Beam and Slab Modeling when T section is Difficult to Determine

[LanaCiv]

Hi,

I am designing a building where beams are oriented in arbitrary directions, which makes it difficult to determine effective T/L section.

Therefore, I am considering modeling the beams as simple rectangular sections while still ensuring an appropriate beam–slab interaction. Based on the Sofistik tutorials, I am unsure about the correct position of the beam relative to the slab CG. Should the beam CG be aligned with the slab CG (Option 1), should the beam be positioned so that the top edges of the beam and slab coincide (Option 2), or is there another recommended approach?

The two modeling options (Options 1 and 2), along with one of the level layouts, are shown in the figure below for reference.

Additionally, regarding torsional stiffness, my understanding is that the beam torsional stiffness is often reduced to approximately 10% (mostly to prevent the beam from attracting unrealistically large torsional moments, which in reality would be distributed between the beam and the slab). Is this reduction also recommended when modeling in Sofistik, or can the full torsional stiffness (100%) be retained?

[JE170969]

If you calculate internal forces with ASE option 2 can be used.
For this case, the geometry of a T-beam (slab-beam) cross-section should be used, but with a very small slab width, as shown in the 01a.

In the ASE report, you should also verify in the section “Reduced Beam-Stiffness in a Slab” that this cross-section has been correctly identified (see 01b).