Gust buffeting and turbulence uncertainties

Literature on turbulence modeling is rich in empirical, semi-empirical and theoretical spectral equations whose parameters assume deterministic values. Starting from a critical review of the state of the art, a unitary model of the 3-D atmospheric turbulence was proposed where all parameters are defined by first- and second-order statistical moments derived from a wide selection of experimental measurements. Analyses based on Monte Carlo simulation of the model parameters show that the propagation of turbulence uncertainties over the wind-excited response of structures gives rise to a relevant scatter. The passage from Monte Carlo simulations to Taylor series expansions addressed to apply the first-order second-moment technique provides systematic rules to deal with the structural response in a full probabilistic environment. It also allows appreciation of the role of each uncertain parameter. The joint application of this procedure and a closed form solution of the response drastically reduces the computational burden and makes these analyses operatively possible. The proposed approach is a step forward towards the formulation of a unitary probabilistic model of the wind-excited response of structures, comprehensive of model errors, parameter uncertainties and parameter randomness.