Full-scale long-term experiments of simply supported composite beams with solid slabs

This paper presents an experimental study aimed at the evaluation of the long-term behaviour of composite steel-concrete beams designed with partial shear connection formed by a steel joist and a solid concrete slab. Three full-scale simply supported beams with identical spans and cross-sections were prepared and tested. These specimens were designed as secondary beams of a typical composite flooring system based on Australian guidelines with the lowest permitted level of degree of shear connection of 0.5. They were cast simultaneously to enable comparisons with respect to pouring and loading conditions. One beam was cast un-propped and was kept unloaded for the whole duration of the long-term tests to measure shrinkage effects. The remaining two beams, cast under un-propped and propped conditions, respectively, were subjected to a sustained uniformly distributed load. Standard short-term and long-term tests were carried out to obtain the relevant material properties of both the steel and the concrete. Short-term and long-term push-out tests were carried out to obtain information on the response of the shear connectors. The experimental results were modelled by means of the finite element method. The time-dependent behaviour was depicted using a step-by-step procedure, while the steel joist and reinforcement were assumed to remain linear elastic. Two constitutive relationships were adopted for the shear connection, i.e., a linear-elastic one, and a new time-dependent one, to account for the long-term effects produced in the complex stress state of the concrete surrounding the shear connectors. The latter representation is intended to fall within the framework of simplified approaches suitable for design applications. Considerations of the accuracy of the numerical predictions are presented based on the two shear connection models. (C) 2010 Elsevier Ltd. All rights reserved.