Seismic response of single tower cable-stayed bridges

Little is known about the structural behavior of cable-stayed bridges subjected to large magnitude, near fault earthquake ground motions. The Ji Lu bridge is located near the epicenter of the Chi-Chi Earthquake that shook northwestern Taiwan on 21 September, 1999. This pretressed concrete, cable-stayed bridge was constructed using recent design and analysis methods. The Ji Lu bridge thus presents a unique opportunity for improving current knowledge about seismic behavior of cable-stayed bridges.; The Ji Lu bridge has two symmetric 120-meter spans supported by a single pylon with harped cables extending from the pylon to the center of the box girder. Bents at the ends of the spans provide restraint vertically for gravity load and transversely for wind and seismic loads.; To calculate the forces on these critical end span connections for design, sophisticated nonlinear modeling of the cables is essential. Detailed simulations show that it is necessary to incorporate this nonlinearity into the analytical model to adequately estimate bridge damage and overall behavior. Results of a detailed nonlinear analysis show that during strong ground shaking from the 1999 Chi-Chi Earthquake, the ends of the bridge rose vertically out of the substructure supports resulting in a release of the transverse restraint. With the end spans free, the structure became particularly vulnerable to damage from the near-fault earthquake excitation. Simulation with the Ji Lu cable-stayed bridge model shows that had the end spans remained restrained during the seismic event, only minor damage would have occurred.; This report shows that while transverse restraint of the superstructure end spans is critical, vertical restraint is also important to reduce overall system demands. However, with the end spans displacement restrained in the vertical and transverse directions, seismic demands on these connections may become very large. In an effort to mitigate seismic demands, simulation results show that increasing the end span connection flexibility alone is not effective. Rather, the increased flexibility causes a systematic increase in super-structure demands. Simulation results suggest that isolation of the entire system, however, could be an effective seismic mitigation strategy for the reduction in overall system demands, particularly for the end span connections.