What's behind the scene of using semi regid diaphragm 1

[dollarbulldog]

Dear members

In ETABS v.9, it allows to use semi-rigid diaphragm to model a floor diaphragm. CSI said that the semi rigid diaphragm considers actual stiffness of elements rather than assuming a kinematic assumption as rigid diaphragm does. The question is, if so, what is the advantage of rigid diaphragm over the flexible model (without assigning any diaphragm)? Second, if so, do we need to consider the accidental torsion as stated by most design code when a semi-rigid diaphragm option is selected? Can anyone suggest.

Thanks

 

[structuresguy]

Well, i always use semi rigid diaphragm option. That is more accurate than rigid, and will avoid some of the problems that rigid diaphragm could introduce.

As for no diaphragm, I don't know how you would have a stable model for typical office type building, where all the floor beams are pinned to the columns. The diaphragm is required to take wind/seismic loads and apply them to the lateral resisting system.

IMO, you should still consider accidental torsion, even with semi rigid, unless the code says that analysis based on actual stiffness does not require accidental torsion. But even then, are you rally modeling "actual stiffness"? I still think not, as we have assumed pinned and assumed fixed conditions, arbitrary stiffness reduction for assumed cracking, etc...