Checks Required at Concrete Column Transfer Slabs (Walking Columns) 5

[KootK]

Please review the sketch below (and attached). I am curious what others do to check slabs at column transfer locations. I'm thinking of "walking" columns in thinner slabs but the discussion is equally germane to full blown podium style transfer slabs as well.

The options:

1) Individual column punching shear with small moment.
2) Combined column punching shear with large moment.
3) Shear friction in the slab between columns.
4) Bearing stresses.
5) P x e moment in slab.
6) Vertical seismic acceleration effects.
7) One way shear between columns, similar to #3 (forgot to draw this)

I believe that the following checks are fairly standard: 1,4,5,6. Combined punching shear (#2) is a KootK invention as far as I know. Shear friction (#3) is standard practice at one of North Americas top tier structural firms. I feel that it's unnecessary as I can't imagine that ever governing over the diagonal shear failure modes (punching & potentially one way).

So, what do others check? And I'm not looking for FEM solutions. And I know it's awful.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[CELinOttawa]

Sorry about that; Not at all accurate language on my part - I see no imposed (external) moment other than the offset. The column and slab can, of course, have internal moments which I do agree are very hard to quantify.

Progressive rotation of the joint and exacerbated punching shear?

Well now if you're just going to be Mr. Know-it-all and guess correctly every time I bring up something new in a thread, well, well, well than I'm just going to run away and have my Mom make me some coco!

On the serious end of things, the sustained high-shear-load situation at the slab is potentially going to mean aggressive creep conditions - Creep is at its worst when dealing with the "over one-third" stress case. Unfortunately that rule of thumb is from the UK, so it applied to cube strength. It is more like the fifty-precent rule for cylinder strengths... Point being you're likely in the high-stress range where your creep is very bad.

In the case of our slab with serious point load, we also have to worry about the possibility of additional cracking under creep if we load too early... And most construction sites aren't looking to sit around waiting for 28 or more days before building the next floor.

 

[rapt]

The example needs to be more specific than this.

If the column above is relatively close to the column below, it is not a flexure/shear situation, it is a deep beam/strut and tie situation and should be designed and detailed accordingly. In that case, the columns are close together but there is no "combined punching shear".

If the column above is further away, the slab needs to be checked for flexure and shear as it would normally. Plus each column has to be checked individually for punching shear.

 

[RFreund]

I would be thinking something similar to what rapt said. Maybe like a haunch design. Unfortunately I'm no concrete juggernaut.

EIT

http://www.HowToEngineer.com

 

[KootK]

In the case of our slab with serious point load, we also have to worry about the possibility of additional cracking under creep if we load too early... And most construction sites aren't looking to sit around waiting for 28 or more days before building the next floor.

Keen observation. Maybe it would be prudent to run shoring from the center line of the bottom of the upper column to the side of the bottom of the lower column for just this reason. That was a mouthful.

The example needs to be more specific than this.

Not so. I'm seeking generally applicable advice based on general principles. I don't want "I would do X". I want "I would do X in this this situation and Y in that situation". Precisely the kind of advice that you have supplied above.

At a lower level floor slab, where the thickness might be 3', you may indeed get some STM appropriate situations. At a penthouse level transfer it would be pretty rare. There, you've really got two scenarios:

1) Overlap, which is essentially poor man's STM. This was discussed recently at length here:Link
2) A clear separation in excess of 1xD for which STM will no longer be applicable.

The zone of most interest to me is the transition from overlap to where one could confidently say that it truly is just slab flexure and single column punching checks. While I might be out to lunch with my combined punching check, I very much contend that individual punching checks are insufficient. Per my middle left sketch on 11 Dec 14 18:33, I don't see the shear "making its way around" fast enough when columns are tightly spaced. Everything that I know of the logical flow of stresses leads me to believe that model is flawed.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[CELinOttawa]

I completely agree that the shear will not flow around to the far side of the column... This point load is going to have to be handled locally by the concrete and inclusions between the two columns. The far sides, the remaining slab, etc, are all going to feel serious loadings which are secondary to the direct cisaillement of the concrete. I try to make the distinction between a built-up (or distributed) shear and an imposed (or localised) shear, since you can address a built up shear gradually, but must locally reinforce for an imposed shear.

 

[rapt]

KootK,

The shear between the columns is covered by the "beam shear" calculation. It is not a punching shear situation.

 

[CELinOttawa]

Rapt: I agree that the total load along this like (between the two columns) must be checked as a direct shear case, but this does not force the use of beam shear equations. It just has to be directly checked.

I think that we still will need a punching shear check, just a three sided one.

 

[KootK]

@CEL: There may yet be something to discuss here. I believe that the consensus opinion is that the predominant check truly is a one-way beam shear check on the concrete between columns. It leads to these two important questions:

1) How wide of a "beam" should one assume for the one way check and;
2) At what magnitude of column separation does it switch back to two way punching?

What does the term "direct shear" mean to you in this context? I'm not familiar with that terminology.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[CELinOttawa]

By direct shear I mean an imposed shear step, rather than a gradual increase through a gradual ramping up such as a traditional UDL.

Indirect shear is where you have an adjacent area, not necessarily engaged in resisting a load, which becomes loaded through a P-Delta effect of gradual yielding/plastic behaviour. That one is a short hand I picked up while working in NZ; You'd occassionally see a notation in another engineer's calculations to the effect of "as adjacent area reaches plastic design strength, additional capacity is available through indirect mobilization of XXX".

I think I just started talking about them in this way as I found it helpful for imagining the real-world behaviour we need to deal with. I tend to use the french "cisaillement" for direct shear, at least in my notes and in my own head.

 

[rapt]

If the gap between the columns is small enough (< 1.5D), there is no shear to check for. It is a pure strut tie situation and the diagonal compression strut transfers the load from the column above to the column below. To be safe, you could also check it for beam shear but would use a very high shear angle as the normal minimum strut angle assumption of 21.7 degrees (EC2) to 45 degrees (some very outdated codes) would not longer apply as the direct strut angle would be higher than this, and it cannot be less than the direct strut angle! That is why some codes give an increased shear strength factor at locations closer than 1.5D from the support for loads applied near the support.

In my younger days I would often check this as a shear friction case as well ensuring that the reinforcement supplied for the tension strut also satisfied the shear friction rules for monolithic concrete. And put some horizontal reinforcement in the top half of the member for this and also some extra ties to give the same area vertically but I think this was probably overkill.

If the gap between the columns is > 1.5D, it is flexure shear and use the appropriate code equations, preferably from a more modern shear design method like the Canadian or the Eurocode which address the strut angle requirements better!

Between D and 2D, I would check it both ways!

 

[CELinOttawa]

I think Rapt may just have this one... Time to start delving into a specific example or lay the thread to rest.

 

[bookowski]

Mixing threads here but walked by this today and snapped a photo, pretty heavily loaded walk (more relevant to the other 'walking' thread).

Kootk - I fell behind here, just saw your question about a screen shot - will do that. Yes, the steel can get confusing if you have rails as well but not that tricky really - it just needs to be scheduled and detailed that the rails are only on the non "beam" sides. For a calc I just assume the stirrups replace the rails, it might not be completely accurate but it's as close as I can wrap my head around. On the type of mats you are talking about I don't feel too bad about wasting a little steel in the form of stirrups (if it is a waste).

 

[snowmachine88]

Let me preface my response by saying I typically over check failure modes.

I definitely believe the 1-way shear model should be checked. Then the question becomes how long of a line of the slab should be used for the check? I would use the distance between your 2-way shear lines that you drew at the beginning of the post. T would add the capacity from a 3-sided 2-way shear check to the capacity for the 1-way shear between the columns. Check whichever column has the higher 2-way shear stresses and place the 1-way shear failure line half-way between the columns. With this method you would need to use capital Vu and Vc since you need the whole perimeter to breakout for a failure. Based on this, you no longer need to check 1-way shear when the capacity of your Vc from the 3-sided 2-way shear plus 1 way shear capacity exceed the shear capacity of the 2-way shear for a single column.

I would check situation 2 as well since adding a beam within the slab between the columns would be of no benefit in this situation. The failure planes are entirely outside the columns. You could theoretically run a beam between the columns and have very thin slabs outside of the columns. However, in this case the shear has to change signs so that part of the slab can punch out down and the other part up. Therefore, the parts of the slab that are parallel to the elevation views you have drawn will need to fail by torsion. I'd grab the torsional components of the slabs from ACI 421.1R-99. If both the torsional and 2-way shear are greater than allowable, the section fails. Now here's where it gets a bit convoluted. If either the 2-way shear capacity or the end faces or the torsional capacity of the side faces are greater than the allowable, "borrow" some of the orthogonal faces. Eg. if the torsional side face resistance of the slab fails use enough of the end face until the torsion is satisfied. Then recheck the 2-way shear with the "borrowed" portion of the end faces no longer included in the 2-way shear check. If you can get both torsion and 2-way shear to pass then the section will be okay. If you can not get both to pass the section fails.

A word of note in ACI 318 Art. 13.5 you can use Direct Design of slabs if columns are offset by up to .1L. I believe that is more of a rule of thumb than justification to say all your "close column" concerns can be ignored because by the time you get an offset of .1L you are in the realm of flexure/shear controlling.

For #4, I have never seen bearing stresses control unless the column is a different concrete strength than the slab.

I can dream up any inventive ways for the slab to fail, I think you have it pretty well covered.

 

[paddingtongreen]

While this has now run its course to the probable best solution, I believe that it is important that KootK prove to his own satisfaction that the combined shear case is not real, he should not just take our words for it.

Michael.
"Science adjusts its views based on what's observed. Faith is the denial of observation so that belief can be preserved." ~ Tim Minchin

 

[KootK]

I believe that it is important that KootK prove to his own satisfaction that the combined shear case is not real, he should not just take our words for it.

I still wholeheartedly believe in the combined punching shear check (#2). I only stopped selling it because it became clear that others were not being persuaded by my arguments.

In my younger days I would often check this as a shear friction case as well

Thanks for your excellent comments Rapt. This is very interesting as you are the first poster to validate check #3 at the top. I agree. Where the shear becomes a strut and tie situation rather than a diagonal tension situation, I think that shear friction warrants attention. In one of my previous threads, I pitched a theory that shear friction is self-satisfying in these scenarios without the need for rebar (Link).

T would add the capacity from a 3-sided 2-way shear check to the capacity for the 1-way shear between the columns. Check whichever column has the higher 2-way shear stresses and place the 1-way shear failure line half-way between the columns. With this method you would need to use capital Vu and Vc since you need the whole perimeter to breakout for a failure. Based on this, you no longer need to check 1-way shear when the capacity of your Vc from the 3-sided 2-way shear plus 1 way shear capacity exceed the shear capacity of the 2-way shear for a single column.

This touches on an important point that I've been wondering about as well. We all seem to agree that it's not appropriate to perform the one way shear check using the entire width of slab. However, does one way shear failure over the shorter width proposed constitute failure on its own? Or can that one way capacity be combined with the three sided punching shear capacity as you have proposed? I'm really not sure. I worry that the one way shear failure would initiate an unzipping failure around the two way perimeter and, thus, the one and two way shear capacities would not be additive. Unzipping aside, is it even correct to combine one and two way shear capacities? Allowable stresses for two way shear are in excess of twice the values for one way shear. This may be because, as Rapt suggested above, punching shear is more of a "direct shear" situation (i.e. shear through a compression strut).

I would check situation 2 as well since adding a beam within the slab between the columns would be of no benefit in this situation.

Yes! Finally, an advocate for the combined punching shear check (#2)! I agree completely and should have thought of this argument myself. The one way shear resisting beam element envisioned, on it's own, does nothing to resolve the P x e rotation problem. The rotation of the combined perimeter encompassing both columns can only be addressed through slab moments and/or eccentric shear around the combined shear perimeter (effectively torsion when the stresses are taken in aggregate).

A word of note in ACI 318 Art. 13.5 you can use Direct Design of slabs if columns are offset by up to .1L. I believe that is more of a rule of thumb than justification to say all your "close column" concerns can be ignored because by the time you get an offset of .1L you are in the realm of flexure/shear controlling.

I believe that provision pertains to column offsets in plan rather than elevation.

For #4, I have never seen bearing stresses control unless the column is a different concrete strength than the slab.

Agreed. My concern was that, in addition to the column f'c potentially being higher, the column may also use the rebar to transmit compression. I believe that it breaks down into two cases:

1) At thick lower level transfer slabs, column bars can be developed in compression and bearing isn't generally an issue.

2) At thin upper level slabs, column rebar likely cannot be developed in compression. However, were a column loaded heavily enough for this to be a concern at a thin upper floor slab, other failure modes would surely precede a bearing failure. As such, a bearing check is moot.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[rapt]

KootK,

I have always thought of shear friction as a poor mans strut and tie anyway! Just gives a way of allowing for a construction joint rather than monolithic!

 

[KootK]

Agreed Rapt. I've actually been longing for the day when an accepted strut and tie model for punching shear becomes available. As an Xmas gift to myself, I picked up this book: Link. It deals with punching shear using alternate formulations and, supposedly, presents methods that are code approved in Europe. We shall see.

I've met with a professor by the name of Amin Ghali on several occasions. He did most of the research that got us using stud rails in North America. He tells me that people in the know don't consider two way stresses, as calculated, to be "real" stresses. At least not with respect to magnitude. He said that it was better to consider the calculated stresses as "indexes". You can compare them to code limits to see if you're in trouble or not but that's about it. So... take that as you will.

Since this thread seems to be winding down, I'd like to thank everyone for their participation. I've learned a good deal with this.

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 that I want to either change it or adopt it.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[Brad805]

Dr. Ghali was a professor of mine years and years ago. Great prof.

Good thread koot. I have never had the chance to work on a structure with this condition, but it has been an interesting read.

 

[CELinOttawa]

I wonder if Dr. Ghali was meaning that these are over simplifications of the truth... Something like Von Mises stress?

 

[KootK]

That's exactly what I think he meant CEL. Kind of like how ASD steel beam design in the US makes you feel like you're just doing M/Sx, like in college. I wish code committees wouldn't do that. It only serves to obfuscate the reality.

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 that I want to either change it or adopt it.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.