Open Canopy Risa 3d model 1

[Hemi79]

So attached is a model I did in Risa V13. I have gone back and forth about how to model the loads. In the end I applied member area loads as opposed to plate loads. With plate loads, the transverse beams all have to be hot rolled W8x18 for them to work. IF I keep the Cee Channel 8x2.5x.07(14gauge) they all fail. However, now that i modeled the canopy with member area loads, I have the end members exceeding unity. I can't understand why. What I don't understand most is Member M67A for example shows Unity beyond 1.0 (i believe its over 4) and nominal moment easily exceeds applied moment. I cant find my error.

 

[JoshPlumSE]

Hemi -

There are a few issues that you may be dealing with:

Issue #1: Meshing of the plates.
The plate elements are not meshed properly. The loading goes directly into the plates and only transfers to members that connect to the corner joints of the plates. This will bypass the 1/3 and 2/3 span joists and go only to the end members. That may explain why the plate loading model did not have issues.

Issue #2: The members are all "fixed end" members. I would think with cold formed members, most of these would have BenPin end releases for the members.

Issue #3: Area load mesh size is set to 144in^2. For newer versions of RISA-3D this may not be as much of a problem. But, in version 13, the program was very dependent on the area load mesh size to get a reasonable distribution between members. General rule of thumb was 1/10th the span squared. So, you've got about a 4 ft span so, 4.8^2 = 23in^2.

 

[Hemi79]

Thanks for the Reply @JoshPlum, do you have any idea on the end members showing a code check of 4.303 for bending when the applied moment is minimal compared to capacity? Also, transient loads seem so small compared to what actual distributive loads would be based on tributary widths. Any thoughts?

Thanks, I will change the connections. I have already changed the meshing size,but for now I am applying area loads to members for the mode.

 

[JoshPlumSE]

Look at the member torsion results spreadsheet. You've got 60 ksi of shear stress due to torsion. This is most likely due to Issue#2 that I mentioned in my first response.

Now, a good question would be why the code check of 4.303 is given as a "bending check" instead of a "shear check". It's really a combined stress code check. It's usually meant for combined code check due to axial and bending. But, some codes have a shear component in these combined stress code checks. So, that's included as well.

I don't have the AISI code in front of me, but the program identifies the equation as C3.3.2-1. If we looked that up, I'm sure we'd see that shear is included in the code check.

 

[Hemi79]

It is and sometimes its Equation 5.2.2-1 which considers My,Mz and Pu. However, calculating the transient loads seen on the model, I realize it is the applied Area Member load x Width of the member, and not the width of midpoint between parallel members. I can't use this approach. That loading is not accurate to what the members would take if the cladding was placed. I applied it in a one way direction so that longitudinal members distribute to transverse beams and from there to columns. Thanks for your help and patience. I did make changes to mesh, I will post model later today in hopes you may be able to look at it.

Thanks,

 

[JoshPlumSE]

I noticed that you used the "open structure" area load option when I first looked at it. But, for some reason it didn't immediately click in my head that this was wrong. You wanted the regular member area loads instead.

The open structure loads are meant only to apply to the member's cross sectional area itself. Like when you've got an open structure subjected to a wind pressure. It is NOT intended to represent area load applied to decking on floor joists or such.