System Identification of Highway Bridges using Long-Term Vibration Monitoring Data

The need for maintenance and protection of critical highway infrastructure links has led in recent years to significant developments in the area of bridge structural health monitoring. Taking advantage of the instrumentation of three highway bridges located in Southern California, this dissertation focuses on the application of system identification (SI) techniques to a database consisting of more than 2,000 data sets collected over a period of nine years. The data sets consist of ambient and traffic induced vibration records, earthquake records, and controlled vehicles test data. During the controlled vehicle tests acceleration transducers were mounted on the vehicle chassis to study the potential for bridge identification using instrumented vehicles.;It is demonstrated long-term monitoring needs to be conducted in a permanent or periodic basis in combination with the application of SI techniques for the purpose of identifying changes in structural response characteristics and material properties due to aging factors and earthquake intensity. The ambient and traffic-induced vibration data were analyzed using the frequency domain decomposition technique for identification of modal parameters, i.e. natural frequencies and mode shapes. For a five-year period, a reduction of the first natural frequency is found to be approximately 2% for a 3-span straight bridge, 5% for a 3-span highly curved bridge and 3% for a four-span slightly curved bridge.;Using seismic acceleration records, the bridge modal parameters are identified using time domain SI techniques. It is found the identified first frequency of one of the bridges decreases up to 20% during an earthquake of moderate intensity (PGA > 0.37g) with epicenter close (< 15miles) to the bridge site. In general, larger earthquake intensities result in reduced identified frequencies and higher damping ratios of bridges. It is found soil softening at the abutments during earthquakes considerably contributes to the variation in frequencies due to changes in the condition supports and ultimately in the global stiffness of the structure. State-space models generated from the measurements successfully replicate the bridge response and show a fair enough prediction of the maximum structural response.;Also, this study develops a relatively simple approach for the identification of structural parameters and evaluation of the current state of highway bridges based on long-term vibration monitoring and Finite Element (FE) model updating. The physical parameters of the FE model, i.e. Young's modulus and boundary stiffness are identified. It is found the identified parameters slightly vary depending on the search domain applied during the updating process. Thus, engineering judgment is required in order to select an appropriate search domain. It is also found both identified stiffness and damping depend on the earthquake ground motion intensity. Finally, seismic vulnerability of the bridges is estimated using fragility curves. It is shown the seismic vulnerability increases with time.