ABSTRACT
Prior finite element models of composite beams have shown good agreement with experiments when validating the evolution of deflections in time. However, a close look at the techniques used by various research groups show a wide range of inconsistencies, particularly when the system is modeled using a combination of beam and shell elements. In this study, new light is shed on the modeling of composite beams by conducting an analysis of axial forces in a restrained composite beam under fire exposure. It is shown that the relative position of the reference plane within the member cross-section drastically affects the computation of axial forces in a restrained composite beam. As the experiments that are commonly used for model validation rarely report axial force data, this inaccuracy has been largely unreported in the literature. The aim of this paper is to demonstrate that structural elements are often used unsatisfactorily when modeling restrained composite beams and recommendations are given for improving axial force predictions.
