Rebar 90 degrees hook length 3

[milkshakelake]

Question:

Why do some engineers use significantly longer 90 degrees hook lengths than is required by ACI code? For example, having the leg be 24" long for #5 bar, when only 10" is needed.


Background:

Per ACI 318-14,

So #5 needs 10". This is also shown in CRSI manual. However, I've seen two different firms use 24" for both #4 and #5. It seems excessive. I think there must be a reason for it, like it's easier to cut or something. Or maybe it's a regional thing, where "my grandfather did it this way for 1000 years" so everyone does it. I'm in the east coast.

 

[gte447f]

I've never encountered folks using anything other than "standard hooks" for development. Maybe if the bar is needed for reinforcement on both sides of the bend?

 

[Celt83]

Usually use a longer Ldh distance to bring the hook down to the bottom reinf. elevation of foundations so they have something to support the bar and tie too.

Some detailing conditions you are trying to pass the tension around the joint rather than just develop the bar so need the Lext dimension to lap with the bars you are passing the tension to.

 

[Brad805]

There could be technical reasons, but I suspect the vast majority do this for the same reason we all like 3/4" plate or thicker.

 

[BridgeSmith]

We extend the bar after the hook in our RC retaining walls because it's a convenient way to reinforce the bottom of the toe. I suspect that the reasons are similar for the places you've seen this. Could you post some example details of what you're referring to?

 

[skeletron]

We extend hooks in cases where nominal bottom reinforcing is not provided separately. Most of the time I show up and it’s an 8” hook unless I specify something on the details.

 

[Tomfh]

24” gives you a decent lap.

Post examples of what you’re talking about so we’re all on the same page.

 

[milkshakelake]

Sorry, I should've clarified what I'm looking at. It's dowels between two pieces of concrete, like foundation wall to footing or shear wall to slab. (I know there are other issues with the details below. I'm not looking for that kind of feedback; I'm working with the original engineer to fix it.)

@gte447f I get what you mean, using the non-Ldh part of the hook as reinforcement. Similar to what Celt83 said.

@Celt83 You mean as a moment connection, like passing the tension into the perpendicular direction as a moment couple?

@Brad805 I'm suspecting it's something like that. Maybe it's just easy and brainless to cut a piece of rebar to 2'x2' or something.

@BridgeSmith I see, so you'd be double counting the hook for development and for actual tension reinforcement.

@skeletron I don't really do that for footings. I mean bending the bottom bar up into a column or wall. I'd rather add a dowel because I'm more used to it, I guess. But either way works.

 

[TRAK.Structural]

With that detailing, the bars in question may be to transfer diaphragm shear into the supporting walls. The slabs are likely not thick enough to justify that the bar is fully developed in the slab section with std hook geometry

 

[bugbus]

There's a common misunderstanding that I've noticed, whereby as long as the total length of the hook and the straight extension exceeds the development length, you can consider the bar fully anchored.

Which is obviously not correct and I see it being abused all the time in situations where there is not enough room to get a bar to develop, so people will just keep extending the tail of the bar until they get enough overall length.

Unless you are providing a larger-than-normal bend diameter (>10*db if I remember correctly), the reinforcement can't be considered to continue developing along the tail of the bar.

 

[Greenalleycat]

I'm with Trak and Bugbus. I don't think this works.

Putting a critical engineering hat on, do we think that the stiffer load path for development is going to be the inside of the hook, or the long ass tail that someone stuck on it?
Obviously it's going to be the hook so the peak load will be inside the hook
If Ldh isn't provided then you risk breaking the concrete IMO
This is relevant to the detail you provided - the bar from the wall to slab is way underdeveloped inside the slab as the hook is so close to the bottom

The advantage of the longer tail is realised at that at least there is something still anchored into the concrete if a cone blows out
But then you have some sort of bar in bending-catenary situation going on, so it's not a primary situation

To me, the advantage of using a hooked bar with long legs is all about construction
I commonly do it for retaining walls - it saves them a bunch of steel and makes it way easier to support all the bars
The horizontal legs support the vertical legs and vise versa, whereas using straight footing bars tied to a short-hooked vertical starter is a floppy system with way more tying required

 

[milkshakelake]

Agreed with bugbus. Only the vertical part of the dowel would be in development (Ldh) so the long horizontal leg doesn't do anything for development. And yes, it's not fully developed, but it's not really possible to fully develop #5 bars in an 8" or 10" slab. This is going into the weeds a bit, but I had a debate with another engineer about dowel size. We never really resolved it. But the general idea is that if you're using a #5 instead of a #4, you won't get full development, but you can use the ACI 318 strength reduction (L[sub]actual[/sub]/L[sub]required[/sub] or something like that) allowed for partially developed bars, and the #5 will function the same way as a #4.

This is relevant to the detail you provided - the bar from the wall to slab is way underdeveloped inside the slab as the hook is so close to the bottom

Agreed, that's one of the issues with the detail that I'm working with the original engineer to fix. I'm hired to do quality control and calculations.

The advantage of the longer tail is realised at that at least there is something still anchored into the concrete if a cone blows out

That's actually a good argument for doing the longer tail. I can buy that. I don't remember the actual name, but I read a forensic engineering report where a parking garage had a pancaking collapse due to punching shear failure. The investigators noted that if the hooks between column and slab were extended, they would have saved the day. That being said, I don't see hooks on a continuous wall or footing to act in the same way as in a punching shear failure, so maybe it's not necessary. But then again, you don't want to be mixing dowel sizes on a jobsite, so just use the longer ones.

The horizontal legs support the vertical legs and vise versa, whereas using straight footing bars tied to a short-hooked vertical starter is a floppy system with way more tying required

I'm not following. Can you explain this again?

 

[TRAK.Structural]

I've used the (actual/required) ratio concept before and think it is applicable here.

If this dowel is intended to work in shear for some sort of lateral load than the calculation should follow the shear friction provisions of ACI (if this is IBC design). You could intentionally roughen the top surface of the walls to get more capacity out of this joint.

 

[Celt83]

...but you can use the ACI 318 strength reduction (Lactual/Lrequired or something like that)...

Technically this reduction is not permitted for hooked bars.

ACI 318-14, 25.4.10.2 exception (b) triggers for dowel conditions as these are for shear friction transfer and need to develop Fy on both sides of the slip plane.

ACI 318-19, 25.4.10.2 amended exception (d) to include hooked reinforcement.

 

[milkshakelake]

Thanks for the code references, Celt!

In that case, though, I don't think even #4 can be fully developed. I ran a quick calculation using 25.4.3.1 and I got 9.5" for 4000 psi concrete.

 

[BridgeSmith]

@BridgeSmith I see, so you'd be double counting the hook for development and for actual tension reinforcement.

I'm not sure what you mean by "double-counting". It does perform 2 functions, if that's what you mean. the hook would theoretically develop sufficient tension capacity for the vertical leg, and the extension gives it sufficient development to resist the moment in the footing. I don't se it as being substantially different than a straight bar extending through a tension zone - there's tension force pulling both directions away from the middle.

In the case of your dowels, if the dowel is supposed to carry moment around the corner, the horizontal leg of the dowel has to have a lap length to the top horizontal bar past the point where both bars are needed to resist the negative moment in the slab.

 

[Greenalleycat]

Milkshake, I've quickly knocked together some sketches to show what I mean
In general terms, both are fine ways to build a foundation (though in the 2nd one my footing bar should probably have hooks)

 

[Tomfh]

This is a bit off-topic and a somewhat controversial topic to discuss, but these joints are generally much stronger if you turn the cog back towards a slightly extended toe (image below). The two methods illustrated above can result in the cogs ripping out well before the wall or footing reaches its flexural capacity. Often that’s fine and these walls have plenty of capacity anyway; but not always.

 

[milkshakelake]

@Bridgesmith I understand, thanks for the explanation. It's for diaphragm shear transfer, not moment. I prefer to keep this pinned, so maybe I'll shorten the bars to the ACI minimum. I'll keep it in mind for real moment connections.

@Greenalleycat Thanks for sketching it out! I get what you mean. And if the retaining wall is taller, that big piece can still function as a dowel.