Broadband Trailing-Edge Noise as a Canonical Problem for Airframe Noise Predictions

Kurzfassung

Since 2010, the American Institute of Aeronautics and Astronautics (AIAA) and the Council of European Aerospace Societies (CEAS) have organized an ongoing series of workshops dedicated to Benchmark Problems for Airframe Noise Computations (BANC). The BANC workshops are aimed at enabling a systematic progress in the understanding and high-fidelity predictions of airframe noise. These collaborative investigations integrate state of the art computational fluid dynamics, computational aeroacoustics, and comprehensive, holistic measurements in multiple facilities targeting a selected set of canonical yet realistic configurations. The current contribution reviews the activity conducted during the past years within the BANC category 1 framework. Category 1 focuses on the prediction of broadband turbulent boundary-layer trailing-edge noise (TEN) and related source quantities. The previous Third Workshop on Benchmark Problems for Airframe Noise Computations, BANC-III, was held on 14–15 June 2014 in Atlanta, Georgia, USA. Since the forerunner BANC-II workshop identified some room for improvements in the achieved prediction quality in category 1, BANC-III-1 relies on the same five test cases, namely 2D NACA0012 and DU96-W-180 airfoil sections at different angles of attack in a uniform flow at selected test velocities. Participants were requested to calculate the following relevant acoustic and aerodynamic quantities: • farfield one-third-octave band TEN spectra for a 90-deg overflight observation angle, referenced to a unit span and observer distance, and corresponding directivities; • unsteady surface pressure (point) power spectral densities at both the suction side and pressure side of the airfoil at 99% of the chord length; • chordwise distributions of the static surface pressure- and wall friction coefficients; • chord-normal profiles of the normalized mean flow velocity, normal Reynolds stresses (if available), and kinetic energy of turbulence in the near wake at 100.38% of the chord length; • similarly, chord-normal distributions of the isotropic turbulence mean dissipation rate and longitudinal integral length at 100.38% of the chord length; • integral 'boundary-layer' parameters, derived from the near-wake mean flow profiles, e. g. the boundary layer displacement thickness or momentum loss thickness. Contributions of four scientific groups were submitted. As already experienced during BANC-I/II-1 mainly users of relatively fast hybrid prediction approaches, developed in an industrial context, submitted data for code-to-code comparisons and still, these contributions are less than representative of the broad TEN simulation community. Therefore, it is hoped to also motivate future contributions from the European X-Noise network. BANC-III-1 results are surveyed from the following simulation methods and the following institutions: • The University of Stuttgart (Institute of Aerodynamics & Gas Dynamics, IAG) and the Technical University of Denmark (DTU) applied 2D-RANS-based simplified theoretical surface pressure models of the 'Blake-TNO'-type. Here, turbulent boundary-layer quantities are extracted from a CFD/RANS computation and correspondingly predicted surface pressure spectra are fed into a classical (diffraction theory) farfield noise prediction. • The German Aerospace Center (DLR) contributed results from hybrid RANS-based 2D-CAA simulations, coupled with 3D stochastic source modeling, where fluctuating sources are reconstructed based on the RANS statistics. • The Politecnico di Torino in cooperation with wavePRO S.R.L. and Metacomp Technologies, Inc. considered a new strategy, based on a numerical, hybrid RANS/LES (IDDES) of the nearfield noise, coupled with a FWH solution in the farfield. Synthetic turbulence is used to seed resolved-scale turbulence in the boundary layer to provoke the wake unsteadiness which is otherwise not self-initiating for the selected NACA0012 test case. A nominally 3D FWH solver was applied to propagate the noise from a single (effectively 2D-IDDES) slice. The achieved results display a high scientific quality level. In most of the cases farfield TEN predictions are within or very close to the provided data scatter band (reproducing systematic errors between multiple test facilities), see Figure 1. Compared to BANC-II particularly deviations among predictions for the DU96-W-180 test case could be significantly reduced. TEN maxima are principally well-represented in terms of frequency and level. However, mutual comparisons show individual code-specific advantages and disadvantages, indicating that a methodology which comprehensively predicts all of the requested nearfield and farfield quantities is not available to date. General trends (like the dependence of farfield noise or surface pressure spectra on angle-of-attack and freestream velocity) are not always correctly represented. Overall, the category 1 workshop problem remains a challenging simulation task due to its high requirements on resolving and modeling of turbulent boundary-layer source quantities. Finally, the updated category 1 problem statement for the forthcoming BANC-IV workshop will be introduced. Particularly, the BANC experimental database is currently being extended by additional test cases and supplementing datasets, provided by DTU Wind Energy and DLR. A mid- to long term perspective for future workshops is seen in the consideration of noise reduction devices as a demanding task for the CAA community. For developers of enhanced 'Blake-TNO' models the treatment of airfoils with moderate flow separation could be another selectable field.