During spheromak merging, magnetic reconnection releases stored magnetic energy, and much of it ends up as heat. So, understanding the temperature distribution will help us understand magnetic reconnection. We already had one tool to measure this -- the VUV monochometer. However, we are attempting to developed another sensor -- the soft x-ray detector.
The soft x-ray detector (SXR) is comprised of photodiodes with thin metal film filters in front of them. The below graph shows the response functions of the unfiltered and filtered photodiodes.
By measuring the photodetector output, we can determine the photon energy, and thus the plasma's temperature.
However, SXR measured anomalously strong signals -- typical Sn signals were nearly twice as strong as Al signals, and 5 times stronger than Ti and Zr. One possibility was that the Sn filter allowed lots of UV or visible light through, and this lollipop like device was built to test that hypothesis. The two filters are UV-fused silica (cutoff above 7.3 eV) and sapphire (cutoff above 8.3 eV), either could be rotated to be in front of the SXR sensors. However, using this device, we found that less than 2% of the Sn filter signal came from UV or visible light, ruling that out as a cause for the anomaly.
This is the sort of data we could get from SXF -- here we show the ratios of signals from different diodes. It took a lot of computation to turn this in to temperature measurements.
This is the best fit temperature profile for counter-helecity merging calculated from SXR data.
For comparison, PrismSPECT was used to model the spectra for a variety of plasma conditions. These simulations showed that the plasma reached a steady state conditions about 10μs after leaving the gun -- before reconnection which occurred at around 40μs.
Finally, the VUV monochrometer (center) was used to measure temperatures for comparison.
This is the temperature profile for counter-helecity merging calculated from VUV data averaged over 25 runs. Uncertainty is shown with dotted lines. The large uncertainty towards the end is due to variations in the plasma structure from shot to shot.
The average temperature of ~40eV measured with SXR is within the uncertainty range for the VUV measurements. However, the expected spike in the electron temperature during reconnection is absent from SXR data -- but clearly visible in VUV data. It is unclear if we will be able to use SXR for accurate temperature measurements. On possibility, which we are investigating, is that the anomalous signals are due to a non-Maxwellian electron velocity distribution.
Vernon's APS-DPP poster.
December 6, 2006 /doc@swarthmore.edu
