Comparative study of evaluation methods for existing portal frame bridges: Evaluating the difference in using a 3D FEM-design approach and a one-way slab analysis when assessing existing bridge structures

Abstract

The bridge population in Sweden is ageing and the demands on our infrastructure are increasing. Therefore, there exists a demand for assessing the existing bridge population. The current practice for capacity assessments of old bridges in Sweden is based on two different modelling choices. Either the bridge is analysed in a two-dimensional (2D) frame analysis, as it was usually designed, or the bridge is modelled in a more modern 3D analysis. Both modelling choices have their advantages and disadvantages, and depending on what bridge is to be assessed the choice of modelling could influence the results of the capacity assessment as well as the overall workload associated with capacity assessments. A study of the difference between the two modelling choices was performed on a portal frame bridge, to compare the results of different bridge geometries when using the 2D frame method and the three-dimensional finite element method. The study examined what parameters of the existing bridge design would have an influence on the difference in results. The study also varied different parameters on modelled bridges to examine the difference in results based on the two different analysis methods. Also included in the thesis was a study of existing bridges in Sweden that would relate the results of the study to common bridge geometries, to establish a connection between the study’s findings and common assessment scenarios. The study found that the width of the bridge becomes decisive for which analysis methods are best suited. In the case of wider bridges, the difference is not so great, and to some extent, the 3D models give a higher capacity classification than 2D models. For narrower bridges, the difference in results between 2D and 3D is much greater, where 2D in this case offers a higher capacity rating. However, an error regarding distribution widths was discovered, making the results of the study somewhat unreliable. General guidelines are presented to simplify the choice of analysis method when a capacity assessment of an existing bridge is performed, in the capacity that is possible regarding the found error.