Behaviour and design of optimised cold-formed steel structures

Cold-formed steel (CFS) structural members are fabricated at room temperature from thin gauge steel sheets. They offer a higher strength-to-weight ratio than conventional hot-rolled steel products and have lower embodied carbon. Moreover, they are light-weight and easier to handle, transport and install. Due to their suitability of prefabrication into panels and standardized systems, they are suitable for modular systems and offer unrivalled construction speeds. However, literature study exposes a need for more efficient CFS structural components, particularly on bolted-moment connections, to improve performance, minimize material use and increase their environmental and economic benefits. This study aims to develop more efficient CFS structures at three different levels (i.e. element, connection, and frame), and to investigate the behaviour and design of CFS bolted connections subjected to both static and seismic loads.
At the element level, stiffness and buckling strength of CFS elements at Serviceability Limit State (SLS) and Ultimate Limit State (ULS), determined in accordance with the Eurocode 3 effective width method, were first optimised, respectively, and the efficiency of the optimum results were assessed using experimentally validated Finite Element (FE) models. CFS elements were then optimised based on their post-buckling behaviour by establishing a link between an optimisation algorithm and Python script in ABAQUS FE analysis. The optimisation process was carried out using either Particle Swarm Optimisation (PSO) algorithm, Genetic Algorithm (GA), or Big Bang-Big Crunch optimisation (BB-BC). The total coil width of the steel plate and its thickness were kept constant during the optimisation procedure to use the same amount of material in all cross-sections. The results demonstrated that noticeable improvements were achieved in stiffness, buckling strength and post-buckling behaviour of CFS element.
At the connection level, the behaviour of various CFS bolted connections under both monotonic and cyclic loading conditions was investigated using experimentally validated FE models. Extensive parametric studies were also conducted using the validated FE models by considering a wide range of design variables in order to eventually propose general design equations for the capacity of CFS bolted connections and identify the most efficient design solutions which considerably improve their seismic performance. In addition, the seismic performance of CFS bolted moment connections was further improved using optimised CFS beam elements with enhanced non-linear post-buckling behaviour. PSO algorithm was innovatively linked to Python script in ABAQUS FE analysis to optimise CFS bolted connections with respect to their energy dissipation capacity and ductility.
Finally, the efficiency of the optimised elements was then assessed at the frame level by incorporating them into CFS structural systems. It was shown that the proposed optimum design methodology at the element and connection levels leads to more resilient CFS moment-resisting portal frames with minimum structural weight.