Creep and Shrinkage Effects on Steel-Concrete Composite Beams

Predicting the long-term behavior of steel-concrete composite structures is a very complex systems problem, both because obtaining reliable information on material properties related to creep and shrinkage is not straightforward and because it is not easy to clearly determine the correlation between the effects of creep and shrinkage and the resultant structural response. Slip occurring at the interface between the steel and concrete may also make prediction more complicated. While the short-term deflection of composite beams may be easily predicted from fundamental theories of structural mechanics, calculating the long-term deflection is complicated by creep and shrinkage effects on the concrete deck varying over time. There are as yet no comprehensive ways for engineers to reliably deal with these issues, and the development of a set of justifiable numerical standards and equations for composite structures that goes beyond a simple commentary is well overdue. As the first step towards meeting this objective, this research is designed to identify a simple method for calculating the long-term deformations of steel-concrete composite members based on existing models to predict concrete creep and shrinkage and to estimate the time-varying deflection of steel-composite beams for design purposes. A brief reexamination of four existing models to predict creep and shrinkage was first conducted, after which an analytical approach using the age-adjusted effective modulus method (AEMM) was used to calculate the long-term deflection of a simply-supported steel-concrete composite beam. The ACI 209R-92 and CEB MC90-99 models, which adopt the concept of an ultimate coefficient, formed the basis of the models developed and examples of the application of the two models are included to provide a better understanding of the process involved. For the analytical approach using the AEMM, the entire process of calculating the long-term deflections with respect to both full and partial shear interactions is presented here, and the accuracy of the calculation validated by comparing the model predictions with experimental data. Lastly, the way the time-dependent deflection varies with various combinations of creep coefficient, shrinkage strain, the size of the beam, and the span length, was analyzed in a parametric study. The results indicate that the long-term deflection due to creep and shrinkage is generally 1.5 ~ 2.5 times its short-term deflection, and the effects of shrinkage may contribute much more to the time-dependent deformation than the effect of creep for cases where the sustained live load is quite small. In addition, the composite beam with a partial interaction exhibits a larger mid-span deflection for both the short- and long-term deflections than a beam with a full shear interaction. When it comes to the deflection limitations, it turned out that although the short-term deflections due to immediate design live load satisfy the deflection criteria well, its long-term deflections can exceed the deflection limitations.