Aeroacoustic Measurements of Airframe Components: in an Open-jet, a Hard-wall and a Hybrid Wind Tunnel Test Section
Marijn Sanders is a PhD student in the research group Engineering Fluid Dynamics. Supervisor is prof.dr.ir. C.H. Venner and co-supervisor is dr. L.D. De Santana from the Faculty of Engineering Technology.
Aircraft noise is a major source of environmental noise that negatively affects human health. Stakeholders and policymakers are nowadays increasingly confronted with protests from residents exposed to high aircraft noise levels, forcing them to change their noise policies. A solution to the aircraft noise problem is the further development and validation of more accurate noise prediction tools and noise reduction technologies. These tools and technologies are crucial to aircraft manufacturers because they help reduce uncertainties arising during the design process of a new aircraft. Wind tunnel measurements play an essential role in the development of noise prediction tools and noise reduction technologies because they provide valuable physics-based data in a controlled environment in a cost efficient way. This data is then be used to develop or validate new (semi-)analytical noise prediction models or extend existing semi-empirical noise prediction models for new or even disruptive aircraft designs. Moreover, computational aeroacoustic (CAA) simulation tools still rely heavily on wind tunnel data for validation.
Aeroacoustic measurements in wind tunnels nowadays commonly use the microphone phased array technique for the localization and quantification of sound sources. However, this type of measurement is subject to considerable uncertainty. This is mainly due to a lack of: (1) an understanding of how the wind tunnel affects the aeroacoustic measurements, (2) a general framework to correct aeroacoustic measurement to a standard free-field condition and (3) an identification and reduction of systematic errors. The work in this thesis aims at understanding and reducing the uncertainty in wind tunnel testing, thereby facilitating the development of more accurate noise prediction tools and noise reduction technologies.
Firstly, the aeroacoustic facility and experimental methods used are described comprehensively. A significant part of the experimental setups and methods used were newly developed as part of this PhD project and are therefore discussed in detail. An open jet, a hard-wall, and a hybrid (Kevlar-walled) test section were designed and are described. A detailed description of the airfoil models and developed instrumentation is also given. The microphone phased array technique and beamforming algorithms are discussed, and a benchmark validation is presented with data from an array benchmark database. Finally, a general framework to correct single microphone and microphone array data to a standard free-field condition is presented.
Secondly, a study is presented comparing trailing edge noise measurements in an open-jet, a hard-wall, and a hybrid wind tunnel test section. This type of comparative study, involving three different test section configurations, is unique in the current literature. The results in this study show that the experimental hardware and acoustic corrections framework used will yield comparable noise levels for microphone phased array measurements in different wind tunnel configurations.
Thirdly, The effect of sweep angle on slat noise characteristics from a common high-lift research model, the 30P30N, was investigated. Aeroacoustic measurements were performed at a 0 degrees and 30 degrees sweep angle. These results showed that aeroacoustic measurements of an unswept highlift model yield the same noise characteristics of a wing with a practical (i.e. 30 degrees) sweep angle.
Lastly, The comparability of slat noise measurements from a common high-lift research model in an open-jet, a hard-wall, and a hybrid wind tunnel test sections was lastly investigated. PIV measurements were performed to capture the 2D time-averaged velocity field in the slat cove flow. The PIV flow measurements showed that the mean flow characteristics in the slat cove were comparable when a similar aerodynamic condition was set based on the Cp distribution of the model. Microphone phased array measurements were also performed to compare the far-field noise levels. This study showed that beamforming maps and integrated ranges for slat noise measurements are comparable, provided that the geometric angles of attack are adjusted to fit the mean flow in the slat cove region. Moreover, it was shown that coherence loss had to be accounted for in the open-jet configuration.
In conclusion, the results of these studies contribute to a better understanding of the uncertainties of aeroacoustic measurements for airframe noise in a small-scale facility. The developed framework for benchmarking, correcting, and presenting microphone phased array measurements shown can be used to reduce systematic measurement errors. In addition, the work helps to better select a test section configuration for a specific measurement campaign. Finally, the measurement data will be made available in an open database for benchmark purposes, with measurements performed in other facilities. In particular, the NACA-63018 and 30P30N datasets are available for benchmarking in the Hybrid Anechoic Wind Tunnel Workshop. In addition, the 30P30N data can be used in the Benchmark for Airframe Noise Computations Workshop to validate numerical simulations.