Wavelet-Based Separation Methods Assessment on the Near Pressure Field of a Landing Gear Subcomponent

Abstract : The pressure field in the vicinity of a shear layer developing over the circular cavity installed in a landing gear wheel is investigated numerically by means of wavelet-based filtering algorithms originally developed for the analysis of turbulent jet noise by Mancinelli et al. (J. Fluid Mech. (2017), vol. 813, pp. 716-749). First, the well-known wavenumber-frequency filtering in the Fourier space is performed on a linear array of microphones over the shear layer. Then, four separation algorithms are successively tested at a given observer location on this array. We show that in the specific case of CFD signals, when the mesh cutoff frequency is lower than the Nyquist frequency, the recursive filtering method only extracts the background numerical noise of the simulation due to the band-pass filter behavior of the wavelet daughters at the smallest scales. Resampling the data yields more physical results by suppressing the Gaussian but nonphysical noise of the signal. Overall, important scatter is found between the methods in the low frequency range. An analysis of the filtered pressure decay with distance is finally carried out on the basis of expected near and far-field behavior of the acoustic field. I. Nomenclature (e x , e y , e z) = Cartesian reference frame (x, y, z) = Spatial coordinates L = Pressure field reference length D w = Wheel diameter H w = Wheel thickness R = 0.27D w = Cavity radius κ = H 2R = Cavity depth-to-diameter ratio U ∞ = Free-stream velocity c 0 = Speed of sound M = U ∞ c 0 = Free-stream Mach number y + = Dimensionless wall distance f = Frequency f s = Sampling frequency f m = Nyquist frequency St D w = f D w U ∞ = Strouhal number ∆t = Simulation time step N it = 2 N = Number of samples δ ω = Vorticity thickness ω = Vorticity vector Ψ = Fourier transformed wavelet function ψ = Wavelet in the time domain s, b = Dilation and translation parameters γ j, k = Wavelet coefficients T = Threshold value k = Wavenumber ω = 2π f = Pulsation χ = Goodness-of-fit test variable v ϕ = k −1 ω = Phase velocity τ = Correlation time-delay˜∆ delay˜ delay˜∆ = Dimensionless Laplace operator
Document type :
Conference papers
Complete list of metadatas

Cited literature [20 references]  Display  Hide  Download

https://hal.sorbonne-universite.fr/hal-02140504
Contributor : Philippe Druault <>
Submitted on : Monday, May 27, 2019 - 1:58:13 PM
Last modification on : Wednesday, July 31, 2019 - 11:25:21 AM

File

AIAA2019_060519.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02140504, version 1

Citation

Antoine Hajczak, Laurent Sanders, Vuillot François, Philippe Druault. Wavelet-Based Separation Methods Assessment on the Near Pressure Field of a Landing Gear Subcomponent. 25th AIAA/CEAS Aeroacoustic Conference, May 2019, Delf, Netherlands. ⟨hal-02140504⟩

Share

Metrics

Record views

42

Files downloads

37