Chauffage de plasma par ondes électromagnétiques à la troisième harmonique de la fréquence cyclotron des électrons dans le tokamak TCV
The Tokamak à Configuration Variable (TCV) programme is based on flexible plasma shaping capabilities together with a powerful electron cyclotron wave (ECW) additional heating for studies of stability, confinement, transport, control and power exhaust. In particular, ECW heating system allows an extended study of the β limit, defined as the ratio between the plasma kinetic pressure and the magnetic pressure, which is attributed to MHD (MagnetoHydroDynamic) instabilities. The ECW heating (ECH) is based on the resonant interaction between the electrons and an electromagnetic (EM) wave in a region of the plasma where the wave frequency is an harmonic of the electron cyclotron frequency. The TCV ECH system is composed of 6 gyrotrons (high power radio frequency sources), at the frequency of 82.7 GHz for second harmonic X-mode heating (X2) and 3 gyrotrons at the frequency of 118 GHz for third harmonic X-mode heating (X3), providing each a nominal power of 0.5 MW. In the moderate magnetic field of TCV (1.45 T), the X2 system is able to heat plasmas up to the X2 cutoff density (4.2 · 1019 m-3) above which the wave cannot propagate. The X3 system extends the accessible density range for ECH up to the X3 cutoff density (11.2 · 1019 m-3) and allows in particular the heating of plasmas in high confinement regime (H mode), most appropriate candidates to reach the β limit. The X2 wave is totally absorbed (plasma optically thick) being launched from the lateral side of TCV and crossing the vertical resonance layer. With this launching configuration, the X2 wave can also be used for non-inductive generation of the plasma current. Since the X3 absorption coefficient is weaker than the X2 absorption coefficient, the X3 wave is injected vertically in order to increase the beam path within the resonance layer, therefore maximizing the X3 optical depth. The present work, based on experiments and simulation, is the first detailed study of the X3 absorption properties in a top-launch configuration. The X3 absorption is shown to mainly depend on the wave injection conditions and the electron temperature. Full single-pass absorption is measured increasing nearly threefold the central electron temperature (1 keV –> 2.7 keV) when 1.35 MW of RF power is injected in low confinement regime plasmas (L-mode) with a central density of 4.0·1019 m-3. Experimental evidences show that a fraction of the power is absorbed on suprathermal electrons generated by the X3 wave itself. An absorption level of 85% is measured increasing threefold the central temperature by injecting 1.35 MW of X3 in H-mode plasmas with a central density of 8.2 · 1019 m-3. A new plasma dynamics is observed for the first time on TCV in these experiments. The X3 absorption is shown to depend strongly on the wave injection angle. In order to maximize the absorption during a plasma discharge by optimizing the injection angle, a real time feedback control has been developed and used. The system is based on the synchronous demodulation technique and uses a PI controller. In order to simulate the X3 wave propagation and absorption, the linear ray-tracing code TORAY-GA is used. These simulations predict an absorption dependence on the temperature and the injection conditions in agreement with the experimental results. Since TORAY-GA does not take into account the diffraction effects on the beam propagation, a comparison with the beam tracing code ECWGB which includes diffraction is discussed. The present results on X3 absorption properties demonstrate the efficiency of the X3 heating system on TCV, therefore extending the β limits study capabilities in elongated plasmas.