"Investigations of the magnetic structure and the decay of a plasma-gun-generated compact torus"
W. C. Turner et al.

Summary:

This paper contains A LOT of information, so I'll only summarize parts that are somewhat relevant to my project.

Compact toroidal (CT) plasmas are potentially advantageous configurations for fusion reactors for a number of reasons, including a topologically simple magnetic coil structure, small size, and relatively high ratio of plasma thermal energy to magnetic energy. Two types of CT plasmas have been identified, those with no toroidal field (field-reversed theta pinch) and those with finite toroidal field (spheromak). This paper concerns spheromaks produced in the laboratory using magnetized coaxial guns. Unlike SSX, all shots contained single spheromaks (no spheromak merging), although magnetic reconnection could be achieved by inserting a "pinching coil" between the gun and vacuum chamber. The lifetime of plasmas studied was less than 500 μs. The decay time was lengthened somewhat by improving the vacuum in the gun and flux conserver, thereby lowering the impurity ion concentration and decreasing the amount of radiation emitted. Experimental diagnostics included a magnetic probe array, an HeNe interferometer for measuring electron line density, a pyroelectric detector for measuring radiated power, a piezoelectric probe for measuring plasma flow pressure and thermal pressure, and a vacuum ultraviolet (VUV) spectrograph for looking at spectral lines.

Sections are included deriving some theoretical relationships for magnetic relaxation, analyzing the magnetic probe data, and applying these results to a discussion of the stability of the plasma configuration. Analysis of the power output and magnetic field decay lead to the conclusion that the most important plamsa energy loss process is impurity radiation. Although glow discharge cleaning improves plasma lifetime, this improvement is limited even after many hours of cleaning, suggesting that radiation persists as an important factor in field decay.

A complete VUV spectrum between 150-3200 Angstroms is presented. The strongest lines observed were CIII 117.5 nm, CIV 154.8 and 155.1 nm, and CIII 229.7 nm. Nearly all of the lines were from carbon or oxygen, with no lines from metal ions such as Ni, Fe, or Cu that might have been released from the plasma gun of flux consever walls. There are a few weak N and Si lines. No strong CV lines were seen in the wavelength range considered. Detailed time-evolutions of the 229.7 nm and 155 nm lines are presented, and the ratio line ratio appears to be approximately 1/4, although the monochromator was not absolutely calibrated.

A 0-D simulation model was developed to compute plasma properties under the assumption that the plasma might not have time to reach coronal equilibrium. The total radiation output calculated by these simulations with an electron density of 2.6 x 10^14 cm^-3 and an initial electron temperature of 10 eV is consistent with the experimental results. Under these conditions, the simulation predicts that the ratio of the CIII [? nm resonance line] to the CIV [155 nm line?] is about 10/1. Very little CV emission is predicted; this is confirmed by observations. The plasma decay time drops sharply as the carbon impurity fraction increases above about 20%, but is roughly constant for low levels of impurities.

Measuring the plasma density with the interferometer, the authors found only a weak linear dependence on the gas valve filling pressure, implying that only a fraction of the gas originally in the plenum chambers becomes ionized plasma. The density likewise increased only slightly when the time delay between the gas valve opening and the capacitor discharge was increased, but there was a strong dependence on the electrical energy input to the gun discharge.

Applications to my research:

This paper is highly relevant to my research because the authors used a VUV monochromator to look at the same carbon emission lines that we have been observing in SSX. In particular, the results presented provide an independent confirmation of the CIII 229.7 nm to CIV 155 nm line ratio, which has been measured in SSX to be in the vicinity of 1/5. This ratio is of particular interest to us because the experimental results have been in large disagreement with the model spectra I've calculated using PrismSPECT. Another point of interest is that the CIII 117.5 nm line, which we have been unable to detect in SSX, was quite strong, while the CIII 97.7 line (which we have seen in SSX) was weak. The authors used a simulation code to predict line ratios and other plasma properties, but their code was less sophisticated than PrismSPECT, and they only present limited results.