Partially Connected Column & Punching Shear 2

[Trenno]

Hi all,

I'm trying to find ways of making the below detail work.

I'm using AS3600 currently, but am definitely open to other code's justification. I know we are pretty conservative when it comes to punching shear.

At this stage the Arch. is aiming for "X" to be 200mm, however I can only seem to squeeze 400kN punching shear capacity out of the configuration below.

Does anyone have any suggestions or point me towards material that can help me justify the smallest possible "X" dimension?

Thanks

"Shibby right..."

 

[rapt]

If you are going to base it on one way shear, the whole load is going to have to transfer through the 300mm width of slab at the connection and can only include the reinforcement that is developed into the back of the column, i.e. about 2 bars. The results may actually be worse than the punching shear result with has a longer shear perimeter! And depending on the actual arrangement, there may also be torsion involved!

 

[KootK]

@Rapt: I suggested one way because a) I expect it to be worse b) it can be made to jive with the limited developed top steel that will be available and c) given the geometry, I feel that it's a better representation of expected behaviour.

I would argue that the length of the one way shear failure plane could be taken as b + d at minimum.



I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Trenno]

Ask and you shall receive...

Edge beam granted.

 

[hokie66]

Good decision, and you will be able to sleep now.

 

[KootK]

I do love a happy ending.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Trenno]

The one hand giveth; the other hand taketh away.

They are pushing for precast columns now. Which isn't the end of the world I guess...

 

[rapt]

KootK,

Not sure where you get the +D from, I would be using the width of the connecting face, so b = 300. But you can still only include reinforcement that will develop to the back of the column!

And a win for common sense! As long as the precast columns stop at the bottom of the edge beam (fat chance of them agreeing to that!)!

 

[KootK]

@Rapt: The +D comes from my feeling that this would be analogous to the one way shear situation where a wide beam bears on a narrower column. While North American codes don't have much to say about that, most firms have internal procedures that allow shear to be taken on a width that is often less than the full beam width but also greater than the width of the column. And that makes sense to me.

The limited space for top steel does not play as important a role in the method that I proposed, It would really be bottom steel that would be of interest.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[rapt]

KootK,

The critical shear plane is at the face of the column where the width is 300. That is what you have to design for. The code logic of allowing the first check at d or D from the face of the support does not apply if the section shape changes as it effectively does in this case.

There is negative moment there, so the top reinforcement is critical, not the bottom.

Even if you could wish away the negative moment and assume a pin support (which you cannot), the bottom reinforcement assumed in the shear calculations must develop past that point by at least D + development length! It has to develop fully into the column. You have exactly the same problem.

 

[Lion06]

If this is a project designed to ACI 318, how would you get two cables within the column cage in both directions?

 

[Trenno]

Nope, AS3600.

 

[KootK]

The critical shear plane is at the face of the column where the width is 300. That is what you have to design for.

Not in my neck of the woods. The sketch below illustrates how a lot of design firms in my area handle wide beam one-way shear at narrow columns. And it makes sense to me. When one-way shear failure occurs, a frustum of concrete doesn't pop out with vertical sides. Rather, the effective shearing surface will broaden a bit, consistent with the fact that the shear is strutting into the joint from multiple directions. This is why the punching shear surface is permitted to be wider than the column face after all.

The code logic of allowing the first check at d or D from the face of the support does not apply if the section shape changes as it effectively does in this case

I believe this to be a different animal altogether. That code provision acknowledges that loads applied within "d" of the support probably strut their way to the support directly rather than induce shearing stresses into the incoming flexural member. The "d" that is referenced is a dimension parallel to the direction of the slab/beam span. The "d" that we're talking about here is perpendicular to the beam/slab span.

Even if you could wish away the negative moment and assume a pin support (which you cannot)

The method that I proposed does not ignore negative moment or wish it away. Rather, it explicitly accounts for the over-strength top steel moment that can be developed within the joint and prescribes that it be included in the design of elements where that action is important.

There is negative moment there, so the top reinforcement is critical, not the bottom...the bottom reinforcement assumed in the shear calculations must develop past that point by at least D + development length! It has to develop fully into the column. You have exactly the same problem.

Yes, there would be negative moment. But not much for the reasons that you've stated above (limited space etc). So, for the sake of one way shear assessment, my method would assume a true pin at the connection and reinforce accordingly. In my country's code, that means that designers must develop enough bottom steel into the support such that it it provides an effective tie for the final compression strut that dives from the beam/slab compression block down into the support. It generally take very little steel. And that steel doesn't have to be developed for fy; it's strength can be prorated based on available development length if necessary. Failing that, you get a pass if you can simply provide a hook or head inboard of the support face. The support reaction provides substantial confinement in this situation which make developing the ties less critical than in other strut and tie applications.

In a situation like this, Canadian integrity steel requirements would lead to more than enough developed bottom steel for the one-way shear check. Integrity steel would be two or three 20/25M bars which I would develop within the column using headed bar anchors.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[rapt]

Kootk,

Considering the concrete edge is near the front face of the column, not well past the back face as you have shown, I would still use 300 at the face of the column! Some sensible codes tell you to check to the face of the support if that is a worse case, and that the D from the support rule is only where you have constant conditions into the support.

If there is negative moment at the critical section, then the shear calculations must be based on the top steel in tension, even in your code. If there are different loading conditions that cause situations where there are both negative and positive moment then both must be checked. But you cannot base it on bottom reinforcement if the moment is negative.

You can get 2 or possibly 3 bars into the column. Do you really think that 3 bars will be sufficient Ast in the shear calculations with a slab width of 300 or 600 or whatever you want to assume to provide a shear capacity to carry about 30m2 of 275 thick slab plus the added loads? If the bars are only partly developed, you can only use that percentage of their area in the calculations.

And the horizontal bars need provide the tension force to the back of the column, not just inboard of the support face.

 

[KootK]

@Rapt: I've reached my three volley maximum. We'll have to just agree to disagree on this one. Thanks for the debate.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Trenno]

I think will have to agree with Rapt on this one, I think the logic is more concrete.

However, the real problem is now the connection between the precast column and the edge beam.

I'm thinking recess the top portion of the column such that it creates a nice bearing surface for the edge beam to sit on.

 

[KootK]

That's a very common method for parking structure column to spandrel connections in my area. I've seen it go up with six story precast columns. Pretty efficient.[pre][/pre]

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[hokie66]

Yes, you can just use polystyrene to block out the part of the column where the beam penetrates. Allow the bars to go through. Requires a bit of cleanup before casting the beam.

 

[Trenno]

Yep, sounds good. Thanks.

Now I'm trying to get my head about what punching shear with a spandrel looks like (see sketch below). I'm wondering if you need to reduce the extends of the critical shear perimeter from dom/2 to something more aligns with the crack angle shown. However I can see in Figure 9.2.1 (B) AS3600, that the shear perimeter for my situation does in fact extend past the width of the beam.

Now I'm also thinking that by introducing a spandrel beam, we are now looking at mostly just a one way beam shear problem. As load will be attracted to the edge beams, then through the beams to the columns?

 

[Trenno]

I've also checked the bearing surface for crushing and it is fine ^^^

 

[rapt]

I hope you have some really good column ties over the depth of the beam connection, and especially over the slab depth, to transfer the forces to the column, just so that it does not tear itself apart!