A Mixed Model Reduction Technique for Aeroservoelasticity

Charles Hindman, University of Colorado

Active flutter suppression has been a topic of interest for some time and computational methods have been used to simulate the time response of aeroservoelastic systems. Though these simulations are valuable for validations and other non-linear problems, time-integration methods do not easily provide all of the information required for controller design for an aeroservoelastic system. For such purposes it is necessary to produce a reduced order model of the system to be controlled. In the current context, one of the requirements for a reduction technique is that it be derived from a computational model analogous to the ones used for time integration. Several approaches have been recently developed to tackle this problem, notably methods designed to either reduce the fluid part of the coupled system or the complete aeroelastic system.

The purpose of an active control system for flutter suppression is to stabilize the aeroelastic coupled modes that are unstable. Therefore it seems natural to use the second model reduction technique as the basis for the representation of the complete system to be controlled. However, the aerodynamic of the actuation by way of the motion of a control surface has a characteristic time that is rather small when compared to the period of one oscillation of the coupled system. Therefore we suggest that the two aforementioned model reduction techniques can be successfully combined for the analysis and design of an aeroelastic system. In this poster, we present an approach that allows the designer of an active control system, or active flutter suppression system, to obtain a reduced order system that combines the representation of the aeroelastic modes to be controlled and of the aerodynamic of an actuation by control surface motion. We also present a simple example of the technique, as well as our current progress regarding theoretic stability analysis and a more complex aeroservoelastic system example.

Abstract Author(s): Charles Hindman