Bridge bearing fixity evaluation using ANN and low-cost wireless sensing systems

Le résumé fourni par la source

Many bridges, particularly in developed nations, are reaching or surpassing their intended lifespan, creating significant challenges in ensuring public safety and serviceability while accommodating increasing traffic demands within budgetary constraints. To overcome such challenges, structural health monitoring systems and methods have drawn significant attention in recent decades to facilitate the timely identification of structural deficiencies and implementation of proper maintenance mainly focused on superstructures. However, the bridge substructure, including supports and joints, could be more susceptible to damage due to increased environmental loads, use of harsh chemicals, and unexpected manmade impacts. Once damaged, substructure damage can involve more serious impacts on the bridge structure, including total collapse. This paper reports on the development of a novel low-cost wireless joint monitoring system (wJMS) to monitor the behavior of bridge bearing of in-service bridges and the formulation of an Artificial Neural Network (ANN)-based bearing health assessment algorithm. The wireless sensing system was optimized for cost, power, and Ingress Protection (IP) rating and assembled using commercially available sensors with a specially designed casing system and configuration for real-time remote monitoring. The system was deployed in two in-service bridges in the state of Connecticut to monitor the longitudinal displacement and girder temperature of steel girder composite bridges. Finally, an ANN model was developed using finite element simulated data to estimate the bearing stiffness of the bridge based on measured longitudinal displacements and girder temperatures. The predicted stiffness values were found to be in agreement with the visual condition assessment results of both bridges.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.