Negative Moment Transfer in Typical Concrete Beam/Column Connection - Standard Hook Anchorage 1

[dijonnaise]

I was having a fun discussion with a junior designer today and came away with a few questions about how beam/column connections are typically detailed.

If we model the standard hook connection at a beam/column as a strut-and-tie, in many cases the standard hook would not be developed within the extended nodal region at the hook node (see attached sketch).
Is there some other factor at play that gives us the certainty that the hooked bar is anchored at this node?

My code (CSA A23.3) indicates that negative moment reinforcement is developed so long as the hook is developed (ldh) before reaching the face of the column (Cl. N12.12), however this doesn't seem to satisfy the tie anchorage requirements of Cl. 11.4.3.2 (Anchorage of ties in node regions) if we consider this a D-region strut & tie.

Thanks!

 

[KootK]

1) I agree, this is an interesting idiosyncrasy.

2) As you've rightly identified, many joint configurations will not pass a strut and tie check.

3) At conditions where the detail is meant to apply, you have supports above and below the beam that create the beneficial clamping action shown below. This is key. This clamping mechanism confines the snot out of the bit of negative steel being anchored and thereby shortens it's required development length in a manner that, to my knowledge, goes unquantified.

4) At conditions where you don't have the clamping action -- such as roof beams -- the typical detail does not apply and you're supposed to wrap the negative steel around the corner and lap it with the outer column bars. This, effectively, becomes a curved bar node transferring moment around the joint corner rather than "anchorage" per se.

5) As we get better at seismic design, ACI has generated documents that detail procedures for checking stuff like this and for checking the very high shears that occur within the joints. As with the strut and tie stuff, this will tend to govern the proportioning of beam and column joints. It is my expectation that, eventually, we'll have to explicitly design all beam and column joints rather than just "detailing" them. 2050 maybe.

Is there some other factor at play that gives us the certainty that the hooked bar is anchored at this node?

To be clear, I do not believe that we actually have "certainty". What we have is moderately thoughtful detailing and questionable historical precedent.

 

[dijonnaise]

Fantastic, this tracks with a discussion I was having with a clever colleague later on in the day. I agree that the axial load from above fundamentally changes the strut-and-tie model in question, and agree that we should be careful to lap a hooked bar into the beam at the top of roof columns (looking into it, this is already incorporated into our top of column typical details).

Thanks!

 

[KootK]

You're welcome. We Canucks gotsta look out for one another.

There's another permutation of this that may be of interest to you. It is my belief that the standard detailing is actually predicated upon certain member proportions. If you imagine a deep beam as shown below, a single strut through the joint would be too steep to calc out. So, instead, we replace the single strut with multiple struts which then creates a requirement to run the column ties through the joint. This is also part of some firms' typical detail.

I sketched this as a roof joint because it makes the salient features pop. I expect a similar result at deep floor beams too however. At any joint, the primary job of the joint is to connect the beam tension steel to the beam compression block with respect to horizontal shear.

 

[dijonnaise]

We do carry that detail as well. Actually, my interpretation of Cl. 7.7.3 is that these ties are basically required according to the standard for any end column condition.

 

[Lomarandil]

You beat me to it Dijonnaise (great name)... I was about to say the same thing coming from ACI 352R

----
just call me Lo.