No description available.

Very Low Frequency (VLF) radio waves propagate within the Earth-ionosphere waveguide with very little attenuation. Modifications of the waveguide geometry affect the propagation conditions, and hence, the amplitude and phase of VLF signals. Changes in the ionosphere, such as the presence of the D-region during the day, or the precipitation of energetic particles, are the main causes of this modification. Using narrowband receivers monitoring remote VLF transmitters, the amplitude and phase of these signals are recorded. A multivariate data analysis technique, Principal Component Analysis (PCA), is applied to the data in order to determine parameters such as seasonal and diurnal changes which affect the variation of these signals. Data was then analysed for effects from extragalactic gamma ray bursts, terrestrial gamma ray flashes and solar flares. Only X-rays from solar flares were shown to have an appreciable affect on ionospheric propagation. / Thesis (M.Sc.)-University of KwaZulu-Natal, Westville, 2009.

Cette thèse de doctorat est consacrée à la modélisation des écoulements turbulents réactifs pré-mélangées. Une analyse des flammes monodimensionnelles est tout d'abord conduite pour examiner les caractéristiques propagatives du front de flamme analytiquement et numériquement. Nous introduisons ensuite, dans le régime des flammes plissées(flammelettes), une nouvelle définition de l'épaisseur de flamme qui peut être différente de l'épaisseur caractéristique de la flamme laminaire. Une forme généralisée de la fonction de densité probabilité (PDF) présumée est alors proposée pour tenir compte des écarts possibles avec la limite de flamme mince. Les paramètres qui caractérisent la forme de cette PDF dépendent du rapport de l'échelle de longueur de Kolmogorov à l'épaisseur dela flamme laminaire. Le cadre de travail correspondant peut être associé à une description de type couche limite appliquée à la flamme turbulente pour laquelle les épaisseurs des différentes sous-couches sont déterminées à partir de la seule connaissance des nombres de Reynolds turbulent Re, et de Karlovitz Ka. Cette description conduit à une proposition de modélisation (i) pour le taux de conversion des réactifs en produits ainsi que (ii) pour le flux turbulent scalaire. Les valeurs des caractéristiques chimiques s'appuient sur des calculs de flammes monodimensionnelles utilisant le logiciel Mathematica TM ou reposant sur l'emploi d'une description détaillée de la cinétique chimique.Les modèles proposés ici sont finalement implantés dans le logiciel de calcul numérique Code-Saturne et validés sur des configurations expérimentales de référence : flammes en V turbulentes et flamme stabilisée dans un écoulement divergent. Les simulations numériques de ces deux configurations conduisent à des résultats satisfaisants. / The present study is devoted to the modeling of turbulent reactive flows. In a first step of the analysis, one-dimensional flames are considered to scrutinize analytically and numerically the characteristics of the propagating flame front. Then, we introduce a new definition for the flamelet thickness that may differ from the laminar flame thickness. Accordingly a generalized and consistent form of the presumed probability density function(PDF) is proposed. The parameters of this PDF depend on the ratio of the Kolmo-gorov length scale to laminar flame thickness. The corresponding developments can be associated with a boundary layer description applied to the turbulent flame in which the thicknesses of the different sub-layers are determined from the sole knowledge of the turbulent Reynolds number Re and Karlovitz number Ka. This description leads to amodeling proposal (i) for the mean reaction rate as well as (ii) for the turbulent scalar flux. The values of all chemical characteristics are calculated from one-dimensional flame susing either the MathematicaTM software or detailed chemical kinetics computations.Numerical simulations are finally performed with the CFD code Code-Saturne and the model is validated through comparisons with reference experimental configurations :turbulent V-shaped flames and impinging flames. Satisfactory results are obtained for these two distinct configurations.

Les vibrations dans les sols constituent un problème environnemental de plus en plus important. Cette étude adresse plusieurs aspects des vibrations du sol. Les conditions locales du site et l'amplification des ondes sismiques représentent un sujet largement étudié en sismologie et en ingénierie sismique. Bien que la plupart des études soient consacrées à la sismologie d'ingénierie et à l'ingénierie sismique, des approches similaires pour la classification des sols et l'amplification du site n'ont pas encore été pleinement établies dans le domaine des vibrations urbaines.D'abord, une stratégie d’optimisation pour la méthode des couches absorbantes a été développée afin d'améliorer la précision des modèles par éléments finis. La conception de couches absorbantes simples en éléments finis par l'annulation de la partie réelle des ondes réfléchies dans le domaine du nombre complexe s'est révélée efficace lorsqu'elle est couplée à la réduction de la rigidité de la couche absorbante. Une réduction excessive de la rigidité ainsi que l'augmentation excessive du facteur de rigidité dans la matrice d'atténuation ont permis une grande réduction de la taille de la couche absorbante et par suite la conception d'une couche absorbante moins coûteuse.Ensuite, une partie importante du travail a été consacrée à la dérivation d'un nouvel ensemble de paramètres du type gradient de vitesse qui contrôle le transfert des vibrations à travers des sols multicouches. L'absence d'une approche compréhensive et bien structurée pour la prédiction et la classification des sites pour les problèmes de vibration laisse le problème large et compliqué. Différents sites ayant différentes propriétés mécaniques et géométriques ont été examinés à l'aide des éléments finis. Les réponses des sites ont été formulées dans des approches spectrales et temporelles simples. Les paramètres nouvellement dérivées ainsi que les lois spectrales servent de moyen de classification des sols multicouches pour les problèmes de vibrations et peuvent même être utilisées à des fins de conception.D’ailleurs, un outil d'intelligence artificielle pour prédire la réponse du sol en utilisant les paramètres précédemment dérivées associées aux propriétés géométriques de la couche de surface a été développé. L'outil des réseaux neurone a été utilisé pour analyser les effets paramétriques des paramètres de gradient de vitesse par rapport à la profondeur de la couche de surface. Des conclusions importantes ont été tirées de l'analyse concernant les propriétés mécaniques et géométriques des couches multiples et leurs effets variantes avec la distance de la source.Enfin, les enregistrements de sources multiples ont été étudiés en les comparants aux réponses spectrales des différents sites définis dans les sections précédentes. Le taux d'appariement entre le contenu spectral d'une source particulière et un site donné sert à évaluer l’aléa vibratoire causé par cette source au site correspondant. L'évaluation des risques de vibrations conduit à un lien de classification entre les sources d'une part et les sites caractérisés par des paramètres à gradient de vitesse d'autre part / Ground vibration is an increasingly important environmental problem. This study investigates multiple aspects of ground vibration. Local site conditions and the related amplification of seismic waves represent a widely studied topic in seismology and earthquake engineering. While most of the studies are dedicated to engineering seismology and earthquake engineering, similar approaches for soil classification and site amplification have not been yet fully established in the field of urban vibrations.First an improvement strategy for absorbing layer method was developed in order to enhance precision of the FEM models. The design of simple absorbing layers in FEM through the nullification of the real part of reflected waves in the complex number domain proved to be efficient when coupled with the stiffness reduction of the absorbing layer. Excessive reduction of the stiffness along with the excessive increase of the stiffness factor in the attenuation matrix enabled a large reduction in the size of the absorbing layer and therefore the design of an inexpensive absorbing layer.Afterwards an important part of the work was dedicated to the derivation of a new set of parameters of the velocity-gradient type that controls the vibration transfer through multilayered soil. The absence of a well-structured comprehensive approach for prediction and site classification for vibration problems leaves the problem broad and complicated. Different sites with different mechanical and geometrical properties were examined using FEM and their surface response was studied. Sites responses were formulated in simple time domain and spectral approaches. The newly derived proxies along with the spectral laws serve as a classification mean for multilayered soils in the vibration problem and may even be used for design purposes.An artificial intelligence tool for predicting soil response using the previously derived proxies coupled with the geometrical properties of the surface layer was later developed. The neural networks tool was used to analyze the parametric effects of the velocity-gradient proxies versus that of the surface layer’s depth. Important conclusions were derived from the analysis regarding the mechanical and geometrical properties of multiple layers and their varying effects with distance from the source.Finally multiple sources recordings were studied through comparing them with the spectral responses of different sites defined in the previous sections. The rate of matching between the spectral content of a particular source and a given site serves as a mean to assess the vibration hazard caused by this source to the corresponding site. The vibration hazard assessment leads to a classification link between sources in one hand and sites characterized by velocity-gradient proxies on the other hand

Cette thèse présente une étude sur la mesure des réseaux sociaux en ligne avec un accent particulier sur Twitter qui est l'un des plus grands réseaux sociaux. Twitter utilise exclusivement des liens dirigés entre les comptes. Cela rend le graphe social de Twitter beaucoup plus proche que Facebok du graphe social représentant les communications dans la vie réelle. Par conséquent, la compréhension de la structure du graphe social de Twitter et de la manière dont les informations se propagent dans le graphe est intéressant non seulement pour les informaticiens, mais aussi pour les chercheurs dans d'autres domaines, tels que la sociologie. Cependant, on sait peu de choses sur la propagation de l'information sur Twitter. / This thesis presents the measurement study of Online Social Networks focusing on Twitter. Twitter is one of the largest social networks using exclusively directed links among accounts. This makes the Twitter social graph much closer to the social graph supporting real life communications than, for instance, Facebook. Therefore, understanding the structure of the Twitter social graph and the way information propagates through it is interesting not only for computer scientists, but also for researchers in other fields, such as sociologists. However, littles is known about the information propagation in Twitter. In the first part, we present an in-depth study of the macroscopic structure of the Twitter social graph. In the second part, we study the propagation of the news media articles shared on Twitter. In the third part we present an experimental study of graph sampling.

The system of sparse gridpoints was used to propagate uncertainty forward in time through orbital mechanics simulations. Propagation of initial uncertainty through a nonlinear dynamic model is examined in regards to the uncertainty of orbit estimation. The necessary underlying mechanics of orbital mechanics, probability, and nonlinear estimation theory are reviewed to allow greater understanding of the problem. The sparse grid method itself and its implementation is covered in detail, along with the necessary properties and how to best it to a given problem based on inputs and desired outputs. Three test cases were run in the form of a restricted two-body problem, a perturbed two-body problem, and a three-body problem in which the orbiting body is positioned at a Lagrange point. It is shown that the sparse grid method shows sufficient accuracy for all mean calculations in the given problems and that higher accuracy levels allow for accurate estimation of higher moments such as the covariance.

No description available.

Master of Science / Department of Computing and Information Sciences / Rodney R. Howell / Nowadays, there is a rapid increase in the complexity of large automotive and control systems because of the integration of external software modules in them. Many of these systems are based on sampled-data control theory. And because of the different timing constraints of individual modules, each module has a different sampling rate. Typically, these systems operate with periodic task sequences and the information flows between the tasks. Generally the information propagates from tasks operating at one period to tasks operating at different periods. When this happens, unusually long information propagation delays can occur which can be disastrous because the system cannot respond to the changes until this delay has been elapsed. Although for arbitrary set of task sets the delays can be very long, Howell and Mizuno (2010) considers a set of task sequences with special constraints and some very useful bounds are derived for the worst case occurrences of them. Howell and Mizuno (2011) has laid out algorithms that compute the delays for certain special cases of task sequences considered in Howell and Mizuno (2010). The purpose of this project is to understand and implement the algorithms from Howell and Mizuno (2011). The implementation is done so that it avoids the manual computation of the delays and helps in better understanding the ideas presented in Howell and Mizuno (2011). The application can be tested against any valid input that meets our assumptions, and it constructs a schedule that exhibits the worst case behavior and from the schedule it computes the worst case information propagation delays.

Call number: LD2668 .T4 1979 A432 / Master of Science

Outdoor to indoor mobile coverage is evaluated for differentfrequencies in two scenarios, a single building scenario and a cityenvironment. A new model for outdoor to indoor propagation issuggested, connecting existing, highly detailed indoor and outdoorray-tracing propagation models. The model is compared to previous,site specific as well as statistical, propagation models. Resultsconclude that the new model gives higher path gain for edge users inthe single building scenario, whereas results from the city scenarioare inconclusive. Furthermore, results from the single buildingscenario suggest that indoor coverage is possible at 5 GHz and belowfor most buildings, whereas for the city scenario indoor coverage at5 GHz will be possible only for buildings without metally coatedwindows. Achieving indoor coverage at 30 GHz is highly problematicfor all cases, and it is concluded that indoor base stations arenecessary if frequencies of 10 GHz and above are to be used infuture mobile networks. In addition, an indoor analysis is made to verify existing lossper meter indoor models. It is concluded that such models are oftenoptimistic, although their assumption of log-normal shadowing remainvalid, at least for closed offices. Furthermore, the assumption ofloss as a linear function of distance might be unfeasible for higherfrequencies, where a breakpoint in the linear model was observed ata distance of roughly 10 meters.