Abstract:To improve the stability bearing capacity of cold-formed steel composite columns and further expand the application scenarios of cold-formed steel components, an innovative cold-formed steel-solid waste foam concrete (CFS-SWFC) special-shaped composite edge column was proposed. Unlike traditional welded steel tube concrete components, cold-formed steel composite columns are susceptible to buckling due to their varying wall thickness and are assembled using self-tapping screws, making the interaction between the cold-formed steel and the core concrete unclear. Four special-shaped hollow columns and six CFS-SWFC special-shaped composite section columns were tested under axial compression to compare and analyze their buckling mechanisms and failure modes. A numerical analysis model of CFS-SWFC was established. Based on experimental validation, a multi-parameter extended analysis was carried out to study the effects of strength, wall thickness, and cross-sectional dimensions on the bearing capacity of the specimens. A calculation method for the bearing capacity of CFS-SWFC was proposed based on the current code GB 50936—2014. The results indicate that the use of solid waste foam concrete enhances the stability and bearing capacity of the specimens by 271%. Although the increase in concrete strength leads to a maximum decrease in deformation capacity of 18%, the final failure mode remains largely unchanged. The strength of SWFC, the thickness of CFS, and the cross-sectional dimensions have a significant impact on the ultimate bearing capacity. Notably, for larger cross-sectional components, an increase in SWFC strength results in a relatively higher enhancement of the ultimate load-carrying capacity. There is a discernible interaction effect between the composite edge columns and the solid waste foam concrete. The current calculation method, which takes the yield of the steel tube as a prerequisite, is not applicable to this type of cross-section. After modifications, the proposed formula demonstrates a good correlation with the experimental results, with a maximum error of 13%.