Flutter Boundary Identification for Time-Domain Computational Aeroelasticity

Abstract

Three time-domain damping/frequency/flutter identification techniques are discussed; namely, the moving-block approach, the least-squares curve-fitting method, and a system-identification technique using an autoregressive moving-average model of the aeroelastic system. These methods are evaluated for use with time-intensive computational aeroelastic simulations, represented by the aeroelastic transient responses of a double-wedge airfoil and three-dimensional wing in hypersonic flow. The responses are generated using the NASA Langley CFL3D computational aeroelastic code, in which the aerodynamic loads are computed from the unsteady Navier-Stokes equations. In general, the methods agree well. The system-identification technique, however, provided quick damping and frequency estimates with minimal response-record length. In the present case, the computational cost required to generate each aeroelastic transient was reduced by 75%. Finally, a flutter margin for discrete-time systems, constructed using the autoregressive moving-average approach, is evaluated for use in the hypersonic flow regime for the first time. For the binary-mode case, the flutter margin exhibited a linear correlation with dynamic pressure, minimizing the number of responses required to locate flutter. However, the flutter margin was not linear for the multimode system, indicating that it does not perform as expected in all cases.