[en] Modern transport aircraft generally fly in the transonic flow regime. This regime is characterized by complex flow phenomena such as moving shock waves, shock-boundary layer interactions and separated flows. These aerodynamic nonlinearities can significantly impact the prediction of the flutter aeroelastic instability, which is of great importance for aircraft safety and performance.
This thesis presents a novel unsteady aerodynamic modeling methodology for predicting the transonic flutter of 2D and 3D configurations such as aircraft wings. The main idea of this methodology is to obtain the flow response to small amplitude periodic deformations of a structure over a range of frequencies through the interpolation of few dominant flow modes. These flow modes can be obtained by dynamic mode decomposition (DMD) of unsteady Euler or RANS simulations at different oscillation frequencies or by harmonic balance (HB) simulations. This methodology based on dynamic mode interpolation (DMI) can then be used to obtain a generalized aerodynamic force matrix in the frequency domain, and aeroelastic stability analysis can be performed using industry-standard flutter analysis techniques such as the p-k method.
The methodology is demonstrated for 2D and 3D benchmark transonic aeroelastic configurations. It highlights that the nonlinear aspects of transonic flows are important for studying the aeroelastic stability of aircraft wings. The methodology is successfully applied to calculate the transonic flutter of a realistic airliner wing and its performance is compared to standard industrial methods. The methodology is more accurate than the panel methods primarily used in the aerospace industry leading to safer wing designs while being faster than the higher fidelity fluid-structure interaction (FSI) simulations, offering thus a promising technique for solving dynamic aeroelasticity problems.
