Probabilistic versus Deterministic Assessment of the Minimum Structural Damping Required to Prevent Galloping of Dry Bridge Hangers

Abstract

The risk of large-amplitude vibrations of bridge hangers due to galloping instabilities has posed a challenge to the engineering and research communities. Galloping vibrations can lead to serviceability problems and reduce fatigue life. A number of aeroelastic models have been developed to predict the unstable behavior and to design counteracting measures, i.e., shape modifications and structural damping addition. All the proposed procedures generally consider that the parameters are assigned according to deterministic values. A framework is proposed for the deterministic and probabilistic assessment of the minimum structural damping required to prevent galloping of bridge hangers based on the output of a two-degree-of-freedom sectional quasi-steady aeroelastic model. All the variables required to define the hanger dynamics, the sheath aerodynamics, and the local wind climate are considered. Because of the large uncertainties and of the nonlinear nature of the problem, the distribution of the minimum structural damping required to prevent galloping is obtained by Monte Carlo simulations. An application of the method to the proposed Messina Straits crossing bridge is presented. Starting from wind tunnel measurements of the aerodynamic coefficients of a real plain high-density polyethylene (HDPE) cable sheath, the random nature of the aerodynamics is shown, which is ascribed to the cable irregularities (surface roughness, section distortion, and axis curvature). Then the statistical variation of the parameters of the model is considered, based on real data derived from wind tunnel tests and from the wind climate measured at the site. The distributions of the parameters defining the hanger dynamics are assigned according to typical values. Using a normal and a log-normal model of capacity, the probability of failure is calculated. The results are discussed and compared with conventional approaches based on the definition of the deterministic variables. Finally, using a deterministic model of capacity, a simplified closed-form equation for the evaluation of the structural damping needed to prevent galloping in a probabilistic-based performance approach is proposed.