The ITER Experimental Reactor will be equipped with a 24 MW Electron Cyclotron Resonance Heating (ECRH) system operating at 170 GHz. The unique ECRH properties of very localised heating and current drive in the plasma are well suited to satisfying various physics requirements such as start-up assist, bulk-heating, current drive and control of magnetohydrodynamic (MHD) instabilities. A versatile and flexible multi-purpose system is under design to meet the ITER requirements.

The suppression of plasma instabilities, in particular the socalled neoclassical tearing modes (NTMs) which are likely to appear in reference operating scenarios of ITER, is a particularly demanding objective. Instabilities are localised at major rational magnetic surfaces corresponding to distinct radial positions within the plasma cross section. These positions can move during the development of a plasma discharge as the profile of the safety factor, q, changes. The instabilities degrade the plasma performance, therefore mitigation and/or suppression is required. Efficient control was demonstrated in various tokamak experiments by driving a localised current by EC waves in the narrow region where the instabilities are excited. In operation, the location of the instabilities must be identified by proper diagnostics and the narrow ECRH beams must be actively steered towards their exact location during the pulse. The ECRH launcher must therefore be capable of steering the narrow RF beams over a significant range of angles.

Present-day experiments are performed with a so-called Front Steering Launcher (FSL), which has a movable mirror at the end of the transmission line inside the torus. An alternative concept, the Remote Steering Launcher (RSL), is presently under investigation for the ITER upper ports as seen from Fig.1. The main technical advantage of this approach is the avoidance of movable parts and the required driving mechanism close to the plasma, which simplifies the construction, simplifies maintenance and is expected to improve the reliability.

The launcher basically consists of a square corrugated wave-guide with a steerable mirror at the entrance of the wave guide (rather than on the plasma-facing end), which is then several meters away from the plasma and outside of the primary vacuum window. The RF-beam is launched at the waveguide entrance at variable angles and, with the dimensions of the waveguide properly chosen, is then radiated from the waveguide mouth under the same but opposite angle. A sketch of this principle is seen in Fig.1. Fig 2 shows the calculated microwave beam propagation in an optimised waveguide: the coloured areas indicate the reflection-patterns of the wave field at the waveguide walls. It is clearly seen that the propagation in the waveguide occurs in a zig-zag manner. A full-scale prototype of the launcher is close to completion and will undergo various low- and high-power tests. As an intermediate step, a simplified uncooled RSL mock-up was designed at the Institut für Plasmaforschung (IPF), Universität Stuttgart and tested at high power in IPP Greifswald. These measurements confirmed the principles of the RSL-concept and demonstrated the expected transmission performance. The partners to the design activity in 2004 were: ENEA/CNR, CRPP, FOM, FZK, IPP, IPP/IPF and UKAEA.

The 140 GHz, 10 MW, CW, ECRH system for the Wendelstein 7-X Stellarator, which is presently being installed and partially operational, provides an ideal test bed for the RSL tests, although the frequency is lower than the ITER frequency (140 vs. 170 GHz). The test arrangement was readily implemented in the IPP installation, because the Wendelstein 7-X transmission line is an open, purely optical system, which runs at atmospheric pressure. The beam from the first 1 MW, CW gyrotron, which became operational at the end of 2003, was directed into the squared corrugated waveguide by a set of mirrors. The waveguide is mounted on a stable frame construction with steerable optics at the entrance and various diagnostics along and at the exit of the waveguide. The high-power tests were performed with a power of up to 700 kW (typically 500 kW) and pulse lengths of up to 10 seconds. The RF beams at the exit of the waveguide were steered into and absorbed by a special dummy load developed and provided by CNRMilano. Due to the finite losses of the waveguide, the uncooled waveguide walls are heated. The losses, and thus the wall heating, is particularly pronounced in the areas where the waves are reflected from the walls along their ‘bounce-motion’. The heat distribution along the waveguide can be measured from outside with an infrared (IR)-camera, as seen in Fig. 3. The hot-spots (the 2nd, 4th and 6th reflection are seen) agree well with the calculated reflection areas of the propagating wave shown in Fig. 2. This technique also allows the experimental verification of calculation of detailed wave patterns. Although the experiments were performed under ambient atmospheric conditions, this waveguide arrangement (reproducing the reference design for the waveguide system in the RS Upper Launcher for ITER) showed no breakdowns and performed well. The detailed measurements on the high-power transmission characteristics supported the low-power measurements. The far field radiation patterns of the RF-beams were measured about 2.2 meters away from the waveguide mouth with a heat-target and an IR camera in short pulses. Beam pattern measurements at different launch angles showed that a high quality Gaussian beam was obtained within the steering range of +/- 12 deg. For larger steering angles, however, beam splitting is expected from theory and was observed in the experiments. These results increase the confidence on the applicability of the RS principle to the design of the ITER EC Upper Launcher.