Aerodynamic uncertainty propagation in bridge flutter analysis

This paper presents a method to approach flutter instability in a probabilistic way and to calculate the critical wind speed, starting from the probability distribution of the flutter derivatives. Uncertainty propagation is studied and the results can be used for risk-assessment purposes. The statistical properties of experimental flutter derivatives were investigated with ad-hoc wind tunnel tests performed on a bridge deck model of common geometry. The probability distribution of the flutter critical wind speed can be analytically calculated if a simplified approach to flutter is followed, while Monte Carlo methods have to be utilized in the general case. Several application examples are presented and both well-behaving and particularly critical cases of uncertainty propagation are discussed. Finally, the effect of partial correlation between flutter derivatives is studied and its non-negligible role in the definition of the probability distribution of the flutter wind speed is underscored.