Installation effects on airfoil self-noise estimated by direct numerical simulations

Abstract

Several Direct Numerical Simulations (DNS) have been achieved using the Lattice-Boltzmann Method (LBM) on a Controlled-Diffusion (CD) airfoil at $R{e}_c=1.5\times 10^5$ and $\alpha =8°$ to evaluate installation effects on airfoil self-noise. 2D DNS provides unrealistic airfoil noise sources and a significant overprediction of the far-field noise with possibly a change in the ranking of the noise sources. Conversely, all 3D simulations of the airfoil immersed in the wind tunnel jet show excellent agreement with the flow around the airfoil and the far-field noise. Two new 3D DNS with some shear-layer refinement and possibly an additional zigzag trip to trigger turbulence have better resolved the jet development. Even though all 3D DNS identify two main noise sources, the laminar separation bubble (LSB) at the leading edge and the turbulent eddies at the trailing edge, their intensity varies because of a strong coupling between the jet and the airfoil that yields different LSB sizes, increased inlet turbulent intensity, oscillations of the angle-of-attack, and variations in the noise sources. The state of the jet shear layer at the nozzle exit of a given open-jet anechoic wind tunnel is seen to influence the flow field around the airfoil and its noise radiation, with potentially a larger jet contribution.