Aeroelastic Flutter Prediction Using Multifidelity Modeling of the Generalized Aerodynamic Influence Coefficients

This work proposes a multifidelity modeling approach for predicting aeroelastic flutter of airfoils and wings. Using aerodynamic models based on the doublet-lattice method and time-accurate Euler equations, cokriging-based surrogates of the generalized aerodynamic influence coefficients are generated as functions of Mach number and reduced frequency. The surrogate-based matrix terms are then used in the p-k method to determine flow conditions at flutter onset. To demonstrate the multifidelity process, a widely used pitching and plunging airfoil case is considered. Verification of the approach is done by comparing with results from a mode-based time-domain aeroelastic solver, as well as data from the literature. The approach draws inspiration from Timme et al., but focuses more on widely used industry tools (namely, the p-k method, panel-based aerodynamics, and time-domain computational fluid dynamics). The benefit of using multiple aerodynamic fidelities, rather than high-fidelity kriging models, is also quantified by examining a flutter speed error metric as the number of high-fidelity samples is varied.