Buckling curves for heavy wide flange QST columns based on statistical evaluation

•Existing buckling curves were proposed to check heavy QST steel columns.•The assignment procedure was based on FE analyses and according to EN 1990.•A distinction was made between stocky (HD) and more slender (HL) sections.•Buckling curve a0 (strong-axis) or b (weak-axis) should be used for HD columns.•HL columns can be designed with buckling curve a (strong-axis) or b (weak-axis).

This paper proposes the use of existing European buckling curves to check the resistance of heavy wide flange quenched and self-tempered (QST) sections made from high-strength steel, failing by flexural buckling. The buckling curves are evaluated according to the statistical procedure given in Annex D of EN 1990 using finite element analyses. The residual stress model as described in a related paper was used to define the initial stress state of the column in the finite element model. A large database was created containing the ratio between the elastic–plastic buckling resistance obtained from finite element analysis and the buckling resistance obtained from the proposed buckling curve for a wide set of column configurations from which a partial factor γRd was deduced. Two different section types were investigated: the stocky HD and more slender HL type, featuring a height-to-width ratio (h/b) of approximately 1.23 and 2.35, respectively. Based on the criterion that the partial (safety) factor γRd should not exceed 1.05 it is suggested to check the buckling response of a heavy HD section according to buckling curve “a0” or “b” when failing by strong-axis or weak-axis buckling, respectively. HL sections are to be designed according to buckling curve “a” for strong-axis buckling and buckling curve “b” for weak-axis buckling.