FLUTTER RISK ASSESSMENT OF LARGE WIND TURBINE BLADES CONSIDERING STIFFNESS UNCERTAINTY

To quantify the impact of blade stiffness uncertainty on the aeroelastic stability assessment of large-scale wind turbines, taking the IEA 15MW floating offshore wind turbine as the research object, this paper proposes an uncertainty quantification and risk assessment method based on multidisciplinary modeling and evidence theory. An aero- hydro-structural fully coupled dynamic model is established, and an efficient prediction model for the flutter limit boundary is constructed by combining the Kriging surrogate model with the SSA optimization algorithm. Based on evidence theory and Mahalanobis distance, quantitative evaluation of the belief and plausibility of the flutter limit under coupled flap-wise, edgewise, and torsional stiffness uncertainties is achieved, the evolution law of how stiffness uncertainty influences flutter limit predictions is revealed. This research provides an important theoretical basis and methodological reference for research on the reliability and robustness of large-scale wind turbine blades. The results indicate that flutter risk assessment can be divided into three characteristic intervals: an unconditional safe operation zone, a dynamic development zone, and a high-risk zone. The critical flutter speed is most sensitive to torsional stiffness uncertainty, with a maximum relative difference of 43.4% between the upper and lower limit under uncertainty coupling conditions for the cases. An increase in the stiffness uncertainty interval amplitude significantly reduces the lower limit of system stability, while the upper limit changes minimally due to stiffness saturation effects. When the discrete width of the torsional stiffness interval is refined from 0.05 to 0.025, the epistemic uncertainty is reduced by 21.46%, indicating that increasing the discretization level of the interval helps improve prediction accuracy, when the stiffness is uncertain by ± 10%, the upper and lower limits of the flutter limit speed deviate from the initial value are ± 6.6%.