Flutter Boundary Prediction in a Global Stochastic Framework

Abstract

A novel global parametric system identification framework is introduced in this work for aeroelastic modeling under varying flight states. The presented framework is based on the functionally pooled (FP) time-series models that enable explicit analytical inclusion of any admissible flight states into the model parameters and thus the system dynamics. In this paper, the autoregressive (AR) type of the FP model, which is designated as the FP-AR, is employed to interpret the aerodynamics of a wing structure. The data were recorded by accelerometers during a dedicated wind-tunnel flutter test with the airspeed increasing all the way to the flutter boundary. Including the varying flight states defined by the increasing airspeed into the system modeling via the FP technique, the global framework provides a more sophisticated identification results compared with the traditional nonparametric Welch-based spectral estimation. Flutter occurrence is indicated by the stability margin of the aeroelastic system evaluated by the estimated FP-AR parameters based on Jury’s stability criterion. For online application purpose, an iterative state-based FP-AR process is also proposed in this paper. Compared with the standard FP-AR modeling, the iterative realization is a developing process that provides a global approximation of the system dynamics at each flight state. Experimental evaluation demonstrates the feasibility and effectiveness of the proposed global framework in aeroelastic modeling while facilitating an accurate flutter boundary prediction.