This diploma thesis studies magneto-optical (MO) response of epitaxial thin films of Co-doped ceria. Thin films were characterized by XPS, LEED, STM, spectroscopic ellipsometry and measurement of MO activity. The work focuses on studying MO response of the films depending on film thickness, cobalt concentration, oxidation state of cerium and chemical state of cobalt. Spectra of MO response consist of low energy region where the MO activity is mediated by transitions from defect induced states to conduction band and high energy region where a peak of MO activity appears which we attribute to transitions from valence band to conduction band. In this work we qualitatively explain the effects of the physico- chemical states of the thin films on the structure of the obtained MO spectra, mainly on the appearance of the MO activity in the low energy region, and on the changes of the position of the MO peak. Compared to other preparation methods the epitaxial thin films allow achieving a shift of the MO peak in the direction of higher photon energy.

The aim of this master thesis is to prepare and systematically characterize two groups of ultra-thin films of iron garnets with different compositions and deposition conditions. The first set consists of approximately 10 nm thick samples of gallium-doped yttrium-iron garnet on gadolinium gallium garnet (Gd3Ga5O12) (GGG) substrates with different crys- tallographic orientations and different annealing temperatures. The second group con- tains gallium-bismuth-doped neodymium-iron garnet on GGG substrates. These samples have different thicknesses, the concentration of individual elements and different annealing temperatures. All samples were prepared by the metal-organic decomposition method. The analysis of their physical properties was done by magneto-optical Kerr effect (MOKE) and optical experiments. Specifically, MOKE spectroscopy, Faraday hystere- sis loops measurements and spectroscopic ellipsometry. Obtained experimental results were further used to deduce the spectral dependence of complete permittivity tensor. Its spectral dependence was than discussed with relation to the electronic structure of in- vestigated materials which helped to select the best deposition conditions and substrate orientation. 1

This diploma thesis studies magneto-optical (MO) response of epitaxial thin films of Co-doped ceria. Thin films were characterized by XPS, LEED, STM, spectroscopic ellipsometry and measurement of MO activity. The work focuses on studying MO response of the films depending on film thickness, cobalt concentration, oxidation state of cerium and chemical state of cobalt. Spectra of MO response consist of low energy region where the MO activity is mediated by transitions from defect induced states to conduction band and high energy region where a peak of MO activity appears which we attribute to transitions from valence band to conduction band. In this work we qualitatively explain the effects of the physico- chemical states of the thin films on the structure of the obtained MO spectra, mainly on the appearance of the MO activity in the low energy region, and on the changes of the position of the MO peak. Compared to other preparation methods the epitaxial thin films allow achieving a shift of the MO peak in the direction of higher photon energy.

This thesis aims to (i) design and optimize the geometry of magneto-photonic crystal based on ferromagnetic garnet in order to resonantly enhance the magneto-optical re- sponse, (ii) to explore the possibility of using magnetic shape memory alloy to build an optically active photonic element, using advanced FDTD modeling. A Faraday rotation of 180◦ was reached but with low values of transmissivity. An investigation of the origin of such high values of Faraday rotation led to a conclusion that such structure has to be highly sensitive towards a change of a refractive index of its surroundings. This was confirmed, and so further development of this structure can lead to an efficient concentra- tion detector. Three designs of optically active element utilizing deformation of magnetic shape memory material in the external magnetic field were numerically simulated. Two designs (photonic crystal with cylindric holes in a hexagonal lattice and self-standing foil with cylindric holes in a square lattice) proved to be efficient and worth of further development. 1

No description available.

Magnetoactive elastomers (MAEs) are polymer composites containing magnetically soft or hard particles incorporated into an elastomer matrix during the crosslinking procedure. In the presence of a magnetic field, the induced magnetic interactions and the corresponding particle rearrangements significantly alter the mechanical properties in dependence on the initial particle distribution and sample shape. In addition, applying magnetic fields also changes the macroscopic shape of an MAE. This thesis investigates the magneto-mechanical coupled behaviour of MAEs by means of analytical and numerical methods. The effects of particle distribution and sample shape have been studied with the help of a physically motivated model of MAEs that considers dipole-dipole interactions between magnetizable particles. The presence of a magnetic field leads to a mechanical anisotropy in MAEs with isotropic particle distribution, and the induced anisotropy is directed along the orientation of the field. Thus, MAEs exhibit direction-dependent mechanical properties with distinct elastic moduli along and perpendicular to the field direction when the MAE sample is subjected to uniaxial deformation. A good agreement is reported between the physically motivated approach and conventional transversely isotropic material models. Furthermore, we investigate the important interplay between the particle distribution and the sample shape of MAEs, where a simple analytical expression is derived based on geometrical arguments to describe the particle distribution inside MAEs. We show that the enhancement of elastic moduli arises not only from the induced dipole-dipole interactions but also considerably from the change in the particle microstructure. Moreover, the magneto-mechanical behaviour of isotropic MAEs under shear deformations is studied. Three principal geometries of shear deformation are investigated with respect to the orientation of the applied magnetic field. We show that the Cauchy stress tensor of MAEs is not always symmetric due to the generation of a magnetic torque acting on an anisometric MAE sample under shear loadings. The theoretical study of magneto-mechanical behaviour of MAEs confirms that the effect of sample shape is quite significant and cannot be neglected. On the other hand, the initial particle distribution and presumed rearrangements due to the magnetic field additionally influence the material response of MAEs. Finally, the physically motivated model of MAEs could be transformed into an invariants-based model enabling its implementation in commercial finite element software. Therefore, we have uncovered a new pathway to model MAEs based on dipole-dipole interactions, leading to a constitutive relation analogous to the macro-scale continuum approach and revealing a synergy between both modelling strategies.

Le vieillissement de la population dans notre société moderne va entraîner de nouveaux besoins pour l’assistance aux personnes âgées et pour la réhabilitation. Les exosquelettes sont une piste de recherche prenant de plus en plus d’importance pour répondre à ces nouveaux challenges. Deux de ces challenges sont, la réalisation d’un contrôle naturel pour l’utilisateur et la sécurité. Cette maîtrise cherche à répondre à ces deux problématiques. Nous avons donc développé un outil de travail informatique utilisant les décharges électriques produites par les neurones moteurs pour contracter les muscles et un modèle des os et des muscles du bras. Cet outil utilise la librairie informatique ROS et OpenSim. Elle permet de connaître la force et le mouvement développés par le coude. De plus, un autre outil informatique a été développé pour optimiser le modèle des os et des muscles du bras pour le personnaliser à l’utilisateur pour un meilleur résultat. Une carte d’acquisition utilisant des électrodes de surface pouvant être reliées avec un ordinateur par USB et compatible avec ROS a été développée. Pour tester les algorithmes développés, un exosquelette pour le coude utilisant un actionneur compliant et contrôlé en force a été conçu. Pour compenser le poids de l’exosquelette et l’effet d’amortissement passif de l’actionneur, une compensation de gravité dynamique a été développée. Finalement, des expérimentations ont été menées sur l’efficacité de l’optimisation du modèle et sur l’exosquelette avec les différents algorithmes.

The advancement of photonics technologies depends on synthesis of novel materials and processes for device fabrication. The characterization techniques of the optical, electrical and magnetic properties of the synthesized materials and devices, by non-contact, non-invasive and nondestructive methods plays a significant role in development of new photonics technologies. The research reported in this thesis focuses on two such aspects of photonic materials characterization: Magneto-Optic characterization and Spectroscopic Ellipsometry. The theoretical and experimental basis of these two techniques, and experimental data analysis are presented in two parts. In Part 1, the changes in magneto-optic parameters of FePT PS-P2VP block copolymer nanocomposites with increasing concentrations of FePt nanoparticles in the block copolymer are analyzed. We present the results of change in MO anisotropy factor with the wt% of FePt and try to analyze these changes with further experimentation. Part 2 presents the results of spectroscopic ellipsometry of group III-IV multilayered thin film materials to give their precise thicknesses and optical constants. Both these techniques are unique ways to understand novel material characteristics for future use in device development.

Thesis (PhDEng)--University of Stellenbosch, 2003. / ENGLISH ABSTRACT: The advent of new high energy product permanent magnet materials has opened great opportunities for novel electrical machine topologies with advantageous features such as high efficiency and high power/weight ratio. Amongst others, axial field permanent magnet (AFPM) machines with ironless stators are increasingly being used in power generation applications. Because of the absence of the core losses, a generator with this type of design can operate at a substantially high efficiency. Besides, the high compactness and disc-shaped profile make this type of machine particularly suitable for compact integrated power generation systems. Due to construction problems, the generator application of this type of machine has been limited to quite a low power range. There is a need to investigate the performance capability of this type of AFPM machine in the upper medium power level. The focus of this thesis is on the design optimisation of the air-cooled AFPM generator with an ironless stator. A design approach that directly incorporates the finite element field solution in a multi-dimensional optimisation procedure is developed and applied to the design optimisation of a 300 kW (at unity power factor) AFPM generator. To enable an overall design optimisation of the machine, different design aspects, such as the cooling capacity, the mechanical strength and eddy loss, are also studied in this research. To enable the free movement of the rotor mesh with respect to the stator mesh, the air-gap element originally proposed by Razek et. al. is derived for Cartesian coordinate systems. For minimising the large computation overhead associated with this macro element, a number of existing time-saving schemes have been utilised together with the derived Cartesian air-gap element. The developed finite element time-step model is applied to calculating the steadystate performance of the AFPM machine. Since the flux distribution in an AFPM machine is three dimensional by nature, calculating the eddy current loss by merely using a simple analytical method may be subject to a significant error. To overcome this problem, the two dimensional finite element field modelling is introduced to perform accurate field analysis. To exploit the full advantages of the twodimensional finite element modelling, a multi-layer approach is proposed, which takes into account the variation of the air-gap flux density in the conductors with regard to their relative positions in the air-gap. To account for the radial variation of the field, a multi-slice finite element modelling scheme is devised. The thermal analysis is an important aspect of the design optimisation of AFPM machines. From a design point of view, it is preferable to have a simple but effective method for cooling analysis and design, which can easily be adapted to a wide range of AFPM machines. In this thesis a thermofluid model of the AFPM machine is developed. The fluid flow model is needed for calculating the air flow rate, which is then used to find the convective heat transfer coefficients. These are important parameters in the subsequent thermal calculations. Experimental investigations have been carried out to verify each of the above-mentioned models/methods. With these models implemented, the design optimisation of an air-cooled ironless stator 300 kW (at unity power factor) AFPM generator is carried out. The performance measurements done on the fabricated prototype are compared in this thesis with predicted results. The study shows that the proposed design approach can be applied with success to optimise the design of the AFPM machine. The advantages of high power density, high efficiency, no cogging torque and good voltage regulation make this type of AFPM machine very suitable for power generator applications. The optimum steady-state performance of the AFPM machine shows that this machine with an ironless stator is an excellent candidate for high speed power generator applications, even in the upper medium power level. The good cooling capacity of this type of machine holds the promise of its being a self-cooled generator at high power ratings. / AFRIKAANSE OPSOMMING: Die uitvinding van nuwe hoë energiedigtheid permanent magneet materiale het groot geleenthede vir nuwe elektriese masjien topologië laat ontstaan met voordelige eienskappe soos hoë benuttingsgraad en hoë drywing/gewig verhouding. Onder andere word die aksiaalveld permanente magneet (AVPM) masjiene met kernlose stators toenemend gebruik vir elektriese generator toepassings. As gevolg van die afwesigheid van kernverliese kan 'n generator met hierdie tipe ontwerp teen 'n aansienlik hoë benuttingsgraad werk. Daarbenewens maak die hoë kompaktheid en skyfvorm-profiel die masjien in besonder geskik vir die ontwikkeling van kompak geïntegreerde drywing generator stelsels. As gevolg van konstruksie probleme is die toepassing van hierdie tipe masjien as generator beperk tot redelik lae drywingsgebiede. Dit is nodig om die werkverrigtingsvermoë van hierdie tipe AVPM masjien in die boonste medium drywingsgebied te ondersoek. Die fokus van hierdie tesis is op die ontwerp-optimering van 'n lugverkoelde AVPM generator met 'n kernlose stator. 'n Ontwerpsbenadering wat die eindige element veldoplossing in 'n multi-dimensionele optimeringsprosedure insluit, is ontwikkel en toegepas op die ontwerpsoptimering van 'n 300 kW (by eenheidsarbeidsfaktor) AVPM generator. Om 'n globale ontwerpsoptimering van die masjien te kan doen is verskillende ontwerpsaspekte soos die verkoelingskapasiteit, meganiese sterkte en werwelverliese ook in hierdie navorsing bestudeer. Om die vrye beweging van die rotormaas ten opsigte van die statormaas te verseker is die lugspleet-element, soos oorspronklik deur Razek et al voorgestel, afgelei vir Cartesiaanse koórdinaat stelsels. Om die lang berekeningstyd geassosieer met hierdie makro-element te minimaliseer is 'n aantal bestaande tydbesparende metodes saam met die ontwikkelde Cartesiaanse lugspleet-element gebruik. Die ontwikkelde eindige element tydstapmodel is toegepas om die bestendige werkverrigting van die AVPM masjien te bereken. Aangesien die vloedverspreiding in 'n AVPM masjien van nature drie-dimensioneel is, kan die berekening van die werwelstroomverliese tot aansienlike foute lei as eenvoudige analitiese metodes gebruik word. Om hierdie probleem te oorkom is twee-dimensionele eindige element modellering gebruik om akkurate veld-analise te doen. Om die volle voordele van die twee- dimensionele eindige element modellering te eksploiteer is 'n multi-laag benadering voorgestel wat die variasie van die lugspleetvloeddigtheid in die geleiers met betrekking tot hulle relatiewe lugspleetposisies in ag neem. Om voorsiening te maak vir die radiale variasie van die veld, is 'n multi-skyf eindige element modelleringstegniek ontwikkel. Die termiese analise is 'n belangrike aspek van die ontwerpsoptimering van AVPM masjiene. Vanuit 'n ontwerpsoogpunt is dit verkieslik om 'n eenvoudige maar tog effektiewe metode van verkoelingsanalise en -ontwerp te hê wat maklik toegepas kan word op 'n wye reeks van AVPM masjiene. In hierdie tesis word 'n termovloeimodel van die AVPM masjien ontwikkel. Hierdie vloeimodel is nodig vir die berekening van die lugvloeitempo, wat op sy beurt weer nodig is om die konveksie hitte-oordrag koëffisiënte te bepaal. Hierdie is belangrike parameters in die opvolgende termiese berekeninge. Eksperimentele ondersoek is uitgevoer om elkeen van die bogenoemde modelle en metodes te verifieer. Nadat hierdie modelle geïmplimenteer is, is die ontwerpsoptimering van 'n 300 kW (by eenheidsarbeidsfaktor) lugverkoelde kernlose stator AVPM generator uitgevoer. Die werkverrigtingmetings gedoen op 'n vervaardigde prototipe masjien, word in hierdie tesis vergelyk met voorspelde resultate. Daar word getoon dat die voorgestelde ontwerpsbenadering met sukses toegepas kan word om die ontwerp van die AVPM masjien te optimeer. Die voordele van hoë drywingsdigtheid, hoë benuttingsgraad, geen vertandingsdraaimomente en goeie spanningsregulasie maak hierdie masjien baie aantreklik vir generator toepassings. Die optimum bestendige werkverrigting van die AVPM masjien toon dat hierdie masjien met 'n kernlose stator 'n goeie kandidaat is vir hoë spoed generator toepassings, selfs in die boonste medium drywingsgebied. Die goeie verkoelingskapasiteit van hierdie tipe masjien hou die belofte in van'n selfverkoelde generator by hoë drywing aanslae.

Radical aircraft and propulsion system architecture changes may be required to continue historic performance improvement rates as current civil aircraft and engine technologies mature. Significant fuel-burn savings are predicted to be achieved through the Distributed Propulsion concept, where an array of propulsors is distributed along the span of an aircraft to ingest boundary layer air and increase propulsive efficiency. Studies such as those by NASA predict large performance benefits when integrating Distributed Propulsion with the Blended Wing Body aircraft configuration, as this planform geometry is particularly suited to the ingestion of boundary layer air and the fans can be redesigned to reduce the detrimental distortion effects on performance. Additionally, a conventional aircraft with Distributed Propulsion has not been assessed in public domain literature and may also provide substantial benefits. A conceptual aircraft design code has been developed to enable the modelling of conventional and novel aircraft. A distributed fan tool has been developed to model fan performance, and a mathematical derivation was created and integrated with the fan tool to enable the boundary layer ingestion modelling. A tube & wing Distributed Propulsion aircraft with boundary layer ingestion has been compared with a current technology reference aircraft and an advanced turbofan aircraft of 2035 technology. The advanced tube & wing aircraft achieved a 27.5% fuel-burn reduction relative to the baseline aircraft and the Distributed Propulsion variant showed fuel efficiency gains of 4.1% relative to the advanced turbofan variant due to a reduced specific fuel consumption, produced through a reduction in distributed fan power requirement. The Blended Wing Body with Distributed Propulsion was compared with a turbofan variant reference aircraft and a 5.3% fuel-burn reduction was shown to be achievable through reduced core engine size and weight. The Distributed Propulsion system was shown to be particularly sensitive to inlet duct losses. Further investigation into the parametric sensitivity of the system revealed that duct loss could be mitigated by altering the mass flow and the percentage thrust produced by the distributed fans. Fuel-burn could be further reduced bydecreasing component weight and drag, through decreasing the fan and electrical system size to below that necessary for optimum power or specific fuel consumption.