Control of Aeroelastic Systems in the Presence of Uncertainty

Suppressing undesirable vibrations in aerostructures is crucial for ensuring the safety and longevity of aircraft. Strict legislation that governs the design of modern jetliners mandates that designs maintain a 15% factor of safety between operating conditions and the conditions at which unstable oscillations may occur. Conventionally, manufacturers have modified structures in the design phase to satisfy such regulations. This is, however, problematic in that the modifications can be expensive and can significantly alter the weight of the aircraft, thus affecting its efficiency. An alternative approach is to use active control as the means by which to perform structural modifications. Using existing control surfaces, a wing's stiffness and damping can be modified actively in a way that reduces or mitigates vibrations. The advantage here is that there is no need to incorporate new structural elements in the wing and it can be implemented on existing hardware, thus not requiring an entire re-design.
In both of the above techniques, it is necessary to predict the conditions at which the unstable vibrations occur. This, however, requires the study of the coupled effects of aerodynamic and structural dynamics, known as aeroelasticity, which is difficult. Indeed, due to manufacturing tolerances, un-modelled dynamics and unknown environmental conditions, it is almost impossible to estimate these conditions with a high degree of accuracy and thus there exists uncertainty.
In recent years, various authors have suggested using uncertainty quantification techniques in the study of aeroelasticity. In this way, the conditions at which aeroelastic phenomena occur are treated as random variables and probabilistic quantities are used.
This project will investigate the applicability of combining uncertainty quantification techniques with active control, for use in aeroelastic systems. Structural and/or aerodynamic properties will be modelled as random variables and response statistics will be used to perform active control in a way that reduces the probability of undesirable aeroelastic phenomena. It is anticipated that new techniques developed will be tested both numerically and experimentally at the University of Liverpool's wind-tunnel model.