Local stability of high-strength aluminium alloy T-sections

• The local buckling behaviour of high-strength aluminium alloy T-section stub columns was studied. • Tests were conducted on 16 high-strength aluminium alloy T-section stub column specimens. • The developed FE models were validated against test results and then used for parametric studies. • The codified design rules were assessed based on test and FE data. High-strength aluminium alloys have been attracting increasing attention in structural applications owing to their outstanding mechanical properties and corrosion resistance. T-sections, commonly used in purlins and lattice tower braces, are typical monosymmetric profiles prone to complex buckling and instability issues. Given the limited investigations into the local stability and design of aluminium T-section columns, this paper explores the local buckling behaviour and resistance of high-strength aluminium alloy extruded T-section stub columns through both experimental and numerical studies. An experimental programme was conducted, comprising sixteen stub column tests, material coupon tests, and measurements of membrane residual stresses and initial local imperfections. Subsequently, finite element models were developed and validated using the test data, which were then employed for parametric studies across a broad range of cross-sections. The accuracy of local buckling design rules in current specifications for aluminium alloy stub columns, including the Effective Width Method specified in the European code and the Weighted Average Method and Direct Strength Method outlined in the U.S. standard, was evaluated using experimental and FE data. The findings revealed that (i) the Eurocode exhibited a high level of conservatism for aluminium alloy T-section stub columns that failing by local buckling; (ii) the codified Weighted Average Method provided more accurate and consistent results compared to the Effective Width Method and the Direct Strength Method; and (iii) the codified Direct Strength Method yielded accurate and consistent failure load predictions for stocky cross-sections with small local slenderness but more conservative and scattered predictions for slender cross-sections.