Abstract:
The scope of the work described in this thesis is studying airfoil noise reduction using porous treatments applied to the airfoil trailing edges. To that end, airfoil noise
was predicted using Computational Aeroacoustics (CAA) and analytical methods. The
porous material was modelled using the Darcy Forchheimer model and an impedance wall
boundary condition. The Darcy-F orchheimer model implemented in the OpenFOAM
computational fluid dynamics (CFD) code was verified numerically against available experimental results of the lift and drag coefficients for completely porous SD7003 airfoil
and theoretical results of the pressure drop across porous material for pipe flow blocked
by porous material. OpenFOAM was utilised for conducting a study of the aeroacoustic
noise produced by airflow over a solid NACA 0018 airfoil using a large eddy simulation
(LES) to calculate the noise sources, then a numerical implementation to the solution
of the Ffowcs Williams & Hawkings (FW-H) equation was used to predict the radiated
noise. The predicted noise spectra were underpredicted in the low-frequency range as
the surface pressure fluctuation spectra near the trailing edge showed the same behaviour
in that frequency range. Further analysis was performed to investigate different factors
that could affect the noise prediction (the numerical divergence scheme, the LES sub-grid
scale (SGS) model, grid refinement, the geometry of the turbulent boundary layer trip,
the size of the computational domain and the use of an acoustic sponge zone). Some
factors, such as using a linear divergence scheme, grid refinement in the spanwise direction and the shape of the domain, improved noise predictions in the low-frequency range,
but the predicted levels were still lower than the experimental data. The analysis proved
that OpenFOAM is not a suitable CFD code for aeroacoustic studies, likely due to the
low order of the numerical spatial and temporal discretisation schemes utilised. Then, a
High-Performance Solver for Turbulence and Aeroacoustic (HiPST AR) code was proposed
for conducting an aeroacoustic study for a solid controlled diffusion (CD) airfoil, using
the same methodology adopted in OpenFOAM studies. The surface pressure fluctuations
spectra showed a good match with direct numerical simulation (DNS) numerical results.
The noise reduction effect of applying an impedance boundary condition to 20 % of the
chord length from the trailing edge was studied. Applying the impedance boundary condition produced no significant noise reduction. Finally , the trailing edge noise prediction
was calculated using analytical models for the NACA 0018 airfoil case. The predictions
calculated using these models were compared against the numerical results calculated
from the OpenFOAM simulations. The noise levels predicted using the analytical models
were lower in the low-frequency range than the experimental data and OpenFOAM numerical result, as these models depend on surface pressure fluctuation spectra near the
airfoil trailing edge, which were underestimated in the same frequency range.