Modeling low rise boxed, C-shaped and L-shaped Shear walls in Etabs

[Rajshrn06]

Hi

When modelling boxed, C-shpaed and L-shaped shear walls in Etabs for low rise buildings, there are large tension and compression forces on the flanges when primary seismic loads are applied in the web's direction. This is because Etabs treats it as coupled wall. If these walls are concrete with reasonable reinforcing, then I would make use of the coupled effect, however this means that my footings under flanges would be large and can get uneconomical for a low rise building.

My question is that if we are using masonry, would it still be reasonable to consider this coupling effect. If I specifcally disconnect the walls at the junctions in Etabs, there will be no tension and compression in the flanges. Are the tension and compression loads in the flanges valid for really long flanges, where from a design perspective, normally only a 45 degree projection of the flange will be used? Can I treat these walls as independent of each other?

 

[sticksandtriangles]

Its a good question and I have not always liked how I handle it myself.

Within the seismic portion of ACI (18.10.5.2) they mention rules for effective flange width from the face of the web (min of 25% of height of wall or 1/2 dist to face of next web).
I think there is something similiar in TMS on CMU walls. ACI also mentions that this can be increased if a more detailed analysis is performed... not sure what this means ...

In my designs I have either felt comfortable with a fully composite C or T shape concrete shearwall if the C and T was "small". If they became large, I would often disconnect the flanges and webs. My definition of "small" and "large" are arbitrary though and could vary from project to project.

I would interested to hear how others handle this.
It gets even more fuzzy on how to handle this stuff when lots of openings come into play as well.

I see that you mention treating it as coupled yields larger footings, how different are we talking? I would think that the results would be somewhat similar.



S&T

 

[r13]

I don't know Etabs, but I think you shall model the way that you would do if design by hand.