Hi all,
I am at a total loss on how to properly model a flexible metal deck diaphragm in RAM Elements 14.0. My understanding from reading other posts on this topic is that using shell elements are a bad idea (I can link these posts if desired). Using area loads also doesn't work because I need to essentially have all the nodes on my roof level to behave as if they are restrained from the diaphragm, and area loads only give me a way to apply roof dead/live load.
To clarify, I do not want/need RAM to design my diaphragm for me. That is easy enough by hand and makes things complicated. I just want to model my structure such that when I apply, for example, a uniformly distributed LATERAL load to the roof beams of my structure, that the load will not bend those beams about their weak axis, but instead transfer that load over to my collecting elements as distributed axial load (diaphragm shear).
If this is not possible, then what is the easiest alternative to achieve a similar result? How do other engineers work around this?
OK, that's what I thought. We normally use RSS, but there was some concern about the validity of results for horizontal braces in RSS, and I was asked to make an alternate model in RAM Elements to crosscheck and determine the load on them. The building has an untopped metal deck with no interior lateral frame lines, so the idea was to do a horizontal truss along the long side of the building only. But the axial loads in the horizontal truss adds significant load to the vertical braces in the exterior walls making it more of 3D problem. For instance, a horizontal truss web might have a 100k-200k axial load in Elements under wind perpendicular to the long side of the building. That then induces a large load in the vertical bracing in the case where the wind is perpendicular to the brace line, which would not normally be the case. But the horizontal brace loads are MUCH less in RSS. In RSS, I have a semirigid diaphragm set for the roof so I can play with the stiffness and go from essentially flexible to essentially rigid and see the impact on brace loads and drift. In Elements, I'm using area loads to distribute my roof loads and wall pressures, and a thin shell with a custom material using a modified modulus of elasticity for the roof membrane, which seems to mimic the semirigid diaphragm of RSS qualitatively, but not quantitatively. For instance, if I use a modified E of 1,000ksi up to 24,000ksi in my noncomposite deck properties in RSS and the same values for E in my custom material in Elements, I get similar decreases in brace force and building drift as I increase deck/shell stiffness, but horizontal truss forces in RSS are maybe only 10-20% of the values in Elements. Vertical brace forces seem to be closer, around 70% of the Elements values. Lower brace forces would be nice but I want to make sure that's justifiable and that the deck really is contributing as much as RSS indicates. Also, in RSS, we've been modifying E to E' as per the example on this site when using semirigid diaphragms. But do we need to be using the stiffness modification values as well to account for the one-way action of the bare steel deck or is it still accounting for the defined deck direction in the semirigid analysis when those factors are not enabled? I've never used those before and just wondering if I should be. Would RSS be treating the semirigid deck as uniformly bi-directional (like Elements) if I don't adjust those stiffness modifiers?Thanks.
Not in Ram Elements. Diaphragm shells with non-uniform bi-directional behavior are possible in RAM SS and Ram Concept though.
Is there a way to model shells in Elements with different properties in orthogonal directions or are they only homogeneous? For instance, if you model an untopped metal roof deck as a thin shell in Elements, it would be just as stiff both directions, would it not?
If you have already selected a deck for your diaphragm and the attachment pattern, most deck manufacturers publish tables to calculate the diaphragm shear stiffness (G' in Vulcraft tables e.g.). I would then determine the size of fictional flat plate elements which match the stiffness. The plates will be extremely thin, but it will capture the load path.
Any suggestions for when the diaphragm is sloping? I have the same issue of wanting to apply a lateral load to the perimeter beams, but my roof slopes, and Elements won't allow different elevations for nodes in a rigid diaphragm. If I model the roof as flat at the max roof height, my braces on the low end overstress in compression with the size brace I'd like to stay within. If I model it flat at the low height everything works, so I can envelope it that way, but is there a better way in that case? Translate my wind loads to nodal loads instead distributed loads?