La maîtrise du comportement en vol d’un projectile est, en partie, conditionnée par la connaissance des coefficients associés à chaque effort aérodynamique. Différents outils sont utilisés dans l’industrie, tels que les codes numériques aérodynamiques ou les essais en soufflerie, afin d’obtenir une première estimation des coefficients en phase d’avant-projet. Il est ensuite nécessaire de vérifier la valeur des coefficients et de valider le comportement du projectile en vol au moyen de tirs instrumentés. Un outil automatisé est donc proposé afin d’identifier les coefficients aérodynamiques d’un projectile à partir des mesures issues d’un vol. La technique d’identification est pensée pour être applicable à une gamme la plus large de projectiles. Elle introduit un problème d’optimisation non linéaire en dimension finie. La fonctionnelle du problème contient deux termes : un terme d’écart entre les paramètres d’état et les mesures, de sorte à s’approcher au mieux des mesures et à les relaxer, et un terme de pénalisation prenant en compte les équations de la mécanique du vol. L’outil proposé est testé, pour un projectile de type flèche, avec des données simulées ou avec des données issues de tirs. Il permet l’identification des coefficients aérodynamiques recherchés. L’algorithme est robuste face au bruit et permet également la reconstruction d’une trajectoire débruitée. / The control of the flight behaviour of a projectile partly depends on the knowledge of the coefficients associated with each aerodynamic loading. Different tools are used in the industry, such as numerical aerodynamic codes or wind tests in order to obtain a first estimate of the coefficients during the stage of pilot study. It is then necessary to verify the value of the coefficients and to validate the behaviour of the projectile thanks to scored fires. An automated tool is consequently proposed in order to identify the aerodynamic coefficients of a projectile from flight data. The identification technique is designed so as to be applicable to the widest range of projectiles. It presents a nonlinear optimization problem in finite dimension. The functional of the problem contains two terms : the first one is a gap between the state parameters and the measurements, in order to approach the measurements at best and to relax them, and the second one is a penalization term which takes the flight mechanics equations into account. The proposed tool is tested, for a Kinectic Energy projectile, with simulated data or real flight data. It enables the identification of the searched out aerodynamic coefficients. The algorithm is robust in a noisy environment and also enables the reconstruction of a denoised trajectory.

Yes / Literature data on solubilities of adamantane in organic solvents have been used to obtain properties, or descriptors, of adamantane. There is much less data on substituted adamantanes but we have been able to obtain descriptors for some 40 substituted adamantanes. These descriptors can then be used to estimate a wide range of physicochemical, environmental and other properties of the adamantanes. For the first time, the water-solvent partition coefficient and the gas-solvent partition coefficient into a large range of solvents, can be estimated, the latter being equivalent to Henry's Law constants. A variety of other important properties can also be estimated. These include vapor pressures, enthalpies of vaporization and sublimation, partitions from air and from blood into biological tissues, and skin permeability from water. The descriptors themselves are not exceptional. Adamantane itself has a rather low dipolarity, zero hydrogen bond acidity and a very low hydrogen bond basicity, in common with other multicyclic aliphatic compounds. These lead to adamantane being a very hydrophobic compound, as is evident from our estimated water-octanol partition coefficient.

L'effet de la rotation aux ancrages est connu comme la principale raison engendrant la fatigue des haubans. Malgré la grande variation de cet effet d'un hauban à l'autre, tous les haubans des ponts haubanés sont généralement conçus par une même règle proposée dans plusieurs normes et recommandations. Cette règle semble excessive pour le cas des haubans dont l'effet de la rotation n'est pas important (tablier très rigide, espacement des haubans petit, etc.), mais probablement pas suffisamment sécuritaire pour le cas des haubans dont cet effet devient critique (tablier très flexible, grand espacement des haubans, haubans très longs, etc.). Afin d'optimiser la conception des ponts haubanés, il est donc intéressant d'estimer la demande de la fatigue qui consiste elle-même à estimer la demande de rotation aux ancrages des haubans. Ce document présente un projet de recherche visant à évaluer la rotation aux ancrages des haubans sous les charges de trafic et de vent. Un indice d'amplification de fatigue qui relie la rigidité du tablier, l'espacement des haubans et leur rigidité verticale à l'amplitude de rotation aux ancrages bas des haubans sous les charges routières est développé. En se basant sur des calculs analytiques et numériques (méthode des éléments finis), des équations sont proposées pour prédire l'amplitude de rotation aux ancrages bas des haubans sous charge de fatigue. Il est démontré que la contrainte axiale maximale due aux charges d'exploitation n'est pas directement liée à la rotation aux ancrages bas des haubans. En combinant avec un autre indice, \Omega, il est possible de relier la contrainte axiale maximale dans le hauban et l'amplitude de rotation à son ancrage bas. Afin d'évaluer la rotation des haubans sous les charges de vent, des essais en soufflerie ont d'abord été réalisés pour évaluer les coefficients aérodynamiques d'un cylindre incliné (représentant un hauban) dans un écoulement. Le cylindre a été positionné à différents angles d'inclinaisons, par rapport à la verticale et l'horizontal, relatifs au vent. Deux balances dynamiques à trois composantes ont été conçues pour mesurer les forces aérodynamiques. Les trois coefficients des forces aérodynamiques sont normalisés en se basant sur la vitesse de l'écoulement. Les résultats montrent que le coefficient de traînée et la résultante du coefficient de portance ainsi que celui de côté peuvent être exprimés par une fonction empirique en fonction de l'angle d'incidence. Le coefficient de portance et celui de côté restent cependant des fonctions d'angles vertical et horizontal, et ne peuvent pas s'exprimer en fonction de l'angle d'incidence. Il a été observé expérimentalement que le principe d'indépendance devient inexact pour les angles d'incidence supérieurs à 40◦. Les résultats expérimentaux ont ensuite été appliqués pour déterminer les charges de vent sur les haubans. La rotation aux ancrages des haubans due aux charges de vent comprend deux parties: l'une est due à la variation de la vitesse moyenne du vent et l'autre, à ses fluctuations. L'effet de la variation de la vitesse moyenne a été déterminée par des analyses statiques non-linéaires des haubans sous charges de vent moyen correspondant aux moyennes horaires de vitesse de vent extraite à partir des données enregistrées à une station météorologique. L'effet des fluctuations du vent a été déterminée par des analyses dynamiques linéaires des haubans en utilisant la méthode spectrale. Seulement la fluctuation longitudinale de la vitesse du vent a été considérée dans cette étude. Ces rotations ont enfin été comparées à la rotation des haubans sous les charges routières. Les résultats de cette étude permettent de constater la sensibilité des haubans à chaque source de fatigue.

Dissertation submitted in fulfilment of the requirements for the Masters Degree in Technology: Chemistry, Durban University of Technology, 2009. / The activity coefficients at infinite dilution ( γ13 ) were calculated for alkanes, alkenes, cycloalkanes, alkynes, ketones, alcohols and aromatic compounds from gas liquid chromatography (glc) measurements at three temperatures (303.15 and 313.15 and 323.15) K. The γ13 values were calculated from the retention data for two ionic liquids (ILs) with the same cation but different anions. The ionic liquids: [methyltrioctylammonium bis(trifluoromethylsulfonyl)imide and methyltrioctylammonium thiosalicylate] were used as the stationary phase. The γ13 data for methyltrioctylammonium thiosalicylate were higher than for methyltrioctylammonium bis(trifluoromethylsulfonyl)imide. For each temperature and each ionic liquid, γ13 values were determined for two columns and the average γ13 values were used for the calculation of the partial molar excess enthalpies at infinite dilution, H E 1 Δ , using the Gibbs-Helmholtz equation. Graphs of In γ13 vs 1/T were plotted to determine H E 1 Δ at T = 303.15 K. The selectivity, S12 , value for the hexane/benzene separation was calculated from the ratio of the avarage γ13 values to determine the suitability of the ILs as an entrainer for extractive distillation in the separation of aromatic and aliphatic compounds.The S12 values were also compared to the literature values for other ionic liquids. It was found that the longer chain alkyl group attached to the cation or anion gave lower S12 values. Both ILs have S12 values greater than one and in theory can be used as an entrainer for the hexane/ benzene separation. The capacities at infinite dilutions, k1 , were also calculated from the inverse of the avarage γ13 values for each IL at each temperature. / National Research Fund; Postgraduate Development and Support Directorate.

PART I: This work is part of an investigation to determine activity coefficients at infinite dilution (1'73) of hydrocarbons dissolved in the industrially important polar solvent tetrahydrothiophene-l, l-dioxide (sulfolane), by medium pressure gas liquid chromatography (g.l.c.). In this work the activity coefficients at infinite dilution for a series of l-alkenes (C6-CS) , l-alkynes (C6-CS) , and cycloalkanes (C7 and Cs) have been measured in the polar solvent, sulfolane, at 303.15 K and 313.15 K. The activity coefficients of some of the solutes discussed in this work would be difficult to determine by any other method because of their low solubility in sulfolane. The mixed second virial coefficients used in this work were determined assuming the principle of corresponding states, the Hudson and McCoubrey combining rules for T~2' the Lorentz rule for V~2' and the McGlashan-Potter equation. PART II: Unrefined sugar contains organic colour material originating In the sugarcane or formed during the extraction and purification processes. Sugar colour must be within the limits of acceptability for direct or indirect consumption. In this work, a cost effective technique to separate colourants from sugar through a sucrose packed medium pressure chromatographic column was investigated. Three dimensional perspective plots of wavelength/absorbance/time were developed to provide insight into the nature of the sugar colourants and to provide a means of investigating various decolourisation systems. In addition to the above experiment a procedure was developed to remove colourant species from unrefined sugar samples and from samples taken during the refining process for chemical analysis. In this work only one technique - gas chromatography-mass spectroscopy was used to identify the species. / Thesis (M.Sc.)-University of Natal, Durban, 1996.

The specific goal of the investigation was the measurement, as a function of temperature, of the self-diffusion coefficients of Pb210 and Cl36 in PbCl2-KCl compositions in the region of the first compound, and to calculate from these data the activation energy necessary for the diffusion of these ions.

Few studies have been conducted on the effectiveness of phytoremediation of pharmaceutical compounds, although the persistent and non-acutely toxic nature of many of these compounds in today's water bodies may yield an ideal application for this practice. To quantify the potential effectiveness of plant uptake, kinetic and proportional bioconcentration factors (BCFk, and BCFp, respectively) in nanograms (ng) carbamazepine and amiodarone per gram (g) wet weight plant tissue for Lemna spp. were determined utilizing a 14-day continuous flow-through study. Samples were analyzed using isotope dilution liquid chromatography-tandem mass spectrometry (ID-LC-MS/MS) running in positive ion mode. Kinetic BCF was estimated at 0.538, while proportional BCF was estimated at 0.485. Kinetic BCF for the amiodarone study was estimated at 23.033, whereas proportional BCF was estimated at 41.340. Possible contamination of the C18 column and peristaltic pump failure may have impacted uptake results. In light of variability and current lack of research in the field, this work should be considered exploratory rather than conclusive.

We evaluate the security of human voice password databases from an information theoretical point of view. More specifically, we provide a theoretical estimation on the amount of entropy in human voice when processed using the conventional GMM-UBM technologies and the MFCCs as the acoustic features. The theoretical estimation gives rise to a methodology for analyzing the security level in a corpus of human voice. That is, given a database containing speech signals, we provide a method for estimating the relative entropy (Kullback-Leibler divergence) of the database thereby establishing the security level of the speaker verification system. To demonstrate this, we analyze the YOHO database, a corpus of voice samples collected from 138 speakers and show that the amount of entropy extracted is less than 14-bits. We also present a practical attack that succeeds in impersonating the voice of any speaker within the corpus with a 98% success probability with as little as 9 trials. The attack will still succeed with a rate of 62.50% if 4 attempts are permitted. Further, based on the same attack rationale, we mount an attack on the ALIZE speaker verification system. We show through experimentation that the attacker can impersonate any user in the database of 69 people with about 25% success rate with only 5 trials. The success rate can achieve more than 50% by increasing the allowed authentication attempts to 20. Finally, when the practical attack is cast in terms of an entropy metric, we find that the theoretical entropy estimate almost perfectly predicts the success rate of the practical attack, giving further credence to the theoretical model and the associated entropy estimation technique.

The positive temperature coefficient (PTC) effect in conductive polymer composites (CPC) are still poorly understood with the thermal expansion of the polymer matrix accepted as the main cause. This thesis aims to study a model system able to explain the effect of the filler size and shape on the PTC behaviour of CPCs. Silver coated glass spheres and flakes are used as conductive fillers due to the ease in controlling uniform size and shape. In a controlled system it was demonstrated that the PTC intensity increases with increasing filler size and with decreasing filler content, both for conductive fillers. Combinations of different conductive fillers were investigated to explore the possibility to obtain both low percolation thresholds and high PTC intensities. Model systems in which at least one of the two conductive fillers is of relatively homogenous size and shape were used to facilitate unravelling some of the complicated relationships between (mixed) conductive fillers and the PTC effect. The PTC intensity of mixed fillers composites were dominated by the filler with the lowest PTC intensity, even at very low volume fractions. The PTC intensity was not only influenced by the conductive particle size but also by its size distribution. The effect of difference in linear coefficient of thermal expansion (CTE) of conductive fillers and polymer matrix based on a change in filler core on PTC behaviour was investigated. Damage to the particles due to the poor adhesion between the silver coating and the PMMA bead lead to the composite behaving like mixed filler composite. Hybrid polymers filled with silver coated glass flakes was also examined in order to enhance the PTC intensity. The PTC intensity of the composite increased with increasing PPE content but the negative temperature coefficient (NTC) effect was observed in all the composites.

We employ some of the modern tools of economic theory and time series econometrics to consider a number of economic problems. The communication and coordination problems we study are relevant in accounting, business, economics and finance. The thesis begins by examining the behaviour of people and organisations, who are supposed to share a common goal. Then it considers the equilibriating mechanisms of behaviour by groups of economic agents, who usually have conflicting interests. We apply the tools of non-cooperative game theory, which constitutes a large part of modern economic theory. In the sequel, we address the question of why people behave the way they do in their economic a↵airs. Peoples' economic behaviour is mirrored in the aggregates of macroeconomics. We propose a Time Varying Autoregressive model to study the movements in the five main macroeconomic variables. The methods come from standard Time Series Analysis, but we do introduce some innovative time series techniques. Finally, we conduct an empirical investigation of the movements in one of the five main macroeconomic variables, the rate of inflation. Among the econometric tools employed are standard Autoregressive models (AR), Autoregressive Distributed Lag models (ADL) and the more recent Generalised Autoregressive Conditional Heteroskedasticity (GARCH) methodology.