Nuclear fusion is a research topic that attracts a lot of interest. If properly harnessed, it promises to be an energy source that circumvents problems that current energy sources have. As such, fusion warrants research aimed at understanding and dealing with its varied issues. Fusion is regularly recreated on earth by heating a hydrogen plasma to around a hundred million degrees Celsius. Confining this plasma requires special machines due to the extreme heat. There are multiple types of machines recreating plasma for research purposes, the most common is called tokamak. Tokamaks confine the plasma in a toroidal shape using powerful magnetic fields that prevent particles from escaping. Relevant for this work is the tokamak JET, where the treated experiment has been conducted, and ITER, which is currently under construction and meant to be the next step in fusion research. An important phenomenon are so called Edge Localized Modes (ELMs). ELMs are short bursts of energy expulsions from the plasma that results in a loss of energy and can cause damage to components facing the plasma. While not necessarily present in all operational modes ELMs are present in JET and will be present in ITER. Therefore it is very important to understand ELMs and how they are affected by certain parameters. Especially important is the dependency of ELM size on collisionality, a measurement on how much particles in the plasma interact with each other. Due to how ITER is supposed to operate it will have a very low collisionality, something that previous studies have linked with large ELM energy losses. This work investigates how parameters, plasma density, gas fueling rate, effective mass, strength of the magnetic field and collisionality affect ELMs.  This work calculates the energy losses for ELMs and investigates whether they are related to certain parameters. To calculate the energy loss two methods are deployed. One method relies on measurements of the diamagnetic flux. The other utilizes measurements of temperature and density with thomson scattering, electron cyclotron emission and reflectometry. Both techniques compare the energy in the plasma before and after an ELM to deduce the energy loss. For both methods, ELMs in a time interval are grouped and their data is used to calculate a typical energy loss. The results show that the energy losses from both methods are comparable with previous measurements at similar collisionality. The methods produce comparable results although the results for singular cases are not always in agreement. Ion cyclotron resonance heating is identified as worsening the agreement. A combination of the results being too noisy and there not being enough data means that no clear trends were observed in the investigated parameters. / Fusion är ett intressant forskningsområde. Om det nyttjas på rätt sätt, kan fusion bli en energikälla kringgår problem som nuvarande energikällor har. Därför finns det mycket forskning om fusion med målet att förstå och hantera de problem som idag stoppar fusion från att användas som energikälla. Fusion återskapas på jorden genom att värma väteplasma till cirka hundra miljoner grader Celsius. Att hålla plasmat kräver speciella maskiner på grund av den extrema värmen som lätt smälter alla material. Det finns flera olika maskiner som kan upprätthålla plasma för forskningsändamål, även om de ännu inte kan utvinna energi. Den vanligaste kallas tokamak. Tokamaker håller plasmat i en toroidal form med hjälp av kraftiga magnetfält. För detta arbete är tokamakerna JET och ITER relevanta. Datan som behandlas i detta arbete kommer från JET. ITER är en forskningsreaktor som är under konstruktion och är ämnad att vara nästa steg inom fusionforskning. Ett viktigt phenomen är Edge Localized Modes (ELMs). ELMs är korta energipulser från plasmat som kan orsaka skador på komponenter vända inåt mot plasmat. ELMs är inte nödvändigtvis närvarande men de är närvarande i JET och kommer att vara närvarande i ITER. Därför är det viktigt att förstå dem. Särskilt viktigt är hur ELM-storleken ändras beroende på kollisionalitet, ett mått på hur mycket partiklar i plasmat interagerar med varandra. På grund av hur ITER ska köras kommer kollisionaliteten vara mycket låg, något som tidigare studier har kopplat till stora ELMs. Beroende på vad som utlöser en ELM säger man att de är peeling eller ballooning begränsade. De flesta experimenten idag är ballooning-limited, vilket betyder att ELMs utlöses på grund av en för hög tryckgradient. På grund av den låga kollisionaliteten tros ITER bli peeling-limited, vilket betyder att ELMs utlöses av för höga strömmar i plasmat. I ett försök att härma ITERs operationstillstånd har experimentet som undersöks i detta arbete låg kollisionalitet. De parametrar vars inflytande på ELMs undersöks är plasmats densitet, bränsletillförsel, effektiv massa, styrkan av magnetfältet och kollisionalitet. För att beräkna energiförlusten används två metoder, en använder en mätningar av magnetflödet i plasmat. Den andra metoden använder mätningar av temperaturen och densiteten vid punkter i plasmat. Båda teknikerna jämför energin i plasmat före och efter en ELM för att fastställa energiförlusten. För båda metoderna används alla ELMs under en period för att beräkna en karaktäristisk energiförlust. Energiförlusterna är jämförbara med tidigare mätningar vid liknande kollisionalitet. De använda metoderna ger överlag liknande resultat för de olika undersökta intervallen. Ion Cyclotron Resonance Heating (ICRH) identifieras förvärra överensstämmelsen avsevärt. En kombination av att resultaten har hög osäkerhet och att det finns få datapunkter innebär att tydliga trender inte observerades i de undersökta parametrarna.

La résolution d'un certain nombre de problèmes physiques est nécessaire au développement de réacteurs électrogènes utilisant la fusion thermonucléaire contrôlée. Le travail présenté ici traite du contrôle du rayonnement dans le cadre général de la tenue des matériaux aux flux de chaleur, ainsi que de l'opération d'un tokamak à forte densité. Ces deux points concernent plus particulièrement le futur réacteur de démonstration, appelé DEMO, pas intermédiaire entre ITER et un réacteur commercial. L'opération de DEMO sera contrainte par la nécessité de rayonner dans un large volume afin de limiter le dépôt de puissance localisé sur les plaques du divertor. Une grande fraction de rayonnement (80 à 90% de la puissance extraite) devra être obtenue tout en conservant un excellent confinement et une pollution réduite au cúur du plasma. Les études actuelles montrent que cette fraction de rayonnement est atteignable tout en limitant la contamination du plasma, mais l'amélioration des modèles de rayonnement est indispensable, tout comme les études concernant la compatibilité entre un bord fortement rayonnant et l'existence d'une barrière de transport permettant l'accès à un régime de confinement amélioré, le mode H. En parallèle, une forte densité (fraction de Greenwald supérieure à l'unité) est aussi indispensable pour atteindre la fraction de rayonnement désirée. De plus, la puissance fusion, et donc le bilan économique d'un réacteur est directement liée à la densité dans la zone de réaction, au centre du plasma. / The route presently envisaged towards the development of a commercial fusion power plant includes that a few remaining physics issues are solved. The present work addresses two of them: plasma radiation control, as a part of the more general power handling issue, and high density tokamak operation. These two issues will be most critical in the demonstration reactor, called DEMO, intermediate step between ITER and a future commercial reactor. For DEMO, the need to radiate a large fraction of the power so as to limit the peak power load on the divertor will be a key constraint. High confinement will have to be combined with high radiated power fraction, and the required level of plasma purity. A fractional radiated power, including bremsstrahlung radiation, of 80-90% of the total power loss will be required. Present studies suggest that this level of radiation could be achieved with acceptable levels of plasma contamination, but improvements are required in models of plasma radiation, and compatibility with the edge transport barrier of the H-mode has to be further assessed. Correlatively, high plasma density (typically with a Greenwald fraction above unity) is required, both because it allows efficient radiation of exhaust power to the reactor walls, and because the final cost of electricity is directly influenced by the achieved Greenwald value.

Le développement de la source d'ions négatifs pour l’injecteur de particules d’ITER constitue une des étapes essentielles pour générer des neutres de haute énergie . Pour remplir les caractéristiques requises pour ITER en termes de chauffage et de courant à l'intérieur du réacteur principal, la source d'ions négatifs doit délivrer 40A de D-. La création d'une telle source représente un défi tant technique que scientifique et demande une meilleure compréhension des phénomènes physiques impliquées . Les connaissances actuelles sur le méchanisme d'extraction d'ion négatifs d’un plasma électronégatif sont limitées, spécialement concernant la compréhension des caractéristiques d'une gaine de plasma magnétisé dans la région d’intérêt où on constante également l’extraction des électrons simultanément avec les ions négatifs qui. De plus, l'asymétrie due à la configuration croisée du champ magnétique pour piéger les électrons nécessite une étude du problème en trois dimensions. Un code 3D Particle-In-Cell électrostatique a été spécialement développé pour étudier ce problème. Le code utilise les coordonnées cartésiennes et peut prendre en compte des géométries complexes. Le code nommé ONIX étudie les propriétés du plasma et le transport des électrons et des ions négatifs au niveau de la zone d'extraction. Les résultats sur la formation d'un ménisque de plasma et l'écrantage du champ d'extraction par ce plasma, ainsi que les trajectoires des ions négatifs, sont présentés. L'efficacité de l'extraction d'ions négatifs du volume et de la surface est investiguée et on trouve que les processus de création en surface des ions négatifs jouent un rôle capital. / The development of the negative ion source constitutes a crucial step in the construction of the neutral beam injector of ITER. To fulfil the ITER requirements in terms of heating and current drive, the negative ion source should deliver 40 A of D-. The achievement of such a source is challenging from technical and scientific points, and it requires a deeper understanding of the underlying physics. The present knowledge of the ion extraction mechanism from the negative ion source is limited due to the complexity of the problem that involves the comprehension of the behaviour of magnetized plasma sheaths when negative ions and electrons are pulled out from the plasma. Moreover, due to the asymmetry induced by the crossed magnetic configuration used to filter the electrons, any realistic study of this problem must consider the three spatial dimensions. To address this problem in a realistic way, a 3D Particles-in-Cell electrostatic code specifically designed for this system was developed. The code uses Cartesian coordinate system and it can deal with complex boundary geometry as it is the case of the extraction apertures. The complex magnetic field that is applied to deflect electrons is also taken into account. This code, called ONIX, was used to investigate the plasma properties and the transport of negative ions and electrons close to a source extraction aperture. Results on the formation of the plasma meniscus and the screening of the extraction field by the plasma are presented here, as well as negative ions trajectories. Negative ion extraction efficiency from volume and surfaces was investigated showing the capital importance of the surface negative ion production.

Barrières de transport (TB) sont un élément clé dans le contrôle de transport turbulent et atteindre la haute performance des ‘plasmas brûlants’. Les études théoriques abordent l’autorégulation de la turbulence comme une explication possible pour la formation de TB, mais une compréhension complète de ces dynamiques complexes est toujours manquante. Dans ce contexte, nous abordons l’auto-organisation dans le transport turbulent dans les plasmas de fusion dans le but de présenter une nouvelle compréhension de la dynamique des TB. Les outils numériques que nous utilisons des simulations de portée de la turbulence gyrocinétique plus complexe à simple turbulence des fluides 2D et prédateur-proie comme modèles.Deux principales caractéristiques de l'auto-organisation, les avalanches et les flux zonal (ZF), semblent contrôler transport à grande échelle. Dans la région de SOL (Scrape Off layer), événements avalancheux intermittents ne permettent pas séparation d'échelle dans le temps ou l'espace entre champs moyens et les modalités de fluctuation. Dans le bord (edge), la génération des doubles couches de cisaillement dans les profils de vitesse réduit le transport turbulent. Un modèle turbulent 2D pour la génération de ‘piedestal’, qui est non spécifique des plasmas de tokamak, a été mis au point, le piedestal étant localisée à l'interface entre les régions a différent amortissement d'écoulement zonal: edge et SOL. Les événements de relaxation quasi-périodiques sont étudiés réduisent le modèle à trois couplage des modes pour identifier l'interaction entre les streamers et les ZF et le rôle du Reynolds stress dans la génération et la saturation du TBs / Transport barriers (TB) are a key element in controlling turbulent transport and achieving high performance burning plasmas. Theoretical studies are addressing the turbulence self-regulation as a possible explanation for transport barrier formation but a complete understanding of such complex dynamics is still missing. In this context, we address self-organized turbulent transport in fusion plasmas with the aim of presenting a novel understanding of transport barriers dynamics. The numerical tools we use span simulations from the most complex gyrokinetic turbulence to simpler 2D fluid turbulence and predator-prey like models.Two features of self-organizations, avalanches and zonal flows (ZFs), appear to control large scale transport. In the SOL (Scrape Off Layer) , intermittent avalanche events do not allow for time or space scale separation between mean fields and fluctuation terms. In the edge, the generation of long living double shear layers in the profiles of the velocity reduces radial turbulent transport. Such radially distributed barriers govern profile corrugations. A 2D turbulent model for pedestal generation, which is not specific of Tokamak plasmas, has been developed, the pedestal being localized at the interface between regions with different zonal flow damping: the edge region, where zonal flows are weakly damped by collisions, and the SOL region characterized by zonal flow damping due to boundary conditions. Quasi-periodic relaxation events are studied reducing the model to three modes coupling to identify the interplay between streamers and ZFs and the role of Reynolds stress in the generation and saturation of TBs.

Beryllium and tungsten dust transport is simulated for ITER and DEMO using the dust transport code MIGRAINe. By combining many single-particle simulations a statistical analysis is performed. From that, the evolution of the dust inventories in the reactors is studied. Additionally, predictions are made for the locations at which dust evaporates. Finally, simplified analytical models with a few parameters for fast estimations of the dust inventory evolution are proposed. In the future, these could be integrated into more global models that also take into account dust generation. / Dammtransporten för beryllium och volfram simuleras för ITER och DEMO med dammtransportkoden MIGRAINe. En statistisk analys utförs genom att kombinera många enkelpartikelsimuleringar varpå utvecklingen av damminventeringe i reaktorerna studeras. Dessutom genomförs förutsägelser för de platser där damm evaporerar. Slutligen föreslås förenklade analytiska modeller med få parametrar för snabba uppskattningar av utvecklingen av damminventering. I framtiden kan dessa integreras i mer globala modeller som också tar hänsyn till dammgenerering.

Les conditions pour la fusion DT peuvent-être obtenues dans les tokamaks. Dans ces machines, l'interaction plasma-paroi et l'extraction de puissance sont gérées dans une cavité appelée divertor. Toutefois, les hautes puissances impliquées et les limitations des composants face au plasma (CFP) sont problématiques. Ce domaine fait l'objet de nombreuses recherches dans le contexte de ITER qui doit démontrer 500 MW de puissance fusion durant 400 s. Ces opérations nécessitent sur la réduction des flux de chaleur sur les CFP à un niveau gérable et repose sur le détachement du plasma dans le divertor qui implique la décroissance des flux de particules et de chaleur. Malheureusement, ce processus demeure difficile à modéliser. Le but the ce doctorat est d'utiliser la modélisation d'expériences de JET avec EDGE2D-EIRENE pour faire des progrès dans la compréhension du détachement. Les simulations reproduisent le détachement observé en environnement C comme Be/W. La distribution du rayonnement est reproduite par le code en C mais des écarts subsistent en Be/W. La comparaison entre différents processus de physique atomique montre que les collisions élastiques ion-molécule sont responsables du détachement. Ce processus permet le confinement des neutres dans le divertor ainsi que des pertes de moments significatives à basse température lorsque le plasma est recombinant. La comparaison entre EDGE2D-EIRENE et SOLPS4.3 montre des tendances similaires pour le détachement. Les deux codes suggèrent que tout processus capable d'améliorer le confinement des neutres dans le divertor devrait faciliter la modélisation du détachement. / The conditions required for fusion can be obtained in tokamaks. In most of these machines, the plasma wall-interaction and the exhaust of heating power are handled in a cavity called divertor. However, the high heat flux involved and the limitations of the materials of the plasma facing components (PFC) are problematic. Many researches are done this field in the context of ITER which should demonstrate 500 MW of DT fusion power during ~ 400 s. Such operations could bring the heat flux on the PFC too high to be handled. Its reduction to manageable levels relies on the divertor detachment involving the reduction of the particle and heat fluxes on the PFC. Unfortunately, this phenomenon is still difficult to model. The aim of this PhD is to use the modelling of JET experiments with EDGE2D-EIRENE to make some progress in the understanding of the detachment. The simulations reproduce the observed detachment in C and Be/W environments. The distribution of the radiation is well reproduced by the code for C but with some discrepancies in Be/W. The comparison between different sets of atomic physics processes shows that ion-molecule elastic collisions are responsible for the detachment seen in EDGE2D-EIRENE. This process provides good neutral confinement in the divertor and significant momentum losses at low temperature, when the plasma is recombining. Comparison between EDGE2D-EIRENE and SOLPS4.3 shows similar detachment trends but the importance of the ion-molecule elastic collisions is reduced in SOLPS4.3. Both codes suggest that any process capable of improving the neutral confinement in the divertor should help to improve the modelling of the detachment.

Le quench d'un système magnétique d'ITER est une transition irréversible d'un conducteur, de l'état supraconducteur à l'état normal résistif. Cette zone normale se propage le long du câble au cours du temps, en dissipant une grande quantité d'énergie. La détection se doit d'être suffisamment rapide afin de permettre une décharge de l'énergie magnétique et éviter un endommagement permanent du système. La détection primaire de quench d'ITER est basée sur la détection de la tension due au quench, qui est le moyen le plus rapide. L'environnement magnétique perturbé pendant le scenario plasma rend la détection de cette tension très difficile, à cause des hautes tensions inductives qu'il génère dans les bobinages. En conséquence, des compensations de tension sont nécessaires afin de discriminer la tension résistive due au quench.Une solution conceptuelle de la détection de quench basée sur la mesure des tensions est proposée pour les trois grands systèmes magnétiques d'ITER. Pour ceci, une méthodologie claire est développée, incluant le calcul classique selon le critère du point chaud, l'étude de la propagation de quench grâce au code commercial Gandalf, et l'estimation des perturbations inductives, grâce au développement du code TrapsAV. Des solutions adaptées sont proposée pour ces systèmes ainsi que les paramètres de cette détection, qui sont le seuil de détection (entre 0.1 V et 0.55 V) et le temps de discrimination (entre 1 s et 1.2 s). Les valeurs choisies, et en particulier le temps de discrimination, sont suffisamment élevées pour garantir la fiabilité du système, et pour éviter le déclenchement intempestif de décharges rapides non nécessaires. / The quench of one of the ITER magnet system is an irreversible transition from superconducting to normal resistive state, of a conductor. This normal zone propagates along the cable in conduit conductor dissipating a large power. The detection has to be fast enough to dump out the magnetic energy and avoid irreversible damage of the systems. The primary quench detection in ITER is based on voltage detection which is the most rapid detection. The very magnetically disturbed environment during the plasma scenario, makes the voltage detection particularly difficult, inducing large inductive components in the coils and voltage compensations have to be designed to discriminate the resistive voltage associated with the quench. A conceptual design of the quench detection based on voltage measurements is proposed for the three majors magnet systems of ITER. For this, a clear methodology was developed. It includes the classical hot spot criterion, the quench propagation study using the commercial code Gandalf and the careful estimation of the inductive disturbances by developing the TrapsAV code.Specific solutions have been proposed for the compensation in the three ITER magnet systems and for the quench detection parameters which are the voltage threshold (in the range of 0.1 V- 0.55 V) and the holding time (in the range of 1 -1.4 s). The selected values, in particular the holding time, are sufficiently high to ensure the reliability of the system and avoid fast safety discharges not induced by a quench which is a classical problem.

[pt] O processo de formação presbiteral é constituído por um itinerário que corresponde à vida do ministro sacerdotal desde o seu despertar vocacional até a conclusão de seus dias sobre a terra. Nesse percurso, encontram-se as etapas de formação inicial e permanente que se complementam, por se caracterizar como um processo que, além de unitário, é integral, enquanto inter-relaciona as dimensões humano afetiva, espiritual, pastoral e intelectual, do seminarista ao sacerdócio ministerial, num iter dinâmico, de modo a lhes favorecer o amadurecimento necessário para cumprir sua missão. O Seminário Arquidiocesano de São José do Rio de Janeiro possui um itinerário formativo elaborado a partir da Ratio Institutionis Sacerdotalis: O dom da vocação presbiteral, sobre o qual esta pesquisa se detém como o seu objeto material, elucidando a contribuição das etapas e das dimensões da formação para o desenvolvimento da antropologia da vocação presbiteral. Nesse sentido, a presente dissertação perpassa alguns autores patrísticos acerca da teologia e da práxis sacerdotal, os atuais desafios antropológicos para a formação presbiteral e o progressivo desenvolvimento humano e espiritual dos formandos ao longo do iter formativo. Distribuída em cinco partes, a pesquisa tem início, identificando a teologia do ministério presbiteral e seus traços antropológicos no testemunho patrístico da Didaqué, Clemente de Roma, Inácio de Antioquia, Policarpo de Esmirna, Papias de Hierápolis, Hermas, Barnabé e Justino de Roma, prosseguindo por meio dos aspectos unitários e integrais do processo de formação, em que são desenvolvidos os temas das dimensões da formação e das etapas formativas da pastoral vocacional, do seminário menor, do propedêutico, do discipulado, da configuração e da síntese. Os aspectos teológicos e antropológicos do processo formativo encerram a pesquisa, ressaltando os desafios para o desenvolvimento do formando e as propostas para a sua maturação humana e vocacional. / [en] The process of priestly formation is constituted by an itinerary that corresponds to a priest s life from his vocation awakening to the end of his days on Earth. In this path, there are the stages of initial and permanent formation that complement each other. This can be characterized as a process that besides being unitary is also wholesome insofar as it interrelates the affective, spiritual, pastoral, and intellectual human dimensions in a seminarian life up to the priesthood within a dynamic iter whose aim is to encourage him towards the necessary maturity to accomplish his mission. The Archdiocesan Seminary of Saint Joseph in Rio de Janeiro has a formative pathway based upon the Ratio Institutionis Sacerdotalis: The gift of the priestly vocation, in which this research focuses as its material object, elucidating the contribution of the stages and dimensions of formation towards the development of an anthropology of the priestly vocation. In this context, this present work runs through some patristic authors’ theologies and their related priestly praxis, the current anthropological challenges for priestly formation, as well as seminarians’ onward human and spiritual development throughout their formation iter. Divided in five parts, this research’s starting point describes priesthood theologies, and anthropological traits in the patristic testimony of Didache, Clement of Rome, Ignatius of Antioch, Polycarp of Smyrna, Papias of Hierapolis, The Shepherd of Hermas, Barnabas, and Justin of Rome, undertaking the unitary and integral aspects of the formation process in which dimension elements of formation, vocation ministries formative stages, minor seminary, propaedeutic, discipleship, configuration and synthesis are developed. Theological and anthropological aspects of the formative process conclude this research as it highlights challenges regarding the development of the person being formed and proposals whose aims are human and vocation maturation.

The International Thermonuclear Experimental Reactor (ITER) will be a major step towards controlled energy fusion in tokamaks. Operation of the hot confined plasma inside the tokamak will have to be optimized and simulations will have to prove that each pulse conducted is feasible under the operational limits of the reactor. For such purpose a Pulse Design Simulator is developed at ITER. This workflow lacks a transport model to simulate the dynamics of the plasma caused by micro-instabilities driven by turbulences. The purpose of this thesis is the adaptation of such model into the Integrated Modelling and Analysis Suite (IMAS), namely mapping the inputs and outputs of an existing code for its integration to the workflow. This work presents a fast 1D core transport code capable of simulating the evolution of the poloidal flux, the temperature evolution of both ions and electrons and the particle density transport. The model is coupled to a neural network regression of the transport model QuaLiKiz for the computation of first-principle based turbulent heat and particle transport coefficients. / Internationella Termonukleära Experimentella Reaktorn (ITER) kommer att vara ett stort steg mot kontrollerad energifusion i tokamaker. Driften av det varma, instängda plasmaet inne i tokamaken måste optimeras, och simuleringar måste bevisa att varje pulsskötning är genomförbar inom reaktorns driftgränser. För detta ändamål utvecklas en pulsdessignsimulator vid ITER. Denna arbetsflöde saknar en transportmodell för att simulera plasmaets dynamik orsakad av mikroinstabiliteter drivna av turbulenser. Syftet med denna avhandling är anpassningen av en sådan modell till Integrated Modelling and Analysis Suite (IMAS), nämligen att kartlägga in- och utdata av en befintlig kod för dess integration i arbetsflödet. Denna arbete presenterar en snabb 1D-kärntransportkod som kan simulera utvecklingen av den poloidala flödet, temperaturutvecklingen för både joner och elektroner samt partikeltäthetstransporten. Modellen är kopplad till en neural nätverksregression av transportmodellen QuaLiKiz för beräkning av första principbaserade turbulenta värme- och partikeltransportkoefficienter.

The development of the negative ion source constitutes a crucial step in the construction of the neutral beam injector of ITER. To fulfil the ITER requirements in terms of heating and current drive, the negative ion source should deliver 40 A of D-. The achievement of such a source is challenging from technical and scientific points, and it requires a deeper understanding of the underlying physics. The present knowledge of the ion extraction mechanism from the negative ion source is limited due to the complexity of the problem that involves the comprehension of the behaviour of magnetized plasma sheaths when negative ions and electrons are pulled out from the plasma. Moreover, due to the asymmetry induced by the crossed magnetic configuration used to filter the electrons, any realistic study of this problem must consider the three spatial dimensions. To address this problem in a realistic way, a 3D Particles-in-Cell electrostatic code specifically designed for this system was developed. The code uses Cartesian coordinate system and it can deal with complex boundary geometry as it is the case of the extraction apertures. The complex magnetic field that is applied to deflect electrons is also taken into account. This code, called ONIX, was used to investigate the plasma properties and the transport of negative ions and electrons close to a source extraction aperture. Results on the formation of the plasma meniscus and the screening of the extraction field by the plasma are presented here, as well as negative ions trajectories. Negative ion extraction efficiency from volume and surfaces was investigated showing the capital importance of the surface negative ion production.