Horizontally offset concrete columns 7

MSL, I know you have already thought about a lot of these I am sure but my biggest comments here are:

1. The eccentricity of the load going into the column below, especially in case 4. Just designing the column to take most of the column above with the eccentricity along the weak axes gives me the eebie jeebies.

2. This will end up more of a hinge than a splice, imo. Detailing the rebar correctly and having it installed correctly seem suspect to get a true fixed-fixed splice.

3. For the punching shear check, it seems to me that you would have half of the upper column very-very rigidly supported and the other half would have a very little "spring" under it. This could lead to a "ponding" effect where your M/S aspect of the column load coming down could get some second order effects. this is where the punching shear could become a huge issue as it gets very concentrated in one spot. It depends on your loads and scale of your columns if this would be an actual issue or not.

1. Eccentricity - yes, I'll make sure the column is designed for that. I'm using P x e.

2. I think the splice would be in the shared part of the column, i.e. something like a 12"x12" area. For the rest of it, I'll have a hard time justifying a hooked development length into the slab. It won't be a proper moment splice by transferring it through the slab.

3. If I consider the punching shear from the column itself, there's pretty much no way to make it work. It'll punch through the slab like paper, especially since it will be carrying loads from slabs above. But I'm thinking that the slab above and below will provide rotational restraint through diaphragm action, preventing the column from rotation. Where I'm getting stuck is how much punching shear to take, because I think the real amount will be somewhere between these two extremes.

If you can't put numbers to it and don't feel the best, have you thought about detailing some corbels? It shouldn't impede on any floor space and I am sure it would make the construction workers (and anyone that goes back and reviews your work later) feel a lot more safe and as if you did your due diligence.

Now that I think about it, I would provide a corbel regardless just as a sanity check. Pennies on the dollar and would help me rest well at night.

It may be a matter of making the transition over a couple of stories. With one building I did in Winnipeg, they wanted to move one of the main building columns over several feet and using high strength concrete. This was done over 3 or 4 stories (don't recall the exact values). The concrete was 8 or 10 Ksi this was high 30 years ago.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

We like to add beams or drop panels whenever we have situations like this, especially #4. Punching shear is never something we want to mess around with.

@WesternJeb That makes sense, I think I'd add a drop panel instead to get more punching shear resistance but a corbel seems like a good solution as well. The advantage of a corbel is that you're making sure any eccentric forces, if they exist, would be dumped into the column below. I'm just afraid it doesn't do anything for punching shear, so I might make a corbel inside a drop panel. Thanks for the suggestion.

@dik Unfortunately for my building, it's short (5 stories) and the transition happens suddenly. But I'll keep it in mind if I need to do it in the future.

@jerseyshore Initially I had beams, but the architect sent me samples where there's a sudden transition without beams or anything, from which I got the cases in the picture above. I was arguing that it doesn't make sense, but he said if another engineer is doing it, I can too. I bet if I talked to this engineer (I know him), he wouldn't be able to justify why he does it; he's average at best. It puts me in a tricky position because this is a huge client and I don't want them saying I'm a bad engineer, but doing what they ask makes me feel uncomfortable. Drop panels might be the solution here. I'm still faced with the issue of how much punching shear to consider. It's like I'd be adding a drop panel to be safe, but for an unspecified load.


Calling KootK and bookowski, y'all were involved in the previous threads. Do you have a definitive way to calculate the punching shear in these cases?

Milk-shake-lake, you could look at it as mechanical engineer.
small toy where their head twisted !!! what keep it standing and strong.
pin-mechanism, where you could add (+) structural plated section inside the edge of bot column and extended to the above column
then add some close spaced stirrups (for lateral confinement). and you will get highly efficient joint interim of strength vertically and laterally.

If you add a corbel then to me you get more punching shear area, which helps alleviate a little of the issue. You also take away the excess of load from the side of the column.. so it is just the true slab load that you would design punching shear for with a little odd geometry from the corbel.

Am I misunderstanding?

eww.
try pushing the architect to align the columns. they look like only very small variations of location so why can't they align and deal with it architecturally?
not only do you now have to justify this, but it introduces complications in the constructability, detailing, it adds congestion, and so on. i would hate to review the rebar shop drawings for these.
I'm thinking simplicity is the best answer.

Bearing stress on the overlapped portion need to be checked (look up Opal Tower in Sydney to see a failure!).

If bearing is ok, how can it punch? The load is transferred directly from the column above to the column below.

I will add that you need sufficient depth of slab/beam to develop the bars that do not continue through, remembering that bars do not develop compression around corners, so in most cases some sort of thickening will be required for compression development.

And you will need to allow for load eccentricity in the load effects on the columns.

Design as though column tapers down to the overlap cross section. Put some extra reo in the interior of the cross section near the floor level where the overlap is. Don't need to develop the bars in the slab because you're assuming the force doesn't go that way.

its quit interesting to see that someone (A.E) fault is other gain (S.E)
Above readings indicate that research door (MSc & PhD) is widely open for this problem and needs verified design procedure.

Verify the bearing stress, as mentioned by @rapt. Additionally, assess the punching shear of the slab, considering the heightened shear stress resulting from eccentricity. The moment induced by the eccentricity will be distributed between the columns and the effective width of the slab.

To confirm the additional reinforcement needed in the slab to counteract the shear, the strut-and-tie model can be employed.

The load will disperse at about 30-45 degrees from the vertical, so a bar just outside will still need full development and those further away will need to develop from the dispersion line. So you still will probably need to thicken it, depending on depth and bar diameter.

And you will have to allow for transverse bursting over the dispersion length.

Not sure about the eccentricity moment effect on punching shear as it is not a slab moment transferring into the column, it is by direct eccentricity of the load from above, unless the column above is actually being carried by the slab with a complete offset. A full FEM analysis might show how much of that moment actually gets into the slab connection. But there will be a horizontal thrust from the offset that has to be accounted for.

This is fairly common in our area. I agree that it feels pretty suspect at first glance, but I think the reason that you're stuck is that it isn't actually so bad. Bearing on the overlap (or include the shared steel if in a pinch) + check the column for some moment + look at it like a small walk and make sure the thrusts are ok, and you're good. This should have very little, if any, impact on punching. The moment portion of the punching is the unbalanced moment transfer from slab to column, or vice versa. Any p x e moment here shouldn't be involving the slab. There must be some difference than if the condition didn't occur at all but you'd have to get really into it and granular to discern what that is, and it should be small. As you said, the perimeter could even be considered to be increased if you think the slab can hang off the column above a bit. In terms of pushing back on the arch., the best you can probably do is make sure you get what you need below to make that overlap feel ok to you.

Agree with Bookowski. Draw strut and tie. I reckon there won't be reo outside the overlap so no need to develop like elastic analysis would. The non overlap part only pushes the slab as far as column vert shear which is nothing. Slab bending from column flexing will be in the analysis results already.

Thanks everyone for responding. I spent quite a bit of my headspace over the holidays thinking about this, and my conclusion is to use a concrete corbel. I don't have enough research materials and literature I can assess to justify not adding one (maybe because that research doesn't exist), but also not enough to determine a structurally justifiable punching shear. A corbel is a decent compromise. I will also consider eccentricity of the column above. To address some specific points:

@ALK2415 That's a cool detail, thanks for taking time to sketch it. I believe I can rely on column rebar to achieve the same thing. I'm not particularly worried about lateral shear, which is the main thing it solves. And I agree that there needs to be more research done about this, especially when other engineers are walking the column with seemingly no consideration.

@bobbyboucher I tried asking architect to line it up, didn't work. I pissed them off a little bit. In the end, the architect is the king and structural engineers work under them to a large extent in my area. Part of it is due to how damn expensive the real estate is. It might be easier to line up in a different market, where you can just build more to cover up slight losses of floor area.

@rapt I'm not sure how an FEM analysis can see how much of the eccentricity of the column will go into the punching shear. It will tell me how much will go into the slab. Maybe I can draw a punching shear perimeter and use the slab moments. It will be very complicated and a bit out of my depth.

@bookowski I'm still not sure how much of the column eccentricity will go into the slab, if any, and how to determine that. I'll trust your judgment that it's probably a small effect.

Any difference in the slab moments compared to the case with no column eccentricity will have to transfer into the slab through the slab/column interface, so would affect punching shear.

Or just look at a case with eccentricity as the only loading. Any slab moments in the slab will have been transferred the the slab through the slab/column interface.