bending moment bars at column to roof slab 7

[ajk1]

Question about the horizontal length of the "L" shape bars between the column and the roof slab, that are there to transfer unbalanced moment between the column and slab: Should that horizontal length be a tension lap length of the L bar with the top bars in the roof slab, or should it be a tension development length? See attached.

 

[JAE]

In my view the horizontal leg of the bars should simply be a hook length, not Ld.

You have to have enough vertical distance from bottom of drop panel to top of hook to get Ldh.

Now some have suggested that extending the bar farther (as shown - to Ld) helps but I've never done that and don't believe it helps all that much.

You also would need continuous top bars over the column but I think this detail is only concerned with the column-to-slab bars.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[DaveAtkins]

I agree with JAE--the horizontal length should not be a tension lap length, nor should it be a development length. It should only be a hook extension (12*d_b, I believe).

Hopefully, KootK will also answer. He has put a lot of thought into developing reinforcing bars around corners, and has found some research on the topic.

DaveAtkins

 

[Yao1989]

I ran into this issue before too. I think an really important question is if you actually expect the unbalanced moment to be so large that it causes tension in the dowels.

In theory you can still transfer unbalanced moment to column, even if entire column section is still under compression. In that scenario, your rebar is only just helping resist interface shear.

TL DR unless the moment is really large that you need to do strut and tie model at intersection, I would just do a standard hook just to have bars developed for interface shear.

 

[DaveAtkins]

I just went into ACI318-14 to verify the bar extension--it is indeed 12*d_b.

But I found an interesting footnote to Table 25.3.1. "It shall be permitted to use a longer straight extension at the end of a hook. A longer extension shall not be considered to increase the anchorage capacity of the hook."

DaveAtkins

 

[KootK]

Hopefully, KootK will also answer. He has put a lot of thought into developing reinforcing bars around corners, and has found some research on the topic.

What's this now?? An actual request for my ridiculous pedantry? Sweet... I'm in.

1) As much as possible, and for any kind of slab, I simply run the column verticals bars up to the top of the slab and hook them outwards. I don't do this because it's mechanically awesome because it's clearly not. I do this because contractor's hate the mechanically correct detail and 1/2 of the time they wind up cutting off the hooks anyhow for various construct ability reasons. And this is just not the hill that I want do die on. When I use this detail I will correspondingly design the slab to column joint as pinned for the sake of bending in the slab. I'll still do the punching shear and column checks assuming fixity however. In thin, flat plate slabs, the notion that undeveloped dowels might pull out from the slab does bother me some. I worry that the character of the punching shear demand changes in unhealthy ways if the slab is actually permitted to lift off of the column on the tension side. In some instances, I've lapped #5 dowels onto larger column verticals just to feel good that the bars would yield before they pulled out. This feels much better to me that trying to bullshit develop some #9 verts into an 8" slab somehow. It's also fairly constructable. Contractor's would still prefer to just run the column bars up and have no hook but I haven't gotten any complaints on this. It also seems to greatly improve my odds of not having the hooks cut off in the field.

2) For most flat plate slabs, I'll go with the approach described in #1 for the reasons described there AND because I think that it's utterly impractical to try to get anything resembling a true moment connection given the geometric constraints. No sense fooling myself.

3) For a flat slab with drop panels, if I feel that I really need moment transfer, I'll use the detail shown below. In this case, I believe that horizontal dowel extensions needs to tension lap with the slab top bars. Moreover, in instance, I feel that it is fundamentally incorrect to turn the dowels outwards. Without question, the detail below sucks to build, particularly when considered in two dimensions.

 

[ajk1]

I feel that it is fundamentally incorrect to turn the dowels outwards.

a) Why do you feel outward turning of the bars is incorrect? We don't at the moment see anything incorrect about it, and we know that is what at least one of our competitor(s) does too.

b) We do not turn the dowels inward over the top of the column. Our chief field engineer with 40 years field experience says that it makes it difficult to place the column concrete.

c) Our company practice has always been to use dowels; never to hook the column verticals. Why would you? The column verticals would generally be significantly larger size than would be necessary for the calculated dowels to transfer the unbalanced moment.

 

[KootK]

a) Why do you feel outward turning of the bars is incorrect?

Because it's fundamentally incorrect detailing with respect to moment transfer. The definitive thread on this subject is surely this one: Link. I practically died of exhaustion fighting the good fighter there so I'll ask you to review that rather than ask me to elaborate here.

and we know that is what at least one of our competitor(s) does too.

I don't doubt that for a second. Like I said previously, I do similar things myself, almost exclusively. It all comes down to what YOUR expectations are for the joint. If your expectation is reliable moment transfer, then I don't believe that you have that. If your expectation is only nominal, dowel based shear transfer, you should be fine with your detail (overkill actually).

b) We do not turn the dowels inward over the top of the column. Our chief field engineer with 40 years field experience says that it makes it difficult to place the column concrete.

Without question, the detail below sucks to build, particularly when considered in two dimensions.

I said as much myself. No offense to your rock star field guy but I'd be willing to bet a thumb that I know more about concrete joint moment transfer than she does. Just sayin...

c) Our company practice has always been to use dowels; never to hook the column verticals. Why would you?

Again, it depends on what your goals for the joint are. If you want to anchor the dowels in the slab for fy and there is insufficient slab thickness for straight bar development, then you'll want hooks to shorten the development length. If you want reliable moment transfer at the joint, you'll want the detail that I showed above. If all you want is nominal shear transfer, then you probably don't even need dowels as the columns are usually cast a little high.

The column verticals would generally be significantly larger size than would be necessary for the calculated dowels to transfer the unbalanced moment

And those large bars are all for naught if they will just tear out of the slab when you try to mobilize them for flexure. In this respect, you're almost better of with smaller bars as I mentioned in my previous post.

As I mentioned previously, I do very similar things to what your firm is doing. The only thing that I'm NOT doing is fooling myself into thinking that these joints are reliable moment connections. I'm not proposing that you change your ways.

 

[ajk1]

I am clearly not as learned as you. If you could perhaps include a free-body diagram of why turning the bars outward does not accomplish moment transfer just as well as turning them inward, I would be appreciative.

(On another matter, I never referred to our chief field engineer as a rock star, nor do I know what that means. He is an unusual engineer in that he is very good at both calculation and design. Most engineers prefer to do design. He has also kept a lot of engineers out of trouble. All this while always with a smile on his face and never ever a word of anger in his voice when explaining things. He is a gem of a person in every way. This does not mean that he is in disagreement with what you say...I will check with him. If you think that long good experience is not meaningful, just look to POTUS).

 

[ajk1]

I have not yet looked at your link. I will do that this evening, so don't spend time replying until I do that. Perhaps the free body diagrams there answer my questions.

 

[KootK]

I am clearly not as learned as you.

Now, now... let's not be overly dramatic. Everybody has a wheelhouse that they can claim as their own. Moment connections in reinforced concrete just happens to be mine (according to me at least).

On another matter, I never referred to our chief field engineer as a rock star, nor do I know what that means.

Rockstar = one who is exceptional at their vocation. It sure sounds like your guy is a rockstar chief field engineer.

If you think that long good experience is not meaningful,

I don't think that. Nor did I say that. I myself am 113 years old. My intent, at the risk of seeming/being boastful, was merely to suggest that it's a rare field engineer that would be able to go toe to toe with me on this particular issue. Or, at the least, that's been my experience to date. But who knows, maybe your guy is just the guy for the job. He is the chief after all. One of the best technical engineers that I know has kind of aged out of design and now oversees a field engineering team as part of his responsibilities. So yeah, there are exceptions to most rules and assumptions. On the other hand, seeing a bunch of stuff in the field doesn't necessarily mean that you understand how it works. I've had a phone glued to me for over a decade I still don't understand how my voice gets dumped into a signal stream with everybody else's and somehow then gets parsed back out on the other end. Magic!

 

[DaveAtkins]

I myself am 113 years old.

From your posts, I got the impression that you are not a day over 112

DaveAtkins

 

[ajk1]

Just one clarification: the engineer that I referred to as our chief field engineer has been doing mostly design for at least the last 20 years. He was too valuable to keep solely in the field. So he is not a field engineer in the sense that you probably have experienced field engineers.

I am somewhat unclear about what you would actually design, since you say you do recognize the problems with turning the bars inward over the column, and the problems that can present. I am sure you know as well as I that the best calculation on paper does not minimize the field problems and the reduction in strength that may occur if they cannot build it, or if the problems associated with its construction mean that it is unlikely to be built right.

What do you do on an interior column when turning the bars inward over the column, from each face of the column, may result in the bars being crowded too close together?

 

[ajk1]

Below is what the engineer to whom I referred, says. If you disagree with this, perhaps you could send some free body diagram sketches illustrating what the mode(s) of failure you perceive could happen if bars are bent outward for interior columns with unbalanced moments. Thanks for your comments and help.

"I think that the BM bars on the exterior side of exterior and corner columns should bend across the columns since hooked top bars at the edge would not be fully developed at the exterior face unless the slab cantilevers beyond the column face. This could potentially result in “splitting” of the column. For interior columns, I do not think it matters which way the bars extend to develop the moment in the column and clamp the interface. It would be less congested not to cross over the column and I was always concerned that concrete may not get properly consolidated below the congestion of bars over the column when bending moment dowel (BM) bars crossed over in all directions. In fact, I think in many cases the minimum spacing between bars cannot be maintained if bars from both sides of the column cross each other along with the slab top steel. Perhaps they would need to be considered bundled bars at that point.



Unless there is a compelling reason that BM bars need to cross the column I would not show it that way. I do not see how the interaction of the BM bars with the slab reinforcement would require this except at exterior faces of exterior and corner columns but perhaps I am missing something".

 

[KootK]

I am somewhat unclear about what you would actually design, since you say you do recognize the problems with turning the bars inward over the column, and the problems that can present.

As I mentioned in my first comment:

1) I almost never call these joints rigid for the sake of my slab design. Correspondingly, I don't attempt to transfer appreciable moment through the joint and I turn the bars outwards.

2) On the rare occasion that I do attempt to transfer appreciable moment. I'll use small-ish diameter dowels lapped to the column verticals and have them field bent so as not to mess with the concrete placement any more than absolutely necessary. In, say, an 8" slab with an 8" drop panel (where I'd consider doing this) it is doable in the field in my opinion.

I think that the BM bars on the exterior side of exterior and corner columns should bend across the columns since hooked top bars at the edge would not be fully developed at the exterior face unless the slab cantilevers beyond the column face.

That surprises me as I wouldn't think that bending the bars across a corner/edge column would be all that much easier than it is at an interior column.

If you disagree with this, perhaps you could send some free body diagram sketches illustrating what the mode(s) of failure you perceive could happen if bars are bent outward for interior columns with unbalanced moments.

As you wish.

 

[jayrod12]

If you've gone through the other thread that koot linked, I believe most people's concerns with turning them away from the column would be the high compression stresses on the inside of the bar bend potentially busting out the bottom of the slab where it interfaces with the column. This is obviously a concern with shallow flat plates more so than thick plates and drop panels.

Damn, koot beat me to the explanation, and his is always much clearer.

 

[KootK]

I think that it's worth noting that:

1) Most roof slab to column connections have the bars turned outwards and;

2) These connections will attempt to take some amount of moment no matter what they're told and;

2) To my knowledge, the failure mode that I illustrated above does not happen.

One of the most unfortunate aspects of structural engineering, in my opinion, is that hardly anything ever fails regardless of how badly it's designed or detailed. It's something of a consequence free profession which, I suspect, is why it doesn't pay nearly as well as it ought to given our level of training and responsibility. If 0.05% of these connections failed as I've proposed, I'd have lucrative clients beating down my door.

 

[jayrod12]

One of the most unfortunate aspects of structural engineering, in my opinion, is that hardly anything ever fails regardless of how badly it's designed or detailed.

Why do you think I get to keep my job???

 

[Celt83]

Have referenced that retaining wall thread a few times.

KootK:
Would the presence of a bottom reinforcing mat or concentrated bottom bars create the necessary compression strut support?

Edit:
I had the second compression strut in the wrong location should start from the top of the first tie.

Open Source Structural Applications:

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

 

[KootK]

Have referenced that retaining wall thread a few times.

Sadly, that thread may well turn out to be my life's great work. I should force my kids to read it.

Would the presence of a bottom reinforcing mat or concentrated bottom bars create the necessary compression strut support?

Not in my estimation. It still lacks a mechanism for resisting the vertical component of the compression strut that is something other than diagonal tension in the concrete. You'd have something with a bottom mat and studrails in play. But then you'd have to actually go to the trouble of calc'ing that out. And even I'm not willing to waste that kind of time on a pseudo-imaginary failure mode.

In most cases, you'll actually have highly compressed concrete at the bottom of the slab to begin with. As such, you've got restraint for the lateral component of the offending compression strut in spades even without the bottom steel.