Our research study focuses on exploring how 3D-students view AI image generation tools usefullness as concept images in 3D production. We conducted interviews with five 3D students to gain insight into how useful they find AI-generated images in 3D production. By collecting information through interview method, participants' responses were analyzed using thematic analysis. Their responses were combined with existing theories in concept art, disruptive technologies, and sparing technologies. Our main findings show that students found images made by using AI can be used as inspiration images and moodboards but have difficulty being used for anything else in a 3D production, as AI was poor at reproducing the same object multiple times. The same object couldn't be obtained from multiple angles which limits usability. However the 3D-students think there is potential in using the images as a foundation that is then further developed by a human. / Vår forskningsstudie fokuserar på att utforska hur 3D-studenter ser på användbarheten av AI-bildgenereringsverktyg som konceptbilder i 3D-produktion. Vi genomförde intervjuer med fem 3D-studenter för att få insikt i hur användbara de tycker att AI-genererade bilder är i 3D-produktion. Genom att samla in information genom intervjuer analyserades deltagarnas svar med tematisk analys. Deras svar kombinerades med befintliga teorier inom konstkoncept, störande teknologier och sparsamma teknologier. Våra huvudsakliga resultat visar att studenterna fann att bilder skapade med AI kan användas som inspirationsbilder och moodboards men har svårigheter att användas för något annat i 3D-produktion, eftersom AI hade svårt att reproducera samma objekt flera gånger. Samma objekt kunde inte erhållas från olika vinklar vilket begränsar användbarheten. Dock tror 3D-studenterna att det finns potential att använda bilderna som en grund som sedan utvecklas ytterligare av en människa.

Cone-sheet swarms are magmatic sheet intrusions and part of volcanic plumbing systems and are pathways for magma to the Earth’s surface, where they feed volcanic eruptions. The analysis of cone-sheets provides information on the geometry of the magmatic plumbing system of a volcano and allows to understand processes and dynamics of magma transport. This is important to interpret information during a volcanic crisis and to help reduce risks to humans and infrastructure. In order to create a realistic model, the structure and shape of cone-sheet complexes can be reconstructed in three-dimensional space. Most cone-sheet swarms are not sufficiently exposed to allow such a reconstruction. The Tejeda cone-sheet swarm on Gran Canaria (Canary Islands, Spain), however, is an excellent location to study a cone-sheet complex in great detail, as it is exposed over 15 kmhorizontally and 1000 m vertically. This allows to determine its geometry in 3D space. The felsic deposits of the Miocene Tejeda caldera were intruded by cone-sheets between 11.7 and 7.3 Ma. Schirnick et al. (1999) assumed straight cone-sheets, based on 2D projections, and suggested that the Tejeda cone-sheet swarm is formed by a stack of uniformly dipping, parallel intrusive sheets that converge towards a common, static, laccolith-like source, forming a concentric structure around acentral axis that has the geometry of a truncated cone. This hypothesis was tested in this study, using structural data from Schirnick (1996) as well as additional data collected in the field. Using the software Move™, the extensive data set was visualized and projected in three dimensional space. The underlying magmatic source of the cone-sheets was reconstructed using two different approaches, with the first one based on sets of cross-sections to select intersection points, following an approach prognosed by Burchardt et al. (2013a). To improve the quality of the reconstruction of the magma chamber, a second method was developed using geometric calculations in MATLAB. The results indicate that individual cone-sheets are straight with parallel to slightly fanning dips, which can be steeper in the central part of the cone-sheet complex. They converge towards a common centre, creating a sub-spherical geometry of the source of the cone-sheet complex. Comparison of the two approaches used for the magma chamber reconstruction indicate that the second approach (geometric calculations) produces less uncertainties in data interpretation. The modelling results lead to the proposition of a dynamic model for the emplacement of the Tejeda cone-sheet complex. Cone-sheets would start to intrude from a reservoir situated at about 4500m below sea level that became successively shallower with time. / Inverterade koniska intrusionssvärmar är en del av det underjordiska vulkaniska systemet som möjliggör vägar för magma till jordens yta, där de livnär vulkaniska utbrott. Genom analys av inverterade koniska intrusioner kan information om geometrin hos magmatiska system erhållas vilket gör det möjligt att förstå magmans processer och transportdynamik. Detta är viktigt då det hjälper att tolka information under vulkaniska kriser och kan bidra till att minska risker för människor och infrastruktur. För att skapa en realistisk modell, kan strukturer och former av komplexa inverterade koniska intrusionssvärmar rekonstrueras i ett tredimensionellt utrymme. De flesta inverterade koniska intrusionssvärmar är inte tillräckligt blottade på jordens yta för att möjliggöra en sådan rekonstruktion.Tejedas inverterade koniska intrusionssvärm på Gran Canaria (Kanarieöarna, Spanien) är dock utmärktbelägen för att studera ett inverterat koniskt intrusionskomplex i detalj, detta då den är blottad i över 15 km horisontell och 1000 m i vertikal utsträckning. Detta gör det möjligt att bestämma dess geometrii tre dimensioner. De felsiska avlagringarna av den Miocena Tejeda kalderan blev intruderade av inverterade koniskaintrusioner mellan 11,7 och 7,3 Ma. Schirnick et al. (1999) antog att dessa intrusioner var raka, baseratpå 2D-projektioner, och föreslog att Tejedas inverterade koniska intrusionssvärm bildades som enlikformigt stupande stapel av parallella intruderande plan som konvergerar mot en gemensam, statiskoch lakkolitisk källa, vilken i sin tur bildar en koncentrisk struktur runt en central axel med samma geometri som en inverterad stympad kon. Denna hypotes undersöktes i detta arbete, med hjälp avstrukturell data från Schirnick (1996) samt ytterligare data insamlat från fält. Den omfattande datamängden visualiserades och projicerades i tre dimensioner med hjälp av mjukvaran Move™. Denunderliggande magmatiska källan till det inverterade koniska intrusionskomplexet rekonstruerades medhjälp av två olika metoder, den första är baserad på tvärsnitt där planens skärningspunkter kan studeras, följt av ett tillvägagångssätt framställt av Burchardt et al. (2013a). För att förbättra kvalitén pårekonstruktionen av magmakammaren utvecklades en andra metod med hjälp av geometriskaberäkningar i MATLAB. Resultaten tyder på att enskilda inverterade koniska intrusioner är raka med parallellt till svagt flacktstupning, vilka kan vara brantare mot den centrala delen av komplexet. De konvergerar mot ettgemensamt centrum, vilket skapar en sub-sfärisk geometri hos källan till det inverterade koniskaintrusionskomplexet. Jämförelse av de två metoderna som används för magmakammarens rekonstruktion tyder på att denandra metoden (geometriska beräkningar) ger färre osäkerheter i tolkningen. Modelleringsresultatettyder på en dynamisk modell för bildningen av Tejedas inverterade koniska intrusionskomplex. Enligtdessa resultat skulle de inverterade koniska intrusionerna till en början ha utgått från en reservoarungefär 4500 m under havsytan som med tiden förflyttade sig mot grundare nivåer.

Facial Ageing (FA) is a very fundamental issue, as ageing in general, is part of our daily life process. FA is used in security, finding missing children and other applications. It is also a form of Facial Recognition (FR) that helps identifying suspects. FA affects several parts of the human face under the influence of different biological and environmental factors. One of the major facial feature changes that occur as a result of ageing is the appearance and development of wrinkles. Facial wrinkles are skin folds; their shapes and numbers differ from one person to another, therefore, an advantage can be taken over these characteristics if a system is implemented to extract the facial wrinkles in a form of maps. This thesis is presenting a new technique for three-dimensional facial wrinkle pattern information that can also be utilised for biometric applications, which will back up the system for further increase of security. The procedural approaches adopted for investigating this new technique are the extraction of two-dimensional wrinkle maps of frontal human faces for digital images and the design of three-dimensional wrinkle pattern formation system that utilises the generated wrinkle maps. The first approach is carried out using image processing tools so that for any given individual, two wrinkle maps are produced; the first map is in a binary form that shows the positions of the wrinkles on the face while the other map is a coloured version that indicates the different intensities of the wrinkles. The second approach of the 3D system development involves the alignment of the binary wrinkle maps on the corresponding 3D face models, followed by the projection of 3D curves in order to acquire 3D representations of the wrinkles. With the aid of the coloured wrinkle maps as well as some ageing parameters, simulations and predictions for the 3D wrinkles are performed.

L’objectif de ce travail est d’étudier la dynamique du bouchon vaseux, ainsi que la dynamique saline, dans l’estuaire de la Gironde. La stratégie adoptée consiste à examiner, par le biais du modèle de simulation SIAM-3D, l’impact du paramétrage hydro-sédimentaire, du débit fluvial et de la bathymétrie, sur le comportement des sédiments et sur l’évolution de la salinité. L’effet de la vitesse de chute sur la dynamique sédimentaire est l’un des cas étudiés ; son impact sur la distribution des suspensions est considérable. Ainsi, la diminution de ce paramètre entraîne l’homogénéisation de la concentration sur la verticale, mais aussi l’étalement du bouchon vaseux et l’évacuation d’une masse de sédiment hors de l’estuaire. Nous avons également mis en évidence lors d’une période de débits moyens ou faibles précédée par une crue, la partition du bouchon vaseux simulé en deux masses turbides, l’une remonte l’estuaire alors que l’autre se maintient en aval. La dynamique saline a été étudiée, entre autres, par l’analyse des gradients de salinité, calculés dans l’estuaire aval et en situation de forts débits. On observe, la diminution de ces gradients avec l’augmentation de l’intensité de la marée à l’échelle des cycles vives-eaux/mortes-eaux, toutefois, de fortes valeurs sont maintenues lors de la phase d’augmentation de cette grandeur à l’échelle de la marée. D’autres résultats ont permis de vérifier la capacité du modèle à simuler des dynamiques sédimentaires observées in-situ ; on peut citer entre autres, le détachement du bouchon vaseux de la rive gauche dans l’estuaire aval et son transfert vers la rive droite, en situation de forts débits. Afin d’évaluer l’impact de l’évolution du fond des fleuves de 1959 à 2002 sur l’hydrodynamique et les dynamiques sédimentaire et saline, nous avons comparé deux simulations, l’une effectuée avec l’ancienne bathymétrie de 1959, l’autre avec la nouvelle bathymétrie de 2002. Les résultats montrent l’augmentation de la masse sédimentaire dans les fleuves avec la bathymétrie de 2002 comparée à celle de 1959, ainsi que la hausse de la salinité dans la Garonne et l’amplification de la marée dans la Dordogne. Ces résultats sont en accord avec les mesures et observations in-situ. / The aim of this work is to study the turbidity maximum dynamics as well as the salinity dynamics in the Gironde estuary. The adopted strategy consists in analysing, by means of the SIAM-3D simulation model, the impact of hydro-sedimentary parameters, the fluvial flow and the bathymetry on sediment behaviour and salinity evolution. The effect of settling velocity on sediment dynamics is one of the studied cases; its impact on suspended sediment distribution is goodly. Thus, the diminution of this parameter leads to homogenisation of concentration in water column, but also spreading the turbidity maximum and evacuation of a sediment mass outside the Gironde estuary. We gave also prominence to the division of the turbidity maximum into two parts during low or mean water flow preceded by peak high water discharge, one of the parts migrates upstream whereas the other remains downstream. We studied the salinity dynamics, among others things, by means of analysing the calculated salinity gradients in the lower estuary during high water discharge. We observe the decrease of these gradients with increase of the tide intensity on a scale of spring tides/neap tides cycles, however, on a scale of tide, these gradients reach high values during the augmentation phase. Others results indicate the model capacity to simulate in-situ observations of the sediment dynamics; for example, the detachment of turbidity maximum from the left bank in the lower estuary and its transfer to the right bank during high water discharge. In order to evaluate the impact of the bed river evolution from 1959 to 2002 on hydrodynamics, sediment and salinity dynamics, we compared two simulations performed using the former (1959) and recent (2002) bathymetries. The results demonstrate that the sediment mass increase in the rivers with recent bathymetry, as well as, the salinity in the Garonne river and the tide amplitude in the Dordogne river. These results are in accordance with the in-situ measures and observations.

Det har blivit mer populärt att skapa 3D-modeller bland kommuner och företag idag. Det går att visualisera de flesta föremål i 3D, vilket kan skapa större tydlighet och därmed ökad förståelse. Lidköpings kommun vill kunna använda sin LiDAR-data till att visualisera byggnader samt kunna kombinera flera datalager med 3D-visualiseringar för att kunna skapa användbara karttjänster. I dagsläget finns det ingen programvara som kan skapa 3D-modeller av LiDAR-data i kommunen. Lidköpings kommun är intresserade av vilka alternativ för GIS-programvaror som finns. De programvaror som valts ut för att skapa 3D-modeller i studien är ArcGIS Pro och QGIS. ArcGIS Pro är en erkänd GIS-programvara på marknaden. Den är licensierad och kostsam. QGIS däremot är en GIS-programvara som är gratis och uppbyggd på fri källkod, vilket gör det möjligt att själv skapa algoritmer som sedan kan användas i programmet. Studien är en jämförelse mellan en programvara som är kostsam och en som är gratis. Metoden som används för att jämföra de båda programvarorna består av tre steg. Steg 1 är en programvarujämförelse, där olika påståenden ställer programvarornas kapacitet mot varandra. Steg 2 är en kontroll av lägesosäkerhet där koordinaterna av 3D-modellernas byggnadshörn jämförs med koordinater inmätta med totalstation. RMSE används för att jämföra lägesosäkerheten. Steg 3 är en visualiseringsanalys som sker igenom ett frågeformulär, vilket går ut till personer som är bekanta med GIS och 3D. Frågeformuläret ställer bilder med 3D-modeller från båda programvarorna mot varandra. Resultatet av samtliga tre steg visar, enligt metoden som har utförts, att ArcGIS Pro är den bättre programvaran. I steg 1 hade ArcGIS Pro fördelar på fem av åtta påståenden jämfört med QGIS. QGIS hade dock mindre tidsåtgång än ArcGIS Pro. I steg 2 hade ArcGIS Pro RMSE för höjd värdet; 1,617, vilket var lägre än RMSE-värdet; 1,961 i QGIS. Enligt svaren från enkäten i steg 3 bedömde ca 90 % av de svarande att ArcGIS Pro var det bättre programmet i de flesta avseenden för de avsedda ändamålen. Vissa jämförelser blev inte helt jämna. QGIS hade minst tidsåtgång för att det inte blev lika mycket felkoder som behövde undersökas som i ArcGIS Pro, även om processen för skapandet av 3D-modeller gick snabbare i ArcGIS Pro. Lägesosäkerhetsjämförelsen på höjd i ArcGIS pro var närmare de inmätta punkterna, då takmodellen skapade större tydlighet. Det blev lättare att se var taket började. I QGIS fanns det ingen takmodell, vilket gjorde det svårare att få höjden exakt. / It has become more popular to create 3D models among local governments and companies today. Now it´s possible to visualize most objects in 3D, which can give an increased understanding. The local government in Lidköping wants to be able to use their LiDAR data to visualize buildings and to combine several data layers with 3D visualizations to create useful web map services. At the present, there is no software that are able create 3D models of LiDAR data at the Local government. The local government in Lidköping is interested in testing different options for modelling in 3D. Two software packages were chosen; ArcGIS Pro and QGIS. ArcGIS Pro is a GIS software that is highly acknowledged in the market. It´s licensed and can be considered expensive. QGIS is a software that is completely free to use and uses open source code, which makes it possible to create algorithms that can then be used in the program. In this view, there will be comparison between a software that is expensive and a software that is completely free. The method used to compare the two software consists of three stages. Stage 1 is a software comparison, where the software capabilities are set against each other in different scenarios. Stage 2 is an investigation of positional accuracy where the coordinates of the 3D model's building corners are compared with coordinates measured with Total Station as reference. RMSE is used to compare positional accuracy. Stage 3 is a visualization analysis that takes place through a questionnaire answered by people who are familiar with GIS and 3D. The result of all three stages, according to the method performed, shows that ArcGIS Pro is the better software. At stage 1, ArcGIS Pro won five out of eight scenarios against QGIS, though QGIS was more time efficient than ArcGIS Pro. At stage 2, ArcGIS Pro had a RMSE value for height of 1,617, which is lower than the 1,961 RMSE value that was on QGIS. At stage 3, ArcGIS Pro had over 90% of the reply rate for most of the questions, representing that ArcGIS Pro’s images were better to visualize the different scenarios. Some comparisons were not completely even though. QGIS had better time accuracy because it did not get as much errors which needed further investigations as in ArcGIS Pro. Although ArcGIS Pro’s processes for creating 3D models were faster. The positional accuracy comparison for the building’s height shows higher accuracy between ArcGIS Pro and the Total Station’s measurements. In QGIS the 3D-modell had no roof model, which made it harder to get accurate height of the building.

La numérisation est un procédé qui consiste à enregistrer un objet dans un ordinateur pour pouvoir ensuite le manipuler à l'aide d'outils informatiques. Nous nous intéressons dans ce manuscrit à la numérisation d'objets tridimensionnels. Il s'agit tout d'abord d'enregistrer leur forme. De nombreuses méthodes ont été développées pour répondre à ce problème, et nous nous concentrerons sur les objets représentés par des maillages. Sur ces objets, il est alors utile de pouvoir représenter des attributs tels que la couleur, la température ou la charge électrique, selon l'application. Nous proposons deux approches complémentaires pour aborder ce problème. La première est fondée sur le placage de textures. Cette technique consiste à déplier (paramétrer) le maillage à plat sur une image dans laquelle l'attribut est stocké. Une valeur récupérée dans l'image est ainsi associée à chaque point de l'objet. Nous proposerons une méthode permettant de masquer l'artéfact des coutures qui est inhérent à cette technique. Déplier le maillage nécessite qu'il soit de bonne qualité, ce qui n'est pas toujours le cas. Nous décrivons donc également dans un second temps une approche de l'échantillonnage d'une surface via un diagramme de Voronoï restreint. Nous expliquons en particulier comment calculer efficacement un tel objet et comment l'optimiser par rapport à un critère de qualité. Ces résultats sont ensuite appliqués au problème de l'ajustement de surfaces / Digitalisation is an operation which consists in storing an object in a computer for further manipulation using data processing tools. In this document, we are interested in the digitalisation of three-dimensional objects. It is first a matter of recording the shape of the object. Many methods have been developed to address this problem, and we will focus on objects described as meshes. On such objects the storage of attributes like colour, temperature or electrical charge is often useful, depending on the application. We will describe two complementary approaches to deal with this issue. The first one relies on texture mapping. This technique consists in unfolding ? parametrising ? the mesh on a flat image in which the attribute is stored. A value recovered from the image can therefore be associated with each point of the object. We will describe a method which hides the seam artifact, commonly encountered using this technique. Unfolding the mesh demands that its quality be good, which is not always the case. We thus secondly describe a surface sampling method based on a restricted Voronoï diagram. We especially detail how to efficiently compute such an object and how to optimise it with respect to some quality measure. These results are then applied to the surface fitting problem

Dans le cadre d’un projet d’aménagement, la reconnaissance du sous-sol est une étape prépondérante pour la construction d’un bâtiment. Le géotechnicien se doit de définir le comportement mécanique des sols à partir d’un nombre limité de sondages. La problématique soulevée par les prospections géotechniques est qu’il est difficile d’une part d’étendre une information ponctuelle dans l’environnement proche des sondages, et d’autre part d’estimer la variabilité latérale des matériaux. Dans cette optique, cette thèse développe une approche novatrice de combinaison de méthodes géophysiques aidant à la quantification de la variabilité spatiale des structures géoélectriques du sous-sol. Dans un second temps, au regard des limites rencontrées dans la modélisation 3D des structures géoélectriques du sous-sol, nous avons mis au point un processus innovant de modélisation 3D des propriétés physiques du sous-sol par combinaison de méthodes géophysiques. Les travaux menés ont permis de caractériser les paramètres régissant le positionnement et la quantité de données sources nécessaires.Enfin, un dispositif innovant de mesure de Tomographie de Résistivité Electrique en Cinématique (TREC) a été développé. La validation du dispositif sur trois sites différents a permis de mettre en évidence sa capacité à détecter les structures géoélectriques des sols sur un linéaire de plusieurs centaines de mètres. Ce dispositif constitue un outil précieux pour l’étude des structures géoélectriques du sous-sol dans le cadre de prospections géophysiques à grand rendement. / The geotechnical study of a building project is a decisive step for the construction of it. The geotechnical engineer must define the mechanical behavior of soils from a limited number of drills. The problem of the geotechnical surveys is, on the one hand that difficult to extend a punctual data in the near space to the drills,and on the other hand to estimate the lateral variability of materials. In this context, this thesis develops an innovative approach by combination of geophysical methods to quantify the spatial variability of geoelectrical structures. In asecond time, against the known limits of the 3D modeling of soil geoelectrical structures, we have developed an innovative process of 3D modeling of the physical properties of soil by geophysical methods combination. These works allow characterizing the parameters who governing the positioning and the number of necessary primary data.Finally, an innovative measuring device of Kinematic Electrical Resistivity Tomography (KERT) has been developed. The validation of the device on three different site has allowed to highlight its ability to detect the soil geoelectrical structures within a linear of hundreds meters. This device is a valuable tool for studying soil geoelectrical structures in the context of high yield geophysical prospection.

This thesis relates to research and development in the field of 3D mesh data for computer graphics. A review of existing storage and manipulation techniques for mesh data is given followed by a framework for mesh editing. The proposed framework combines complex mesh editing techniques, automatic level of detail generation and mesh compression for storage. These methods work coherently due to the underlying data structure. The problem of storing and manipulating data for 3D models is a highly researched field. Models are usually represented by sparse mesh data which consists of vertex position information, the connectivity information to generate faces from those vertices, surface normal data and texture coordinate information. This sparse data is sent to the graphics hardware for rendering but must be manipulated on the CPU. The proposed framework is based upon geometry images and is designed to store and manipulate the mesh data entirely on the graphics hardware. By utilizing the highly parallel nature of current graphics hardware and new hardware features, new levels of interactivity with large meshes can be gained. Automatic level of detail rendering can be used to allow models upwards of 2 million polygons to be manipulated in real time while viewing a lower level of detail. Through the use of pixels shaders the high detail is preserved in the surface normals while geometric detail is reduced. A compression scheme is then introduced which utilizes the regular structure of the geometry image to compress the floating point data. A number of existing compression schemes are compared as well as custom bit packing. This is a TIF funded project which is partnered with Unlimited Realities, a Palmerston North software development company. The project was to design a system to create, manipulate and store 3D meshes in a compressed and easy to manipulate manner. The goal is to create the underlying technologies to allow for a 3D modelling system to become integrated into the Umajin engine, not to create a user interface/stand alone modelling program. The Umajin engine is a 3D engine created by Unlimited Realities which has a strong focus on multimedia. More information on the Umajin engine can be found at www.umajin.com. In this project we propose a method which gives the user the ability to model with the high level of detail found in packages aimed at creating offline renders but create models which are designed for real time rendering.

Spatial ability and the skill to visualize objects is necessary for earth scientist, especially structural geologist. For this reason, undergraduate students within earth science needs to learn this skills, and how is the best way to teach this skills. To implement 3D thinking in undergraduate teaching, three strategies are presented. The first one is to separate and combine objects, which includes to see geological differences and categorize them. Secondly, visualize objects, both many and single ones and see which ones who are connected. Third and lastly is still and moving objects which involves geological processes. It is important to give students time to develop their spatial abilities and help them during the learning process. Another aspect to this is the path of information within the human brain when visualizing an object. The result shows that the process starts with the human eye which perceive the object and its attributes, then the geniculate nucleus sorts the information and directs it toward the visual primary cortex located in the occipital lobe. The primary visual cortex then send the information though the ventral- and dorsal steam which produces a visualization. Aspects which can affect the spatial ability may be earlier experiences, age and the way each person perceive the object. These factors effects how hard humans think the process of visualization is. The questionnaire showed that 3D models help students to visualize and should be used more in undergraduate teaching. / Förmågan att visualisera objekt och förstå dem är nödvändig för en geovetare, speciellt inom strukturgeologi. Därför är det viktigt att studenter på kandidatprogram i geovetenskap får tillfälle att lära sig den förmågan. För att implementera 3D-tänkande i undervisningen har tre strategier tagits fram. Den första är att separera och kombinera objekt, vilket inkluderar att se geologiska skillnader och kategorisera dem. Den andra är att visualisera ett eller flera objekt och se vilka som är kopplade till varandra. Den tredje är att se objekt som är i stilla eller i rörelse, vilket involverar geologiska processer. Det är viktigt att ge studenterna tid för att utveckla sin spatiala förmåga och att hjälpa dem under utvecklingen. Arbetet syftar också till att ta reda på hur visuell information bearbetas av den mänskliga hjärnan. Först bearbetas informationen av ögat, sen till geniculate nucleus som sorterar informationen och skickar den till det primära visuella cortex som ligger i occipital-loben. Därefter processas informationen av ventrala och dorsala stammen och det är här som en visualisering produceras. Faktorer så som ålder, tidigare erfarenheter och hur ögat uppfattar objekten påverkar hur visualiseringen blir och hur svårt individen uppfattar processen. Enkätundersökningen visade på att 3D-modeller hjälper studenter att visualisera och borde användas mer i undervisning på kandidatnivå.

Le travail présenté dans ce mémoire traite de la problématique du dimensionnement de structures innovantes de génératrices, à savoir les machines à flux axial ou encore discoïdes, destinées aux éoliennes de grande puissance allant jusqu'à deux dizaines de mégawatts. Les machines synchrones à aimants permanents à flux axial étant intrinsèquement tridimensionnelles, l'optimisation de leur dimensionnement nécessite un modèle dimensionnant tridimensionnel et multiphysique qui peut être, de par sa nature 3D, très consommateur en mémoire et en temps de calcul. On se propose donc, dans ce mémoire, de développer une approche de modélisation par constantes localisées offrant un très bon rapport entre vitesse de calcul et précision afin d'accélérer de façon significative les premières phases de la procédure de dimensionnement. L'approche par constantes localisées peut par ailleurs s'appliquer de manière similaire aux principales physiques concernées (magnétique, thermique et mécanique). Plus particulièrement pour la partie magnétique, l'approche par constantes localisées est mise en œuvre par le biais de réseaux réluctants magnétiques génériques. Le principe de la méthode développée repose sur le découpage de la géométrie de la machine en volumes élémentaires finis parallélépipédiques à l'instar de qui est fait avec la méthode des éléments finis. En effet, chaque volume élémentaire étant modélisé par un réseau réluctant local, le découpage de la géométrie conduit à la génération automatique du système algébrique magnétique de la structure et dont la solution fournit la distribution locale du potentiel scalaire magnétique aux nœuds des volumes élémentaires. Cette approche de modélisation a été implémentée dans un code de calcul interne au GREAH. La modélisation de la génératrice à flux axial par réseaux réluctants a été validée par des simulations 2D et 3D par éléments finis. Le code de calcul développé intègrera les travaux en cours sur les modèles thermique et mécanique au GREAH afin d'aboutir à un code de calcul multi-physique par constantes localisées / The aim of this PhD thesis is to propose a design approach of the axial flux permanent magnet synchronous generator for multi megawatts wind converter application. The axial flux permanent magnet synchronous machine being intrinsically three dimensional, a 3D multiphysics sizing model is needed in order to conduct an efficient optimal design of this type of machines. Unfortunately, 3D multiphysic models are generally issued from finite element method. So, they are very heavy to be handled in a reduced time from an engineer point of view and especially at the first stages of the design procedure. In order to address the faster axial flux machine design, this PhD thesis develops a lumped parameters approach modelling offering a good computation time to precision ratio and thus helping to reduce considerably the spent time in the first stages design procedure. The lumped parameters modelling approach is also well suited for the modelling of the three main physics to be considered in an electrical generator (magnetic model, thermal model and mechanical model). For the magnetic model, the lumped parameters approach consists, in this thesis, in the development of a generic magnetic reluctance network linked to the machine geometry. In fact, the latter is divided into parallelepipedic elementary volumes, each of them is modelled by a local reluctance network. The reluctance network of the hole machine geometry results then automatically in an algebraic magnetic system, the solution of which provides the distribution of the scalar magnetic potential at the elementary volume nodes. The described lumped parameters modelling was implemented in a computational code developed at the GREAH Lab. The effectiveness of the reluctance network axial flux machine modelling was validated by 2D and 3D finite element simulations. The lumped parameters computational code will regroup the ongoing developments of thermal and mechanical models at the GREAH Lab in order to reach the completion of a mutliphysics lumped parameters computational code.