Punching shear at corner column with odd drop panel 3

[Lion06]

I have a corner column in 2-way flat plate construction, that has an odd drop cap condition. This is my first time doing flat plate, so this is my first real world experience checking things like punching shear for the slabs.

I have a corner condition that has a drop cap in just one of the four quadrants around the column (see attached sketch).

The software I'm using is giving no benefit to the punching shear capacity from this portion of a drop cap. That doesn't seem to make sense to me, but I'm not quite sure how to approach checking this.

ACI and PCA Notes don't really give any guidance (that I can find) on how to handle a drop panel like this. I can't find anything good in CRSI, either. What's really kind of throwing me for a loop is that the software is not only neglecting the drop cap, but treating the critical section like it has only 3 sides. I don't see anything in ACI that would limit me to 3 sides on the critical section.

Can anyone see what I'm missing?

 

[Ron]

Is this an existing condition or did you design it this way? Is there any logical reason for the drop panel to be offset? Is there a clearance issue? Is there a higher loading in that quadrant with no need for a drop panel otherwise? Is the column misplaced? Odd.

 

[hokie66]

I think the drop panel arranged that way does make sense, as the moments will cause shear to be larger on those two sides. You definitely have 4 sides...possibly there is a problem with the inputs. May be worth a chat to the software techs. RAPT would handle it, but I don't suppose that is what you are using.

 

[Lion06]

The drop is that way because of architectural requirements. I inherited this project and I'm doing some spot checks.

The loading is uniform. All columns have drop caps, and this is the only location with this unique drop.

I'll talk to the software rep tomorrow. No, we're not using RAPT.

 

[csd72]

Lion06 (AKAStructalEIT,AKAPAStructuralPE),

Surely it would be easier and safer to check this by hand, it is not that difficult.

I am always of the opinion that anything a bit radical like this should be checked by two independent methods.

 

[ron9876]

I agree with csd72. I recommend not depending on a computer program to do something that you can't do by hand. How could you have a feel for the accuracy of the results?

Don't think a condition like this is addressed by ACI. Especially if you are using the equivalent frame method. The capacity would have to be greater than if it wasn't there. I would determine the stress level without the drop and see how close it is and use judgement from there.

A similiar condition that comes up is a top step at balconies. Do you use an average "d" or do you use different depths, calculate c.g. and then calculate properties? In that case I have used the more detailed calc and checked it with the average.

 

[msquared48]

Lion06:

Although it may be an architectural requirement, Hokie's comment is not without merit. Don't ignore the possibility.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[Lion06]

I can do a punching shear check by hand, but I wanted to figure out if there was a reason that the software was ignoring the drop in one quadrant.

 

[a2mfk]

I have only limited flat plate experience, but I like Hokie's point. I assume this corner is a cantilever, and like Hokie said the shear forces will be highest on those two sides. With that drop panel arrangement, I see no benefit coming on those two sides, its just a column. So you have no drop panel in the most critical shear zone, which is giving you no benefit for the controlling sections of the shear. I'd have to think about it more what the other two sides are doing for you in this arrangement, but I can see why you may be getting those results from the software...

I don't think this is your standard punching shear problem with this configuration.

 

[hokie66]

a2mfk,
The shear will be greatest on the inside, not the cantilever sides. How much greater is the problem. The partial drop will help to equalize the stress, but will also draw more shear to that side.

 

[rapt]

I am not even sure if RAPT would handle that correctly. It would pick up the different effective depths on the different faces. But I would have to check how it then applied that with the ACI code polar moment of inertia logic. The logic is always that a drop panel is on all sides of a column!

I notice that you refer to this as a drop panel and a drop cap in different places. It is one or the other, but in this case I would say neither.

Not sure whether I agree with you either Hokie. In terms of slab moments, yes, the drop panel is where the larger shears are coming from, but the column moment will probably be controlled by the cantilever.
But thinking of purely punching shear effects, you do not normally differentiate which sides loads come from. There is a shear head transferring a reaction and a moment into the column. Using the ACI polar moment of inertia methodology, you will get much larger shear stresses on the outside faces of the column and they will control the design.

 

[msquared48]

The shear in the slab at the backside of the column will be the simple span shear plus Mcant/Lbackspan. It should be greater than the cantilever shear, depending on the length of the backspan.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[hokie66]

Haven't done it recently, but surely the ACI Code has provisions for checking shear at a corner column. What if we analyzed the plate without the cantilever, taking all the shear as for a purely corner column, then check the cantilever separately for shear? Then distribute the varying shear stresses along the critical plane. Just thinking...

I would first check to see if your slab works without any drop, and if it does, then the drop can only help.

 

[rapt]

Kokie,

Yes, ACOI code covers corner columns.

For punching shear at this column, ignoring the cantilever will result in a lower axial load and higher moment transfer, probably giving higher shear stresses on a much reduced shear perimiter. The result is not realistic. The column is not a corner column by definition because it has a shear perimiter on all 4 sides.

With punching shear you need to consider the whole thing working together.

 

[hokie66]

OK, that's not a good approach.

Lion06, did you check to see if the slab needs a drop at all?

 

[Lion06]

I did. I checked it by hand with no drop and a 4-sided failure perimeter for shear and moment and it's not close to working.

I know there will be some benefit from the drop, but I'm convinced it won't get me over the hump just because of the amount it's failing by.

I've designed stud rails for this condition.

Thanks for the help, guys.

 

[ishvaaag]

To this kind of problem I remind a trick that is working when using CYPECAD. If you find that the thickness is not enough on punching shear, you include embedded beams in cross shape over the column; then marvelously the thickness complies with the code!

The reason for this must be that the provisions for shear within beams seem to be somewhat less stringent (on an available area for shear) most surely because it is thought it is better known their behaviour, and the shear reinforcing scheme way more reliable.

In any case since the program is able to treat precise areas in areas and elevations, it would give you either a reinforcement or a note on the section being insufficient.

The question of effects of relative position in height of parts of one structure is a quite interesting thing and not a well trodden one; in general, if other restrictions help, at the service level the things seem to work as if aligned in the plane and so on; think of rotationally closely restricted channels, for example. CYPECAD for example makes a very good work of extending the (longitudinal) reinforcement in stepped levels, other thing is how successful the assumptions within the program are for a proper portrait of what leads to the shear reinforcing scheme.

For a critical case of difficult geometry, and because available, I would make a 3D FEM analysis; it would give me by inspection the maximum stress at compressive struts, and the maximum (principal) stresses orthogonal to them; so one can devise "shear" inclined rebar or struts able to deal with such tensile stresses. Apart from restricting the compressive stress to some level, reinforced concrete is exactly that everywhere: provide along the tensile stresses enough capacity to forestall the excessive aperture of the cracks, if any. One might need to examine to what stress such steel is let to work to restrain the width of the transverse crack; if low, it will be small.

 

[ishvaaag]

errata

where "or struts" read "or stirrrups"

 

[csd72]

Not sure why this cannot be analysed by hand from first principles.

Calculate the shear perimeter.
Calculate the average stree due to total slab reaction
Then calculate the 'triangular' stress due to the bending moment.

This will give you a maximum shear stress on each side that can be checked using the basic slab shear formula.

Two sides will be checked with the deeper depth and two sides with the shallower depth.

 

[Lion06]

csd-
The failure perimeter (if you include the drop) lends itself to a more time consuming calc of section properties for the polar moment of inertia for the moment contribution to punching shear with my belief being that it won't get me out of the woods anyway.

Saving a stud rail or two is not worth the time it would take me to justify it and I feel more comfortable ignoring the benefit (I'll just call it "belts and suspenders"). A punching shear failure is no joke when I'm at a stress ratio of 1.89 ignoring the drop, well..... I want to be able to sleep well at night.