Shear wall with or without flanges? 1

[drile007]

Hi,

Did anybody find out the difference in (concrete) shear wall reinforcement if you model them with or without flanges?
Thank you for any commnet.

Drile007

 

[iacobul]

It's a big difference between shear walls with or without flanges.A big flange gives the wall a great increase in momentum capacity, and you have a lot more space (and more economical way) for putting the bars. Write the equlibrum equation for a shear wall and you will understand what i am saing .

 

[drile007]

Yes, I know there is a big difference, but how big is it (approximately). I wondering what is the common engineer praxis...modeling whit or without flanges?

Drile007

 

[RARWOOD]

Often you don't have much of a choice, as the amount of wall available is established by the architect.

 

[drile007]

I agree, but do you take them into account if there are any?

 

[RARWOOD]

I would!

 

[drile007]

How? What’s the width of the flange, is that constant through the height of the building, modeling two piers for every direction and combining the results, etc., etc.. I know we should do it, but I think in the common praxis nobody plays with that?

 

[slickdeals]

Think of it as a cantilevered wide flange beam. Are you better off with just the web or the web in addition to the flanges. Or think of it as a truss with super strong diagonals but weak chords (in other words a small flange).

 

[Lion06]

One thing to be cautious of when doing this - If you have a shearwall with a flange, the flange is going to resist lateral forces in the orthogonal direction. If your center of rigidity and resultant lateral force do not coincide (even if they do, for seismis you must assume 5% accidental eccentricity) you could end up with the direct shear causing compression in the flange, and the torsional portion causing compression in one portion of the flange (for shear in the orthogonal direction). Come to think of it, this is going to be a problem even for wind with all of the quatering wind load cases and required torsional moments on the building per ASCE 7.

 

[drile007]

Interested point of view StructuralEIT (I think that's the real behavior)...
But still nobody answer me, are they modeled in common praxis?

PS: Maybe is my question too "dangerous" for open forum?

 

[Lion06]

I have never seen this done in practice. I'm not saying some don't, I just have never seen it.

 

[drile007]

Thanks StructuralEIT...you confirm me my thought!
But is there any clause in US codes which force to model them? There is a clause in Eurocode!

 

[Lion06]

I don't believe there is otherwise almost all shearwalls occurring at the corners of buildings would fall into this category (even though it would be a flange on one side only).

 

[prsconsultant]

Recently I went through an excercise to compare the wall reinforcing in a core shear wall utilizing flanges. By utilizing the flanges, the reduction in overall vertical reinforcing was about 30%. This number could vary case by case depending on the wall geometry. If you have PCAWALL or PCACOL it would be worth checking the difference.

 

[drile007]

Recently I went through an exercise to compare the wall reinforcing in a core shear wall utilizing flanges. By utilizing the flanges, the reduction in overall vertical reinforcing was about 30%.

But on which load you design them. If you design them on the same load that's the logical result. If you modeled the whole structure with flanges (and done dynamic analysis) you should get completely different forces on the wall from the model without flanges...and there's my point?

StructuralEIT and others:
Will you model the flanges if there is the clause in your codes to do so?

 

[prsconsultant]

We always do a 3D model as most of our structures are very complicated. It is incorrect to use a 2-D equivalent structure when you have interconnected shear walls, where torsional forces are predominant. To get the analysis forces it is better to use a 3-D model.

Once you get the analysis forces from a 3-D model, you can either design the entire wall as one section - most optimal solution (all loads are applied at the center of rigidity of the wall) or design each wall for its resolved component of the total force. In the second approach you can use the flange effects for design or ignore the flange effects.

 

[Lion06]

prsconultant-
We use RAM and I don't think you can model a flanged shearwall. Does your program allow you to do this? Even at the corners of the building, you model two seperate walls, not one wall going around the corner.

drile007-
If the code tells you that you must model it that way, then I guess you must. That being said, I don't think the IBC says that. Our projects use the IBC typically.

 

[drile007]

If the code tells you that you must model it that way, then I guess you must.

That's very easy to say or prescribe, but far far more difficult to achieve, especially without software who will do it automatically.

 

[prsconsultant]

For elastic analysis and design, there is nothing different in modeling of flanged shear walls vs non-flanged walls.
Take an example a elevator core - rectangular box, you can define each face as one or multiple wall elements, interconnected at the corners. If you are doing a second order analyses wherein you are required to account for multiple modes of bending, this method might not be correct as shell elements used for walls does not account for these effects. Typically in such cases the entire core is modeled as a flextural element.

For the design, you can define the entire core as one wall group and design this section for biaxial bending, shear and torsion (RAM reports the wall group forces at the wall CG). RAM version 13 (I am one of the alpha evaluator) is in the process of implementing this to their design. If you are running RAM V12, you can import the model to RAMADVANSE and do the design. The design in RAMADVANSE is capable of utilizing the flanges per ACI code.

Currently the only commercial program I am aware of is ETABS which has capability to design a 3-D section. This design is very optimal and can make a huge difference in a high rise building.

 

[Lion06]

Nice. We had our upgrade to RAM v13 just a couple of weeks ago. There are some good additions with it. The Direct Analysis Method is incorporated, as well as the 13th edition steel manual. RAM Beam will also design transfer beams now. It will take the column load reaction and put it on the beam in the gravity module. You used to have to play a RAM Frame trick on it.
I may come back and ask you another question about that at some point.