Full-scale experimental study on the stability of chords of cold-formed steel c-section roof trusses

Abstract

The use of cold-formed steel (CFS) roof trusses is growing as a substitute for wood because of their advantages in quick fabrication, high strength-to-weight ratio, and lightweight nature [1]. However, the torsional buckling behaviour of these trusses remains inadequately understood. There have been experimental studies conducted for the behaviour of CFS roof trusses [2], wide-span roof trusses [1] and small-scale roof trusses [3]. The common failures in the previous studies of CFS roof trusses include distortion of heel plates, local buckling of the top chords [3] and flexural-torsional buckling becoming a concern in elevated temperatures [4]. The behaviour of CFS under loads applied away from the shear centre requires more testing due to its thin nature. Through eight full-scale experiments, different truss configurations (back-to-back and linear), different lateral restraint spacing and internal support inclusions, this study investigates the structural behaviour of lipped channel (C-section) chords in CFS Howe roof truss assemblies. The experimental total load at failure exceeds factored predicted capacities by 12% and 34%, and factored design capacities by 34% and 60%, depending on lateral restraint spacing. Design equations, however, are conservative with predicted-to-experimental capacity ratios as low as 0.7 for wider spaced lateral restraints and 0.5 for closely spaced restraints. In 37.5% of cases, the design standards fail to predict the correct failure modes. Observed failures during the experiments include lateral-torsional buckling, out-of-plane buckling, and inward torsional buckling. The single-channel linear truss system (face of web connected to back of chords) proved more robust than back-to-back system (back of web connected to back of chords), offering better torsional restraint and load-bearing capacity post-failure. Truss strength is enhanced by increased lateral restraints, but current design standards lack provisions for calculating member lengths with such restraints. Therefore, further research, including FEM analysis, is needed to address this gap and improve design accuracy.