Dissertations / Theses on the topic 'HFHC'

Des revêtements céramiques sont obtenus par un procédé de dépôt chimique en phase vapeur assisté par injection liquide pulsée (DLICVD) de précurseurs organométalliques. Des dépôts de carbure de chrome (CrCx) sont élaborés dans un réacteur tubulaire à paroi chaude à partir d’une solution de bis(benzène) chrome dans du toluène pour des températures de 475 °C et sous pression partielle d’azote (pression totale 50 Torr). Une couche d’accroche pouvant être nécessaire pour revêtir des pièces métalliques, tels des aciers et alliages, par un revêtement céramique non-oxyde de type CrCx, des couches de chrome métallique (Cr) et des carbures mixtes Cr-Si-C ont également été élaborées par ce procédé DLICVD. Ainsi, l’ajout d’un additif à base de chlore ou de soufre (par exemple l’hexachlorobenzène ou le thiophénol) dans la solution BBC/toluène permet la déposition de films de chrome métallique (Cr) à 475 °C. De plus, l’utilisation d’une solution de précurseur contenant simultanément du Si et du Cr tel que le tetrakis(trimethylsilylmethyl)chromium dans du toluène mène au dépôt d’un carbure mixte Cr-Si-C pouvant jouer le rôle d’interphase dans des assemblage céramique-métal. Des films de carbure de silicium (SiC) sont obtenus à partir de deux précurseurs (1,3 disilabutane et polysilyléthylène) injectés purs ou en solution également dans du toluène. Les dépôts sont faits dans une gamme de température comprise entre 700 et 800 °C, sous pression partielle d’azote (pression totale 50 Torr). Les films obtenus sont des films amorphes de SiC contenant une faible quantité d’hydrogène (provenant du mécanisme de décomposition des précurseurs) : a-SiC:H. Les films sont stœchiométriques dans le cas de l’injection de précurseur pur, et quasi stœchiométrique lorsque les précurseurs sont dilués dans du toluène. Les films amorphes tels que déposés deviennent nanocristallins en présentant la structure cubique du SiC après recuit sous vide à 1000 °C. L’influence du solvant (toluène) sur la composition, la morphologie et la vitesse de croissance des dépôts est discutée en fonction des systèmes chimiques étudiés et des conditions expérimentales, en particulier les conditions locales dans le réacteur DLICVD telles que les gradients de température et de concentration. Des films de carbure de hafnium (HfC) sont également élaborés par le même procédé à partir d’une solution de bis(cyclopentadiényl)diméthyl hafnium dans du toluène après avoir testé plusieurs précurseurs. Une température de 750 °C est utilisée et l’utilité d’une pression partielle de dihydrogène dans le gaz vecteur azote est démontrée (pression totale 50 Torr, 423 sccm de N2 et 77 sccm de H2). Tels que déposés, ces films sont riches en carbone (C-rich HfCx) et ont une structure quasi-amorphe. Ils deviennent nanocristallins après recuit sous vide à 1000 °C. Enfin, la mise en œuvre de films multicouches céramiques par DLICVD à paroi chaude est mise en évidence par l’élaboration de revêtements multicouches HfC/SiC à 750 °C, sous pression partielle d’un mélange de gaz vecteur N2/H2. Le contrôle du procédé permet une nano structuration de ces revêtements multicouches jusqu’à une bi-période de 100 nm (empilement de 100 couches d’environ 50 nm chacune). La stabilité thermique de ces architectures et des tests préliminaires de résistance à l’oxydation à haute température des films de SiC et HfC/SiC sont discutés
Ceramic coatings are made from metalorganic precursors by a chemical vapour deposition process assisted by pulsed liquid injection (DLICVD). Chromium carbide (CrCx) films are grown in a tubular hot wall reactor from a solution of bis(benzene)chromium in toluene under partial pressure of nitrogen at 475 °C (total pressure set at 50 Torr). Bonding layers are useful on metallic components, such as steels and alloys, with non-oxide ceramic films such as CrCx, to that purpose metallic chromium (Cr) and mixed carbides Cr-Si-C have been made by DLICVD. Furthermore, adding a chlorinated or sulfur based additive (e.g. hexachlorobenzene or thiophenol) in the BBC/toluene solution allows depositing metallic chromium (Cr) at 475 °C. Moreover, using a precursor containing Si and Cr as tetrakis(trimethylsilylmethyl)-chromium in toluene leads to the deposition of Cr-Si-C mixed carbide. Silicon carbide films are made from two precursors (1,3-disilabutane and polysilylethylene) that have been injected either pure or diluted in toluene. A temperature range of 700 to 800 °C has been used under a partial pressure of nitrogen (total pressure of 50 Torr). SiC films are amorphous and contain a small quantity of hydrogen (hydrogen comes from precursor pyrolysis mechanism): a-SiC:H. Films are stoichiometric when pure precursors are injected, and quasi stoichiometric when precursors are diluted in toluene. As deposited coatings are amorphous and become nanocristalline (cubic SiC structure) after annealing at 1000 °C under vacuum. The influences of the solvent (toluene) on the composition, morphology and growth rate are discussed as a function of the chemical system and experimental conditions, in particular reactor gradient conditions such as temperature and precursors concentration in gas phase. Hafnium carbide films are also made using a solution of bis(cyclopentadiényl)diméthyl hafnium in toluene by the same process. Temperature is set to 750 °C and hydrogen partial pressure has been shown useful (total pressure of 50 Torr, 423 sccm of N2 and 77 sccm of H2). As-deposited films are C-rich HfCx and quasi amorphous. They become nanocristalline after annealing at 1000 °C under vacuum. Finally, ceramics multilayer HfC/SiC coatings were deposited by DLICVD at 750 °C under a partial pressure of a mixture of N2/H2. The process allows a good control of the multilayer nanostructure. Thermal stability and high temperature oxidation preliminary tests on SiC and HfC/SiC films are discussed

In Vietnam today there is an urgent demand for well-designed tests of listening comprehension. Little attention has been given to this problem. This field study is intended to provide guidance on the design of listening comprehension tests and in particular for intermediate level students at the Hanoi Foreign Language College (HFLC). The Field Study Report consists of six chapters. Chapter One gives a brief introduction which covers the problem, aims, subjects (testees) and the background of the Field Study Report. Chapter Two deals with the purposes of testing in some detail, setting this in the framework of the teaching-testing link, teachers and testing, and students and testing. Test characteristics are considered and the problem of sampling for test content is addressed. Chapter Three concerns test items for listening comprehension. The chapter begins with a short description of listening comprehension and is followed by a survey of theorists on listening comprehension together with the test items for listening that they suggest. Some commonly-used standardised tests and their listening items are discussed. The chapter ends with a checklist of selected listening items suitable for students at HFLC in Vietnam. Chapter Four discusses the designing and trialling of test items for listening comprehension, such as multiple choice, completion of a taped talk, matching pictures with statements etc. The chapter presents the results of the trialling of sample items and also deals with correlations between the sample tests used. Chapter Five deals with test design and development in general. It clarifies the bases for test design and provides a checklist of steps in the development of tests. In addition, the chapter includes a resources inventory for listening test items. Chapter Six presents the conclusions of the Field Study Report.

Magíster en Gestión de Operaciones
Este proyecto, formó parte del plan de mitigación de clientes morosos implementado por una empresa de telecomunicaciones en el año 2009, dónde una de las líneas de acción estaba enfocada en disminuir la morosidad de los nuevos clientes en su primera factura. Se propuso un modelo que clasifique a los nuevos clientes en pagadores y no pagadores, permitiendo con ello, reorientar recursos y realizar acciones preventivas de morosidad de manera más focalizada y efectiva, las cuales estaban siendo hasta ese momento aplicadas al universo total de nuevos clientes. El éxito de este proyecto requería el compromiso del área de cobranzas, usuaria del modelo, por esto, la elección del modelo consideró aspectos tanto teóricos como prácticos. La simpleza y la fácil ejecución del modelo, idealmente programable, eran variables deseables y decidoras del éxito de éste. Se ejecutó basado en Clasificación Bayesiana por su forma simple de utilizar y presentar una confiabilidad aceptable [17]. Los métodos de inducción bayesiana han demostrado ser una clase de algoritmos tan competitivos como los métodos árboles de decisión y redes neuronales [29], que permiten implementar algoritmos en Wolfram Mathematica 6.0 cuyas sentencias de programación son muy similares al lenguaje C++, permitiendo su programación posterior. El modelo se confeccionó con 42.087 clientes, contratantes entre el 15 de julio y el 15 de octubre del 2009, de estos, 40.087 formaron la base de aprendizaje para calibrar y realizar los cálculos de probabilidades, y 2.000 como base de testing. Los clientes a clasificar en pagadores y no pagadores correspondieron a 9.328, quienes ingresaron a la empresa entre el 16 de Octubre y el 15 de Noviembre del 2009. En base a establecer una probabilidad a priori, se plantea en valores estimados de corte probabilístico basado en la experiencia o un método de clasificación [19], por una probabilidad de corte de 0,4 para clasificar a un cliente pagador por la necesidad que enfrenta la empresa de mejorar el % de clientes morosos en su primera boleta y a la vez reducir los costos actuales involucrados en la cobranza. El modelo clasificó a 2.910 clientes como no pagadores, 2.179 de ellos efectivamente no cancelaron su boleta al vencimiento (75%). Este resultado es compatible con el esperado en el análisis teórico, debido a la eficiencia de los estimadores estimados, como también, el tamaño de los set de datos empleados que en forma teórica respalda los resultados obtenidos con un 80% de confiabilidad. Como trabajo futuro, resulta interesante evaluar el aporte del modelo en la reducción real del porcentaje de clientes morosos al permitir reorientar recursos y focalizarse en ciertos clientes, como a su vez, actualizar la base de aprendizaje de tal forma ir considerando la actualidad al modelo.

En la presente tesis se estudia el diseño del servicio IPTV para brindar una solución a la digitalización de la televisión en el mundo IP. Se plantea dos propuestas de tecnología para el servicio de IPTV: HFC y FTTH, logrando así la interactividad entre la televisión y el usuario. En el capítulo 1 se detalla el origen IP, IP multicast, las aplicaciones de IP, IPTV en el mercado mundial, las aplicaciones de IPTV, comparación entre IPTV y CATV, el marco problemático y los objetivos. En el capítulo 2 se muestra el servicio de IPTV en tecnología HFC y FTTH, así como también las tecnologías FTTH y HFC con sus elementos de red. En el capítulo 3 se analiza los diseños de la cabecera IPTV y su transporte mediante tecnología HFC y FTTH. Finalmente, en el capítulo 4 se detallan los presupuestos del servicio IPTV, así como la comparación de calidad del servicio entre ambas tecnologías. Comparando los resultados de ambas tecnologías, se recomendará la mejor para el despliegue del servicio IPTV.
Tesis

1,1,1,2-tetrafluoroethane (HFC 134a) is one of the major replacement compounds for chlorofluorocarbons and is therefore of great industrial importance. This work describes a method of characterisation for HFC 134a. The method profiled the levels of synthesis by-products present in samples of HFC 134a using gas chromatography (GC) with detection by flame ionisation detector (FID) or electron capture detector (ECD). The principal method for identification of the by-products was EI GC-mass spectrometry. The multi-variate data produced by the profiling of samples were analysed using chemometric techniques. A training set of samples of HFC 134a, with known origins of production, was analysed by both GC-FID and GC-ECD. This training data set was used to investigate the different methods of chemometric analysis as applied to the raw data, normalised data and principal component analysed data. K-means clustering and Hierarchical clustering were investigated to find the optimum methods for the identification of samples' origins of production based on their chromatographic profiles. The FID chromatographic traces could be correctly identified by applying a two step principal component analysis (PCA) using a hierarchical clustering method to classify the samples. The ECD chromatographic data could be correctly identified by applying a PCA followed by classification using a hierarchical clustering method. The two classification techniques were used to identify further samples of HFC 134a into clusters belonging to the known origins of production or into new clusters representing samples of unknown origin. Classification of samples using the ECD data required the least amount of operator interpretation and provided the least amount of ambiguity in sample identification.

There is a great need for competent teaching of Business English in tertiary institutions in Vietnam. At the Hanoi Foreign Trade College (HFTC) alone, about 150 students per year seek training to equip them to work in export and import organizations, banks for foreign trade and customs departments. In teaching such students, one problem is the provision of interesting and relevant materials. This study addresses the question of selection of materials and techniques for teaching business English listening skills. This Field Study Report consists of five chapters. Chapter 1 describes the purposes and aims of study, the background to this study, dealing with students, the teaching and the curriculum. Chapter 2 looks at the general theory about listening and in particular listening in business. The chapter also describes listening requirements for business graduates from the HFTC. Chapter 3 discusses criteria for selection of textbooks for language teaching in general and for teaching listening skills in business English in particular. Chapter 4 surveys the teaching of business English in two institutes of Technical and Further Education in the Australian Capital Territory (ACT). The final chapter makes suggestions for priorities to improve the teaching of business English. This Field Study Report should be regarded as an exploratory attempt in choosing classroom techniques and materials for teaching listening skills to students of commerce at the Hanoi Foreign Trade College.

Background: HFNC system is a novel device used with aerosol therapy and seems to be rapidly accepted. Although there are some studies conducted on HFNC and vibrating mesh nebulizer, the effect of HFNC on aerosol delivery using jet nebulizer or pressurized metered-dose inhaler (pMDI) has not been reported. In an effort to examine the effect of HFNC on aerosol deposition, this study was conducted to quantify aerosol drug delivery with or without a HFNC using either pMDI or jet nebulizer. Methodology: The SAINT model, attached to an absolute filter (Respirgard II, Vital Signs Colorado Inc., Englewood, CO, USA) for aerosol collection, was connected to a pediatric breathing simulator (Harvard Apparatus, Model 613, South Natick, MA, USA). To keep the filter and the SAINT model in upright position to collect aerosolized drug, an elbow adapter was connected between the absolute filter and the breathing simulator. An infant HFNC (Optiflow, Fisher & Paykel Healthcare LTD., Auckland, New Zealand) ran at 3 l/min O2 was attached to the nares of the SAINT model. Breathing parameters used in this study were Vt of 100 mL, RR of 30 breaths/min, and I:E ratio of 1: 1.4. Aerosol drug was administered using: 1) Misty-neb jet nebulizer (Allegiance Healthcare, McGaw Park, Illinois, USA) powered by air at 8 l/min using pediatric aerosol facemask (B&F Medical, Allied Healthcare Products, Saint Louis, MO, USA) to deliver albuterol sulfate (2.5 mg/3 mL NS), and 2) Four actuations of Ventolin HFA pMDI (90 μg/puff) (GlaxoSmithKline, Research Triangle Park, NC, USA) combined with VHC (AeroChamber plus with Flow-Vu, Monaghan Medical, Plattsburgh, NY, USA). Aerosol was administered to the model with and without the HFNC and another without (n=3). Drug was collected on an absolute filter, eluted and measured using spectrophotometry. Independent t tests were performed for data analysis. Statistical significance was determined with a p value of <0.05. Results: The mean inhaled mass percent was greatest for pMDI with (p = 0.0001) or without HFNC (p = 0.003). Removing HFNC from the nares before aerosol treatment trended to increase drug delivery with the jet nebulizer (p = 0.024), and increased drug delivery by 6 fold with pMDI (p = 0.003). Conclusions: Aerosol drug may be administered in pediatrics receiving HFNC therapy using either jet nebulizer or pMDI. However, using pMDI, either with or without HFNC, is the best option. When delivering medical aerosol by mask, whether by jet nebulizer or pMDI, removing HFNC led to an increase in inhaled mass percent. However, the benefit of increased aerosol delivery must be weighed against the risk of lung derecruitment when nasal prongs are removed.

The diploma thesis addresses facility of QoS control with GNU/Linux tools iptables and iproute. An attention is focused especially on HTB and HFSC traffic shaping methods with regard to utilization in wireless networks. The paper also includes a simulation of ensuring QoS in wireless network with 802.11e amendment.

Las redes híbridas de fibra óptica y cable coaxial, HFC, fueron implementadas en un principio por operadores de TV Cable, los que posteriormente incluyeron servicios como Video-on-Demand, Pay-Per-View, etc. Con el avance de la tecnología, las redes de TV Cable fueron capaces de ofrecer otros servicios multimedia como telefonía y acceso a Internet de Banda Ancha. Para esto, modificaron las mismas redes existentes, transformándose en Operadores Multi-Servicio, MSO. La red sufrió modificaciones importantes, pasando de ser una red prácticamente unidireccional a ser una red bidireccional desbalanceada. En la discusión actual sobre telecomunicaciones se encuentran habitualmente términos como convergencia IMS (IP Multimedia Subsystem) y Redes de Nueva Generación, NGN (Next Generation Networking). HFC representa un bloque fundamental en la comprensión de estos conceptos. Un factor clave para el éxito de los operadores de cable que pretenden adecuar sus redes para la próxima generación de arquitecturas de comunicaciones, será la capacidad del personal técnico para evaluar las diversas opciones disponibles. Por otro lado, los MSO, se enfrentan a otras consideraciones como el manejo de la compatibilidad de las diversas normas de próxima generación con las arquitecturas existentes y cómo determinar el momento óptimo para realizar este cambio. El objetivo de esta memoria fue diseñar un curso teórico y práctico sobre servicios soportados por redes HFC de nueva generación, basándose en los atributos de escalabilidad y disponibilidad que esta tecnología presenta frente a otras. Para guiar el desarrollo de este proyecto se utilizó una metodología que consiste en la recopilación de información sobre tecnología HFC en la etapa inicial, seguido de un análisis de los datos obtenidos. Con esta información se generó un programa docente y una planificación que permitió desarrollar el curso. En este proyecto de título se diseñó e implementó un curso de servicios teórico y práctico. En lo teórico se cubren temas como: Broadcasting de TV, Internet/Datos, Telefonía, IPTV, eTV, Video-on-Demand, Pay-Per-View, VoIP, videoconferencias y Juegos On Line. En lo práctico, se presenta una propuesta preeliminar de laboratorio docente, que contempla una red HFC a escala y guías prácticas sobre la evaluación de servicio y la vía de retorno. El curso se diseñó para hacer uso de recursos docentes presenciales e interactivos a través de Internet. Este curso representa un aporte, tanto fuera como dentro del Departamento de Ingeniería Eléctrica de la Universidad de Chile, al tratar temáticas actuales con un enfoque orientado a la industria de los servicios ofrecidos por los MSO. El curso diseñado es una propuesta orientada a completar la instrucción de un ingeniero civil electricista.

Les paramètres physiques à la base des fuites du fluide frigorigène HFC-134a dans les systèmes de climatisation automobile ne sont pas encore pleinement compris. Le but de ce travail de recherche est d'établir une méthode de mesures des débits de fuite de fluide frigorigène des systèmes de climatisation automobile ainsi que des composants de ces systèmes, et aussi de développer une approche générique de prévision de ces émissions. Dans la thèse, les fuites chroniques des différents composants des systèmes de climatisation automobile sont évaluées et hiérarchisées. Une méthode d'essais de laboratoire, basée sur la mesure de concentration dans un volume d'accumulation, est présentée pour déterminer les débits de fuite de systèmes et de composants. La précision de la mesure est aussi justifiée. Des mesures en régime permanent et pour plusieurs températures contrôlées sont effectuées afin de comparer le débit de fuite de l'ensemble du système et la somme des débits de fuite de tous les composants. Les simulations de la variation de température permettent de prédire les impacts des conditions climatiques annuelles quel que soit le climat. Les essais en régime dynamique sont également traités pour analyser la contribution du temps de fonctionnement du système aux émissions annuelles du système. Afin de vérifier la méthode d'essai en laboratoire, des opérations de récupération du fluide frigorigène ont été effectuées sur une quarantaine de véhicules avec une précision de +0 / -1 g. Sur la base des résultats des essais en laboratoire et de ceux obtenus sur la flotte de véhicules, un facteur de corrélation a été établi pour corréler les tests en laboratoire aux émissions mesurées sur le terrain. Les prévisions des émissions de tuyauteries flexibles utilisées dans les systèmes de climatisation automobile ont été développées en prenant en compte les effets de la température et de la pression. Les joints toriques typiques sont étudiés et deux modes de fuite: la perméation à travers des matériaux polymères et l'écoulement du gaz dans les micro-canaux sont distingués. Les performances d'étanchéité d'un joint torique radial sont étudiées en utilisant la méthode des éléments finis. Le comportement non-linéaire des déformations des polymères est pris en compte. L'analyse des facteurs principaux tels que la contrainte, la pression de contact maximale et le contact est basée sur les résultats de simulations numériques. Les deux modes de fuites permettent de comprendre les phénomènes clés des émissions et donc d'améliorer les performances d'étanchéité. En résumé, le débit de fuite d'un système de climatisation automobile est la somme des débits de fuite de toutes les sources de fuites. Ces sources sont de deux types: la perméabilité du gaz dans les polymères et l'écoulement du gaz dans les micro-canaux existant entre les joints et les parties métalliques des raccords. Pour chaque mode d'émission, une loi de comportement a été développée et les modèles prédictifs permettent de prévoir les débits de fuite avec un nombre limité de mesures.

Hydraulic devices play an essential role in mechanical engineering due to their high-power density, good controllability, flexible application and high robustness, which expose innovative methods of energy transmission. However, in applications where there is an increased risk of fire or explosion, the commonly used combustible mineral oils represent an unacceptable safety hazard. In such cases, fireresistant, water-based hydraulic fluids are in demand. A special feature of these liquids is their high cavitation tendency and the associated strong erosion wear. The aim of this research is to predict the cavitation behaviour of HFC and the subsequent erosion phenomena using numerical methods and to validate the results with experiments. Additionally, experimental results for HFC were compared with HLP. The findings help to implement further developments to decrease the erosive effect of cavitation in high-pressure differences in hydraulic components. For this purpose, flow geometries of typical hydraulic components, e.g. valve and pump, are used for experimental and numerical investigation. The large-eddy simulation (LES) turbulent modelling is used with Zwart-Gerber cavitation model. The cavitation aggressiveness is quantified by cavitation erosion indices according to Nohmi.

Abstract In this work the LCVD of copper and nickel from the precursor gases Cu(hfac)tmvs and Ni(CO)4 has been investigated. The in-house constructed LCVD system and processes and the practical utilisation of these in prototype integrated circuit edit work are described. The investigated process parameters include laser power, laser scan speed, precursor partial pressure and the effect of H2 and He carrier gases. The deposited metal conductor lines have been examined by LIMA, AFM, FIB secondary electron/ion micrography, and by electrical measurements. Furthermore, the study of experimental FIB circuit edit processes is carried out and discussed with particular emphasis on ion beam induced ESD damages. It is shown how the LCVD and FIB methods can be combined to create a novel method to carry out successfully circuit edit cases where both methods alone will fail. The combined FIB/LCVD- method is shown to be highly complementary and effective in practical circuit edit work in terms of reduced process time and improved yield. Circuit edit cases where both technologies are successfully used in a complementary way are presented. Selected examples of some special circuit edit cases include RF- circuit editing, a high resolution method for FIB-deposited tungsten conductor line resistance reduction and large area EMI shielding of IC surfaces. Based on the research it was possible for a formal workflow for the combined process to be developed and this approach was applied to 132 circuit edit cases with 85% yield. The combined method was applied to 30% of the total number of edit cases. Finally, the developed process and constructed system was commercialized.

The refrigerant charge reduction in HVAC\R systems is an important issue because it falls within environmental policies regarding refrigerants contributions to the greenhouse effect. A way to move toward charge reduction is to increase the compactness of heat exchangers, which means more complex designs of the evaporators. Nevertheless, while a large amount of studies have has been published on the thermal and hydraulic analysis of flow boiling of refrigerants in horizontal tubes, very little attention has been given to flow boiling in geometries different from straight tubes. This PhD thesis aims at studying the flow boiling characteristics of third generation (HFC) and fourth generation fluids such as HFO-1234yf in geometries which modify the fluid dynamics and two-phase flow with respect to horizontal straight tubes. To achieve this goal, an experimental test facility was specifically designed and built to conduct refrigerant evaporation experiments. This test facility allowed to perform flow regimes visualizations and pressure drop measurements in singularities (such as sudden contractions and return bends). First, two-phase flow regimes visualizations have been carried out using HFO-1234yf, R-134a and R-410A either in straight tubes or in singularities. A qualitative analysis of the flow behavior and also several comparisons to flow pattern prediction methods from the literature were conducted. The second objective of this work was to characterize the flow disturbances caused by singularities such as sudden contractions and return bends, and to study their effects on the hydrodynamic performance (e.g. pressure drop) of refrigerants.Lastly, pressure drop databases for third and fourth generation refrigerants are presented.

Orientador : Evandro Conforti
Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação
Made available in DSpace on 2018-07-28T20:05:53Z (GMT). No. of bitstreams: 1 Mattarredona_EdgarAntonioCosta_M.pdf: 1277176 bytes, checksum: 86ae5ea6da667f07e5225456194a3d3e (MD5) Previous issue date: 2001
Mestrado

This study aims to experimentally investigate the spray cooling characteristics for active two-phase cooling of automotive power electronics. Tests are conducted on a small-scale, closed loop spray cooling system featuring a pressure atomized spray nozzle. Two types of refrigerants, HFC-134a (R-134a) and HFO-1234yf, are selected as the working fluids. The test section (heater), made out of oxygen-free copper, has a 1-cm2 plain, smooth surface prepared following a consistent procedure, and would serve as a baseline case. Matching size thick film resistors, attached onto the copper heaters, generate heat and simulate high heat flux power electronics devices. The tests are conducted by controlling the heat flux in increasing steps, and recording the corresponding steady-state temperatures to obtain cooling curves. The working fluid is kept at room temperature level (22oC). Performance comparisons are made based on heat transfer coefficient (HTC) and critical heat flux (CHF) values. Effects of spray characteristics and liquid flow rates on the cooling performance are investigated with the selected coolants. Three types of commercially available nozzles that generate full-cone sprays with fine droplets are utilized in the tests. Effect of liquid flow rate is evaluated varying flow rates at 2, 3, 4 ml/s. The experimental results obtained from this study provide a framework for spray cooling performance with the current and next-generation refrigerants aimed for advanced thermal management of automotive power electronics.

Thesis (M.Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1997.
In title, on t.p. of thesis and in exercises program, "[mu]" appears as the Greek lower-case letter.
Includes bibliographical references (leaf 95).
by Sofia Gebru.
M.Eng.

Dissertação de Mestrado em Engenharia Electrotécnica e de Computadores - Eletrónica e Computadores
No âmbito da adequação dos ciclos de estudo da Escola Superior de Tecnologia de Setúbal ao designado Processo de Bolonha, pretendo com este ralatório, enquanto licenciado "pré-Bolonha", obter o grau de Mestre em Engenharia Eletrotécnica e de Computadores, com base nas competências adquiridas no decorrer do meu percurso profissional e aqui descritas. Concluo que as funções desempenhadas permitiram-me consolidar e expandir os conhecimentos adquiridos na Licenciatura em Engenharia Eletrotécnica e Computadores

La hipomagnesèmia familiar amb hipercalciúria i nefrocalcinosi (HFHNC) és una tubulopatia minoritària autosòmica recessiva causada per mutacions als gens CLDN16 o CLDN19, que codifiquen per la claudina-16 i -19, respectivament, que s’expressen a la porció gruixuda de la nansa ascendent de Henle, i estan implicades en el transport iònic paracel·lular. Aquesta malaltia es caracteritza per la pèrdua urinària massiva de calci i magnesi, nefrocalcinosi bilateral i progressió inexorable de la malaltia renal crònica, que desencadena en fallida renal. Addicionalment, la majoria de pacients amb mutacions a CLDN19 també desenvolupen alteracions oculars, ja que aquest s’expressa a les cèl·lules epitelials de la retina. Característicament, existeix una gran variabilitat fenotípica entre els pacients, inclús entre aquells que comparteixen la mutació c.59G>A; p.G20D a CLDN19 (mutació Hispànica), i entre membres d’una mateixa família. Aquest fenomen suggereix que possiblement existeixen altres processos moleculars que determinen la progressió clínica dels pacients. Sota aquesta hipòtesi, aquest treball s’ha focalitzat en l’estudi dels factors genètics i epigenètics que podrien modular la progressió de la malaltia renal. Per això, s’han obtingut les dades clíniques i mostres d’orina i sang de 30 pacients afectes d’HFHNC i 6 individus control d’arreu d’Espanya, essent la majoria dels pacients diagnosticats d’HFHNC en aquest territori. L’anàlisi de les dades clíniques va permetre avaluar les diferències en l’evolució de la malaltia renal en funció del sexe i el genotip, classificar els pacients segons la pèrdua anual de funció renal (progressió renal ràpida, moderada i lenta), identificar biomarcadors clínics de pronòstic, i evidenciar l’alta variabilitat fenotípica intrafamiliar. L’estudi de variants gèniques en homozigosi associades als fenotips més extrems (progressió ràpida i lenta) s’ha realitzat amb les dades obtingudes de la seqüenciació massiva de l’exoma dels 30 pacients de la cohort. Seguint aquesta estratègia, s’han identificat un total de 45 variants gèniques. D’entre aquestes, per la funció fisiològica dels gens on es localitzen, en destaquen la rs11207827 (al gen PATJ) i la rs1050171 (al gen EGFR). Per determinar els factors epigenètics implicats en la fisiopatologia de la HFHNC i en la seva progressió, s’han utilitzat els urinary exosome-like vesicles (uEVs) com a font no invasiva d’informació dels processos cel·lulars renals. Mitjançant microarrays, s’ha analitzat el perfil d’expressió dels miRNAs continguts als uEVs dels 20 pacients que mantenien els ronyons natius funcionals identificant-se 24 miRNAs diferencialment expressats en el conjunt de pacients respecte el grup control, i 43 en el subconjunt dels pacients homozigots per la mutació p.G20D a CLDN19. La comparació de pacients d’ambdós sexes, va mostrar únicament la infraexpressió del miR 1915 5p en aquells de sexe masculí. Per últim, s’han identificat 4 miRNAs diferencialment expressats en els pacients d’HFHNC amb progressió renal moderada, en comparació amb el de progressió lenta, i 8 en el subgrup de pacients homozigots per la mutació p.G20D a CLDN19. La biologia de sistemes i l’ús de xarxes neuronals d’intel·ligència artificial han permès associar els resultats obtinguts amb processos biològics crucials en la fisiopatologia de la HFHNC, com la fibrosi renal i el transport de calci i magnesi, entre d’altres. Aquest treball ha permès incrementar el coneixement de la fisiopatologia de la HFHNC i determinar factors clínics, genètics i epigenètics implicats en la progressió de la malaltia renal i identificar nous possibles biomarcadors pronòstics de la HFHNC, i dianes terapèutiques que podrien permetre modular la severitat de la malaltia i, en el millor dels casos, curar-la.
La hipomagnesemia familiar con hipercalciuria y nefrocalcinosis (HFHNC) es una tubulopatía minoritaria autosómica recesiva causada por mutaciones en los genes CLDN16 o CLDN19, que codifican para la claudina-16 y -19, respectivamente, expresadas en la porción gruesa del asa ascendente de Henle, e implicadas en el transporte iónico paracelular. Esta enfermedad se caracteriza por la pérdida urinaria de calcio y magnesio, nefrocalcinosis bilateral y progresión inexorable de la enfermedad renal crónica, que desencadena en fallo renal. Adicionalmente, la mayoría de pacientes con mutaciones en CLDN19 también desarrollan alteraciones oculares, ya que este se expresa en las células epiteliales de la retina. Existe una gran variabilidad fenotípica entre pacientes, incluso entre aquellos que comparten la mutación c.59G>A; p.G20D en CLDN19 (mutación Hispánica), y entre miembros de una misma familia. Este fenómeno sugiere que, más allá de la mutación causante de la enfermedad, posiblemente existen otros procesos moleculares que determinan la progresión clínica de los pacientes. Bajo esta hipótesis, este trabajo se ha focalizado en el estudio de los factores genéticos y epigenéticos que podrían modular la progresión de la enfermedad renal. Para ello, se han obtenido los datos clínicos, orina y sangre de 30 pacientes afectos de HFHNC y 6 individuos control de diferentes lugares de España, siendo estos la mayoría de pacientes diagnosticados de HFHNC en este territorio. El análisis de los datos clínicos permitió evaluar las diferencias en la evolución de la enfermedad renal en función del sexo y del genotipo, clasificar los pacientes según la pérdida anual de función renal (progresión renal rápida, moderada y lenta), identificar biomarcadores clínicos de pronóstico, y evidenciar la alta variabilidad fenotípica intrafamiliar. El estudio para la identificación de variantes génicas en homocigosis asociadas a los fenotipos más extremos (progresión renal rápida y lenta) se realizó con los datos obtenidos de la secuenciación masiva del exoma de los 30 pacientes de la cohorte. Con esta estrategia, se identificaron 45 variantes génicas. De entre estas, por la función fisiológica de los genes donde se localizan, destacan la rs11207827 (en el gen PATJ) y la rs1050171 (en el gen EGFR). Para determinar los factores epigenéticos implicados en la fisiopatología de la HFHNC y en su progresión, se utilizaron los urinary exosome-like vesicles (uEVs) como fuente no invasiva de información de procesos celulares renales. Mediante microarrays, se analizó el perfil de expresión de los miRNAs contenidos en los uEVs de los 20 pacientes que mantenían los riñones nativos funcionales, identificándose 24 miRNAs diferencialmente expresados en el total de pacientes respecto al grupo control, y 43 en el subconjunto de pacientes homocigotos para la mutación p.G20D en CLDN19. La comparación de pacientes de ambos sexos mostro únicamente la infraexpresión del miR-1915-5p en aquellos de sexo masculino. Finalmente, se identificaron 4 miRNAs diferencialmente expresados en los pacientes con progresión renal moderada, en comparación con el de progresión lenta, y 8 en el subgrupo de pacientes homocigotos para la mutación p.G20D en CLDN19. La biología de sistemas y el uso de redes neuronales de inteligencia artificial han permitido asociar los resultados obtenidos con procesos biológicos cruciales en la fisiopatología de la HFHNC, como la fibrosis renal y el transporte de calcio y magnesio, entre otros. Este trabajo ha permitido incrementar el conocimiento de la fisiopatología de la HFHNC y determinar factores clínicos, genéticos y epigenéticos implicados en la disfunción renal. Desde el punto de vista traslacional, se han identificado nuevos posibles biomarcadores pronósticos de la HFHNC, y dianas terapéuticas que podrían permitir modular la severidad de la enfermedad y, en el mejor de los casos, curarla.
Familial hypomagnesemia with hypercalciuria and nephrocalcinosis (FHHNC) is a rare autosomal recessive tubulopathy caused by mutations in either CLDN16 or CLDN19 genes, which encode for claudin-16 and -19, respectively, expressed in the thick ascending loop of Henle, and involved in paracellular ion transport. This disease is characterized by urinary loss of calcium and magnesium, bilateral nephrocalcinosis, and inexorable progression of chronic renal disease leading to renal failure. Besides, most patients with CLDN19 mutations also develop ocular anomalies, as it is expressed in the retinal epithelial cells. There is great phenotypic variability among patients, even in those who share the c.59G>A; p.G20D mutation in CLDN19 (Hispanic mutation), and also among members of the same family. This phenomenon suggests that, beyond the mutation causing the disease, other molecular processes could determine the clinical progression of patients. Under this hypothesis, this work has focused on the study of genetic and epigenetic factors that could modulate renal disease progression. For this purpose, clinical data, urine and blood from 30 patients affected by FHHNC and 6 control individuals were collected from around Spain, being them the majority of patients diagnosed of FHHNC in this territory. The analysis of the clinical data allowed evaluating the differences in the evolution of the renal disease according to sex and genotype, classifying patients according to the annual decline of renal function (fast, moderate and slow renal progression), identifying clinical biomarkers of prognosis, and evidencing the high intrafamilial phenotypic variability. The study for the identification of gene variants in homozygosis associated with both extreme phenotypes (fast and slow renal progression) was carried out with data obtained from exome sequencing of the 30 patients included in the cohort. With this strategy, 45 gene variants were identified. Among these, due to the physiological function of the genes where they are located, the rs11207827 (in the PATJ gene) and the rs1050171 (in the EGFR gene) were highlighted. To determine the epigenetic factors involved in FHHNC physiopathology and disease progression, the urinary exosome-like vesicles (uEVs) were used as a non-invasive source of information on renal cellular processes. Microarray technique was used to analyze the expression pattern of miRNAs contained in uEVs of the 20 patients who maintained functional native kidneys. Twenty-four miRNAs were identified differentially expressed in all patients when compared with the control group, and 43 in the subset of patients homozygous for the p.G20D mutation in CLDN19. The comparison of patients of both sexes showed only the under-expression of the miR-1915-5p in males. Finally, 4 miRNAs were differentially expressed in patients with moderate to slow progression of renal disease, and 8 within the subgroup of patients homozygous for the p.G20D mutation in CLDN19. Systems biology and the use of artificial neuronal networks have allowed to associate the results obtained with crucial biological processes in FHHNC physiopathology such as renal fibrosis and calcium and magnesium transport, among others. This work, through two complementary strategies, has allowed increasing the knowledge of the physiopathology of FHHNC and, for the first time, determining clinical, genetic and epigenetic factors involved in renal disease progression. From a translational perspective, this thesis has identified new possible prognostic biomarkers of FHHNC as well as novel therapeutic targets that could allow modulating the severity of the disease and, in the best case, to cure it.

This study examines the problems of constructing and administering a test of spoken English for Vietnamese EFL teacher-trainees. In an attempt to standardize the assessment, a planned oral interview was pilottested with a group of ten Vietnamese EFL teachers currently enrolled in a Graduate Diploma Course in TESOL at the Canberra College of Advanced Education, Australia. Results of the study indicate that the validity and reliability of such measurement can be achieved if certain carefully outlined procedures in planning the test and training the testers are carefully followed. Given the close relationship between testing and teaching, it is suggested in this study that there could be an improvement in the teaching of spoken English to Vietnamese EFL teacher-trainees if (i) the amount of time allocated to testing oral proficiency in the curriculum was increased, (ii) Vietnamese EFL teachers were provided with formal training in language test construction, and (iii) research on EFL oral testing was encouraged. Further, this study recommends co-operation between TEFL institutions in Vietnam to develop standard instruments for the assessment of spoken English of EFL teacher-trainees on a national level.

L'utilisation de nouveaux fluides frigorigenes respectueux de notre environnement remet en cause les connaissances etablies pour le dimensionnement des installations frigorifiques et de conditionnement d'air. Certains de ces fluides - les melanges non azeotropes - presentent lors de leur condensation des particularites (condensation anisotherme, apparition d'une resistance au transfert de chaleur et de masse dans la phase gazeuse) qui rendent delicat le dimensionnement des condenseurs. L'etude experimentale de la condensation de fluide pur hfc134a et de differents melanges de hfc134a-hfc23 dans un condenseur de type tema x a permis de caracteriser les echanges de chaleur sur quatre types de tubes (un tube lisse et trois tubes a ailettes). Le coefficient d'echange du melange binaire hfc134a-hfc23 est tres different de celui du fluide pur hfc134a. La comparaison des donnees experimentales de la condensation du fluide pur avec les differentes correlations de la litterature a determine le choix le mieux adapte au dimensionnement des condenseurs de ce type. Les mesures effectuees lors de la condensation du melange non azeotrope ont ete confrontees a deux modeles classiques de la condensation de melange : le modele de la courbe d'equilibre et la theorie du film. Pour la modelisation du premier tube du faisceau, le modele de la courbe d'equilibre, pourtant plus simple, donne d'aussi bon resultats que la theorie du film. Par contre, les echanges de chaleur dans le faisceau lors de la condensation de melanges sont difficiles a prendre en compte avec les correlations classiques.

碩士
義守大學
資訊工程學系
90
Providing QoS guarantees in Hybrid Fiber Coaxial (HFC) networks is a very important and challenging problem. Medium Access Control (MAC) protocol and upstream channel scheduling algorithm are two key mechanisms that will affect the degree of QoS guarantees of HFC networks. Due to the long round-trip propagation delay, uncertain request access delay, and resource reservation without time information, the upstream channel scheduler in the headend cannot obtain the real cell arrival time at the stations. Thus, with the same input traffic, the cell transmission order in HFC networks would be different to the order in traditional switching networks. Therefore, this thesis proposes a new MAC protocol, called Time-stamp General-Contention Resolution Algorithm (TG-CRA), to provide the cell arrival time information to the upstream scheduler in the headend. In order to reduce the overhead of the time information, a linear offset scheme is used to stamp the arrival time information of new arrival cells on the request. Compare to the G-CRA algorithm, simulation result shows that the TG-CRA algorithm can yield better performance in reducing the uncertain request delay and cell access delay. The variance of cell access delay is also reduced, thus the TG-CRA algorithm can provide better mean delay jitter.

碩士
國立交通大學
資訊科學學系
86
To provide residential users with multimedia interactive services, ahigh bandwidth access network solution is required. In this work, we focus on using existing CATV networks to provide high bandwidth interactive services. We first introduce the HFC(Hybrid FiberCoax) architecture of CATV network adnd idscuss the problems in providingtwo way communication over HFC. Second, we discuss the important issues in upstream channel protcol design: 1) overlapping the collision resolution anddata transmission to archive best usage of bandwidth, in which we examine how to arrange the request mini-slots and data transmission; 2) resolving collision, in which we discuss several traditional collision algorithms; 3)resolving the synchronization problem, in which we propose our solution to getthe RTC(Round Trip Correction) parameter. Third, we introduce several proposed IEEE 802.14 protocols for upstream channel and compar these protocols.

碩士
國立臺北科技大學
能源與冷凍空調工程系碩士班
101
The study intends to select from the market a compressor that meets the operation specifications of HFC-245fa refrigerant, and use the compressor to achieve effectiveness of HFC-245fa. After experiments are conducted, the study employs theoretical analysis and actual compression system to compare the difference among refrigerating and heating capacities, COPL and COPH under different evaporation temperatures and condensing temperatures. Research results show that HFC-134a refrigerant is most similar to HFC-245fa refrigerant. The study conducts actual simulation tests and measurement of these 2 refrigerants under 9 conditions (TYPE 1 ~ TYPE 9) that use 3 different condensing temperatures 7.2oC, 12.2oC, 17.2oC (HFC-134a) and 49oC, 55oC, 60oC (HFC-245fa) and 3 different evaporator temperatures 35oC, 45oC, 55oC (HFC-134a) and 80oC, 92oC, 103oC (HFC-245fa). Test results show that CAPH, CAPL, COPH and COPL of HFC-245fa are 74%~85% of those of HFC-134a. As to compressor, compressor power loss actually produced by HFC-245fa is 110~112% of that of HFC-134a. These results show that overall efficiency in terms of COP can reach as high as 80%; power loss of compressor increases by 10%; and its operating state was close to that of HFC-134a. Therefore, the compressor can be selected to load HFC-245fa.

碩士
國立交通大學
資訊科學學系
84
While ATM (Asynchronous Transfer Mode) serves as the backbone of our NII (National Information Infrastructure), the prevailing cable TV networks may serve as the NII community networks that span the infrastructure into the homes. However, in order to support these services to the residential users, the community cable TV networks have to be redesigned to support 2-way interactive communication. In this work, we propose PCUP (Pipelined Cyclic Upstream Protocol) as the upstream MAC (Medium Access Control) protocol for HFC (Hybrid Fiber Coax) community access network. PCUP is designed with the intention of pipelining the upstream channel. This is achieved by proper station positioning, which measures the station propagation offset from the headend, and transmission scheduling, which assigns each station the transmission starting time and duration in a cycle. By taking into account the propagation offsets and the transmission times, transmitted cells can appear back-to-back, i.e. pipelined, at the headend. Since only the active stations are scheduled to transmit in a cycle, a membership control mechanism, which runs a contention-based tree walk algorithm, is executed periodically to allow the stations to join or leave. We evaluate the behavior of our protocol through simulations to show the effectiveness with respect to loss ratio, throughput, delay, and setup latency. Comparisons with R-ALOHA (Reservation Slotted ALOHA) and CSMA/CD are also made. The results confirm PCUP's excellent performance.

碩士
國立交通大學
電信工程系
87
In this thesis, we have designed and implemented EDFAs for HFC networks applications. We first studied the properties of EDFA, especially the gain model and gain dynamics. Finally, we designed and implemented two kind of EDFAs, one for power amplifiers in the head end and another one is suitably used with DWDM techniques.

This master thesis deals with methods of automatic recognition of bird species by their voices. In first, I defined the database of records and created a reference data by handmade evaluation. The next step is to find the optimal features for describing a bird singing. I use a Human Frequency cepstral Coefficients (HFCC). For the best accuracy of recognition is necessary to correctly classify a bird's vocalization from a non-vocalization segments. The VAD system is based on an algorithm k-Nearest Neighbours. The last step describes the system based on Hidden Markov Models which allows to recognize the concrete bird species from the parts of bird's singing.

碩士
元智大學
生物科技與工程研究所
99
A Pseudomonas strain TeU resistant to tellurite (TeO32-) and cadmium (Cd2+) ions was isolated from heavy-metal contaminated sediments by enrichment. Transposon mutagenesis of strain TeU resulted in mutants exhibiting Cd2+ sensitivity (Strain BU21) and one with decreased ability to reduce tellurite (strain AU08). Genes encoding an HflKC complex and an putative metallopeptidase were identified to be associated with the bacterium’s ability for tellurite reduction and cadmium resistance, respectively. We have cloned and sequenced the hflC and metallopeptidase gene of strain TeU. Metallopeptidase gene of strain TeU was cloned and expressed in E. coli in order to investigate if the protein could improve the bacterium’s ability to resist cadmium.Transmission electron microcopy was also employed for to observe metal precipoitation when strain TeU grown in the medium containing tellurite.

碩士
國立臺灣科技大學
化學工程系
90
The purpose of this paper is to investigate the two-step growth of MOCVD copper using Cu(hfac)2 as the precursor. In the first step, we pre-deposit Cu2O films with water as the additive because precursor’s partial pressure is unstable during deposition without water additive. In contrast to H2, H atoms which are produced due to dehydrogenation reaction of ethanol alcohol can be used to reduce Cu2O films at lower temperatures and enhance the reaction rate .The Cu atom in Cu2O films(24％) is the catalyst. The mixture of ethanol alcohol and water can be used to increase Cu atoms in Cu2O films (50％). More Cu atoms in the Cu2O films can cause the increase of nucleation sites, decrease of grain size, and then increase of grain density.

The demands facing the refrigeration industry are minimal usage of conventional energy sources for compression and avoidance of ozone depleting substances. One of the approaches to combat these issues is the use of thermally driven solid sorption compression with non-ozone depleting refrigerant. In this context, the research work presented in this thesis is devoted to a comprehensive thermodynamic analysis and development of a laboratory model of an activated carbon+ HFC 134a adsorption refrigeration system. The cooling load catered to by the laboratory model is 2-5 W, mainly for thermal management of electronics. A complete thermodynamic analysis is carried out for the desorption temperatures varying from 75 to 90 oC, evaporating temperatures from -20 to 15oC and adsorption/condensing temperatures from 25 to 40 oC. A program on MatLab platform is developed for theoretical modeling. A new concept of thermal compression uptake efficiency (u) which is analogous to volumetric efficiency of a positive displacement compressor is introduced to consider the effect of void volume. The thesis also covers an investigation of two-stage and hybrid (thermal+ mechanical) cycle compression systems. It is possible to identify the conditions under which a two-stage gives a better performance than a single-stage one. It also shows that hybrid cycle system gives the best performance and saves ~40% of power compared to operation under the same conditions run with a single-stage mechanical compression refrigeration system. A heat transfer analysis of the thermal compressor is carried out to evaluate non-uniformities in bed temperature. As a part of it, the thermal conductivity of the bed under adsorbed state has been calculated. A laboratory model of activated carbon+ HFC 134a adsorption refrigeration system is fabricated to meet a 2-5 Watts cooling load based on the results from theoretical calculations. Experimental results show a fair match in the trends for the COP with analysis. The main aim of the research was to examine how effective the adsorption refrigeration system is in reducing the temperature rise of the heater used to simulate the electronic component. The heater that would have stabilized at 81, 97, 103 and 112 oC without any cooling for heat inputs of 3, 4, 4.4 and 4.9 W, respectively, would attain a cyclic steady state around 24, 26, 28, 31 oC. The influence of cycle time on the performance of the systems is also investigated. It is concluded that an activated carbon+ HFC 134a adsorption refrigeration system can be a good supplement to conventional compression refrigeration systems. In situations where heat recovery imminent this system could be a good choice. For waste heat recovery and suppression of infrared signatures of electronic components, it is ideally suited where COP becomes immaterial.

碩士
國立交通大學
資訊科學系
88
Data-Over-Cable Service Interface Specifications v1.1 (DOCSIS1.1), developed for data transmission over the Hybrid Fiber Coaxial (HFC) networks, defines five upstream services to support per-flow Quality of Services (QoS). The cable modem termination system (CMTS) has to periodically grant, based on the QoS parameters, upstream transmission opportunities to the QoS flows. However, packets may be dropped while several flows demand the same interval for transmission. In this paper, we propose the two-phase, the scheduling sequence determination phase and the minislot assignment phase, minislots scheduling algorithm to reduce the QoS violation rate. In the scheduling sequence determination phase, the flow whose packets are with the highest probability of being dropped is scheduled first. Then, in the minislot assignment phase, the scheduler allocates the interval with minimal probability of QoS violation. The simulation results demonstrate that our scheduling algorithm can reduce QoS violation rate up to 70% compared with first-come-first-serve-random-selection algorithm. The two-phase minislots scheduling algorithm can work within the framework of DOCSIS1.1 standard.

碩士
國立臺北科技大學
能源與冷凍空調工程系碩士班
102
The main purpose of this study is to analyze and discuss the effectiveness of the green refrigerant HFC-152a as a replacement for the refrigerant HFC-134a. Two experiments were conducted, involving one refrigerator and one dehumidifier, respectively, both of which are commercially available and vapor compression cycle system based. In the experiments, these appliances were initially tested for refrigerant performance with their original HFC-134 refrigerant systems and then tested with the alternative refrigerant HFC-152a. Subsequently, analysis was performed to compare the differences in performance between the two types of refrigerant systems.   Experiment results show that in the experiment where 30g HFC-152a replaced the originally used 50g HFC-134a in one 95L household refrigerator without changing system components and replacing the refrigeration oil, HFC-152a exhibits overall performance close to that of the refrigerant R134a and superior performance with about 11% less in average power consumption and a significant decrease of about 40% in refrigerant filling amount relative to the refrigerant HFC-134 originally used in the vapor compression cycle system. In the other experiment where 100g HFC-152a replaced the originally used 125g HFC-134a in one household dehumidifier with 6L dehumidification capacity, the system exhibits optimal overall performance exhibits optimal performance across parameters with about 20% less in refrigerant filling amount, an increase of about 7% in dehumidification capacity and about 5% less in average power consumption. All of these results prove that HFC-152a is able to replace HFC-134a.

碩士
國立臺灣大學
電機工程學研究所
89
Hybrid-Fiber Coaxial networks have emerged as one of the premier technologies to deliver broadband access to the homes, and the lots of applications on it using TCP as transport layer protocol. In this thesis, we investigate the effects of HFC MAC control protocol on TCP performance and propose a new adaptive FCFS MAC scheduling and allocation algorithm which comprising "Fast Request Transmission" and "Long Packet Deferring" mechanisms, which can help improve the TCP performance substantially in terms of downstream throughput, upstream channel access delay and give good balance on downstream and upstream traffic. In addition, we also introduced an analytical model for TCP Reno on DOCSIS HFC networks, which can analyze and quantify the TCP performance characteristics on DOCSIS, and we developed a DOCSIS v1.1 MAC layer module for ns2, which can be further used to conduct simulation using different traffic sources on DOCSIS HFC networks. The mechanism "Fast Request Transmission" mainly designed to degrade the asymmetric characteristics in HFC networks for improving the downstream throughput and shorten the upstream access delay, while "Long Packet Deferring" mechanism is mainly designed to balance the downstream TCP traffic and upstream TCP traffic for achieving good and fair TCP performance in both direction.

博士
國立臺灣科技大學
化學工程系
97
The feasibility of using a metal-organic chemical vapor deposition (MOCVD) technique with hexafluoroacetylacetonate-copper(I)- cycloocta-1,5-diene (hfac)CuI(COD) as a precursor to achieve the deposition of a thin and conformal copper film was examined. Self-synthesis of [(hfac)CuI(COD)] and characterization with H1-Nuclear Magnetic Resonance Spectrometer (H1-NMR) and Fourier-Transform Infrared Absorption Spectrophotometer (FTIR)、differential scanning calorimetry (DSC) where performed and (hfac)CuI(COD) was confirmed to be the product. The kinetics of disproportionation reaction of (hfac)CuI(COD) was investigated by the use of DSC with different heating rates in dynamic nitrogen atmosphere. First, the activation energies (Eas) of the disproportionation reaction were estimated with model-free isoconversional methods respectively. The Eas were found to fall within the range between 17.6 and 18.7 kJmol-1, with no temperature and heating rate effects observed. Then, when the Ea was ascertained, the model-fitting methods with least square fitting procedure were adopted to determine the kinetic model for the disproportionation reaction. As a result, the disproportionation reaction follows second-order reaction kinetics. Second, the deposition of thin films on TaN0.46 (-Ta(N)(bcc))、TaN0.82(f.c.c-TaN) and TaN1.12 (amorphous+f.c.c-TaN) substrates was achieved in a cold-wall CVD reactor. On TaN0.46 substrate, the surface morphology of the Cu growth at deposition temperature 140, 190 and 230℃ were followed Volmer-Weber type to form a smooth and continuous film in 5minute of deposition time. On TaN0.82 substrate, the surface morphology of the Cu growth at deposition temperature 190℃ were followed Frank-Van der Merwethe type. It was found that the flatter and denser copper films, and stronger preferred orientation of Cu(111) were obtained after 5minute of deposition time. On TaN1.12 substrate, the surface morphology of the Cu growth at deposition temperature 190℃ were also followed Frank-Van der Merwethe type and formed the flatter and denser copper films, and stronger preferred orientation of Cu(111) after 7min. Then, the deposition of Cu thin films on TaN0.46、TaN0.82 and TaN1.12 trenches were investigated. On TaN0.46 substrate, a continuous and good step coverage Cu film was obtained at ambient deposition temperature. On TaN0.82 and TaN1.12 trenches, the deposition temperature were decreased down to 140℃. Because the growth rate of copper at 140℃ was followed the surface reaction limited regime, that could be formed smooth and continuous thin copper films. Third, growth kinetics of copper films with MOCVD reaction system using (hfac)CuI(COD) as the precursor was studied. In this research, the kinetic data of MOCVD Cu thin films as a function of deposition temperature and partial pressure of precursor were investigated. It was found that the growth rate of copper between 140~190℃ was written surface reaction limited regime with the value of activation energy as 14.82kJ/mol. Through the analysis on the growth kinetics, the kinetic model and mechanism of chemical vapor deposition as follows: Step1: Adsorption of Step2: Disproportionation reaction to form copper and Cu(hfac)2 And the reaction model: Finally, a novel electro-enhanced metalorganic chemical vapor deposition (EEMOCVD) technique for producing copper (Cu) thin films on TaN1.03/Si substrates with (hfac)CuI(COD) as a precursor was investigated in this research. This novel technique features supplying a direct current (DC) to TaN1.03/Si substrates while the deposition of Cu thin films is in progress. Experiments on EEMOCVD yielded fortuitously positive results: (1) the deposited Cu films were superior in quality; and (2) the growth rate of Cu film deposition increased. The above results are more desirable than those achieved through the conventional MOCVD (CMOCVD) technique. The proposed EEMOCVD technique hence proves to be more effective in forming smooth and continuous thin copper films.

碩士
逢甲大學
資訊工程所
90
With the development of Internet, the demand of bandwidth from user is more and more. To evaluate these resolving plans and the current environment of Taiwan, CATV network is a good resolving plan indeed. In recent standards of CATV, DOCSIS 1.1 is a generally acknowledge standard. There are many mechanisms of standard ensuring QoS (Quality of Service) and the idea of priority. We can promise that those higher priority users could get the higher transmitting performance through giving the different priority to different Cable Modem (CM). We can achieve some level of QoS to protect the base transmitting demand of QoS to users with different level. Data Over Cable Service Interface Specifications v1.1 (DOCSIS 1.1), developed for data transmission over the HFC networks, defines five upstream services to support per-flow QoS. However, it doesn’t define how to schedule these upstream services. In this paper, we study the relationship between DOCSIS 1.1 and QoS, and proposed some advanced methods. In our scheme, the five upstream services are classified into two services — periodic services and aperiodic services. The periodic services are time-sensitive services, and they must provide bandwidth and access delay guarantee. Then the proposed admission control mechanism can handle them. Finally, using EDF、AWFQ algorithm to schedule all services in system effectively and fairly. The analysis and simulation results confirm that the throughput and delay are better than several proposed scheme.

碩士
國立成功大學
資訊工程研究所
88
Hybrid Fiber Coaxial (HFC) system provides a bi-directional cable TV network. HFC network has several characteristics：long propagation delay, indirect collision detection, asymmetric bandwidth, and non-uniform user distribution. Hence a good contention resolution algorithm is important and necessary in HFC networks. For obtaining a better performance, the contention resolution of HFC networks is usually divided into first transmission rule (FTR), which determines when to transmit for new requests, and collision resolution algorithm (CRA), which determines how to resolve the collisions. In IEEE 802.14 Draft, Admission Time Boundary (ATB) is used as the standard of FTR and Q-ary tree algorithm is used as the standard of CRA. In this thesis, a novel contention resolution algorithm, queue-based algorithm, is proposed. By using a queue, collided queue, all collisions in the pervious period are all recorded. The headend then schedule the consequent transmissions according on this collided queue. A measurement-based ATB setting is proposed. It can estimate a more correct number of new requests and set a more proper ATB. From simulation results, its performance is the closest to the ideal among all ATB setting algorithms. Finally, the effect of Q in the Q-ary tree algorithm is also investigated. A dynamic Q setting is necessary when a poor ATB setting algorithm is used.