Search

Microphone array measurements with the DLR research aircraft ATRA flying in different configurations are presented with a focus on the analysis of sources in the engine jets. The free jets exiting the engine nozzles are large distributed source regions with locally and temporally fluctuating source distributions that evolve in the downstream direction. The source maps from the array data calculated with a deconvolution method for moving sources are divided into integration regions for the different noise sources of the engine and the free jet downstream. Frequency spectra of the integrated noise powers from the source regions of the engine inlets, the nozzles, and the jets for fly-overs in three different configurations with high engine speeds are presented and compared.

The integration of advanced propulsion systems into the aircraft, and the assessment of their benefits related to e.g. fuel burn, noise emission and potential further measures, is one of the major activities of the Large Passenger Aircraft (LPA) Platform 1 of the Clean Sky 2 initiative. In order to optimize and assess the potential of advanced propulsion integration, it is necessary to establish suitable design, evaluation, and measurement tools. The so-called Cross-Demonstrator-Capabilities (XDC-) activities have been set up to develop and demonstrate powerful numerical and experimental methods for aerodynamic, aeroacoustics and aeroelastic simulation and measurement tasks, supporting the dedicated demonstrators of LPA. The paper will give an overview of the XDC activities and some examples of the developments related to this demonstrator.

Simultaneous flow and acoustic measurements have been acquired for an aircraft, during ground operation. Measurements of velocity field data with stereoscopic particle image velocimetry (SPIV) on the engine-exhaust jet are correlated with the acoustic-pressure signals of a 120 microphone linear array. The turbulent velocity and the applied-stress components of the Lighthill- and Lilley-analogy source terms in the SPIV region are correlated with the microphone array. Convection effects on the propagating sound waves are neglected in the present study. The self-noise term in the Lighthill quadrupole, shows a hot spot for the correlation in the by-pass to ambient shear layer. This high correlation region is clearly visible in one of the analysed test points only, while in general a poor statistical convergence seems to affect the test-source correlation.

This report by Max Nussbaumer is the result of his summer internship from 1. July 2015 to 11. September 2015 at DLR engine acoustics in Berlin. Under the supervision of Henri Siller, he performed acoustic measurements with a microphone array in the JExTRA jet facility. The jet facility was provided by Robert Meyer, the anechoic set-up of the microphone array has been conceived and constructed by Alessandro Bassetti with the support of Wolfram Hage and Larisa Hritsevskyy, who together with Stefan Funke supported the experiments. Stefan Funke also consulted Max Nussbaumer during the data reduction. Karsten Liesner supported the operation of the jet facility and adapted the control software.

Diese Arbeit zeigt die Entwicklung und die Validierung eines Mikrofonarray-Verfahrens, mit dem erstmals die Untersuchung der gerichteten Schallabstrahlung von Turbofantriebwerken und die Trennung der Schallfeldanteile des Einlaufs, der Düse(n) und des Strahls möglich ist. Das inverse Verfahren mit dem Namen SODIX (source directivity modeling in the cross-spectral matrix) reproduziert die Kreuzspektralmatrix eines Mikrofonarrays unter Verwendung eines neuartigen Ersatzschallquellenmodells, das den Multipolcharakter der aeroakustischen Schallquellen, die Abstrahlcharakteristiken des Triebwerks, konvektive Verstärkungseffekte im Strahl und die Richtwirkung kohärenter Quellgebiete im Strahl abbilden kann. Das Quellmodell basiert auf der Annahme, dass die Amplitude jeder punktförmigen Ersatzschallquelle in Richtung eines jeden Mikrofons unterschiedlich ist und dass sich die Phasen der Schallwellen kugelförmig ausbreiten. Durch die Berücksichtigung der physikalischen Nebenbedingung mathematisch positiver Quellstärken konvergiert das Verfahren unabhängig von der Startlösung zu plausiblen Ergebnissen. Die Dynamik der Methode ist größer als 20 dB und das räumliche Auflösevermögen ist vergleichbar mit dem von Entfaltungsverfahren. Für die Validierung von SODIX wurden Messungen mit einem Turbofantriebwerk und einem linearen Array mit 248 Mikrofonen auf einem Freifeldprüfstand durchgeführt. Die SODIX-Ergebnisse bestätigen, dass die breitbandigen Schallfeldanteile stark gerichtet sind. Es werden Verarbeitungsschritte gezeigt, mit denen die Fernfeldrichtcharakteristiken der einzelnen Quellregionen am Triebwerk und im Strahl aus den SODIX-Ergebnissen berechnet werden können. Ein Vergleich dieser Berechnungen mit Messdaten aus dem Fernfeld belegt die hohe Genauigkeit des Verfahrens: Die Ersatzschallquellen reproduzieren die Referenzmessungen im Fernfeld im Mittel auf 0,5 dB genau. Weiterhin wird gezeigt, dass eine Modifikation des Triebwerks im Bereich der Düse akustisch nachgewiesen werden kann. Strahllärm und Triebwerkstöne werden ebenfalls untersucht. Zusätzlich zu den Freifeldtests wurden Messungen mit demselben Triebwerk in einer Prüfstandshalle durchgeführt. Akustische Messungen in Prüfstandshallen – die bisher nur für Leistungstests verwendet werden – versprechen deutliche Zeit- und Kosteneinsparungen gegenüber den aufwändigen Freifeldtests. Die akustischen Randbedingungen in der Halle werden untersucht und SODIX wird mit Modellen zur Korrektur der Strömung und einer dominanten Reflexion erweitert. Die SODIX-Ergebnisse aus der Prüfstandshalle werden ausführlich mit denen aus dem Freifeld verglichen. Insbesondere für die Berechnungen der breitbandigen Schallabstrahlung aus der Düse sind die Abweichungen kleiner als 2,5 dB. Die Ergebnisse dieser Arbeit belegen, dass SODIX ein stabiles und gut validiertes Verfahren zur Untersuchung der Schallabstrahlung von Turbofantriebwerken ist, das zukünftig von den Herstellern für die effiziente Entwicklung und Validierung neuer, lärmarmer Triebwerkstechnologien im Freifeld und in Prüfstandshallen eingesetzt werden kann. This thesis shows the development and validation of a microphone-array method, which allows for the first time the detailed investigation of the directional sound radiation of turbofan engines and the separation of the sound-field components from the inlet, the nozzle(s), and the jet. The method called SODIX (source directivity modeling in the cross-spectral matrix) is an inverse method that uses point sound-sources to reproduce the cross-spectral matrix of a microphone array. SODIX uses a novel point-source model which takes into account the multipole character of the aeroacoustic sound sources, the radiation characteristics of the engine, convective amplification in the jet, and the directivity of coherent source regions in the jet. The source model is based on the assumption that the amplitude of each replacement sound source is different in the direction of each microphone and that the phases of the sound waves propagate spherically. By taking into account the physical constraint of mathematically positive source strengths, the process converges to plausible results independent of the starting solution. The dynamic range of the method is greater than 20 dB and the spatial resolution is comparable to that of deconvolution methods. For the validation of SODIX, measurements were performed with a turbofan engine and a linear array with 248 microphones on a free-field test stand. The SODIX results confirm that the broadband sound-field components are highly directional. Processing steps are shown which allow the far-field directional characteristics of the individual source regions on the engine and the jet to be calculated from the SODIX results. The comparison of these calculations with measured data from the far field shows the high accuracy of the method. On average, the replacement sound-sources calculated with SODIX reproduce the reference measurements in the far field within 0.5 dB. Furthermore, it is shown that a modification of the engine in the region of the nozzle can be determined acoustically. Investigations of jet noise and engine tones are also performed. In addition to the free-field tests, measurements were made with the same engine in an indoor test-bed. Acoustic measurements in indoor test-beds - previously used only for performance tests - promise significant time and cost savings compared to the extensive outdoor tests. The acoustic boundary conditions in the test hall are investigated and SODIX is extended with models for the flow and a dominant reflection. The SODIX results from the indoor test-bed are compared in detail with those from the free field. In particular, the results of the broadband sound radiation from the nozzle shows deviations less than 2.5 dB. The results of this thesis prove that SODIX is a stable and well-validated method for the evaluation of the sound radiation of turbofan engines, which can be used in the future by the manufacturers for the efficient development and validation of new, low-noise engine technologies both in the free field and in indoor test-beds.