Traffic Safety Evaluation for Railway Bridges Using Expanded Multisensor Data Fusion

Abstract

The traffic safety of a railway bridge is generally evaluated by levels of structural responses such as acceleration, vertical displacement, and deck twist. Whereas acceleration can be readily measured in general, acquiring displacement and twist responses in field testing is often a challenging task due to lack of appropriate sensors. As most existing displacement transducers are designed to measure at a single location, the deck twist which is calculated from four displacements requires costly and labor-intensive sensor instrumentation. To effectively address the issue, this study proposes an integral strategy for the traffic safety evaluation of railway bridges using multisensor data. The proposed approach provides a formulation to estimate the dense displacement necessary for obtaining twist responses using acceleration and strain measurements. Wireless sensors are adopted because of their intrinsic advantages in multimetric sensing of heterogeneous data, convenient sensor instrumentation, and high-fidelity time-synchronized data acquisition. The proposed approach for dense displacement estimation is numerically and experimentally validated using beam models. Subsequently, a full-scale experiment on a railway bridge is conducted to evaluate the traffic safety for high-speed trains at three different speeds of 280 km/h, 300 km/h, and 400 km/h. The acceleration, vertical displacement, and twist are obtained and compared with design limits to determine the traffic safety of the railway bridge.