Skip to main content

Experimental and Numerical Investigation of Turbulent Boundary Layer Effects on Noise Propagation in High Speed Conditions

Final Report Summary - ENITEP (Experimental and Numerical Investigation of Turbulent Boundary Layer Effects on Noise Propagation in High Speed Conditions)

Executive Summary:
New aircraft concepts incorporating ultra-high bypass ratio turbofan or open rotor engine configurations offer significant increases in operational efficiency and economy with potential reductions in environmental impact due to reduced fuel burn. However, the open rotor configurations in particular pose an additional challenge, imposing relatively high noise levels at the surface of the aircraft fuselage, directly influencing passenger comfort and potentially requiring additional noise treatments with associated penalties of increased weight, fuel burn and environmental impact. The noise levels experienced at the fuselage surface can, however, be modified by acoustic refraction occurring during propagation through the boundary layer. In this context, the influence of turbulent boundary layer refraction on the propagation of noise to the surface of a three dimensional rear fuselage geometry has been investigated during the ENITEP project by the generation of an extensive, high speed experimental data base and comparison of these data with the results of computational aero-acoustic (CAA) simulations.
An existing, high speed wind tunnel model instrumented for the measurement of surface acoustics was modified to permit measurement of the steady and unsteady boundary layer aerodynamics. An in-flow noise source was employed to inject broadband and tonal noise at frequencies and separation distances representative of typical contra-rotating open rotor (CROR) configurations. In addition, the noise source was subject to an isolated, wind-on characterisation exercise. A qualitative insight into the refraction phenomena was gained by testing over a wide speed range up to a fuselage reference local Mach number of 0.78. Selected experimental test cases were simulated using Euler and RANS computational fluid dynamics (CFD) approaches to generate flow field inputs for the CAA tasks.
The noise source output was modelled numerically and the accuracy of existing CAA prediction tools was investigated by experimental/numerical comparisons of fuselage peak noise levels and directivities for the selected test cases. In addition, the steady refraction effect was estimated by comparison of CAA results obtained with and without the fuselage turbulent boundary layer mean flow represented.
CAA-derived peak noise levels generally exceeded those measured during test, although comparison of surface acoustic pressure distributions showed a generally good match in terms of directivity. Local acoustic attenuation by the fuselage boundary layer over the speed range M = 0.40 to 0.75 falling in the approximate range 1-12dB has been estimated from the CAA results. A dependence on Mach number and also noise source separation distance, particularly at low speed, has been observed.
In addition to continued development of existing acoustic prediction tools addressing the steady refraction phenomenon, the existing data set will also support future investigation of unsteady (scattering) refraction effects.

Project Context and Objectives:
Project Context:
The noise generated by aircraft has been an important research topic for several decades. Research projects lead to new methods and techniques, improving the understanding of noise generation and propagation mechanisms. This understanding naturally leads to new aircraft concepts and to the reduction of the noise level generated by their engines.
To reduce fuel consumption, the new trends in engine design lead to contra-rotating open rotor configurations (CROR). Despite strong reductions in terms of fuel consumption, the removal of the casing and the associated acoustic treatments leads to unprecedented noise levels impinging upon the aircraft fuselage.
The methods and techniques developed in the past research projects mainly addressed acoustic propagation in steady, complex mean flows. These techniques are relevant for current aircraft designs as the nacelle casing and the wing deviate the acoustic propagation to the ground. For open rotor configurations, the engine placement leads to very high acoustic levels at the aircraft fuselage, and therefore to a new requirement for methods designed to capture the refraction effects through the fuselage turbulent boundary layer. This effect directly impacts the acoustic performance of the aircraft in terms of passenger comfort.
The goal of the current project was to analyse the effects of the steady boundary layer refraction on a simplified three dimensional section of a fuselage. The experimental setup provided a qualitative insight by covering a wide range of experimental flow configurations, while the numerical tasks applied and developed existing methods in order to predict and understand the refraction phenomenon.
Main Objectives:
The overall objectives of ENITEP, summarised in terms of the two principal work packages were as follows:
• WP1: acquisition of experimental acoustic data; utilisation of these data to validate and develop numerical simulation methods in order to better understand the complex phenomena of noise transmission from contra-rotating open rotor (CROR) engines into the aircraft cabin at cruise conditions.
• WP2: analysis of the refraction effects of a steady, mean flow representing the wind tunnel experiment.
The technical work packages were conducted over three project periods with principal objectives as follows:
• Period 1 - preparation for the wind tunnel test campaign through commencement of design and manufacture (D&M) of the test hardware.
• Period 2 - completion of detail design and manufacture of new test hardware, acoustic wind tunnel tests and supporting CFD simulations of selected test configurations in support of WP2
• Period 3 - aero-acoustic simulation of selected wind tunnel test cases through the application and development of existing tools. The success of these tools in simulating the steady boundary layer acoustic refraction phenomenon was then assessed by comparison with the experimental results.

Project Results:
General Approach to the Project:
The ENITEP project comprised two technical work packages: the first aimed at generating an extensive experimental database characterising the fuselage model boundary layer aerodynamics and measuring the surface acoustics, the second comparing these measurements with the results of computational aero-acoustics (CAA) simulations and using predictive tools to estimate the steady refraction effect of the turbulent boundary layer mean flow.
Acoustic propagation through a turbulent boundary layer is also subject to an unsteady refraction phenomenon (scattering). In order to support research in this area beyond the frame of ENITEP, considerable effort was expended early in the project in order to ensure that the experimental measurements would capture this effect also. Thus, a significant proportion of the high speed acoustic wind tunnel testing was devoted to measurement of the fuselage boundary layer unsteady aerodynamics using the constant temperature anemometry (CTA) technique with paired hot wire probes, traversed with multiple degrees of freedom.
The detail design and manufacture (D&M) of the semi-automated, highly productive CTA traversing systems was completed outside of ENITEP, with the added benefit that an additional task could be included to further enhance the wind tunnel data set by characterising the wind-on output of the noise source employed to study the fuselage acoustics.
Following a risk mitigation exercise, the support system for the rear fuselage model was modified in order to minimise aerodynamic blockage and potential flow unsteadiness with the probe traversing systems installed. The noise source was integrated with the ARA TWT by means of the dedicated support system and D&M of pressure rakes for survey of the fuselage boundary layer mean flow were completed. Particular attention was paid to the data logging set-up in order to provide a fully synchronised data set for all acoustic sensors employed. It was also possible to include the hot wire signals in this approach.
High Speed Acoustic Wind Tunnel Tests:
Following Mach number calibrations for tunnel blockage, pressure rake surveys of the fuselage boundary layer were conducted. Boundary layer turbulence was surveyed by means of the traversed CTA probes, the acoustic baseline was established by testing the fuselage in the clean configuration and the noise source output was characterised by in-flow measurements in the absence of the fuselage.
Computational Fluid Dynamics (CFD) Simulations
A subset of four experimental test cases representing the fuselage acoustic baseline configuration were simulated using CFD tools in order to generate flow field definition for input to the CAA tasks. Mach numbers of 0.40 and 0.75 and noise source separation distances of 17cm and 35cm were chosen in order to highlight potential sensitivity of the steady refraction mechanism to speed and noise source proximity. RANS and Euler solutions were obtained, providing a total of eight flow solutions. The RANS results would support CAA solutions for direct comparison with the experimental data. The Euler solutions would support further CAA solutions which could be used to estimate the magnitude of the steady refraction effect.
Numerical Investigations using CAA Tools:
The experimental characterisation of the noise source was employed to build a numerical source model that could be applied across a wide speed range. Using a compact grouping of monopoles to represent the noise source, it was possible to obtain an acceptable match with experiment, in terms of peak levels but especially wind-on directivity at low and high speeds. The average discrepancy was <0.8dB at M = 0.40 and <1.8dB at M = 0.75. The source model was then used in the acoustic simulation of the four, previously-selected fuselage test cases, assuming a steady mean flow for the fuselage boundary layer.
Comparison was made between the experimentally-measured and RANS CFD-based, CAA-derived acoustic pressure distributions on the fuselage surface. Noise source directivity was generally well-simulated, demonstrating the narrowing seen from experiment at the higher speed, although peak levels were significantly higher than the experimental data suggested. A poor match was obtained far upstream on the fuselage surface, where the reduced level of noise propagated from the source was insufficient to overcome the wind tunnel background noise. There were also indications that the presence of the fuselage may have measurably influenced the output of the noise source, the experimental characterisation of which was conducted in isolation.
Comparison between results from the RANS- and Euler-based CAA solutions enabled the character and magnitude of the steady refraction effect to be estimated. Inspection of the fuselage surface acoustic pressure distributions showed that refractive attenuation by the boundary layer, particularly forward of the noise source (CROR rotor) plane, caused significant reductions in noise level at the surface, increasing in the upstream direction. The attenuation has been estimated in the range 1.6-8.0dB at M = 0.40 and 1.1-12.4dB at M = 0.75. The refraction effect also appeared to be more sensitive to source position for the M = 0.40 case. It should be noted that the CAA results did not include unsteady refraction effects (scattering) caused by boundary layer turbulence, although these were expected to be of a lower order than the steady effect.

Potential Impact:
Potential Impact of the Project:
The immediate impact arising from this project is the availability of a unique experimental database which can be used in the future assessment of aero-acoustic noise prediction methods. The wind tunnel model and measurement techniques developed during this programme of work will be transferable to future wind tunnel projects, providing the European industry with increased capability. Furthermore, the numerical tools applied and developed within the frame of ENITEP will assist in modelling the steady refraction effects of turbulent boundary layers.
Advances in the accuracy of noise prediction will aid the design of future aircraft, reducing the noise impact of Counter Rotating Open Rotor Engines while maintaining or improving their efficiency. Improved acoustic performance of future aircraft will result in a reduction in noise generation which will benefit both on-board passenger comfort and the local environment. Reduced acoustic treatment requirements will help to minimise associated weight penalties and fuel consumption and result in a cleaner atmosphere, providing the wider community with a better living environment.
Dissemination and Exploitation of Results:
The results of the ENITEP project have been disseminated internally by the consortium partners via the periodic and work package-related reporting throughout the duration of the project. Comparisons between the experimental and CAA results and an assessment of the steady refraction effect have also been presented at the 2016 AeroTraNet aerodynamics training network meeting (see "Aerotranet2016-FFT_public.pdf", attached). Presentation of these results at the 3AF Greener Aviation conference is also scheduled for 2016.
The experimental data set generated by the ENITEP project would support further investigation and CAA tools development in connection with the steady refraction effect. The current data set, by virtue of the unsteady turbulent boundary layer characterisation, is also ideally suited to improve understanding of the unsteady refraction (scattering) phenomenon. Furthermore, the extensive CTA hot wire results generated within ENITEP may be applied to improve algorithms for the correction of hot wire data in compressible viscous flows.

List of Websites: