Large-scale Lateral–Torsional Buckling Tests of Welded Girders

Abstract

Lateral–torsional buckling is a stability-related failure mode of unbraced or insufficiently braced flexural members. In recent years, several concerns have been raised regarding the adequacy of the Canadian steel design standard, CSA S16-14, in characterising the lateral–torsional buckling resistance of members made up of three plates welded together into an I-shaped section, as is commonly done for deep girders. The large heat input from welding results in residual stress distributions that may cause these welded sections to be more susceptible to lateral–torsional buckling than their rolled counterparts, particularly in the inelastic region. However, the paucity of up-to-date physical test data for modern welded girders considerably limits the potential for an adequate evaluation of the current design provisions. To improve the understanding of lateral–torsional buckling behaviour, an experimental programme was developed, consisting of 11 welded girders with unbraced spans of 9.75 m (32 ft). Test specimens were selected and designed by considering various parameters that may influence lateral–torsional buckling resistance. Preliminary finite element simulations were conducted to anticipate displacements and rotations of the test specimens, which were used to inform the design of an experimental test set-up capable of accommodating the full range of expected movement. The observed moment resistances and stability response from seven tests were analysed, and the effects of initial geometric imperfections, residual stresses, and geometric parameters were considered. Test results were compared against predictions from CSA S16-14, with adjustments applied to account for load height effects from the gravity load mechanism, and the adequacy of the Canadian provisions for lateral–torsional buckling resistance was evaluated.