Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.796912
Title: Field aligned flow in 2-dimensional magnetofluids and the genetic algorithmic solution of Poisson's equation
Author: Ireland, Jack
Awarding Body: University of Glasgow
Current Institution: University of Glasgow
Date of Award: 1994
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
This thesis splits naturally into two topics, field aligned flow in a two dimensional plasma, and the application of genetic algorithms to the solution of Poisson equations. Genetic algorithmic techniques were developed as a new method of numerical solution to a problem arising in field aligned flow. The relation between plasma physics and computing (particularly novel computing methods) is introduced in chapter I. In chapter 2, we begin with Maxwell's equations and a fluid description of a plasma, and derive under various assumptions, equations governing the structure of a field aligned two dimensional plasma. The appearance of field aligned flow in the Earth's magnetotail is discussed along with some treatments in the literature. Chapter 3 examines the fields arising from having the fluid flow along the field lines time independent. It is shown that only very special fields support exact field aligned flow. These fields can be classed by their corresponding flow function. An equation is derived that describes fields where the flow function is a constant everywhere, which provides the spur for genetic algorithm application. Some magnetotail relevant solutions of this equation are presented. Chapter 4 investigates time dependent field aligned flow. It is shown that this situation is somewhat more complicated than the time independent case, and that a singularity in the flow may appear, indicating the presence of a large fluid acceleration and the breakdown of the present model. In chapter 5, the basic concepts of genetic algorithms are introduced. An algorithm is developed to test the efficacy of this method for application to the solution of a class of ordinary differential equations. This work is built on in chapter 6, where a Poisson equation solver is constructed. Comparisons are made between this and other more traditional methods. Finally, chapter 7 describes some possible extensions to the work presented. Suggestions for both genetic algorithms and field aligned flow are discussed. Appendices A and B contain a listing of the Poisson solver POISGEN and sample input files respectively.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.796912  DOI: Not available
Share: