Numerical investigations of bio-inspired blade designs to reduce broadband noise in aircraft engines and wind turbines

Bodling, Andrew
Major Professor
Anupam Sharma
Committee Member
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Aerospace Engineering
Organizational Unit
Journal Issue
Aerospace Engineering

This work presents numerical analysis of airfoil geometries inspired by the down coat of the night owl. The objective is to understand the mechanisms of airfoil trailing edge noise reduction that has been observed with such designs in previous experiments. To reduce the computational complexity, first the NACA 0012 airfoil is selected as the baseline airfoil. The bioinspired geometry consists of an array of ``finlets'' that are applied near the trailing edge of the baseline airfoil and are aligned with the flow direction. Wall-resolved large eddy simulations are performed over the baseline and the bioinspired airfoil geometries and the aerodynamic and aeroacoustic performance of the two geometries are contrasted. Both models are simulated at chord-based Reynolds number, $Re_c=5\times10^5$ , flow Mach number, $M = 0.2$, and angle of attack, $\alpha = 0\degree$. The boundary layer in the simulations is tripped with a geometry-resolved trip wire in order to compare with experiments that are at much higher $Re_c$ (of the order of 2 M).

Comparisons with experimental data show good agreement for aerodynamic pressure coefficient ($C_p$) and skin friction coefficient distributions ($C_f$) for the baseline airfoil. The time-averaged wall-normal velocity and Reynolds stresses for the baseline airfoil also agree well with experimental and direct numerical simulations (DNS) data. Farfield noise spectra comparisons between the baseline and the bioinspired airfoil near the airfoil trailing edge show reductions of up to 10 dB with the finlets. The simulations reveal that the finlets lift turbulence eddies away from the airfoil trailing (scattering) edge hence reducing the scattering efficiency. These findings suggest that one of the mechanisms of noise reduction is the increased source-scattering edge separation distance. Reductions in the unsteady surface pressure and velocity fluctuations near the airfoil surface are observed primarily at high frequencies which suggests that the increased source-scattering edge separation distance mainly influences the high-frequency noise. The simulations also show that the finlets reduce spanwise coherence, particularly at low frequencies. This is attributed to be the mechanism of low-frequency noise reduction with the finlets.

Two additional finlet designs are tested on the NACA 0012 airfoil. In one design, the leading edge of the finlet is changed so that the final height is reached in a single ``step''. In the other design, the finlet height is increased from $H=1.00\delta^*$ to $H=2.26\delta^*$, where $\delta^*$ is the displacement thickness. The lack of farfield noise reduction from the single-step finlet demonstrates the importance of having

a finlet leading edge that is highly skewed to the incoming flow. Increasing the finlet height is shown to further reduce the high-frequency noise, but with a higher aerodynamic drag penalty.

The finlets are then applied to a DU96-W-180 baseline airfoil at a non-zero lift condition to permit direct comparisons with the experimental data and to evaluate the hypotheses of the noise reduction mechanisms at realistic operating conditions. Two fences with nondimensional thicknesses, $d/\delta^*=0.107$ and

$0.214$ are investigated. Wall-resolved large eddy

simulations are performed at chord-based Reynolds number, $Re_c=6\times10^5$,

flow Mach number, $M = 0.146$, and angle of attack, $\alpha = -0.2\degree$. The simulation results suggest that the fences should be as thin as possible to minimize the adverse impact on drag and lift. On the suction side near the maximum fence height, there are reductions in the high-frequency surface pressure spectra. However, close to the trailing edge the simulations show no reductions in the high-frequency surface pressure spectra and therefore, no high-frequency farfield noise reduction is predicted. This shows that the reductions in the surface pressure fluctuations \textit{closest} to the trailing edge are ultimately what leads to farfield noise reductions. Larger velocity deficit below the fence height is shown to lead to more reductions in the surface pressure fluctuations. The velocity deficit and source-trailing edge separation distance is shown to work in tandem to reduce the pressure fluctuations near the trailing edge. Although both the pressure and suction side has an increase in velocity near the airfoil surface, since the source-trailing edge separation distance on the pressure side is larger (or large enough), there are still reductions in the surface pressure fluctuations on the pressure side. Higher farfield noise of thicker fence is attributed to scattering of sound from the top surfaces of the fences. Thinner fences are therefore both aerodynamically and aeroacoustically better than thicker fences.