Tools for Analyzing Flutter in the Presence of Aeroelastic Uncertainty

The potentially catastrophic results of entering the flutter regime mean that flutter flight testing of aircraft carries significant risk, and accurate predictions of flutter behavior are highly valuable. Recent advances have yielded aeroelastic models that are quite accurate, but these models are highly complex and very computationally intensive. Using these models to analyze uncertainty has thus been difficult, and the impact of uncertainty on quantities of interest, such as the probability of flutter, is not well understood. New tools are needed to provide the desired understanding of the impact of uncertainty on flutter behavior. The proposed Phase II research will develop the needed analysis tools based on a generalized polynomial chaos (gPC) representation of uncertainty. The team will leverage recent advances in reduced-order modeling approaches, which provide improved accuracy, and will develop an efficient experimental design methodology for characterizing how uncertainty propagates from underlying sources in the high-fidelity simulation to the reduced-order models. After reduced-order models that characterize uncertainty have been developed, outputs of interest such as probabilistic flutter bounds can be readily characterized, and sensitivity to specific uncertainties can be quantified. The ultimate result of the research will be an efficient testing methodology that minimizes risk.