Characterization of the mutual influence of Ion Cyclotron and Lower Hybrid Range of frequencies systems on EAST

Waves in the Ion Cyclotron (ICRF) and Lower Hybrid (LH) Range of Frequencies are efficient techniques respectively to heat the plasma and drive current. Main difficulties come from a trade-off between good RF coupling and acceptable level of impurities release. The mutual influence of both systems makes such equilibrium often hard to reach [1]. In order to investigate those interactions based on ScrapeOff Layer (SOL) plasma parameters, a new reciprocating probe was designed allying a three tips Langmuir probe with an emissive wire. The emissive filament provides a precise measure of plasma potential [2], which can be used to calibrate Langmuir probe’s results. This paper reports on experimental results obtained on EAST, where there are two ICRF antennas and two LH launchers. Among others diagnostics, the new reciprocating probe enabled to evidence the deleterious influence of ICRF power on LHWs coupling in Lmode plasmas. In areas connected with an active ICRF antenna, SOL potentials increase while densities tend to decrease, respectively enhancing impurities release and deteriorating LHWs coupling. This phenomenon has mostly been attributed to RF sheath; the one that forms on top of Plasma Facing Components (PFCs) and causes ExB density convections [3]. From those experiments it seems ICRF has a strong influence on magnetically connected areas, both in the near field – influencing ICRF waves coupling – and in farther locations such as in front of LH grills. Moreover, influence of ICRF on LH system was observed both in L and H modes. Those results are consistent with RF sheath rectification process. Concerning the influence of LHWs on ICRF coupling, nothing was observed in L-mode. Besides during Hmode experiments, LHWs have been identified as having a mitigating effect on ELMs [4], which on average lowers the pedestal, increasing edge densities to the profit of ICRF waves coupling.


Introduction
Lower Hybrid Current Drive (LHCD) consists in launching waves at a frequency lying in the GHz range of frequencies.
Excited by klystrons, waves are transported into waveguides and launched in the plasma from a grill. Since LH Waves (LHWs) are evanescent at too low densities (bellow 7.10 17 m -3 at 4.6GHz), gas puff is often used to increase electron densities locally in front of the grill to improve coupling efficiency. Thanks to their large parallel electric field, LHWs have the interesting property of damping very efficiently at high parallel phase velocities to the electron thermal speed. They are consequently able to drive off-axis current, allowing good control of plasma current profile, enabling to build internal transport barriers and access attractive steady-state operations with high bootstrap current. On ITER, LHCD system should enable to reach more than 70% of bootstrap current, with high β N (~3) and good confinement (H ITER-89~2 .5) [5]. Ion Cyclotron Range of Frequencies (ICRF) goes from 10 to 100MHz. ICRF waves are mostly being used for heating plasmas on tokamaks. Unlike LHWs, ICRF waves are excited from an antenna composed of current straps as close as possible to the plasma. ICRF waves are also evanescent at low densities and must tunnel through a thin evanescent layer before reaching higher density plasma where they can propagate normally until reaching a resonance layer and essentially exchange their energy with ions before being spread to the whole plasma by collisions. Unfortunately, single-pass absorption and coupling efficiency are never perfect, so that waves undergo several reflections in the Scrape-Off Layer (SOL), enhancing dramatically interactions with Plasma Facing Components (PFCs). Those interactions are generally the result of RF sheaths which characterization from a global point of view can be done through an increase of impurities production and consequently an increase in radiations. From a local point of view, ExB convections induce strong modifications on density profiles at the edge, often degrading both ICRF and LH waves coupling efficiencies. By presenting experiments recently conducted on Experimental Advanced Superconducting Tokamak, we aimed at characterizing interaction processes during ICRH and LH operations. A brief overview of EAST and the position of objects and diagnostics of interests for this study are given in section 2. Further details on the design of a newly designed reciprocating probe which enabled to collect key information

EAST presentation
Two ICRF antennas and two LH grills are available on EAST.
In order to minimize interactions between each launcher, their positions were changed after the 2014 campaign to maximize distances between them ( Fig. 1.). B and I ports ICRF antennas such as E and N ports LH grills are almost diametrically opposite in the torus to minimize mutual influences. While 6MW power is available on each ICRF antenna roughly working at same frequencies (34 and 35 MHz), each LH grill has different power, frequency, and wave number range; respectively 4MW for N-port at 2.45GHz with 2< n|| <3, and 6MW for E-port at 4.6GHz with 1.8< n|| <2.5. Langmuir probes were also fixed on each LH grill, respectively on top of the N-port and on the right side of the E-port. Reciprocating probes can be inserted from J and K ports, with an X-mode reflectometer in J-port [7] providing density profiles potentially useful to benchmark both probes results. However, the reflectometer's frequency range (from 4 to 6 GHz) is not appropriate to measurements in the SOL; it can only provide good results at densities above 10 18 m -3 . On the C-port, an Extreme UV spectrometer enables measurement of high-Z impurities in the plasma core [7].

Reciprocating Emissive and Langmuir probe
This probe (Fig. 2.) was specifically designed for the study of ICRF waves influence on the SOL. It consists in a combination of a 2cm tungsten wire of 150μm diameter, doped with thorium  to boost its emissive properties and a three tips Langmuir probes. Because emissive probe is hot, it enables to measure the potential of the plasma with much better precision than cold Langmuir probes. The idea is then to use emissive wire floating potential instead of Langmuir tips one, to calculate more precise values of plasma potential, electrons temperature and density. Reciprocations are fully controlled shot by shot from the main control room. Plunges timing, deepness, stationary laps inside the plasma are all set before each shot. Installation of the probe can conveniently be done in a small chamber which can be isolated from tokamak vacuum. Once the probe is ready, vacuum is restored after about two days pumping, and the probe can enter in the tokamak. This was particularly useful for experiments with the emissive probe, because unlike in machines with short discharges, EAST provides several seconds plasmas so that tungsten filament rapidly breaks after few plunges. Probes can be inserted into the middle plane of J and K ports, both well connected magnetically to I-port and not connected to B-port ICRF antennas (Fig. 3.).

Experimental results
Series of experiments on EAST aimed at characterizing the influence of ICRH on the SOL, by comparing the values of parameters measured by the reciprocating probes with and without ICRH, on L-mode plasma with currents of 450kA (Fig.  4). The probe was inserted two times per shot materialized by red and blue areas on Fig. 5, and remained deep in the SOL, behing the ICRF antenna limiter not to break the filament by plunging too deep. This explains why parameters measured are so small. Both ICRF antennas were also alternatively activated (Fig. 5), so that their respective influence can be compared. Figure 3 clearly shows that the K-port reciprocating probe was close and connected to the I-port antenna (red curves on Fig.  4), but far and not connected to the B-port antenna (blue curves on Fig. 4). Profiles in Fig. 4 show that ICRH significantly enhances potentials on connected regions where RF fields are more intense (red curve gets above the blue one as the probe approaches the plasma), presumably leading to ExB density convections consistent with pump-out effects observed (red curves). SOL rectifications in regions far and not connected to the active antenna are much smaller than in connected cases, yet those rectifications are significant as can be seen comparing blue (non-connected ICRF antenna) with dashed black curves (no ICRH).  Purple curve corresponds to plasma oscillations D α and radiations are in magenta.
During another series of experiments (Fig. 6), we tried to characterize the sheath formation at proximity and faraway of ICRF antennas for different power levels using a Langmuir reciprocating probe on the J-port and plunging until the LCFS. Fig. 7 shows floating potentials radial profiles measured for each reciprocation. Blue curves correspond to B-port launcher, which was launching relatively low power, was not magnetically connected and too far to significantly influence SOL parameters of the J-port. Red curves corresponding to Iport antenna launching much higher power and being very close and connected show much different trends. In the far SOL, oscillations between positive and negative values over one to two centimeters are characteristics of strong electric fields and convections. A typical point of the graph is the hump around the antenna limiter. This zone is subject to significantly higher potential gradients than others as it has been observed on most devices equipped with ICRH. Amplitude of oscillations also increases with ICRF power, indicating good correlation with RF sheath-induced mechanism.   As a consequence, potential differences result in electric fields and ExB density convections. Also known as convective cells, those may locally increase evanescent layers width and have deleterious effect on the coupling efficiency of surrounding current straps as can be seen in Fig. 8.
In order to investigate even more the role of ICRF power, series of experiments with power modulations (Fig. 9) not only enabled to evidence the role of the power on near field sheath looking at ICRF coupling resistances (bottom right graph), but also on farther regions looking at LH reflection coefficients [7] (bottom left graph). ICRF-induced sheath basically enhances density convections in the SOL making waves coupling more difficult. Those observations remain true both in L and H modes.  Fig. 9. Overview of the shot N°69949 on top (same signals as in figure 5). Bottom left bar chart shows reflection coefficients of the E-port 4.6GHz LH launcher at three different times, respectively without ICRH and both in non-connected and connected cases. Bottom right graph finally shows the coupling resistance of an I-port ICRF antenna current strap. All show deleterious impact of ICRH power. While no influence of LHWs on ICRF was noticed in L-mode, on the contrary in H-mode, LHWs were found to have some positive effects [8,4]. Density gradient being stiff, in H-mode, EPJ Web of Conferences 157, 03057 (2017) DOI: 10.1051/epjconf/201715703057 22 Topical Conference on Radio-Frequency Power in Plasmas pedestal tends to become sharper, density increases in the center more than at the edge and ICRF waves coupling becomes more complex. Fortunately, LHWs have been found to have a mitigating effect on ELMs [4] -changing type I into type III -diminishing their intensity and increasing their frequency. As can be seen in Fig. 10, this changes pedestal shape, it becomes softer, regulating edge densities, shortening ICRH waves evanescent length and improving the coupling. Such process could be very useful on ITER where it is planned to achieve high power H-mode plasmas.

Conclusions and Prospects
Over the EAST 2016 campaign, series of experiments in Lmode were carried out with the aim to characterize the mutual influence of ICRF and LH systems. In areas both magnetically connected and disconnected to active ICRF antennas, RF sheath was evidenced using a reciprocating probe allying both three tips Langmuir and emissive wire methods. No obvious influence was observed in non-connected regions, whereas in connected areas, SOL potential and temperature (cf formulas in Appendix) tend to increase with RF power. Electric fields appear at the edge and seem to create density convections, consistent with both global density decrease and ICRF deleterious influence on LHWs coupling. Reciprocally, whereas no influence of LHWs on SOL parameters was observed in L-mode, they can mitigate ELMs in H-mode, which influences the plasma edge by softening the pedestal and improving ICRF coupling.