Stability interaction effects in 3D steel frames : a case study

3D steel frames are usually assessed by checking their columns and beams separately using design rules, e.g. EN 1993-1-1 (Eurocode 3). For this, the force distribution in the 3D steel frame is determined first, followed by cross-sectional resistance and member stability checks. When increasing the load on the frame, the first critical check defines the ultimate load. This paper presents a case study in which a 3D steel frame is assessed both by code checking and by more realistic numerical simulations by the finite element method (FEM). It shows that the code checking approach overestimates the ultimate load of the FEM approach. The FEM simulations show simultaneous failure of beams and columns and naturally take into account the mutual stiffness interaction of unstable beams and columns. As the latter is not the case for the design rules used in the code checking approach, the suggestion is made that the current design rules predict too high values for this reason. As similar findings are obtained for 2.5D and 2D frames, further research is needed, including full scale experiments.