MSE
The current density (or safety factor) profile at TEXTOR is measured with a newly installed Motional Stark Effect (MSE) diagnostic. This measurement relies on the spectral shape of the Balmer-a emission of the hydrogen or deuterium neutral beam. Injected fast neutral atoms with velocity vb moving in a magnetic field B experience a Lorentz electric field El = v x B, which causes a Stark splitting of the emitted Balmer-a line (see Figure 1). This light is either polarised parallel to the electric field (p polarised component, for the Dm = 0 transition, m being the projection of the total orbital momentum on the electric field direction) or polarised perpendicular to the electric field (s polarised component, for Dm = ±1). The geometry of the measurement system secures clear separation of the plasma emission from beam emission by the Doppler shift. (The three beam components are separated from each other as well). Now by determining either the polarisation direction of the p or s components, or the ratio between the p and s intensities, the direction of the magnetic field and hence the pitch of the field lines can be determined. If in addition the radial position of the measurement is included in the analysis, the q-profile or the current density can be determined.
Figure 1: Typical spectra in the plasma core (left) and edge (right) as recorded by the MSE diagnostic (full line) and the theoretical description of the spectrum (fit) by the dashed line. The three groups of emission correspond to the full, half and third energy components of the neutral beam. Within each group the ratio of s (center line) to p (left and right lines) intensities provide a measure for the magnetic pitch angle.
At most existing tokamaks narrow band interference filter are used in the MSE systems to detect one of the lines and measure its polarisation direction by photoelastic modulators. In contrast, in the MSE system operational at TEXTOR the full polarization spectrum of the emitted light is measured. The advantages of exploiting the full spectral information are obvious:
- the system is insensitive to variations in beam velocity or magnetic field,
- the Doppler shift is measured as well, allowing to determine the observation volume directly from the spectrum,
- the polarisation is measured at several lines simultaneously, increasing the accuracy of the deduced direction of the magnetic field,
- the Stark splitting is obtained allowing the compute the magnitude of the magnetic field and finally,
- if in addition to Motional Stark electric field a radial electric field is present in the plasma this can, albeit with limited accuracy, be determined as well by comparing the polarisation direction of two different energy components of the neutral beam.
A trade-off of spectra-polarimetry compared to the traditional approach is that the optical throughput is reduced. However, in the present case this does not limit the time resolution since the signal level is high enough. A second trade-off is that the data analysis of full spectra-polarimetry is more complex.
An example of a measured MSE spectrum is shown in Figure 1 along with the fitted theoretical curve. Two q-profiles measured in a single TEXTOR discharge, but at different values of the plasma current are depicted in Figure 2. The q-values in the figure have been derived from fits to the spectra. The curves in the figure are quadratic q-profiles with the position of the q=1 radius and the value of q at the edge as constraint.
Figure 2: Two q-profiles measured in the same plasma discharge but at different values of the plasma current.
Lately a collaboration has been ste up with the group of Prof. John Howard from Australian National Universtiy (ANU). They exploit a system based on imaging the coherence of the MSE light. This system has now been put in place at TEXTOR and first results are encouraging (see Figure 3).
Figure 3: (a) Image of interior of TEXTOR vacuum vessel as recorded by the spectro-polarimeter. (b) Plasma emission within the 660nm filter pass-band (no neutral beam emission) for discharge #108248. Notice the absence ofinterference fringes. (c) Image of the neutral beam. Visible interference fringes indicate that the light is polarized. Successive images can be demodulated for the polarization orientation.
