Does SAP2000 account its shear deformation in Shear Wall Analysis? 2

[wisnu666]

Hello guys.

Firstly, i am trying to verify about shear wall analysis that SAP2000 or ETABS did with manual calculation method suggested by Kiyoshi Muto. As we know that Shear Wall has two deformation shape, By flexure and By shear. so, if we calculate the deformation theoretically, by summing flexure deformation and shear deformation, we get the real deformation of shear wall.

Can you explain to me that SAP2000 or ETABS really account for its shear deformation? since it's about verification of FEA calculation vs Manual calculation, in the next, i can make a conclusion whether FEA or manual calculation too conservative, or far beyond from conservative.

btw this is for my thesis

so your help will be appreciated

thanks

 

[JoshPlumSE]

In plane or out of plane shear deformation? You'd have to look up what exact element formulation SAP uses. However, my impression is that it will account for both.

The real key (especially with ETABs where the walls are created automatically) is meshing. Is the shear wall meshed adequately to capture the shear walls deflected shape? Sometimes you want very little meshing (for large models where you're trying to keep the solution time down). Other times you want more meshing to get you greater accuracy.

FEM solutions tend to be to stiff if they are not meshed enough. But, then as they get more meshed, they will become more flexible than traditional hand calcs. I'm not familiar with Kiyosi Muto, but traditional beam theory relies on various assumptions (plane sections remaining plane) that may not be true in the well meshed FEM solution.

 

[wisnu666]

@JoshPlum

Really thanks for your explanation btw that's make sense for me.

about your question, i modelled a shear wall, 2D Analysis. shear deformation that i mean is in plane deformation.

i already modelled the shear wall with refine mesh. 2D modelled. seems the SAP2000 Shear Wall deformation result is smaller than traditional calculation, which the different is not to much, arround 0.01- 0.03 mm for 3 model sample. is that make sense? or we have to check the traditional calc assumption first?

 

[JoshPlumSE]

Usually, an "under meshed" FEM wall will provide somewhat stiffer results especially for short / squat walls. Then when you have tall skinny walls (where flexural deformation becomes more and more important) a well meshed wall will provide slighter more flexible results than a hand calc. But, for a well meshed wall, it should always be a small percentage.

For differences that small, you are fine. What's the construction tolerance for out of plumbness? We can only get so accurate with our modeling when the dimensions we input contain inherent inaccuracies as well.

 

[wisnu666]

@JoshPlum

so, i am modelling a one story of free standing shear wall. 3 models shear wall, same height, same force that i applied, but with variant length (the concept is squat wall, as we know squat wall is dominated by shear deformation). the idea is to know :

1. does sap2000 calculate shear deformation?
2. if number 1 is calculated, what is the value of shear correction factor?

because, my professor believed that there is a mistake (error) in calculating deformation with SAP2000 (we know that deformation has relation to internal force) so we have to check whether is SAP2000 true or not.

by the way, last saturday, i re-evaluted my work. as we know, there are 2 assumptions to calculate deformation. like you said, Euler-Bernoulli assumption (plane section remain plane after deformed) and Timoshenko assumption (After bending deformation, there is another force transfered that induced "shear deformation").

then, i checked my traditional calc. The value of Bending deformation from my traditional calc and the deformation from SAP2000.

the result is quite close!

let see this 3 models sample :

1. SW-1
bending Deform Trad Calc | bending deform SAP2000
0.0508 mm | 0.0489 mm

2. SW-2
bending Deform Trad Calc | bending deform SAP2000
0.0185 mm | 0.0155 mm

3. SW-3
bending Deform Trad Calc | bending deform SAP2000
0.0063 mm | -0.006 mm

SAP2000 value is come from my NON MESHED model.

so i from the result, the conclusion come to my mind. probably, SAP2000 calculate the deformation is based on euler-bernoulli theory, which the shear deformation is not accounted.

how about your opinion @JoshPlum ?

 

[JoshPlumSE]

When you say a "Non-Meshed" model, do you mean that wall is modeled as a single plate element? If so, then it is amazing that your results are as close as they are. The issue is that the shape function of a plate element is not generally sophisticated enough to capture the deflected shape of the wall in a single element. Which is the reason why walls have to use a mesh of plate elements. A three by three element tall should be sufficient to capture shear wall behavior. But, a single element won't be very accurate on its own.

Honestly, if the results are as close as you are suggesting then it suggests the plate element formulation is really very good. To be so close with a single plate element is impressive. Though it is still poor modeling practice to leave walls unmeshed.

 

[Bridge Freak]

Make sure that you have opted for thick shell elements: Define > Section Properties > Area Sections. Select "Shell" and click to "Add New Section". Select "Shell - Thick".