Calibration of the Resistance Factor for Beam-Columns in Cold-Formed Steel Members

Resumo

The Direct Strength Method (DSM), used for the design of cold-formed steel structural members, is based on the analysis of local, distortional, and global elastic buckling of the cross-section, combined with empirically derived direct expressions to estimate ultimate strength. This method is widely accepted and incorporated into national design standards, as it enables a unified, robust, and flexible approach to structural design. Specifically for beam-columns, current DSM-based design specifications still rely on simplified linear interaction expressions, which combine axial and bending strength responses separately. With the advancement of computational tools—such as the finite strip method—it is now feasible to accurately determine local, distortional, and global elastic buckling under any combination of axial load and bending moment. This allows for a more realistic stability assessment that directly accounts for combined actions. In this context, Torabian and Schafer (2018) proposed new DSM expressions specifically tailored for the design of beam-columns, incorporating the effects of combined compression and bending directly into the stability evaluation. The authors also conducted an extensive experimental program, complemented by data from the literature, to validate the performance of the proposed formulations. The development of the DSM for beam-columns represents a significant step forward in achieving a more realistic and mechanically consistent design approach, while also reducing the excessive conservatism inherent in traditional interaction-based expressions. This study presents a structural reliability analysis based on the First-Order Reliability Method (FORM), applied to the enhanced DSM model proposed by Torabian and Schafer for stiffened channel beam-columns under combined axial compression and bending. Using the experimental and theoretical dataset provided by the authors, a statistical analysis of model error is performed, followed by the calibration of the resistance factor, assuming a target reliability index of 2.5. The aim is to contribute to the refinement of design codes through statistically grounded safety calibration, aligned with structural reliability principles.