Les fibres de carbone sont actuellement utilisées dans les matériaux composites pour les secteurs de l'aérospatiale, l’aéronautique et les sports de compétition. Leur application sur les marchés de grande consommation est toutefois entravée par le coût élevé des matières premières et le procédé de carbonisation, notamment l’étape de stabilisation. Par conséquent,les matériaux précurseurs alternatifs et peu coûteux sont très demandés. La lignine, une ressource naturelle très abondante contenant de grandes quantités de carbone, est considérée comme un bon candidat. Jusqu'à présent, les fibres de lignine ont été principalement préparées par filage en voie fondue et en mélange avec des polymères thermoplastiques pour améliorer leur aptitude au filage et leurs propriétés mécaniques, mais en réduisant fortement leurs rendements de carbonisation et en augmentant leur prix. Nous proposons dans cette thèse des fibres précurseur à base de lignine obtenues par un procédé de filage en continu par coagulation. En combinaison avec de petites proportions d'alcool polyvinylique, on obtient des fibres composites hautement flexibles et infusibles, contenant jusqu'à 70-90 % de lignine industrielle. Notre développement nous permet de fabriquer des fibres de carbone avec des rendements de l’ordre de 30 % qui présentent des propriétés prometteuses. Jusqu'à présent, les fibres de carbone à base de lignine mentionnées dans la littérature n'atteignent pas les propriétés mécaniques requises pour des applications à hautes performances en raison de leur structure de carbone amorphe. Cependant, en incorporant des cristaux liquides de feuillets d'oxyde de graphène ou des nanocristaux de cellulose dans nos fibres précurseurs de lignine, nous pouvons améliorer l'orientation des plans de carbone obtenus après la carbonisation. Nos systèmes de fibres à base de lignine avec une structuration améliorée représentent donc une étape importante vers la mise en oeuvre industrielle de la lignine en tant que matériau précurseur «vert» pour les fibres de carbone à faible coût et à haute résistance. / Carbon fibres are currently used in composite materials for the aerospace, transportation and energy sectors. Their application in mass markets however is hindered by the high cost of the fibre raw materials. Therefore, alternative and inexpensive precursor materials are in high demand. Especially lignin, a widely abundant natural resource containing high quantities of carbon, is considered as an important candidate. So far, lignin fibres have mostly been prepared by melt-spinning and by blending with thermoplastic polymers to enhance their spinnability and mechanical properties, but strongly lowering their carbonization yields and raising their price. We propose lignin-based precursor fibres obtained through a continuous wet-spinning process. In combination with only small ratios of polyvinyl alcohol, highly flexible and infusible composite fibres, containing up to 70-90 % of industrial lignin, can be obtained.Our development enables us to manufacture carbon fibres in high yields which exhibit promising properties. Until now, lignin-based carbon fibres reported in literature do not reach the mechanical properties required for high-performance applications due to their amorphous carbon structure. However, by incorporation of liquid crystalline graphene oxide flakes or cellulose nanocrystals into our lignin precursor fibres, we are able to improve the orientation of the carbon planes obtained after carbonization. Our lignin-based fibre systems with enhanced structuration thus represent an important step towards the industrial implementation of lignin as “green” precursor material for low-cost and high-strength carbon fibres.

<div> <p><a>Polymer nanocomposites are envisioned for use in many advanced applications, such as structural industries, aerospace, automotive technology and electronic materials, due to the improved properties like mechanical strengthening, thermal and chemical stability, easy bulk processing, and/or light-weight instigated by the filler-matrix combination compared to the neat matrix. In recent years, due to increasing environmental concerns, many industries are inclining towards developing sustainable and renewable polymer nanocomposites. Cellulose nanomaterials (CNs), including cellulose nanocrystals (CNCs) and cellulose nanofibrils (CNFs), have gained popularity due to their excellent mechanical properties and eco-friendliness (extracted from trees, algae, plants etc.). However, to develop CN-reinforced nanocomposites with industrial applications it is necessary to understand impact of hygroscopic swelling (which has very limited </a>quantitative study at present), aspect ratio, orientation, and content of CNs on the overall performance of nanocomposites; and overcome the low dispersibility of CNs and improve their compatibility with hydrophobic matrix. In this work, we attempt to understand the influence of single nanocrystals in the hygroscopic and optical response exhibited by nanostructured films; effect of CNCs on the properties of PVA/CNC fibers by experimental evidence with mathematical modeling predictions; and hydrophobized CNFs using a facile, aqueous surface modification to improve interfacial compatibility with epoxy. </p><p><br></p> <p>To evaluate the effect of CNC alignment in the bulk response to hygroscopic expansion, self-organized and shear-oriented CNC films were prepared under two different mechanisms. The coefficient of hygroscopic swelling (CHS) of these films was determined by using a new contact-free method of Contrast Enhanced Microscopy Digital Image Correlation (CEMDIC) that enabled the characterization of dimensional changes induced by hygroscopic swelling of the films. This method can be readily used for other soft materials to accurately measure hygroscopic strain in a non-destructive way. By calculating the CHS values of CNC films, it was determined that hygroscopic swelling is highly dependent on the alignment of nanocrystals within the films, with aligned CNC films showing dramatically reduced hygroscopic expansion than randomly oriented films. Finite element analysis was used to simulate moisture sorption and kinetics profile which further predicted moisture diffusion as the predominant mechanism for swelling of CNC films. </p> <p><br></p><p>To study the effects of different types and aspect ratios of CNCs on mechanical, thermal and morphological properties of polyvinyl alcohol (PVA) composite <a>fibers, CNCs extracted from wood pulp and cotton were reinforced into PVA to produce fibers by dry-jet-wet spinning. The fibers were collected as-spun and with first stage drawing up to draw ratio 2. </a>The elastic modulus and tensile strength of the fibers improved with increasing CNC content (5 – 15 wt. %) at the expense of their strain-to-failure. The mechanical properties of fibers with cotton CNC were higher than the fibers with wood CNC when the same amount of CNCs were added due to their higher aspect ratio. The degree of orientation along the spun fiber axis was quantified by 2D X-ray diffraction. As expected, the CNC orientation correlates to the mechanical properties of the composite fibers. Micromechanical models were used to predict the fiber performance and compare with experimental results. Finally, surface and cross-sectional morphologies of fibers were analyzed by scanning electron microscopy and optical microscopy.</p><p><br></p> <p>To improve the dispersibility and compatibility of CNFs with epoxy, CNFs were modified by using a two-step water-based method where tannic acid (TA) acts as a primer with CNF suspension and reacts with hexadecylamine (HDA), forming the modified product as CNF-TA-HDA. The modified (-m) and unmodified (-um) CNFs were filled into hydrophobic epoxy resin with a co-solvent (acetone), which was subsequently removed to form a solvent-free two component epoxy system, followed by addition of hardener to cure the resin. Better dispersion and stronger adhesion between fillers and epoxy were obtained for m-CNF than the um-CNF, resulting in better mechanical properties of nanocomposites at the same loading. Thermal stability and the degradation temperature of m-CNF/epoxy improved when compared to neat epoxy. </p> </div> <br>

Putting on or taking off clothes helps the body to stay within the comfortable temperature range (toavoid shivering or sweating) at different activity levels and ambient conditions. Clothes with built-inthermo-regulating properties would mean maintained comfort without putting on or taking off clothesthat frequently. Integration of phase change materials (PCMs) in clothes is one way of achievingthermo-regulating properties. When the body temperature goes up, the PCM melts and absorbs theheat from the body in the form of latent heat (cooling effect). When the temperature drops, the PCMcrystallizes and the stored heat is released again (warming effect).Research on thermo regulating fibres of the bi-component type containing PCM in the core has beenconducted at Swerea IVF in Mölndal, Sweden, for some time. It has been found that high molecularweight HDPE is a suitable viscosity modifier for hydrocarbon waxes used as PCM. The preparation ofcore materials has so far been done in a batch wise fashion in the way that molten wax has beensoaked into pelletized HDPE at around 180°C during prolonged times followed by melt compoundingin a Brabender batch kneader (0.3 kg per batch). Besides being very impractical for larger productionvolumes the method involves long residence times at high temperatures which may induce thermaldegradation reactions. The objective of the present diploma (master’s thesis) work was to develop acontinuous mixing method to produce PCM/HDPE blends and to test the resulting material in bicomponentfibers with a Nylon (PA6) sheath and to characterize the resulting fiber properties in termsof strength and latent heat.It was proven possible to compound HDPE with large amounts (70%) of octadecane (PCM) on aBrabender twin screw extruder. HDPE was metered to the extruder hoper by means of a screw feederand wax was continuously fed to the hoper in the liquid state by means of a heated membrane pump.To facilitate mixing HDPE in form of powder instead of pellets was used. The extruded threads weresolidified in a water bath followed by granulation. Bi-component fibers were successfully producedfrom such materials. Fibers containing 15 to 42% Octadecane were produced showing heat of fusionsin the range 26 to 86 J/g and tenacities in the range 33 to 16 cN/tex. The heat of fusion of the fiberscompares favorable with existing commercial products showing values in the range 5-15 J/g (acrylicand cellulosic fibres containing microencapsulated hydrocarbon waxes). The peak melting point ofoctadecane measured by DSC was found to be depressed some 4-5°C in the fibers compared to pureoctadecane (28°C). Such a melting point depression is important to consider when choosing type ofhydrocarbon wax. / Program: Magisterutbildning i textilteknologi

Les prothèses vasculaires textiles en polyéthylène téréphtalate (PET) présentent souvent des occlusions après implantation pour les petits diamètres (6-8mm) car la surface des prothèses est peu hémocompatible. Or, l'hémocompatibilité des prothèses serait largement améliorée si ces dernières se recouvraient d'une couche de cellules endothéliales qui tapissent naturellement les vaisseaux sanguins. Ce projet vise à mettre en évidence que les textiles bruts ne sont pas un support viable pour le développement de ces cellules endothéliales, puis propose de remplacer les matrices protéiniques par un recouvrement synthétique. Pour ce faire, de l'acide poly-L-lactique (PLA) solubilisé a été filé sous forme de nanofibres déposées sur la surface luminale de la prothèse. L'étirage par jet d'air a été caractérisé selon un modèle plan, puis adapté à la fon11e tubulaire des prothèses. Les nanofibres ont été évaluées sur le plan de la cytocompatibilité, de l'adhérence et de la prolifération avec un modèle de cellules endothéliales animales. Ce travail vise aussi à optimiser l'adhérence de ces fibres sur le PET par l'emploi d'une technique de modification de surface par plasma. Les résultats montrent qu'il est possible de produire des nanofibres de PLA et de contrôler leur diamètre, et de sceller la paroi de la prothèse textile. Enfin, les cellules endothéliales prolifèrent en monocouche sur des prothèses recouve1tes de nanofibres. Il est possible d'optimiser l'adhérence des nanofibres sur le PET avec un traitement par plasma. En conclusion nous avons proposé une alternative à l'enduction traditionnelle des prothèses permettant la prolifération en monocouche des cellules endothéliales. / Textile vascular prostheses show poor patency rate for smaller diameter grafts (6-8mm). Mainly due to thrombosis or hyperplasia, graft failures can be explained by meagre hemocompatibility. Lack of neoendothelialization of the inner wall of the graft can be one reason explaining this poor hemocompatibility, This project aimed to prove that bare textiles are not a good support to stand endothelial cells' proliferation. Poly(L-latic) acid was therefore chosen to replace protein coating by being formed as a nanofibres mesh on the PET textile prostheses luminal surface. Air jet spinning process was first evaluated in a basic planar model to determined optimal parameters for nanofibres production. Endothelial cells compatibility, adhesion and proliferation were tested. Then air jet spinning was dedicated to tubular shape of textile vascular prostheses. Nanofibres mesh were analysed for chemical and physical properties, and covered graft were tested for water permeability. Lastly, atmospheric pressure plasma treatment was performed to optimize PLLA nanofibres adhesion on PET. Results showed that nanofibre diameters were controlled by polymer concentration. Nanofibre cristallinities depend of spinning parameters. Air jet spinning allows quick covering of textile surfaces with a dense net of nanofibre scelling the inner wall of the prosthesis, even in tubular samples. Moreover, endothelial ceIls show monolayer proliferation on these nanofibres. Finally, Polylactic acid adhesion on PET was optimized with atmospheric pressure plasma. In conclusion, we bring a new solution to cover inner wall of prostheses, allowing a monolayer proliferation of endothelial cells.

Hochleistungsfähige Lithium-Ionen-Batterien sind insbesondere von der hohen spezifischen Kapazität ihrer Elektrodenmaterialien abhängig. Intermetallische Phasen sind vielversprechende Kandidaten für alternative Anodenmaterialien mit verbesserten spezifischen Kapazitäten (LiAl: 993 Ah/kg; Li22Si5: 4191 Ah/kg) gegenüber den derzeit vielfach verwendeten Kohlenstoff-Materialien (LiC6: 372 Ah/kg). Nachteilig ist jedoch, dass die kristallinen Phasenumwandlungen während der Lade-Entlade-Prozesse Volumenänderungen von 100-300% verursachen. Durch die Sprödigkeit der intermetallischen Phasen führt dies zum Zerbrechen des Elektrodenmaterials und damit zum Kontaktverlust. Um Lithiierungs- und Delithiierunsprozesse ohne kristalline Phasenumwandlungen zu realisieren und somit große Volumenänderungen zu vermeiden, wurden amorphe Al-Legierungen untersucht. In amorphe, mittels Schmelzspinnen hergestellte Legierungen (Al86Ni8La6 und Al86Ni8Y6) kann beim galvanostatischen Zyklieren nur sehr wenig Li eingelagert werden. Da kristalline Phasenumwandlungen im amorphen Zustand nicht möglich sind, wird für die Diffusion und Einlagerung von Li-Ionen ein ausreichendes freies Volumen im amorphen Atomgerüst benötigt. Die Dichtemessung der Legierungen zeigt, dass dieses freie Volumen für eine signifikante Lithiierung nicht ausreichend ist. Wird Li bereits in die amorphe Ausgangslegierung integriert, können Li-Ionen auf elektrochemischem Wege aus ihr entfernt und auch wieder eingebaut werden. Die neuartige Legierung Al43Li43Ni8Y6, die Li bereits im Ausgangszustand enthält, konnte mittels Hochenergiemahlung als amorphes Pulver hergestellt werden. Verglichen mit den Li-freien amorphen Legierungen Al86Ni8La6 bzw. Al86Ni8Y6 und ihren kristallisierten Pendants zeigt diese neu entwickelte, amorphe Legierung eine signifikant höhere Lithiierungsfähigkeit und erreicht damit eine spezifische Kapazität von ca. 800 Ah/kg bezogen auf den Al-Anteil. Durch den Abrieb des Stahlmahlbechers enthält das Pulver Al43Li43Ni8Y6 einen Fe-Anteil von ca. 15 Masse%. Dieses mit Fe verunreinigte Material zeigt besonders bei niedrigen Laderaten eine bessere Zyklenstabilität als ein im abriebfesten Siliziumnitrid-Becher gemahlenes Pulver der gleichen Zusammensetzung. Mittels Mössbauerspektroskopie wurde nachgewiesen, dass das Pulver z.T. oxidisches Fe enthält. Dieses kann über Konversionsmechanismen einen Beitrag zur spezifischen Kapazität leisten. / High-energy Li-ion batteries exceedingly depend on the high specific capacity of electrode materials. Intermetallic alloys are promising candidates to be alternative anode materials with enhanced specific capacities (LiAl: 993 Ah/kg; Li22Si5: 4191 Ah/kg) in contrast to state-of-the-art techniques, dominated by carbon materials (LiC6: 372 Ah/kg). Disadvantageously the phase transitions during the charge-discharge processes, induced by the lithiation process, cause volume changes of 100-300 %. Due to the brittleness of intermetallic phases, the fracturing of the electrode material leads to the loss of the electrical contact. In order to overcome the huge volume changes amorphous Al-based alloys were investigated with the intension to realize the lithiation process without a phase transformation. Amorphous powders (Al86Ni8La6 and Al86Ni8Y6) produced via melt spinning and subsequent ball milling only show a minor lithiation during the electrochemical cycling process. This is mainly caused by the insufficient free volume, which is necessary to transfer and store Li-ions, since phase transitions are impossible in the amorphous state. If Li is already integrated into the amorphous alloy, Li-ions can easily be removed and inserted electrochemically. The new alloy Al43Li43Ni8Y6 contains Li already in its initial state and could be prepared by high energy milling as an amorphous powder. Compared with the Li-free amorphous alloys Al86Ni8La6 or Al86Ni8Y6 and their crystalline counterparts, this newly developed amorphous alloy achieves a significantly higher lithiation and therefore reaches a specific capacity of 800 Ah/kg, based on the Al-content. By the abrasion of the steel milling vials the powder contains a wear debris of 15 mass% Fe. This contaminated material shows a better cycling stability than a powder of the same composition, milled in a non-abrasive silicon nitride vial. By means of Mössbauer spectroscopy has been shown that the wear debris contains Fe oxides. This may contribute to the enhancement of the specific capacity about conversion mechanisms.

In this article, we look at a model with (independent) system operator who faces stochastic but growing transmission demand and a penalty if frequency is not balanced. In this set up, we derive an optimal grid expansion investment strategy and analyze its value and risk implications. It turns out that the firm value is strictly concave in the level of transmission demand. Firm value, however, increases with optimal investment for any level of demand. Moreover, firm risk is decreasing in the level of demand and higher when the firm has an investment option. The risk increase corresponds to the exercise of the call option and is stronger, the closer the firm approaches its exercise trigger. (author's abstract) / Series: Working Papers / Research Institute for Regulatory Economics

Tissue engineering is a promising alternative strategy to produce artificial bone substitutes / however, the control of the cell organization and cell behavior to create fully functional 3-D constructs has not yet been achieved. To overcome these, activities have been concentrated on the development of multi-functional tissue engineering scaffolds capable of delivering the required bioactive agents to initiate and control cellular activities. The aim of this study was to prepare tissue engineered constructs composed of polymeric scaffolds seeded with mesenchymal stem cells (MSCs) carrying a nanoparticulate growth factor delivery system that would sequentially deliver the growth factors in order to mimic the natural bone healing process. To achieve this, BMP-2 and BMP-7, the osteogenic growth factors, were encapsulated in different polymeric nanocapsules (poly(lactic acid-co-glycolic acid) (PLGA) and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV)) with different properties (degradation rates, crystallinity) and, therefore, different release rates to achieve the early release of BMP-2 followed by the release of BMP-7, as it is in nature. Initially, these nanoparticulate delivery systems were characterized and then the effect of single, simultaneous and sequential delivery of BMP-2 and BMP-7 from these delivery systems was studied in vitro using rat bone marrow MSCs. The effect of using these two growth factors in a sequential manner by mimicking their natural bioavailability timing was shown with maximized osteogenic activity results. BMP-2 loaded PLGA nanocapsules were subcutaneously implanted into Wistar rats and according to initial results, their biocompatibility as well as the positive effect of BMP-2 release on the formation of osteoclast-like cells was shown. To complete the construction of the bioactive scaffold, this nanoparticulate sequential delivery system was incorporated into two different types of polymeric systems / natural (chitosan) and synthetic (poly(&amp / #949 / -caprolactone) (PCL)). 3-D fibrous scaffolds were produced using these materials by wet spinning and 3-D plotting. Incorporation of nanocapsules into 3-D chitosan scaffolds was studied by two different methods: incorporation within and onto chitosan fibers. Incorporation into 3-D PCL scaffolds was achieved by coating the nanocapsules onto the fibers of the scaffolds in an alginate layer. With both scaffold systems, incorporation of nanocapsule populations capable of delivering BMP-2 and BMP-7 in single, simultaneous and sequential fashion was achieved. As with free nanocapsules, the positive effect of sequential delivery on the osteogenic differentiation of MSCs was shown with both scaffold systems, creating multi-functional scaffolds capable of inducing bone healing.

Im Rahmen der Dissertation wurden Komplexe Intermetallische Phasen (KIP) in den Systemen Ag-Mg und Ag-Ga-Mg dargestellt und charakterisiert. KIP sind Verbindungen, die sich grundlegend von einfachen Metallen unterscheiden. Große Elementarzellen, ein hierarchisch strukturierter Aufbau und inhärente Fehlordnung sind wesentliche Charakteristika. Empirisch wird zudem eine Häufung von strukturchemisch verwandten KIP (Ähnlichkeitsregel) in der Nähe von definierten Zusammensetzungen beobachtet (Häufungsregel). Obwohl nur wenig über die physikalischen Eigenschaften dieser Verbindungsklasse bekannt ist, zeigen neueste Untersuchungen, dass sie interessante Eigenschaften wie ungewöhnliches plastisches Verhalten und Pseudo-Bandlücken in der elektronischen Zustandsdichte in Höhe der Fermi-Energie aufweisen können. Diese Arbeit zeigt exemplarisch, dass in der Chemie der KIP durchaus einfache Regeln (Häufungs-, Ähnlichkeits-, Valenzelektronenkonzentrationsregel) genutzt werden können, um neue Verbindungen mit vorgegebenen geometrischen Baueinheiten aufzufinden. Vereinfachende Annahmen, wie die Aussage, dass zweikomponentige Mackay-Cluster keine Fehlordnung aufweisen oder dass ein Mackay-Cluster maximal 92 Valenzelektronen enthält, erwiesen sich hingegen als falsch. Die Untersuchungen haben gezeigt, dass eine Kombination verschiedener Synthesemethoden notwendig ist. Insbesondere hat sich das Schmelzspinnverfahren, das zur Darstellung kristalliner magnesiumreicher Legierungen angewandt und modifiziert wurde, bei der phasenreinen Synthese der Verbindungen bewährt. Die Entwicklung von Spritzdüsen aus Tantal ermöglichte die kontaminationsfreie Verarbeitung der Mg-haltigen Legierungen. Die Bestimmung der Kristallstrukturen inklusive der Fehlordnungsphänomene war für das Verständnis der Stabilität der KIP entscheidend. Da häufig zwischen verschiedenen Modellen zur Beschreibung der Fehlordnung entschieden werden musste, waren neben präzisen Beugungsdaten genaue Untersuchungen der Präparate mittels chemischer Analytik, Metallographie und WDX hinsichtlich der Zusammensetzung aller Phasen erforderlich. Der magnesiumreiche Teil des binären Phasendiagramms Ag-Mg wurde neu bestimmt. Dabei wurden zwei bislang unbekannte Phasen dargestellt. Die fünf magnesiumreichen Phasen kristallisieren innerhalb eines schmalen Bereichs von nur 9 At.-% Mg. Die Kristallstrukturen dieser Phasen wurden unter besonderer Berücksichtigung der Fehlordnungsphänomene untersucht. Die Verbindungen sind strukturchemisch verwandt und lassen sich den I3-Cluster-Phasen zuordnen. Ag2Mg5 kristallisiert ohne Fehlordnung im Al5Co2-Typ. Die Kristallstrukturen von Ag7Mg26 und Ag17Mg54 lassen sich als fcc bzw. bcc Anordnungen von Mackay-Clustern beschreiben. Es handelt sich um die ersten bekannten binären Phasen, in denen innerhalb von isolierten Mackay-Clustern Substitutionsfehlordnung auftritt. AgMg4 kristallisiert hexagonal in einem eigenen Strukturtyp. Das I3-Cluster-Netzwerk füllt den gesamten Raum bis auf einen annähernd zylindrischen Bereich um 0, 0, z, in dem eine Atom-Split-Position aus drei Lagen vorliegt. Lokal liegen drei unterschiedliche Koordinationspolyeder vor, deren Stapelabfolge in der Kristallstruktur von AgMg4 zufällig ist, jedoch mit kurzreichweitiger Korrelation. Es konnte gezeigt werden, dass sich AgMg4 in eine Tieftemperaturphase umwandelt, in der die Polyeder vermutlich langreichweitig unter Bildung einer Überstruktur ordnen. Der Phasenbestand des ternären Systems Ag-Ga-Mg wurde untersucht und die Kristallstrukturen von sechs neuen Phasen bestimmt. Des weitern wurde eine Verbindung im System Ga-Mg-Pd charakterisiert. Anhand der Strukturtypen Al3Ir und Cu3P sowie den Verbindungen Ag0,55Ga0,45Mg3, Ga4,62Mg13,38Pd7 und Ag1,31Ga1,89Mg7,80 wurde gezeigt, wie die Variation von Strukturmotiven durch geringe Abweichungen von idealer Symmetrie zu zunehmend komplexeren Kristallstrukturen führt, die sich stets von Packungen des Edshammarpolyeders ableiten lassen. Drei ternäre Phasen vom I3-Cluster-Typ konnten identifiziert werden: Neben ternären Varianten der Phasen Ag7Mg26 und Ag17Mg54 kristallisiert Ag0,59Ga0,41Mg2 metastabil im NiTi2-Typ. In den strukturchemisch verwandten Phasen Ag6Ga12Mg11 und Ag21Ga74Mg44 bilden die Mg Atome Netzwerke mit Clatrath-II- bzw. Clatrath-IV-Topologie, die mit Ikosaedern und Frank-Kasper-Polyedern aus Ag und Ga gefüllt sind. Diese Phasen werden wahrscheinlich durch das e/a-Verhältnis der gesamten Struktur im Sinne einer Hume-Rothery-Regel stabilisiert.

The production of fine filaments using the melt spinning process needs considerable effort. A thermoplastic melt is stretched from the spinneret under a constant take-up speed. The high performance thermoplastic PEEK is solidified in the melt spinning process in a small distance and short time. Therefore, the fine PEEK filaments in the fibre formation zone underwent a high deformation and cooling rate. To make the melt spinning process stable and to produce the fine PEEK filaments, material properties and material behaviour are examined using on-line and off-line measurements. The fibre speed measured using Laser Doppler Anemometry and simultaneous temperature measured using infrared thermography enable both the strain rate and consequently the apparent extensional viscosity to be estimated. This provides the apparent extensional viscosity over the spinning line, which can itself show the structural development of PEEK fibres in the fibre formation zone, i.e. necking and solidification phenomena. The one-dimensional fibre formation model must include both procedural and material parameters. The heat transfer coefficient was estimated using the filament temperature measurement and showed a relatively high contribution of radiation and free convection in comparison to forced convection near the spinneret. The improved model of PEEK fibre formation gave a good agreement to both temperature and speed measurements, and also confirmed the high deformation rate effect on the extensional viscosity, which could be simulated with a properly generalised Newtonian constitutive equation. The end properties of the fibres, such as as-spun filament fineness, orientation (expressed using total birefringence) and total crystallisation (examined using DSC) are investigated in relation to different spinning conditions, i.e. take-up speed, throughput and the draw down ratio. The tensile test diagram results, measuring phenomena such as the elongation at break, tenacity, and the Young modulus of elasticity are also analysed in order to complete the correlation of the above-mentioned spinning conditions to the structural properties of as-spun fine PEEK filaments. The melt spinning of fine PEEK fibres under different spinning conditions is examined with the purpose of finding the optimum take-up speed and throughputs. Other spinning conditions, such as the temperature of melt processing, and the arrangement and diameter of the spinneret holes, are changed in order to make the process more stable. The recommendations for further study can be used to further examine some aspects of this work; however, this work presents a new concept for fine PEEK melt spinning supported by spinnability examinations under different spinning conditions and the improved model of fibre formation, which is also relevant for typical industrial processing applications.

Deux voies ont été explorées pour rechercher des matériaux pouvant être utilisées comme aimants permanents: a) soit l'amélioration des performances magnétiques de composes connus. Ainsi d'une part l'insertion d'environ 120 K et d'autre part l'enrichissement en fer a l'aide d'une méthode d'hypertrempe de la solution solide GdFe12-xAlx conduit à des composés ferromagnétiques (TC=500 K pour x=2), b) soit l'obtention de nouveaux composés à base d'uranium qui permet d'induire une anisotropie magnétocristalline. L'étude des propriétés structurales des siliciures et germaniures UM10Si2 (M=Fe, Co) et U2M17-yXy (M=Fe, Co et X=Si, Ge) montre une occupation préférentielle du silicium ou du germanium de certains sites cristallographiques des structures types ThMn12 et Th2Ni17. Les propriétés magnétiques de ces composés sont analysees en fonction des distances M-M.