We report the parameter estimation results of a self-sustaining aeroelastic oscillator. The system is composed of a rigid wing that is elastically mounted on a rig, which in turn is fixed in a wind tunnel. For certain flow conditions, in particular dictated by the Reynolds number in the transitional regime, the wing extracts energy from the flow leading to a stable limit cycle oscillation. The basic physical mechanism at the origin of the oscillations is laminar boundary layer separation, which leads to negative aerodynamic damping. An empirical model of the aeroelastic system is proposed in the form of a generalized Duffing-van der Pol oscillator, whereby the linear and nonlinear aeroelastic terms are unknowns to be estimated. The model (input) noise process accounting for the amplitude modulation observed from experiments will also be estimated. We apply a Bayesian inference based batch data assimilation method in tackling this strongly nonlinear and non-Gaussian model. In particular, Markov Chain Monte Carlo sampling technique is used to generate samples from the joint distribution of the unknown parameters given noisy measurement data. The extended Kalman filter is utilized to obtain the conditional distribution of the model state given the noisy measurements. The parameter estimates for a third order generalized Duffing-van der Pol oscillator are obtained and marginal and joint probability density functions for the parameters will be presented for both a numerical model and a rigid wing that is elastically mounted on a rig in a wind tunnel.

This content is only available via PDF.
You do not currently have access to this content.