Nonlinear Earthquake Analysis of Bridges in Time Domain
Master thesis
Permanent lenke
http://hdl.handle.net/11250/236876Utgivelsesdato
2011Metadata
Vis full innførselSamlinger
Sammendrag
Asynchronous seismic excitation has in previous studies been shown to significantly alter the response of long-span bridges. In order to further investigate this phenomenon, this master`s thesis studies the seismic behavior of a cable-stayed reinforced concrete bridge. A 3-D finite element model of the cable-stayed bridge is employed to conduct the seismic analyses. The model includes both geometric nonlinearities, from large displacement theory, and material nonlinearities, which are present in the soil springs connected to the bridge foundations, in the stay cables and in the simplified plastic hinges in the bridge tower legs. The case study bridge, which has a total deck length of almost 300 m, is subjected to two different earthquake accelerograms in the transverse direction. The accelerograms, which are based on the N-S component of the 1985 Nahanni and the 1976 Friuli earthquakes, are baseline corrected and matched against the elastic design spectrum of EC8-1. The analyses are carried out in time domain, in order to include the nonlinear effects, while frequency domain calculations are used for verifications.
The results support previous studies regarding the effects of asynchronous earthquake motion. Significant changes in response variables are observed when including spatial variability of ground motions. The three main factors causing asynchronous motion are each investigated. The results suggest that the effect of time delay of seismic waves is less decisive than the effects of local soil conditions and loss of coherency. It is further concluded that the simplified method of EC8-2 for asynchronous ground motion underestimates the internal forces of the bridge. Hence, based on the relatively few analyses conducted, a review of the simplified method is suggested. The effect of soil-structure interaction is evaluated in a simplified manner by connecting soil springs to the bridge supports. The results are, except for at the viaduct column studied, elongation of natural periods, larger displacements and lower internal forces.
An inquiry into the effects of nonlinear material models for the soil springs and the reinforced concrete is also made. Due to the modest design spectrum for the bridge, the nonlinearities prove to have no effect on the response. Scaled analyses show that the nonlinear concrete model yields larger displacements and lower internal forces, as expected, except for at the top of the bridge tower. A sensitivity analysis regarding the soil spring stiffness more surprisingly shows that lowering the soil stiffness not necessarily yields lower bending moments in the bridge deck, as one would normally expect.