Reduced order modelling and control for 2D and 3D airfoils

Charles Hindman, University of Colorado

A method of rapidly developing linearized models of aeroelastic airfoils in 2 and 3 dimensional Euler flows is presented, along with a methodology for creating reduced order models from these linearized models. A 2 dimensional NACA airfoil with 3 degrees of freedom (DOF) and a 3 dimensional Agard wing modeled by the first 8 structural modes are the basis of these reduced order models. These models have 20 - 50,000 DOF in the fluid domain, and require substantial computing time on a parallel computer for time integration. After linearization and reduction of order, only a few degrees of freedom are required for controller design. Controllers are developed based on these low order models and then applied to the nonlinear systems. The controllers are tested against several different grid resolutions. This method also allows the effect of grid resolution on the dominant eigenvalues of the fluid system to be computed.

Abstract Author(s): Charles W. Hindman<br />with Mark Balas and Michel Lesoinne