Shape optimization of cold-formed steel beam-columns with practical and manufacturing constraints

This study aims to present a practical method for optimization of symmetric cold-formed steel (CFS) beam-column members using Genetic Algorithm (GA). To eliminate impractical cross-section shapes from the optimization results, a range of manufacturing and construction constraints are incorporated into the optimization process. Axial forces are applied with different eccentricities (0 to 30 mm) to cover the full spectrum of beam-column actions from pure axial compression to pure bending. The effect of element length on the optimization results is investigated by using short, intermediate and long beam-column members. A total of 132 beam-columns with different cross section shape complexity (4 to 12 rollers/nodes and 1 to 3 lips) are optimized. The compression and bending moment strengths are obtained based on direct strength method (DSM) using CUFSM software by accounting for local, distortional and global buckling modes. The results show that using more complex shapes does not necessarily lead to better design solutions. Increasing the eccentricity generally leads to more spread optimum sections particularly when distortional buckling is the predominant mode in short and intermediate-length beam-columns. In cases where local and global buckling modes govern the design, however, less spread sections with higher turn angles generally provide higher strength capacities. With the variation of eccentricity, the ultimate strength of optimum beam-column sections normalized by the strength of a reference lipped-channel are in the range of 110–163%, 128–194% and 160–222% for short, medium and long members, respectively. The results of this study, should prove useful in more efficient design of CFS beam-column elements in practice.