Unbraced Length of Members w/ Intermediate Nodes

Say I have columns with intermediate nodes that are not stability bracing points (for example - nodes where cold formed girts are connected). When using SELECT command - how to instruct program to account for the actual height of columns and to ignore intermediate nodes?

I am using AISC LRFD so the design of physical members is not supported in this case.

thanks

  • Len,

    In earlier versions of STAAD, there used to be 4 parameters for defining unbraced lengths - LX, LY, LZ and UNL. The first three were for column buckling as explained in my earlier response, the 4th for beam bending. At some point during the program's evolution, UNL was replaced by UNT and UNB to cater for the possibility that the top and bottom flange would not be braced in the same manner, such as in a metal deck where the top flange of the beam is fully laterally supported but the bottom flange is braced only where it frames into a column.

    Thus, there are a total of 5 parameters now.

    Regarding the last sentence in your response - why are there 2 parameters - one called UNT/UNB, the other called LX (STAAD terminology), both representing the same distance?

    Answer:

    2 reasons.

    1) Remember that LX is a term used to identify torsional buckling as a column. UNT/UNB is a term for identifying the lateral torsional buckling of the compression flange when the beam is under bending. So, conceptually, using a single variable to identify two distinct behavior patterns (axial compression versus bending) would not be right.

    2) As many steel design documents point out, the flexural-torsional buckling mode (for column compression capacity checks) should be considered only when the shear centre does not coincide with the centroid. Consider an American or Canadian W shape. LX should be zero because it does not have the FTB mode as a column. But UNT/UNB need not be zero since the code clearly says it (Lb) represents the distance between points where the compression flange is braced against lateral displacement.

    So, here we have a case where LX=0 but Lb is non-zero.

    Thus, it is necessary and convenient to have separate parameters to represent those distinct terms, at least from the standpoint of the AISC and Canadian S16 codes.