Experimental and Numerical Investigation of Cold-Formed Steel Storage Platforms with Perforated Channel Beams

This paper presents the results of experimental investigations of the flexural capacity of the main beams in a storage platform with plan dimensions of 5.0 m × 2.0 m. The platform was designed as a grid structure, with main beams made of cold-formed channel sections and crossbeams made of sigma-section members. The crossbeams, spaced at 0.6 m, were connected to the main beams using M12 bolts and angle cleats. Eight test specimens were examined, with section heights ranging from 250 to 500 mm and wall thicknesses of 3 or 4 mm, fabricated from S350GD+Z steel. Loading was applied in a four-point bending scheme until failure of one of the main beams, while recording the moment–deflection relationship. The obtained failure loads were compared with the design resistances calculated in accordance with EN 1993-1-3. Additionally, GMNIA analyses were performed for the tested storage-platform structures using the Idea StatiCa Member version 24.1 software, incorporating measured material properties and equivalent geometric imperfections in accordance with prEN 1993-1-14. The adopted procedure included a sensitivity study to investigate the influence of different combinations of local and distortional buckling mode imperfections on the numerical results. The observed behaviour was characterized by interaction between distortional and local buckling modes, accompanied by yielding in the compression zone. The GMNIA analyses predicted the ultimate bending resistance with good accuracy, yielding FEM-to-test resistance ratios between 0.93 and 0.98. The sensitivity study indicated that the predicted ultimate resistance was only weakly affected by the assumed combination of local and distortional imperfection modes. For the investigated platform systems, the effective section approach according to Eurocode 3 provided accurate predictions of the ultimate resistance, with calculated-to-test resistance ratios ranging from 0.96 to 1.01. This agreement is discussed in the context of the possible stiffening effect of crossbeam-to-web connections.