Buckling Length of Unbraced Frame Columns

Abstract

In the design of steel columns in unbraced frames, the current AISC specification commentaries from both LRFD and ASD contain an alignment chart to determine the K factor for a particular column. The K factor is based on the effective length concept where K factors are used to equate the strength of a compression member of length L to an equivalent pinended member oflength KL subjected to axial load only. The unbraced frame alignment chart is a graphical representation of a transcendental equation of a buckling solution of a subassemblage. This solution involves several assumptions limiting the use of the alignment chart to idealized cases not necessarily satisfying a particular practical situation. The aim of this study is to I) compare K factor values from frame instability analysis usmg structural software with values from the alignment chart in situations where the assumptions of the alignment chart are violated and 2) suggest application of appropriate known solutions to particular situations in which violations of the assumptions occur. Situations investigated are: variations in bay width, variations in column moment of inertia, variations in loading, and variations in column height. The nomograph perfom1ance was found to be relatively insensitive to bay width variation. Variations in column moment of inertia and column loading lead to large inaccuracies in the nomograph K factor values but Lui's method handled these cases well. The nomograph performance was found to be most sensitive to column height variation. Configurations with large variation in column height require system stability analysis to obtain accurate K factors.