Nous avons développé une méthode d'identification de courts-circuits et d'excentricités statiques d'un turboalternateur en fonctionnement. Cette méthode doit être non seulement capable d'identifier le type de défaut, mais également sa gravité et sa localisation. Pour cela, la force électromotrice issue d'une sonde de flux radiale est considérée pour identifier les défauts de la machine. Les méthodes d'apprentissage ayant fait leur preuve sur ce type de problématique, un ensemble d'échantillons des défauts, i.e. matrice des prototypes, est construit grâce à des simulations par la méthode des éléments finis de la machine dans un grand nombre d'états de fonctionnement. Le but est alors d'identifier la classe (le défaut) de mesures expérimentales en considérant les sorties de simulations. Pour faire cela, une maquette de turboalternateur est modélisée et les sorties de simulation sont comparées aux mesures expérimentales. L'application de méthodes de classification supervisée choisies selon la forme des données montre qu'on peut atteindre des taux de bonne classification de 79 % pour les courts-circuits et 93 % pour les excentricités après un traitement spécifique des attributs et un recalage des sorties de simulation. Une règle d'identification du type de défaut est également mise en place au travers d'une classification hiérarchique. Celle-ci donne d'excellents résultats pour lesquels toutes les mesures expérimentales sont assignées au bon type de défaut. Enfin, un automate à états finis permet d'améliorer les résultats précédents en tenant compte de l'évolution temporelle de l'état de fonctionnement de la machine. Ainsi, un taux de bonne classification de 91.5 % pour l'identification de courts-circuits est obtenu. / An identification method of rotor inter-turn short-circuits and static eccentricities of an operating turboalternator is developed. This approach provides the type of the machine functional state such as its seriousness and location. The electromotive force obtained from a radial flux sensor is considered in order to identify the machine faults. Some previous works have shown that learning methods are efficient to characterize precisely the fault of a machine. Thus, a fault signatures data set, i.e. prototypes matrix, is built thanks to simulations using finite element method applied to the machine considering a great number of functional states. Therefore, the goal of this work is to identify the class (fault) of experimental measurements using the simulated output. To do that, a small scale alternator is modelled and simulation outputs are compared to the experimental measurements. The application of supervised classification method chosen beyond the shape of data shows that a good classification rate of 79 % for short-circuits and 93 % for eccentricities can be achieved thanks to specific features and simulation output treatments. An identification rule of the fault type is designed by considering a hierarchical classification approach. It achieves excellent results since all experimental measurements are assigned to the right type of fault. Finally, the use of a finite state automaton achieves better results for short-circuits identification by taking into account the temporal evolution of the machine functional state.

Cette thèse expose les travaux effectués au sein du laboratoire central de l’IEMN. La finalité de ce travail est de participer a l’optimisation des transistors HEMTs de la filière AlGaN/GaN sur substrat Si(111), au niveau de leur fabrication et de leurs propriétés électroniques, qui seront a terme intégrés dans des dispositifs de puissance hyperfréquence. Ce manuscrit expose dans un premier chapitre les principales propriétés physiques, électriques et mécaniques des matériaux choisis, ainsi que le principe de fonctionnement du HEMT. Ensuite, toutes les étapes technologiques permettant la conception et la fabrication du HEMT seront décrites dans le deuxième chapitre, parmi lesquelles ont été menées plusieurs études sur leur optimisation. Une attention particulière sera portée sur les verrous technologiques rencontrés durant cette thèse. Le troisième chapitre présente l’ensemble des études effectuées sur les composants fabriqués au laboratoire. Elles s’appuient sur l’analyse des caractéristiques électriques en régime statique, pulsé, et hyperfréquence. Le dernier chapitre concerne l’étude de la fiabilité de HEMT AlGaN/GaN sur substrat SiC de même topologie. Les défauts induits par l’utilisation prolongée de ces transistors seront mis en exergue. / This thesis describes the work performed within iemn laboratory. The design, fabrication and improvement of AlGaN/GaN HEMTs on Si(111) substrate is the main goal of this work, as these transistors are aimed to be integrated in devices for microwave power applications. In the first chapter, this manuscript shows the worthwhile physical, electrical and mechanical properties of the used materials. HEMT working principle is also exposed in this part. Next, every step of the HEMT manufacturing process are described in the second chapter, including several optimization studies. Particular attention will be paid to bottlenecks encountered during the device making. The third chapter presents the whole studies done on these electronic components. Each study relies on electrical characterization in DC, pulsed and microwave modes. The final chapter regards reliability studies of AlGaN/GaN HEMT on SiC substrate having the same topology. Defects induced during more than 3000h ageing in high operating temperature will be highlighted.

Les micros résonateurs électromécaniques de type MEMS sont aujourd’hui étudiés pour leur intérêt dans les applications nécessitant des fonctions de capteurs ou d’actionneurs de petites dimensions co-intégrées avec des fonctions électroniques. Actuellement, la majorité des résonateurs sont issus des filières silicium et quartz. Cependant, les propriétés du silicium se dégradent lorsque la température dépasse 200°C alors que la co-intégration monolithique des composants en quartz est impossible. Parmi les matériaux possibles pour pallier ces limites, le nitrure de gallium (GaN) semble être un matériau prometteur. Ses propriétés piézoélectriques ainsi que la possibilité de l’intégrer avec des transistors à haute mobilité électronique AlGaN/GaN sont particulièrement intéressantes. Afin d’améliorer les performances des résonateurs électromécaniques à base d’hétérostructures AlGaN/GaN nous proposons de les miniaturiser. Dans ce contexte, nous avons proposé de développer la croissance épitaxiale de plusieurs structures de seulement quelques centaines de nanomètres. Leurs caractéristiques structurales, électriques et mécaniques ont été étudiées afin de définir la structure optimale pour les applications MEMS. Des architectures d’actionneurs piézoélectriques et électrothermiques ainsi que des détecteurs sans grille pouvant répondre aux critères de miniaturisation ont été étudiées. Les étapes du procédé technologique spécifique à ces composants ont développées. Pour finir, les performances des résonateurs fabriqués sur ces couches minces ont été mesurées afin de mettre en avant l’utilité des structures minces et les performances des différentes architectures des transducteurs. / Micro-electro-mechanical resonant systems are widely investigated for their applications in actuators and sensors of small dimensions and co-integrated with electronic functions. In the area of vibrating resonant devices, most are based on silicon and quartz technologies. Even if Si based resonators exhibit ultrasensitive mass/force detection they lose their mechanical and electrical properties for temperature higher than 200°C whereas quartz devices are not easily co-integrated. To overcome these intrinsic limitations, other approaches such as wide bandgap semiconductor have been investigated. Especially GaN exhibits good piezoelectric properties and benefits of co-integration opportunity with AlGaN/GaN high electron mobility transistors. Therefore it seems to be an ideal candidate to address new generation of MEMS sensors that withstand harsh environment. To optimize the actual resonator based on AlGaN/GaN hétérostructures, we have chosen to study the device downscaling process. In this view, we developped the epitaxial growth, by molecular beam epitaxy, of AlGaN/GaN heterostructures on 3 kinds of thin buffers. Structural, electrical and mechanical characterizations were carried out in order to select the best buffer for MEMS applications. Then, we suggested several designs of piezoelectric and electrothermal actuators as well as a gateless detector that are compatible with downscaling. Specific fabrication steps were developed and optimized. Finally, we compare the performances of resonators fabricated on thin buffers with one processed on thick commercial structure in order to bring forward the advantages of thin buffer and the performances of the transductor design.

Les cristaux liquides nématiques ancrés constituent un milieu propice à l'étude de la propagation non linéaire des faisceaux lumineux - que ce soit en termes d'études fondamentales ou plus appliquées comme le routage de l'information. Ce sont des milieux plus complexes que les systèmes "modèles" usuellement abordés (type équation de Schrödinger non linéaire) pour étudier les comportements non linéaires. En effet, la non-linéarité a une réponse non locale et est beaucoup plus lente que le temps d'évolution de l'onde. Enfin, le milieu est stochastique.Dans ce cadre, cette thèse étudie l'influence de la non-localité sur les singularités d'ondes obtenues lors de la propagation d'un paquet d'ondes optiques dans un cristal liquide nématique. C'est un travail expérimental et numérique. Nous montrons que le type de singularité obtenu est complètement régi par le taux de non-localité du système,- en régime fortement non local, une structure localisée, connue sous le nom de "nematicon", présente un profil d'intensité ultra-piqué (type "point de rebroussement") avec une trajectoire constamment fluctuante- en régime faiblement non local, le paquet d'ondes subit toujours un phénomène de type "effondrement d'ondes". Il est accompagné d'un auto-raidissement de l'enveloppe du paquet d'ondes et d'un profil en double exponentielle au point de compression.Enfin, nous mettons en évidence la présence d'ondes de chocs diffractives dans notre système en appliquant une condition initiale possédant une discontinuité (demi-faisceau Gaussien). C'est la première mise en évidence expérimentale de ce type d'ondes dans un milieu focalisant, stochastique et non local. / Anchored nematic liquid crystals is a suitable medium and an excellent playground for nonlinear beam propagation studies - for fundamental as well as for applied studies such as optical routing. They are more complex systems than the "ideal" ones that are usually referred to investigate nonlinear phenomena (such as the nonlinear Schrödinger equation). Indeed, our nonlinearity is characterized by a nonlocal response and its response time is much slower than that of the wave. At last, the medium is stochastic.In this context, this thesis deals with the effects of non locality on wave singularities that occur during the propagation of optical wave packets in nematic liquid crystals. The studies are experimental and numerical.We show that the type of singularity that emerges is completely driven by the amount of non locality,• for the highly nonlocal regime, a localized state, known as "nematicon" exhibits an ultra-sharp intensity profile (cusp like) with a continuously fluctuating trajectory.• for the weakly nonlocal regime, the wave packet is always subject to a collapse-like phenomenon. Its envelope self-steepens and its profile depicts an exponential shape where the beam is the most focused.Finally, we evidence diffractive shock waves thanks to an initial condition displaying a discontinuity (half Gaussian beam). This is the first experimental evidence of such phenomenon in a focusing non local and stochastic medium.

Thesis (M.S.)--Ohio State University, 1984. / Typescript. Vita. Includes bibliographical references (leaves 72-75). Available online via OhioLINK's ETD Center.

Thesis (M. Mus.)--Ohio State University, 1982. / Available online via OhioLINK's ETD Center.

The purpose of the study was to (1) identify major historical antecedents of urban 4-H and youth programs by the Cooperative Extension Service, (2) describe the evolution of urban 4-H and youth programs in Indiana, (3) illustrate the emergence of 4--H and youth programs in Indiana cities from 1968-1977, (4) investigate and clarify the perceived contemporary role of Extension Agents-Youth, and (5) identify leadership training needs as well as provide information to staff of Cooperative Extension that should assist in training agents corking in urban regions.The population for the study included all Extension AgentsYouth, Extension Boards, and Administrators from Lake, Marion and Vigo Counties that are currently active. The population also included the administrative staff of the Cooperative Extension Service from the State Office. The above counties were selected because of their highly urbanized populations and the program efforts that were initiated in the three counties.For the purpose of this study, the Director of Cooperative Extension Service in Indiana, Dr. Howard Diesslin, provided the approval to work with the three county staffs and administration. An additional Extension Staff was utilized fran Madison County to field test the questionnaire for clarity and relevance.The director of the Indiana Cooperative Extension Service granted approval for the questionnaire to be mailed under Extension mailing priviliges. The Associate Director, Dr. Paul Crooks, scrutinized the language of the instrument to determine whether the franking privilege available to Cooperative Extension Service could be used. Fifteen doctoral candidates and one professor of Educational Administration further critiqued the instrument for clarity, language and style.The questionnaire was mailed or handed to 75 Extension persons from Lake, Marion and Vigo Counties. Fifty-seven questionnaires were returned and 56 were usable. The data obtained from the questionnaires were analyzed, summarized and presented in narrative form. A frequency distribution was utilized for reporting numbers and percentages for each item in the sections relating to program, program responsibility, skills and training.The findings and conclusions of the study support the following recommendations:1. The Extension Service should improve communications with urban audiences through the efforts of Extension professionals, including youth agents, by effective means such as public meetings, media and wholesome programs.2. Program responsibilities should be clarified to increase effectiveness and accountability of Extension Agents-Youth.3. Goals and purposes of programs should receive a high level of agreement among regional staff.4. Professional Extension Agents-Youth employed to work in urban regions should have strong educational preparation in behavioral and social sciences.5. Agents working in urban regions should become familiar and work with other agencies to enhance efforts for meeting the needs of all people of the region.6. Agents in urban regions should have some knowledge in agriculture and related fields.7. A study should be made to determine whether training needsof personnel are being met. If not, immediate efforts should be directed toward such a goal.8. A follow-up study should be conducted to determine the kinds of training opportunities being pursued by the national staff and the state staff.9. A follow-up study should be conducted to compare attitudes of Extension personnel from different urban regions.10. Methods and techniques need to be developed to improve communications between administrators, Extension agents and board members.

We synthesized a series of enediyne containing Combretastatin A-4 analogues and evaluated the growth inhibitory activity of analogues against two human hepatoma cell lines (Hep G2, Hep 3B), two breast cancer cell lines (MDA-MB-231, MCF-7) and human lung cancer cell line (A-549). It was found that compounds 39a, 39d, 39e, 39f and 40a produced growth inhibitory activity for the five human cancer cell lines. Among them, compound 40a showed the best inhibitory activity (IC50 = 0.8 £gM for A-549 and IC50 = 8.53 £gM for Hep 3B) which were comparable to the Combretastatin A-4 (IC50 = 1.57 £gM for A-549 and IC50 = 14.32 £gM for Hep 3B).

A total of 27 new enediyne analogs of combretastatin A-4 (CA-4), with different substituents (Cl, Br, OCH3, SCH3 or N(CH3)2 ) of the aryl A-ring, were designed, synthesized, and evaluated for their growth inhibition activity against human tumor cell lines. Among them, compound 14a showed most significant inhibition activity against Hep3B (0.22 £gM), A-549 (0.49 £gM) and MCF-7 (0.22 £gM), and these activities were greater than CA-4. The enediyne analog 14a may be consider as a new drug candidate worthy of further investigation and development as a potential antitumor agent.

The goal of this experiment is to explore the absorption and reflection of UV light by TiO2/MCA in physical mixing and chemical synthesis. The experiment shows that TiO2/MCA in anatase structure is better than in amorphous structure (with FT-IR. UV-VIS. XRD and Raman spectrum). The first experiment shows that the small TiO2 particle size (290 nm) has better UV absorption than large TiO2 particle (400 nm) in the same mole. After TiO2/MCA(P) exposed to the UV source, the concentration of MCA in TiO2 anatase structure will decrease much than in TiO2 amorphous structure. In the second experiment, the raman peak in 1070 cm-1 proofs that the existence of the Ti-O-C structure. In addition, the FT-IR spectrums peck shows a part of TiO2/MCA has no chemical bonding, and also produces sparse ester (1100 cm-1 . 1300 cm-1). If the peak (1100 cm-1 . 1300 cm-1) were not exist, compare with the previous step, the level of chemical bonding will increase, and it will produce even less ester. In conclusion, The direction of UV absorption of TiO2/MCA(C1. C2. C3) and TiO2/MCA(P) are extremely similar.