Equivalent imperfections for the out-of-plane stability design of steel beams by second-order inelastic analysis

Abstract

In current structural design specifications, such as EN 1993-1-1 for steel and EN 1993-1-4 for stainless steel, the stability of members is typically assessed through the use of buckling curves, which consider the influence of initial geometric imperfections and residual stresses. An alternative, more direct, approach is to perform either an elastic or inelastic second-order analysis of the member or structure with imperfections. For modelling convenience, so-called ‘equivalent’ imperfections are typically utilised, which consider the combined influence of both geometric imperfections and residual stresses. Equivalent imperfections for the design of columns and beams by second-order elastic analysis, also referred to as geometrically nonlinear analysis with imperfection (GNIA), are provided in the current design specifications. For columns, equivalent imperfections for design by second-order inelastic analysis, also referred to as geometrically and materially nonlinear analysis with imperfections (GMNIA), were recently developed, but for beams that are susceptible to lateral-torsional buckling (LTB), there are currently no appropriate provisions. The aim of this study is therefore to develop equivalent imperfections for use in out-of-plane stability design of steel and stainless steel members by GMNIA. The proposals are calibrated against the results of benchmark finite element (FE) simulations performed on a large number of steel and stainless steel members with geometric imperfections and residual stresses subjected to major axis bending. Two proposals for equivalent imperfection amplitudes are developed: (1) e0,mod, for use with eigenmode-affine imperfections and (2) e0,bow, for use with sinusoidal bow imperfections. The latter is applied solely in the lateral direction and as a summation of a half-sine wave and a full-sine wave. Relative to the traditional Eurocode design calculations, employing the proposed LTB imperfections in GMNIA provides generally more accurate member resistance predictions, while remaining safe-sided relative to the benchmark FE results. The reliability of the design provisions is demonstrated through statistical analysis, where it is shown that partial safety factors of 1.0 for steel and 1.1 for stainless steel can be safely adopted.