Helicon Waves in High Density Plasmas
FTP the Postscript version of my thesis
- Preamble
- Chapter 1 Introduction
- Chapter 2 Experimental Apparatus and Diagnostics
- Chapter 3 Theoretical Treatment of Helicon Waves
- Chapter 4 Plasma Formation Using a Double Saddle Coil Antenna
- Chapter 5 Helical and Phased Antennas
- Chapter 6 Diffusion during Plasma Formation
- Appendix A Data Management
- Bibliography
Abstract
This thesis investigates helicon wave propagation in a 1.5m long, 2.5cm radius, high density (ne > 5x1018m-3), high static magnetic field (B0 > 0.05T) plasma, where the plasma has been produced by the helicon wave. Helicon waves launched with a double saddle coil antenna were found to consist of the first radial, m=+1 azimuthal mode, with no observable m=-1 mode under any conditions. The wave was found to propagate with the infinite plane wave dispersion relation k|| \propto (ne/B0)½ as if the plasma had no boundaries. Comparisons of experimental measurements of wavelength and radiation resistance with predictions of a magnetohydrodynamic numerical model show very good agreement. The numerical model also indicates that, for the conditions used in the experiment, the radiation resistance of the m=-1 mode is too low for it to be significantly excited.
Plasma formation by a helical antenna was also investigated.This antenna can excite m=+1, and m=-1 modes in opposite directions along the field. The helical antenna produced an axially asymmetric plasma, with the plasma density highest along the axial direction of propagation of the m=+1 mode. No m=-1 mode was again observed and the plasma extended only a short distance in the direction where the m=-1 mode would be expected to propagate. The helical antenna was found to produce plasma more efficiently than the double saddle coil antenna, however it had a more limited range of operating conditions due to its high k|| selectivity.
Attempts to produce plasma using helicon waves of azimuthal modes other than m=+1, by means of a phased double saddle coil antenna were unsuccessful. However, by launching waves into a pre-formed plasma it was possible to control the azimuthal mode content of the wave launched. When the antenna was phased for excitation of m=-1, significant m=+3 was observed and no m=+1 mode, as expected.
Observations were made of the plasma formation process in the early phase of the discharge. During the initial stages of the discharge, as the plasma diffuses away from the antenna, a high density peak was observed downstream from the antenna, near the end of the uniform field region. A 1-D diffusion model was used to investigate this phenomenon, and predicted increased downstream ionisation due to a directed flux of neutrals created by the expansion of the plasma from the source region. After a period of about 10msec this density peak collapses and the plasma density drops to a uniform lower value that typifies the static phase of the discharge.