Flutter-boundary prediction using system identification-based reduced-order aeroelasticity analysis

Abstract

The paper is concerned with the prediction of flutter boundary using computational aeroelastic techniques and system identification-based reduced-order aeroelasticity analysis. The method consists in constructing a continuous time state-space model of the closed-loop aeroelastic system using numerical simulation results, and followed by performing eigenvalue analysis of the resulting state-space model to determine the flutter boundary. The work presented here is developed based upon our previous work.20, 26In the present paper, we consider CFD data for the training of ROM and nonlinear flutter dynamics represented by a Hammerstein model comprising a cubic static nonlinearity followed by a matrix ARMA linear block, and develop the correlation least-squares method of identification. Prediction of the flutter boundary of the AGARD 445.6 wing by eigenvalue analysis yields results that are very close to experimentally obtained flutter parameters, and are an improvement over method based on only linear models. © 2012 by Lai Kwok Leung.