Adaptive identification method for aerodynamics applicable to aeroelasticity

The reduced-order models for unsteady aerodynamics can substitute for computational fluid dynamics solver in the analysis of aeroelasticity. Usually most reduced-order models lack robustness against parameter changes and therefore need to be rebuilt frequently for new parameters. Furthermore such construction can be computationally expensive. This paper investigates the adaptive identification method for unsteady aerodynamics applicable to aeroelasticity based on Volterra theory. By using Haar scaled function, Volterra kernel is extracted from system input/output data, which is expanded in terms of series. Then the coefficients in kernel are expressed as a function of single varying parameter. Finally, a reduced-order model with regard to new parameters is obtained through interpolation. The proposed method is applied to reduced-order modeling of unsteady aerodynamic of a two-dimensional NACA64a010 airfoil undergoing pitch motion. The numerical results show that the reduced-order model accommodating to parameter variations can more accurately predict the unsteady aerodynamic induced by small movement of the structure.