Single Curtain Reinf. in 10" Shear Wall

[lutein]

ACI 318 allows the use of single layer wall reinforcing (single curtain) for walls less than or equal to 10" thick. anything thicker than 10" shall have (2) layers.

We did an inspection on a post-tension concrete slab on cast-in-place concrete shear walls structure. the concrete shear walls are the (2) stair cores at each end. they are 10" wall with single layer of reinforcing. we found diagonal cracks along the 2 faces of the shear wall cores from 1st to 5th floor consistently. these cracks were formed during construction on the 6th floor. no exterior envelope walls were constructed during that time, i.e. minimal wind.

In my opinion, these cracks are not due to wind given the heavy reinforcing in the walls and % of wind load the structure was seeing when the cracks formed. The structure is intended to be 16 stories tall, so at 6th floor construction with no envelope walls, the wall reinforcing, per our calculation, was way more than sufficient. Additionally, per our calculation, the overall design was in compliance to the code.

And that's what get us puzzled!

We suspect they are due to combined reasons:
1. the post-tension effect (precompression) of the slab, which caused forced displacement to the walls.
2. single layer of reinforcing which causes the wall to have 5" concrete until the stress hits the steel.
3. with early strength mix of PT slab, the wall strength may still be in 60% range, while the slab is being stressed.

Any input would be greatly appreciated. We are really puzzled by this phenomenon, and are wondering if any of you have seen this, that could share some insights.

Thanks.

 

[Teguci]

Is there a pour joint in the PT slab that separates the two halves of the building?

If not, how is the volume change between the rigid shear walls accounted for?

How long is the building?

Are the stair walls oriented in both directions, with the length, width, etc..?

Love sketches

 

[lutein]

the building is 180' long x 70' wide, with the 2 stairs at each end (160' apart from center of core to center of core).
the overall lateral system is composed of 2 stairs at each end, elevator at middle, and some blade walls in between.

it is obvious that the restraint is from the 2 stair cores.

to answer your question: no, there is not slab joint specified or observed in the slab.

 

[hokie66]

There's your problem. Shortening due partly to PT, but mostly due to shrinkage, means the slab has to move toward the middle. The walls don't want to go along, so something has to give, either the walls or the slab. It happens both ways, but in your case, sounds like the walls are the weak link. Not much to do with whether there is one or two layers of steel. This type movement has to be accommodated by either pour strips or slip joints at the walls which are tied in later.

 

[lutein]

You are exactly right, slab-wall construction should have been composed of 10# felt paper as bond breaker under the slab, with PVC pipe in wall vertical to allow movement from PT precompression. The PVC pipe will be grouted solid when pouring the wall above. Additionally, the slab should have a construction joint somewhere in the middle.

 

[hokie66]

That detail at the wall has problems with construction sequencing and scheduling. If you really want to allow the slab to shrink and tie it in later, supporting the slab on a corbel is preferable.

 

[ron9876]

Are the stair shearwalls C shaped? Are the cracks in the returns or in the main walls? As Teguci said a sketch would be helpfull.

Also it sounds like there may be more shearwalls than needed. A consequence of this is that the walls will be relatively lightly reinforced and more prone to cracking.

 

[Lion06]

lutein-
I don't think that detail works. How are you handling the shear transfer in the wall across the plane with the felt paper in it? The bars can be used for shear friction, but there is no coefficient of friction value for concrete to felt paper.

hokie-
Do you think the PT has more to do with the cracking in the walls than shrinkage of the slab? I would think shrinkage of the slab would manifest itself as cracking in the slab, not cracking in the wall.

 

[hokie66]

Lion,
Both are components of the slab shortening. The elastic shortening and creep due to post-tensioning typically comprises only about 1/3 the shortening due to shrinkage. So it's just a matter of the weaker element cracking first. This is just as much an issue in conventionally reinforced structures as it is in post-tensioned.

 

[Lion06]

hokie-
Agreed. I was thinking that the shearwalls would have much greater capacity in shear than the slab would in tension and the shrinkage would cause cracks in the slab. When I think about the PT, though, it's driving the two walls together and because of the nature of the PT, couldn't manifest itself as cracking in the slab.

 

[hokie66]

The PT itself won't cause the slabs to crack, but PT slabs can crack due to tension caused by shrinkage. These cracks can be quite wide, especially with unbonded strands and with little mild steel.

 

[TXStructural]

When in the sequence of construction were the slab tendons tensioned? Depending on how the PT was used/placed, wouldn't it counter the shrinkage, resulting even more slab shortening?

I agree that the walls are cracked due to slab shortening. It probably manifest as it did because the older, lower slabs were shrinking while there was restraint by the newer, upper slab(s). Or simply because the force required to crack the walls wasn't reached until several floors were pulling on them.

 

[lutein]

Lion06-the wall vert in PVC will be grouted after stressing of pt, along with shear friction bar from slab, that will form the shear transfer mechanism. We adopted the detail from PTI reference book.

TXstructural-you brought up a great point about the restraint from slab below and above. We reinspected the structure after all floors were built, we noticed that the wall cracks only happen near middle third of the height, which kind of match what you said.

With the nature of post tension construction, the precompression is applied at slab edges at each end from tendon anchors. The 2 stair cores that cracked are at each end of slab where the precompression is applied. Hence, the slab will not be in compression until the wall gives.

I wonder a slab construction joint in the middle of slab would help.

I will scan the plan later the day for you all to look at.

 

[Lion06]

lutein-
I'm just not seeing it. Per ACI, there is no shear friction for the case of concrete cast against felt paper. What coefficient of friction would you use? Additionally, bars that are completely perpendicular to the shear force (as the vertical bars would be in this case) have no do no contribute to the shear capacity. In reality, they do, but ACI doesn't recognize the contribution, so how do you define it? Especially since the bars are doing double duty now.
Are you saying that the 3" diameter grout crossing the plane counts as a shear friction surface?

 

[lutein]

Lion06
Shear friction is purely the shear capacity of the bars. Slab will bear on top of wall with dowels into walls. Wall verticals extend from tier below to above slab engaging the slab dowels. No contribution from concrete friction is considered, just like hollow core plank on cmu wall detail, with bent dowels into hollow core planks to achieve diaphragm shear transfer to shear walls.

 

[hokie66]

I think lutein is relying on dowel action, but calling it shear friction. Two different things. Shear friction relies on clamping, and is witchcraft, in my opinion only of course.

A plan of the floor showing the stair cores would help, or some photos. But yes, a slab construction joint would help, but not if construction continued at pace. What you really need is a delayed pour strip, left for as long as possible.

 

[lutein]

Thanks hokie66, dowel action is the correct term.

see attached simplified diagram for confidentiality of the project.
the clouded areas are the cracked region.

as you can see, the stair cores are located at each end of the banded tendon. banded tendons are running plan left-right direction with uniform tendons running plan top-bottom.

 

[hokie66]

With the core in the middle to serve as an anchor, my approach would have been to provide corbels on all of the walls at the ends, with slide bearings and dowels in sleeves to be grouted later. The slab then shortens towards the middle, and the cores are not unduly stressed. I did a parking garage that way a few years ago, and there were no cracking problems that I know of. This approach is not effective without the corbels.

 

[Lion06]

That is my point. Doweling action is not shear friction and ACI gives no capacities for that kind of shear reinforcement.

 

[hokie66]

I don't know about ACI...haven't looked into it. But you can certainly get capacity with dowels. Dowels are not really reinforcement, they are shear connectors.