RISA Shear Wall Module

[CBSE]

I have started using RISA for doing my wood shear wall calcs, and so far I like what the module has to offer. I do have some heart-burn over some results.

Scenario: I have a 20ft long wall, 12ft plate heights, and (3) 3'-6" shear wall segments. (1) shear wall at each end, and (1) shear wall in the middle. The opening heights are 7ft tall, and are 4'-9" each. I have a seismic load of 3,400 lbs roughly at this wall line. I'm using 7/16" OSB sheathing (which, apparently I'm limited to "S1" which must be structural 1?). When I place my point load at the top of the wall, the first wall segmented gets blown out of the water for capacity. This intuitively does not make sense to me, but when I place my loads in my spreadsheets, I place it as a load and it is evenly distributed to my walls.

Now, if I divided that load into a uniform load across the wall, then I get results that are more in line with what I'm going to say is the accepted design standard for distributing forces to shear walls for a simple wall line.

What gives? Do I need to divide my load into a uniform load every time I do a wood shear wall? It's not much more time, but kind of a pain in the butt!

Thanks

 

[JAE]

Isn't the load actually coming into the system along the full length?

If you lump it all into a point load at one end then the relative stiffnesses and the finite element analysis probably presumes the load all goes to the front panel.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[CBSE]

That's what I'm assuming is happening with the relative stiffness's. It just kind of threw me for a loop when I had one wall that wouldn't work and I spent about 3 hours trying to figure it out, and then decided to do a distributed load along the top and got the results I expected. I guess I shouldn't have assumed that if I put a point load at the top of the wall it would automatically distribute the way I "think" it should...dang those "ass"umptions!

 

[JoshPlumSE]

I was away from the office for a few day or else I would have responded sooner. Previous responders have it basically correct. But, replied to this same question in tech support for awhile. So, hopefully my practiced response is a bit more clear. When you apply the load as a point load at one corner, the following happens:
1) Initially all the load goes into the top of that wall where the point load is. That's likely fairly obvious.
2) The load the proceeds to travel based on relative stiffness. Whether the force moves down to the base of the wall or laterally through that first wall into the other walls is based on relative stiffness. The stiffer component will received more of that load.
3) So, if you look at your deflected shape, (probably at an amplification factor higher than the default 40) you should see some local squishing of the wall in that region right next to that point load. This will not happen in reality. However, it is a result of the way you applied the load as well as a result of the way RISA models the walls. That squishing shows you the localized softness to point loading in that wall which is preventing more of that load from getting into the other walls.

The reality is that a shear wall is a combination of studs, top plates, sill plates, and shear panels. RISA doesn't model all of this. Instead, the program uses a series of orthotropic plate elements so that the total lateral stiffness (caused by the shear panels) is represented by the shear stiffness components. And, the vertical stud stiffness is represented by the vertical/ axial stiffness of the plates. But, it's important to realize that the discrete stiffness that occurs at stud locations or the top plate and sill plate locations is not directly modeled. So, if you apply a vertical point load where the top plate would be, the program resists this through the lateral resistance of the plates. Which were developed more for the proper shear resistance of the overall wall rather than the compressive resistance the top plate.

The solution to this is usually the following (which I think you have already figured out):
A) Apply the load as a distributed load along that shear line.
B) Add a rigid diaphragm at the top of the wall which will ensure that there is no localized squishing of any of these walls.
C) Add in members at the top of the walls that represents the top plates. This could be a doubled up 2x4 or 2x6 to represent the real stiffness of those top plates (which is what I would personally do). Or, it could be a rigid link if you want to get results similar to what you would get for the rigid diaphragm... But, limited to this single shear line rather than the whole floor. This will probably get you closest to your hand calc assumptions.