Flutter Prediction Using Continuous Wavelet Transform

This paper proposes a new method to predict flutter based on the wavelet transform, which transforms time signals into the localized time-frequency domain. A flutter test was conducted in a low speed-wind tunnel to obtain experimental data in the sub-critical region. The prediction technique was applied to the response signals of a low-aspect-ratio wing model subjected to the wind-tunnel turbulence. It was found that the linear extrapolation of a new index defined in the transformed domain could well predict the flutter speed. Other prediction methods were also examined using the same signals to compare predictability.