Abstract
This paper presents a numerical investigation into the effect of construction methods to the structural behaviour of simply supported composite beams. Based on advanced three dimensional finite element models with material, geometrical and interfacial non-linearity, the structural behaviour of composite beams under experimental investigation has been obtained by the authors and other researchers with a high level of accuracy. By considering the construction method of composite beams in practice, it is important to realize that composite action is only required to be developed during the composite stage, and to resist primarily the imposed loads of the floors. Hence, the shear connectors are not required to resist the dead loads of the composite beams, and this represents typically a 30% reduction on the forces acting onto the shear connectors. Thus, the structural demand on the shear connectors in composite beams in practice is considerably smaller than that in those composite beams which are fully supported during concrete casting. Through a numerical parametric study, the deformation characteristics of a total of 12 simply supported composite beams are presented, and the differences in the structural behaviour of composite beams due to the construction methods are discussed. It is found that the required slippage of the shear connectors in most simply supported composite beams built on sites are 4 to 5 mm smaller than that in those composite beams built in laboratory; this corresponds to a 30 to 50 % reduction. The findings are very important in quantifying the practical ranges of ductile slippage of shear studs in composite beams, and also in explaining the discrepancies between the slippage requirements of shear studs obtained from previous numerical investigations and the ductile slippages of shear studs measured from standard push-out tests.