Numerical Study on the Buckling Behavior of Austenitic Stainless Steel Unequal-Leg Angle Columns

A computational-based parametric study on stainless steel unequal-leg angles with various end conditions was conducted to assess the buckling behavior and comparison to ANSI/AISC 370 provisions. Experimental work completed at the University of Wisconsin–Madison subjected hot-rolled Grade 304 austenitic stainless steel unequal-leg angles, ranging in length from 10 to 148 in., to uniform compression with fixed supports. The angles failed in flexural-torsional buckling with variable degrees of flexural and torsional deformations. This paper reports on the finite element modeling validation of the flexural-torsional buckling failures and subsequent 2,880-model parametric study. The finite element analysis utilized a simplified approach that isolated the angle column with perfect fixed-fixed boundary conditions and incorporated measured material properties, cross-section dimensions, and geometric imperfections. This method accurately simulated the appropriate nonlinear stiffness, deflection patterns, and ultimate capacities associated with the torsion-dominated buckling failures.Flexure-dominated failures of the long specimens were not as accurately predicted by this modeling; however, the assumption of continuous positive contact and a perfect bearing surface was noted to be inaccurate. Further investigation noted that accounting for imperfect bearing and plastic deformation of the base plates from previous tests captured the reduced ultimate capacity and ductility of the column, similar to what was observed in testing. Therefore, the base modeling technique was adequate to conduct a parametric study that utilizes perfect supports. The evaluation of 12 nonslender-element unequal-leg angles, considering both nominal and measured material properties supported by fixed-fixed or pinned-pinned boundary conditions, provided additional data to qualify the behavior of unequal-leg single angles. Considering the nominal stress-strain response indicated in AISC 370, flexural-torsional buckling is consistently observed as the overall failure mode with transitions from flexural buckling dominated to torsional buckling dominated behavior. Comparisons to existing AISC 370 strength provisions indicated that the direct consideration of flexural-torsional buckling was necessary to conservatively predict the capacity. However, accounting for the measured stress-strain response resulted in a substantially higher tangent stiffness before yielding compared to the nominal response, which led to a change in overall behavior. Similar flexural-torsional buckling failures were observed, but the increased stiffness reduced the impact from the onset of increased torsional buckling participation at maximum load. The reduced design capacity from considering flexural-torsional buckling was not needed to obtain a conservative result. Instead, the flexural buckling provisions alone resulted in reasonable predictions of strength for the nonslender-element cross sections. This result is not enough evidence to change ANSI/AISC 370 provisions; however, it highlights the importance in confirming the nominal stress-strain behavior of hot-rolled stainless steel angles for compression design as this assumption could alter the importance of considering flexural-torsional buckling.