The Effects of Damage and Uncertainty on the Aeroelastic / Aeroservoelastic Behavior and Safety of Composite Aircraft

An experimental model of a typical three degree-of-freedom airfoil section with composite control surface subjected to a two-dimensional, incompressible air-flow has been created. The flutter response of the system with both, a defect-free and with damaged control surfaces, were measured experimentally. An analytical model was implemented to predict the flutter response in the case of a defect-free control surface. Flutter predictions were obtained using both the U-G method and the Root Locus Technique. Measurements obtained using the defect-free rudder compared well with predictions, validating the analytical model and flutter predictions. Measurements with the damaged rudders show that some forms of damage may have little impact on flutter speed, whereas other forms of damage may have a pronounced effect. In general, damage to the rudder decreases the flutter speed. INTRODUCTION Air flowing over a flexible surface, for example an aircraft wing, results in aerodynamic forces applied to the surface. The resulting surface deformations may increase or decrease the initial aerodynamic force. An increase in aerodynamic force will cause still greater surface deformations, whereas a decrease will cause an elastic recovery of surface deformations. Hence airflow can cause a feedback process involving aerodynamic, elastic, and inertial forces. Aeroelasticity is the study of this feedback process. Flutter is one of several important aeroelastic phenomena. Flutter is a structural vibration caused by a constant, steady-state air flow over the surface. In extreme cases flutter can cause serious structural damage. Although most commonly associated with aircraft, flutter can occur for any elastic structure exposed to a steady-state airstream. This paper describes a laboratory model representing a three degree-of-freedom airfoil section with composite control surface that has been created to study aeroelastic effects observed during wind-tunnel tests. The focus of current studies is on flutter, although the model has been designed to allow interrogation of additional aeroelastic factors such as limit cycle oscillation, buffeting, or control system reversal. Previous studies [1] compared flutter speeds measured experimentally to predictions, obtained using both the U-G method [2] and the Root Locus Technique [3], for the system with defect-free control surfaces. Both these numerical methods and the laboratoty model developed in the present study are based on the statespace model developed by Edwards et al. [4] and summarized in Figure 1. During the present work flutter speeds exhibited by an airfoil with an intentionally damaged composite control surface is measured experimentally.