Modal identification and health monitoring of bridges using seismic acceleration records

This dissertation is built on two complementary tracks. In the first, use of earthquake records for obtaining modal parameters of instrumented bridge systems is addressed. In the second, subsequent use of modal parameters to understand the dynamic behavior of several bridge systems in more detail is evaluated.; Regarding the first track of this dissertation, the problem of System Identification (SI) from earthquake records is addressed with various methods and different configurations of real or envisioned accelerometer arrays. This effort is aiming at setting the stage for future SI applications using large networks of sensors. In this context, strong motion data from several bridges are analyzed first using different SI methods. The objective is to assess the state-of-the-art of identification methods applied to bridges. Subsequently, to assess the effect of increasing the number of sensors in acceleration recording arrays, a simulation study based on finite element modeling of a large bridge system is undertaken to generate "virtual" seismic acceleration data. Acceleration records from this simulation for various configurations of sensors are used to understand the effects of increasing number of sensors on SI results.; Within the context of the second track of this dissertation, the acceptability of all obtained SI results, somehow always presumed in the past literature, is challenged. This is conducted by acquiring the confidence intervals of the identified modal parameters for the purpose of selecting the reliable ones. A framework of sensitivity analysis and Monte Carlo simulations is developed for this purpose to formulate a robust computational procedure for obtaining these confidence intervals. A challenge to the accuracy of the identified modal parameters involves consequences regarding the damage detection and health monitoring, as some of the identified modal information is essentially not useable for acquiring a reliable damage diagnosis of bridge systems.; In summary, earthquake shaking is a convenient way of obtaining input-output data from bridge structures. However, use of this data has limitations that should not be ignored. The results and conclusions of this dissertation underline these limitations while presenting the opportunities offered by SI for better understanding of bridge systems.