INFLUENCE OF LOADING WIDTH ON WEB COMPRESSION BUCKLING OF STEEL BEAMS

This paper presents an experimental and numerical study of the behavior of steel wide flange sections subjected to loads causing compression buckling in the web. This research includes experimental investigation of the effects of load width and duration on web compression buckling. This data is then used to calibrate numerical models. Experimental investigations were conducted on specimens with load widths of approximately 2.5, 1.75, and 1.5 times their section depth. Loads sustained on the specimens had a magnitude of about 85% of the expected buckling strength to investigate creep effects near failure. Results of these experiments were used to calibrate numerical models for a parametric study.

The numerical parametric study examined 60 specimens of four wide flange sections, investigating the effects of loaded width and angle of load application on web compression buckling. The numerical models accounted for initial imperfections in the specimens by applying imperfections with a magnitude of 0.13*t_w to the first mode shape obtained from a linear perturbation analysis. This value of imperfection was chosen because it is the average imperfection measured in the experimental specimens and is likely a good representation of a typical wide flange section.

A prediction method is provided based on the data obtained from the numerical parametric study. This prediction method is derived from rectangular plate buckling solutions and considers the cases where the width of the concentrated load is a function of the section depth, and when the applied load is not orthogonal to the specimen. The current AISC 360-16 provisions do not directly address the influence of load width on the calculation of web compression buckling strength and refer to the design of compression members when the loaded width is greater than or equal to the section depth. The AISC approach was also evaluated and deemed conservative for design.