Tipping point analysis of the NPL footbridge

Abstract

The NPL footbridge was used for a pedestrian crossing in the past but now serves as an experimental test ground for assessing structural monitoring sensors, in a regular regime and under various structure-stressing experiments. Algorithms from the tipping point toolbox are applied to study sensor data recorded at the footbridge. Tipping point analysis is an established methodology in environmental sciences, and due to its general approach in studying dynamical systems of arbitrary origin, it can be applied to sensor data as well, which is demonstrated in the present study. Two time series techniques are applied, degenerate fingerprinting and potential analysis, to assess the proximity of critical behaviour in material temperature and tilt records obtained by two sets of sensors in time period 2009–2012. Various detected transitions and bifurcations are reported, and their implications for the structural evolution of the bridge are discussed. The proposed methodology, being computationally fast and easy to implement on a single computer desktop, might be used for real-time asset health monitoring for early warning signals of critical behaviour and for detection of dangerous structural changes. The results of the applied time series techniques characterising the behaviour of the structure are very encouraging. Further collaboration with structural engineers to assess the full-scale dynamics of other bridges and structures is expected.