Starred rolled stainless steel angle sections under compression: An experimental and numerical investigation. (January 2021)

• Record Type: Journal Article
• Authors: Botelho, I.S.; Kayser, A.M.G.; Sarquis, F.R.; Lima, L.R.O.de; Vellasco, P.C.G.da S.; da Silva, A.T.; Rodrigues, M.C.; Diniz, M.G.
• Is Part Of: Thin-walled structures. Volume 158(2021)
• Publication Date: 2021-01
• Subjects: Austenitic stainless steel -- Starred angle sections -- Torsional buckling -- Experimental analysis -- Non-linear finite element simulations; Thin-walled structures -- Periodicals; 690.1
• DOI: 10.1016/j.tws.2020.107177
• Languages: English
• ISSNs: 0263-8231

Abstract

Abstract: Angle sections have been extensively used in bridge bracings, communication and power towers, trusses and bracings diaphragms. Within these cases, a commonly adopted configuration is the starred section, usually achieved with an association of two equal-leg angles connected by welds or bolted plates. On the other hand, stainless steel has been increasingly being adopted in structural engineering due to its high versatility, high corrosion and excellent mechanical properties. This paper focus on the torsional buckling behaviour of A304 austenitic stainless steel starred angle rolled sections behaviour under compression. The adopted methodology involved an experimental campaign totalling eight tests. These tests were performed on L63.5 × 63.5 × 4.76 rolled sections in starred angle section assembled in three configurations: angles without internal connections; angles connected by welding; and angles connected by bolted plates. The experimental campaign was followed by comprehensive finite element simulations, considering geometrical and material nonlinearities, made to enlarge the database involving a torsional buckling failure mode. Subsequently, the results were also compared with Eurocode 3 (EC3) and two alternatives (Alpha and Lambda) design methods. The EC3 prediction presented more significant discrepancies compared to experimental and numerical results, confirming this code's natural conservatism. These discrepancies were not observed when the alternative (Alpha and Lambda) design methods were utilised to forecast the torsional buckling ultimate loads indicating a significant performance improvement. Highlights: An experimental programme was carried out comprising starred rolled stainless steel angle cross-sections columns. A numerical model based on finite element method was calibrated against experimental results. An extensive sensitivity analysis was performed in order to enlarge the database of the studied structural elements. The Eurocode 3 (EC3) and two novel design procedures based on the Eurocode 3 – part 1.4 methodology were proposed.