Unsteady Aerodynamic Model Order Reduction for Aeroservoelastic Optimisation by Balanced Proper Orthogonal Decomposition and the use of Synthetic Mode Shapes
Simon Binder, Andreas Wildschek, Roeland De Breuker
Abstract
The combined optimisation of aircraft structures and active control systems in early design stages requires low-order unsteady aerodynamic models that are robust to structural and control surface layout modifications.
A combination of the balanced proper orthogonal decomposition with the concept of synthetic modes is proposed and different formulations for the generation of synthetic modes are given.
The models resulting from the proposed procedure are tested for their suitability and accuracy in aero(servo)elastic analyses carried out during regular sizing-type optimisation of aircraft (stability assessment, continuous turbulence loads and calculation of control surface transfer functions).
Applied to a full aircraft aerodynamic model, the results indicate that the number and type of synthetic modes have a significant influence on the achievable accuracy and the accuracy per order of the resulting reduced order model.
At a required accuracy of 10⁻³, the most suitable set of synthetic modes is based on radial basis functions and reduces the aerodynamic model order by about two orders of magnitude while still being able to handle structural and control surface layout modifications.
A combination of the balanced proper orthogonal decomposition with the concept of synthetic modes is proposed and different formulations for the generation of synthetic modes are given.
The models resulting from the proposed procedure are tested for their suitability and accuracy in aero(servo)elastic analyses carried out during regular sizing-type optimisation of aircraft (stability assessment, continuous turbulence loads and calculation of control surface transfer functions).
Applied to a full aircraft aerodynamic model, the results indicate that the number and type of synthetic modes have a significant influence on the achievable accuracy and the accuracy per order of the resulting reduced order model.
At a required accuracy of 10⁻³, the most suitable set of synthetic modes is based on radial basis functions and reduces the aerodynamic model order by about two orders of magnitude while still being able to handle structural and control surface layout modifications.
Full Text: PDF
ISSN 1974-5117