Statistical characterisation of parameters defining human walking as observed on an indoor passerelle

Increased slenderness of footbridge structures in the last few decades has led to an increasing number of vibration serviceability problems under human induced dynamic load, such as walking. These problems are typically related to increased discomfort of footbridge users due to perceptible vibrations during footbridge crossing. The current codes of practice dealing with the vibration serviceability of footbridges often fail to assess reliably the vibration level induced by humans. This is because they model the load induced by pedestrians by considering only an ‘average’ walker. However, because of natural diversity in human population, so called inter-subject variability, it is necessary to model the walking force in a probabilistic way taking into account this type of variability. To do this, a large database of parameters (such as walking frequency, step length and weight) describing walking force induced by different individuals is required. Currently, only limited amount of data to populate this database is available. To generate more data of this kind, pedestrian traffic on an indoor passerelle inside a University campus in Sheffield was monitored using video cameras for 6.5 hours. The data that characterise human walking (such as walking frequency, step length and arrival time) were collected and analysed statistically. Based on this, the probability density functions for the parameters analysed were constructed. The results of this study could be used when defining a badly needed probabilistic force models for vibration serviceability assessment of indoor passerelles of this kind.