Moment Transfer in slab-column connections

[MattJM]

I need help in understanding.

In general, for punching shear at slabs/columns, I definitely need to be concerned about the total shear force over the total shear area (bo * d). The source of shears are from direct shears and moment transfer.

However, in section 11.12.6 of ACI 318-05, I need to go back and check maximum shear STRESSES at points along the critical section. This is compared against sort of an average of capacities of the area for steel and concrete (per eqn 11-41).

Question(s) 1: If the TOTAL shear capacity around the critical section is capable, is it significant that stresses in certain locations are over the limits? Can concrete fail progressively around the perimeter of the critical section - like unzipping?

Question(s) 2: Imagine a perimeter column with perimeter beams only. The beams may provide a lot of shear capacity for the TOTAL punching shear perimeter of the column, but it is possible that stresses out in the slab are still exceeding the limits. Do I need drop panels (for example) and beams?

Thanks

 

[hokie66]

With beam and slab construction, you generally take the total shear in the beams. If you are using a flat slab or flat plate without beams, you have to restrict the stress in accordance with the code, so you have to allow for the unbalanced shear. Consequently, most flat plates have either spandrel beams or cantilevers. I don't know if it is exactly "unzipping", but concrete does fail due to stress concentrations.

 

[JAE]

Agreeing with all hokie66 said...

1) The total shear capacity is for direct downward shear, evenly distributed around your critical section. When you have unbalanced moments (i.e. moments that are transferred from the slab to the column) you will get higher stresses (higher than the averagae) on one side and lower on the other.

From this - you need to convert your [φ]V_n (a force) to an [φ]v_n (a stress).

One way to look at this is that with a wide column strip you analyze it as though 100% of the column strip unbalanced moment is transferred directly to the column. The problem is that the column strip is wide and the column relatively small. Thus, some of the moment is narrowed down as it approaches the column and is taken into the column in a flexural way (compression block with tension rebar mechanism).

But some of it is taken through torsional twisting of the orthogonal column strip and this results in a transfer of the force through the shear stresses across the slab itself. The figure R11.12.6.2 shows the stress distribution.

2) Per hokie66.

 

[MattJM]

Thanks hokie66 and JAE.

ACI commentary on the gamma factors does say it is inteded for slabs without beams. And that is how I design it.

However, in theory, a stress concentration is a stress concentration. So when beams are present at the edge (or some other means to handle shears), how do they know to pick up the extra shear that is out in the slab. I can say to myself that the beams have more than enough capacity - it can't possibly punch through THEM.

But have I neglected some behavior happening out in my slab? And doesn't the slab deal with this "stress at a point" in a similar way in both systems?

In theory, it seems that if you can't create a total mechanism for punching, in both flat plate/slab or with beams, you can't really punch through. Something else must happen. Like local failure out in the slab where the stress is high that slowly progresses back toward the edge? (unzipping)

In that case, if I have beams, by the time my beam picks up the balance of shear, I still have a cracked column strip.

Seems inconsitent to me.

Thanks again.