Prediction of long-span bridge response to turbulent wind

A comprehensive numerical solution to investigate flutter instability and buffeting response of long-span bridges under turbulent wind excitation is developed. Using flutter derivatives in a self-excited force model, finite element formulation makes possible establishing a general aerodynamic motion equation of the bridge system. Based on this formulation, the self-excited force can be conveniently treated as deterministic or random.; In the case of deterministic self-excited force, spectral analysis is used for the buffeting problem and eigenvalue analysis for flutter problem. In the case of random self-excited force, random parametric excitation method is used to conduct both buffeting and flutter analyses. In the random parametric excitation analysis, a new approach is proposed to identify the flutter velocity for the second moment instability. Though the random parametric excitation analysis is still complicated in this study, its computational effort is significantly reduced, compared with those of previous studies.; Deer Isle bridge, because of its very poor aerodynamic behavior, is a subject of many previous experimental and theoretical analyses. In this study, a comprehensive numerical investigation of flutter instability and buffeting response is conducted. The effectivenesses of the stiffening system installed in the 1940's and the planned fairing system are evaluated. The findings are as follows: (1) Critical flutter occurs in the first vertical mode instead of in the first torsional mode as assumed by previous researchers. (2) The critical flutter velocity of the existing bridge is 82 km/hr at 2% damping ratio, a value much lower than the design wind velocity of 175 km/hr. This explains the observed poor aerodynamic behavior of the bridge. (3) The fairings to be installed on all 72 panels raises the critical flutter velocity to more than 400 km/hr. The fairings are much more efficient than the stiffening system installed in the 1940's. (4) Placing fairings on only 32 panels gives a critical flutter velocity of 218 km/hr, a value still higher than the design wind velocity of 175 km/hr. Partial fairings are sufficient from a purely structural viewpoint. Full-length fairings may be desirable for aesthetic reasons.