Design method for transverse load resistance of web-posts in cellular beams

The effect of transverse loads on the buckling resistance of the web-post between circular openings is assessed by finite element (FE) models that are calibrated against the results of a test on a cellular beam with transverse loads applied above two adjacent openings. The test failure load of the web-post was 250 kN of which 96 kN is attributed to web-post buckling and the rest to the transfer of load by local bending of the web-flange T-sections. The FE modelling showed that buckling of the web-post subject to transverse load can be predicted using the method of BS EN 1993-1-5 with a buckling coefficient that is a function of the edge spacing of the openings to the beam web depth. The FE results considered a range of geometric parameters and two steel strengths. It was found that the stresses across the web-post are non-uniform and a formula for the effective web-post width is presented. With this modification, the ratio of the predicted web-post buckling resistance to the isolated web-post buckling load obtained from the FEA was in the range of 0.80 to 0.99. The load transfer by local bending of the top Tee was determined from the elastic bending resistance of the Tee and the ratio to the failure load from the FEA was conservatively in the range of 0.62 to 0.77. Based on this study, a design formula for the web-post buckling resistance of cellular beams to transverse loads is derived. • First study on transverse load buckling of cellular beam web-posts. • Validated FE models predict buckling with Eurocode plate method. • Proposed formula using effective web-post width improves accuracy. • New design rule derived for safe web-post resistance to transverse loads. • The range of application of the study was for typical cellular beams (ho = 0.7–0.8 h).