Abstract
The objective of the work reported in this thesis is to study the behaviour of stiffened compression flanges of steel box girder bridges. Various collapse modes and many different parameters that effect the behaviour of the stiffened panels which make up the flanges have been considered. In order to achieve this, a series of three parametric studies have been carried out using a finite element package (LUSAS), which allows for both material and geometric non-linearities. The aim of the work was, in the main, to provide a datum for assessing design recommendations being prepared for the ECCS (European Convention for Constructional Steelwork), and to help in their development. The recommendations were developed as part of this research programme. The first parametric study looked at a range of isolated stiffened panels subjected to uniform uniaxial compression, representing the compression flange of a box girder subjected to pure moment. This validated the basic requirements within the recommendations for the strength of longitudinally stiffened flange panels. The second study was aimed at examining the effect of the shear flexibility of the flange of a girder loaded by vertical shear. In order to study the influence of this flexibility which results in a lag in stresses across the box width relative to the edges (shear lag), finite element results were generated for panels loaded along their longitudinal edges by in-plane uniform shear stress. A new model has been established from these results to predict the effect of shear lag. The model consists of two equations. The first is concerned with the prediction of the elastic shear lag effective breadth ratio while the second is concerned with the effect of shear lag at collapse. The latter is substantially different because of the ability for plastic redistribution of direct stress to occur across the flange prior to collapse. Finally, a third study looked at the forces on, and the behaviour of the cross-frames in orthogonally stiffened flanges. The results of the study showed that the flange applied a more complex lateral load distribution to these stiffeners than currently considered in design practice with, for some parameters, substantially higher load levels than required in the strength requirement of the British code for the design of bridges. A modification is proposed to the rules to provide transverse stiffeners which satisfy both stiffness and strength considerations.