Noise is a major environmental pollutant, which can inict physical and psychologicalinjury. An important noise contribution stems from aeroacousticsources, which are found in e.g., ventilation ducts, engine exhaust systems andairplane engines.In this thesis, research methods for low Mach number aeroacoustic sourcesin ducts are developed. The basis of the methods is the ability to describe theintrinsic linear properties of the source as an N-port (multiport), where theoutput sound eld is related to the input sound eld and the generated soundeld, all consisting of plane waves. The methods presented are both numericaland experimental. The numerical method treats the passive properties, i.e.,scattering, attenuation and amplication of incident sound, while the experimentalmethod treats the active part (intrinsic sound generation) as well. Themethods are applied in the study of noise generation by a vortex mixer plate,placed in an airow of Mach 0.2, and in the study of acoustic-hydrodynamicinteraction in a T-junction of rectangular ducts.It is found that the accuracy of the experimental methods is signicantly increasedwhen the equations are over-determined, which is achieved by addingadditional microphones to the test rig. In the frequency range studied, themixer plate is found to generate less sound when made exible, without disturbingthe mixing quality.For the numerical method { based on the linearized Navier-Stokes equations,a model of the oscillation of the Reynolds stress (\turbulent damping")due to the acoustic eld is introduced. By comparing with experimental results,it is found that not using this model results in an over-prediction of theamplication at higher frequencies with several factors in magnitude, whileimplementing the model results in a much better agreement. / <p>QC 20121123</p>

Continued growth of the aviation industry and increasingly strict noise requirements set by international bodies and airport authorities alike means that novel methods of reducing aircraft noise must be found. Engine noise represents a majority contribution to total aircraft noise during take-off and turbulent mixing of the exhaust gases is the dominant noise source of the engine at take-off. While bypass ratio has been the historical, and rather convenient means, of reducing jet noise, an upper limit to bypass ratio is now being approached and additional means of reducing jet noise must be found. One method that has shown potential for reducing aeroacoustic jet noise is the application of small, high pressure jets to the circumference of the jet nozzle. These jets, termed microjets, have the advantage over static devices that the microjets can be activated only when the noise benefit is required and deactivated when emitted noise is not an issue, such as in cruise, thereby reducing the thrust penalty associated with the devices over the majority of the flight. Large eddy simulations have been performed to investigate the impact that the addition of microjets has on the aerodynamic flowfield and radiated far-field noise of a high Reynolds number, Mach 0.9, propulsive, laboratory scale jet. Far-field noise was predicted through a new implementation of the permeable Ffowcs Williams Hawkings surface method in the solver. In addition to single-point flowfield statistics and far-field noise, spatio-temporal second- and fourth-order correlations are investigated. Two pairs of simulations were conducted, a coarse mesh containing 100 million elements and a fine mesh with 200 million elements. The coarse mesh included an azimuthal clustering of the cells in the near-microjet region. The non-uniformity of the azimuthal cell size was shown to adversely affect the development of the initial shear layer, yielding a delay in transition to a fully turbulent state and larger coherent structures in regions with larger cells. Radial velocity and turbulent kinetic energy profiles show good agreement with experimental results. A previously unidentified periodic interaction between the main jet and microjets was found. The dynamic interaction gives rise to velocity and pressure fluctuations in the near microjet region that match a tonal frequency found in the microjet far-field spectra that is absent from the clean jet case. Second- and fourth-order correlation distributions show large periodic regions of high correlation amplitude in the near microjet region. The evidence demonstrates that the main-microjet interaction is a clear high-frequency noise source. Despite the high-frequency noise associated with the main-microjet interaction, the addition of microjets yields a 1-2 dB reduction in overall sound pressure level. Additionally, over a significant portion of the length of the potential core the microjets reduce the amplitude of the majority of the six main correlation amplitudes that can be used in far-field noise prediction. Finally, the generation of the counter-rotating vortex pair downstream of the microjets was investigated. It is commonly presumed that this vortex pair is similar in origin to the counter-rotating vortex pair present in a jet in a crossflow. Vortex identification methods, velocity vectors and streamlines in the near microjet region demonstrate that the horseshoe-like vortex is the source of the counter rotating vortex pair that is present downstream of the microjets. The horseshoe-like vortex in the microjet case has the same sense as the vortices in the microjet shear layer and appears to be generated by the development of a recirculation region of microjet fluid during the main-microjet interaction.

Cette thèse propose différentes méthodologies de calcul du bruit à large bande générée par l’interaction rotor-stator d’une soufflante de turboréacteur et ce de la génération des sources acoustiques jusqu’au rayonnement en champ lointain. En premier lieu, le mécanisme d’interaction rotor-stator est étudié à travers une turbulence de grille homogène et isotrope (THI) impactant un profil isolé. Le cas de l’interaction rotor-stator en milieu guidé est ensuite traité et appliqué à une maquette de compresseur axial. Enfin, la dernière partie traite de la simulation du rayonnement aval en sortie de conduit. Dans chacune de ces parties, des approches analytiques et numériques sont proposées. Les méthodes analytiques permettentune prévision rapide du problème simplifié, et les méthodes numériques permettent de leverles hypothèses au prix d’un temps de calcul plus long.Le modèle d’Amiet est mis en œuvre pour simuler le bruit d’interaction THI-profil àtravers une expérience en soufflerie anéchoïque. Ce modèle simplifiant le profil en une plaque plane non portante, une approche numérique est également mise en place. La convection de perturbations de vitesse incidentes synthétisant une THI et impactant un profil isolé est simulée à l’aide d’un code CAA (Computational AeroAcoustics) résolvant les équations d’Euler en deux dimensions. Les différences de rayonnement entre une plaque plane et un profil cambré épais observées expérimentalement sont en partie retrouvées. Une simulation plus avancée d’un calcul LES (Large Eddy Simulation) 3D est également abordée.Dans la deuxième partie du mémoire, le modèle d’Amiet est étendu au problème de l’interaction rotor-stator en conduit. Différentes formulations sont proposées et discutées. Elles sont appliquées au cas d’un banc d’essai du DLR (centre de recherche aérospatiale allemand) avec des données d’entrée provenant d’un calcul RANS (Reynolds Averaged Navier-Stockes) ou bien directement de mesures. Les prévisions acoustiques ainsi obtenues sont satisfaisantes (+/-2 dB avec des données d’entrée expérimentales) compte tenu de la relative simplicité des modèles. Un calcul LES est exploité pour fournir directement les sources de bruit (fluctuations de pression pariétale). Les spectres obtenus montrent des tendances similaires à ceux issus du modèle d’Amiet malgré certains phénomènes non-physiques encore présents au niveau du calcul LES.Pour terminer, le rayonnement en champ libre est traité. Une manière simple et rapidede résoudre ce problème consiste à utiliser une intégrale de Kirchhoff, en supposant unécoulement moyen uniforme. Une comparaison avec une solution analytique (technique deWiener-Hopf) exacte pour les mêmes conditions montre que le rayonnement du bruit largebande est bien prévu par la méthode de Kirchhoff pour des angles de rayonnement inférieursà 90°. Cependant, pour prendre en compte la géométrie d’éjection de la nacelle et l’effet d’un écoulement hétérogène (cisaillement) sur le rayonnement en champ libre, une approche numérique est mise en place. Une technique permettant de simuler le caractère aléatoire des sources turbulentes ainsi que de satisfaire l’hypothèse de modes acoustiques incohérents est développée. Cette méthodologie est appliquée au cas simple du conduit semi-infini, puis à une tuyère réaliste. Les calculs sont validés à l’aide de solutions analytiques sur les configurations simplifiées. Une analyse critique des instabilités créées dans la couche de cisaillement et de leur influence sur les formulations intégrales couplées au calcul Euler pour obtenir le bruit en champ lointain complète ce dernier chapitre. / This report presents several methodologies aiming to predict broadband noise generatedby the interaction between the rotor and the stator of a turbofan. The main objective is toobtain a complete prediction procedure from the acoustic sources generation to the far fieldnoise radiation. As a first step, the rotor-stator mechanism is studied as the interaction betweena homogeneous and isotropic turbulence (HIT) and an isolated airfoil. Then, the ductedrotor/stator interaction case is treated. Finally, the last part is devoted to the aft radiationthrough the bypass duct. In each part, analytical and numerical approaches are investigated.On the one hand, analytical models allow a fast prediction but on a simplified problem, onthe other hand numerical methods remove many assumptions with a longer computationaltime.The Amiet theory is applied to simulate the airfoil-HIT interaction and comparisons withexperimental results obtained in an anechoic wind tunnel are discussed. Since this modelsimplifies the airfoil to a non-lifting flat plate, a numerical approach has also been investigatedusing a CAA (Computational AeroAcoustics) code solving the Euler equations in twodimensions. The incoming HIT is synthesized by a sum of convected harmonic velocity perturbationsinjected through the inlet boundary. The acoustic radiation discrepancies betweena flat plate and a lifting airfoil observed between experimental results and the Amiet theoryare partly verified. Finally, the possibility to use a 3D LES (Large Eddy Simulation) is alsodiscussed.In the second part, the Amiet model for an isolated airfoil is extended in order to solvethe turbofan interaction noise problem. Several formulations are proposed and are validatedby comparisons with measurements obtained from an axial compressor test bed. Input dataare provided either by a CFD (Computational Fluid Dynamics) or experiment. Acoustic resultsshown to be quite reasonable (�}2 dB in the later case) considering the simplicity of themodel. A LES computation is also studied in order to directly obtain the acoustic sources(vane surface pressure fluctuations). The resulting acoustic spectra shape are in relativelygood agreement with the Amiet model even if some non-physical phenomena are observed inthe LES data.The acoustic radiation in the free field is studied in the last part. A simple way to solvethe problem is to use a Kirchhoff integral along the outlet duct section assuming an uniformmean flow. A comparison with an exact(under uniform mean flow assumption) analytical modelinvolving a Wiener-Hopf technique demonstrates that good results can be obtained withthe Kirchhoff method up to 90°of radiation angle. However, in order to take in considerationthe effect of nacelle geometry and heterogeneous flow (shear) on the acoustic radiation, anumerical approach is set up. This method allows to simulate the non-deterministic and thenon-coherent mode behavior of a fan broadband noise. Simplified configurations are used topartly validate these numerical simulations by comparison with analytical solutions. This approachis finally applied to a realistic nozzle case. The instability waves created in the shearlayer and their effects on the integral formulations used to obtain the far field radiation arecarefully analyzed.

Les écoulements instationnaires au-dessus de surfaces discontinues produisent d'important bruit aérodynamique. L'objectif de ce travail de thèse est l'étude aéroacoustique d'écoulement au-dessus de cavités bidimensionnelles rectangulaires, et de trouver des stratégies de réduction du bruit. Des simulations numériques directes des équations bidimensionnelles de Navier-Stokes compressibles ont été réalisées afin d'étudier l'influence des conditions initiales sur le mode d'oscillation de l'écoulement pour des cavités profonde et peu profonde. Les résultats montrent que dans le cas de cavités profondes, l'écoulement oscille selon un régime de couche de cisaillement suivant le second mode de Rossiter, et ce quelle que soit la condition initiale choisie. En revanche, dans le cas de cavités peu profondes, le régime d'oscillation observé peut être en couche de cisaillement ou bien en mode de sillage suivant la condition initiale choisie. Une analyse de sensibilité d'écoulement dans le cas de cavités profondes a été réalisé en utilisant une méthode adjointe. Les équations adjointes ont été forcées par une perturbation localisée sinusoïdale soit de la quantité de mouvement suivant x adjointe (au voisinage de la couche de cisaillement), soit de la densité adjointe (loin de la cavité). Les résultats désignent une région de l'écoulement très sensible à l'ajout de masse, et localisée au voisinage du coin supérieur amont de la cavité. Par conséquent, un actionneur de type soufflage/aspiration placé au bord d'attaque de la cavité agira sur les fluctuations de quantité de mouvement suivant x au voisinage de la couche de cisaillement et sur les fluctuations de pression au loin. / The unsteady flow over surface discontinuities produces high aerodynamic noise. The aim of this thesis is to study the aeroacoustics of two-dimensional rectangular cavities and to find strategies for noise reduction. Direct Numerical Simulation of the compressible Navier-Stokes equations is performed to investigate the influence of the initial condition on the oscillation modes in deep and shallow cavities. Results show that the deep cavity oscillates in shear layer regime at the second Rossiter mode regardless of the initial condition. On the other hand different initial conditions lead to a shear layer or wake mode in the shallow cavity case. A sensitivity analysis of the deep cavity is done by the use of adjoint methods. Local sinusoidal perturbations of x-momentum and density are applied to the adjoint equations. The results show a high sensitivity region to mass injection at the upstream corner. Therefore an actuator placed at the leading edge will modify the velocity fluctuations reaching the trailing edge and hence the pressure fluctuations in the far-field.

Three models for the generation and propagation of hydrodynamic noise near the ocean surface are presented, and are compared for their ability to predict hull noise generated by CFAV Quest. The simulated fluctuating pressure field on the hull is also validated against experimental results. The near field flow is first solved using the NWT CFD package, and the hydrodynamic noise is calculated using the Lighthill-Curle acoustic analogy. The far field sound is obtained using three methods: a method of images solution to the Lighthill-Curle equations, a simple source model of the transmission loss, or a normal mode model of the transmission loss. Both the simple source and method of images models improve the SPL predictions of the Lighthil-Curle equations. Best performance is obtained from the method of images, improving predictions by approximately 40 dB. The normal mode model is shown to give poor results, due to assumed sea-floor boundary conditions.

A mixing layer is generally considered as a prototype of free shear flows whichoccur in a very broad spectrum of applications from natural phenomena to the engineeringscience. The increasing development of the aeroacoustics, more particularly thanks to theprogress in numerical simulations, has allowed to identify the coherent structures which evolvein the flow as the main sound sources. Furthermore, the emergence of the instability wavestaking the form of the coherent structures can be characterized by the stability analysis. In thiscontext, we propose through this work to perform a global stability analysis, in order tounderstand the emergence of fundamental frequencies, as well as a modal decompositionwithin both a linear and nonlinear framework, to characterize the coherent structures primarilyresponsible for the sound generation. In particular, we propose to develop the methods of theglobal stability with respect to the compressible flows as well as a technique of calculation ofcoherent structures applied to the nonlinear regimes based on the DMD method. Such analysistools are validated and illustrated on an academic configuration of a co-flowing mixing layer, atypical noise amplifier, from a simple case where a single source is implicated in the acousticradiation to a more complex case where two sources are present. Both the temporal and spatialDMD are shown capable of describing the characteristics of the coherent structures in the nearfield and the behaviour of the acoustic waves in the far field. Finally, the methods of the modaldecomposition have proven themselves as a relevant model reduction aiming at designing anefficient control strategy.

Une couche de mélange est généralement considérée comme un prototyped'écoulements libres cisaillés trouve sa présence au sein des nombreuses phénomènesnaturelles et applications industrielles. Le fort développement de l'aéroacoustique, notammentgrâce aux progrès des simulations numériques, a permis d'identifier les structures cohérentesqui se développent dans l'écoulement comme les principales sources de bruit. En outre,l'analyse de stabilité permet de caractériser l'émergence des ondes d'instabilités sous forme deces structures. Dans ce cadre, nous proposons de réaliser à travers ce mémoire une analysede stabilité globale, afin de comprendre l'émergence de fréquences fondamentales, ainsi qu'unedécomposition modale linéaire et non-linéaire, afin de caractériser les structures cohérentesessentiellement responsables du rayonnement acoustique. En particulier, nous proposons dedévelopper les méthodes de stabilité globale en régime compressible ainsi qu'une technique decalcul de structures cohérentes appliquée en régime non-linéaire basée sur la méthode« DMD ». De tels outils d'analyse sont validés et illustrés sur une configuration académiqued'une couche de mélange à co-courant, de type d'amplificateur de bruit, d'un simple cas où uneseule source est impliquée dans le rayonnement acoustique à un cas plus complexe où deuxsources sont présentes. Aussi bien la DMD temporelle que la DMD spatiale sont montréscapables de décrire le comportement des structures cohérentes en champ proche et lescaractéristiques des ondes acoustiques en champ lointain. Enfin, les méthodes dedécomposition modale s'avèrent une base de réduction de modèle pertinente, dans l'objectifd'un contrôle efficace. / A mixing layer is generally considered as a prototype of free shear flows whichoccur in a very broad spectrum of applications from natural phenomena to the engineeringscience. The increasing development of the aeroacoustics, more particularly thanks to theprogress in numerical simulations, has allowed to identify the coherent structures which evolvein the flow as the main sound sources. Furthermore, the emergence of the instability wavestaking the form of the coherent structures can be characterized by the stability analysis. In thiscontext, we propose through this work to perform a global stability analysis, in order tounderstand the emergence of fundamental frequencies, as well as a modal decompositionwithin both a linear and nonlinear framework, to characterize the coherent structures primarilyresponsible for the sound generation. In particular, we propose to develop the methods of theglobal stability with respect to the compressible flows as well as a technique of calculation ofcoherent structures applied to the nonlinear regimes based on the DMD method. Such analysistools are validated and illustrated on an academic configuration of a co-flowing mixing layer, atypical noise amplifier, from a simple case where a single source is implicated in the acousticradiation to a more complex case where two sources are present. Both the temporal and spatialDMD are shown capable of describing the characteristics of the coherent structures in the nearfield and the behaviour of the acoustic waves in the far field. Finally, the methods of the modaldecomposition have proven themselves as a relevant model reduction aiming at designing anefficient control strategy.

No description available.

The successful application of Computational Aeroacoustics (CAA) requires high accuracy numerical schemes with good dissipation and dispersion characteristics. Unstructured meshes have a greater geometrical flexibility than existing high order structured mesh methods. This work investigates the suitability of unstructured mesh techniques by computing a two-dimensionallinearised Euler problem with various discretisation schemes and different mesh types. The goal of the present work is the development of an unstructured numerical method with the high accuracy, low dissipation and low dispersion required to be an effective tool in the study of aeroacoustics. The suitability of the unstructured method is investigated using aeroacoustic test cases taken from CAA Benchmark Workshop proceedings. Comparisons are made with exact solutions and a high order structured method. The higher order structured method was based upon a standard central differencing spatial discretisation. For the unstructured method a vertex-based data structure is employed. A median-dual control volume is used for the finite volume approximation with the option of using a Green-Gauss gradient approximation technique or a Least Squares approximation. The temporal discretisation used for both the structured and unstructured numerical methods is an explicit Runge-Kutta method with local timestepping. For the unstructured method, the gradient approximation technique is used to compute gradients at each vertex, these are then used to reconstruct the fluxes at the control volume faces. The unstructured mesh types used to evaluate the numerical method include semi-structured and purely unstructured triangular meshes. The semi-structured meshes were created directly from the associated structured mesh. The purely unstructured meshes were created using a commercial paving algorithm. The Least Squares method has the potential to allow high order reconstruction. Results show that a Weighted Least gradient approximation gives better solutions than unweighted and Green-Gauss gradient computation. The solutions are of acceptable accuracy on these problems with the absolute error of the unstructured method approaching that of a high order structured solution on an equivalent mesh for specific aeroacoustic scenarios.

As improving the performance of turbofan designs becomes increasingly difficult, manufacturers are looking to new technologies for the next generation of jet engines. An 'open rotor' replaces the fan of the turbofan with a set of external rotors. This has the potential to offer a significant improvement in propulsive efficiency, but the design for low noise is a key challenge. Hence, high fidelity noise prediction methods are needed to accurately predict and compare the noise of different designs. This thesis focuses on one set of methods based on the Ffowcs Williams-Hawkings (\fwh) equation. This equation is considered to be the most realistic description of aeroacoustic noise generation, as it is a direct rearrangement of the Navier-Stokes equations. The \fwh\ equation is difficult to solve for realistic test cases such as an open rotor, and is susceptible to several types of error. This thesis categorises these errors as ``input'', ``neglection'' and ``discretisation'' errors. Discretisation errors arise from the need to integrate a discretised source field for the total noise, neglection errors result from needing to ignore part of the source field for practical reasons, and input errors relate to any errors caused by inaccurate input to the solver. The fundamental motivation of this thesis is to advance the understanding of neglection and discretisation errors and how they can be mitigated, in order to develop best practice solvers and methodologies for application to open rotors. Dimensional analysis is combined with analytical flow solutions to develop a process for isolating and quantifying discretisation errors. This process is used to study a wide range of solver methodologies and select a best practice solver methodology for open rotor noise prediction. This first-of-a-kind study produces a solver methodology that reduces discretisation errors by an order of magnitude compared to an industry standard solver. Previous research into neglection errors has shown that avoiding density perturbations in acoustic source terms can be beneficial. This thesis uses a generic aeroacoustic analogy to provide a new, physically intuitive method of incorporating a surface discontinuity that enables density perturbations to be avoided in a far more elegant manner than previous research. The above method improvements are investigated using a modern open rotor rig test case. The results demonstrate that discretisation and neglection errors can be severe in realistic cases and the potential of the method improvements to significantly mitigate them.