Darbe analizuojamas metaduomenų saugyklų kūrimas pritaikant veiklos modeliavimo metodo metamodelį. Apžvelgtos metaduomenų saugyklų ir veiklos modeliavimo metodų bei notacijų BPMN, ARIS, GRAI tinklo, UEML savybės, jų pritaikymo metaduomenų saugykloms projektuoti galimybės. Atrinktas tinkamiausias veiklos modeliavimo metodas ARIS, kuris apima daugiausiai organizacijos veiklos aspektų. Sukurta metaduomenų saugykla panaudojant ARIS veiklos modeliavimo metodo metamodelį, kuri pritaikyta įmonės UAB „Klaipėdos Konteinerių Terminalas“ duomenų srautams valdyti ir integruota su įmonės naudojama informacine sistema. Šios saugyklos pagalba metaduomenys, išgauti iš įvairių įmonės dokumentų, suskirstomi pagal skirtingus veiklos aspektus: darbo vietas, veiklos funkcijas, padalinius. Išnagrinėtas metaduomenų saugyklų kūrimas UEML, GRAI tinklo, BPMN metamodelių pagrindu bei jų savybės ir palygintos su ARIS metodo metamodelio pagrindu sukurtos metaduomenų saugyklos savybėmis. Nustatyta, kad veiklos modeliavimo metodo ARIS taikymas leidžia sukurti metaduomenų saugyklą, apimančią daugiau įmonės veiklos aspektų. Tai leidžia geriau identifikuoti įmonės duomenų pokyčius ir vietą veiklos valdymo procese. / This paper analyses creation of a metadata repository by implementing the meta-model of enterprise modelling methods and notations. It encompasses overviews of metadata repository qualities of UEML, ARIS, GRAI Grid, BPMN enterprise modelling methods and notations and their application in designing metadata repositories. ARIS method was selected as the most suitable enterprise modelling method because it encompasses the largest number of aspects of the enterprise activity. Employing this method meta-model, a metadata repository was created and integrated with the informational system of the company UAB „Klaipėdos Konteinerių Terminalas“ for more efficient management of the company’s data flows. With the help of this metadata repository data extracted from various company’s documents is grouped according to different aspects of the company’s activity: workplaces, departments, job functions, etc. The paper investigates creation of metadata repositories on the basis of UEML, GRAI Grid and BPMN meta-models and compares their qualities with those of a metadata repository created on the basis of ARIS meta-model. It has been established that application of ARIS enterprise modelling method enables to create a metadata repository encompassing more aspects of the enterprise’s activity. It allows to better identify changes in the enterprise’s documents and their place in the activity management process.

The thesis outlines a hierarchical modelling methodology for investigation in vehicle dynamics, in particular for combined ride and handling manoeuvres. The methodology involves the use of detailed multi-degrees of freedom models of vehicles with the inclusion of sources of non-linearity, using a multi-body approach, based on Lagrangian dynamics for constrained systems. It also includes the use of simpler and task-specific models, formulated in Newton-Euler approach. These simpler models with lower degrees of freedom, but with appropriate level of detail are more efficient in the study of specific, but non-trivial problems such as transient behaviour of vehicles in combined ride and handling, as encountered in many routine daily manoeuvres. The modelling methodology is supported by careful vehicle testing, both for validation of the proposed approach, and assessment of the extent of applicability of simple, intermediate and multi-degrees of freedom full-vehicle models. Certain important vehicle handling and ride characteristics in pitch plane dynamics, roll behaviour, vehicle body bounce and combination of these have been studied, as well as the effectiveness of restraining action of chassis elements, such as the semileading and trailing arms for passive control of vehicle squat and dive motions, arising from acceleration from coast to drive and deceleration/brake of vehicle from drive to coast. Combined pitch and bounce motions have been studied when negotiating speed traps such as bumps, which also combine with significant body roll when single event obstacles of this kind are introduced. The novelty of the research is in the detailed integrative numerical-experimental approach, and the development of intermediate models that adequately predict vehicle behaviour under steady and non-steady conditions for a wide range of ride and handling manoeuvres. The investigations have culminated in a significant number of findings of practical use, particularly the ineffectiveness of anti-squat and dive features when combined pitch and bounce motions limit the usefulness of these devices. On the contrary, excessive roll dynamic behaviour of the vehicle is effectively palliated by the anti-roll bar, even under complex combined pitch, roll and body bounce such as those experienced in negotiating single event speed bumps. Good agreement is found between the predictions of the intermediate model and those of the multi-body model and the actual vehicle tests, particularly for pitch and bounce dynamics.

Three-dimensionally woven composites are a relatively new class of material that offer improved out-of-plane performance by including through-the-thickness mechanical reinforcement compared to traditional laminated composite structures. The mechanical properties are highly dependent upon the weave architecture as this dictates the nature of the through the thickness reinforcement and its effect in improving out-of-plane shear strength. A comparison of two testing methods, Short Beam Strength, and Five Point Bending was conducted over a range of span to thickness ratios with the latter found to be more consistent at producing shear failure over a greater range of span to thickness ratios, although evidence of matrix crushing was present in both, and flexural failure in the Short Beam Strength test. Two weave architectures, the orthogonal and angle weave were subjected to the Five Point Bending test and the failure and damage progression behaviour of both weave architectures were characterised using Digital Image Correlation analysis to measure the edge strain through the thickness of the specimens. This testing showed the angle weave architecture had in general a higher failure strength, and more gradual failure due to longer debonding cracks. The orthogonal weave architecture showed a characteristic post-failure response indicative of crack bridging with discrete load recovery and load drop phases. A numerical model developed from previous work builds on the mosaic modelling method and was modified to include cohesive elements in order to simulate interface debonding via the maximum stress criterion. The simulations are consistently 15 20% greater in failure loads, and 8 - 12% greater in failure shear stresses than those found from the averaged experimental results over the range of tested span to thickness ratios. Post failure response was not modelled. The work presented in this thesis is another step towards gaining a thorough understanding of the mechanical properties of 3D woven composite structures, focussing in particular on out of plane shear strength. The modified mosaic modelling method used showed it is effective at modelling the out of plane testing of orthogonal 3D woven composite structures, and offer the potential to predict the failure of larger composite structures of the same construction and 3D woven architecture although developments are still needed in modelling the post failure response.

Le bâtiment quel qu'il soit (industrie, tertiaire, domestique), s'électrifie toujours plus pour répondre entre autre au besoin de l'efficacité énergétique. Ainsi de nombreux équipements, en particulier d'électronique de puissance ou encore d'actionneurs sont présents et produisent des perturbations électromagnétiques tant conduites que rayonnées véhiculées ou propagées par les câblages électriques et dispositifs de protection, dimensionnés pour 50Hz et non les hautes fréquences. Ceci est à l’origine de la problématique CEM dans le bâtiment. La méthodologie de modélisation proposée consiste à chercher un schéma équivalent pour chaque constituant ou composant d’une installation. L’accès aux paramètres localisés de chacun est un des points fort de la méthode, tout en permettant l’association avec d’autres composants (couplages). Les modèles ainsi générés sont implantés automatiquement dans un logiciel circuit, en l’occurrence Spice, permettant la résolution et l’association d’autres équipements électriques complexes. La modélisation des différents composants, principalement le circuit de mise à la terre, permet la simulation et l’analyse de plusieurs configurations qui devraient nous permettre de dicter des règles de conception pour rendre plus robuste le bâtiment moderne fortement électrifié. / In order to answer the need for energy efficiency, building, whatever it is (industrial, commercial, domestic), still electrifies more and more. Hence many equipment, in particular power electronic or actuators are introduced and produce conducted and radiated electromagnetic disturbances. They propagate via cabling and protection devices which have been sized for 50Hz and not for high frequencies generated by the equipment. This constitutes the EMC problematic for building.The proposed modeling methodology is to evaluate an electrical equivalent circuit for each installation’s component. Access to lumped parameters of each component is one of the highlights of this method, allowing the other components association (couplings). The generated models are implemented in a circuit software such as Spice, enabling the resolution and the association of other electrical components.The modeling of the components, mainly the grounding circuit, allows the simulation of different configurations. The analysis of several configurations should help the elaboration of design rules to make today buildings more efficient.

Ce travail concerne la synthèse et la caractérisation structurale de nanocristaux de faible taille (~1-3 nm) d’oxydes métalliques simples, à savoir le dioxyde de titane (TiO2), le dioxyde d’étain (SnO2) et le monoxyde de zinc (ZnO). Les synthèses ont été réalisées au moyen de méthodes sol-gel non-aqueuses voire strictement non-hydrolytiques sous contrôle cinétique. La caractérisation structurale s’est principalement appuyée sur la diffraction des rayons X, la microscopie électronique en transmission et la méthode des fonctions de distribution de paires atomiques, obtenues grâce à la diffusion totale des rayons X, couplées à des méthodes de modélisation à l’échelle atomique. Dans le cas de TiO2, des nanoparticules d’anatase bien cristallisées de 4 nm à 8 nm ont été synthétisées. Le ratio molaire de donneur d’oxygène par rapport au titane s’est avéré être un paramètre influençant fortement la taille des particules. Nous avons également mis en évidence la formation d’une phase intermédiaire caractérisée par des nanoparticules faiblement cristallisées de très faible taille dont la structure pourrait s’apparenter à une structure brookite désordonnée. Pour SnO2, des nanocristaux présentant une structure rutile ont été obtenus avec des tailles comprises entre 2 nm et 4 nm. Dans le cas de l’utilisation d’un éther, nous avons mis en évidence la formation concomitante d’une phase organique polymérisée et de nanoparticules primaires dont la structure intermédiaire présente de fortes similitudes avec la structure rutile. L’utilisation de solvants possédant une fonction benzyle en présence de tétrachlorure d’étain a conduit à la formation d’eau dans le système. Dans le cas de ZnO, nous avons montré que l’utilisation d’une base organique pour initier la formation du réseau oxyde dans une solution méthanolique d’acétate de zinc en présence d’un agent complexant du zinc permettait d’obtenir des nanoparticules de l’ordre de 1 nm. Même pour les faibles valeurs de taille, les nanoparticules présentent une structure très proche de la wurtzite avec un désordre croissant au niveau du réseau cationique. / The aim of this work was to synthesize small size (~1-3 nm) metal oxide nanocrystals namely titanium dioxide (TiO2), tin dioxide (SnO2) and zinc oxide (ZnO), and to study their structure. Syntheses were conducted via non-aqueous or even strictly non-hydrolytic sol-gel methods under kinetic control. The structural characterization was mainly carried out by X-Ray diffraction methods, transmission electronic microscopy and the study of pair distribution functions, obtained by X Ray total scattering, coupled with atomic scale modelling methods. In the case of TiO2, anatase nanocrystals were obtained with sizes ranging between 4 nm and 8 nm. The molar ratio of the oxygen donor with respect to titanium was shown to be an important parameter to control the nanoparticle size. In peculiar conditions we have been able to isolate an intermediate phase characterized by very small sized and poorly crystallized nanoparticles which the structure can be assimilated to a disordered brookite structure. Concerning SnO2, rutile-type nanocrystals were synthesized with sizes ranging between 2 nm and 4 nm. The use of an ether as oxygen donor led to the simultaneous formation of an organic polymeric phase and of primary nanoparticles characterized by an intermediate structure close but still different from the rutile-type structure. Moreover, the use of benzyl-type solvents in the presence of tin tetrachloride led to the formation of water in the system. Lastly, for ZnO, we have shown that using an organic base to induce the formation of the metal oxide network in a methanolic solution of zinc acetate in the presence of a strong complexing agent of the zinc allowed us to obtain wurtzite nanocrystals of ultrasmall sizes around 1 nm. Even for the smallest sizes, the nanoparticles exhibit a structure very close to that of wurtzite with an increasing disorder of the cationic network.