Highly loaded single stage gas turbines are being developed to minimize the turbine size and weight. Such highly loaded turbines often result in transonic flows, which imply a reduction in the efficiency due to the shock losses. The efficiency of a turbine is defined as the ratio between the real work extracted by the turbine rotor from the fluid and the maximum available enthalpy for a given pressure ratio. The relationship between turbine performance and design parameters is not yet fully comprehended due to the complexity of the flow field and unsteady flow field interactions. Hence, experimental and numerical studies remain necessary to understand the flow behavior at different conditions to advance the state of the art of the prediction tools. The purpose of the current research is to develop a methodology to determine the efficiency with an accuracy better than 1 % in a cooled and uncooled high pressure (HP) turbine tested in a short duration facility with a running time of about 0.4s. Such low level of uncertainty requires the accurate evaluation of a large number of quantities simultaneously, namely the mass flow of the mainstream, the coolant, and leakage flows properties, the inlet total pressure and total temperature, the stage exit total pressure, the shaft power, the mechanical losses and the heat transfer. The experimental work is carried out in a compression tube facility that allows testing the turbine at the temperature ratios, Re and Mach numbers encountered in real engines. The stage mass flow is controlled by a variable sonic throat located downstream of the stage exit. Due to the absence of any brake, the turbine power is converted into rotor acceleration. The accurate measurement of this acceleration as well as those of the inertia and the rotational speed provides the shaft power. The inertia of the whole rotating assembly was accurately determined by accelerating and decelerating the shaft with a known energy. The mass-flow is derived from the measured turbine inlet total pressure and the vane sonic throat. The turbine sonic throat was evaluated based on a zero-dimensional model of the turbine. The efficiencies of two transonic turbines are measured at design and off-design conditions. The turbine design efficiency is obtained as 91.8 %. The repeatability of the measurements for 95% confidence level varies between 0.3 % and 1.1 % of the efficiency depending on the test case. The theoretical uncertainty level of 1.2 % is mainly affected by the uncertainty of exit total pressure measurements. Additionally, the effect of vane trailing edge shock formations and their interactions with the rotor blade are analyzed based on the experimental data, the numerical tools and the loss correlations. The changes of blade and vane performances are measured at mid-span for three different pressure ratios which influence the vane and rotor shock mechanisms. Moreover, the unsteady forces on the rotor blades and the rotor disk were calculated by integration of the unsteady static pressure field on the rotor surface.

This dissertation thesis deals with the computational modelling of the dynamic response of the rotating electrical machine structure on the application of the magnetic forces. Apart from the dynamic response of the ideal symmetrical machine, the influence of the air gap eccentricity on the dynamics response is studied in this work. A basic type of the air gap eccentricity, which is caused by eccentric mounting of the rotor pack on the shaft of the rotor, is considered. The calculations the dependence of the magnetic forces on the time and a misalignment of the rotor pack are performed as first. The computational model of the magnetic field of the rotating electrical machine, which is based on solution of the electromagnetic coupled field analysis by finite element method, is used for this purpose. An analysis of the influence of the unbalanced magnetic pull and the stiffness of some parts of the machine on the modal properties of the machine is performed in the second part of this thesis. A third part of this thesis is focused on the calculation of the dynamic response of the machine during the steady state operation of the machine and the influence of the rotor pack misalignment on the dynamic response is studied. The obtained results showed that the tangential components of the magnetic forces, which act on the stator pack, excite significant torsional vibration of the stator. Besides the vibration of the stator of the machine, the influence of the rotor pack misalignment on the sound power of the machine, vibration of the rotor, loads of rotor bearings and air gap eccentricity is studied in this thesis.

The study of turbofan aeroacoustics has become important in academia and industry as noise from other aircraft sources, as jet noise, have been reduced. The EESC-USP Fan Rig is a long-duct low-speed fan experimental setup recently built at the Department of Aeronautical Engineering of the University of São Paulo. The objective is to provide a facility for studying fan aeroacoustics with a flexible configuration that allows changes in operational conditions and geometry of the rig so that each of the noise sources can be treated separately. In this work, three experimental campaign were taken aiming to extend the knowledge of this workbench capabilities and to observe the effect that some parameters can have on noise generation. A performance campaign was carried out to characterise the effects of controllable parameters on the aerodynamic characteristics of the flow. Starting from unrestricted flow to 70% area constraint, and a range of fan speeds, several flow measurements that could be translated into performance quantities were taken: volume flow rate, fan total and static pressures, compression ratio, total and net power, and efficiency. The measured compression ratio ranged from 1.00 to 1.02, with maximum axial Mach of 0.13 and maximum total efficiency of 65%. The surge effect, an unstable configuration characterised by rotor blades stall was also observed. Significant differences in fan performance due to the installation of an Inlet Control Device (ICD) and the rotor-stator spacing were not found, except under surge conditions. A parametric campaign was carried out exploring the effects of fan rotational speed, fan loading and rotor-stator spacing. Acoustic measurements were taken using an array of 77 wall-mounted microphones to provide a baseline data set for future comparisons. Hereby, data was processed to obtain the modal decomposition and power spectrum for each configuration. The last proved useful to compare tonal and broadband noise for each configuration. Experimental results indicate that changes in fan rotational speed scale noise generation mechanisms proportionally, do not affect noise spectral shape, and in consequence, are not useful to distinguish noise mechanisms. Although throttling does not seem to exhibit a clear effect on fan noise, it turns out that it is a good approach for cross-comparisons of other parameters\' effect on fan noise, because of its direct modification on the flow structure within the duct. Results also showed that increasing rotor-stator spacing reduces both blade passing frequencies tone levels and the acoustic power of the interaction modes, which are in agreement with results obtained by similar test facilities. In this work a instrumented stator vane was designed using recently available technologies, as 3D printing and the use of MEMS microphones, to measure this pressure fluctuations. Phase-averaging and the signal from a hall sensor were used to separate broadband content in time series. Broadband level distribution were analysed for two span lines of the instrumented stator vane and for a reference microphone located in the inlet antenna. Broadband levels increased with the increase in fan speed and its distribution over these span lines maintained its shape for different speeds. Cross-correlation of the microphones were calculated and showed decay of zero-delay cross-correlation with increase in distance between microphones over the vane. And, finally, the integral length scale, obtained by the integration of zero-delay cross-correlation curve, showed decay with increase in fan speed. The data generated by this work showed good agreements with what was expected from the literature and will help as input data to the semi-analytical and semi-empirical models that are being developed in parallel by the research group. / O estudo da aeroacústica de fan tornou-se importante na academia e na indústria à medida que o ruído de outras fontes nas aeronaves, como o ruído de jato, foram reduzidas. O EESCUSP Fan Rig é um túnel de vento aeroacústico voltado para fans de baixa velocidade, que foi construído recentemente no Departamento de Engenharia Aeronáutica da Universidade de São Paulo. Seu objetivo é prover uma bancada experimental para o estudo dos mecanismos de geração de ruído de fan com uma configuração flexível que permita mudanças nas condições operacionais e geométricas de forma a isolar tanto quanto possível cada uma das fontes. Neste trabalho foram realizadas três campanhas experimentais visando aprofundar os conhecimentos sobre as capacidades da bancada e observar os efeitos no ruído que algumas variações paramétricas podem gerar. Uma campanha de performance de fan foi realizada para quantificar os efeitos dos parâmetros controláveis sobre o as características aerodinâmicas do escoamento. Partindo de escoamento sem restrição até restrição de área de 70%, e para diversas velocidades de rotação do fan, diversas medidas do escoamento foram feitas para que se pudesse calcular as métricas de performance: vazão volumétrica, pressão total e estática do fan, taxa de compressão, potência útil, potência elétrica de entrada e eficiência. Os resultados apontaram para uma taxa de compressão entre 1.0 e 1.02, com Mach axial máximo de 0.13 e eficiência máxima de 65%. Os efeitos de \"surge\", uma configuração instável caracterizada pelo estol das pás do rotor também foi observado. Diferenças significativas na performance do fan devido à instalação de uma Inlet Control Device (ICD) e pelo espaçamento entre rotor e estator não foram encontradas, com exceção em condições de \"surge\". Uma campanha paramétrica para estudo de ruído foi realizada explorando os efeitos da velocidade de rotação e carregamento no fan, espaçamento entre rotor-estator e condição de entrada do escoamento no duto. Os resultados mostraram que o aumento do espaçamento do rotor-estator reduz os níveis tonais nas frequências de passagem das pás (BPFs) a uma taxa de aproximadamente 4dB quanto o espaçamento é duplicado. Os resultados experimentais indicam que as mudanças nos mecanismos de geração de ruído com a velocidade rotacional do fan escalam-se proporcionalmente, não afetam a forma espectral do ruído e, por consequência, não são úteis para distinguir os mecanismos de ruído. Embora a variação no carregamento não pareça exibir um efeito claro no ruído do fan, verifica-se que é uma boa abordagem para comparações cruzadas, em conjunto com a velocidade de rotação, do efeito de outros parâmetros no ruído do fan, já que, desta forma, é possível fazer uma modificação direta na estrutura do escoamento dentro do duto. Os modos Tyler-Sofrin, esperados pela contagem de pás e aletas usadas no conjunto rotor-estator, foram identificados e a potência acústica desses modos de interação acompanham a potências dos respectivos tons no espectro, levando à conclusão que a maior contribuição para o ruído tonal é, de fato, a interação rotor-estator. Também neste trabalho, também foi projetada uma aleta de estator instrumentada, usando tecnologias recentemente disponíveis, como a impressão 3D e o uso de microfones MEMS, para realizar medições de flutuação de pressão na superfície superior da aleta. Um método de processamento de sinal foi desenvolvido, usando a técnica de \"phase averaging\" combinada com o sinal de um sensor Hall para separar o conteúdo de banda larga em séries temporais. A distribuição do nível de ruido banda larga foi analisado para duas linhas na envergadura da aleta instrumentada e em um microfone de referência localizado na antena de microfones. Os níveis de banda larga aumentaram com o aumento da velocidade do fan e a sua distribuição ao longo destas linhas de alcance manteve a sua forma para diferentes velocidades. As correlações cruzadas dos microfones foram calculadas e mostraram o decaimento da correlação cruzada de zero-atraso com o aumento da distância entre os microfones sobre a aleta. Finalmente, a integral do comprimento de escala, obtida pela integração da curva de correlação cruzada de zero-atraso, apresentou decaimento com aumento na velocidade do fan. Os dados gerados por esse trabalho se mostraram de acordo com o esperado na literatura da área e servem de dados de entrada para modelos semi-empíricos e semi-analíticos que vem sendo desenvolvidos paralelamente pelo grupo de pesquisa.

Dans le cadre de la conception de turbo-machines aéronautiques, l'amélioration du rendement est assurée, entre autres, par la diminution du jeu fonctionnel entre les parties tournantes et statiques, telles que les roues aubagées et le carter qui les entoure. Cette stratégie a des conséquences sur le comportement vibratoire des turbo-machines en favorisant l'apparition de contacts structuraux entre les sommets d'aubes et le carter, par exemple. Lorsque plusieurs points de contact co-existent, des phénomènes d'interaction modale peuvent apparaître. La simulation de ces phénomènes, potentiellement dangereux, dans un cadre indutriel est au cœur de notre étude. Du fait de la dimension du problème à traiter, des méthodes de réduction modale sont utilisées : une méthode à interfaces fixes, la méthode de Craig-Bampton, et une méthode à interfaces libres, la méthode de Craig-Martinez. L'étude porte notamment sur la performance de ces méthodes de réduction lorsqu'elles sont associées à une non linéarité de type contact. Le phénomène d'interaction modale est tout d'abord étudié sur des modèles 2D simplifiés avec l'analyse de la sensibilité des régimes d'interaction à la taille des modèles réduits, puis sur les modèles 3D industriels pour lesquels la solution vraie n'est pas accessible. La gestion du contact dans le 3D fait appel aux B-splines bicubiques surfaciques afin d'assurer de meilleures propriétés à la surface de contact et de faciliter la résolution numérique. Les résultats obtenus à l'aide des deux méthodes de réduction modale sont comparés et permettent de déterminer la méthode la plus appropriée pour l'industrialisation du code présentée en annexe de ce mémoire.

Les propriétés de l'écoulement turbulent confiné dans une cavité délimitée par deux disques parallèles coaxiaux dont l'un est en rotation sont très sensibles à la géométrie périphérique du système. De récentes études ont montré que l'écoulement en bloc prévu par la théorie de Batchelor n'est pas systématiquement observé lorsque la cavité est isolée, c'est à dire non soumise à un flux forcé. L'objectif de cette thèse est de mieux comprendre ce phénomène et d'étendre le cadre de l'étude au cas d'un écoulement centripète forcé. La première partie du travail aborde l'aspect théorique du problème. Un premier bilan des échanges de fluide dans les couches limites est réalisé à partir des solutions stationnaires de l'écoulement sur un disque, développées par von Kármán et Bödewadt. Ces solutions sont couplées à de nouvelles hypothèses permettant de tenir compte des échanges de flux hors couches limites et du taux de pré-rotation du fluide. Nous obtenons ainsi des solutions analytiques originales pour les distributions radiales du coefficient d'entraînement du fluide dans le noyau central, de la pression statique pariétale et de la pression totale, que la cavité soit isolée ou soumise à un flux radial forcé. La seconde partie est consacrée aux moyens expérimentaux et numériques nécessaires à la validation de ces solutions théoriques. Elle inclut d'abord une description de l'installation expérimentale, des techniques de mesure, et du programme d'essais réalisé dans le cadre de cette thèse. Les simulations numériques sont réalisées à l'aide du code de calcul industriel FLUENT. Les détails concernant le choix du domaine de calcul, le maillage, les conditions aux limites et le modèle de turbulence retenus sont ensuite fournis. La validation des solutions théoriques fait l'objet des deux derniers chapitres, respectivement consacrés aux cas de la cavité isolée et soumise à un écoulement centripète forcé. L'accord obtenu entre l'expérience, les résultats numériques et la théorie est généralement très bon, ce qui nous conforte dans le choix de nos hypothèses. L'analyse des résultats permet en particulier de comprendre le mécanisme d'apparition de l'écoulement en bloc prévu par Batchelor. Toutefois, l'utilisation des modèles n'est pas universelle car elle nécessite l'ajustement de constantes, pour lesquelles nous tentons de dégager des relations empiriques les plus générales, dans la mesure du possible.

In this thesis, a numerical approach based on a potential flow method has been developed in order to simulate unsteady rotor/stator interaction, and to predict the unsteady performance of a propeller and its rudder. The method is first developed and tested in two-dimensions by using a boundary element method in which a front hydrofoil is moving downward, while a back hydrofoil is stationary. The wakes of the two hydrofoils are modeled by continuous dipole sheets and determined in time by applying a force free-condition on each wake surface. The wake/hydrofoil interaction is de-singularized by applying a numerical fence on the surface of the back hydrofoil. The viscous wake/hydrofoil interaction is considered by employing a viscous wake vorticity model on the rotor's wake surface. The present method is validated by comparison with analytical solutions, experimental data and by using the results from a commercial Reynolds Averaged Navier-Stokes (RANS) solver for the same set-up and conditions. The numerical approach is further extended to three-dimensions to predict the mutual interaction between a propeller and rudder. A fully unsteady wake alignment algorithm is implemented into a Vortex Lattice Method to simulate the unsteady propeller flow. The interaction between propeller and rudder is investigated in a fully unsteady manner, where a panel method is used to solve the flow around the rudder, and a vortex lattice method is used to solve the flow around the propeller. The interaction between a propeller and its rudder is considered in an iterative manner by solving the propeller and the rudder problems separately and by including the unsteady effects of one component on the other. The effect of the unsteady propeller-rudder interaction on the mean and on the unsteady propeller/rudder performance, including sheet cavitation on the rudder, is studied. / text

The present work is a result of collaboration between the LMFA (Laboratoire de Mécanique des Fluides et d'Acoustique, Ecole Centrale de Lyon - France), Snecma and the Cerfacs. It aims at studying the flow in the 3.5-stages high-speed axial compressor CREATE (Compresseur de Recherche pour l'Etude des effets Aérodynamique et TEchnologique - rotation speed: 11543 RPM, Rotor 1 tip speed: 313 m/s), designed and built by Snecma and investigated at LMFA on a 2-MW test rig. Steady measurements, as well as laser velocimetry, fast-response wall static and total pressure measurements have been used to experimentally investigate the flow. The analysis focuses on two main aspects: the study of the flow at stable operating points, with a special interest on the rotor-stator interactions, and the study of the instabilities arising in the machine at low mass flow rates.The description of the unsteady flow field at stable operating points is done through measurements of wall-static pressure, total pressure and velocity, but also total temperature, entropy and angle of the fluid. It is shown that the complexity and unsteadiness of the flow in a multistage compressor strongly increases in the rear part of the machine, because of the interactions between steady and rotating rows. Therefore, a modal analysis method developed at LMFA and based on the decomposition of Tyler and Sofrin is presented to analyze these interactions. It is first applied to the pressure measurements, in order to extract the contributions of each row. It shows that all the complex pressure interactions in CREATE can be reduced to three main types of interactions. The decomposition method is then applied to the entropy field extracted from URANS CFD calculations performed by the Cerfacs, in order to evaluate the impact of the interactions on the performance of the machine in term of production of losses.The last part of this work is devoted to the analysis of the instabilities arising in CREATE at low mass flows. It shows that rotating pressure waves appear at stable operating points, and increase in amplitude when going towards the surge line, until reaching a critical size provoking the onset a full span stall cell bringing the machine to surge within a few rotor revolutions. The study of these pressure waves, and the understanding of their true nature is achieved through the experimental results and the use of some analytical models. A precise description of the surge transient through wall-static pressure measurements above the rotors is also provided, as well as a description of a complete surge cycle. An anti-surge control system based on the detection of the amplitude of the pressure waves is finally proposed.

L'écoulement dans les turbomachines est tri-dimensionnel et instationnaire. Actuellement, les concepteurs de moteurs cherchent à réduire l'encombrement et le poids des machines. En conséquence, les interactions entre les roues, appelées interactions rotor-stator, sont renforcées. Parmi elles, l'effet potentiel remontant n'est désormais plus négligeable malgré sa rapide atténuation spatiale. Dans cette étude, cet effet potentiel remontant a été analysé sur une configuration spécialement conçue : une grille linéaire d'aubes de turbine, suivie de barreaux défilants en aval à une distance de 20% de corde axiale, simulant des aubes de rotor en aval. La grande échelle du banc d'essais facilite l'étude du comportement de la couche limite des aubes de la grille. Des mesures de pression et d'anémométrie laser à deux composantes, synchronisées avec le défilement des barreaux aval sont réalisées. Le nombre de Reynolds, basé sur la corde, est 1.6 × 105. Une grille de turbulence placée en amont de la grille d'Aube afin de pouvoir augmenter le taux de turbulence amont a été utilisée. Des résultats de mesures en absence de cette grille (faible taux de turbulence amont) sont également présentés et analysés. Une modélisation numérique, basée sur un calcul laminaire avec un préconditionnement basse vitesse pour la même configuration, a montré la déformation des lignes de courant de l'écoulement dans le canal inter-aubes, en fonction de la position du barreau aval. La distribution de pression autour de l'aube est également périodiquement modifiée. Les résultats stationnaires expérimentaux, en absence de tout barreau aval, ont révélé un décollement de la couche limite à l'extrados de l'aube à bas taux de turbulence amont(Tu−am = 1.2%) qui est supprimé à haut taux de turbulence amont (Tu−am = 4.2%) ; la couche limite commence alors sa transition par un mode by-pass. Aucun effet instationnaire dans la couche limite n'a été observé à l'intrados, quel que soit le taux de turbulence amont. L'étude instationnaire, avec le défilement des barreaux en aval, a permis de mettre en évidence un décollement périodique de la couche limite à l'extrados à bas taux de turbulence amont (Tu−am = 1.8%). Dans ce cas, la couche limite suit deux modes de transition au cours d'une période : une transition par décollement et une transition bypass. Au contraire, dans le cas à fort taux de turbulence amont (Tu−am = 4.0%), aucun décollement de la couche limite n'a été décelé. La couche limite est sujette à l'effet instationnaire à l'extrados. Elle est devenue turbulente au bord de fuite à tout instant par un mode by-pass. Cette étude a montré que l'effet potentiel issu d'un roue en aval est du même ordre de grandeur que les effets de sillage et doit être pris en compte dans l'analyse des phénomènes. Par des méthodes d'indexation de roues, le décollement de la couche limite pourrait être supprimé.

L’amélioration des performances des turboréacteurs actuels est un enjeu crucial dans un contexte de contraintes économiques et environnementales fortes. Au sein du turboréacteur, le canal inter-turbines, localisé à l’interface entre la turbine Haute Pression (HP) et le premier distributeur Basse Pression (BP), est le siège d’écoulements très complexes. Ainsi, les structures aérodynamiques issues de la turbine HP (sillages, tourbillons et ondes de choc) interagissent fortement entre elles et impactent l’écoulement du distributeur BP, engendrant ainsi des pertes de rendement de l’ensemble de la configuration. Ce travail de thèse s’attache à étudier les phénomènes d’interactions aérodynamiques entre une turbine HP et le premier distributeur BP et à analyser les mécanismes à l’origine des pertes aérodynamiques dans le distributeur BP. Une campagne expérimentale antérieure, réalisée sur un banc d’essai comprenant une turbine HP couplée à un distributeur BP, avait permis de recueillir des mesures de l’écoulement dans des plans situés dans le canal inter-turbines et à l’aval du distributeur BP. En lien avec ces résultats expérimentaux, les simulations numériques menées dans cette étude avec le logiciel elsA s’attachent à restituer précisément la nature tridimensionnelle, instationnaire et turbulente de l’écoulement au sein de cette même configuration. Ces travaux se développent alors en trois étapes principales. Dans un premier temps, une étude stationnaire avec traitement plan de mélange permet de comprendre et quantifier les aspects généraux de l’écoulement. Une évaluation de l’effet de la modélisation turbulente RANS (Reynolds-Averaged Navier-Stokes) et du schéma numérique spatial sur les structures aérodynamiques présentes dans la configuration est réalisée. Dans un deuxième temps, une modélisation turbulente avancée de type ZDES (Zonal Detached-Eddy Simulation) est employée pour la résolution de l’écoulement dans le distributeur BP. Les structures aérodynamiques instationnaires issues de la roue HP amont sont modélisées par une condition limite à l’entrée du domaine de calcul. L’approche ZDES est comparée à une approche Unsteady RANS (URANS) sur la même configuration. La formation et la dissipation des sillages et des tourbillons est significativement différente entre les deux modélisations, ce qui impacte de manière importante la génération des pertes aérodynamiques. Enfin, des simulations URANS de plusieurs configurations permettent de mieux comprendre les effets d’interaction entre les différentes rangées d’aubes. Ainsi, les approches instationnaires chorochroniques prenant en compte un seul rotor et un seul stator évaluent des effets instationnaires importants dans le canal inter-turbines. Ces approches conduisent à la mise en oeuvre d’un calcul sur une configuration multipassages-chorochronique prenant en compte les deux stators et le rotor afin de modéliser complètement les interactions déterministes existantes. Afin de quantifier celles-ci avec précision, une décomposition modale du champ instationnaire est mise en place. Les niveaux d’interactions liées aux différentes roues sont alors quantifiés et l’impact sur les pertes aérodynamiques est évalué. / Improving the performance of current aeronautical turbines is an important issue in a context of severe economical and environmental constraints. In a turbofan, the inter-turbine channel which is located between the High-Pressure (HP) turbine and the first Low Pressure (LP) vane is characterized by a complex flow. Therefore aerodynamic structures coming from the HP turbine (wakes, vortices and showkwaves) strongly interact between each other and affect the LP vane flow field. This generates efficiency losses of the overall configuration. This PhD thesis aims at studying the aerodynamic phenomena between a HP turbine and the first LP vane and at analyzing the mechanisms creating aerodynamic losses. A previous experimental campaign, which was carried out on a facility including a HP turbine coupled to a LP vane, enabled to gather flow field measurements in planes located in the inter-turbine channel and downstream of the LP vane. In comparison with these experimental data, the numerical simulations done with elsA software intend to reproduce accurately the 3D, unsteady and turbulent nature of the flow within this configuration. The work can be divided into three mains steps. As a first step, steady simulations with a sliding mesh treatment enable to understand the general aspects of the flow. An assessment of the effects of RANS (Reynolds-Averaged Navier-Stokes) turbulent predictions and of spatial numerical schemes on the aerodynamic structures present in the configuration is carried out. As a second step, the advanced turbulence approach ZDES (Zonal Detached-Eddy Simulation) is considered for the LP vane flow prediction. The unsteady aerodynamic structures coming from the upstream HP rotor are set as an inlet boundary condition of the computational domain. The ZDES approach is compared to a URANS (Unsteady RANS) approach on the same computational domain. The generation and dissipation of the wakes and vortices are significantly different on the two simulations, and thus impact the creation of aerodynamic losses. Finally, URANS simulations enable to better understand the interaction effects between the different blade rows. First, the unsteady phase-lagged approaches that take into account a single rotor and stator assess the important unsteady effects in the inter-turbine channel. They finally lead to the implementation of a multipassages phase-lagged computation that takes into account the two stators and the rotor in order to model all the existing determinist interactions. In order to quantify them accurately, a modal decomposition of the unsteady flow field is set up. The interaction levels linked to the different blade rows are therefore quantified and the impact of the aerodynamic losses is evaluated.

The thesis focuses on constructing dynamic model - synchronous generator with using magnetic equivalent circuit diagrams. The model construction is based on knowledge of geometric parameters, magnetic characteristic of used material and fyzical deployment of vindings placed in stator slots. Computed reults are compared with results optained from model created with finite element method.