Unrealistic Shear Diagram in ETABS

[Yousef ZAA]

Good day,
I have a problem justifying the shear diagram in one of the shear walls; The structure consists of shear walls and since the slab qualifies to a rigid diaphragm, a rigid constraints has been assigned to all shear walls joints. Of course the expected shear diagram of each shear wall is simply a step diagram with the shear forces increasing from the top level downward, but one one of the shear walls has its shear on the 1st level less than the second level, that gives the impression that there is a negative reaction or something! I could not justify this shear distribution. This particular shear wall have openings at 1st level and I understand that considering the relative stiffness of this shear wall at the 1st level it won't attract that much of forces but still could not justify how would the base shear (shear at 1st level) be less than the shear the 2nd level! The torsional component on shear walls does not justify this as well!

 

[dold]

Rigid diaphragm is taking force from the bottom of level 2 shear wall into the diaphragm and back into other, stiffer walls. Fairly common issue when modeling with rigid diaphragm assumptions. The base shear of the entire structure should not / cannot be less than level 2 story shear, but the force in a shear wall can drop from one level to another when using rigid diaphragms.

 

[chris3eb]

I agree with dold that the diaphragm can be transferring shear from the upper levels of the wall in question to other, stiffer walls. I wouldn't necessarily describe it as an "issue when modeling with rigid diaphragms" however. If you switch to a semi-rigid diaphragm which essentially models the diaphragm's stiffness, it can still happen. This phenomenon is known as the backstay effect.

 

[JoshPlumSE]

but one one of the shear walls has its shear on the 1st level less the second level, that gives the impression that there is a negative reaction or something!!

In my experience, this is relatively common "issue" for a rigid diaphragm constraint. dold and chris3eb have it right (IMO).

In my experience, most engineers will design that shear wall (the one with the backstay effect) for the larger shear all the way down to the base of the wall. But, will design he other shear walls as if they really do receive extra load from the backstay wall. This is inconsistent, but conservative.

Chris3eb has a good suggestion. If you switch from a rigid to semi-rigid diaphragm at this level, it might mean that the problem will go away. Maybe not completely. But, the diaphragm will behave less rigidly, so it should get closer to what you're expecting to see.

That being said, there is a computational issue when you do this. You're dramatically increasing the degrees of freedom at that level, so you'll probably slow down the solution.

 

[dold]

I concur with chris3eb. For example, if you define the diaphragm as semi-rigid and manipulate the diaphragm stiffness properties from "infinitely" rigid to very flexible you'll see this backstay behavior diminish. I.e., the less rigid your diaphragm is the more force your lowest wall will take from the wall above.

Side note: rigid links/diaphragms are never "infinitely" rigid, but defined as orders of magnitude more rigid than the other elements in the model. In RAM and RISA (not sure about etabs but i'd imagine it's similar) rigid diaphragms are internally constructed by connecting all of the nodes at a diaphragm level with rigid links. FWIW.

 

[Yousef ZAA]

Thank you all for your helpful input.