Torsion of flexible diaphragms

[JIMEY]

Almost every source that I have read, including posts on this forum, has said that torsion can be ignored for a flexible diaphragm. However, when I model a flexible diaphragm in structural analysis software (I use ETABS), the results I get seem to treat torsion similarly to a rigid diaphragm. The only difference is that shear wall stiffness does not affect the load going to it.

Can someone please clarify for me how torsion is to be treated when dealing with a flexible diaphragm, because I can't seem to justify ignoring it at this point. If there is an eccentric load, it only seems logical to me that more load would go to the shear walls at one side than the other. Thanks in advance.

 

[msquared48]

Maybe I'm way off base here, but I always thought that ETABS was for concrete, implying rigid diaphragms only.

Professionally, I have never used it, but have not seen it used in any other circumstance... yet.

Probably all wet here... (curing)

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[JIMEY]

Just to clarify, what I did was I created a new material in ETABS which I made ridiculously unstiff to simulate a very flexible diaphragm. When I assigned it as a diaphragm I picked "semi-rigid". I believe that means that it will take into account how stiff it is rather than just assuming it is rigid.

The ETABS model aside, in my head I can't rationalize how eccentric loads can be ignored. If the resultant load is off-center then doesn't that mean that the shear wall it is closest to will see more of the load than the other shear walls?

 

[JoshPlumSE]

I don't believe that torsion should be ignored for rigid diaprhagms. However, ACCIDENTAL torsion is allowed to be ignored.

To me, if your are expecting significant torsion in your model (which isn't accounted for in your current load distribution), then you just add in that torque moment by introducing equal and opposite shears at a moment arm distance.

 

[honat]

A couple of clarification. Yes ETABS is a building structural design tool just like STAAD would be.
The other concern is that you are probably ready too much into the purpose of your DIAPHRAGM. Its primary pupose is to reduce torional distortion in your structural frame system. For that reason a certain degree of rigidity is needed. If too flexible, horizontal bracing wiil be required (i.e most long span shallow metal decks type in industrial setting). Its secondary purpose is to redistribute the horizontal shear forces through your vertical bracing concept. Eccentricity in your diaphram loading will of course determine which shear wall or vertical bracing system will see the lion share of the diaphram load. The most important thing to keep in mind is how to correctly fasten the diaphragm system to the respective chords. True oil cane type flexible roof system should not be used as diaphragm system. In such case brace all four sides or brace 3 sides with proper bracing diaghragm connection to tie the vertical brace system. Simplicity and effectivenss goes a long way.
Good luck

 

[MTJJ]

I think some of the trouble you are having is that software really doesn't model flexible diphrams correctly, (or atleast not how it would be done traditionaly with paper and pencil, ie. "Semi-Rigid" is not flexible in any software package). And as always there is no such thing as a turly rigid or flexible diaphragm.

The decision of wether a diphragm is flexible or ridid is based on the deflection of the diphram versus the defelction of the lateral supports. When a flexible diphragm is loaded its displacement is so much more than its support displacement that it can be thought of as being fixed at the supports (ie. shear walls are way stiffer than the floor).

While its true there is no torsion in a flexiable diaphragm, it is not ture that the shear wall forces would be the same for an eccentric load. Think of the diphragm as a simple beam supported at the locations of the shear walls. The lateral shear wall forces are found through statics. The wall line closet to the eccentric load gets more load.

Hope this helps. I think "Design of Wood Structures" by Breyer has a nice discussion of flexible diaphramgs.

 

[ash060]

There is not really torsion in the flexible diaphragm. In ETABS, if I remember, the forces it generates are applied at the c.g. of the slab and it applies point loads and moments at the c.g. The moments being your torsion moments.

I would not do this if I had a flexible diaphragm because the load is not a point load. So you could take the point loads and moments and break them down into linear varing line loads, and than do the analysis as MTJJ says.

Also, if you look if the PCI Design Handbook they have a great example of the two types. Look at example 3.8.3.1 it covers both.

 

[JIMEY]

Thanks you guys. I think I may look into that "Design of Wood Structures" book. I've tried searching the internet on this issue, and information seems to be scarce.