Experimental and numerical study on reinforced single-coped beam with slender web connections

Abstract

A full-scale test programme for reinforced single-coped beams with slender web (RSCBSW) connections is reported in this paper. The test matrix included reinforcement types and coping details. The adequacy of the two reinforcement strategies was confirmed by the test results, and it was found that all specimens failed by buckling of the plate girder section away from the coped region. The test results also showed that for the reinforcement strategy with double full-depth transverse stiffeners, the decrease in the strength of the specimen in the post-ultimate stage was comparatively less significant compared with the specimens strengthened by a single full-depth transverse stiffener, and this phenomenon was more apparent in the case of a deep cope. Subsequently, finite element (FE) models of test specimens were developed to replicate the behaviour of the RSCBSW connections, based on which a parametric study covering a wider range of factors was conducted. The numerical results further confirmed the effectiveness of the reinforcement strategy, and no models experienced local web buckling in the vicinity of the cope. The available design equations to quantify the buckling resistance of the plate girders were applied to the test specimens and the numerical database. It was found that for cases with double full-depth transverse stiffeners, the design predictions were generally conservative. In contrast, unsafe predictions were obtained for several cases with a deep cope which was enhanced by a single full-depth transverse stiffener.