond the depth of the slab. Maybe that's not the end of the world. In the past, I've done these checks:

1) I check slab shear using the usual provisions but assuming that "d" is from the top of the slab to where I reasonably expect the tie reinforcement to land.

2) I check the tie reinforcement for it's ability to resist the tie force generated by some kind of reasonable arch behavior.

3) I check bearing on my assumed arch / comp. strut.

4) Shear friction for sport.

For the particular joint that I sketched, the notion of using two smooth, unkeyed surfaces and dowel/SF only gives me the willies in the worst way.

LTwine: You're quite right, the tension in one part of a SF connected section is compensated for by the added compression elsewhere in the section. It is common practice -- and necessity really -- to take advantage of this. Shear wall design is a prime example.

MIECZ: I'm not so sure that spalling doesn't occur for that case too (pushing the dowels towards the centroid of the cross section). Do an FBD on the rebar on one side of the joint. Any compression against the concrete on one side of the rebar ultimately has to be balanced by compression on the other side. That balancing compression imposed on the rebar is seen by the cover concrete as spalling inducing tension. This is a better situation, to be sure, but I still suspect that spalling could be an issue.

 

[miecz]

Kootenay Kid

I don't see what you're getting at. Maybe a sketch would help.

 

[hokie66]

Paddington,

Just to clarify, the 10" depth in my example goes outside the wall, all the way to the heel. That way, the step is made with just a lump of timber tied to the inside face reinforcement.

The design of the heel may well be like typical corbel design, depending on the loading.

On the shear friction issue, I wouldn't like to use a flush joint at a suspended slab to wall connection. A wall to footing connection is different because of the wall load normal to the joint, but I still prefer bearing rather than clamping.

 

[rowingengineer]

http://www.concretees.com/people/bruce/pubs/C870620.pdf

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