Analysis of Footbridge Comfort Vibrations

Abstract: The aim of this master's thesis was to investigate and increase the knowledge of how different parameters affect the dynamic properties of steel beam bridges used for pedestrian traffic. Steel footbridges are commonly used due to their high strength and low weight. Due to the lower static loads caused by pedestrians, the dynamic loads are usually decisive. For lightweight bridges with a low ratio of stiffness to mass, the natural frequencies can coincide with the walking frequency of pedestrians. The dynamic properties of a bridge are usually more difficult to determine at an early design phase than for the static case. To be on the safe side, this generally results in a bridge with a low utilization ratio, and oversized dimensions. This comes at a great cost, both environmentally and financially. There are different guides and methods on determining the dynamic properties of a bridge. The one used in this thesis, the Sétra technical guide on footbridges, is commonly used in the industry. It is based on different bridge classes depending on crowd density and use. Depending on the natural frequency of the bridge, different dynamic load cases are achieved. This guide was used to evaluate a large number of footbridges for their dynamic properties. The different bridges varied both in geometry, stiffness, and mass. This was to see if there were any thresholds for where the dynamic criteria could be met reliably, and to compare the dynamic and static criteria. The analysis was achieved by writing and using a Python script to create and analyze each numerical model in the finite element analysis tool Abaqus. This allowed for a study of the different parameters. The analysis concluded that the choice of bridge class had the individually largest impact on the results according to the Sétra guide. It also concluded that for bridges with a relatively small bearing area of less than 140 square meters, it is not possible to achieve a high utilization ratio. The natural frequencies had to be close to threshold values for the acceleration criteria to be met.