A wing was designed with a moving surface high-lift device in the form of a rotating cylinder at the leading edge to improve low speed flight characteristics. This rotating cylinder accelerates the air flow over the top of the wing, effectively combining the concept of lift generated by an airfoil and lift generated by a rotating cylinder. This faster moving air over the top of the wing increases the pressure differential, thus increasing lift. The added momentum to the air flow results in delayed flow separation and a decrease in drag. For experimental testing, a wing was built using balsawood, basswood and MonoKote and was tested in a subsonic wind tunnel using two different tests stands. For validation and further testing, the high-lift device was modeled in Gambit and numerical simulations were performed using ANSYS Fluent. Experimental and numerical data show the high-lift device to be effective.

No description available.

<p>Many flow cases in fluid dynamics face undesirable flow separation due torising static pressure on wall boundaries. This occurs e.g. due to geometry as ina highly curved turbine inlet duct or e.g. on flow control surfaces such as wingtrailing edge flaps within a certain angle of attack range. Here, flow controldevices are often used in order to enhance the flow and delay or even totallyeliminate flow separation. Flow control can e.g. be achieved by using passiveor active vortex generators (VG) that enable momentum mixing in such flows.This thesis focusses on passive VGs, represented by VG vanes that are mountedupright on the surface in wall-bounded flows. They typically have an angle ofincidence to the mean flow and, by that, generate vortex structures that in turnallow for the desired momentum mixing in order to prevent flow separation.A statistical VG model approach, developed by KTH Stockholm and FOI,the Swedish Defence Research Agency, has been evaluated computationally.Such a statistical VG model approach removes the need to build fully resolvedthree-dimensional geometries of VGs in a computational fluid dynamics mesh.Usually, the generation of these fully resolved geometries is rather costly interms of preprocessing and computations. By applying this VG model, thecosts reduce to computations without VG effects included. Nevertheless, theVG model needs to be set up in order to define the modelled VG geometry inan easy and fast preprocessing step. The presented model has shown sensitivityfor parameter variations such as the modelled VG geometry and the VG modellocation in wall-bounded zero pressure gradient and adverse pressure gradientflows on a flat plate, in a diffuser, and on an airfoil with its high-lift systemextracted. It could be proven that the VG model qualitatively describes correcttrends and tendencies for these different applications.</p>

Simulation and modelling of turbulent flows under influence of streamline curvature and system rotation have been considered. Direct numerical simulations have been performed for fully developed rotating turbulent channel flow using a pseudo-spectral code. The rotation numbers considered are larger than unity. For the range of rotation numbers studied, an increase in rotation number has a damping effect on the turbulence. DNS-data obtained from previous simulations are used to perform a priori tests of different pressure-strain and dissipation rate models. Furthermore, the ideal behaviour of the coefficients of different model formulations is investigated. The main part of the modelling is focused on explicit algebraic Reynolds stress models (EARSMs). An EARSM based on a pressure strain rate model including terms that are tensorially nonlinear in the mean velocity gradients is proposed. The new model is tested for a number of flows including a high-lift aeronautics application. The linear extensions are demonstrated to have a significant effect on the predictions. Representation techniques for EARSMs based on incomplete sets of basis tensors are also considered. It is shown that a least-squares approach is favourable compared to the Galerkin method. The corresponding optimality aspects are considered and it is deduced that Galerkin based EARSMs are not optimal in a more strict sense. EARSMs derived with the least-squares method are, on the other hand, optimal in the sense that the error of the underlying implicit relation is minimized. It is further demonstrated that the predictions of the least-squares EARSMs are in significantly better agreement with the corresponding complete EARSMs when tested for fully developed rotating turbulent pipe flow. / QC 20100825

Over the last few decades the design of high lift devices has become a very important part of the total aircraft design process. Reviews of the design process are performed on a regular basis, with the intent to improve and optimize the design process. This thesis describes a new and innovative methodology for the design and evaluation of mechanisms for Trailing Edge High-Lift devices. The initial research reviewed existing High-Lift device design methodologies and current flap systems used on existing commercial transport aircraft. This revealed the need for a design methodology that could improve the design process of High-Lift devices, moving away from the conventional "trial and error" design approach, and cover a wider range of design attributes. This new methodology includes the use of the innovative design tool called SYNAMEC. This is a state-of-the-art engineering design tool for the synthesis and optimizations of aeronautical mechanisms. The new multidisciplinary design methodology also looks into issues not usually associated with the initial stages of the design process, such as Maintainability, Reliability, Weight and Cost. The availability of the SYNAMEC design tool and its ability to perform Synthesis and Optimization of mechanisms led to it being used as an important module in the development of the new design methodology. The SYNAMEC tool allows designers to assess more mechanisms in a given time than the traditional design methodologies. A validation of the new methodology was performed and showed that creditable results were achieved. A case study was performed on the ATRA - Advance Transport Regional Aircraft, a Cranfield University design project, to apply the design methodology and select from within a group of viable solutions the most suitable type of mechanism for the Variable Camber Wing concept initially defined for the aircraft. The results show that the most appropriate mechanism type for the ATRA Variable Camber Wing is the Link /Track Mechanism. It also demonstrated how a wide range of design attributes can now be considered at a much earlier stage of the design.

Over the last few decades the design of high lift devices has become a very important part of the total aircraft design process. Reviews of the design process are performed on a regular basis, with the intent to improve and optimize the design process. This thesis describes a new and innovative methodology for the design and evaluation of mechanisms for Trailing Edge High-Lift devices. The initial research reviewed existing High-Lift device design methodologies and current flap systems used on existing commercial transport aircraft. This revealed the need for a design methodology that could improve the design process of High-Lift devices, moving away from the conventional "trial and error" design approach, and cover a wider range of design attributes. This new methodology includes the use of the innovative design tool called SYNAMEC. This is a state-of-the-art engineering design tool for the synthesis and optimizations of aeronautical mechanisms. The new multidisciplinary design methodology also looks into issues not usually associated with the initial stages of the design process, such as Maintainability, Reliability, Weight and Cost. The availability of the SYNAMEC design tool and its ability to perform Synthesis and Optimization of mechanisms led to it being used as an important module in the development of the new design methodology. The SYNAMEC tool allows designers to assess more mechanisms in a given time than the traditional design methodologies. A validation of the new methodology was performed and showed that creditable results were achieved. A case study was performed on the ATRA - Advance Transport Regional Aircraft, a Cranfield University design project, to apply the design methodology and select from within a group of viable solutions the most suitable type of mechanism for the Variable Camber Wing concept initially defined for the aircraft. The results show that the most appropriate mechanism type for the ATRA Variable Camber Wing is the Link /Track Mechanism. It also demonstrated how a wide range of design attributes can now be considered at a much earlier stage of the design.

Depuis plusieurs décennies, le trafic aérien ne cesse de croître. Ainsi, près de 6 milliards de passagers transitent dans le monde par an. Les objectifs européens à l’horizon 2020 en terme d’émission sonore des aéronefs imposent une réduction de 10 dB par point de mesure par rapport aux aéronefs de l’an 2000. Dans ce contexte, le projet européen VALIANT (VALidation and Improvement of Airframe Noise prediction Tools) a pour but principal de tester, valider et améliorer les codes numériques et les modèles de prédiction du bruit de cellule (trains d’atterrissage + voilure) sur des géométries simplifiées afin de disposer de cas tests pour les recherches futures. L’objectif de la thèse, associé à la contribution de l’ECL dans ce projet, est de créer des bases de données expérimentales fiables sur des systèmes à deux éléments – bec/aile et aile/volet – et de modéliser analytiquement le bruit issu de tels systèmes. La thèse s’est concentrée sur un système aile/volet non porteur et parallèle dans un écoulement de soufflerie à veine ouverte, en configuration d’alignement ou de recouvrement partiel, menant à de possibles interactions aérodynamiques et/ou acoustiques. Les mesures ont été faites pour différentes vitesses d’écoulement (30 − 100 m~s), avec une attention particulière à 50 m~s (M0 ∼ 0, 15). Le taux de turbulence de l’écoulement incident est modifiable par l’ajout d’une grille de turbulence à maille large placée dans la section de sortie du convergent. Les résultats aérodynamiques (fil chaud, pression en paroi) ont révélé la présence d’une forte interaction lorsque la distance entre les deux corps est de l’ordre de grandeur de la couche limite turbulente au bord de fuite de l’aile. De plus, le couplage acoustique a lieu lorsque la longueur de recouvrement est positive ou nulle. Des mesures de localisation de sources menées par l’ONERA/DSNA ont permis de valider les mesures de champ lointain en confirmant l’absence de sources de bruit d’installation en dessous de 10 kHz. Par ailleurs, des comparaisons avec les simulations numériques donnent de bons accords. Du point de vue analytique, le problème mathématique de deux plaques planes en recouvrement partiel dans un écoulement uniforme a été posé et une réduction bidimensionnelle a été justifiée. Le problème n’ayant pas de solution exacte, plusieurs modèles issus de la littérature – théories de Howe et d’Amiet – ont été étudiés. Les plus pertinents ont été confrontés aux résultats expérimentaux, révélant les limites asymptotiques de ces modèles. Un modèle original est alors proposé pour la géométrie du problème posé, sans hypothèse restrictive. La démarche est basée sur une procédure de diffraction itérative permettant de prendre en compte la proximité des deux corps et utilisant la fonction de Green exacte du demi-plan en écoulement uniforme. Le modèle prédit des comportements qualitatifs angle/fréquence proches des résultats expérimentaux. La prise en compte de la statistique des rafales incidentes reste néanmoins à effectuer afin de procéder à des comparaisons quantitatives. Une campagne expérimentale complémentaire avec une marche descendante permet de mettre en évidence les écoulements de cavité arrière d’une aile, plus proche de la réalité. De même, des mesures sur une configuration bec/aile a été testée et la prise en compte de la déflexion du jet de la soufflerie pour la réfraction des ondes sonores par la couche de cisaillement a été proposée. / Air traffic still grows from decades, with yearly 6 billion passengers nowadays in the world. By 2020, the EC imposes aircraft noise reductions by 10 dB per measuring point with respect to the status in 2000. In this context, VALIANT (VALidation and Improvement of Airframe Noise prediction Tools) is an EC-supported project that aims at testing, validating and improving numerical codes and analytical/theoretical models for the prediction of airframe noise (landing gears + high-lift devices) in simplified configurations in order to generate test cases for research needs. The main objective of the thesis in connection with ECL contribution in the scope of VALIANT project is to generate reliable experimental databases for 2-element systems – slatwing and wing-flap – as well as to analytically model overlapping configurations. It is particulary focussed on the experimental and analytical studies of a non lifting wing-flap system in a parallel flow, in aligned and overlapping arrangements. This is aimed at discussing likely aerodynamic and/or acoustic interactions. The tests have been carried out in an open-jet anechoic wind-tunnel for each arrangement and for several flow speeds (30-100 m/s), with main interest on 50 m/s (M0 ∼ 0, 15). The turbulence rate could be changed by fixing a removable turbulence grid with a large mesh at the outlet cross-section of the duct. Aerodynamic results (hot-wire anemometry, wall-pressure) show a strong interaction when the wing-flap distance is about the wing trailingedge boundary layer thickness. Acoustic coupling can be pointed out in cases of overlap. Source localization tests performed by ONERA/DSNA have permitted to validate far-field acoustic tests since no installation effect source seems to strongly radiate below 10 kHz. A good agreement with numerical simulations has been shown for every test. In order to predict noise analytically for a 2-element system in case of overlap in a uniform flow, the mathematical statement has been defined and a 2D-reduction of the equation system can be justified. Since no exact solution exists, several models from the literature – Howe’s and Amiet’s theories – have been studied. Comparisons between overlapping half-planes and slotted trailing-edge models proposed by Howe and experimental results show obvious limitations in the predictions. Then, an original model is proposed involving two bodies in close overlap arrangement, with no assumption. It is based on an iterative scattering procedure to take into account the close vicinity of the two bodies, using the exact half-plane Green’s function in a uniform flow. Convergence is relatively quick and qualitative predictions in angle/frequency behaviour show a good agreement with experiments. However, the statistics of the vortical flows responsible for the sound must be implemented for better comparisons. Other experiments have been done with a backward-facing step – it represents wing trailing-edge cove – to make cavity flow mechanisms appear, such as what is observed in real HLD. Finally, a series of tests has been performed involving a slat-wing system and an angular correction due to refraction of sound waves in shear-layer for a deflected jet has been proposed.

Esta dissertação apresenta o desenvolvimento e otimização de uma metodologia baseada em código comercial para previsão de ruído em um eslate, além do estudo da dependência do ruído ao variar a camada limite do aerofólio. Com a contínua redução do ruído produzido por motores em aeronaves e as sucessivas restrições nos níveis de certificação, o ruído produzido por dispositivos aerodinâmicos tem ganhado importância no projeto de uma aeronave. Durante o pouso, o ruído gerado pelos dispositivos hiper-sustentadores é classificado dentre os mais relevantes, sendo o eslate um de seus componentes. Este trabalho busca criar e otimizar uma metodologia baseada no código PowerFLOW, assim como estudar a influência das camadas limite do aerofólio na geração de ruído. Tal código é baseado na formulação de Lattice- Boltzmann. As fontes acústicas simuladas são propagadas utilizando uma analogia acústica de Ffowcs-Williams e Hawkings e, então, analisadas utilizando métodos estatísticos de análise de sinais. Estudos de validação e verificação do código baseados em soluções analíticas são apresentados, tais como uma camada de mistura periódica no espaço e a solução dos vórtices de Taylor-Green. A seguir, o aerofólio 30P30N é utilizado em todo o estudo relacionado a eslates, analisando primeiramente a independência da solução em relação ao nível de refinamento da malha e do tamanho do domínio empregados. Baseado nas recomendações de tal estudo, o resultado é comparado com simulações disponíveis na literatura. Com uma maior confiança na metodologia, o trabalho então apresenta estudos variando a camada limite em regiões do eslate, assim como removendo-a completamente em certas regiões do aerofólio ao empregar condições de contorno de livre-escorregamento. O trabalho mostra que a influência das camadas limite do aerofólio é desprezível em relação ao erro do método. Isso é causado pela aparente independência do ruído do eslate em relação ao escoamento perto de sua cúspide. Tal independência permite que a malha computacional seja otimizada, reduzindo o custo da simulação em até 60%. / This dissertation presents the development and optimization of a methodology based on a commercial software to predict slat noise, also studying noise dependency when varying airfoil boundary layers. Due to continuous reduction on aircraft engine noise and successive restrictions on noise certification levels, airframe noise has been gaining importance on aircraft design. During landing, high-lift noise is ranked as one of the most relevant ones, being slat noise one of its components. This work focuses on creating and optimizing a noise prediction methodology based on the software PowerFLOW, and also on studying the influence of airfoil boundary layers on noise generation. Such software is based on Lattice-Boltzmann formulation. The simulated sound sources are propagated using Ffowcs-Williams and Hawkings acoustic analogy and then analyzed by signal analysis methods. Code validation and verification studies based on analytical solutions are presented, such as the spacially-periodic mixing layer and the Taylor-Green vortices solutions. Following, the 30P30N airfoil is employed through the rest of this work, firstly studying the solution independency related to mesh refinement level and computational domain size. Based on recommendations from this study, the results are compared to simulations from the literature. With higher confidence levels on this methodology, the work then presents studies varying the slat boundary layer and also removing it completely by employing free-slip boundary conditions on certain airfoil regions. This work presents that the airfoil boundary layer influence is neglectable when compared to the method error. This is caused by the apparent slat noise independency related to the flowfield near the slat cusp. Such independency allows the computational mesh to be optimized, reducing the simulation cost by up to 60%.

Com o desenvolvimento para aeronaves de propulsores turbo-fan com elevada razão de derivação, componentes da estrutura do avião passaram a ter relevância na geração de ruído aerodinâmico, principalmente durante a aproximação e o pouso. Dentre esses componentes, o eslate se destaca por ser uma fonte que se estende ao longo de praticamente toda a envergadura da asa. Neste trabalho, simulações numéricas foram feitas no intuito de considerar configurações práticas nas análises do ruído aeroacústico gerado pelo eslate. Um código comercial baseado no Método Lattice-Boltzmann foi usado no cálculo do escoamento transiente em torno do aerofólio MD30P30N. O domínio computacional simulado imitou a configuração geométrica de um túnel de vento. Foi levado em consideração o efeito da presença de duas formas de excrescência que são comuns na cova do eslate de aeronaves comerciais. Uma delas foi um selo que fica posicionado na parede da cova e a outra, um tubo do sistema anti-gelo. Tanto o escoamento transiente na região da cova quanto as características do ruído aeroacústico propagado para o campo distante foram analisados. Uma metodologia que impõe condição de parede com escorregamento livre no es- late e elemento principal, permitindo assim uma redução do custo computacional, foi usada. A abordagem foi ainda testada para a condição de um aerofólio submetido a escoamento cruzado, simulando uma asa infinita com enflechamento. Também uma modificação na metodologia, para que ela possa ser empregada em aerofólios com elevados ângulos de ataque, foi proposta e testada. O código híbrido MSES foi usado para o cálculo da espessura de deslocamento na camada limite do aerofólio. A modificação na geometria baseada em \'delta\'* causou uma melhora da solução aeroacústica de uma simulação empregando paredes com escorregamento livre, tomando como base de comparação a solução com paredes sem escorregamento. Simulações com selo dentro da cova, perto do recolamento, mostraram que, em certas circunstâncias, há um bloqueio dos vórtices da camada de mistura, intensificando picos tonais no espectro do ruído. A variação da posição do selo mostrou um efeito significativo no ruído do eslate, de forma que um selo suficientemente afastado do recolamento modificou o espectro do ruído do eslate. Os resultados com o aerofólio enflechado indicam que, também neste caso, o ruído do eslate não depende diretamente da camada limite na cúspide, mas da circulação do aerofólio. Por sua vez, a presença do tubo na cova aumenta significativamente a intensidade do ruído de banda larga produzido pelo eslate. Em uma asa sem enflechamento, o tubo causa também um aumento substancial na intensidade de picos tonais de baixa frequência. / The development of high by-pass ratio turbo-fan engine turned the airframe noise into an important component in a commercial airplanes\' noise characteristics. Between the airframe noise sources the slat can be highlighted as it extends almost along the whole wing span. Numerical simulation was carried out in order to consider practical configuration in the aeroacoutic noise generated by the slat. The effects of two different excrescences, which are normally present in commercial airplanes\' slat cove, were taken into account. One of them was a seal attached to the cove wall and the outher one was a tube that compose the anti-icing system. Both unsteady flow in cove region and far-field noise characteristics were analysed. A methodology that impose free-slip wall boundary condition on slat and main element surfaces was employed, which allowed the reduction of computational requirements. This approach was also tested for airfoil with crossflow, which simulates an infinite swept wing. Also a modification of the methodology was proposed and tested to extend its application in high-lift airfoils under higher angle of attack. A commercial code based on the Lattice-Boltzmann Method was used to compute the unsteady flow over the MD30P30N airfoil. The simulated computational domain imitates the geometry of a wind tunnel. The hybrid Euler/IBL code MSES was employed to calculate the displacement thickness of the airfoil\'s boundary layers. The geometry modification based on \'delta\'* caused a improvement on the aeroacoustic solution of a free-slip simulation, the no-slip simulation results being taken as reference. Simulations of geometries with relatively small seal close to the reattachment point showed that a blockage of the mixing layer vortices hapens and tonal peaks are intensified in the far-field noise spectrum. The variation of the seal position showed a significant effect on the slat noise, so that a seal farther from the reattachment modified affected both the shape and intensity of the noise spectrum. Results with the swept airfoil indicates that, even in the presence of crossflow, the slat noise does not depend on the cusp boundary layer, namely it is more sensitive to the airfoil circulation. The tube crossing the slat cove augmented significantly the broadband noise generated by the slat. In an unswept wing it also caused a substantial increase in the low-frequency tonal peaks.

A redução do ruído externo gerado por aeronaves operando nas proximidades de grandes centros urbanos é apontada como uma questão vital para a manutenção e expansão sustentável das atividades da aviação civil. Nas últimas décadas, reduções significativas no ruído gerado pelos sistemas de propulsão da aeronave tornaram relevantes as contribuições do trem de pouso e dos dispositivos de hiper-sustentação (flapes e eslates) para o ruído global da aeronave. A caracterização do espectro acústico de cada componente hiper-sustentador é necessária para o desenvolvimento de métodos preditivos de ruído e projetos aerodinâmicos que viabilizem a redução de ruído sem penalizações severas para o desempenho e a segurança da aeronave. Experimentos com modelos em escalas mostram que a contribuição de cada elemento hiper-sustentador para o ruído global é determinada pelo tamanho e modelo da aeronave. Tal fato dificulta a generalização dos resultados experimentais e determina a caracterização do espectro acústico de cada componente de um aerofólio em termos de sua geometria e configuração operacional. Este trabalho tem como objetivo principal a caracterização do ruído do eslate a partir de medições experimentais com um aerofólio hiper-sustentador McDonnell Douglas (30P30N), composto por flape, eslate e elemento principal. Os experimentos foram realizados em túnel de vento de seção fechada, e as medições acústicas contaram com o uso de uma antena composta por 62 microfones. Os dados acústicos foram processados com algoritmos de beam-forming convencional e deconvolução DAMAS (Deconvolution Approach for the Mapping of Acoustic Sources). A aplicação de técnicas de beam-forming permite representar uma distribuição espacial de fontes na forma de um mapa acústico e determinar o nível de ruído gerado por fontes que concorrem de forma independente para o ruído global. A base de dados experimentais permite o estudo do ruído do eslate sob diferentes configurações operacionais e geométricas do aerofólio. A análise do espectro acústico do eslate revela a ocorrência de ruído tonal em baixa e alta frequências, e ruído de banda larga em média frequência. Os mapas de beam-forming obtidos associam o ruído de banda larga com uma distribuição bidimensional de fontes ao longo da envergadura do eslate. O ruído do eslate aumenta com a velocidade de escoamento livre, enquanto que os picos tonais de baixa frequência e o ruído de banda larga decrescem com o aumento do ângulo de ataque do aerofólio de 2° para 10°. Os espectros de ruído do eslate colapsam quando reescalados pelo número de Mach do escoamento livre elevado a uma potência entre 4 e 5, e o ruído tonal colapsa em Strouhal dado pela corda do eslate e pela velocidade do escoamento base. Os resultados mostram que o ruído do eslate é fortemente dependente da geometria do aerofólio, particularmente para variações de overlap. Uma boa correspondência quantitativa foi obtida comparando-se espectros experimentais de ruído do eslate com espectros numéricos, obtidos a partir de um modelo com a mesma geometria e em condições de teste idênticas, o que indica a viabilidade do uso de túneis de vento de secção fechada para a realização de experimentos aeroacústicos. / The reduction in the noise produced by aircraft operating in the vicinity of large urban centers is an important issue for a sustainable growth in the civil aviation activities. Over the last decades, from a signicant reduction achieved in the noise generated by aircraft propulsion systems, the contribution of both landing gears and high-lift devices (flaps and slats) has become important to the aircraft overall noise. The identication of the noise signature of each high-lift component is required for the development of both noise prediction methods and new aerodynamic design concepts toward achieving a noise reduction without severe penalty over the aircraft performance and safety. Scaled model experiments have shown that the importance of each airframe component to the overall noise is determined by particularities in both aircraft geometry and size. Such noise model dependence hampers the generalization of experimental results from a reference testing model and leads to the necessity of assessing noise generation according to the testing model geometry and operational condition. This study focuses mainly on the characterization of slat noise from experimental measurements on a high-lift Mcdonnell Douglas (30P30N) airfoil, composed of a slat, a ap and a main element. Measurements were performed in a closed-section wind tunnel by a 62-microphone array and the acoustic data were processed with in-house codes based on conventional beam-forming and DAMAS (Deconvolution Approach for theMapping of Acoustic Sources) algorithms. Beam-forming techniques potentially enable the representation of a spatial source distribution as an acoustic map, from which the contribution of independent sources to the overall noise can be estimated. The experimental database enables the study of the slat noise from dierent airfoil operational conditions and geometrical settings. The slat noise spectral signature reveals the occurrence of tonal noise over both low- and high-frequency bands and also broadband noise over a mid-frequency range. Beam-forming maps indicate the slat broad-band noise originates from a source spatially distributed along the slat span. The slat noise increases in function of the ow speed, whereas low-frequency tonal peaks and the broadband noise decrease as the airfoil angle of attack increases from 2 to 10. The slat noise spectra scalle when the Mach number is raised to a power between 4 and 5, and the tonal noise collapses with Strouhal based on the slat chord and the ow speed. Results show the slat noise is strongly in uenced by the airfoil geometry, particularly for variations in the overlap. A good quantitative agreement was achieved through the comparison between the experimental and numerical slat noise spectra for the same model geometry and test conditions, which indicates the viability of performing aeroacoustic experiments in closed-section wind tunnels.