Optimization of Cold-Formed Steel Portal Frame Topography Using Real-Coded Genetic Algorithm

Abstract

Cold-formed steel portal frames are a popular form of construction for single-storey buildings of spans of up to 30 m. Such buildings typically use cold-formed steel channel-sections for the columns and rafters, with joints formed through back-to-back gusset plates bolted to the webs of the channel-sections. For such buildings, for frames of a given topography (where the span, height, pitch and frame spacing are all known) the choice of the cold-formed steel channel-section that results in either the minimum weight or minimum cost for the building can be determined easily for a given cold-formed steel manufacturer. This is because a typical cold-formed steel manufacturer would typically only have a finite number of discrete sections in their catalogue, and so in conjunction with frame analysis and design member checks, all the different permutations of sections can be considered. However, it is well-known that the minimum weight or minimum cost of a frame will not necessarily yield the most economical solution of the whole of building. In this paper, a study is conducted to investigate effects of frame topography on either the frame weight or cost per meter length of building. Thereafter, an optimization technique that uses a real-coded genetic algorithm is applied to search for the optimum topography of steel portal frame for a building. The advantage of using a realcoded genetic algorithm is that a high precision optimum solution can be achieved easily when the optimization process works in the continuous search space. The result of a parametric study that investigates the effect of typical wind zones in Australia on the optimum parameters of building geometry is presented.