Can I release the columns suppoting a 2 way flat slab without drop panels? 1

[T2ioTD]

I tried to model a two way flat slab without drop panels, and since I understand that the engineer can decide whether to release or no the columns directly under the slab, I did that, but to my surprise the model went unstable.

What is normally done for such cases? my slab is relatively thin (20 cm) and the columns are 25x80 cm

 

[jayrod12]

What is your lateral system? Is it modeled appropriately? Is it sized appropriately?

What is the connection from column to slab going to look like? Often it's more difficult to detail a pinned connection in concrete than just design for the rigid case.

 

[KootK]

It is likely that the column to slab fixity was the only thing laterally stabilizing your slab laterally in your original model and, thus, your model became unstable when you took that away. If you are presently modelling the gravity system only, you can probably just add some lateral restraints to your floor deck to stabilize the system. Those restraints would serve as stand-ins for your real building lateral system, whatever that is.

 

[CURVEB]

I would study the code on this carefully. We came across this issue recently and I found many instances in ACI318 that indicate you need to consider slab moments transferred to the columns. Obviously, there is some engineering judgment that comes into play here, as the connection between slab and column likely falls somewhere between fully fixed and fully pinned. However as others have said it's probably going to be tough to justify a true "pinned" condition in concrete construction. To envelope the possible controlling conditions you might need to run the model both ways:
- Pinned columns is likely going to cause the largest deflections in the slabs and the highest moments, except for at perimeter columns where the slab moment would obviously 0 as there is nothing to carry the moment.
- Fixed columns will add moments to your columns and the moment in the slab will increase your punching shear requirements. This will also capture the negative bending in the slab at the perimeter columns.

 

[KootK]

I agree with CUVEB and really should have mentioned the same in my original post. In my opinion, a pinned column model is unsafe for:

1) Column design in many instances.

2) Punching shear capacity.

3) Column top steel as it is intimately related to punching shear capacity.