Predictions of footbridge vibrations and influencing load model decisions

Prior to erecting a footbridge it may be useful to quantify the future and to predict vibration levels of the footbridge, since the vibration levels to come will determine the serviceability of the bridge throughout its service life. For design stage predictions of pedestrian-generated vibrations of a footbridge, decisions need to be made in terms of how to model the load. If the load is modelled as being stochastic it entails that a set of walking parameters are to be modelled as random variables for the predictions. Fundamentally, walking parameters are load amplification factors, step frequency, walking speed, pedestrian weight etc. and the paper adapts this line of thinking and presents results in terms of footbridge vibration levels computed under various calculation assumptions. Since the studies treat walking parameters as random variables, it is the stochastic nature of footbridge vibrations, which is in focus when comparing vibrations. The stochastic nature is brought about by Monte Carlo simulations, and a central aim of the studies of the paper is to examine how two different load models perform in terms of predicting selected stochastic features of footbridge vibrations when subjected to single-person traffic.