The external envelope of steel framed industrial buildings normally involves the use of purlins and rails spanning between the main hot-rolled frames to support the roofing/cladding. These purlins are typically light-gauge cold-formed steel members of complex shape for which the thinness of the material means that local instabilities will significantly influence their structural behaviour. In this thesis, the finite element (FE) method (ABAQUS) has been used to develop numerical analyses to study the buckling behaviour and degree of moment redistribution in continuous and sleeved cold-formed steel 2-span purlin systems. Five types of nonlinear FE analyses have been validated against reported physical tests: (i) continuous 2-span beams subjected to uniformly distributed load (UDL), (ii) single span beams subjected to a moment gradient, (iii) single span beams subjected to pure bending (iv) sleeved 2-span beams subjected to a UDL and (v) single span sleeved sections subjected to a moment gradient. The FE analyses were used to generate a large portfolio of new results for gravity and uplift loading for continuous and sleeved 2-span arrangements covering a wide range of cross-sections by varying the flange and web dimensions and material thickness. The effects of local and distortional buckling and limited rotational capacities for single span FE models were investigated. The 2-span FE results formed the basis for a simple modification to conventional plastic design that recognises the possibility of a reduction in moment with increasing rotation in the interior support region. The assumption of full moment redistribution for gravity loading was found to be only valid for stocky sections but not for slender sections. For uplift loading in addition to the potential reductions in moment at the interior support, limitations in the span moment due to lateral torsional buckling (LTB) for slender members were also accounted for. Based on the FE results, an α-reduction framework was established to predict the collapse load for continuous and sleeved 2-span systems. It was assumed that the cross-section or LTB resistance was achieved in the span while a reduced cross-section resistance allowing for the post-peak fall in capacity was achieved at the interior support. The accuracy of the proposed design method was compared against elastic and full plastic design cases by considering their ultimate load carrying capacities. Whereas the elastic design method provides overly-conservative results and plastic design overestimates the capacity of slender sections, the proposed design method gave accurate predictions of the failure load with minimal scatter for all cases. The developed α-reduction framework provides a foundation for allowing the use of other purlin sections and interior support connections by inserting alternative cross-sectional moment capacity inputs obtained from several sources such as physical testing, hand calculations from design codes and FE analyses.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:659497 |
| Date | January 2014 |
| Creators | Hui, Chi |
| Contributors | Gardner, Leroy; Nethercot, David |
| Publisher | Imperial College London |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/10044/1/25533 |
Page generated in 0.0021 seconds