Civil Engineering Departmental Seminar - CONSIDERATIONS ON BRIDGE MONITORING AND DAMAGE DETECTION

CONSIDERATIONS ON BRIDGE MONITORING AND DAMAGE DETECTION

Some thoughts about the importance of bridge monitoring (BM) and damage detection in existing bridges taking into consideration the actual state of transportation infrastructure will be discussed.

The main objectives and need for bridge monitoring are introduced first, following with a description of the data collection techniques used in BM, either in the stage of the creation of the virtual asset or during the operational phase. From them, strain and temperature monitoring by Distributed Fiber Optic Sensors (DFOS) is selected for an in-depth description, main advantages, deployment possibilities in concrete bridges as surface-bonded or embedded in the concrete matrix variants, and field of application. Several practical applications in real-world structures and bridges are also described, showing the excellent performance of this technique for an accurate BM. The focus is also in the image-based data collection, in particular on the Digital Image Correlation (DIC) technique and remote sensing by satelite-based InSAR (Interferometric Synthetic Aperture Radar). The importance of Data Fusion (DT) of data streams coming from different sensing techniques for a better decision-making based on the results of the monitoring is also emphasized, by presenting a successful application of combined DFOS+DIC+AE (Acoustic Emission) to characterize complex resisting mechanisms in structural concrete. Finally, damage detection in bridges under the challenging effects of operational (traffic) and environmental (temperature) variability is considered and an un-supervised methodology for damage detection and location by using only monitored acceleration data is derived. The cases with ambient vibration and forced vibration by traffic and with and without environmental changes are differentiated and appropriate damage features proposed. The application to real bridges is presented, showing the correct performance in damage detection in real-world scenarios as presented in the 3 case studies of bridges where acceleration records in un-damaged and damaged conditions are available. The damaged state is only used to check the correct performance of the un-supervised damage detection methodology.