An investigation on structural identification (ST-ID) of a long-span bridge for performance prediction

Of all the civil infrastructures, long-span cable-supported bridges are the lifeline structures for the transportation network in a country/state. These bridges are not only a substantial component in transportation system, but also offer a respectable contribution to economic and cultural growth of their regions. Compared to other infrastructures, long-span bridges are not budget-friendly transportation items. Along with the construction cost, the cost of maintenance, management, and safety of longspan bridges is also relatively high. Many long-span bridges in service may also be vulnerable to various unpredicted load events and environmental conditions. Comprehensive investigations on long-span bridges therefore needs to be conducted before and after construction. Structural Identification (St-Id) presents an effective solution to better structural assessment of long-span bridges utilizing Structural Health Monitoring (SHM) data. This approach provides the bridge authorities/owners with an integrated concept consisting of six-step framework: (i) visualization and conceptualization (ii) priori modeling (iii) experimentation (iv) data analysis (v) calibration/verification analysis (vi) utilization. Even though containing limitations, uncertainties and challenges in each step, St-Id process yields to significant decision making-based outcomes for long-span bridges. The main aim of the dissertation is to make a comprehensive investigation on structural response of the long-span bridges in Turkey adopting the integrated approach of St-Id. For this objective, the Boshporus Suspension Bridge is considered as a real-life civil infrastructure in this dissertation. Based on the conclusions of the investigation on the Bosphorus Bridge, it is also aimed to present a framework for integrated solution to safety and proper structural condition assessment of other existing or new long-span bridges in Turkey. In the first part of the dissertation, the first two steps of St-Id, which are visualization / conceptualization and priori modeling, are presented. For this aim, original project drawings, retrofit projects and experiences from the bridge's official authority KGM (General Directorate of Turkish State Highways) are identified and compiled properly. Taking these documents and experiences into consideration, an elaborate study is conducted to establish priori finite element model (FEM) of the bridge. Based on the conclusions obtained from the analyses results of the priori model, more advanced FE models of improved spine-beam and 3-D full-scale shell are developed for further investigations on the bridge. In this section of the dissertation, similar efforts are also made for FE modeling of the bridge with vertical hangers. For this purpose, the hanger replacement project considerations that also include certain retrofitting works are precisely specified. New locations of vertical hangers at the deck and main cables are first determined. Besides, physical and technical properties of new rocker bearings, wind tongues, viscous dampers and tower base-section retrofitting are presented in detail. Considering these specifications, improved spine-beam and 3-D full-scale shell FE model of the Bosphorus Bridge with inclined and vertical hangers are prepared to perform the modal analysis. Making a comparison between the results of the FE models, updates for the models and vertical hanger effects are determined in terms of natural vibration frequency. In the second part of this dissertation, the SHM system of the Bosphorus Bridge is presented in detail in accordance with the requirements of St-Id approach in order to conduct the second steps of St-Id: experimentation and data analysis. General design considerations, monitoring objectives and parameters of the SHM system are clarified depending on the bridge's own characteristics. The monitoring specifications of the bridge are also determined to be in close agreement with those of the well-known long-span bridges in Hong Kong and Japan. Considering these specifications and recommendations of the KGM, critical points and components to be monitored are specified and preliminary performance criteria is proposed for the Bosphorus Bridge. For data analysis, the extreme wind event is considered. Before the data analysis, this event is accurately identified and presented. Depending on the preliminary results, the need to develop a new Matlab code for averaging extreme wind data is stated. The frequency-domain results are then obtained to be clear and the first nine dominant (deck and tower modes) modes of the bridge are identified. Moreover, influence of the extreme wind event on the bridge is determined on the basis of its dynamic properties through SHM data. The last part of the dissertation includes updating of the developed FE models with the experimental frequencies obtained from the extreme wind event through the SHM system and the utilization for seismic performance prediction of the bridge under the multi-point scenario earthquake motion. The conclusions obtained from St-Id process of the Bosphorus Bridge indicate that the developed FE models can be utilized for advanced analysis and investigations of the bridge. Particularly, the updated 3-D full-scale shell model of the bridge can be considered for various further objectives. The Bosphorus Bridge is determined to show elastic performance under the extreme wind load based on the comparative experimental results obtained from "During" and "After" data analysis. Visual inspection of the Power Spectral Densities (PSDs) of the dynamic response also demonstrates an increase in the damping capacity of the lateral mode under the extreme wind load. This increase can be attributed to activated friction at the expansion joints and the change in the boundary conditions of the rocker bearings that restrained lateral movement. From the multi-point and uniform-point earthquake analyses under the scenario earthquakes for Istanbul, the bridge with both inclined and vertical hangers is not likely to be damaged and no repair and strengthening intervention can be required for the bridge. Under the DD-1 spectrum compatible earthquakes, damage is estimated for the tower base-section and approach viaduct box-section of the bridge with both inclined and vertical hangers. For retrofit strategy, sectional capacity is recommended to be increased with additional longitudinal stiffener of doubly symmetric W100x499 I-beam. The comparative results also yield to necessity of the consideration of the multi-point earthquake analysis rather than the uniform-point earthquake analysis. Using simplified approach, structural response of the bridge to the multi-point earthquakes is identified with the deck and tower vibration modes. Some recommendations are also made depending on the outcomes of the detailed analyses. With the increase in the number of long-span bridges in Turkey, especially in Istanbul, there is need for a command center where the all SHM systems are integrated for rapid decision-making. Due to high seismicity of Istanbul, the connection of the earthquake warning system with the SHM systems of the bridges and other important infrastructures should also be integrated. Along with the preliminary attempt to determine the operational performance criteria of the Bosphorus Bridge in this dissertation, developing extended SHM-based operational performance criteria is also necessary to take preventive measures for operational and extreme load events.