different period and story draft using shell and membrane

[useetabs]

I use etabs 8.5.6 to model a 35 story building. I model the slabs using shell and membrane and obtain different period.
torsion x y
shell 4.6 3.5 4.15
membrane 20.3 7.5 15.6

the drafts using membrance are much larger than using shell.

please whether it makes sense or the mode has some problem?

 

[Murali27]

It looks like there is a problem with local modes. Please verify that whethe you had applied rigid diaphragm.

-Murali G

 

[useetabs]

Hi Murali:

Thank you! this mode had been applied rigid diaphagm.could you give more suggestion

-useetabs

 

[jen4950]

This makes perfect sense. Membranes have no bending stiffness, whereas shell elements introduce the stiffness of the slab into the system.

Be careful here- say for instance you have a shear wall primary lateral system and have the gravity columns modeled in the ETABS model, they will attract moment and forces if connected to a shell floor element, or if geometric or mass floor to floor inconsistencies exist and you have a rigid diaphram. The differential intrastory movement will engage the gravity columns.

 

[useetabs]

Hi,jen:
thank you. then which mode should I use to design shear wall?

 

[jen4950]

Well, it depends on what type of structure you are designing. Dual system? Devil is in the details..

I'd recommend a membrane element for a strictly shearwall lateral system. There are several different opinions about the subject though.

One guy I know would say model the whole damn thing and let the force go where it wants to; But this is risky- if you have a flat plate system your punching shear might control the capacity of the system and you might get unrealistic results.

Think pancake failure = bad.

But if you have a pan joist or slab and beam system, it might make sense to model elements that in reality have the capacity and stiffness to contribute to the lateral effectiveness of the system.

"Engineering Judgement in spades."

 

[useetabs]

Hi, jen:

You are right. As you said, I have a shear wall primary lateral system , the gravity columns system and a flat plate system modeled in the ETABS model. I export the slab model with safe export option "Export Floor Loads and Load from Above" and the punch shear obtained from safe is very large and unrealistic. could you give me further suggestion to do the ETABS model? Thank you very much!

useetabs.

 

[jen4950]

I'd model the plate as a membrane element and do the flat plate as a seperate analysis in something like PTData or ADAPT if it is a flat PT plate.

Do not rely on the stiffness of the flat plate for your lateral system IMHO. It is very rigid in shear and as a collector mechanism but bending is not effective. I.E. gravity- it works; stiffness in regards to the lateral system- think paper with a toothpick thru it. It's just not realistic or safe. You don't see any flat plate moment frames without beams- so why include them as a contributing element just because you have shear walls?

This is the fundamental arguement for rigid diaphrams- Of course if you have a crazy geometry and structure like I'm dealing with right now it becomes clear that a rigid diaphram just might not be conservative.. but that's another thread all in it's own..

I'd only use the ETABS model for a gravity load takedown and column design; as well as shear walls. PT plates are an art that is out of ETABS land.. be sure to use dedicated PT software like PTData or ADAPT.