Confused about support condition of narrow concrete slab supported on steel beam 2

[Aakalim103]

We have a narrow concrete slab supported on three sides on a WQ steel beam (Box beam). The slab is part of a hollow core floor system and on the fourth side, it is connected to the hollow core slab. I have attached a sketch to explain the situation. On the longer side, the slab is only resting on the flange of the WQ beam and is not connected to it in any other way. On the other side, it is just connected to the hollow core slab through holes which will be filled with concrete and rebar ties going from the cast in place concrete to the hollow core.

I am confused as to what support condition i should take to design this slab/beam. I am inclined to design it as a beam simply supported on the shorter edges as shown in the sketch. However one of my team mate thinks that it should be designed as a slab simply supported on all four sides. My rational for designing it as a beam is that the connection to the hollow core is usually provided only to cater for the horizontal shear between the hollow core and slab or wall, so i don't think it will work for the vertical shear, unless it is designed as such. Right now we just have two 12mm diameter rebar loop in for each hole as shown in the sketch. On the other longer edge it is just resting on the steel beam without any shear connectors or embedded bolts, which means that in case the steel beam deflects more than the concrete slab, the slab will not be supported.

I would appreciate your opinion on this.

Thank you.

 

[KootK]

I am inclined to design it as a beam simply supported on the shorter edges as shown in the sketch.

Do the reverse:

- design it as a one way slab supported on the long edges as that's how it will behave based on the proportions and the relative stiffness of a CIP thing compared to a precast, prestressed thing.

- design the adjacent plank to accommodate the support reactions on one of the long sides.

- design the adjacent steel beam to accommodate the support reactions on the other of the long sides.

which means that in case the steel beam deflects more than the concrete slab, the slab will not be supported

I recommend a positive connection between the slab and the steel beam. It would help with differential floor deflections, diaphragm transfer, and possibly code compliance.

 

[BridgeSmith]

For sure, considering it as a one-way slab is the simplest approach, but probably at least a little conservative. How conservative will depend on a number of factors, primarily the aspect ratio of the short span length to the long side length, and the support conditions of the hollow core that it's attached to. If it was supported by steel beams on all sides, it would be fairly straightforward, but the hollowcore throws a wrench into the analysis.

For a one-off design, it's likely not worth the time to do the refined analysis for it as a two-way slab with the varying support conditions. If there's 20 or 50 of these slabs, that's a different situation. Then, you're probably creating an FEM model or some of the more involved bi-axial bending analysis methods.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[BridgeSmith]

Quote (OP)
I am inclined to design it as a beam simply supported on the shorter edges as shown in the sketch.

Do the reverse:

Assumed the OP was suggesting having it slab span in the short direction. Yes, if designing as a one-way slab, it should span the shorter dimension.

which means that in case the steel beam deflects more than the concrete slab, the slab will not be supported.

I'm struggling to understand how that would be possible, with the steel beams supporting the slab, unless the steel beams support other loads not applied to the slab.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[Aakalim103]

Thank you BridgeSmith and Kootk for you helpful replies.

I cannot design the Hollow core to support the support reaction since the hollow core is already sent to the factory. The steel beam has also been ordered, although the steel beam is able to take the support reaction from one side.

The hollow core itself is simply supported but i am not sure if the provided holes and the rebar loops would enough to transfer the support reaction to the hollow core and how do check or calculate it.

I'm struggling to understand how that would be possible, with the steel beams supporting the slab, unless the steel beams support other loads not applied to the slab.

This is an internal beam and it supports hollow-core slabs on other side too. There is also another small beam spanning onto it.

 

[KootK]

Look...I design miles of hollow core each and every year. I can make this issue go away for you nice and easy but, for that to happen, you're going to have to unclench a bit and let yourself come along for the ride.

I cannot design the Hollow core to support the support reaction since the hollow core is already sent to the factory.

1) This is no impediment. Hollow core is standardized stuff that is "checked" to a far greater extent than it is "designed". So you or the fabricator can check it before the plank is cast, while the plank is cast, or after the plan is cast.

2) I guarantee you that this is going to work with respect to plank capacity. If it were me, I wouldn't even bother to run the numbers. Consider:

a) You're only adding 12" of tributary load to your plank. Not even that when load sharing is considered.

b) These planks not excessively long planks (24'?). Unless your loads are monstrous, the standard issue planks that you're dealing with for this condition will likely have considerable reserve capacity.

c) Based on your proportions, it looks as though you'll be able to count on the neighboring plank to help resist your edge load a bit.

d) If, somehow, the edge plank is over capacity, you can just have a core or two adjacent to the infill grouted solid an, perhaps reinforced with mild bar. We do this all the time, it's not a big deal.

3) Based on your proportions, the infill IS going to span one way and be carried by the adjacent plank. This is a function of relative stiffness and, within reason, there's not a damn thing that you can do about it. Your only choices here are to deny it or deal with it.

The hollow core itself is simply supported but i am not sure if the provided holes and the rebar loops would enough to transfer the support reaction to the hollow core and how do check or calculate it.

The rebar are loops are not the vertical shear transfer mechanism. The loops are for transferring diaphragm shear. The vertical shear transfer mechanism here will be the miles of interlock you'll have at the plank keyway joint which PCI hyper-conservatively allows you to take as 80 psi. And you'll find that even that modest shear capacity will be gobs more than you need here.

4) Your CIP span is so short that you don't even need to reinforce it. In fact, you probably couldn't reinforce it reasonably because you'd never get your bars developed ahead of the peak moment with just conventional rebar. I'd just install the loops, run four longitudinal bars through the corners, and leave it at that. I suppose that you could stuff #4 @12" in there but that's bound to feel pretty sill to the contractor installing it.

 

[Aakalim103]

Thank you Kootk for the detailed explanation. It is now clear to me that the slab will perform as a simply supported slab in the shorter direction.

But there is one thing i need to ask just for my own understanding. I read the PCI manual and it says that the longitudinal shear capacity of the keyway joint can be taken as 80psi (or at least that is what i understood, correct me if i am wrong). I could not find any recommended value for the vertical shear transfer capacity of the grouted key-way. Off course the keyway will provide some vertical shear capacity, and in my case, it most probably will be enough, but still but i cannot find any reference as to how much would it be.

 

[KootK]

Thank you Kootk for the detailed explanation.

You are most welcome.

I could not find any recommended value for the vertical shear transfer capacity of the grouted key-way. Off course the keyway will provide some vertical shear capacity, and in my case, it most probably will be enough, but still but i cannot find any reference as to how much would it be.

To find more detail and the vertical shear capacities, you'd have to turn to the references listed in the PCI Hollowcore Manual. That said:

1) If you look at the proposed load/response sharing mechanism suggestions shown in the PCI manual, you'll come to the conclusion that a significant amount of shear transfer capability is assumed.

2) It is generally acknowledged that the grout key probably has more shear capacity than the planks themselves about a plane taken parallel to the long axis of the planks. You get into all that vierendeel effect across the cores etc.