La majorité des produits verriers du marché sont issus d’une opération de mise en forme à hautes températures, suivie d’une phase de refroidissement contrôlé afin d’éliminer (verre recuit) ou générer (verre trempé) des contraintes résiduelles. Le comportement mécanique du verre étant fortement thermo-dépendant, le contrôle des températures est un élément déterminant pour le succès du procédé de fabrication. Lors de la simulation numérique, pour ce milieu semi-transparent, les échanges thermiques par conduction et par rayonnement sont à considérer. La résolution de l’ETR (Equation de Transfert Radiatif) est menée dans cette thèse par le biais de la « Méthode P1 » et le « Back Ray Tracing » (BRT). Les deux codes développés ont été validés par l’étude comparative avec les données en températures et en contraintes résiduelles issues de la littérature sur le refroidissement dans l’épaisseur du verre soumis à des conditions variées en convection naturelle et forcée. Une expérimentation qui consiste à refroidir un disque de verre sur un support métallique a été développée dans le but de comparer les températures et contraintes générées par l’expérimentation et par la modélisation issue du couplage thermomécanique avec les deux codes P1 et BRT. De manière plus originale, la méthode BRT a été étendue à des géométries de révolution. Une première approche a consisté à étudier le fluage d’une goutte de verre et à analyser l’influence du choix du modèle de résolution de l’ETR sur les températures et les géométries au cours de la mise en forme. / Most of glass products on the market come from a high-temperature forming operation, followed by a controlled cooling phase to remove (annealed glass) or generate (tempered glass) residual stresses. Since the mechanical behaviour of the glass is highly thermo-dependent, temperature control is a determining factor for the success of the manufacturing process. During numerical simulations, for this semi-transparent medium, heat exchanges by conduction and radiation have to be considered. In this work, the resolution of the ETR (radiative transfer Equation) is carried out using the "P1 method" and the "Back Ray tracing" (BRT). The two developed codes were validated by the comparative study with the temperature and residual stresses data from the literature on cooling in the thickness of the glass subject to various conditions in natural and forced convection. An experimentation which consists in cooling a glass disk on a metal support was developed in order to compare the temperatures and stresses generated by the testing and by the modelling resulting from the thermomechanical coupling with the two codes P1 and BRT. In a more original way, the BRT method was extended to revolving geometries. A first approach was to study the creep of a glass gob and to analyze the influence of the choice of the ETR's resolution model on the temperatures and geometries during the forming.

Background. Ray tracing is a tool that can be used to increase the quality of the graphics in games. One application in graphics that ray tracing excels in is generating shadows because ray tracing can simulate how shadows are generated in real life more accurately than rasterization techniques can. With the release of GPUs with hardware support for ray tracing, it can now be used in real-time graphics applications to some extent. However, it is still a computationally heavy task requiring performance improvements. Objectives. This thesis will evaluate the difference in performance of three raytracing methods in DXR Tier 1.1, namely dynamic ray tracing and two forms of inline ray tracing. To further investigate the ray-tracing performance, soft shadows will be implemented to see if the driver can perform optimizations differently (depending on the choice of ray-tracing method) on the subsequent and/or preceding API interactions. With the pipelines implemented, benchmarks will be performed using different GPUs, scenes, and a varying amount of shadow-casting lights. Methods. The scientific method is based on an experimental approach, using both implementation and performance tests. The experimental approach will begin by extending an in-house DirectX 12 renderer. The extension includes ray-tracing functionality, so that hard shadows can be generated using both dynamic- and the inline forms ray tracing. Afterwards, soft shadows are generated by implementing a state-of-the-art-denoiser with some modifications, which will be added to each ray-tracing method. Finally, the renderer is used to perform benchmarks of various scenes with varying amounts of shadow-casting lights and object complexity to cover a broad area of scenarios that could occur in a game and/or in other similar applications. Results and Conclusions. The results gathered in this experiment suggest that under the experimental conditions of the chosen scenes, objects, and number of lights, AMD’s GPUs were faster in performance when using dynamic ray tracing than using inline ray tracing, whilst Nvidia’s GPUs were faster when using inline ray tracing compared to when using dynamic ray tracing. Also, with an increasing amount of shadow-casting lights, the choice of ray-tracing method had low to no impact except for linearly increasing the execution time in each test. Finally, adding soft shadows(subsequent and preceding API interactions) also had low to no relative impact on the results depending on the different ray-tracing methods. / Bakgrund. Strålspårning (ray tracing) är ett verktyg som kan användas för att öka kvalitén på grafiken i spel. En tillämpning i grafik som strålspårning utmärker sig i är när skuggor ska skapas eftersom att strålspårning lättare kan simulera hur skuggor skapas i verkligheten, vilket tidigare tekniker i rasterisering inte hade möjlighet för. Med ny hårdvara där det finns support för strålspårning inbyggt i grafikkorten finns det nu möjligheter att använda strålspårning i realtids-applikationer inom vissa gränser. Det är fortfarande tunga beräkningar som behöver slutföras och det är därav att det finns behov av förbättringar.  Syfte. Denna uppsats kommer att utvärdera skillnaderna i prestanda mellan tre olika strålspårningsmetoder i DXR nivå 1.1, nämligen dynamisk strålspårning och två olika former av inline strålspårning. För att ge en bredare utredning på prestandan mellan strålspårningsmetoderna kommer mjuka skuggor att implementeras för att se om drivrutinen kan göra olika optimiseringar (beroende på valet av strålspårningsmetod) på de efterföljande och/eller föregående API anropen. Efter att dessa rörledningar (pipelines) är implementerade kommer prestandatester att utföras med olika grafikkort, scener, och antal ljus som kastar skuggor. Metod. Den vetenskapliga metoden är baserat på ett experimentellt tillvägagångssätt, som kommer innehålla både ett experiment och ett flertal prestandatester. Det experimentella tillvägagångssättet kommer att börja med att utöka en egenskapad DirectX 12 renderare. Utökningen kommer tillföra ny funktionalitet för att kunna hantera strålspårning så att hårda skuggor ska kunna genereras med både dynamisk och de olika formerna av inline strålspårning. Efter det kommer mjuka skuggor att skapas genom att implementera en väletablerad avbrusningsteknik med några modifikationer, vilket kommer att bli tillagt på varje strålspårningssteg. Till slut kommer olika prestandatester att mätas med olika grafikkort, olika antal ljus, och olika scener för att täcka olika scenarion som skulle kunna uppstå i ett spel och/eller i andra liknande applikationer.  Resultat och Slutsatser. De resultat från testerna i detta experiment påvisar att under dessa förutsättningar så är AMD’s grafikkort snabbare på dynamisk strålspårning än på inline strålspårning, samtidigt som Nvidias grafikkort är snabbare på inline strålspårning än på den dynamiska varianten. Ökandet av ljus som kastar skuggor påvisade låg till ingen förändring förutom ett linjärt ökande av exekveringstiden i de flesta testerna. Slutligen så visade det sig även att tillägget av mjuka skuggor (efterföljande och föregående API interaktioner) hade låg till ingen påverkan på valet av strålspårningsmetod.

Background. Digital simulations of physical scenarios are becoming increasingly feasible and driverless vehicles are playing an ever growing part in contemporary mining operations and have the potential to increase productivity and worker safety. Such vehicles require sensors to detect their environments, two of the most common types being LiDAR and RADAR sensors. LiDAR sensors are sensitive to atmospheric sensory pollutants whereas RADAR sensors typically are not, but more susceptible to echoes. As such, digital simulations of such sensors seem a viable alternative to reduce costs and risks in testing new hardware. Objectives. This thesis aims to adapt existing models for CPU simulated LiDAR and RADAR sensors to the GPU as well as to further develop their functionality. These models will then be evaluated against one another according to their performance and scalability. Methods. The stated goals are achieved through the method of literary research, implementation, experimentation, and gathering of data. This data will then be structured, analyzed, and discussed to reach conclusions about the developed software models. Results. The results show that GPU accelerated sensor models have a high overhead cost compared to CPU implementation which hampers performance for low intensity simulations. GPU implementations do however scale more efficiently in many scenarios and achieved speedups of up to 650 times when executed on DXR shaders with heavy workloads than equivalent tests on the CPU. Likewise, low workloads appear unfit for GPU accelerations as the overhead cost of streaming data and instructions between the CPU and GPU can take over twice as long as merely executing the same instructions on the CPU. Conclusions. In conclusion, GPU accelerated ray tracing sensor simulations can be highly efficient compared to CPU implementations when tracing large numbers of rays or simulating many concurrent sensors, but may result in increased execution time if the workload is not high enough to justify the additional overhead cost of CPU-to-GPU communication. / Bakgrund. Digitala simuleringar av fysikaliska fenomen blir mer och mer görbara och självgående fordon spelar en allt större roll i dagens gruvoperationer. Dessa fordon har möjligheten att öka produktiviteten för företaget och säkerheten för arbetarna. Sådana fordon behöver sensorer för att finna sig i sin omgivning och LiDAR- och RADAR-sensorer är två av de vanligaste alternativen. LiDAR-sensorer är känsliga för luftburna störningsmoment medan RADAR-sensorer är jämförelsevis opåverkade men känsliga mot ekon. Med detta i åtanke verkar digitala simuleringar av sådana sensorer vara ett lovande alternativ för att sänka kostnader och risker med att testa ny hårdvara. Syfte. Syftet med arbetet är att översätta befintliga CPU modeller för att simulera LiDAR- och RADAR-sensorer till GPU-programvara såväl som att vidareutveckla deras funktionalitet. Dessa modeller kommer sedan att bli utvärderade gentemot varandra vad gäller deras prestanda och skalbarhet. Metod. De givna målen kommer att uppfyllas via litterära studier, implementation, experimentering och datainsamling. Denna data kommer sedan att bli omstrukturerad, analyserad, diskuteras, och få slutsatser dragna kring sig gällande den mjukvara som utvecklats och framförts. Resultat. Resultaten visar att GPU-accelererade sensormodeller har en hög overheadkostnad jämfört med CPU implementationerna, vilket sänker deras relativa prestanda i lågintensitetssimuleringar. GPU-implementationerna har dock högre skalbarhet i många situationer och kan uppnå resultat upp till 650 gånger snabbare än originalkoden när de istället utförs via DXR-shaders med hög arbetsbörda. Likaså är framstår låg arbetsbörda som opassande scenarion för GPU accelererad mjukvara då exekveringstiden på CPUn kan komma att bli snabbare än overheadkostnaden av att strömma datan och instruktionerna till GPUn. Slutsatser. GPU accelererad raytracing med stora antal strålar eller sensorer som exekveras samtidigt ger upphov till mycket tidseffektiva simuleringar men kan leda till ökad total exekveringstid om arbetsbördan inte blir tillräckligt hög för att rättfärdiga overheadkostnaden av GPU-till-CPU kommunikation.

De nos jours, la problématique du bruit généré par les avions est devenue un point de développement important dans le domaine de l'aéronautique. C'est ainsi que de nombreuses études sont faites dans le domaine et une première approche consiste à modéliser de façon numérique ce bruit de manière à réduire de façon conséquente les coûts lors de la conception. C'est dans ce contexte qu'un motoriste a demandé à l'université de Sherbrooke, et plus particulièrement au groupe d'acoustique de l'Université de Sherbrooke (GAUS), de développer un outil de calcul de la propagation des ondes acoustiques dans les nacelles mais aussi pour l'étude des effets d'installation. Cet outil de prédiction leur permet de réaliser des études afin d'optimiser les traitements acoustiques (« liners »), la géométrie de ces nacelles pour des études portant sur l'intérieur de la nacelle et des études de positionnement des moteurs et de design pour les effets d'installation. L'objectif de ce projet de maîtrise était donc de poursuivre le travail réalisé par [gousset, 2011] sur l'utilisation d'une méthode de lancer de rayons pour l'étude des effets d'installation des moteurs d'avion. L'amélioration du code, sa rapidité, sa fiabilité et sa généralité étaient les objectifs principaux. Le code peut être utilisé avec des traitements acoustiques de surfaces («liners») et peut prendre en compte le phénomène de la diffraction par les arêtes et enfin peut être utilisé pour réaliser des études dans des environnements complexes tels que les nacelles d'avion. Le code développé fonctionne en 3D et procéde en 3 étapes : (1) Calcul des faisceaux initiaux (division d'une sphère, demi-sphère, maillage des surfaces de la géométrie) (2) Propagation des faisceaux dans l'environnement d'étude : calcul de toutes les caractéristiques des rayons convergents (amplitude, phase, nombre de réflexions, ...) (3) Reconstruction du champ de pression en un ou plusieurs points de l'espace à partir de rayons convergents (sommation des contributions de chaque rayon) : sommation cohérente. Le code (GA3DP) permet de prendre en compte les traitements de surface des parois, la directivité de la source, l'atténuation atmosphérique et la diffraction d'ordre 1. Le code a été validé en utilisant différentes méthodes telles que la méthode des sources-images, la méthode d'analyse modale ou encore la méthode des éléments finis de frontière. Un module Matlab a été créé spécialement pour l'étude des effets d'installation et intégré au code existant chez Pratt & Whitney Canada.

International Telemetering Conference Proceedings / October 21, 2002 / Town & Country Hotel and Conference Center, San Diego, California / The objective of the project is to continuously track a handheld device in an office, with centimeter accuracy in the three dimensions. A 3-D ray-tracing algorithm has been developed to simulate the impulse response of the indoor channel. The algorithm can evaluate the impulse response at multiple receiver locations. Non-linear optimization has been used to eliminate the need for multiple runs of simulation. The optimization program also significantly reduces the number of rays launched. The algorithm incorporates bandwidth effects on multipath resolution of the system.

Ghost images are caused by the inter-reflections of light from optical surfaces that have transmittances less than unity. Ghosts can reduce contrast, provide misleading information, and if severe can veil parts of the nominal image. This dissertation develops several methodologies to simulate ghost effects arising from an even number of light reflections between the surfaces of multi-element lens systems. We present an algorithm to generate the ghost layout that is generated by two, four and up to N (even) reflections. For each possible ghost layout, paraxial ray tracing is performed to calculate the locations of the Gaussian cardinal points, the locations and diameters of the ghost entrance and exit pupils, the locations and diameters of the ghost entrance and exit windows, and the ghost chief and marginal ray heights and angles at each surface in the ghost layout. The paraxial ray trace data is used to estimate the fourth order ghost aberration coefficients. Petzval, tangential, and sagittal ghost image surfaces are introduced. Potential ghosts are formed at the intersection points between the ghost image surfaces and the Gaussian nominal image plane. Paraxial radiometric methodology is developed to estimate the ghost irradiance point spread function at the nominal image plane. Contrast reduction by ghosts can cause a reduction in the depth of field, and a simulation model and experimental technique that can be used to measure the depth of field is presented. Finally, ghost simulation examples are provided and discussed.

An investigation is presented into methods of characterising cirrus ice crystals from in-situ light scattering data. A database of scattering patterns from modelled crystals was created using the Ray Tracing with Diffraction on Facets (RTDF) model from the University of Hertfordshire, to which experimental and modelled data was fitted. Experimental data was gathered in the form of scattering patterns from ice analogue crystals with similar optical properties and hexagonal symmetry to ice, yet stable at room temperature. A laboratory rig is described which images scattering patterns from single particles while allowing precise control over the orientation of the particle with respect to the incident beam. Images of scattering patterns were captured and compared to patterns from modelled crystals with similar geometry. Methods for introducing particles en-masse and individually to the Small Ice Detector (SID) instruments are discussed, with particular emphasis on the calibration of the gain of the SID-2 instrument. The variation in gain between detector elements is found to be significant, variable over the life of the detector, and different for different detectors. Fitting was performed by comparison of test scattering patterns (either modelled or experimental) to the reference database. Representation of the two dimensional scattering patterns by asymmetry factor, moment invariants, azimuthal intensity patterns (AIP) and the Fourier transform of the AIP are compared for fitting accuracy. Direct comparison of the AIP is found to be the most accurate method. Increased resolution of the AIP is shown to improve the fitting substantially. Case studies are presented for the fitting of two ice analogue crystals to the modelled database. Fitting accuracy is found to be negatively influenced by small amounts of surface roughness and detail not currently considered by the RTDF model. Fitting of in-situ data gathered by the SID-3 instrument during the HALO 02 campaign at the AIDA cloud chamber in Germany is presented and discussed. Saturation of detector pixels is shown to affect pattern fitting. In-flight operation of the instrument involves the variation of gain of the whole detector (as opposed to individual elements) in order to obtain unsaturated images of both large and small particles.

QC 351 A7 no. 40 / The aim of this study was to determine to what extent a constant spectral response can be obtained for wide-passband interference filters used with wide -angle lenses. We investigated the possibility of using the curvature of the lens surfaces to reduce the shift in the filter passband for large field angles and found that locating the filter on the proper surface will considerably reduce the shift of the passband. Specifically, we determined the distribution of angles of incidence for full aperture pencils incident at several field angles on the second and fourth surfaces of the 90° Geocon IV, the 90° Paxar, and the 125° Pleogon. We then calculated the spectral transmittance of each lens when a wide passband interference filter was located on its second or fourth surface. We also calculated the degree of polarization introduced. From the cases considered, we found that the tracing of an upper and lower marginal (rim) ray at maximum field angle is sufficient to determine the suitability of a surface, the criterion being that, the smaller the angle of incidence at the surface, the better. In addition, we found that, with the filter on the second surface of the Paxar, spectral transmittance changes with field angle were negligible and the modulation due to polarization was about 1 %.

Neste trabalho, modelos de pedestres, utilizados para simular a caracterização da radiopropagação em interiores de edifícios, são reproduzidos, analisados e comparados em diversos ambientes e com diferentes fluxos de pedestres. Estes modelos têm como base o método de traçado de raios (imagens), e se diferenciam em relação ao formato (lâmina, paralelepípedo e cilindro), constantes eletromagnéticas (material absorvente, condutor e dielétrico) e mecanismos de espalhamento de onda eletromagnética (difração, reflexão ou ambos) considerados sobre o pedestre. Para cada um dos modelos, um algoritmo foi criado e detalhado através da apresentação de equações e estrutura dos dados. A análise dos modelos foi realizada em duas etapas de comparação: uma com dados empíricos e outra entre parâmetros de caracterização do canal, como desvanecimentos e dispersão no tempo, obtidos através de simulações com cada tipo de modelo de pedestre. Dentre os vinte e nove modelos ensaiados, os resultados da análise mostraram que o pedestre modelado por um cilindro condutor é aquele que apresenta resultados mais satisfatórios. / In this work, pedestrian models, used to simulate characterization of indoor radio propagation are reproduced, analyzed and compared in different environments with different pedestrian flows. These models are based on the image ray-tracing method, and differs themselves on shape (plate, cylinder and cuboid), electromagnetic constant (absorber, conductive and dielectric materials) and considered spread mechanisms (diffraction, reflection, or both). For each model, an algorithm is created and detailed through the presentation of equations and data structure. The models analysis were made in two comparison steps: one with empirical data and the other with the environment characterization parameters, like fading and time spread, obtained through simulations of each pedestrian model. Within twenty nine models simulation, the results analysis show that the most satisfactory results are given by the model that considers the pedestrian as a conducting cylinder.

The mobility that wireless communication offers to users, added to the ease of installation have increased the demand on such communication systems. However, the main drawback of wireless communication is the degradation of the signal as it travels through the channel due to the different propagation mechanisms the signal undergoes. To minimise the effect of the channel and get the best service, the base stations must be appropriately located within the environment. This requires proper knowledge of the channel characteristics. Ray tracing software is used throughout this work to generate the channel characteristics of an indoor environment. After getting the channel characteristics, a novel cost function is defined based on the path loss values and it is then optimised. Once the optimal base stations’ positions are found, the minimal amount of power required to cover a predefined percentage of the possible receivers’ locations is calculated. On the other hand, a receiver’s position acquiring enough field strength does not necessarily enjoy the service. This depends on the time dispersion parameters values relative to the symbol rate. The time dispersion parameters have always been ignored in the literature while finding the optimal base stations’ locations. Three cost functions that take into consideration both the path loss and rms delay spread, for the first time in the literature, are therefore defined. The cost functions are optimised and their corresponding results are compared. Furthermore, indoor environments have always been considered static which is never realistic. They are subject to continuous changes such as opening doors and windows as well as the presence of people. The first detailed analysis and quantified results of the effect of a dynamic environment on the optimal base stations’ positions and minimal emitted power are presented. It is shown that the optimal base stations’ locations and minimal emitted power are sensitive to such environment changes. The environment changes can also disturb the service for active receivers. Three techniques to overcome the effect of environment changes and bring the disturbed service back to receivers are proposed. The first two techniques rely on increasing the emitted power or changing the antenna polarisation. The third technique is a novel technique that gives the base station the ability to automatically move in various directions within a limited distance. The techniques are tested and their efficiency and limitations are discussed.