TRIGGERING OF ITBS BY LOW-ORDER RATIONAL qmin AND ALFVEN GRAND CASCADES

In this section, we first consider in detail the spectroscopic pattern of an AC in reversed-shear tokamak discharges. Figure 10.5 shows the pattern consisting of many ACs with different toroidal mode numbers n.

In accordance with the theory/modelling presented in Chapter 9, the time appearance of every AC is associated with q_mm passing through the relevant rational number. The Alfven spectroscopy in this case is based on the clustering of ACs with different n’s in time [10.2], that is,

n = 1 ACs occur when q_mm = 1, 2, 3...; n = 2 ACs occur when q_mm = 1, 3/2, 2...; n = 3 ACs occur when q_mm = 1,4/3, 5/3...

The bunch of ACs, in which all n’s are present at once is called “grand cascade.” It occurs when q_min passes an integer value. Figure 10.5 shows how the ACs appear one-by-one just before the time when q_min passes an integer value of 2, while they all emerge at once just after such event at /~5.15 s. Because of the easily recognisable characteristic pattern of the grand cascades, it is easy to diagnose experimentally the times when c/_min passes integer values in JET shear-reversed discharges with ICRFI or NBI exciting ACs. By comparing the grand cascade times and the times of ITB formation in shear-reversed JET plasmas, it was established that these two times of very different origin correlate [10.8].

Further investigation has shown that low rational values of q_mi„ in plasmas with non-monotonic <7(r)-profiles cause improved electron confinement seen as an increase by ~10%-20% in the central electron temperature [10.9]. We call this spontaneous improvement of the confinement an

“ITB triggering event.” If no significant heating is applied to the plasma at that time, the ITB triggering event gradually disappears in ~100-200 ms. However, if a significant main heating is applied just before the ITB triggering event, a full-scale ITB could be developed. Therefore, we conclude that the ITB triggering event is a necessary but not sufficient condition for producing an ITB. The timing of the main heating power determines whether the ITB triggering event can end up in a full-scale ITB. Figure 10.6 illustrates how the ITB triggering event looks like in the same JET discharge, as shown in Figure 10.5.

The correlation between Alfven grand cascades and the ITB triggering events was investigated in JET experiments with different types of pre-heating, that is, with LHCD, ICRH+LHCD, ICRH only, NBI only, and with pellets. The correlation was found to exist in JET plasmas with densities up to ~5x 10^l9nr³, indicating that the timing of ITB triggering from the Grand-ACs may facilitate

FIGURE 10.6 Electron temperature traces measured with multi-channel ECE diagnostics. A sudden increase in the slope is observed at 5.1 s, indicating a sudden improvement in electron confinement.

FIGURE 10.7 Correlation between the times of ITB triggering times and grand cascades.

FIGURE 10.8 Time delay between Grand-ACs and ITB triggering events for a set of discharges with interferometry diagnostic used for the AC detection.

ITB scenario development in machines with high densities. Several plasma conditions were investigated too,

• • l.5я<2.2 MA
• • 2.45 T< 3.4 T
• • 3total<17 MW.

The results of these studies are presented in Figure 10.7. Similar observations have been made on DIII-D [10.10].

For using this kind of Alfven spectroscopy in practice, a discharge with ICRH power is run at the beginning of a specific experimental session. All Grand-ACs are identified in such discharge, thus providing the time sequence of ITB triggering events and q_mm (t) = integer surfaces appearing in this particular plasma. The timing of the main heating power is decided then for the rest of the session for making an ITB associated with one of the integer values of q_min.

With the improved time resolution of AE detection (see Figure D.4 in the Appendix D), it also became possible to investigate more accurately whether the ITB triggering event is preceding Grand-АС marking q_min(t)=integer or it occurs just after that. Figure 10.8 strongly suggests that the ITB triggering events occur before the appearance of the integer surface in the plasma. The most plausible theory that may explain such sequence of events is the theory associated with the depletion in time of rational magnetic surfaces of drift waves turbulence [10.11,10.12] just before q_mm reaches an integer value, rather than the presence of an integer q_mm value itself.

PLASMA MASS DETERMINATION FROM ACTIVE AE MEASUREMENTS

The square root dependence of Alfven velocity on the plasma mass has suggested from the very beginning of MHD spectroscopy the diagnostic potential of the AE frequency measurements for determining the plasma isotopic mix [10.1]. After the discovery of the gap AEs, including TAEs and EAEs, this diagnostic potential was demonstrated during JET DT campaign with the use of active TAE diagnostic described in Section 7.1 [10.3]. The measurements of TAE frequency were performed in a series of nearly identical JET discharges with varied D:T concentrations [10.13].

Figure 10.9 presents the results obtained with three different diagnostics including the active TAE for two JET discharges with different range of the D:T ratio. The figure suggests a good

FIGURE 10.9 Estimate of D-T ratio from the measured frequency of an externally excited n= 1 TAE compared with visible spectroscopic data (intensity ratio of T₍, to D₍, lines) and with the neutral particle analyser results*.

agreement with edge spectroscopic measurements, although with a somewhat different time evolution. This method may be useful in a reactor where the same plasma configuration will repeatedly be employed. Direct measurements are possible in real time for similar equilibria, w'ith a time resolution of ~30 ms, which is the time needed to scan the frequency of the probe wave across a TAE resonance. The overall accuracy depends on the reproducibility of the equilibrium. When the equilibrium is significantly varied from shot to shot along with the isotopic mix, one must rely upon a full theoretical analysis.

REFERENCES

• 1. J.P. Goeldbloed et al., Plasma Phys. Control. Fusion 35 (1993) B277.
• 2. S.E. Sharapov et al.. Phys. Lett. A 289 (2001) 127.
• 3. A. Fasoli et al., Plasma Phys. Control. Fusion 44 (2002) B159.
• 4. G.J. Kramer et al.. Nucl. Fusion 40 (2000) 1383.
• 5. S.E. Sharapov et al.. Fusion Energy Proc. 21st Int. Conf. (IAEA, Chengdu. 2006) paper EX/P6-19.
• 6. E.M. Edlund et al.. Phys. Rev. Lett. 102 (2009) 165003.
• 7. P. Sandquist et al., Phys. Plasmas 14 (2007) 122506.
• 8. E. Joffrin et al., Nucl. Fusion 43 (2003) 1167.
• 9. S.E. Sharapov et al.. Nucl. Fusion 46 (2006) S868.
• 10. M.E. Austin et al., Phys. Plasmas 13 (2006) 082502.
• 11. A.D. Beklemishev and W. Horton, Phys. Fluids B4 (1992) 200.
• 12. X. Garbet et al., Nucl. Fusion 43 (2003) 975.
• 13. P.R. Thomas et al., Phys. Rev. Lett. 80 (1998) 5548.

Reproduced from [A. Fasoli et al., Plasma Phys. Control. Fusion 44 (2002) B159], with the permission of IOP Publishing.