|
modelling bracing
i just had a question about how people model cross bracing (vertically and horizontally). do you use just two members? is the connection between the cross braces released of moments? i was going through the staad examples and they use two members and i didn't see any releases at the joint where the two braces intersect? what is the correct way to obtain reasonable results?
model as truss
consult your staad manual. how you actually treat this depends on whether your bracing takes both compression or tension.
if your bracing resist compression then model the joint where the bracing crosses to reflect your joint detail. risa-3d allows you to put a joint in both members where they cross. by slaving the joint to each other both braces will translate together but rotate independent.
in risa-3d if you put one joint in the middle of the x for tension only bracing, and ran four members to it two of your
the way bracing members are modelled will depend on the analysis package you are using so it's best to consult the manual.
the analysis package i use is a 3d-frame element package where it is the users responsibility to define the end-release for each element e.g. to create a bar element, the user must release the element about y and z axis bending (assuming x is torsional).
i would also model the bracing as tension only, if the bracing member is allowed to take compression than you may receive error messages saying that your frame has buckled somewhere.
i would also play around with the flexural stiffness of bracing members as well, if you are using rod bracing, these will be pretensioned by turnbuckles, which help remove any sag under gravity. otherwise the analysis may return ureasonable results on flexural displacements that just aren't realistic. |
|