No description available.

Typically wire antenna structures are modeled by approximating curved structures with straight wire segments. The straight wire approximation yields accurate results, but often requires a large number of segments to adequately approximate the antenna geometry. The large number of straight wire segments or unknowns requires a large amount of memory and time to solve for the currents on the antenna. By using curved segments which exactly describe the contour of the antenna geometry the number of unknowns can be reduced, thus allowing for bigger problems to be solved accurately. This thesis focuses on the analysis of a helix antenna. The Method of Moments is used to solve for the currents on the antenna, and both the triangle basis and pulse testing functions exactly follow the contour of the helix antenna. The thin wire approximation is used throughout the analysis. The helix is assumed to be oriented along the z-axis with an optional perfect electric conductor (PEC) ground plane in the x-y plane. For simplicity, a delta gap source model is used. Straight feed wires may also be added to the helix, and are modeled similarly to the helix by the Method of Moments with triangular basis and pulse testing functions. The primary validation of the curved wire approach is through a comparison with MININEC and NEC of the convergence properties of the input impedance of the antenna versus the number of unknowns. The convergence tests show that significantly fewer unknowns are needed to accurately predict the input impedance of the helix, particularly for the normal mode helix. This approach is also useful in the analysis of the axial mode helix where the current changes significantly around one turn. Because of the varying current distribution, the improvement of impedance convergence with curved segments is not as significant for the axial mode helix. However, radiation pattern convergence improvement is found. Multiple feed structures for the axial mode helix are also investigated. In general, the many straight wire segments, and thus unknowns, that are needed to accurately approximate the current around one turn can be greatly reduced by the using the curved segment method. / Master of Science

Once a keystone for the city of Alexandria, the Old Town Cinema has lost its relevance and spirit. Unable to compete with the prevailing Cineplexes, the Old Town becomes a spectacle of the pornographic: baring all at once, substituting base desire for refined attraction. Out of this disenchantment, an idea for a new cinema emerges- the momentous Cinema. Moments which unfold frame by frame- nurturing anticipation, choreographing the perspectives that the Cineplex has abandoned. A place free of solicitation, free of congestion. An alluring place, revealing itself in measured experiences. / Master of Architecture

published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy. Johannesburg, May 23, 2014. / A systematic experimental investigation was performed of the phenomenon of fine structure, with emphasis on the region of the Isoscalar Giant Quadrupole Resonance (ISGQR), in nuclei across stable even-even neodymium isotopes. The 200 MeV proton beams were delivered by the Separated Sector Cyclotron (SSC) facility of iThemba Laboratory for Accelerator Based Sciences (iThemba LABS). Measurements were made using the state-of-the-art K = 600 magnetic spectrometer, where unique high energy-resolution ( E ≈ 42 − 48 keV FWHM) proton inelastic scattering results were obtained on 142Nd, 144Nd, 146Nd, 148Nd and 150Nd targets. All measurements were taken at θLab = 8◦, where the cross-section of the ISGQR is at a maximum. An additional measurement was also made for the 142Nd at θLab = 7◦. Nuclei with mass number A ≈ 150 and neutron number N ≈ 90 are of special interest since they occupy that region of the nuclide chart wherein the onset of permanent prolate deformation occurs. The stable neodymium (Z = 60) isotopes have been chosen in the present study in order to investigate the effects accompanying the onset of deformation on the excitation energy spectra in the ISGQR region (9 ≤ Ex ≤ 15 MeV). The neodymium isotopes extend from the semi-magic N = 82 nucleus (142Nd) to the permanently deformed N = 90 (150Nd) nucleus. In order to emphasize the ISGQR in the measured excitation energy spectra, a Discrete Wavelet Transform (DWT) background subtraction was carried out. This model independent method for background determination decomposes the spectrum into various approximations and details through the application of high pass and low pass filters. A comparison of the resonance widths extracted shows a systematic broadening of the ISGQR (􀀀 = 3.220 MeV to 5.100 MeV), moving from spherical 142Nd to highly deformed 150Nd nuclei as has already been observed for the Isovector Giant Dipole Resonance (IVGDR) excited by γ-capture. Even though it is known that the IVGDR spectacularly splits and shows a double bump for the deformed 150Nd, no obvious splitting of the ISGQR was observed. In order to investigate the fine structure of the ISGQR, a theoretical microscopic calculation termed the Quasiparticle-Phonon Model (QPM) was applied to predict excitation energy spectra for 142−146Nd targets. These calculations were based on the one- plus two-phonon configuration. Characteristic energy scales were extracted for the resonance region using the Continuous Wavelet Transform (CWT) technique, on both experimental data and theoretical predictions. Comparison of the resulting characteristic energy scales suggests the coupling to low-lying collective vibrations as the dominant contributor to the ISGQR decay width. Level densities of 2+ states were extracted through the application of a fluctuation analysis technique, for full spectra from the ground state upwards in all five Nd targets. Comparisons are made with theoretical predictions from the Back Shifted Fermi Gas, Hartree-Fock-BCS and Hartree-Fock-Bogoluibov models. While there is generally an excellent agreement between experimental level densities and theoretical predictions from the ground state up to less than 10 MeV excitation, there is a marked disagreement beyond 10 MeV in all target nuclei. Comparison of the experimental results for the Nd isotope chain shows a clear systematic trend in which the onset of this disagreement occurs at lower and lower excitation energies, moving from low to high mass. For the spherical 142Nd nucleus the deviation occurs at about 9 MeV while in the case of the deformed 150Nd this occurs much earlier at about 4 MeV, all limited to a maximum of 103 MeV−1 by the energy resolution of the present experiment. Additionally, measurements of elastic scattering and inelastic excitation of lowlying collective states in 144−150Nd has also been possible. Excitation energy spectra in all targets predominantly exhibited various 2+ states, owing to the “spin-filter” effects. A single strong 3− 1 state, together with a weak 4+ 1 state were observed in each target nucleus. Angular distributions were obtained for the various ground and excited states by applying the optical model of elastic scattering and Distorted Wave Born Approximation (DWBA) of inelastic scattering. Deformation lengths δL were obtained for most of the states and these were in good agreement with previously obtained results from the literature.

In this thesis, the Method of Moments has been applied to calculate capacitance between two arbitrary 3D metal conductors or a capacitance matrix for a 3D multi-conductor system. Capacitance extraction has found extensive use for systems involving sets of long par- allel transmission lines in multi-dielectric environment as well as integrated circuit package including three-dimensional conductors located on parallel planes. This paper starts by reviewing fundamental aspects of transient electro-magnetics followed by the governing dif- ferential and integral equations to motivate the application of numerical methods as Method of Moments(MoM), Finite Element Method(FEM), etc. Among these numerical tools, the surface-based integral-equation methodology - MoM is ideally suited to address the prob- lem. It leads to a well-conditioned system with reduced size, as compared to volumetric methods. In this dissertation, the MoM Surface Integral Equation (SIE)-based modeling approach is developed to realize electrostatic capacitance extraction for 3D geometry. MAT- LAB is employed to validate its e?ciency and e?ectiveness along with design of a friendly GUI. As a base example, a parallel-plate capacitor is considered. We evaluate the accu- racy of the method by comparison with FEM simulations as well as the corresponding quasi-analytical solution. We apply this method to the parallel-plate square capacitor and demonstrate how far could the undergraduate result 0C = A ? "=d' be from reality. For the completion of the solver, the same method is applied to the calculation of line capacitance for two- and multi-conductor 2D transmission lines.

Thesis (Ph. D.)--Ohio State University, 2007. / Title from first page of PDF file. Includes bibliographical references (p. 216-221).

Thesis (Ph. D.)--Ohio State University, 2007. / Title from first page of PDF file. Includes bibliographical references (p. 204-208).

Dans cette thèse, on s'intéresse à différentes équations issues de la théorie cinétique. Tout d'abord, on considère une équation de Landau pour les particules de Fermi-Dirac. On montre l'existence d'une solution au problème de Cauchy associé et on détermine les états d'équilibre. Ensuite, dans une deuxième partie, on s'intéresse aux systèmes de moments pour l'équation de Boltzmann en relativité restreinte et on détermine les espaces de moments relativistes adéquats. Dans une troisième partie, on étudie les états stationnaires d'une équation de Kac avec thermostat dans le cas où la section efficace est supposée non-intégrable. Finalement, la quatrième partie est consacrée à l'étude d'une équation issue de la théorie de la coagulation, l'équation de Oort-Hulst-Safronov, qui est approchée par une suite d'équations discrètes.

A three dimensional (3-D) capacitance extraction algorithm based on a kernel independent hierarchical method and geometric moments is described. Several techniques are incorporated, which leads to a better overall performance for arbitrary interconnect systems. First, the new algorithm hierarchically partitions the bounding box of all interconnect panels to build the partition tree. Then it uses simple shapes to match the low order moments of the geometry of each box in the partition tree. Finally, with the help of a fast matrix-vector product, GMRES is used to solve the linear system. Experimental results show that our algorithm reduces the linear system's size greatly and at the same time maintains a satisfying accuracy. Compared with FastCap, the running time of the new algorithm can be reduced more than a magnitude and the memory usage can be reduced more than thirty times.