Progressive flutter prediction using flight data with limited sensors and actuators

Abstract

In this work, the flutter prediction theory and method originally developed by the author based on the dynamic eigen decomposition technique are modified for applications to flight flutter simulation and testing of airplanes. Towards this end, frequency responses to control surface excitations are processed at subcritical conditions and the dynamic eigenvalues calculated herein are extrapolated to higher dynamic pressures. To obtain enough responses, the flight test is repeated progressively such that the data set is augmented as more responses are added. By using a recursive formula that relates frequency responses at different dynamic pressures it is possible to expand the columns in the transfer function matrix by integer multiples, thus making the measurements and the excitations equal in number. For demonstration of the new methodology, a tapered wing with four flaps is modeled using ten flexible wing modes and four rigid flap modes. Theodorsen’s theory accounting for motion of the control surfaces is employed for unsteady aerodynamics. It is shown that the proposed scheme as applied to the simulated flight data not only results in extremely accurate predictions of flutter but also is procedurally compatible with the present-day practices, e.g., aeroelastic damping extrapolation, providing a viable option for flutter prediction based on flight test data.