Flight-Test Evaluation of Flutter Prediction Methods

Several methods have been formulated to predict the onset of flutter during flight testing. These methods have been demonstrated using data from simulations; however, a rigorous evaluation that includes data from flight testing must be performed. The ability of several methods to predict the onset of flutter by analyzing data from flight tests of the aerostructures test wing is evaluated. The evaluated methods include data-based approachesthat use damping extrapolation, an envelope function, the Zimmerman-Weissenburger flutter margin, and a discretetime autoregressive moving-average model. Also, a model-based approach that uses the µ-method flutterometer is evaluated. The data-based methods are demonstrated to be unable to predict flutter accurately using data from low-speed test points, but converge to the accurate solution as airspeed is increased. Conversely, the flutterometer is demonstrated to be immediately conservative using data from low-speed test points, but these predictions remain conservative and do not converge to the true flutter speed as the envelope is expanded. The operation of a flight test should note the properties of each method to perhaps adjust test points based on the predicted utter margins.