Shear strength of unstiffened steel I-section members

Ultimate shear strengths of 20 unstiffened I-girder specimens from the literature and seven new specimens designed to fill gaps in the experimental data are compared with predictions from models recommended by Basler, Höglund, and Lee and colleagues. Basler's model, which is based solely on an idealized web shear buckling resistance, is shown to be accurate for webs with slenderness between 60 and 100. However, it is very conservative for members with typical plate girder web slenderness ratios. Höglund's methods, which aim to characterize the shear postbuckling capacity of unstiffened webs, are found to be accurate to conservative. The method by Lee and colleagues, which aims to predict the shear postbuckling strength of webs with widely spaced stiffeners, is accurate on average. However, this method significantly overpredicts the strength of several specimens. A proposed adaptation of Höglund's recommendations from a 1997 article matches the current AISC shear buckling strength equations for stocky web panels, is very simple to apply, and provides predictions that are in between those of Höglund's methods. This adaptation of Höglund's equations gives shear resistances that are as high as 3.66 times the capacity from Basler's model. The specific characteristics of the different shear resistance equations are discussed, and recommendations for potential further improvement are offered. The interaction between the shear and moment capacities is found to be insignificant, although many of the specimens are subjected simultaneously to shears and moments close to the calculated nominal strengths. Resistance factors are computed to facilitate potential inclusion in modern specifications.