Concrete Column Moment at Transfer Slab

[JSMHA]

Our office is using Ram Structural System and Ram Concept to model and analyze a two way concrete flat slab with 5 stories of wood framing above. With these concrete transfer slab buildings we have noticed a significant amount of moment at the top of the concrete columns. We are essentially following the workflow proposed in this article:

https://www.thestructuralengineer.info/news/concrete-design-workflows-in-ram-structural-system

. We are seeing a particularly high moment in the columns, especially from the Ram Frame analysis, which includes ACI chapter 6 cracking coefficients for the slab and columns. The building is completely laterally supported by concrete shearwalls.

The question I have, is how are people handling this high moment? We have been discussing enveloping our slab design. We would analyze with the walls pinned at the top and the columns fixed at the top. Freeze the reinforcing, and analyze again for the columns pinned at the top. This is due to some concerns on the moment transfer from the slab to the column, due to the dowels not being fully developed without extending the hook, which ACI does not usually recommend. But we would design the columns for the moment based on the full fixed condition. Does anyone design the top of concrete columns for pinned, no moment? Or maybe design the columns for a partially fixed condition?

We are getting some cases with exterior concrete columns, 18"x18", where the moment is more than twice the axial load, which does not seem practical.

 

[HTURKAK]

Modelling of flat slabs always an issue . I have experience with SAP 2000 .

If the column locations reasonably regular, i would check with equivalent frame analysis as suggested by ACI .

For FEM method, IMO, it is not reasonable to model the columns with frame elements and slab with plate/ shell elements which the analysis will produce unrealistic figures..

I will suggest you to make fine mesh around columns ( at least 3 nodes at EF of column) and modell the columns with brick elements or
define master node at center and slaves around and you can define springs to model the column stiffness.

The following snap from the book MOdelling of structural systems by Iain Mc Leod .

 

[cliff234]

When we need to develop a slab-to-column moment connection we typically stop the vertical bars in the columns at the underside of the slab, and then we specify a larger number of smaller hooked column dowels (#5 bars preferably) into the slab. #5 hooked dowels will fully develop in thin slabs if horizontal bars are detailed in the corners of the hooks. If the slab is thick enough, we’ll specify #6 vertical hooked dowels into the slab.