Mat Foundation Can Be Thinner, Why Not? 1

[KootK]

Why not?? Because the concept is deeply flawed, that's why. At the least, that's how it has appeared to me over the last 2+ years that I've been contemplating this.

Please refer to the attached article and the sketches shown below. This is an innovation in foundation punching shear design that's been put into practice on some significant buildings.

What am I missing here?

 

[hokie66]

200 x 200 column stubs shown, at least in the 'one way' test setup.

 

[human909]

I think the collar lets the whole system down by creating a slip plane between column and the very foundation it wants to punch through.

With the generoust amounts of horizontal shear reinforcement then it seems to cover this direct shear plane. I don't see this shear plane being less reinforced in the fabricated assembly vs the regular reinforced concrete example. It is just more delineated.

 

[hokie66]

One thing annoying about the paper is the horrendous mixture of units. Forces in kN and tons (the numbers are actually tonnes); stresses in psi, ksi, MPa; lengths in mm and inches. Makes for tricky reading.

 

[KootK]

A contrarian view, if I may.

By all means, I was hoping for some contrarian views and have been starting with the assumption that it is I who is confused about things. It seem almost unfathomable that this could get designed, peer reviewed, tested, fabricated, and installed in an 80 story building in Korea if it actually does have the fatal flaw in it that I see.

So I'll contest you support of this a bit, as I am want to do...

the 'collar' serves to confine the concrete within.

1) Does a column embedded in a reinforced mat slab need extra confining? That, particularly, given that the column should shed it's axial load as it makes its way through the mat?

2) I contend that a square sleeve would make for pretty crappy confinement. As the side walls bow out, they'd be restrained by the concrete and you'd kind of be back to the no sleeve confinement situation. I suppose that you could get some meaningful confinement in the corners kind of like we do with rectangular column ties.

3) Effectiveness aside, how do you see confinement helping with vertical load transfer? Are you pondering, as I have, a situation in which the column dilates into the sleeve and engages it for shear transfer that way? My issues with that are:

a) That strikes me as almost as dubious as counting on friction for vertical load transfer and;

b) If that's the mechanism, why doesn't the paper say anything about it? The story told is STM, start to finish.

...and I think the setup does resemble real world practice.

I disagree, although it's frankly difficult for me to tell just what the test setup was in the various situations. Yes to the 200x200 stub columns, however:

4) The one way setup had the ribs passing beneath the column which is cheating.

5) For the two way setup, the load applying thing seems to be wider than the sleeve which is cheating.

6) More than anything, the transfer of load into discrete supports over a short shear span is big time cheating. At that point, you're just testing a modified pad footing design, not punching shear force transfer into surrounding concrete. One of my best guesses on this is that, despite contrary language in the paper, they really just mean for this to be an enhanced footing design?? But, then, do we actually have an existing problem with footing design that this would be solving? don't see it.

 

[KootK]

With the generoust amounts of horizontal shear reinforcement then it seems to cover this direct shear plane.

I disagree that the horizontal reinforcing here is significant:

1) Firstly, as an aside, know that for much of the past, in North America, we gave no credit to this reinforcement as enhancing punching shear resistance. We do now but we're a bit late to the party.

2) My understanding is that the horizontal reinforcement would serve to merely enhance what is still, fundamentally, a diagonal tension failure mechanism. It doesn't make that mechanism go away.

3) I'm arguing that the governing failure mode for this thing would be a diagonal tension failure over a frustum geometry that would be less than would be available if the collar thing were absent. As such, the presence of the reinforcing would assist both solutions, not favor the collared solution in any way.

I don't see this shear plane being less reinforced in the fabricated assembly vs the regular reinforced concrete example.

I do. On the vertical planes represented by the blue lines in the first sketch below, you've effectively replaced a nice, frictiony, concrete-concrete surface with a relatively slippery steel-concrete surface.

To be honest I'm confused I would be thought that such an arrangement wouldn't work if appropriately designed.

It would work if appropriately designed. However, in my opinion, "appropriately designed" would mean basing it on the punching shear mechanism shown at right below which would be considerably worse than the punching shear mechanism available in the absence on the collar. To the extent that I'm correct about that, it makes the collar a nonsensical thing to do.

 

[MIStructE_IRE]

I agree completely Koot. I still can’t justify this myself. Your punching diagrams are spot on in my view.

I really can’t see how this got past an entire team for an 80 storey building unless I’m completely missing the point here?!

 

[phamENG]

The authors' contact information is right there on the paper. Maybe you should ask them directly?

(I'm as clueless as everyone else - possibly more so as I don't do a ton of concrete design...)

 

[XR250]

I don't do a whole lot of concrete design either but KootK's punching shear argument certainly makes a lot of sense.
Another CitiCorp?

 

[KootK]

The authors' contact information is right there on the paper. Maybe you should ask them directly?

That was attempted, years ago now, and to no avail. I don't lose a whole lot of sleep over ethical chestnuts like this but I did feel it prudent to reach out to the author before going public with this so to speak.

One of my goals with this thread is to use it to validate or, better still, invalidate my concerns after which I may press the author further on the issue. If I contact the author again, I'll likely reference this thread as a means of explaining my concern in greater detail and, if it pans out that way, demonstrating a critical mass of concern from other folks.

 

[phamENG]

XR - I agree, the arguments are certainly compelling.

KootK - I figured as much. You hadn't mentioned it, though, so I figured I'd bring it up.

 

[KootK]

I probably should have led with the genesis of it given that someone was sure to suggest contacting the author. I was concerned that might taint the discussion but, in retrospect, I doubt that would be the case.

 

[Celt83]

certainly seems like the assembly is getting a lot of help from unaccounted for actions.

I agree that the smooth collar actually makes the punching situation worse.

Some things that are helping the situation, that they may just be getting lucky on or actually took into account hard to tell from that paper:
- It's a mat foundation so the soil is pushing back in that region giving you compression struts to the column center line. (wouldn't show up in their test setup since it only had bearing pads at the "wing" ends)
- The interaction between the steel wings and the lower tension ring restraining cross-section curvature within the collar helping to limit the splitting tension that would lead to the inner punch mechanism.
- The criss-crossing tension rods at the base that wind up going into catenary action to support the inner failure cone, funny enough this likely reduces the tension ring demand above.



My Personal Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

Open Source Structural GitHub Group:

https://github.com/open-struct-engineer