Numerical noise prediction and source identification of a realistic landing gear

Abstract

Noise predictions of realistic landing gear configurations are obtained combining high-fidelity CFD simulations and the Ffowcs Williams–Hawkings (FWH) acoustic analogy. The aeroacoustic prediction of such a complex aeronautical system depends on geometry fidelity and the ability of the mesh and numerical method to resolve important flow features responsible for noise generation. To understand the role of small components in the far-field noise predictions, different setups are analyzed including a detailed full complexity landing gear and a simplified, yet realistic, configuration. For the latter case, two different setups in terms of mesh resolution and topology are analyzed. For the finer mesh, refinement is applied in regions with strong pressure fluctuations and also close to edges where shear layers develop flow instabilities. We also compare noise predictions obtained by a landing gear installed on a full aircraft with those computed only for a setup with the bottom half of the fuselage. An assessment of the solutions obtained from solid and permeable FWH surfaces is presented for different configurations of permeable surfaces. One objective of this work consists in the identification and analysis of noise sources in the landing gear. For this task, we first employ acoustic analogy to individual components of the landing gear to identify potential sources of tonal noise. Then, proper orthogonal decomposition is applied to identify mechanisms of noise generation at specific frequencies. It is shown that turbulent coherent structures are responsible for tonal noise generation by the towing link, wheel cavities and landing gear compartment. For the external cavity of the wheel, a Rossiter mode is excited and leads to resonance. We demonstrate that removing small-scale components of the full complexity geometry affects the flow characteristics inside the landing gear compartment and, consequently, its noise emission. In this case, the presence of small scale components leads to finer turbulent eddies that generate more noise at higher frequencies. These finer scales also reduce the coherence of larger turbulent structures, reducing amplitude of the low frequency band of the spectrum.