Cette thèse traite du problème d’une protection en profondeur qui puisse être assurée par un système d’exploitation. Elle établit la faiblesse des solutions existantes pour l’expression des besoins de sécurité. Les approches supportent en général une seule propriété de sécurité. Nous proposons donc un langage qui permet de formaliser un large ensemble de propriétés de sécurité. Ce langage exprime les activités système directes et transitives. Il permet de formaliser la majorité des propriétés de confidentialité et d’intégrité de la littérature. Il est adapté à l’expression de modèles de protection dynamique. A titre d’exemple, un nouveau modèle dynamique est proposé pour la protection des différents domaines d’usage d’un navigateur Web. Nous définissons une méthode de compilation du langage pour analyser les appels systèmes réalisés par les processus utilisateurs. La compilation repose sur la construction et l’analyse d’un graphe de flux d’information. Nous montrons qu’en pratique la complexité reste faible. Une implantation de ce langage est proposée sous la forme d’un contrôle d’accès mandataire dynamique pour Linux. Une expérimentation à large échelle a été réalisée sur des pots-de-miel à haute interaction. Notre protection a montré son efficacité aussi bien pour les serveurs que les postes client. Il présente des perspectives intéressantes aussi bien pour la protection des systèmes que pour l’analyse de vulnérabilités. Ce travail a contribué au projet SPACLik vainqueur du défi sécurité de l’ANR SEC&SI. / The subject of this thesis is to propose an in-depth protection that can be enforced by the operating system. First, we present that current security solutions are weak in the expression of security. Indeed, most of them support only one security properties. We introduce a language that allows to formalize a large set of security properties. This language expresses directs and transitives system activities. It allows to formalize the majority of integrity and confidentiality security properties introduced in the litterature. Moreover, the language can also expresses dynamic security properties. We introduces a new dynamic security model for the protection of multiple security domains managed by a web browser. We define a method to compil our language. The purpose is to analyze the system call done by the users processes. The compilation process build and analyze an information flow graph. Futhermore, we show that the complexity of our protection solution is low. We propose an implementation of this language as a dynamic mandatory access control for Linux. We experiment it on large scale high interaction honeypots. Our protection shows its efficiency both for clients and servers. Moreover, it presents interesting perspectives for the protection of other systems and for the vulnerability analysis. This work has contributed to the SPACLik project that wins the security contest of the French National Research Agency : ANR SEC&SI.

Dans un contexte de réduction des émissions de polluants et de gaz à effet de serre, mais aussi d’enjeux économiques liés aux ressources en pétrole brut, la nécessité de se diversifier énergétiquement a imposé l’apparition et l’utilisation dans des quantités sans cesse croissantes de carburants de substitution. Ces carburants provenant de nombreuses sources primaires, produits à partir de diverses techniques, possèdent des propriétés physico-chimiques différentes. Les effets de chaque propriété sur le comportement du moteur Diesel (émissions, rendement, etc.) ne sont pas encore compris de manière détaillée. La présente thèse porte sur une étude expérimentale de l’influence des propriétés physico-chimiques des carburants sur les processus physiques se déroulant dans la chambre de combustion pour des applications sur moteurs Diesel. L’originalité de l’approche réside par la définition d’une matrice de carburants (20 carburants au total) dont les propriétés varient de manière indépendante. De plus, les différentes étapes de l’injection Diesel ont été considérées. On s’intéresse par la mise en oeuvre de huit techniques de mesures avancées, à la caractérisation de l’hydraulique dans les orifices de l’injecteur, de la structure du spray, de son atomisation, sa vaporisation et enfin la combustion du mélange résultant. Ces expériences ont été conduites dans des cellules adaptées et sur moteur à accès optiques permettant de reproduire des conditions de fonctionnement représentatives. L’analyse porte essentiellement sur la phase quasi-stationnaire de l’injection pendant laquelle les résultats expérimentaux sont corrélés à des modèles basés sur la physique issus de la littérature et de nouvelles corrélations empiriques sont proposées. / In a context of pollutant emission and green house gas reduction, but also economic issues related to crude oil ressources, the necessity for energy diversification has imposed the emergence and the use in growing quantities of substitute fuels. These fuels derived from numerous feedstocks, produced by several manufacturing processes, have different physico-chemical properties. The effects of each individual property on the Diesel engine behavior (emissions, efficiency, etc) are not well fully understood yet. The present thesis focuses on an experimental study about the influence of physico-chemical properties of the fuels on the processes occuring in the combustion chamber for Diesel engine application. The originality of the approach concerns the definition of a fuel matrix (a total of 20 fuels) for which properties are independently varied. Moreover every stages of the Diesel injection are considered. The interest focuses by the setting of eight advance measurement technics on the caracterization of the flow into the injector orifices, the spray structure and atomization, the spray vaporization and finally the Diesel combustion. Experiments were conducted in dedicated vessels and in an optical access engine allowing to reproduce representive operating conditions. The analysis mainly focuses on the quasi-stationnary phase of the injection event and experimental results are scaled to physic-based models and new empirical correlation are proposed.

The structure of the Finney Random Close Packing (RCP) of equal spheres has been analysed, together with the influence which such structure exerts over the capillary pressure characteristics of geometrically similar sphere packings. The analysis is centred on the simplicial, or Delaunay cell, which is an irregular tetrahedron with apices defined by four immediate neighbour sphere-centres. In terms of using RCP as a model porous medium, an individual simplicial cell is equivalent to an individual pore. A number of measured pore-size distribution parameters are presented for the Finney packing, from which it is shown from first principles that drainage-imbibition hysteresis is not an intrinsic property of the individual pore. The nature and degree of randomness which characterises the Finney packing is evaluated on two levels. First, by classifying edgelengths as either short or long, seven mutually exclusive cell classes are defined. Using the binomial theorem it is shown that cells (pores) are not random on the level of the individual cell. There are less of the extreme cells (with 6 long edges, or with 6 short edges) and more of the bland cells (with 3 short and 3 long edges) in the Finney packing than predicted on the basis of simple random expectations. Second, the distribution of cell classes within the packing is shown to be essentially homogeneously random. Evidence for extremely slight cell class clustering is found. The drainage and imbibition processes within the packing are simulated using pore-level algorithms. The algorithms utilise both the Haines' insphere approximation and the MS-P approximation for critical drainage meniscus curvature, and the cell cavity insphere radius approximation for critical imbibition meniscus curvature. Good agreement with experimental data is obtained, and the results confirm that drainage-imbibition hysteresis is a direct consequence of the connectivity between cells (pores), and is not an intrinsic property of the individual pore. Finally, the drainage and imbibition algorithms are adapted to emulate percolation theory models. The results prove that the classical bond problem of percolation theory does not adequately describe the drainage process for RCP, and that the classical site problem does not adequately describe the imbibition process for RCP.

The influence of pre-metallisation surface preparation on the structural, chemical, and electrical properties of metal-nGaN interfaces has been investigated by X-ray Photoemission Spectroscopy (XPS), current-voltage measurement (I-V) and cross section Transmission Electron Microscopy (TEM). XPS analysis showed that the three GaN substrate treatments investigated, ex-situ HF etch, in-situ anneal in Ultra-High-Vacuum (UHV), and in-situ Ga reflux cleaning in UHV result in surfaces increasingly free of contaminants. Additionally, the three treatments are found to induce increasingly larger upward band bending. Ag-nGaN contacts formed after Ga reflux cleaning exhibit a Schottky barrier height of 0.80 eV and an ideality factor of 1.56, as determined by I-V.XPS and TEM characterisation of Au-nGaN formed after the three pre-metallisation surface treatments show that HF etching and UHV annealing produce abrupt, well-defined interfaces. Conversely, GaN substrate cleaning in a Ga flux results in Au/GaN intermixing. I-V characterisation of Au-nGaN contacts yield a Schottky barrier height of 1.25 eV with very low ideality factor and very good contact uniformity for the pre-metallisation UHV anneal while the Ga reflux cleaning result in a much lower barrier (0.85 eV), with poor ideality and uniformity. I-V and XPS results suggest a high density of acceptor states at the surface, which is further enhanced by UHV annealing. The mechanisms of Ga-nGaN, Ag-nGaN and Au-nGaN Schottky barrier formation are discussed in the context of the Metal-Induced Gap States model (MIGS) Unified Defect Model (UDM) and Cowley-Sze model.

No description available.

No description available.

M.Sc. (Experimental Physics) / Please refer to full text to view abstract

M.Ing. / Grade 350W structural steel was introduced into the South African steel market to replace its predecessor 300W structural steel. 350W structural steel has a 50 MPa increase in its minimum yield strength. Currently, little or no experimental data exists that describes the behaviour of full scale 350W structural sections. No extensive comparisons between 350W and 300W, as to their structural behaviour, exist. The main aim of this investigation was firstly to determine if the minimum yield strength was fully realised throughout typical sections and if the increase in yield strength changed the yielding behaviour of the material. Secondly, the effect that the increased yield strength has on the structural resistance of axial compression members. The investigation largely consisted of an evaluation of firstly the material and then secondly the structural behaviour of full scale member sections as axial members. It was found that all the sections tested complied with the minimum specified yield strength of 350 MPa. In general, 350W hot rolled structural steel yields more gradually than 300W hot rolled structural steel. This increase is small and seems to be of little significance as far as the overall behaviour is concerned. In general, the increase in minimum yield strength from 300 MPa to 350 MPa has realised an increase in critical buckling strength well into the medium slenderness ratio region and is beneficial for members of short and intermediate lengths. The data seems to indicate that SANS 10162-1:2005 can be used with confidence when designing axial compression members in 350W hot rolled structural steel.

M.Sc. (Physics) / Please refer to full text to view abstract

This thesis describes a study of the clustering properties of galaxies and clusters of galaxies as measured by the power spectrum (P(k)) and the counts in cells statistic. The samples used are the optical Stromlo-APM galaxy survey, the APM cluster survey and the IRAS 1.2Jy, QDOT and PSCz surveys. Throughout, N-body simulations, for a variety of cosmological models, are used to test methods and to supplement analytic error estimates. For the Stromlo-APM sample the amplitude of the power spectrum is dependent on galaxy morphology. Early-type galaxies show a higher clustering amplitude than late-type galaxies by a factor of ~ 1.8. There is also tentative evidence for some dependence of the clustering amplitude on galaxy luminosity. The parameter Ω0.6/b is estimated via a comparison with the real-space power spectrum of the two-dimensional APM galaxy survey. For APM clusters the power spectrum is measured to very small wave numbers, with a possible detection of the expected turn-over. The results are inconsistent with the standard cold dark matter model. The shape of P( k) for clusters is approximately the same as that for Stromlo-APM galaxies but amplified by a factor of ~ 3.5. The power spectrum of the QDOT sample depends sensitively on the galaxy weighting scheme, probably due the manner in which the region of the Hercules supercluster is sampled. A best estimate of the power spectrum of IRAS galaxies is computed by combining the IRAS l.2Jy and QDOT samples. The PSCz galaxy power spectrum is also computed. The PSCz galaxies have a clustering amplitude twice that of optical galaxies. A similar result is found from a joint counts in cells analysis. Redshift-space distortions in the PSCz sample are analysed using a spherical harmonic decomposition of the density field. The value of Ω0.6/b = 1 is ruled out by this analysis at the 2σ significance level.