Experimental program on the flexural response of steel joist top chord extensions

Abstract

This article describes an experimental program performed to investigate the stability response of the top chord member of open web steel joists extending as cantilevered members to support floor or roof extensions. The program included 15 full-scale specimens. Four different cross-sections were examined: single-C's, double C's back-to-back, double angle back-to-back, and cross-sections formed by a C-shape and an angle placed back-to-back. C and angle cross-section profiles were varied as well as the boundary conditions used for the top chord extension. The buckled shapes were examined using digitally scanned 3D renderings and monitored image correlation data. The ultimate strength, strains and the displacements of the extensions were measured to characterize their behaviours and be able to reproduce the test results with numerical simulations. Tensile tests and 4-point bending tests were performed to characterize the material properties. The influence of the restraint conditions and the type of section on flexural strength is presented. In all extension tests, failure took place at, or near the support by a combination of flexural yielding, lateral-torsional buckling (LTB) and local buckling. Extensions built with single C shapes without top flange lateral bracing failed by LTB and could not reach their plastic flexural strength. All specimens with top flange bracing could attain their plastic flexural capacity, although local buckling was observed near the bearing seats in most cases. For the deeper 2-C cross-sections, LTB of the individual C shapes also occurred near the support. Design methods for joist top chord extensions must account for these instability mechanisms.