Modelling and Design of Curved Composite I- and Box Girder Bridges

Abstract

Design and modelling of curved composite bridges is a complex art, with few ex isting guidelines in European standards. By introducing a curvature to the bridge,behaviours such as torsion will highly influence its performance and the design de mands will therefore include the torsional stiffness. Naturally, a box girder section would be an appropriate design choice, however, a twin I-girder section could be economically motivated if equipped with a lower plan bracing. This thesis aims to investigate the design and modelling of curved composite bridges. This is done by firstly, discussing appropriate numerical modelling strategies, as there is a need for simplified beam models during the conceptual design phase. Here, the concluding result shows that a spatial beam model is a more suitable beam model, both for the I- and the box girder bridge. Based on these findings, the most advantageous crosssection design at varying radii is investigated by presenting the optimised volumeof the I- or box girder bridge. In summary, this part of the study indicates that an I-girder cross-section is the most material-efficient down to a 300 meters radius of curvature. For bridges more curved than that, the box girder bridge results in theleast amount of steel. The thesis concludes by presenting a design procedure for curved composite bridges that could be utilised in practice within this field.