The literature seems unclear on this one. In the picture below, every other purlin is attached to the bracing system. Do I consider the top chord critical length 2L in this case or 1L (L being the distance between two nodes)?
Thanks for the reply! My cladding is very weak, so I am assuming 2L then...
I was wondering about the roller issue. Why do you say the result is wrong if in reality there is no roller there? To me this is just a matter of statically determinate or indeterminate, not right vs. wrong. What do you think?
If the real structure lacks actual pin supports capable of restraining both vertical and horizontal reactions, modeling them as pins will give you incorrect results. The truss model will arch towards the pins instead of functioning through bending.
If the roof sheathing has low diaphragm stiffness, then only the purlins at the roof braces might laterally brace the truss top chord.
It's not a given that the roof bracing system is strong and stiff enough to brace the truss. It would be evaluated as a panel bracing system per the AISC Specification Appendix 6, Section 6.2.1.
Thanks, what about Eurocode, is there something similar to AISC 6.2.1 in Eurocode? I could find specifications only for lateral torsional buckling of purlins as a function of cladding, which is not the same issue.
I'm not familiar with Eurocode, so I don't know if it has equations for this. If not, then presumably this is a "use a rational analysis" situation. The AISC method, perhaps with factor-of-safety modifications, should qualify.
If you'd like to dig into the background, then Yura and Helwig have several papers and seminar series online.