Joint experiment with CNR-IFP Milan
Additional heating of FTU tokamak with powerful e.m. waves at the Electron Cyclotron frequency, in resonance with the cyclotron motion of the electrons in the toroidal field (Fig. 1) , is performed with 4 gyrotrons oscillating at the frequency of 140 Ghz and with a power output capability of 500 kW for 0.5 s each.
Fig.1 Lay-out of the ECRH system on FTU machine
The transmission line couples the radiating antenna to the gyrotrons in the main mode with ~10% line losses (~27% overall, gyrotron-to-plasma), including mode conversion and resistive dissipation, while maintaining full control of the wave polarization, essential for optimum plasma-wave coupling.
The plasma is illuminated with 4 beams, steerable both in the poloidal and toroidal directions for flexible off-axis heating and for current drive experiments. The size of each beam is less than 10% of the plasma minor radius for good localization of the heated volume.
1. System description
The ECRH system includes the millimeterwave equipment, an High Voltage Regulated Power Supply and a complex of Auxiliary Systems.
The RF sub-system is composed by 4 units, each one formed by:
1) a gyrotron, its support and its superconducting magnet;
2) a quasi-optical set at the waveguide input, enclosed in its own Box;
3) a metallic waveguide;
4) a quasi-optical set of mirrors at the waveguide output, including a DC break and the tokamak barrier window;
5) a launching system;
6) a calorimetric matched load.
2. The gyrotron
The total e.m. power for the experiment (2 MW) is generated by 4 conventional cylindrical-cavity gyrotrons, TE22,6 being the cavity mode. An internal quasi-optical converter couples efficiently the cavity mode to a gaussian beam, radiated outside the tube at 90o from the electron beam axis, horizontally, through a BN single-disk, edge-cooled vacuum window.
Each gyrotron is fitted with its own Matching Optics Unit (MOU; not shown in the figure) for filtering out spurious sidelobes and for shaping the main beam to the size required for optimum matching to the coupled millimeterwave system. The power in the beam out of the MOU is ~430 kW.
The cavity magnetic field (5.6 T) is provided by a superconducting magnet, equipped for low consumption of the cryogenic liquids and minimum maintenance.
3. The gyrotron-to-waveguide matching optics
The matching of the waveguide to the gyrotron is accomplished with a set quasi-optical elements. An ellipsoidal Aluminum mirror converges the beam into a waist of 28.6 mm, which is ideally coupled (better than 98%) to the HE1,1 mode in the metallic waveguide. Two flat corrugated mirrors are placed in the beam path form the mirror to the waveguide input, to change the polarization. One mirror changes the ellipticity, and the second one the axis of the polarization ellipse. The series of the two mirrors can ideally change any incident polarization into any reflected polarization. The real mirrors are designed to keep the cross-polarization at a minimum when the incidence angle is 20o.
The group of quasi-optical elements at the waveguide input is completed by a movable mirror, set out of the beam path during normal operation, and inserted when the beam has to be diverted into a local calorimetric matched load.
All these elements are enclosed into a sealed Box for radiation leakage prevention. A shutter automatically isolates the waveguide when RF is not injected in it (the beam is switched into the local matched load).
4. The waveguide
The long runs (~30 m) from the gyrotron room to the tokamak hall are covered by a metallic, circular waveguide with an inner diameter of 88.9 mm . The inner wall is corrugated to provide the anysotropic wall impedance necessary to sustain the propagation of the hybrid, low loss HE1,1 mode. The full length is divided into ~20 straight pieces, mostly 2 m long. All pieces stand on adjustable mountings and are aligned better than 0.5 mm/m. The waveguide joints allow fine adjustments during installation, while providing full sealing against radiation leakage during operation. Thermal expansions are compensated by sliding joints inserted in each bend-to-bend section.
Each straight section can be removed from the line without loss of alignment, since all the adjustable parts of each stand remain fixed to the supporting beams, which are pre-aligned to horizontal before the installation of the waveguide. The line is run in air at atmospheric pressure, with a capability for pressurized operation.
All bends are at 90o, essentially quasi-optical, the two arms being coupled trough a plain mirror. Alignment, mechanical strength, mirror mounting and support for radiation shielding is accomplished by the bend-body, cast in Aluminum and machined to the required precision. An array of holes (in cut-off) in the plane mirrors directively couples to the outside a small fraction (-60 dB) of the propagating e.m. field, allowing a measurement of the transmitted power (Fig.2).
Fig.2 The waveguide system: the bend, with coupling-hole array; waveguide; joint on its adjustable mounting; supporting aluminum beam; and hanger (blue) anchoring the beam to the building.
5. The waveguide-to-launcher matching optics
The section of the system linking the waveguide to the launcher, firmly attached to the vacuum vessel, must deal with thermal contractions (up to ~5 mm) during operation at liquid nitrogen (LN2) temperature, while maintaining strict alignment of the optical axis of the launchers to the correspondent waveguide axis. This mechanical and electromagnetic matching is provided by a mirror, set on a mounting forced to slide only along the incident and reflected beam axis respectively. The focal length of the mirror matches the HE1,1 waveguide mode.
The barrier window is made of a single BN disk, has a diameter of 123 mm and is 11 lambda/2 (5.405 ±0.01 mm) thick. Edge cooling is provided for shot-to-shot heat removal, while within each 0.5s pulse the window heats almost adiabatically up to ~600o C at center.
6. The launching system
The launching system , entirely inside vacuum but outside the magnet, is composed by a group of 4 sets of 3 mirrors, one set for each line.
The first and the last Aluminum mirrors of each series are focussing, providing a converging beam with a radius of ~10 mm at the center of the vessel (in vacuum) . The intermediate mirror, made of copper, is flat and illuminates the final mirror through a vacuum gate valve separating the main vacuum from the appendix containing the first two mirrors and the barrier window. The final mirrors are all independently movable in the poloidal and toroidal directions. Each of them rotates toroidally around the incident beam axis.
All movements are transferred to linear displacement by a stranded metallic wire and are coupled to outside the vacuum through a bellow. Actuators and encoders are in air.
Toroidally oblique launch is obtained by reflection at plane mirrors parallel to the port walls. The shape of the launched beams has been monitored at low power before installation both by taking I.R. pictures of the thermal print left by the beam on absorbing foils, and by accurate 2D pattern measurements . The polarization transfer function was also characterized before installation.
The alignment of the mirrors in the launcher is controlled after installation, under vacuum and with the machine at LN2 temperature, with the aid of a retractable probe using a thermopile as the RF detector.
7. The matched calorimetric loads
Matched loads are used for tube conditioning and power measurements . An integrating sphere is installed in each gyrotron box, having the inner wall sprayed with a mixture of Al and Ti oxides as microwave absorbers. The average wall reflectivity is ~40%. The beam is spread after entrance by a diverging mirror opposite to the input port. The overall reflectivity is lower than ~4%.
A cylindrical version of the matched load is used to terminate the transmission lines for system conditioning and transmitted power measurements. Inside the cylinder an array of irises and tubes, directly cooled by circulating water, and sprayed with Al2O3 and TiO2 mixtures, trap and absorb the RF power after entrance.
Calorimetric measurements of the absorbed RF power are performed on the water cooling circuit in both loads.
8. Arc detection and leakage r.f. monitoring systems
The whole paths of the high power e.m. waves (in the open air between mirrors and inside the waveguides) is monitored against arcing by an array of fiber optics and fast detectors.
A special technique has been developed for monitoring any r.f. leakage from waveguide joints and possible apertures in the enclosures .
9. The Regulated High Voltage Power Supply
The High Voltage DC supply required for gyrotron nominal operation is 70 kV, 25 A, with a voltage stability ± 0.3%. The power supply is composed by a 100 kV, 100 A, stabilized and filtered dodecaphase rectifying unit, and two series tetrode regulators (in parallel) for fine voltage stabilization. Each tetrode feeds two gyrotrons.
Fast (10 ms) removal of the High Voltage is achieved with a Crow-Bar unit, composed by a stack of 5 ignitrons with 40 kV capability each. The rise time of the output voltage (10-90%) is 50 ms, and the settling time to specified stability is less than 100 ms. The output voltage can be modulated (DV=0-70 kV) under controlled feedback up to 10 kHz, sinusoidal.
10. The Auxiliary systems
The gyrotron filament heater needs a stable and well regulated AC power supply at 50 V, 20 A maximum. The heater is fed through a 4:1 isolation transformer by a stabilized 20-220 V, 50 Hz, power supply with 1.4 kW power capability.
Minor supplies (0-30 V; 0-15 A) are employed for the gun coil and for the de-focussing coils at the collector.