Question on punching shear

[Lion06]

Let's say we have a really thick slab with a column on it. It is a structural slab, not a S-O-G. Let's say the column has uplift on it. The rebar is fully developed into the slab. Is there still concern of a punching shear issue for this situation? If so, would the critical perimeter be measured at d/2 to the INSIDE of the column, and not the outside?
I haven't convinved myself that it is even an issue yet, but I'm getting a little resistance to my thinking. My thinking is this. The 45 degree shear plane would have to radiate toward the center of the column, not away from it as is typical for punching shear. The reason for this is the uplift and the direction of principle tensile stresses. Because of the direction that the shear plane wants to go, it will be crossing the tension rebar, which, in my thinking, negates that failure.
I guess my thinking is similar to this. Say you have a concrete beam with a stirrup sticking out of the bottom and you pull on that stirrup with a significant force. The shear crack wants to start right at the stirrup because the shear is highest just to either side. The stirrup which is delivering the load is also acting as the shear reinforcement. The load can never be larger than the bar can support in tension, and, by default, the shear is not an issue. I guess the only difference is the difference in phi factors - tension is 0.9, shear is 0.75.

Any thoughts?

 

[abusementpark]

"Because of the direction that the shear plane wants to go, it will be crossing the tension rebar, which, in my thinking, negates that failure.I guess my thinking is similar to this."


StructuralEIT,

I have has this same question about general shear in a concrete beam. Why do we only consider the steel from the stirrups in resisting the shear and not the longitudinal shear as well? My thinking is that it would have to shear through the longitudinal bars as well to produce a full shear failure of the cross-section.

However, the answer I have been given is that we don't consider the longitudinal bars because they are not in the shear failure plane, which makes sense and I have accepted. Essentially the failure that we are checking is a shear failure between the longitudinal bars, which is usually 80% or so of the cross-section. To me, this is certainly enough to be consider a shear failure, even though we have not completely sheared through the entire cross-section.

 

[Lion06]

It's actually a 3'-0" hydraulic slab supported by caissons. There is roughly 30' of head at the design water level. The uplift is in the slab putting the caissons in tension. I just flipped it and put the column in tension for talking purposes.

I'm having a hard time understanding

1.) Why we would check a breakout cone - I've never seen this done for rebar, you just check development length.

2.) How to do it even if you wanted to - App. D does not apply to rebar.

 

[Lion06]

abusementpark-

regarding the beam example, I agree, but I think one difference is that in a beam the longitudinal steel is perpendicular to the shear force, therefore you only have to pop out the 1.5" of cover on the bottom of the beam to render it useless for shear. That's not true of the reinforcement that is oriented in the same direction of the shear force - it's in tension.

Also, the longitudinal bars are considered somewhat in the Vc calc - that accounts for the shear strength of the concrete, aggregate interlock, and doweling action of the longitudinal bars.

 

[kslee1000]

All I can say is code has something left out for varies reasons. When special situation/application arise, we have to provide judgement using all tools/concepts avaliable to us to get job done. Good luck.

 

[BAretired]

StructuralEIT,

Maybe you are right. If you have developed the vertical bars within the 3' slab, perhaps it is not necessary to consider the stress pyramid for the group of bars. I don't know as I don't recall ever checking it. Because of the unusual force involved, why not check it and if it turns out to be a problem, either increase slab thickness as required or add a base plate welded to the steel.

BA

 

[civilperson]

Th "d" is measured from the tension steel to the compression face of the slab. For uplift, the tension is near the bottom or the connection of the column to the slab. Shear is the same when loads are reversed.

 

[msquared48]

SEIT:

I still ndo not think that this would give a truncated cone, but the sure way to get that is to weld the column rebar to a metal plate placed at the bottom of the slab. I have used this detail many times in the past to develop loads of 50 kips or so for multi story building overturning forces seen at a PT slab.

If you need a greater pullout capacity from the concrete, just increase the size of the plate and add more rebar.

Mike McCann
MMC Engineering

 

[BAretired]

Right Mike, and if you still have problems developing the force, you could use a shearhead with steel beams crossing the column area in both directions.

BA

 

[msquared48]

Another point to consider here...

Just because you have extended straight bars into the 3' slab to develop the bar does not mean that the tension failure plane will be at the tip of the embedded rebar. More likely than not, it will be at some point along the length of the embedement with the lower portion of the bar pulling out of the concrete. Consequently, the failure plane area, and tension capacity, is much less with this scenario than the tip failure one. This is the reason why we put bends on bars - to develop the full length of the vertical embed. This is the reason I use plates for high loads.

Mike McCann
MMC Engineering

 

[hokie66]

SEIT,

BA and Mike are correct. My previous answers assumed the bars were anchored at the far end, but in reading your original post, I see that you said the bars were "developed". Not the same thing. They have to be anchored in order to emulate the support by a column.

 

[ron9876]

I have considered this before in uplift type slabs. Just not that much uplift. I always felt that the cone is under the column. Whether the reinf can resist the reaction depends on developement at the failure plane. With that magnitude of load I would use something positive to resist the load like the plate that has been proposed.

 

[BAretired]

I spoke to my son about this thread yesterday. He is much more up to date on concrete research than I am. He suggests considering a three dimensional strut and tie model. The bars from the column into the slab form a tension member and must be properly anchored at the end. Compression struts would be positioned around the column in a sort of pyramid fashion. Where the compression struts are trying to break out of the concrete, vertical ties are needed to transfer the compression struts into the compression block of the slab.

Another point which he mentioned was scale effects. When you are dealing with a slab this deep, the usual procedures we use for shear are not very reliable. Changes have been made in the latest Canadian code to take this into account.

BA

 

[hokie66]

BA,

Admirable of you to disclose your source. Must be nice to have a son who can give advice to his old man. I trust he gets plenty also.

 

[BAretired]

hokie,

For about the last fifteen years, my son has been involved with the CSA Committee responsible for the concrete code in Canada, so his knowledge in that area goes much deeper than mine. And yes, there was a time when he got advice from his old man but not so much any more.

BA

 

[msquared48]

Same here. Mine's a PE in Washington, and is much more up to date than me. Experience dies count for something though...sometimes.

Mike McCann
MMC Engineering