Browsing by Author "Moodley, Mervlyn."

Now showing 1 - 3 of 3

Corrosive sulphur in transformers.
(2013) Naicker, Sharlene-Asia.; Moodley, Mervlyn.
Failures in power transformers, in the majority of cases, have been linked to the formation of copper sulphide (Cu₂S) and corrosive sulphur. Cu₂S is conductive and affects the dielectric properties of the insulation system in the transformer. The formation of Cu₂S and corrosive sulphur in transformers is of worldwide concern to power utility companies, like Eskom, and large industrial manufacturing plants that maintain their own transformers. This research deals with determining and understanding the formation mechanism of Cu₂S as well as investigating factors that influence the acceleration of the corrosive sulphur formation. Data from oil test results was obtained from an experimental set-up belonging to eThekwini Electricity. The set-up consists of two 100 kVA transformers, one containing corrosive oil and the other containing clean oil. While varying the load and temperature of the transformer, oil samples were taken at various intervals and tested for corrosive sulphur by monitoring the concentration of dibenzyl disulphide (DBDS) and the dielectric strength of the oil. This data was used to investigate the reaction rates, activation energies and various thermodynamic parameters of the corrosive sulphur and Cu₂S, and to establish the factors affecting their formation. At high temperatures, the DBDS concentration was found to reach equilibrium. The activation energy for the DBDS reacting with copper and further formation of DBDS was found to be 47.4 KJmol⁻¹ and 35.2 KJmol⁻¹ respectively. This research also determined that a significant amount of the initial concentration of DBDS was needed in order to react with copper. This experimental study also showed that DBDS and Cu₂S reactions do not play a major part in influencing the physical properties of the transformer and transformer oil. Surface physics concepts were employed to discuss the interaction dynamics of Cu₂S on copper surfaces. The main focus of the surface physics investigation compared the results from this experiment with results from other surface physics investigations in published literature.
Phase transitions in induced lattice gauge models.
(1995) Moodley, Mervlyn.; IIchev, Assen.
The present research is based on the study of the phase structure of lattice models incorporating selfinteracting scalars and gauge background fields otherwise known as induced gauge models. Emphasis is placed on the effect the choice of the integration measure over the radial modes of the scalar fields have on the phase structure of these models. Both numerical simulations and analytical results based on the mean field approximations are presented. In Chapter 1 an introduction to quantum field theory is given leading to the formulation of Euclidean quantum field theory. In Chapter 2 global and local gauge invariance together with the mechanism of spontaneous symmetry breaking are discussed. In Chapter 3 the formulation of quantum field theory on the lattice is introduced. The lattice regularization entails discretizing space and time and presents an elegant approach to studying certain phenomena of the continuum theory which are beyond the reach of standard perturbative analysis. In Chapter 4 the Monte Carlo methods for evaluating the Euclidean Feynman path integral as applied to lattice gauge theory are discussed. In Chapter 5 numerical studies of some lattice gauge models are presented. Both pure lattice gauge models and gauge-Higgs models are examined. In Chapter 6 the Kazakov-Migdal model which presents an interesting approach to inducing QCD is discussed. In Chapter 7 the mixed fundamental-adjoint induced model is introduced. This model succeeds in breaking the local ZN symmetry of the Kazakov-Migdal model by adding to it scalar fields in the fundamental representation of the gauge group. The effect of the choice of the radial integration measure on the phase structure of a class of Abelian induced models is studied.
Two-level system coupled to a structured environment - a computational study.
(2015) Pooran, Nirav.; Moodley, Mervlyn.
It is well known that analytical solutions to large complex problems are not tractable. Computational or numerical simulations of these types of systems allow one to investigate regimes where analytical methods fail. In this thesis the dynamics of an open quantum system which does not have an exact analytical solution is calculated numerically. The model investigated is a two-level system coupled to a non-trivial structured environment. There are two regimes of interaction between the system and environment, which is of interest in this thesis, namely the weak and strong coupling regimes. These different regimes are obtained by tuning the interaction strength parameter in the model that also define whether the model is Markovian or non-Markovian. Other parameters of the model are also varied in order to gauge thermalization and relaxation rates during interaction. Computational or numerical simulations can become a long tedious process when one is running simulations on a single core or even a dual core machine, since these simulations can run for days and in all cases require multiple runs to obtain statistics. With the development of clusters which contain thousands of cores and parallel computing software, running simulations is now very efficient and not a tedious process as before since parts of the code can be run over multiple cores. In this thesis the code is written in Python and uses the parallel computing software library called Message-Passing Interface (MPI). In the first chapter two types of quantum systems, namely closed and open quantum systems, are introduced. The standard equations generally used for these two types of systems are presented and the differences in the treatment of these systems are pointed out. The Heisenberg and interaction pictures are also described in this chapter. The model is then discussed in depth in chapter two where the Hamiltonian of the total system is introduced and the interaction picture Hamiltonian is derived. An overview of approximate analytical solutions to the model is discussed. Chapter three discusses parallel coding specifically MPI and the code for the simulations are presented and explained. In chapter four plots are done simulating the model and discussing the effects and outcome of changing parameters which govern the interaction and thermalization of the system and the environment.