Дисертації з теми "Liquid software"

The development of high quality metal products requires "clean" liquid metals as the base materials. For a large number of application there is a need to quantify the cleanliness of the liquid metals in the sense that the number and size of inclusions have to be controlled to be below some acceptable limits. The demand for quality helped the development of measuring systems that can count the number and size distribution of inclusion. One of the devices, called LiMCA ($ rm underline Liquid underline Metal underline Cleanliness underline Analyzer),$ was developed at McGill University and has been successfully used in the aluminum industry for a number of years.
The LiMCA apparatus is based on the Electric Sensing Zone principle. By maintaining a constant current through a small orifice through which liquid metal passes, non-conductive particles passing through the orifice temporarily increase the electrical resistance of the orifice, therefore increasing the electric potential. The signal processing component of the LiMCA system detects the voltage transients, translates them into particle sizes, and counts them based on their sizes or stores the transients in certain time increments.
The current LiMCA system uses analog electronic components to implement the signal processing part and describes a ransient only by its height or its time of occurrence. This implementation has limited the further development of the system for applications where the particle size distribution and particle occurrence must be counted concurrently.
Digital Signal Processing (DSP) technology has been successfully applied to upgrade the LiMCA system. With this technology, the DSP-based LiMCA system is able to describe each LiMCA transient by a group of seven parameters and to classify it into a certain category with the help of these parameters. Moreover, it counts the classified peaks based on their height (Pulse Height Analysis) and their time of occurrence (Multi-Channel Scan) concurrently for data acquisition.
A conceptually new software was designed to accommodate the DSP-based LiMCA and the Object Oriented Programming technique was used to develop the Graphical User Interface which constitutes the framework of the overall host interface.

The oscillating droplet technique is an experimental technique that is used to measure the surface tension and viscous damping coefficients of a liquid droplet. This technique has been the subject of much analysis; theoretical, numerical, and experimental with a number of different external forces used to confine the droplet. These external forces are found to modify the oscillation frequency and damping rates, which need to be quantified in order for the measurement technique to be used. The dynamics of the droplet are three-dimensional but previous numerical work has largely focused on axisymmetric cases. This work uses numerical techniques to extend the previous analysis to include the full three-dimensional effects. In this work a three-dimensional numerical model is designed, developed and applied to study the dynamics of a liquid droplet both in free space and with a high DC magnetic field used to balance gravitational forces. The numerical model is a grid point formulation of the pseudo-spectral collocation method discretised in a spherical coordinate system with the implicit Euler method used to advance the solution in time. A coordinate transformation method is used to ensure the direct surface tracking required for modelling the surface shape oscillations. The study covers the laminar fluid flow regime within a droplet exhibiting translational and surface shape oscillations providing a greater understanding of the physical behaviour of the droplet along with a qualitative and quantitative comparison with theoretical results. Initially a droplet oscillating in free space is considered, with a range of surface oscillation modes used to demonstrate the three-dimensional dynamics. Then the influence of electromagnetic forces on a diamagnetic droplet is studied, this includes the field from a solenoid magnet used to levitate a diamagnetic droplet. Finally the dynamics of an electrically conducting droplet in an external static magnetic field are modelled. In each case a number of methods are used to analyse the surface displacement in order to determine the surface tension and viscous damping coefficients. The numerical study of a freely oscillating droplet shows good agreement with the low order theoretical results for droplets in the limit of low viscosity. The high accuracy of the surface tracking method allows the non-linear effects of mode coupling and frequency shift with amplitude to be observed. There is good agreement with the theoretical values available for inviscid axisymmetric oscillations and the numerical study provides the opportunity to determine these effects for three-dimensional viscous oscillations. The magnetic field from a solenoid is used to study the levitation of a diamagnetic droplet and the oscillation frequencies of the droplet are compared with a theoretical model. The magnetic field is analysed and the accuracy of the field calculation used when determining the modification to the oscillation frequencies is considered with the use of a theoretical model. Analysis is made into the splitting of the frequency spectrum due to the magnetic field. The theoretical model that is available for an electrically conducting droplet in a static magnetic field predicts different fluid flow within the droplet and oscillation frequency and damping rate changes. These changes are compared qualitatively and quantitatively with the numerical model results with good agreement.

Thesis (MSc. (Hons.))--University of Western Sydney, 2004.
A thesis presented to the University of Western Sydney, College of Science, Technology and Environment, School of Science, Food and Horticulture, in fulfilment of the requirements for the degree of Master of Science (Honours). Includes bibliographies.

In modern chemical laboratories, high performance liquid chromatography (HPLC) has become the technique of choice amongst separation scientists as evidenced by the huge body of published literature that employ the technique as the pre-eminent analytical method. Increasingly analytical chemists are seeking to determine the chemical signature (or fingerprint) of complex and perhaps even simple samples. The analysis of environmental samples, natural products, such as essential oils, the evaluation of illicit drugs, or the quality assurance of pharmaceuticals are just four areas where chemical fingerprinting may be important. The chemical classification of these types of mixtures can be undertaken using chromatographic methods of analysis coupled with detection processes that provide identification ie. mass spectroscopy (MS), infrared spectroscopy (IR), atomic spectroscopy (ICP) and nuclear magnetic resonance spectroscopy (NMR). While these techniques find widespread use and their applications are growing at a rapid rate, their limitations as routine applications for determining chemical signatures, is limited by the initial purchase price and the high associated running costs; especially for hyphenated liquid chromatographic systems. In general, a high degree of operator expertise is also required to ensure correct operation and interpretation of the results. While complete chemical specification may be undertaken by using multiple hyphenated methods, none of the above mentioned techniques provide for a rapid means of detection on a routine or continuous mode of operation. This can limit their application. However, an alternative to monitor and evaluate the chemical signature of complex samples is to employ multidimensional separations. In the expanded multidimensional separation space the probability that two species will elute with exactly the same retention time in both separation dimensions decreases compared to the one-dimensional separation. This probability further decreases as the separation mechanisms become more divergent and the two-dimensional separation space is maximised. Hence, the uniqueness of a two-dimensional retention time increases and thereby the multidimensional separation approach becomes a means of chemical fingerprinting. Hence 2D-HPLC technique is a powerful and less expensive method of separation compared to hyphenated techniques. This thesis describes the development of a stand-alone 2D-HPLC system for the separation of complex samples. A comprehensive 2D-HPLC software package has been developed for data acquisition, hardware control, data processing and graphical presentation. Two data acquisition hardware modules have also been developed. These hardware and software modules have been integrated with existing equipment in our laboratory. The performance of this newly developed 2D-HPLC system has been successfully evaluated. All these details have been described in the five chapters of this thesis.

Approved for public release, distribution unlimited
The Next Generation Internet (NGI) will address increased multi-media Internet service demands, requiring consistent Quality of Service (QoS), similar to the legacy phone system. Server Agent-based Active network Management (SAAM) acts like a rush-hour traffic reporting helicopter. Upon routing request arrivals, SAAM server determines the best, least traffic/resistance route and assembles the routing path, freeing up ¡ʹlight-weight¡· routers to provide faster, more reliable, forwarding services. The SAAM server is a critical network node; therefore, it is imperative to make it extremely robust. With Margulis Agent-Based Mobile Application (MAGMAà EÌ ) liquid software, a SAAM server agent will remain inactive in resident memory of each router until it is stimulated by a message from the departing server. Then the agent will begin running a new server at a starting point determined from the prior server¡Šs recent state information or a pre-determined point if that state information is not available. MAGMAà EÌ will provide SAAM an increased fault tolerance and security against malicious attacks. Liquid software research has taken place since 1996 (University of Arizona/University of Pennsylvania); however, there is no known application currently providing fault tolerance and system security. In this thesis, the foundation for a mobile SAAM server was developed, with the researcher being able to manually move the server from one host to the next. Furthermore, this thesis designed a protocol thatcompresses critical state information, providing condensed messages to efficiently configure the next SAAM server across the network with the state information from the departing server extracts critical state information from the current server and periodically transports a compressed form of the state information to potential next SAAM servers in the network. MAGMAà EÌ will provide a revolution in today¡Šs computer fault tolerance and security paradigms, benefiting industry through more survivable networks with guaranteed QoS.

Cellply è un'azienda del settore biomedicale che sta sviluppando una piattaforma IVD (In-Vitro Diagnostic) in grado di fornire un'analisi personalizzata testando la risposta di farmaci antitumorali su campioni biologici prelevati da pazienti. Il cuore del sistema è un Lab-on-a-chip che si basa su una tecnologia microfluidica brevettata, denominata Open Microwell, che permette di gestire e analizzare i campioni di cellule tramite elaborazione di immagini in modo standardizzato e automatizzato[1]. Per lo sviluppo della piattaforma di Cellply si è reso necessario sviluppare un sistema di liquid handling, in modo da automatizzare e rendere più precisa e ripetibile la gestione dei campioni biologici. La parte fondamentale del sistema di liquid handling è un pipettatore automatico, cioè un componente in grado di automatizzare tutti quei procedimenti che, in un laboratorio biologico, vengono tipicamente realizzati tramite pipette manuali. Il pipettatore viene montato su un sistema meccanico multiassiale (XY Portal) in grado di posizionarlo con precisione in un'area di lavoro (working area) ben definita, all'interno della quale avviene il trasferimento dei liquidi. Il lavoro di tesi ha riguardato lo sviluppo e l'implementazione del software di controllo dell'XY Portal. La progettazione è stata realizzata sfruttando l'ambiente di sviluppo LabVIEW per il linguaggio di programmazione grafico di National Instruments. Dopo una prima fase in cui sono stati definiti i requisiti e analizzati i relativi rischi, si è passati allo sviluppo vero e proprio della libreria software per poi terminare con l'esecuzione dei test necessari alla verifica della corretta implementazione dei suddetti requisiti. La libreria software è stata progettata seguendo le specifiche di progetto necessarie all'integrazione della stessa all'interno del software principale di Cellply che gestisce le varie risorse hardware che compongono il dispositivo diagnostico in fase di sviluppo.

Em inúmeras aplicações práticas da engenharia a camada interfacial entre dois fluidos é de extrema importância para a correta caracterização do escoamento. Especificamente na Engenharia Ambiental e Hidráulica na interface entre o ar e a água existem propriedades capazes de relacionar a aeração e oxigenação. Focado na dinâmica dos processos turbulentos, o estudo proposto se aplica a problemas de sub-pressão em vertedores e visa analisar a troca de gases da interface ar-água em situações de fases dispersas e avaliar a qualidade da água a jusante dos mesmos. O objetivo maior é descrever a relação entre o comportamento da superfície da água, o desenvolvimento da camada limite e os regimes de escoamento de acordo com a vazão volumétrica. Para representar a distribuição da fração de vazios da fase dispersa, foram usados os princípios da formulação dos escoamentos em vertedores escalonados e da formulação para enlaçamento de ar. As argumentações e justificativas das formulações propostas para os perfis de concentrações foram baseadas na distribuição da fração de vazios prevista numericamente. A abordagem numérica do problema fluidodinâmico diferencia da maioria dos estudos concentrados na literatura, por estar baseada na quebra da superfície livre.
In many pratical aplications of engineering the interfacial layer between two fluids has a strong relevance in the process of right measurements of the flow. Espeficically in the enviromental and hydraulics engineering the interface between air and water is the property related with oxigen and areation of the flow. With focus on the dynamics of the turbulent processes, the proposed study is applied to sub-pressure problems in spillways. It investigates the gases exchange in the air- water interface in the dispersed phase and evaluate the water quality downstream. The major goal is to describe a relation between the boundary layer development and the flow regimes with discharge. To represent the distribution of void fractions of the dispersed phase, the physical formulation concepts of stepped spillways and entrapped air were used. The arguments and reasons to justify the proposed concentration profile formulation are based on the numerical provided void fraction distribution. The results as well as the physical model are in good agrement with described literature data. The Numerical approach of the fluid dynamics problem differs from major of described literature studies because it is based on the surface breakup.

The aim of this thesis is to analyse the cryogenic process for cooling superconducting radio frequency accelerator test cavities in the LEP accelerator at CERN. A liquefaction cryoplant is analysed, including the production of liquid helium at 4.5 K, the systems for distribution and regulation of liquid helium, and the radio frequency field used for accelerating particles. After discussing regulation problems and modifications planned for a new cavity installation in 1992, different techniques for specifying the control programs for the new installation are evaluated. Various diagramming techniques, standards and methodologies, and Computer Aided Software Engineering-tools, are compared as to their practical usefulness in this kind of process control. Finally, in accordance with anticipated requirements, possible ways of making high and low level control program specifications are suggested.

Dans la tendance d'Internet, nous commençons à compter nos activités quotidiennes sur les systèmes de base en nuage pour stocker nos données et pour suivre tout notre état et comportement lors de l'utilisation de services Web. Par conséquent, nos données ne sont plus sécurisées comme auparavant, car nous avons délégué la responsabilité de les protéger, de les sécuriser et de les stocker à des tiers. Ces dernières années, de grandes entreprises telles que Yahoo, Facebook ou LinkedIn ont été victimes de plusieurs fuites, ce qui montre à quel point nous sommes vulnérables. Nous disposons aujourd'hui d'appareils mobiles plus puissants que jamais. Cependant, nous n'explorons pas leurs capacités en matière de traitement, de stockage et de mesures de sécurité pour utiliser les services Web localement. De plus, nos appareils sont fondamentalement divisés en deux mondes différents, ceux avec Android et ceux avec un système d'exploitation iOS. Néanmoins, avec la popularité de la technologie mobile, notre comportement a changé. Aujourd'hui, il est courant de trouver un utilisateur utilisant deux ou trois appareils mobiles en même temps, souvent avec un système d'exploitation différent. Par conséquent, nous sommes simultanément confrontés à un environnement à plusieurs périphériques avec le même système d'exploitation ou des systèmes d'exploitation différents, dans lequel la plupart des technologies et des services Web nécessitent le recours à un fournisseur tiers pour la synchronisation et le maintien de l'état de nos tâches quotidiennes. Bien que ce scénario chaotique concerne plusieurs appareils et systèmes d'exploitation, il existe une technologie mentionnée précédemment à la fin des années 90, appelée Liquid Software, qui pourrait être une solution pour harmoniser ce scénario. Le Liquid Software vise à permettre une migration fluide vers différents appareils avec ou sans le même système d'exploitation, tout en préservant l'expérience utilisateur d'une application ou d'un service donné. Dans cette thèse, basée sur les principes de Liquid Software, nous nous concentrons sur un environnement multi-périphériques capable de synchroniser le service Web côté client sans dépendre d'un système de base Cloud. Nous avons proposé une architecture basée sur des modèles, appelée Le CUBE, qui utilise des principes centrés sur l'utilisateur combinés à différentes technologies, telles que les principes REST et RESTful. Le CUBE propose également un modèle à quatre couches dans lequel REST et RESTful fonctionnent indépendamment l'un de l'autre mais en coopération. Un autre élément critique du CUBE est la couche de conversation (CC), modélisée dans le but de permettre la modification des données / états à partir des services Web. La CC utilise des techniques de traitement du langage naturel pour trouver et définir des services Web adaptés à l'utilisateur. Une autre contribution du CUBE est la création d'une sorte de réseau d'utilisateurs appelé INNER CUBE, avec tous les appareils mobiles appartenant à un seul utilisateur. Le but de la combinaison de ces éléments au sein d'une seule unité, le CUBE, est de permettre le transfert fluide de services/données et de sessions d'appareils (ordinateurs de bureau, ordinateurs portables et téléphones mobiles) vers d'autres appareils, quel que soit leur système d'exploitation, et au même temps, entièrement transparent pour l'utilisateur final, tout en réduisant la dépendance du fournisseur tiers. Nous avons également proposé deux tests pour valider le modèle CUBE
Within the trend of the Internet, we start to rely our daily activities on Cloud Base Systems to store our data and keep tracking all our state and behavior while using Web Services. Therefore, our data are no longer safe as before since we have delegated the responsibility to keep them safe, secure, and stored by third-party providers. In the last few years, big companies such as Yahoo, Facebook, or LinkedIn were victims of several leaks, which demonstrates how vulnerable we are. We have nowadays more powerful mobile devices than ever before. However, we do not explore their capacities regarding processing, storage, and security measures to using Web Services locally. Moreover, our devices are basically divided into two different worlds, those with Android, and those with an iOS operating system. Nevertheless, with the popularity of mobile technology, our behavior has changed. Today it is common to find a user using two or three mobile devices at the same time, and often they have a different operating system. Therefore, we are facing simultaneously a multiple-device environment with the same or different operating systems, where most of the technology and Web Services requiring a third-party provider to synchronizing and keep the state of our daily tasks. Although of this chaotic scenario regarding multiples devices and operating systems, there is a technology earlier mentioned at the end of 90's called Liquid Software that may be a solution to bring harmony to this scenario. The Liquid Software aims to allow for fluently migrate towards different devices with or without the same operating system, and at the same time keep the user-experience of a given application or service. In this thesis, based on Liquid Software principles, we focus on multiple devices environment able to synchronize Web Service at the client-side without depending on a Cloud Base System. We proposed the model-based architecture, called the CUBE, that makes use of User-Centric principles combined with different technologies, such as REST and RESTful principles. The CUBE also proposes a four layers model where REST and RESTful work independently from each other but in cooperation. Another critical element of the CUBE is the Conversation Layer (CL), modeled with the purpose to allow change data/state from Web Services. The CL makes use of techniques of Natural Language Processing to find and defining suitable Web Services to the user. Another contribution of the CUBE is creating a kind of user-network called INNER CUBE with all mobile devices that belong to a single user. The aim of combining these elements inside a single unit, the CUBE, is to allow for fluently move services/data and sessions from devices (desktops, laptops, and mobiles phones) to other ones regardless of their operating system, and at the same time, entirely transparent for the final user, and also reducing the dependency of the third-party provider. We have also proposed two tests to validate the CUBE model

Construction d'un dispositif experimental pour etudier l'homogeneisation du cuivre liquide chauffe et brasse par induction et la distribution chimique en cours de reaction avec un jet d'oxygene gazeux. Realisation d'un programme de calcul pour trouver les compositions a l'equilibre interface metal-laitier

This work deals with design of thermal source of panel shape for scientific biological application. Panel has adjustable surface temperature and thermal gradient. The first part of this work concern with design and numerical modeling of the thermal system, application of materials and arrangement of thermal and cooling devices – Peltier elements. Second part of this work is about construction of the panel and experimental measurement.

Cette étude a pour but d'approfondir les connaissances sur les fluctuations turbulentes de vitesse de façon à fournir les données les plus fiables possible pour la modélisation de la turbulence. Une étude expérimentale dans un four à induction est effectuée, il en ressort que l'intensité du courant inducteur contrôlé l'amplitude des vitesses moyenne et turbulente. L'analyse spectrale des fluctuations de vitesse avec un programme utilisant le tfr directe et inverse est proposée.

In this thesis we propose a model for laminar natural convection within a mixture of two cryogenic fluids with preferential evaporation. This full model was developed after a number of smaller models of the behaviour of the surface of the fluid had been examined. Throughout we make careful comparison between our analytical and computational work and existing experimental and theoretical results. The coupled differential equations for the main model were solved using an explicit upwind scheme for the vorticity-transport, temperature and concentration equations and the multigrid method for the Poisson equation. From plots of the evolution of the system, it is found that convection becomes stronger when preferential evaporation is included. This new model demonstrates how to include preferential evaporation, and can be applied to other fluid systems.

In this thesis is described a laboratory process of example of traction control with frequency converter SINAMICS S120. The focus of this thesis is divided into three parts. Modification of the traction to translational motion for transport of the open vessel with liquid. Proposal of the autonomous operation of the traction without connection to other devices such as a personal computer or PLC, including wiring diagram. The simulation of a dynamic movement for transport of the vessel with liquid in the simulation environment MATLAB and implementation using the proposed laboratory traction. Commissioning and activation of each control functions is described step by step.

L'objectif principal de ce travail de thèse concerne l'étude de l' élimination de la DCO et de la nitrification dans une nouveau lit bactérien Multi-Section ( MSB ) . Après une caractérisation de l’hydrodynamique et du transfert d’oxygène de ce lit bactérien, les expériences biologiques menées sous des conditions opératoires contrastées (fortes et faibles charges organiques eteaux usées contenant ou pas des matières particulairs) ont été menées. En parallèle, des simulations avec le logiciel Biowin® ont été réalisées. Les principaux résultats sont résumés en suivant :- La rétention de liquide statique est majoritaire par rapport à la rétention dynamique que ce soit en présence ou en absence de biofilm. Le biofilm joue le rôle d’une "éponge" permettant un maintien de l’humidité du lit même à faible débit. Les expériences de DTS ont montré que le biofilm accroit le temps de séjour du liquide et conduit à une diminution de l’épaisseur du film liquide permettant ainsi de promouvoir le transfert de l'oxygène.- Le réacteur MSB montre une élimination efficace de la DCO (> 95 % ) et de la nitrification ( > 60 % de l’azote entrant), mais une accumulation de DCO particulaire a lieu dans le filtre ce qui conduira à un colmatage à terme. La nitrification cohabite avecl’élimination de la DCO même dans la première section et pour une charge organique élevée ce qui implique une bonne capacité d’oxygénation du MSB par l’aération naturelle.- Un modèle dynamique de MSB a été utilisé implémenté sur le simulateur - BioWin , afin d'obtenir la répartition des biomasses au sein du réacteur et d'évaluer le processus limitant dans chaque section. Le modèle partiellement calibré peut aider à estimer les besoins minimum d'oxygène pour la nitrification et peut rendre compte de la compétition entre la croissance hétérotrophe et la nitrification
The main objective of this PhD work focused on the study of the COD removal and nitrification in a new designed Multi-Section Bioreactor (MSB). Hydrodynamic characterization of the reactor, biological experiments under contrasted conditions and simulations by Biowin® software were carried out:- Firstly, it was found that static liquid retention is the predominant part both without and with the presence of biofilm. Biofilm acts like a "sponge". RTD experiments showed that biofilm can promote liquid residence time, decrease the liquid film andpromote the oxygen transfer consequently.- Secondly, the MSB operated at contrasted organic loading rate (OLRs) and nitrogen loading rate (NLRs) showed that COD can be effectively removed (removal efficiency > 95%) and nitrification (> 60% of the N removal) occurred in this biofilter.Nitrification is efficient even in the first section implying no drastic oxygen limitation though only natural aeration is occurring.- Thirdly, a TF dynamic model has been used from a simulator - BioWin, in order to get more insights on the biomass distribution in the pilot and to assess the limiting process in each section of the bioreactor. Calibration of the model can help us to estimate theminimum oxygen requirement for nitrification for each zone inside the pilot and it can well represent the competition between heterotrophic growth and nitrification

Sélectionner des traitements de surface pour l’industrie nécessite de prendre en compte :les propriétés à conférer au substrat, la nature et la géométrie de celui-ci et les caractéristiques du milieu extérieur. Certaines combinaisons de ces paramètres rendent difficile la sélection d’un traitement unique, d’où le recours à des multitraitements de surface. Dès lors, se posent les questions suivantes :

- Utiliser des multitraitements de surface peut se faire en scindant les différentes requêtes en sous-ensembles, de manière à ce que chaque traitement réponde à l’un d’eux. Dans quel ordre ces requêtes doivent-elles être introduites par rapport au substrat ?

- Comment sélectionner les traitements de surface répondant à chaque requête individuelle ?

- Comment classer des multitraitements en termes d’adéquation au problème posé ?

Dans ce travail, les première et troisième questions sont abordées, en explorant les requêtes concernant habituellement les dispositifs de moulage de l’aluminium :

- Résistance aux contraintes d’origine thermique.

- Résistance à la corrosion par les métaux fondus.

- Résistance au frottement.

L’analyse de la bibliographie relative aux traitements de surface utilisés dans ces systèmes a été analysée et des « architectures »-types ont été identifiées (chapitre 3). On prévoit, par exemple, un traitement conférant la résistance à la fatigue superficielle, ainsi qu’un revêtement étanche et résistant à l’aluminium fondu. Une barrière thermique est parfois préconisée.

Pour chacune des architectures, des traitements de surface individuels peuvent être sélectionnés. Un « facteur de performance » permettant de classer les solutions par rapport au problème de la fatigue thermique a été construit (chapitre 4) et discuté dans deux situations :

- Lorsqu’un revêtement est présent, et que les contraintes d’origine thermique (différence de dilatation thermique couche-substrat) menacent de le rompre lors de l’immersion dans un milieu corrosif à haute température. Des essais de corrosion dans de l’aluminium fondu ont été réalisés sur un acier revêtu par du nitrure de chrome dopé à l’aluminium, synthétisé par déposition physique en phase vapeur (chapitre 5 – collaboration :Inasmet).

- Lorsque des variations thermiques rapides menacent de rompre le substrat et la (les) couches. Des essais de fatigue thermique ont été réalisés sur de l’acier à outils pour travail à chaud non traité, boruré ou recouvert d’un multitraitements (zircone yttriée / NiCrAlY / boruration / acier). Le revêtement en zircone yttriée a été obtenu par projection par plasma. L’essai de fatigue thermique a été modélisé et le facteur de performance, discuté (chapitre 6).

Au chapitre 7, les architectures-types ont été introduites dans une méthodologie de sélection des multi-traitements de surface, qui a été appliquée dans deux cas :

- Celui des moules de fonderie, devant résister à la fatigue thermique et à la corrosion par l’aluminium fondu. Le facteur de performance a été extrapolé à d’autres situations qu’aux chapitres 5 et 6. Les solutions habituellement proposées pour résoudre ce problème sont retrouvées.

- Celui de deux pièces en acier frottant l’une contre l’autre en présence d’aluminium fondu.

To select surface treatments, one must account for the required functional properties, the substrate features and the solicitations the substrate must endure. Certain combinations of these parameters make it difficult to select a single surface treatment, a reason why several successive treatments are preferred. To select them, one needs to determine:

- How to divide the several requests into groups and how to stack up these groups from the substrate to the outer surface, so that each treatment deals with one specific group of requests/properties.

- How to select individual layers for each group of properties.

- How to rank the multi-treatments in terms of relevance for a given application.

In this work, one tries to answer the first and the third questions, by studying the case of aluminium foundry, in which the industrial devices frequently face the following solicitations:

- Thermal stress (thermal fatigue, thermal expansion mismatch).

- Presence of corrosive molten metal.

- Sliding wear.

In the literature, several “standard” architectures are proposed (chapter 3), like a diffusion layer reducing superficial fatigue plus a corrosion barrier layer. A thermal barrier coating is also sometimes proposed.

For each of these architectures, one can select individual treatments. To rank them, one devised a “performance index” for thermal stress (chap.4), which is discussed for two cases:

- For large differences between layer and substrate thermal expansion coefficients, when both are put into contact with a high temperature corrosive medium, the layer may be damaged. One discusses this case by examining the corrosion caused by molten aluminium for a steel substrate coated by anticorrosive chromium nitride doped with aluminium. The layer is produced by physical vapour deposition (chap. 5 – cooperation: Inasmet).

- Repeated fast surface temperature transients can also damage the substrate and/or the layer by thermal fatigue. One conducted thermal fatigue tests with samples of hot work tool steel, respectively untreated, simply borided and protected by a multilayer. In the last case, top coat is yttria stabilised zirconia, followed by a nickel superalloy and then a borided layer (undercoat). One synthesized the zirconia coating by plasma spray and one modelled the thermal fatigue (chap. 6).

In chap. 7, architectures from chap. 2 are introduced in a multi-treatment selection routine, which is applied in two cases:

- Foundry moulds for molten aluminium, withstanding both thermal fatigue and corrosion. The devised performance index is extrapolated beyond the tests of chap. 5 and 6 to treatments for this industrial application, thereby quantifying their respective merits.

- A foundry device exposed to molten metal and sliding wear.

Doctorat en Sciences de l'ingénieur
info:eu-repo/semantics/nonPublished

In modern chemical laboratories, high performance liquid chromatography (HPLC) has become the technique of choice amongst separation scientists as evidenced by the huge body of published literature that employ the technique as the pre-eminent analytical method. Increasingly analytical chemists are seeking to determine the chemical signature (or fingerprint) of complex and perhaps even simple samples. The analysis of environmental samples, natural products, such as essential oils, the evaluation of illicit drugs, or the quality assurance of pharmaceuticals are just four areas where chemical fingerprinting may be important. The chemical classification of these types of mixtures can be undertaken using chromatographic methods of analysis coupled with detection processes that provide identification ie. mass spectroscopy (MS), infrared spectroscopy (IR), atomic spectroscopy (ICP) and nuclear magnetic resonance spectroscopy (NMR). While these techniques find widespread use and their applications are growing at a rapid rate, their limitations as routine applications for determining chemical signatures, is limited by the initial purchase price and the high associated running costs; especially for hyphenated liquid chromatographic systems. In general, a high degree of operator expertise is also required to ensure correct operation and interpretation of the results. While complete chemical specification may be undertaken by using multiple hyphenated methods, none of the above mentioned techniques provide for a rapid means of detection on a routine or continuous mode of operation. This can limit their application. However, an alternative to monitor and evaluate the chemical signature of complex samples is to employ multidimensional separations. In the expanded multidimensional separation space the probability that two species will elute with exactly the same retention time in both separation dimensions decreases compared to the one-dimensional separation. This probability further decreases as the separation mechanisms become more divergent and the two-dimensional separation space is maximised. Hence, the uniqueness of a two-dimensional retention time increases and thereby the multidimensional separation approach becomes a means of chemical fingerprinting. Hence 2D-HPLC technique is a powerful and less expensive method of separation compared to hyphenated techniques. This thesis describes the development of a stand-alone 2D-HPLC system for the separation of complex samples. A comprehensive 2D-HPLC software package has been developed for data acquisition, hardware control, data processing and graphical presentation. Two data acquisition hardware modules have also been developed. These hardware and software modules have been integrated with existing equipment in our laboratory. The performance of this newly developed 2D-HPLC system has been successfully evaluated. All these details have been described in the five chapters of this thesis.
Master of Science (Honours)

Commodity brain-computer interfaces (BCI) are beginning to accompany everything from toys and games to sophisticated health care devices. These contemporary interfaces allow for varying levels of interaction with a computer. Not surprisingly, the more intimately BCIs are integrated into the nervous system, the better the control a user can exert on a system. At one end of the spectrum, implanted systems can enable an individual with full body paralysis to utilize a robot arm and hold hands with their loved ones [28, 62]. On the other end of the spectrum, the untapped potential of commodity devices supporting electroencephalography (EEG) and electromyography (EMG) technologies require innovative approaches and further research. This thesis proposes a modularized software architecture designed to build flexible systems based on input from commodity BCI devices. An exploratory study using a commodity EEG provides concrete assessment of the potential for the modularity of the system to foster innovation and exploration, allowing for a combination of a variety of algorithms for manipulating data and classifying results. Specifically, this study analyzes a pipelined architecture for researchers, starting with the collection of spatio temporal brain data (STBD) from a commodity EEG device and correlating it with intentional behaviour involving keyboard and mouse input. Though classification proves troublesome in the preliminary dataset considered, the architecture demonstrates a unique and flexible combination of a liquid state machine (LSM) and a deep belief network (DBN). Research in methodologies and techniques such as these are required for innovation in BCIs, as commodity devices, processing power, and algorithms continue to improve. Limitations in terms of types of classifiers, their range of expected inputs, discrete versus continuous data, spatial and temporal considerations and alignment with neural networks are also identified.
Graduate
0317
0984
jasoncummer@gmail.com

Deep ‘Analog Artificial Neural Networks’ (AANNs) perform complex classification problems with high accuracy. However, they rely on humongous amount of power to perform the calculations, veiling the accuracy benefits. The biological brain on the other hand is significantly more powerful than such networks and consumes orders of magnitude less power, indicating some conceptual mismatch. Given that the biological neurons are locally connected, communicate using energy efficient trains of spikes, and the behavior is non-deterministic, incorporating these effects in Artificial Neural Networks (ANNs) may drive us few steps towards a more realistic neural networks.

Emerging devices can offer a plethora of benefits including power efficiency, faster operation, low area in a vast array of applications. For example, memristors and Magnetic Tunnel Junctions (MTJs) are suitable for high density, non-volatile Random Access Memories when compared with CMOS implementations. In this work, we analyze the possibility of harnessing the characteristics of such emerging devices, to achieve neuro-inspired solutions to intricate problems.

We propose how the inherent stochasticity of nano-scale resistive devices can be utilized to realize the functionality of spiking neurons and synapses that can be incorporated in deep stochastic Spiking Neural Networks (SNN) for image classification problems. While ANNs mainly dwell in the aforementioned classification problem solving domain, they can be adapted for a variety of other applications. One such neuro-inspired solution is the Cellular Neural Network (CNN) based Boolean satisfiability solver. Boolean satisfiability (k-SAT) is an NP-complete (k≥3) problem that constitute one of the hardest classes of constraint satisfaction problems. We provide a proof of concept hardware based analog k-SAT solver that is built using MTJs. The inherent physics of MTJs, enhanced by device level modifications, is harnessed here to emulate the intricate dynamics of an analog, CNN based, satisfiability (SAT) solver.

Furthermore, in the effort of reaching human level performance in terms of accuracy, increasing the complexity and size of ANNs is crucial. Efficient algorithms for evaluating neural network performance is of significant importance to improve the scalability of networks, in addition to designing hardware accelerators. We propose a scalable approach for evaluating Liquid State Machines: a bio-inspired computing model where the inputs are sparsely connected to a randomly interlinked reservoir (or liquid). It has been shown that biological neurons are more likely to be connected to other neurons in the close proximity, and tend to be disconnected as the neurons are spatially far apart. Inspired by this, we propose a group of locally connected neuron reservoirs, or an ensemble of liquids approach, for LSMs. We analyze how the segmentation of a single large liquid to create an ensemble of multiple smaller liquids affects the latency and accuracy of an LSM. In our analysis, we quantify the ability of the proposed ensemble approach to provide an improved representation of the input using the Separation Property (SP) and Approximation Property (AP). Our results illustrate that the ensemble approach enhances class discrimination (quantified as the ratio between the SP and AP), leading to improved accuracy in speech and image recognition tasks, when compared to a single large liquid. Furthermore, we obtain performance benefits in terms of improved inference time and reduced memory requirements, due to lower number of connections and the freedom to parallelize the liquid evaluation process.