Moment of Inertia for Complex Shear Wall Shapes (C / L / etc.) 1

[KootK]

I'm trying to use a predominantly shear wall system for a thirteen story post tensioned concrete building. I need to assess the stiffness of the shear walls.

I have concrete shear walls in both C & L plan shapes. In both instances, the "flanges" are about as long as the webs.

It seems to me that, because of shear lag effects, it might not be prudent to consider the entire flange length as being effective. The question then, is how much of the flange should I use?

The other engineers in my office seem to use one of two approaches:

1) Conservatively ignore the contribution of the flanges entirely.

2) Use the entire flange as effective. The idea being that concrete is good at transmitting shear and there will not really be much of a shear lag effect.

Does anybody have any advice that they'd be willing to offer?

Is there any code guidance of any kind available?

Thanks for your help,

Adam

 

[UcfSE]

Taking the entire flange as effective would be an unconservative assumption. I would either use no flange or possibly take a small portion of the flange, upto 6t, as effective. 6t is part of the requirement for flanges of "L" beams in chapter 8 of ACI 318 and is also consistent with the effective flange for masonry shear walls of the same section (ACI 530 1.9.4.2.3). It does not specifically say in that section of the 318 that the same should be used for shear wall flanges so you'll have to research it some or use your judgment. In any case the flange farthest from the web wil not "see" much if any load from the perpendicular direction. The simplified design put out by the PCA does not take into account any flange in its examples of shear walls for flexure.

 

[KootK]

Thanks for your response UcfSE,

Yeah, my first instinct was to go with the 6t provision as well. Would that mean that I'd really have to go with 3t then as I've only got flange on one side?

I thought that the 6t provision might be a little too conservative for this application. Here's why:

1) For analysing my PT slab / frame for lateral loads, I get to consider an effective slab width much greater than 6t. So there's some evidence for using greater widths in certain applications. I'm sure that it doesn't hurt that my slab is pre-compressed though either.

2) The shear wall flanges will be well braced at each story which amounts to a bracing spacing of about 9-11 times the thickness of the flange itself. Obviously flange buckling is not much of a concern here.

Any additional thoughts?

Thanks,

AdamP

 

[UcfSE]

I don't think the flange buckling is the concern but rather the non-linear stress distribution across the flange. It will be greatest at the web and reduce at a non-linear rate from there out. Instead of finding the area under the curve we average it out and take an "effective width" over which we assume a linear stress distribution. Once you get far from the web the actual stress is very small, far being defined by the code or by your nonlinear analysis. The 6t provision is for "L" beams that already have flange on just one side. The provision for "T" beams is 8t I believe, but there are also other requirements for finding the effective width for L and T beams, as you know.

 

[blake989]

When the added rigidity is needed, I use the PCI Design Handbook Section 3.7.1.2 and Figure 3.7.2. If you don't have access to the PCI, let me know and I will try to post the effective widths for you for different shapes.