Offset Concrete Columns

[CURVEB]

I've read several posts on the design checks for 2 conditions that are sometimes seen in concrete flat plate construction: walking columns and transfer slabs.

My understanding is that walking columns use a minimum of 1 floor of full overlap between the columns above and below (like a short wall). The eccentricity is taken out by the floor slabs and transferred to the primary lateral resisting system.

Transfer slabs are used when columns do not align above and below a floor. In this scenario, all of the force has to be carried by the slab in bending and shear. Checks would also include an increased punching shear stress, although there doesn't seem to be a great deal of consensus in exactly how this load path would work.

My question is this: there is a 3rd scenario where columns above and below a slab do not align, but only be a small amount. Say they overlap by 50% or more.
In this situation, would you consider that the slab does not carry any of the column load in bending or shear? Does the entire axial force transfer by bearing in the area of overlap between the 2 columns (perhaps with a small increase considering the slab may help engage more of the lower column)? I still think that in this case the eccentricity is taken out by the floor diaphragms, similar to the first situation I described above. There is probably some additional bending moment applied to the lower column due to the offset of the axial force from the column's geometric axis.

I'm curious what other engineers feel are appropriate checks to make.

 

[rapt]

Curveb,

Your 3rd case is really a similar situation to your 1st, except that because there is a partial overlap you do not have the intermediate transfer column.

In the 1st, there is partial overlap between your column above and the "intermediate transferring column" below it. Then there is partial overlap between the "intermediate transferring column" and the column below it.

In your 3rd case,

- the slab thickness needs to be sufficient to develop the bars that are not continuous from above and below.
- the column compression will be much higher due to the reduced overlapping area in each column.
- in 1 and 3, you need to actually design the diaphragm action, not just assume it works.

everything else you have said is ok.

 

[KootK]

My thoughts on the issue are well summarized here: Link. Not sure if that's one of the threads that you've already reviewed.

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.

 

[CURVEB]

Thanks Koot. That is one of the posts I had read through. I thought maybe this is a slightly different situation since the 2 columns are actually stacked.

Rapt:
A couple of thoughts:
- the slab thickness needs to be sufficient to develop the bars that are not continuous from above and below. If the transfer mechanism is direct bearing and you aren't relying on the steel right at the overlap, is this still a requirement? Further, if you transfer the bar force into the slab doesn't that put a really high bending load in the slab itself?
- the column compression will be much higher due to the reduced overlapping area in each column. By this do you mean that the small area of overlap needs to be designed as a column section? If so, how much of the slab outside of the column would you consider effective (if any)?
- in 1 and 3, you need to actually design the diaphragm action, not just assume it works. Agreed. I was thinking some additional bars hair-pinned around my column cage and extended to lap with my slab reinforcement beyond the column.

I've attached a sketch of the column in case that clarifies what I'm asking about.

Thanks for the feedback.

 

[jayrod12]

I'd be tempted to make the argument that the load transfers through an area equal to the overlap plus the slab thickness. Depending on the thickness of the slab, you may then get full load bearing area from column to column.

 

[KootK]

Thanks Koot. That is one of the posts I had read through. I thought maybe this is a slightly different situation since the 2 columns are actually stacked.

I believe that we were discussing exactly your case (sketch below taken from that thread), in intricate detail. Pretty sure we hit upon most of your specific concerns as well. Specifically:

In this situation, would you consider that the slab does not carry any of the column load in bending or shear?

- for shear, consensus seemed to be that there was some shear impact but not much.
- for moment, consensus seemed to be that there was some impact but it wouldn't be critical because of the ductility of that mechanism.

Does the entire axial force transfer by bearing in the area of overlap between the 2 columns (perhaps with a small increase considering the slab may help engage more of the lower column)?

Certainly, stiffness compatibility would usually lead one in that direction.

I still think that in this case the eccentricity is taken out by the floor diaphragms, similar to the first situation I described above. There is probably some additional bending moment applied to the lower column due to the offset of the axial force from the column's geometric axis.

As broached in the other thread, many people deal with the eccentricity via slab moments. I disagree with that as I believe that the columns are the stiffer path for the moment that arises from eccentrically applied column axial loads. Consider:

1) Columns are often uncracked,
2) You've got the whole "moment leakage" thing going on at the joint just as with gamma and regular slab moments.
3) The stiffness of the joint to transfer moment from column to slab is pretty dubious in these situations.

I also feel that there is moment, and additional shear, in all of the walked columns, not just the lower one.

- the slab thickness needs to be sufficient to develop the bars that are not continuous from above and below. If the transfer mechanism is direct bearing and you aren't relying on the steel right at the overlap, is this still a requirement? Further, if you transfer the bar force into the slab doesn't that put a really high bending load in the slab itself?

That we didn't cover in the other thread. And it's a tricky issue. My thoughts:

1) I would guess that most of the load travels to the overlap regions as a result of stiffness compatibility. So probably not much load in the bars that die in the slab. Pretty tough to guarantee no load though.

2) From the perspective of the column axial load only, it's tempting to terminate the bars outside the overlap at the top of the slab. That way it wouldn't generate a compression anchorage failure in the slab. I don't do this however. It just looks too weird and would impact regular slab/column moment and shear transfer. Instead, I terminate the bars in the slab with a standard hook which seems to be a ubiquitous detail.

3) In my market, the slab is rarely of a thickness that it would allow one to develop any typically sized column rebar. At least not not without a drop panel which tends to mess up nice, flat, form work friendly soffits.

Agreed. I was thinking some additional bars hair-pinned around my column cage and extended to lap with my slab reinforcement beyond the column.

As mentioned in the other thread, I think that it's also prudent to give some consideration to strain in those drag struts, all the way back to the lateral resisting system. The more the strut stretches, the more punching shear you potentially add to your slabs. Good sketch of the concept in the other thread.

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.

 

[CURVEB]

Koot - you're correct. It's the same issue - sorry I had too many web tabs open and confused myself as to which link you had referenced.

These answers are pretty much what I would have expected, however I have one final thought/question. In the area of bearing overlap between the high and low columns, I think that you can look at it in 2 ways: a direct bearing transfer or treat it as a short column. For example:
Say Pu = 1000k
Columns above and below are 14x28. Area of overlap is 14x14
Slab strength - 6 ksi

Case 1 - Direct Bearing: phi*Bn = 0.65*0.85*f'c*A1 = 649k (assuming no increase in bearing area due to the slab). Assuming a sqrt(A2/A1) increase, and I would maybe consider half of the slab depth as the maximum extents of the increase, an 8" slab would give an A2 of 484in2, so the new phi*Bn = 1.57 * 649 = 1018k.

Case 2 - Use the amount of steel available through the overlap (basically 1/2 of the steel in the column above or below) and design the 14x14 section as a column. I expect with the added area of steel this would work OK, and since it is at an overlap the steel is fully developed for compression above and below.

I could probably convince myself that either of these transfer mechanisms would work, but is either more or less valid than the other, in your opinion?

Thank you.

 

[KootK]

I feel that both approaches are valid. I usually just go with the column approach and make a typical detail of it for the project.

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.