Assessment of Eurocode methodologies for verification of flexural and lateral torsional buckling of prismatic beam-columns

Abstract

At present, EC3 provides several methodologies for the stability verification of members and frames. Each of these methodologies has its own scope of applicability. The rules provided in clause 6.3.1 are for flexural buckling of uniform members and are based on the famous Ayrton-Perry type of formulation which is of straight-forward application and it has a clear mechanical background. Clause 6.3.2 of the Eurocode deals with lateral torsional buckling of uniform members. In this thesis focus is given to prismatic beam-columns. They can be designed using clause 6.3.3 which gives the interaction formulae for stability verification of a member subject to bending moment and axial force; as an alternative – in clause 6.3.4 the General Method is suggested. However, application of this method has been shown not to be reliable in many scientific studies – either due to the lack of mechanical consistency or due to the lack of clarification in adopting certain decisions for non-standard situations. It was the purpose of this study to review the application of existing procedures for verification of flexural and lateral-torsional buckling of simply supported prismatic beamcolumns. This included recent proposals for lateral-torsional buckling of beams (Taras, 2010), the consideration of cross-section classification along member length (Greiner et al, 2011) and safety assessment and comparison of existing procedures in EC3-1-1 by a parametric study consisting of I-shaped cross sections ranging from class 1 to class 3, with different lengths, loading types (uniaxial bending with axial force) and height over width (h/b) ratios. Results from clauses 6.3.1 to 6.3.3, general method and GMNIA were calculated for 4144 cases. In comparison with results from 6.3.1 to 6.3.3, the general method is more unsafe. Its results were found to vary with increasing cross sectional slenderness and increasing length. On the other hand results from clauses 6.3.1 to 6.3.3 are more conservative. In comparison, the general method was found to have a higher spread as indicated by the standard deviation of its results than from 6.3.1 to 6.3.3. Results from general method varied when calculated from different alternatives in the general method. When interpolation criteria was used in the general method, the results were more un-conservative in comparison to the minimum criteria. The spread of the general method included both safe and unsafe results whereas the results from 6.3.1 to 6.3.3 remained mostly conservative.