The distortion of turbulence interacting with thick airfoils is analyzed with scale-resolved numerical simulations to elucidate its impact on leading-edge-noise generation and prediction. The effect of the leading-edge geometry is investigated by considering two airfoils with different leading-edge radii subjected to grid-generated turbulence. The velocity field is shown to be altered near the stagnation point, in a region whose extension does not depend on the leading-edge radius. Here, the deformation of large-scale turbulence causes the amplitude of the upwash velocity fluctuations to increase in the low-frequency range of the spectrum because of the blockage exerted by the surface. Conversely, the distortion of small-scale structures leads to an exponential decay of the spectrum at high frequencies due to the alteration of the vorticity field. The prevalence of a distortion mechanism over the other is found to depend on the size of the turbulent structures with respect to the curvilinear length from the stagnation point to the location where surface-pressure fluctuations and pressure gradient peak. This occurs at the curvilinear abscissa where the curvature changes the most. The same high-frequency exponential-decay slope observed for the upwash velocity is retrieved for surface-pressure spectra in the leading-edge region, suggesting that the airfoil unsteady response is induced by the distorted velocity field. This physical mechanism can be accounted for in Amiet's model by using a distorted turbulence spectrum as input and accounting for the increased amplitude of the distorted gust in the aeroacoustic transfer function, retrieving an accurate noise prediction for both airfoils.

@en

This study analyzes the alteration undergone by turbulent eddies as they approach a propeller operating at low Reynolds number, with the purpose of investigating the resulting effects on the noise emitted by the propeller. The two mechanisms affecting turbulence distortion, the streamtube contraction and the interaction between the turbulent structures and the blade, have been investigated experimentally. The set-up consists of a propeller with a diameter of 30 cm operating at a 75% chord-based Reynolds number of 10.8 × 104 interacting with the turbulence produced by a rectangular grid. The flow behavior has been studied by particle image velocimetry (PIV) and hot-wire anemometry (HWA), while a microphone arc was installed for the acoustic analysis. The results reveal that the interaction between incoming turbulence and the propeller plays a dominant role in the alteration of turbulence with respect to the streamtube contraction. This is due to the relatively low contraction ratio of the propeller at this regime, equal, in this case, to C.R.=1.3. Turbulence characteristics are used as input for two different analytical noise-prediction models, both based on Amiet’s theory for turbulence-impingement noise. The first implements the original formulation of Amiet for propeller noise, which requires a position along the blade to be specified to define all the inputs. The second has been developed in the present work to account for the variations of the blade geometry and turbulence conditions in the radial direction. The comparison between the noise predictions and the experimental measurements shows that a better agreement can be obtained with the second model. This reveals that noise generation is strongly dependent on the variation of the flow conditions and propeller geometry along the radial direction, confirming that the description of these characteristics can enhance the accuracy of low-fidelity noise-prediction methods.

@en

A physical analysis has been conducted to assess the effects of turbulence distortion on leading-edge noise generation and low-fidelity modeling in the framework of Amiet's theory. This model retrieves the power spectral density (PSD) of far-field noise using as input the upwash velocity spectrum of incoming turbulent flow and an aeroacoustic transfer function modeling the aerodynamic and acoustic response of the airfoil to the perturbation. The study has been carried out by investigating the interaction of grid-generated turbulence with a NACA 0012 and a NACA 0012-103, featuring the same thickness but different leading-edge shape. The alteration of the velocity field has been shown to occur in agreement with the analytical findings of the rapid distortion theory (RDT), identifying in particular an exponential decay for the upwash-velocity component spectrum at high frequencies. The same decay is observed for the surface-pressure spectra close to the leading edge, suggesting that sound is generated by a pressure distribution induced by the altered velocity field and hence proving that turbulence-distortion effects should be included to enhance the noise-prediction accuracy of low-fidelity models. This has been further demonstrated by using as input in Amiet's model the altered upwash velocity spectrum sampled in the immediate vicinity of the leading edge: an improved prediction of the high-frequency decay is yielded, but the evident overestimation of the noise levels with respect to the results provided by the Ffowcs-Williams and Hawkings' (FWH) analogy indicates the necessity of correcting the aeroacoustic transfer function as well.@en

The analytical model for leading-edge noise prediction formulated by Amiet, developed for a flat plate, relates the far-field acoustic pressure to the upstream inflow conditions, modeled by canonical turbulence spectra. The inaccurate results provided by this low-fidelity method when applied to thick airfoils has been attributed to the distortion experienced by turbulent structures when approaching the airfoil, not modeled in the original formulation of Amiet. The first attempts to account for the effects of this physical mechanism consisted of modifying the term representing the incoming turbulence by means of the analytical results of the rapid distortion theory, obtaining a promising improvement of the noise-prediction accuracy. This paper aims to set up the physical framework to investigate the relation between turbulence distortion and noise-generation mechanisms with the purpose of enhancing inflow-turbulence noise modeling. A numerical database obtained for a rod-airfoil configuration has been chosen to allow the analysis of the vortex dynamics when interacting with a body. The analysis of the velocity field near the leading edge has highlighted that the extension of the region where turbulence distortion occurs depends on the size of the incoming turbulence structures. Furthermore, surface pressure fluctuations have been observed to peak at the same position along the airfoil where the pressure gradient in the streamwise direction is maximum. A novel approach has been proposed to account for turbulence distortion in Amiet's model by using as input the turbulence spectrum directly sampled in this position. A satisfactory agreement with the prediction provided by the solid formulation of the Ffowcs-Williams and Hawkings analogy has been obtained.@en

The present study assesses the ability to numerically predict turbulence-interaction noise of a NACA0012 airfoil with grid-generated turbulence by utilizing the Lattice Boltzmann solver PowerFLOW. Both the near-field flow characteristics and far field noise are bench-marked against an existing experimental study. The grid was chosen to match that from the experiment to provide evidence that the present numerical approach in physically placing a grid upstream of the airfoil can reproduce the turbulence characteristics observed from the benchmark experiment and thus accurately capture the turbulence-interaction noise generated. The comparison of the results show that the turbulence statistics, including turbulence intensity, integral length scales and anisotropy are highly consistent with the experiment. Moreover, far field acoustics of the turbulence interaction as well as the near-field flow properties near the leading-edge and the unsteady wall pressure fluctuations of the airfoil are also analyzed and the results agreed well with the experimental measurements. The present study confirms that the grid-generated approach is suitable for numerical investigation of turbulence-interaction noise and its potential mitigation strategies.@en