Structural integrity assessment of railway bridges during the passage of trains

Traditional algorithms for the identification of dynamic parameters based on vibrational structural response generally involve strict assumptions about stationarity and may not be suitable for time-varying systems. Therefore, there is a tendency to discard the measurements of the structural response collected during short-time events, such as train passages on bridges. The response to a strong event may however include valuable information about structural features that cannot be observed in low-excitation scenarios. Moreover, a strong excitation typically generates a higher-amplitude structural response, which is particularly convenient when using low-cost instrumentation for data collection that is usually characterized by low sensitivity. In this paper, dynamic parameters of bridges during the passage of trains are identified by fitting the structural response to a simplified analytical model specialized for railway applications. The efficacy of a damage index based on the fitting residuals is investigated using a numerical study, simulating the use of an event-triggered smart sensing system. Applying the proposed method, damage detection can be achieved using a single low-cost sensing node.