Dynamic Interaction of Walking Humans with Pedestrian Structures in Vertical Direction Experimentally Based Probabilistic Modelling

There is a lack of credible and usable knowledge, specifically related to human-structure interaction in the vertical direction despite of its importance and potentially huge economic impact. The research presented in this thesis addresses this problem via a systematic combined experimental and analytical study of the effects of people on dynamic properties of vibrating structures they excite by walking. Series of extensive frequency response function based modal tests were performed on a full-scale test structure with more than one hundred test subjects walking in different loading scenarios. The experimental results were then used to identify the parameters of a single-degree-of-freedom (SDOF) mass-spring-damper (MSD) model of a walking human. Four different approaches, including agent-based modelling, were used to simulate measured scenarios of multi-pedestrian traffic. It was found that normal distributions with μ=2.864 Hz and σ= 0.191 Hz, and μ=0.295 and σ= 0.023 can describe the natural frequency and damping ratio of the SDOF MSD model of a walking human, respectively, when total mass of the human body is assumed as the mass of the SDOF system. A new vibration serviceability assessment method was proposed that takes into account not only the variability of the human body MSD parameters and the forcing function but also their interaction with the structure. Application of this novel method on two full-scale structures under walking traffic load verified its excellent performance yielding a maximum 10% error in estimating the level of structural response compared to 200-500% error margins when key design guidelines currently used around the world were employed. This method is versatile and, being easy to apply in practice, has the potential to replace the existing methods for simulating single and multi-pedestrian traffic on footbridges and floors.