Towards Full Aircraft Airframe Noise Prediction: Lattice Boltzmann Simulations

Abstract

Computational results for an 18%-scale, semi-span Gulfstream aircraft model are presented. Exa Corporation's lattice Boltzmann PowerFLOW® solver was used to perform time-dependent simulations of the flow field associated with this high-fidelity aircraft model. The simulations were obtained for free-air at a Mach number of 0.2 with the flap deflected at 39º (landing configuration). We focused on accurately predicting the prominent noise sources at the flap tips and main landing gear for the two baseline configurations, namely, landing flap setting without and with gear deployed. Capitalizing on the inherently transient nature of the lattice Boltzmann formulation, the complex time-dependent flow features associated with the flap were resolved very accurately and efficiently. To properly simulate the noise sources over a broad frequency range, the tailored grid was very dense near the flap inboard and outboard tips. Extensive comparison of the computed time-averaged and unsteady surface pressures with wind tunnel measurements showed excellent agreement for the global aerodynamic characteristics and the local flow field at the flap inboard and outboard tips and the main landing gear. In particular, the computed fluctuating surface pressure field for the flap agreed well with the measurements in both amplitude and frequency content, indicating that the prominent airframe noise sources at the tips were captured successfully. Gear-flap interaction effects were remarkably well predicted and were shown to affect only the inboard flap tip, altering the steady and unsteady pressure fields in that region. The simulated farfield noise spectra for both baseline configurations, obtained using a Ffowcs-Williams and Hawkings acoustic analogy approach, were shown to be in close agreement with measured values. NomenclatureAOA = angle-of-attack C L = lift coefficient Cp' rms = unsteady RMS pressure coefficient Cp = pressure coefficient c = local chord DNS = Direct Numerical Simulation DES = Detached Eddy Simulation MDDES = Modified Delayed Detached Eddy Simulation GAC = Gulfstream Aerospace Corporation Hz = Hertz, cycles per second LES = Large Eddy Simulation M = Mach number PSD = power spectral density in [psi 2 /Hz] or [dB/Hz] psi = pounds per square inch Re = Reynolds number RMS = root mean square * Aerospace Engineer, Computational AeroSciences Branch, Associate Fellow AIAA AIAA 2014-2481 AIAA Aviation 2 RNG = Re-Normalization Group s = second URANS = unsteady Reynolds averaged Navier-Stokes VR = variable resolution X, Y, Z = right handed coordinate system

Aeroacoustic Evaluation of Flap and Landing Gear Noise Reduction Concepts

“…Region 4 covers the entire aircraft model, facilitating direct comparisons with the companion computational simulations. 24,25 In this paper we present PSD results obtained from integrating the UDAMAS contour maps over the entire aircraft (Region 4). Results based on Regions 1, 2, and 3 will be presented at a later date.…”

“…The underlying causes for the observed reduction in gear noise (in the presence of the flap) are being investigated using the companion high-fidelity simulations that were recently completed. 24,25 The baseline spectra of Fig. 29 are repeated in Fig.…”

Application of the Lattice Boltzmann Method to Shear Layer Flows

“…4,[6][7][8][9][10][11] Additional validation studies performed with PowerFLOW can be found here. [12][13][14][15][16][17] …”

An Assessment of Flap and Main Landing Gear Noise Abatement Concepts

“…PSD results from Regions 1 and 4 have been used to facilitate direct comparisons with the companion computational simulations. 19,20,21 In this study mostly sample spectra obtained from integrating the UDAMAS contour maps over the wing-flap area (Region 1) are shown. Results based on integration of Regions 2, 3, and 4 are presented occasionally to reinforce certain trends.…”