| Previous research (Bradford 1985) suggests that distortional buckling may contribute to a decrease in strength below that predicted by traditional buckling modes for steel I-girders (i.e., local buckling or lateral-torsional buckling, which may also be referred to as global buckling). However, the term distortional buckling is not well-defined for I-girders. Also, the significance of this failure mode for practical geometries of I-girders is presently unknown. Thus, strength prediction equations for distortional buckling in a format consistent with the current AASHTO (2005) and AISC (2005) design criteria do not presently exist.; The overall goal of the research is to address these present uncertainties regarding the failure mode of distortional buckling. First, a quantitative classification method is applied to cross-section displacement data from previous experimental testing to define distortional buckling. Unbraced lengths where interactions of local and global buckling are significant are then investigated using finite strip analysis. Lastly, the effect of distortional buckling on moment capacity for steel plate girders is studied using finite element analysis. Finite element analysis of modified experimental girders as well as a suite of parametric girders representing cases where distortional buckling is thought to be most significant is performed.; Results of this research indicate that at relatively large unbraced lengths, there is little difference in moment capacity between AISC and FEA results. Higher discrepancies occurred for girders with the highest values of monosymmetry parameters at relatively short unbraced lengths, where current AISC specifications over-predict moment capacity at unbraced lengths of or close to Lp. At this unbraced length, the moment capacity is a function of Mp, My, and web slenderness. However, based on this research it appears that monosymmetry is also an important parameter affecting moment capacity. |
