A MSD model for coupled analysis of pedestrian-footbridge dynamic interaction

Recent footbridges are characterized by long span, light materials and increasing slenderness which make them more sensitive to dynamic forces induced by pedestrians. While walking, the pedestrian adapts his gait to the bridge motion and interacts with the structure. At contact points, the pedestrian transmits contact forces to the bridge that, in turn, imposes a set of displacements and velocities to the pedestrian's feet. The bridge response accounting for the human-structure interaction depends both on reliable estimates of expected loading scenarios and on the accuracy of the models representing pedestrians.This work presents a complete framework for the analysis of the human-structure dynamic interaction. To properly describe this phenomenon, the coupled equations of motion for the two mechanical systems are derived. To this aim the standard FE modeling of the structure is retained, while a new bipedal mass-spring-damper model is adopted for the pedestrian. The model is able to reproduce the sequence of single and double support phases typical of the human gait and is excited by both an equivalent bio-mechanical force and the motion at contact points with the bridge. The solution of the coupled formulation is based on a forced uncoupling of the equations, possibly associated to an iterative integration procedure. At each time instant the two systems are analyzed separately: the bridge subjected to vertical contact forces, the pedestrian to an imposed motion at his feet.The uncoupled strategy of solution is implemented into a research code. In the case study groups of pedestrians cross a lively footbridge, whose modal properties were experimentally identified. The numerical model was developed with ANSYS. To investigate the potential of the proposed approach, numerical analyses have addressed the effect on the bridge response of both the degree of synchronization and the spatial distribution of groups of nine pedestrians. The bridge vertical accelerations reach values out of the range of comfort and show an high dependency on these parameters. The effectiveness of the proposed approach of modeling and analysis is highlighted. (C) 2017 The Authors. Published by Elsevier Ltd.