UC Davis PDG:2-D ECE Images from TEXTOR

The ECE Imaging system forms a 16x8 image of the plasma, with 16 spatial channels and 8 spectral channels. Composite pictures of size 16x24, such as that shown here, can be generated by combining images from multiple plasma discharges in which the image plane is shifted via (a) changes to the TEXTOR toroidal field, or (b) changes in the local oscillator (LO) frequency applied to the array. This system will soon (in 2006) be upgraded to double the plasma coverage, forming a pair of 16x8 electron temperature Te images whose separation can be varied. The diagnostic port access and available window size on TEXTOR, coupled with recent technological developments, permit the formation of Te images as large as 24x48 should additional diagnostic development funds become available.

One of the first physics studies to be performed with the 2-D ECE Imaging system on TEXTOR is that of sawtooth oscillations that are preceded by fast magnetohydro-dynamic (MHD) precursors. The most famous model to explain the sawtooth crash as a periodic magnetic reconnection was introduced by Kadomtsev. Several explanations have subsequently been put forward to clarify experimental features that contradict the Kadomtsev model, but no reliable model for the internal disruption in general and for the sawtooth oscillation in particular has yet been found. In order to reconcile these discrepancies, and to improve the present theoretical understanding of the driven reconnection process, one requires higher dimensional insight than conventional 1-D information into the core plasma dynamics. Such information is now available through the use of the 2-D ECEI system on TEXTOR.

In a sawtooth crash, the core Te profile is first flattened during reconnection and piled up outside of the inversion radius (the radial position where the change of Te is minimum during reconnection). The hot spot, partially shown in whitish yellow color corresponds to a peaked Te profile within the sawtooth inversion radius. Note that the inversion radius drawn here (the double white line) is estimated based only on the ECEI measurements; an independent measurement of the current profile will be necessary to definitively establish this q~1 surface. The observed process is indicative of an “x-point” magnetic reconnection in the partial reconnection model. Here, the hot island approaches the inversion radius and it appears that the heat is crossing the inversion radius at the upper corner of the low field side. Prior to this, there is a signature (mode structures in frames #1, 2) of the resistive ballooning mode. As the heat is crossing the inversion radius, there is a sign of a cold spot in the lower portion of frame #4. Here, the crash time is ~150 msec and one period of m=1 mode is ~170 msec.

In order to visualize the heat transfer beyond the inversion radius to the mixing zone, a sequence of images of the heat transfer processes is illustrated in the figure above right. This set, covering a larger radial range than the figure above left, is produced by interleaving the data from three separate discharges with overlapping radial coverage. This was obtained by producing three discharges with slightly different magnetic field (2.3 T, 2.35 T and 2.4 T) but otherwise identical, and synchronizing the timing of the sawtooth crashes. Since the change of magnetic field was less than ±2%, there was little change in plasma parameters such as temperatures and density. We also believe there was only a minute change in the q~1 layer, if any. The preliminary analysis demonstrates that the temperature in the vicinity of the inversion radius is not perturbed through the crash, whereas the hot spot is clearly present before and after the crash.

For further technical details, and ECE Imaging measurements and results from TEXTOR, please examine the following links:

* Further technical details on TEXTOR, and the TEXTOR ECE Imaging system

The unique features of the ECE Imaging diagnostics derive from the use of wideband, low cost Schottky diode mixer arrays coupled with innovative, low cost electronics. Follow the links below to learn more about the technology employed in ECE Imaging.

* Imaging array design and fabrication
* 2-D ECE Imaging electronics
* Quasi-optical notch filters

UC Davis has fabricated and installed multichannel ECE Imaging systems on other fusion plasma tokamaks across the world. Follow the links below for a description of the systems involved, and to sample data collected with these systems. 

* ECE Imaging on the KSTAR tokamak in Korea
* ECE Imaging on the TEXT-U tokamak in the U.S.A. (reference only)
* ECE Imaging on the RTP tokamak in the Netherlands (reference only)

email Comments to: Calvin Domier