No description available.

No description available.

No description available.

No description available.

One-dimensional electrostatic plasma simulation using the particle-in-cell technique is used to study the spectral features of stimulated electromagnetic emission (SEE). SEE is a potential diagnostic tool to study the ionosphere and its spectrum provides a different view of the heated region from the incoherent scatter radar. At this time, a unified and complete theory which explains the SEE phenomena in detail does not exist. The SEE simulations we discuss are proposed to provide interpretation of many of the past puzzles of the experimental data, as well as to facilitate the design of future SEE experiments and the theoretical development of SEE. In the numerical simulation, only the upper hybrid layer where the geomagnetic field is essentially perpendicular to the density gradient is modelled. Three of the SEE features, namely the downshifted maximum (DM), upshifted maximum (UM) and broad upshifted maximum (BUM), are suggested to be generated at the upper hybrid layer. We observed these three features which have many similarities in the simulation. It is evident that the DM and UM are generated by the same parametric instability involving lower hybrid waves while the BUM is produced by other different mechanisms. Boundary effects are found important on the generation of all three features in the simulation. Moreover, detailed investigation of the simulation results raises a number of questions concerning detailed generation mechanisms of SEE which have not been considered and answered in the past. Besides the DM, UM and BUM features, the quenching of DM is also observed in the simulation when the pump frequency is very close to electron cyclotron harmonics. It is concluded that both the cyclotron damping and mode conversion of the upper hybrid wave into electron Bernstein modes are possible causes. Finally, some suggestions for the future SEE simulation are included. / Master of Science

It has been shown that the energy exchange between two objects can be greatly enhanced when the separation between the objects is on the order of the wavelength of thermal emission. The earliest theoretical and computational work focused on simple planar and spherical geometries, or they resorted to approximations that separated the object to outside of the thermal wavelength \(\lambda_T = hc/(k_BT)\). Since those original works, the study of near-field energy exchange has expanded to object shapes that can be described by a separable coordinate system using a spectral expansion of the dyadic Green function of the system. The boundary element method has also been used to study arbitrary shapes in thermal equilibrium. Application of these new expansion methods to general shapes out of thermal equilibrium will facilitate in the optimization of nanoscale structures. A three step process is used to investigate the effects of object shape on the total and directionality of the energy exchange between objects. First, a general expression for the energy flux between the objects will be formulated. Second, a computational method to evaluate the expression will be implemented. Finally, the effects of varying the surface geometry will be explored. The computational results demonstrate that the total energy exchange between two bodies is influenced by the surface shape of the objects even when the surface areas are held constant. While the primary increase over the classical blackbody energy exchange \(\sigma T^4 A\) is primarily governed by separation of the surfaces, we show that the view factors from classical far-field radiative transfer can be used to predict the change in the total energy exchange from a reference configuration at the same separation when the surface area of the two objects is comparable. Additionally, we demonstrate that the spatial distribution of the energy exchange can be localized into small spatial region with a peak value increased over \SI{30}{\percent} by using two objects with dramatically different projected areas.

published_or_final_version / Physics / Doctoral / Doctor of Philosophy

No description available.

Integrated photonics is the field of shrinking and simplifying the fabrication of devices that guide and manipulate light. It not only offers to greatly lower the size and cost of systems used in optical communications it also offers a platform on which new physical phenomenon can be explored by being able to fabricate and manipulate structures on the scale of the wavelength of light. One such platform in integrated photonics is that of two-dimensional slab photonic crystals. These structures exhibit a photonic band-gap, a band of optical frequencies that are prohibited from propagating within the medium, that can be used to guide and confine light. When used to create photonic crystal waveguides these waveguides exhibit unique dispersion properties that demonstrate very low optical group velocities, so called "slow-light". This dissertation begins with the practical realization of design and fabrication of such waveguides using the silicon-on-insulator material system using conventional deep-UV photolithography fabrication techniques. It will detail and demonstrate the effect physical dimensions have on the optical transmission of these devices as well as their optical dispersion. These photonic crystal waveguides will then be used to demonstrate the enhancement of nonlinear optical phenomenon due to the slow-light phenomenon they exhibit. First spontaneous Raman scattering will be theoretically demonstrated to be enhanced by slow-light and then experimentally shown to be enhanced in a practical realization. The process of four-wave mixing will be demonstrated to be enhanced in these devices and be shown to be greatly affected by the unique optical dispersion within these structures. Additionally, we will examine the dispersion that exists in silicon nitride microring resonators and the effect it has on the use of these devices to generate optical frequency combs. This is done by leveraging the dispersion measurement methods used to characterize photonic crystal waveguides. We conclude this work by examining the avenues of future work that can be explored in the area of photonic crystal waveguides.

A dissertation submitted to the Faculty of Engineering and the Built Environment, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of Science in Engineering, 2016 / The work presented extends and contributes to the research of modelling lightning return stroke (RS) electromagnetic (EM) fields in the time domain. Although previous work in this area has focused on individual lightning electromagnetic pulse (LEMP) modelling techniques, there has not been an investigation into the strengths and weaknesses of different methods, as well as the implementation considerations of the models. This work critically compares three unique techniques (Finite Antenna, FDTD, and Single Cell FDTD) under the same ideal simulation parameters. The research presented will evaluate the EM fields in the range of 50m to 500m from the lightning channel. This range, often referred to as the near field distance, has a significant effect on lightning induced overvoltages on distribution lines, which are primarily created by the horizontal EM fields of the RS channel. These close distances have a significant effect on the model implementations, especially with the FDTD method. Each of these modelling methods is explained and tested through examples. The models are implemented in C++ and have been included in the Appendix to aid in future implementation. From the model simulations it is clear that the FDTD method is the most comprehensive model available. It allows for non-ideal ground planes, as well as complex simulation environments. However, FDTD has a number of numerical related errors that the Finite Antenna method does not suffer from. The Single Cell FDTD method is simple to implement and does not suffer from the same numerical errors as a full FDTD implementation, but is limited to simple simulation environments. This work contributes to the research field by comparing and evaluating three techniques and giving consideration to the implementation and the applicability to lightning EM simulations. / MT2017