Wind turbines have grown both in size and number in the past decades. The taller turbines has made it possible to place them in forest areas which is fortunate for a country like Sweden with lots of forest. An issue with wind turbines is the sound they produce. The sound mainly comes from the rotor blades when they pass through the air. The sound heard some distance away from the turbine is sometimes masked by ambient background noise such as wind induced sound in the vegetation, but this is not always the case. Noise concerns among some people living in the vicinity of wind turbines are sometimes raised. Sound propagation models are used to predict the wind turbine sound level at certain distance. It is important that these models are accurate. Sound propagation is greatly influenced by the meteorological conditions. These conditions change over the day and year and vary a lot depending on the terrain conditions. In the past, large meteorological propagation effects have been found for sound sources close to the ground. Higher elevated sources like wind turbines have not been studied as much. One reason for this is that wind turbines are a relatively new sound source. In this thesis the meteorological influence on the wind turbine sound propagation is studied. Continuous simultaneous acoustic and meteorological measurements are performed at two different wind turbine sites during two years to capture all variations in the weather. The two sites are covered by forest, one is flat and the other has shifting terrain. The sites are representative for many locations in Sweden and around the world. The differences between the measured and expected wind turbine sound levels are established for different meteorological categories. The median of all deviations within each meteorological category is then compared. During no snow cover conditions the variation of the median under different meteorological conditions is 6 dBA and during snow cover the variation of the median is 14 dBA. The variations are due to the combined effect of refraction, ground conditions and terrain shape. The deviations from an expected value are seen for all octave bands from 63 Hz to 1000 Hz but are found to most distinct at low frequencies of around 125Hz. Meteorological effects starts to be important somewhere between 400 m and 1000 m from wind turbines.The characteristic "swish" sound from wind turbines are also studied in this thesis. The swish sound or as it is also called, the amplitude modulated sound, is found to be more common under some meteorological conditions such as temperature inversions and downwind conditions. A metric for detection of amplitude modulation duration and strength is proposed. Amplitude modulation, is according to some, the reason why wind turbine sound is perceived as more annoying than other typical environmental sounds at the same sound level. The swishes probably increase the probability to hear the wind turbine sound in presence of other background noise.

In recent years, the demand for wind energy has dramatically increased as well as the number and size of commercial wind turbines. These large turbines are loud and can cause annoyance to nearby communities. Therefore, the prediction of large wind turbine noise over long distances is critical. The wind turbine noise prediction is a very complex problem since it has to account for atmospheric conditions (wind and temperature), ground absorption, un-even terrain, turbine wake, and blade deformation. In these large turbines, the blade deflection is significant and it can potentially influence the noise emissions. However, the effects of blade flexibility on turbine noise predictions have not been addressed yet, i.e. all previous research efforts have assumed rigid blades. To address this shortcoming, the present work merges a wind turbine aeroelastic code, FAST (Fatigue, Aerodynamics, Structures, and Turbulence) to a wind turbine noise code, WTNoise, to compute turbine noise accounting for blade aeroelasticity. Using the newly developed simulation tool, the effects flexible blades on wind turbine noise are investigated, as well as the effects of turbine parameters, e.g. wind conditions, rotor size, tilt, yaw, and pre-cone angles. The acoustic results are shown as long term average overall sound power level distribution over the rotor, ground noise map over a large flat terrain, and noise spectrum at selected locations downwind. To this end, two large wind turbines are modeled. The first one is the NREL 5MW turbine that has a rotor diameter of 126 m. The second wind turbine, the Sandia 13.2MW, has a rotor diameter of 206 m. The results show that the wind condition has strong effects on the noise propagation over long distances, primarily in the upwind direction. In general, the turbine parameters have no significant effects on the average noise level. However, the turbine yaw impacts significantly the turbine noise footprint by affecting the noise propagation paths. The rotor size is also a dominating factor in the turbine noise level. Finally, the blade aeroelasticity has minor effects on the turbine noise. In summary, a comprehensive tool for wind turbine noise prediction including blade aeroelasticity was developed and it was used to address its impact on modern large turbine noise emissions. / Master of Science / Large wind turbines provide sustainable renewable energy but create loud noise that causes annoyance to nearby communities. Therefore, the prediction of large wind turbine noise is critical, but a complex problem, especially for the propagation over a long distance. The noise prediction needs to account for the turbine design, atmospheric factors, terrain, and airflow. Furthermore, in these large turbines, the blade deflection is significant and it can potentially influence the noise prediction. The present work addresses the above factors in the wind turbine noise prediction by merging a wind turbine structural code, FAST (Fatigue, Aerodynamics, Structures, and Turbulence) to a wind turbine noise code, WTNoise, to compute turbine noise accounting for blade deflection. Two turbines that have rotor diameter larger than 100 m were modeled and studied under different wind turbine design specifications, wind conditions, and blade deflection assumptions. The results showed that the rotor size is one of the dominating factor of turbine noise level. The blade deflection only has minor effects on the turbine noise. In summary, a comprehensive tool for wind turbine noise prediction including blade deflection was developed and used to address its impact on modern large turbine noise emissions.

No / The acoustical effects of hedges result from a combination of physical noise reduction and their influences on perception. This study investigates the physical noise reduction so as to enable estimation of its relative importance. Different in-situ methods have been used to measure noise shielding by hedges. These include a statistical pass-by experiment where the real insertion loss of a hedge could be measured, three controlled pass-by experiments using a reference microphone at close distance, and transmission loss measurements using a point source. Thick dense hedges are found to provide only a small total A-weighted light vehicle noise reduction at low speeds. Measured insertion losses range from 1.1 dBA to 3.6 dBA. The higher noise reductions are found to be associated with an increased ground effect.

Prediction of the sound field in large urban environments has been limited thus far by the heavy computational requirements of conventional numerical methods such as boundary element (BE), finite-difference time-domain (FDTD), or ray-tracing methods. Recently, a considerable amount of work has been devoted to developing energy-based methods for this application, and results have shown the potential to compete with conventional methods. However, these developments have been limited to two-dimensional (2-D) studies (along street axes), and no real description of the phenomena at issue has been exposed (e.g., diffraction effects on the predictions). The main objectives of the present work were (i) to evaluate the feasibility of an energy-based method, the diffusion model (DM), for sound-field predictions in large, 3-D complex urban environments, (ii) to propose a numerical hybrid method that could improve the accuracy and computational time of these predictions, and (iii) to verify the proposed hybrid method against conventional numerical methods. The proposed numerical hybrid method consists of a full-wave model coupled with an energy-based model. The full-wave model is used for predicting sound propagation (i) near the source, where constructive and destructive interactions between waves are substantial, and (ii) outside the cluttered environment, where free-field-like conditions apply. The energy-based model is used in regions where diffusion conditions are met. The hybrid approach, as implemented in this work, is a combination of FDTD and DM models. Results from this work show the role played by diffraction near buildings edges close to the source and near the exterior boundaries of the computational domain, and its impact on the predictions. A wrong modeling of the diffraction effects in the environment leads to significant under or overpredictions of the sound levels in some regions, as compared to conventional numerical methods (in these regions, some differences are as high as 10 dB). The implementation of the hybrid method, verified against a full FDTD model, shows a significant improvement of the predictions. The mean error thus obtained inside the cluttered region of the environment is 1.5 dB. / Master of Science

Il existe aujourd'hui un enjeu sociétal majeur à s'intéresser à la propagation du son en milieu extérieur etnotamment, dans notre contexte, à diminuer l'incertitude sur l'estimation des niveaux sonores et améliorer ainsi laprécision des diverses analyses, du bureau d'étude à l'institut de recherche. Dans le cadre de l'acoustiqueenvironnementale, l'influence des conditions météorologiques sur la propagation acoustique en milieu extérieurpeut être importante. Il est donc nécessaire d'appréhender et de quantifier les phénomènes météorologiques demicro-échelles que l'on observe dans la couche limite atmosphérique.Dans ce but, le modèle météorologique de recherche de Météo-France (Meso-NH) a été utilisé. Après avoircomparé les résultats de ce modèle à très fine résolution (de l'ordre du mètre) à l'aide des bases de données de deuxcampagnes expérimentales (Lannemezan 2005 et la Station de Long Terme), il s'est avéré nécessaire de développercet outil en intégrant la prise en compte de la force de traînée des arbres. Dès lors, les résultats issus de Meso-NH surles champs de vent, de température et d'énergie cinétique turbulente aparraissent satisfaisants. Ces informationssont par la suite utilisées en données d'entrée du modèle de propagation acoustique.Le modèle acoustique temporel utilisé est basé sur la méthode Transmission Line Matrix (TLM). Sondéveloppement a été effectué dans le but d'être appliqué à la propagation acoustique en milieu extérieur : prise encompte du relief, de différents types de sol, des conditions atmosphériques, etc. La validation numérique de laméthode TLM, par comparaison avec d'autres modèles (analytique et numérique de type Equation Parabolique), apermis de montrer la pertinence de son utilisation dans le cadre de l'acoustique environnementale.Enfin, à l'aide de ces modèles, des niveaux sonores simulés sous différentes conditions de propagation(favorables, défavorables, homogènes) ont été comparés aux mesures in-situ réalisées lors de la campagneexpérimentale de Lannemezan 2005. Les résultats se sont avérés très satisfaisants au regard de la variabilité desphénomènes observés. Cependant, l'utilisation des champs issus d'un modèle micro-météorologique de type Meso-NH reste délicate du fait de la forte sensibilité du niveau sonore aux profils verticaux de célérité du son. L'étude defaisabilité sur une expérience plus complexe (la Station de Long Terme) est encourageante et, à condition de disposerd'importants moyens de calculs, elle permet de considérer la TLM comme une nouvelle méthode de référence etainsi, d'envisager d'élargir son domaine d'utilisation à d'autres applications. / Actually, it exists a major societal issue to be interested in outdoor sound propagation and specially, in our context, toreduce the uncertainty in noise levels estimation and thus to improve the analyses accuracy, from engineers toresearch institutes. The influence of meteorology on outdoor sound propagation is significant. It is thereforenecessary to understand and quantify the micro scales phenomena into the atmospheric boundary layer.In this way, the French research meteorological model (Meso-NH) has been used. After comparing resultsof this model at very fine resolution (~1 meter) to measurements issued from the databases of two experimentalcampaigns (Lannemezan 2005 and Long Term Monitoring Station: LTMS), it appeared necessary to develop this toolin order to take into account the drag force of the trees. Finally, the Meso-NH results for the wind, temperature andturbulent kinetic energy fields are satisfactory. Then, theses informations can be used as input data for acousticmodels.Our time domain acoustic model is based on Transmission Line Matrix method (TLM). Its development wasdone in order to be applied to outdoor sound propagation: taking into account topography, soil types, meteorologicalconditions, etc. The numerical validation of the TLM method, by comparison with other models (analytical andnumerical: Parabolic Equation), has shown the relevance of its use in the context of environmental acoustics.Finally, thanks to these models, simulated noise levels in different propagation conditions (downward,upward and homogeneous refraction conditions) were compared to in situ measurements carried out during theLannemezan 2005 experimental campaign. Satisfying results were obtained regarding the observed phenomenavariability. However, using the micro-meteorological model Meso-NH is difficult because of the strong acousticsensitivity to the vertical celerity profiles. A feasibility study on a more complex experience (LTMS) is encouragingand, provided having substantial computing resources, it permits to consider the TLM as an accurate method in thecontext of environmental acoustics.

Les méthodes d'ingénierie acoustique (e.g. ISO 9613-2 ou CNOSSOS-EU) approchent efficacement les niveaux de bruit générés par les routes, les voies ferrées et les sources industrielles en milieu urbain. Cependant, ces approches d'ingénierie sont limitées à des géométries de forme simple, le plus souvent de section rectangulaire. Ce mémoire développe donc, et valide, une approche hybride permettant l'extension des méthodes d'ingénierie à des formes plus complexes, en introduisant un terme d’atténuation supplémentaire qui représente l'effet d'un objet réel comparé à un objet simple.Le calcul de cette atténuation supplémentaire nécessite des calculs de référence, permettant de quantifier la différence entre objets simple et complexe. Dans la mesure où il est trop onéreux, numériquement, '’effectuer ce calcul pour tous les chemins de propagation, l'atténuation supplémentaire est obtenue par interpolation de données stockées dans un tableau et évaluées pour un large jeu de positions de sources, de récepteurs et de fréquences. Dans notre approche, le calcul de référence utilise la méthode BEM en 2.5D, et permet ainsi de produire les niveaux de référence pour les géométries simple et complexe, tout en tabulant leur écart. Sur le principe, d'autres approches de référence pourraient être utilisées.Ce travail valide cette approche hybride pour un écran en forme de T avec un sol rigide, un sol absorbant et un cas avec bâtiments. Ces trois cas démontrent que l'approche hybride est plus précise que l'approche d’ingénierie standard dans des cas complexes. / Noise engineering methods (e.g. ISO 9613-2 or CNOSSOS-EU) efficiently approximate sound levels from roads, railways, and industrial sources in cities. However, engineering methods are limited to only simple box-shaped geometries. This dissertation develops and validates a hybrid method to extend the engineering methods to more complicated geometries by introducing an extra attenuation term that represents the influence of a real object compared to a simplified object.Calculating the extra attenuation term requires reference calculations to quantify the difference between the complex and simplified objects. Since performing a reference computation for each path is too computationally expensive, the extra attenuation term is linearly interpolated from a data table containing the corrections for many source and receiver positions and frequencies. The 2.5D boundary element method produces the levels for the real complex geometry and a simplified geometry, and subtracting these levels yields the corrections in the table.This dissertation validates this hybrid method for a T-barrier with hard ground, soft ground, and buildings. All three cases demonstrate that the hybrid method is more accurate than standard engineering methods for complex cases.

Dans le contexte de l’amélioration des modèles de prévision en acoustique extérieure, ces travaux de thèse se focalisent sur la modélisation des effets des irrégularités de la surface du sol sur la propagation acoustique. Pour ce faire, des méthodes numériques temporelles sont utilisées : d’une part, la méthode FDTD basée sur la résolution des équations de l’acoustique par des schémas aux différences finies, et d’autre part, la méthode des lignes de transmission (TLM). La modélisation des effets de la rugosité de surface est abordée en considérant le formalisme de l’impédance effective. Deux modèles d’impédance effective sont étudiés : le premier caractérise les effets d’une rugosité déterministe constituée de diffuseurs de géométrie constante, le second caractérise les effets moyens d’une rugosité aléatoire définie par un spectre de rugosité. Ce second modèle est validé expérimentalement par une campagne de mesures en salle semi-anéchoïque, audessus de surfaces rugueuses dont la rugosité a été définie très précisément. Les deux modèles d’impédance effective sont également validés par des simulations numériques FDTD et TLM. La possibilité d’implémenter ces conditions d’impédance effective dans les deux codes temporels est ainsi montrée, ce qui permet de modéliser les effets de la rugosité sans avoir à réaliser un maillage précis du profil des hauteurs de la surface du sol. Une campagne de mesures de l’impédance de différents terrains est réalisée afin d’étudier les effets de la variabilité spatiale et saisonnière de l’impédance sur la prévision des niveaux sonores. Les impédances mesurées lors de cette campagne sont également utilisées comme des données d’entrée réalistes pour le code TLM, afin de simuler et d’étudier les effets de la propagation acoustique au-dessus d’un sol hétérogène présentant une impédance spatialement variable. / In the context of prediction models improvement for outdoor sound propagation, this work focuses on the modelling of the effects of ground irregularities on sound propagation. Time-domain numerical methods are used: on one hand, the solving of the governing equations by finite difference schemes (FDTD method), and on the other hand, the transmission line matrix (TLM) method. Effective impedance is considered to model the effects of surface roughness. Two effective impedance models are studied: the first one takes into account the effects of a deterministic roughness formed by scatterers of constant geometry, the second one takes into account the mean effects of a random roughness defined by a roughness spectrum. This second model is validated experimentally by a measurements campaign carried out in a semi-anechoic chamber, above rough surfaces whose roughness profiles were precisely designed. The two effective impedance models are also validated by FDTD and TLM simulations. The possibility to use the effective impedances directly into the numerical methods is then shown, allowing the modelling of roughness effects without meshing the exact height profile of the ground surface. A measurements campaign of the impedance of different grounds is performed in order to assess the effects of space and seasonal variability of the impedance on the sound levels predictions. The results of this campaign are also used as realistic entry data for the TLM code, and the propagation above a heterogeneous ground showing spatially variable impedance is simulated.

Dans cette recherche, un modèle de propagation d’ondes de choc sur grandes distances sur un environnement urbain est construit et validé. L’approche consiste à utiliser l’Equation Parabolique Nonlinéaire (NPE) comme base. Ce modèle est ensuite étendu afin de prendre en compte d’autres effets relatifs à la propagation du son en milieu extérieur (surfaces non planes, couches poreuses, etc.). La NPE est résolue en utilisant la méthode des différences finies et donne des résultats en accord avec d’autres méthodes numériques. Ce modèle déterministe est ensuite utilisé comme base pour la construction d’un modèle stochastique de propagation sur environnements urbains. La Théorie de l’Information et le Principe du Maximum d’Entropie permettent la construction d’un modèle probabiliste d’incertitudes intégrant la variabilité du système dans la NPE. Des résultats de référence sont obtenus grâce à une méthode exacte et permettent ainsi de valider les développements théoriques et l’approche utilisée / This research aims at developing and validating a numerical model for the study of blast wave propagation over large distances and over urban environments. The approach consists in using the Nonlinear Parabolic Equation (NPE) model as a basis. The model is then extended to handle various features of sound propagation outdoors (non-flat ground topographies, porous ground layers, etc.). The NPE is solved using the finite-difference method and is proved to be in good agreement with other numerical methods. This deterministic model is then used as a basis for the construction of a stochastic model for sound propagation over urban environments. Information Theory and the Maximum Entropy Principle enable the construction of a probabilistic model of uncertainties, which takes into account the variability of the urban environment within the NPE model. Reference results are obtained with an exact numerical method and allow us to validate the theoretical developments and the approach used