Alternate Evaluation Procedure for the Compressive Strength of Gusset Plates

Gusset plates are essential components in steel truss bridges and braced frame buildings, connecting structural members at panel points. Current design practices conceptualize a segment of the gusset plate near the connected brace as an equivalent compression member, necessitating assumptions about effective length factors and cross-sectional properties commonly defined by a Whitmore section. Also known as the line-of-action method (LOAM) or column analogy method, this procedure primarily relies on engineering judgment. Following the collapse of the I-35W Bridge in Minneapolis in 2007, extensive research efforts led to new design requirements for gusset plate buckling in the AASHTO LRFD Bridge Design Specifications. However, these requirements are still primarily based on the LOAM and include a new shear yield check, and the scatter between the predicted and experimental gusset buckling strengths remains large because the assumed and actual buckling modes are inconsistent. This paper introduces an alternate approach that models gusset plate buckling as a shear-induced plate buckling problem. The proposed model can capture the sway buckling mode initiating from the free edges of the gusset plate, a failure mode consistently observed in laboratory tests. Through finite element analyses, the proposed approach accounts for critical parameters including the effects of the brace connected length within the gusset region, the inclination angle of the brace, and the brace connection type (welded versus bolted). A correlation study using data collected from seven experimental test programs demonstrates that the proposed method significantly reduces the scatter when compared to conventional LOAM approaches. The proposed procedure also predicts the high likelihood of gusset failure at panel point U10 in the I-35 Bridge that collapsed in 2007.