Comparative Evaluation of Current Pedestrian Traffic Models on Structures

Abstract

Recent vibration serviceability problems of many structures have drawn researchers’ attention to the walking-induced vibration modelling and assessment of floors and footbridges. Stochastic nature of human-induced vibration and its dependency on numerous conditions significantly reduce the accuracy of the current non-stochastic design procedures and consequently the level of confidence on them. This paper tends to study the performance of two of the most recent design guidelines, UK National Annex to Eurocode 1 (2008) [1] and French Sétra guideline (2006) [2], on real structures and to highlight their advantages and disadvantages. These methods are used to calculate dynamic reactions of a slender and lightly damped pre-stressed concrete slab strip and an as-built footbridge using various loading scenarios such as group loading, spatially unrestricted flow of pedestrians, and crowd loading. The calculated structural responses under first and second harmonics of walking load are further compared with their corresponding experimentally acquired reactions to examine their accuracy and efficiency. Various potential sources of discrepancy in the results of these procedures in comparison with the experimental data, such as missing human-structure interaction effects (e.g. damping), invalid ‘perfect periodicity’ assumption, limited applicability and unrealistic people correlation are then discussed. Finally the effects of using stationary pedestrians’ added damping on performance of considered design models are investigated and briefly discussed. The results of this analysis can provide a comprehensive insight into current design procedures’ accuracy and shall be taken as a good starting point for future researches in this area.