Electron Cyclotron Emission and Electron Cyclotron Resonance Heating (EC-15)

The following sections are included:

• PREFACE

• CONTENTS

In this paper we trace the development of measurements of ECE on fusion machines from the early days, when the main interest in this emission was its contribution to the power loss, to the present, where such measurements are employed as a powerful diagnostic of the plasmas produced in most contemporary tokamak and stellerator machines. The development path has not always been smooth but it has always been rich in creativity, surprises, debate and ultimately success. The significant developments in the field are identified and illustrated with examples.

In 1989 it was decided that ASDEX Upgrade should have an ECRH system. The plans for a 140 GHz/2 MW/2 s system were finally approved in 1993. The system comprises 4 gyrotrons with 4 separate transmission lines and launchers. Although a 0.5 s test gyrotron was already installed in autumn 1994, it was only in summer 1997 that the first gyrotron of the final system was ready for use in the experimants, and in spring 2000 the system was completed with all 4 gyrotrons, Since then failure of gyrotrons and a long time needed for their replacement reduced temporarily the available power. This paper reviews the planning, construction and operation of this system, and some unforseen problems encountered.

Astrophysical applications of electron cyclotron maser emission (ECME) are reviewed with emphasis placed on planetary radio emissions. The early history of the field, and the relevant theory of ECME are summarized. Some emphasis is given to the problem of escape of ECME from an astrophysical source.

It was only a few years ago that there was serious discussion about whether or not to include an electron cyclotron heating (ECH) system on ITER. It now appears that the ECH system may be the principal auxiliary heating system when ITER starts, a development which transpired due to recent availability of high power long pulse gyrotrons at frequencies above 100 GHz and the successful use of these tubes in a number of innovative ITER relevant experiments on present devices. This is a brief summary of the ECH experimental results reported at EC-15.

The requirements for ECRH and ECCD in next generation fusion devices (reactor-grade tokamaks, satellite tokamaks, stellarators and spherical tori) are discussed. Based on a review of recent progress in the field, areas of future research in ECRH&CD technology and physics are discussed. In technology, these include development and application of steady state frequency tunable gyrotrons and fast switches for EC waves while in physics, the application to MHD control should be enlarged, synergy with LHCD exploited for current profile control and schemes for access of overdense plasmas have to be developed further. An important element will be feedback controlled deposition of the EC waves.

Second harmonic X-mode (X2) electron cyclotron (EC) heating has been used in DIII-D to examine plasma initiation and burnthrough of low Z impurities. Although the toroidal inductive electric field (E_ɸ) in DIII-D is high enough (0.9-1.0 V/m) to allow robust startup without EC assist, startup in fusion devices such as ITER will have lower fields (E_ɸ = 0.3 V/m) and EC assist can provide an increased margin for burnthrough of low Z impurities. ECH, applied before the inductive electric field, is used to separate the various phases of plasma breakdown and startup. Ionization first occurs near the X2 resonance location and then expands in the vessel volume. Perpendicular launch (k_∥ = 0) is required for X2 pre-breakdown, and the power threshold can be reduced by optimizing prefill and vertical field, although the lowest power threshold is not at the optimum value for Ohmic startup. An orbit following code confirms that cold electrons (0.03 eV) can be sufficiently heated by ECH to energies above the threshold of ionization of hydrogen. This code predicts successful ionization in future tokamaks such as KSTAR and ITER. The ITER startup scenario has been simulated in DIII-D experiments and X2 ECH assist has been applied at reduced toroidal loop voltage to assist burnthrough and plasma current ramp up.

EBW current drive assisted plasma current start-up has been demonstrated for the first time in a tokamak. It was shown that plasma currents up to 17 kA can be generated non-inductively by 100 kW of RF power injected. With optimized vertical field ramps, plasma currents up to 33 kA have been achieved without the use of solenoid flux. With limited solenoid assist (0.2 V × 20 ms, less than 0.5% of total solenoid flux), plasma currents up to 55 kA have been generated and sustained further non-inductively. Experimentally obtained plasma currents are consistent with Fokker-Planck modelling.

Closed loop control of the period of Fast Ion Stabilised Sawtooth has been demonstrated for the first time on Tore Supra by varying the ECCD injection angles in real time. Fast ions generated by 4 MW of central ICRH increased the Sawtooth period to 100 ms. This sawtooth period was reduced to the ohmic period of 30 ms by the addition of only 300 kW of ECCD near the q = 1 surface. A feedback loop allowed the sawtooth period to be switched in real time between short and long sawteeth with a response time of ~1 s.

An extensive database of Tore Supra discharges with Internal Transport Barriers (ITBs) has been analysed. A tight correlation has been found, which links the central value of q and the creation of an ITB, while no correspondence with magnetic shear or q_min values can be inferred. In the case of incomplete transition to ITB (O-regime), modelling in presence of ECCD confirms the experimental observations about triggering/stopping and amplifying the oscillations.

Resistive neoclassical tearing modes (NTMs) are anticipated to be the principal limit on stability and performance in ITER as the resulting islands break up the magnetic surfaces confining the plasma. Drag from island-induced eddy currents in the resistive wall can slow plasma rotation, produce locking to the wall, and cause loss of the high-confinement H-mode and disruption. NTMs are destabilized by helical perturbations to the pressure-gradient-driven "bootstrap" current. NTMs can be stabilized by applying co-electron-cyclotron current drive (ECCD) at the island rational surface. Such stabilization and/or preemption is successful in ASDEX Upgrade, DIII-D, and JT-60U, if the peak off-axis current density is comparable to the local bootstrap current density and well-aligned.

ASDEX Upgrade has used a feed-forward sweep of the toroidal field to get ECCD alignment on the island. JT-60U has used feed-forward sweeps of the launching mirror for the same purpose, followed up by real-time adjustment of the mirror using the electron cyclotron emission (ECE) diagnostic to locate the island rational surface. In DIII-D, ECCD alignment techniques include applying "search and suppress" real-time control to find and lock onto optimum alignment (adjusting the field or shifting the plasma major radius in equivalent small steps).

Most experimental work to date uses narrow, cw ECCD; the relatively wide ECCD in ITER may be less effective if it is also cw: the stabilization effect of replacing the "missing" bootstrap current on the island O-point could be nearly cancelled by the destabilization effect on the island X-point if the ECCD is very broad. Modulating the ECCD so that it is absorbed only on the m/n = 3/2 rotating island O-point is proving successful in recovering ECCD effectiveness in ASDEX Upgrade when the ECCD is configured for wider deposition.

The ECCD in ITER is relatively broad, with current deposition full width half maximum almost twice the marginal island width. This places strict requirements on ECCD alignment, with the cw effectiveness dropping to zero for misalignments as small as 2 cm. Tolerances for misalignment are presented to establish criteria for the alignment by moving mirrors in ITER for both cw and modulated ECCD.

Utilization of electron Bernstein wave (EBW) for tokamak startup and advanced heating in helical device is demonstrated. In the LATE device, plasma current is started up and ramped up to 20 kA by injecting a 5 GHz, 190 kW, 70 ms microwave pulse in the electron cyclotron range of frequency and controlling the applied vertical field. The electron density exceeds the plasma cutoff density. The plasma current is carried by a unidirectional tail electrons against the reverse electric field from self-induction, suggesting the acceleration along the magnetic field via cyclotron absorption of mode-converted electron Bernstein wave (EBW) with high refractive indices along the magnetic field. In LHD, the extraordinary mode is injected to the upper hybrid resonance layer from the high field side and EBW heating is observed at much inner radius away from the fundamental electron cyclotron resonance layer.

Combining the potential of ECRH, ECE-Imaging and the generation of resonant magnetic perturbations (RMP) by the dynamic ergodic divertor provide new insights into the physics of tearing modes. ECE-I measurements reveal the two dimensional propagation of heat pulses around magnetic islands. Inside an island with a flat temperature profile, the electron heat transport is shown to be strongly reduced with respect to the surrounding plasma. ECRH or co-ECCD localized at the resonant surface is found to lead to a higher threshold for the penetration of RMP. Analysis of the generalized Rutherford equation shows that under TEXTOR conditions the heating effect dominates over the effect of current drive in agreement with the experimental observations. On ITER, both are expected to be of the same order of magnitude.

To continue with the characterization of the ECCD in the TJ-II stellarator new experiments have been carried out. The TJ-II ECRH/ECCD system consists of two 53.2 GHz gyrotrons, which deliver a maximum total power of 600 kW transmitted to the vacuum vessel by two quasi-optical lines, which have a steerable mirror placed inside the vessel. The dependence of the plasma current on the injected power has been investigated and the current drive efficiency has been evaluated for different induced currents. The results are being compared with the experiments carried out in other helical devices: Heliotron-J, CHS and LHD. These comparisons gives us a common understanding of the current drive physics in heliac machines.

A 0.25 MW system designed to heat electrons and drive current via the electron Bernstein wave is in operation on the MST reversed field pinch. The antenna is a grill of four half-height S-band waveguides with each arm powered by a separate, phase controlled traveling wave tube amplifier at 3.6 GHz. Coupling to the plasma (as measured by ratio of reflected power) is very dependent on the relative phasing between adjacent waveguides. The total reflected power can be maintained near the 10% level. The antenna face is outfitted with a pair of triple Langmuir probes to measure local electron density; the density gradient at the upper hybrid resonance (typically within 1-2 cm of the antenna) is expected to strongly influence coupling efficiency. Conditioning of the antenna is currently underway (near the 0.2 MW level) and total system power is expected to reach 0.25 MW, or roughly a fourth of the Ohmic input power in target plasmas. The x-ray spectrum (5-200 keV) is monitored as a way to detect modification to the electron distribution as full transmitter power is approached. Particular emphasis is on recent experiments in which toroidally localized soft x-ray emission is measured with a 20 chord camera.

The use of ECRH as a possible technique to mitigate or avoid disruptions on the Frascati Tokamak Upgrade (FTU) has been further developed after the preliminary promising results reported in the last EC-14 [G. Granucci, et al., Proc. of 14th Joint Workshop on ECE and ECRH, Santorini, Greece (2006)]. The crucial role of MHD in the final phase of a disruption and the effectiveness of the EC power in stabilizing the growing modes, thus saving the discharge, has been shown. In the FTU experiments the reduction of one island width, by means of direct EC heating automatically triggered by the V_loop increase at the beginning of the disruption current quench, leads to the reduction of the other(s) coupled mode(s). The strong mode coupling, observed in the phase approaching the disruption, can be used to resolve the problem of the radial tracking of the mode to be suppressed, opening the road to a powerful strategy for ITER.

Electron cyclotron current drive (ECCD) experiments have been made in stellarator/heliotron (S/H) devices including Heliotron J, TJ-II, CHS and LHD. The experimental results show that ECCD can be controlled by the power injection angle, absorption position and magnetic field structure. The current drive efficiency is similar, γ = n_eI_ECR/P_EC = 8-16 × 10¹⁶ A/Wm², ζ = 32.7n_cI_ECR/P_WT_e = 0.03-0.05 when the magnetic field ripple ratio is 0.93 < B_min/B_max < 1.0. The reversal of driven current direction is observed depending on the magnetic field ripple structure, indicating that the amplitude and direction of EC current is determined by the balance between the Fisch-Boozer effect and the Ohkawa effect. The toroidal current can be controlled by ECCD; zero net current is demonstrated by cancelling the bootstrap current, and the EC driven current is balanced by co and ctr-ECCD.

The space movement of plasma oscillations in m/n = 1/1 mode is analyzed. ECE from antennae disposing at the different points of poloidal and toroidal direction are used. On-axis slightly modulated ECH is applied for control of amplitudes and frequency of oscillations. It was found that oscillations occupy total area inside q = 3/2 radius. Maximal amplitudes correspond to ECH absorption zone. Simultaneous analysis of ECE in two polarization and also SXR detects the strong periodical "compressing and widening" of electron distribution function main body in the longitudinal direction. Such phenomenon creates plasma current pulsing at central. area and generation of high electric field splashes. The peculiarities of toroidal and poloidal movement of disturbance show that that is not pure rotating and can be explained as its "swinging" along magnetic force lines direction under continual drift plasma rotation.

ECRH assisted plasma startup at fundamental resonance is investigated in Tore Supra in view of ITER operation. ECRH pre-ionisation is found to be very efficient allowing plasma initiation in a wide range of prefill pressure compared to ohmic startup. Reliable assited startup has been achieved at the ITER reference toroidal electric field (0.3 V/m) with 160 kW of ECRH. Resonance location scan indicates that the plasma is initiated at the resonance location and that the plasma current channel position had to be real-time controlled since the very beginning of the discharge to obtain robust plasma startup.

This paper reports about electron Bernstein wave (EBW) heating and current drive experiments at 2.45 GHz and experimental set-up for 28 GHz at the WEGA Stellarator. The device allows continuous plasma operation for a magnetic field of up to 0.5 T. For the 2.45 GHz the OXB mode conversion could be investigated in detail by HF-probes. The results were consistent with full wave simulations. The propagation of EBWs was calculated by a 3D- ray-tracing code for both 2.45 GHz and 28 GHz. The results were confirmed by the measurement of the local resonant power deposition with the power modulation technique. The predicted EBW driven toroidal current for the 2.45 GHz could be clearly detected experimentally by total and local measurement. For the 28 GHz a waveguide and antenna system, which is optimized for OXB-conversion, was build-up.

Electron Bernstein Wave (EBW) heating is important for high-beta plasma experiments and will be used for heating over-dense plasmas on TJ-II. TJ-II is a medium sized Heliac operating at CIEMAT in Madrid, whose plasmas are created and heated by ECH via two 300 kW gyrotrons at second harmonic X-mode (53.2 GHz), with additional heating provided by two neutral beam injectors. Theoretical work has shown that the most suitable scheme for launching EBWs in TJ-II is O-X-B mode conversion, which has acceptable heating efficiency for central densities above 1.2 × 10¹⁹ m^-3[1]. A system based on a 28 GHz-100ms diode gyrotron will be used to deliver 300 kW through a corrugated waveguide. The microwave heating beam will be directed and focused by a steering mirror located inside the vacuum vessel. Prior to the heating experiments, measurement of the thermal EBW emission (EBE) from the plasma is being made to help determine the optimum launch angle for EBW mode conversion, and also to provide an indication of the electron temperature evolution in over-dense plasmas. A dual-polarized quad-ridged broadband horn is used to measure the EBW emission and polarization at 28 GHz. Initial measurements indicate that the emission in under-dense plasmas corresponds to oblique electron cyclotron emission (ECE) and then converts to EBE when the plasma becomes over-dense during neutral beam injection.

A line-of-sight ECE diagnostic has been taken into operation on the TEXTOR tokamak, which views the plasma along the optical path of the high power ECRH beam. Separation of the low power ECE and high power gyrotron radiation is achieved by a Fabry-Pérot filter in the form of a 25.75 mm quartz plate which is transparent to the gyrotron radiation but reflects selected ECE frequencies with a periodicity of 3 GHz. Tests at low and high power show that the system is behaving in accordance with expectations. ECE spectra have been obtained both in the absence and during high power ECRH. These clearly show the potential of identifying the location of magnetic islands or the sawtooth inversion radius. For some launching conditions the measured ECE spectrum is strongly perturbed. The reason for these perturbations is under investigation.

Measuring ECE in ITER plasmas will present several difficulties in both hardware and physics that will impede efforts to determine T_c(r,t) and characteristics of the electron distribution function. Recent work on the diagnostic systems and techniques to be employed on ITER has clarified some of these issues and found some solutions. Studies show that even with the greater relativistic broadening due to the high electron temperature, T_e measurements with spatial resolution of 6-10 cm are still possible especially with properly designed front-end optics and instruments. ECE will still be able to provide high resolution T_c profiles in both core and edge regions and even follow oscillations of high m, n TAE modes. Of greater concern is the possibility and effects of non-Maxwellian electron distributions created by intense auxiliary heating. In these cases modeling will have to be used to correct and define the valid limits of T_e from ECE and a special oblique viewing antenna is recommended. The planned ECE instruments, heterodyne radiometer and Michelson interferometer, will provide complementary measurement capabilities. A viable in-port-plug calibration source for these instruments is a concern; progress on work being done to develop a prototype is reported.

Profiles (0.3 < ρ < 0.9) of electron temperature and density fluctuations in a tokamak have been measured simultaneously and the results compared to nonlinear gyrokinetic simulations. Electron temperature and density fluctuations measured in neutral beam-heated, sawtooth-free L-mode plasmas in DIII-D are found to be similar in frequency and normalized amplitude, with amplitude increasing with radius. At ρ = 0.5, nonlinear gyrokinetic simulation results match experimental heat diffusivities and density fluctuation amplitude but overestimate electron temperature fluctuation amplitude and particle diffusivity. In contrast, the simulations at ρ = 0.75 do not match either the experimentally derived transport properties or the measured fluctuation levels.

Bernstein-extraordinary-ordinary mode-converted electron Bernstein wave (EBW) emission was angularly scanned at the Mega Ampere Spherical Tokamak (MAST). Radiometry along a rapidly steering line of sight sampled the conversion efficiency contours for various view angles. Here we describe the diagnostic, present the first results and illustrate their potential as an indirect measure of the q profile at the plasma edge.

Electron cyclotron emission spectrum is a rich source of information about thermonuclear plasmas. Perpendicular low field side detection of X mode second harmonic is commonly used to measure electron temperature in toroidal devices. Measurements at different angles, and over a wider range of harmonics, sample different regions of the electron momentum space and can, in principle, give information on it. This is extremely interesting in view of burning plasma experiments, where energy transfer from non-thermal population to thermal particles is expected to be significant.

The long standing disagreement between the central electron temperature measured by Thomson scattering and electron cyclotron emission (2nd harmonic, X-mode) in high performance plasmas observed in JET and TFTR is reexamined. In this paper we report on recent analysis carried out in JET with the aim of investigating the role of different plasma parameters, such as plasma heating scheme and fuelling, on the appearance of this discrepancy. Our analysis shows that the appearance of the discrepancy between the temperature diagnostics is correlated with the appearance of a high energy tail in the ion distribution function.

In this work we provide a full review of the absolute calibration of ECE quasi-optical diagnostics, such as Fourier transform spectrometers. The optimisation of the overall process of averaging the interferograms to extract a reliable calibration curve are discussed. The basic assumptions, for example the linearity of the detector and the achievability of a given signal to noise ratio, are pointed out. Finally the actual physical meaning and the "accuracy versus precision" characteristics of the calibration curve are analysed.

In JET the ECE radiometer is designed to provide combined O- and X-mode operation which has enabled the simultaneous measurements of the pedestal electron temperature profile in both the inboard and outboard region. These measurements are particularly relevant in the study of edge localized modes (ELMs). In general, the characteristics of the pedestal temperature profile are similar on both the inboard and outboard sides, but some apparent asymmetries are observed in the case of big type I ELMs. Some examples of this new set of measurements are presented here.

Temperature fluctuation diagnostics are being used to detect and study the fast particle instabilities, which could destabilize Magnetohydrodynamic (MHD) modes in tokamak plasmas. As the thermal noise is higher than the temperature fluctuation amplitude, correlation of two adjacent Electron Cyclotron Emission (ECE) channels is needed to recover the electron temperature fluctuations using long time integration. On Tore-Supra, we are developing a multi-channel ECE correlation diagnostic to observe MHD modes at 4 plasma positions with a radial resolution of about 2 centimeters and the poloidal resolution of about 4 cm (1/e folding width). A 1-channel prototype using 2 bandpass YIG filters has experimentally identified MHD instabilities related to fast particles dynamics. This first channel is currently being upgraded to reduce crosstalk and a second channel will be operational in 2008. The probing frequencies can be changed during the discharges, with 1s dead time between two modifications. Programmable attenuators are used to optimize the signal dynamic according to the frequency sensitivity. At nominal condition (B = 3.8 T), the diagnostic accessibility covers the outer plasma part up to a normalized radius r/a = 0.2 on the inner side. Two more channels will be implemented in 2009 and will allow us to measure the radial profile of MHD modes, or to study their inner/outer side asymmetry.

High β plasmas in the National Spherical Torus Experiment (NSTX) operate in the overdense regime, allowing the electron Bernstein wave (EBW) to propagate and be strongly absorbed/emitted at the electron cyclotron resonances. As such, EBWs may provide local electron heating and current drive. For these applications, efficient coupling between the EBWs and electromagnetic waves outside the plasma is needed. Thermal EBW emission (EBE) measurements, via oblique B-X-O double mode conversion, have been used to determine the EBW transmission efficiency for a wide range of plasma conditions on NSTX. Initial EBE measurements in H-mode plasmas exhibited strong emission before the L-H transition, but the emission rapidly decayed after the transition. EBE simulations show that collisional damping of the EBW prior to the mode conversion (MC) layer can significantly reduce the measured EBE for T_e < 20 eV, explaining the observations. Lithium evaporation was used to reduce EBE collisional damping near the MC layer. As a result, the measured B-X-O transmission efficiency increased from <10% (no Li) to 60% (with Li), consistent with EBE simulations.

Electron Bernstein Wave (EBW) heating is important for high-beta plasma experiments and will be used for heating over-dense plasmas on TJ-II. TJ-II is a medium sized Heliac operating at CIEMAT in Madrid, whose plasmas are created and heated by ECH via two 300 kW gyrotrons at second harmonic X-mode (53.2 GHz), with additional heating provided by two neutral beam injectors. Theoretical work has shown that the most suitable scheme for launching EBWs in TJ-II is O-X-B mode conversion, which has acceptable heating efficiency for central densities above 1.2 × 10¹⁹ m^-3 [1]. A system based on a 28 GHz-100 ms diode gyrotron will be used to deliver 300 kW through a corrugated waveguide. The microwave-heating beam will be directed and focused by a steering mirror located inside the vacuum vessel. Prior to the heating experiments, measurement of the thermal EBW emission (EBE) from the plasma is being made to help determine the optimum launch angle for EBW mode conversion, and also to provide an indication of the electron temperature evolution in over-dense plasmas. A dual-polarized quad-ridged broadband horn is used to measure the EBW emission and polarization at 28 GHz. Initial measurements indicate that the emission in under-dense plasmas corresponds to oblique electron cyclotron emission (ECE) and then converts to EBE when the plasma becomes over-dense during neutral beam injection.

The 60-channel electron cyclotron emission (ECE) radiometer diagnostic on the ASDEX Upgrade tokamak is presently being upgraded to include a 1 MHz sampling rate data acquisition system. This expanded capability allows electron temperature measurements up to 500 kHz (anti-aliasing filter cut-off) with spatial resolution ~1 cm, and will thus provide measurement of plasma phenomena on the MHD timescale, such as neoclassical tearing modes (NTMs). The upgraded and existing systems may be run in parallel for comparison, and some of the first plasma measurements using the two systems together are presented. A particular planned application of the upgraded radiometer is integration into a real-time NTM stabilization loop using targeted deposition of electron cyclotron resonance heating (ECRH). For this loop, it is necessary to determine the locations of the NTM and ECRH deposition using ECE measurements. As the magnetic island of the NTM repeatedly rotates through the ECE line of sight, electron temperature fluctuations at the NTM frequency are observed. The magnetic perturbation caused by the NTM is independently measured using Mirnov coils, and a correlation profile between these magnetic measurements and the ECE data is constructed. The phase difference between ECE oscillations on opposite sides of the island manifests as a zero-crossing of the correlation profile, which determines the NTM location in ECE channel space. To determine the location of ECRH power deposition, the power from a given gyrotron may be modulated at a particular frequency. Correlation analysis of this modulated signal and the ECE data identifies a particular ECE channel associated with the deposition of that gyrotron. Real time equilibrium reconstruction allows the ECE channels to be translated into flux surface and spatial coordinates for use in the feedback loop.

The diagnostic potential of ECE [A. Costley, Diagnostics for Fusion Reactor Conditions (Varenna 1982), EUR 8351-I, p. 129&149] has been studied in almost every aspect since the dawn of Fusion experiments. It represents currently an almost unique opportunity to provide an established diagnostic system for the next step experiments, but it still has a vast potential to explore. In this heuristic work we introduce a novel approach to physics interpretation, suggesting ECE energy transport between harmonics which in certain plasma regimes can also be relevant from the plasma point of view. In addition we discuss the implementation of an unconventional hybrid hardware solution to extend the capability of the systems by using mixed quasi-optical and electronic techniques.

The 128 channel 2-D electron cyclotron emission imaging system collects time-resolved 16 × 8 images of T_e profiles and fluctuations on the TEXTOR tokamak. This instrument was upgraded in February 2007 with new wideband electronics which increased the instantaneous frequency coverage by >50% to 6.4 GHz with a corresponding increase in horizontal plasma coverage. Frequency extenders have been developed to combine modules together to double the instantaneous coverage to 12.8 GHz. A new lens-antenna configuration to be implemented on TEXTOR, employing an array of miniature substrate lenses coupled with front-side local oscillator pumping, shows a significant increase in both RF bandwidth and RF sensitivity over the lens-antenna geometry currently installed on TEXTOR while reducing the level of RF power required to pump the array.

Common application of X- and O-mode allowed at many experiments to register the emission spectra of super thermal electrons, to evaluate them total, longitudinal and perpendicular energy. The location of the current carrying electrons on the simple rational magnetic surfaces is discovered. Periodic ECE pulsing of high energy electrons got possibility to evaluate the oscillating part of longitudinal electric field in ~1 V/cm. Strong plasma self-organizing is explained by the common action of electric field and potential plasma oscillations to electron component.

The characteristic of ECE spectra in a tokamak has been studied taking account of the torus geometry. Although the structure of the magnetic field (B) in the sight line is different in the cases of changing the angle between the sight line and B, the radiation temperature of ECE are almost same in the low frequency side of 2nd harmonic, electron temperature profile [T_e^th(r)] in low field side can be obtained by changing the angle between the B and sight line. The effect of supra-thermal electron on the radiation temperature is also evaluated in the case of bi-Maxwellian. In the uniform density case of supra-thermal electron (n_e^sp), small amount of supra-thermal affects the radiation temperature. For instance, when the n_e^sp is 0.4% of thermal electron density results in 10% of deviation of electron temperature in the case of T_e^th(0) = 20 keV.

A 2-D plasma visualization system is being developed for the KSTAR tokamak to image electron temperature and electron density fluctuations. Optical designs are being revised to accommodate a new lens-antenna configuration developed for plasma imaging use. This approach employs an array of miniaturized lenses, with a single imaging antenna placed at the center of each lens, resulting in both increased RF bandwidth and RF sensitivity. Progress is also being made in the development of quasi-optical notch filters that protect the imaging arrays from stray ECRH power. High performance 84 GHz and 170 GHz notch filters have been designed which exhibit large rejection (>35 dB) at the resonant (notch) frequency, and retain this high rejection over an incident angle range of ±10°.

A brief review of the theory contributions presented at the EC-15 Workshop is given. The covered spectrum of topics is quite broad including the electron cyclotron wave propagation physics as well the results of numerical modeling of transport and current drive.

A new approach treating an actual narrow microwave beam as a superposition of wide "virtual" beams unaffected by diffraction within a plasma volume is proposed. A solution thus found describes propagation of the microwave beam in the plasma volume taking into account the beam's diffraction, absorption and refraction in the framework of a single procedure.

An overview of the main Electron Bernstein Waves (EBW) theoretical results obtained in the TJ–II stellarator is presented. Firstly, former studies helping us to determine the best excitation scheme are discussed. Next, we describe the procedure used to find the optimum launching position and direction of the injected power, as well as the ray tracing optimization method that was developed to find the optimum launched beam. The 2D full-wave simulations that were performed in order to check the beam ray tracing optimization results are also presented. Final topics are devoted to the comparison between relativistic and non-relativistic ray tracing calculations and to the preliminary transport simulations performed with the ASTRA system. Plans for future work are briefly outlined. [Spanish MCyT project ENE2004-06957/FTN].

The adjoint approach for arbitrary collisionalities with momentum conservation is considered. The results are generally applicable for the parallel conductivity as well as for current drive calculations. In addition, the weakly relativistic extension of the variational principle for the Spitzer function with momentum conservation is described. The models developed are well suited to ray-tracing calculations.

A new general linear calculation of RF current drive has been implemented in the GENRAY all-frequencies RF ray tracing code. This is referred to as the ADJ-QL package, and is based on the Karney, et al. [1] relativistic Green function calculator, ADJ, generalized to non-circular plasmas in toroidal geometry, and coupled with full, bounce-averaged momentum-space RF quasilinear flux [2] expressions calculated at each point along the RF ray trajectories. This approach includes momentum conservation, polarization effects and the influence of trapped electrons. It is assumed that the electron distribution function remains close to a relativistic Maxwellian function. Within the bounds of these assumptions, small banana width, toroidal geometry and low collisionality, the calculation is applicable for all-frequencies RF electron current drive including electron cyclotron, lower hybrid, fast waves and electron Bernstein waves. GENRAY ADJ-QL calculations of the relativistic momentum-conserving current drive have been applied in several cases: benchmarking of electron cyclotron current drive in ITER against other code results; and electron Bernstein and high harmonic fast wave current drive in NSTX. The impacts of momentum conservation on the current drive are also shown for these cases.

ITER hybrid scenarios may require off-axis current drive in order to keep the safety factor above 1. In this type of applications, alignment of the current sources and self-consistency of current and temperature profiles are critical issues, which can only be addressed by integrated modelling. To this end, the CRONOS suite of codes has been applied to the simulation of these scenarios. Results of simulations of ITER hybrid scenarios assisted by ECCD, using the ITER equatorial launcher, for both co- and counter-ECCD, are presented.

A new scenario for ITER has been developed by means of the combined injection of three radio frequency power sources: Ion cyclotron resonant heating, lower hybrid current drive and Electron Cyclotron Heating and Current Drive (ECH/ECCD), which have been used to obtain a steady state plasma with 97% of noninductive current, Q ≈ 6.5 and a burning time of t ≈ 3000 s. In this scenario, the main role is played by ECCD, which is used to trigger and keep an Internal Transport Barrier (ITB) robustly fixed at normalized radius ρ ≈ 0.45. A threshold on the total power from ECH system has been found in order to provide the negative shear needed for sustaining the ITB.

In the electron cyclotron range of frequencies (ECRF), the extraordinary X waves or the ordinary O waves have been successfully used in many conventional tokamaks for generating plasma current and for modifying the current profile. ECRF waves are expected to play a similarly important role in ITER for current profile control. Also, in the EC frequency range, electron Bernstein waves have been used for heating and for current generation in stellarators, conventional tokamaks, and a levitated dipole. EBWs are also highly suitable candidates for current profile control in overdense plasmas encountered in spherical tokamaks (ST). It is well-known that relativistic effects need to be included in a proper description of the propagation and damping of EC waves. We have developed a code R2D2 which numerically solves the fully relativistic dispersion relation for all EC waves. The results from R2D2 provide an insight into the properties of EC waves and the effect of relativity on the damping of these waves. We present similarities between the damping of O waves and EBWs. The physics of the interaction of O waves with electrons in ITER-type plasmas bears similarity to the interaction of EBWs with electrons in STs. So EBW current drive in STs could provide useful insight into O wave physics in ITER.

A review is presented of a new analytical theory of linear o mode coupling in electron cyclotron frequency range in two-dimensionally inhomogeneous configurations of magnetized plasmas. The developed theory describes the conversion of ordinary and extraordinary waves the vicinity of the plasma cutoff surfaces with taking into account variations of a magnetic field on flux surfaces in a toroidal magnetic configuration, what is especially important for electron cyclotron heating and diagnostics of overdense plasmas.

Radial transport of the current carrying electrons can broaden the profile of electron cyclotron current drive (ECCD), with potentially detrimental effects for applications that reply on strong localization of the noninductive current. Early experiments on the DIII-D tokamak did not observe any clear effects of particle transport on the ECCD profile. However, more recent experiments at high ECCD power, low density, and radiation temperatures above 20 keV clearly demonstrate that the ECCD profile is reduced and broadened compared to CQL3D predictions assuming no radial transport. At high relative power densities, a diffusion coefficient of ≈0.4 m²/s is required in CQL3D to reproduce the experimental ECCD profile, while smaller diffusion coefficients are needed to model the ECCD profile at low relative power densities. This level of transport is an order of magnitude less than the electron thermal diffusivity but is comparable to the effective particle transport rate needed to maintain the density profile.

As it was recognised that local electron cyclotron (EC) wave power losses can be a competitive contribution to the 1D electron power balance for reactor-grade tokamak plasmas in regimes as anticipated for steady-state operation, a systematic effort is ongoing to improve the modelling capability for the radial profile of EC wave emission. This effort aims at generating a hierarchy of codes that cover the non-local behaviour of EC wave transport with good accuracy and also provide sufficient computational efficiency for being usable in 1D transport studies. An overview on this activity is given with emphasis on the code RAYTEC which explicitly addresses the geometrical effects present in toroidal plasmas with arbitrary cross-section.

In this report, fully relativistic full wave 3D STELEC code calculations (both damping and propagation from elliptical polarized antennae) of X- and O-modes and electron Bernstein waves (EBWs) in FT-2, NST, ARIES-like and ITER tokamaks are presented.

The SNECTR, CYTRAN, CYNEQ and EXACTEC codes are compared in view of the calculation of the profile of the net electron cyclotron (EC) wave power density emitted for different electron temperature profiles and average temperatures of relevance for reactor-grade magnetoplasmas. The effects of either specularly or diffusely reflecting walls are assessed for a cylindrical plasma with circular cross-section, specular reflection, as assumed in EXACTEC, providing a lower bound to the net EC wave power losses in the hot plasma core (and therefore, as a rule, also to the total EC power loss) as well as to re-absorption in the edge plasma. The assumption of isotropy of the radiation intensity in the plasma underlying both CYTRAN and CYNEQ, is discussed and found to be adequate for strong diffuse reflection, however overestimating the net EC loss in the plasma core for weakly reflecting walls by up to 20%. On the whole, the full transport code SNECTR (no longer in active use), for specular reflection, and the exact cylindrical code EXACTEC are in excellent agreement with each other, CYTRAN and CYNEQ resulting to be fast routines suited for use in systematic transport simulations of fusion plasmas.

The TORAY module in PTRANSP is used to generate self-consistent, time-evolving predictions of electron cyclotron current drive and heating in H-mode and Hybrid ITER plasmas. Various scenarios for steering midplane and upper launching antenna are considered. Electron cyclotron current drive (ECCD) is deposition beyond r/a ≃ 0.3 can delay the decrease of the central safety factor in Hybrid plasmas below unity, maintaining q(0) above unity for long durations.

This work deals with the outward particle flux induced by Electron Cyclotron Resonance Heating (ECRH) in TJ-II, usually known as pump-out, and with the kinetic effects induced on transport. Heat wave experiments have been used in TJ-II to study electron heat transport and to estimate the power deposition profile' and several effects that cannot be attributed to a diffusive behaviour were observed, namely convective transport and a non-monotonic delay of heat pulse. An increase of the positive electric field in the plasma core together with a systematic widening of power deposition profile in comparison with the predicted by WKB theory were also seen.

A quasi-optical description of wave beam propagation is described and applied to the calculation of the ECRH power deposition profile in typical tokamak experiments. The quasi-optical model allows a proper description of the effects on beam propagation and absorption from finite spatial and spectral width of the beam as well as from finite spatial and spectral inhomogeneity. Typically, quasi-optical effects result in broadening of the power deposition profile. In case of the ITER Upper Port ECRH launcher this results in up to a factor of two reduction of the figure of merit for NTM or sawtooth control as compared to previous estimates.

Electron Bernstein waves (EBW) can propagate into the high-density region of a plasma because they don't have a cut-off. This makes them attractive both for heating and for the electron temperature measurement in an overdense plasma. In this paper we mainly explore the O-X-B mode conversion scenario into the HSX stellarator. The analysis uses a 3-D ray tracing code which employs the cold plasma dispersion relation for O-X-wave propagation, while the electron temperature is taken into account for the EBW case. It is shown that O-X-B mode conversion with high efficiency may take place in the plasma region around the HSX box port.

A review of the present status of the performances of Electron Cyclotron Current Drive (ECCD) in ITER is presented. Injection of EC power is considered from both the Upper Launcher and the Equatorial Launcher taking into account the actual design of the launching system, aiming at different physical effects. Key objectives of the whole system are central heating, q profile control and stabilization of magnetohydrodynamic instabilities, mainly Neoclassical Tearing Modes and sawteeth. Investigations relevant to different ITER standard scenarios are reported, as well as to reduced toroidal field scenarios. ECCD sensitivity to the chosen scaling of plasma parameters at reduced fields is discussed. Calculations have been performed using the GRAY code. A discussion on the validity of the theoretical models used for the computation of EC propagation, absorption and current drive is presented with reference to plasma parameters and EC launching conditions such as those foreseen in ITER.

Electron Bernstein waves (EBW) have been confirmed as a suitable choice for plasma heating and current drive generation (EBCD) at densities where the O and X modes find cut-off values. In the present work an estimation of the efficiency function and current generated for a relavistic distribution function of electrons is presented. The arbitrary large values of the refractive index, due to the electrostatic nature of EBW, has made necessary the expansion of our calculation up to any Larmor radius order. Particle trapping has been included thus considering the Okhawa effect. The capability to generate current in a plasma similar to the confined in TJ-II, where an EBW heating system has been recently installed, is also discussed.

The progress on electron cyclotron heating and current drive (EC H&CD) presented in the EC-15 workshop is summarized. Significant progress has been made on gyrotron development. In particular, following a long term R&D effort, gyrotron performance has now exceeded the ITER criteria. Progress of R&D and the design of ITER launchers and transmission lines were presented, which gives a clear prospect for EC H&CD system availability on the first day of operations. In parallel, many efforts devoted to innovation and upgrading of EC H&CD systems on worldwide fusion devices were reported.

The EU programme for the development of the gyrotron for ITER, carried out in cooperation between European laboratories and industrial partners, consists of (i) the manufacturing and testing the 2 MW, CW coaxial cavity gyrotron at 170 GHz meeting both technological constraints and ITER specifications. This goal will be conducted in three different steps: 1 s, 60 s and CW; (ii) in parallel, the development and improvement of critical components for the coaxial concept will continue. While the coaxial concept provides substantial advantages in terms of cost and EC power available for the ITER system with respect to the 1 MW cylindrical cavity gyrotron, the development programme entails additional risks due to both the relatively tight schedule and the developments still required. A fall back option consisting in a more traditional 1 MW gyrotron is foreseen in case of major difficulties in the development of the 2 MW coaxial cavity gyrotron. This alternative development is expected to benefit from the successful development of the 140 GHz gyrotron for W7-X; (iii) a Test Facility for testing at high power, CW the prototype gyrotrons and other EC H&CD system components has been established at the CRPP, in Lausanne. The tests with the first prototype of coaxial gyrotron are currently progressing. A decision point between the coaxial and cylindrical cavity concept is expected by the end of 2009 when the tests with the first and refurbished prototype will be available.

A 95 GHz, multi-megawatt continuous-wave (CW) gyrotron oscillator is currently under development at CPI. The gyrotron consists of a single-anode magnetron injection gun designed to operate at 75 A and 90 kV, a TE_22,6 mode cylindrical interaction cavity, an internal mode converter to transform the TE_22,6 mode to a Gausssian beam, an edge-cooled CVD diamond output window, and a single-stage depressed collector fabricated from a strengthened copper alloy. During the initial experimental campaign, carried out in the Summer of 2007, peak output power levels up to 1 MW at 40 A beam current were demonstrated at pulse lengths up to 5 ms. In addition, pulses up to 15 s in duration at 25 A beam current, the long-pulse limit of the CPI test stand, and 630 kW peak output power were achieved. In the Fall of 2007, modifications to the CPI test stand were made to allow for short-pulse operation up to 75 A. A second test campaign, aimed at demonstrating peak output power in excess of 1 MW, is planned for early in 2008.

ECRH is the main heating system for W7-X. A heating power of 10 MW with CW-capability is required to meet the scientific objectives. The different heating- and current drive scenarios, which support W7-X operation at various magnetic fields and in different density regimes are briefly reviewed. The ECRH plant consists of 10 RF-modules with 1 MW power each. The RF-beams are transmitted to the W7-X torus via two quasi-optical open multi-beam mirror lines. Integrated high-power CW tests of the transmission system including all types of optical elements except the launcher were performed recently and first results are presented.

The ECRH system for ASDEX Upgrade is currently extended by 4 multi-frequency units of 1 MW, 10 s each. The first unit, a two-frequency system is operational since early 2007. The contribution reports on the commissioning of this unit and on two examples of application: at 140 GHz the system is crucial for H-mode operation with the fully W-coated plasma facing components now present in ASDEX Upgrade for moderate values of density and safety-factor, to suppress central accumulation of heavy impurities. At 105 GHz the EC-beam is used as probe for a collective Thomson scattering diagnostic, which uses the launcher of the second unit as receiver. The other 3 units shall additionally operate at intermediate frequencies, requiring alternative solutions for the windows. For the gyrotron window a Brewster-angle set-up is used successfully. At the torus a Fabry-Pérot-type double-disk window is used. It has been tested successfully with low power.

Calculation of the RF propagation inside and outside of the ITER ECH and CD equatorial launcher is a critical task of the design. In the conventional design, the RF power is delivered by 8 waveguides to a steering mirror followed by a focus mirror through a slot of neutron shield. In order to optimize the structures of the launcher, the RF propagation from the waveguide to the RF absorption layer in the plasma is calculated. The optimization is taking account of the neutron shielding, RF propagation efficiency inside the launcher, heat load on the injection mirror and the beam profile at the absorption layer. In this paper, the numerical optimization is performed to obtain the desired beam at the absorption layer. Since the desired beam profile is not clear for the equatorial launcher, the optimization for peaking and broadening profile is carried out. The beam size is successfully reduced from 35.6 cm to 23.7cm for peaking optimization and increased to 62.1 cm for broadening optimization without any reduction of the neutron shielding. The propagation efficiency of RF power through the launcher is 99.5% and the heat load of the injection mirror is less than 2.2 MW/m² for both designs.

A significant upgrade to the 110 GHz DIII-D ECH system was completed last year. Two additional Communication and Power Industries (CPI) diode gyrotrons were installed and tested to half the designed pulse length of 10 s. For the 2008 experimental campaign, the DIII-D system comprised five long pulse CPI gyrotrons. One additional high voltage power supply is being tested to support operation of up to 6 gyrotrons simultaneously at full parameters. The five gyrotrons in operation have chemical-vapor-deposition (CVD) diamond windows that are monitored by infrared camera during full parameter testing and operation during plasma experiments. A sixth CPI gyrotron has been repaired after collector failure and is being conditioned for high power at DIII-D. New equipment for gyrotron collector heat load monitoring was tested and used to measure the collector power deposition profile. A new fast fault processing system based on FPGA technology is being commissioned.

Fast switches for high-power millimeter waves are of interest for electron cyclotron resonance heating (ECRH) systems as they allow sharing of the installed power between different types of launchers or different applications, whichever is given priority during a plasma discharge. The switching can be controlled electronically without moving parts by a small frequency-shift keying of the gyrotron (some 10s of MHz), and a narrow-band diplexer, which directs an input beam to one of the two output channels. Devices of this type can be integrated into corrugated waveguide as well as mirror transmission lines. They can be operated as a power combiner, and can be connected consecutively to combine wave beams from different gyrotrons. In the paper, the principle and the design of a 140 GHz four-port quasi-optical resonator diplexer is presented. Low-power measurements of switching contrast, mode purity and efficiency are shown to agree well with theory. Frequency modulation experiments with gyrotrons are performed. Results from high-power switching of a beam from a gyrotron of the ECRH system for the stellarator W7-X are presented, and the preparations for a power combination experiment with two gyrotrons are discussed.

The ITER EC system is in the process of being revised as a result of the ITER design review that occurred during 2007. This revision is introducing several modifications to the gyrotron and transmission line layout as well as identify several interface issues between the subsystems comprising the EC system. Each modification aims at improving the EC systems performance while reducing the overall cost and increasing reliability. These proposed revisions have now been assembled together into a coherent system design offering an alternative baseline design of the EC system.

The aim of this paper is to review present status of the EC system design highlighting the improvements over the previous baseline design. The proposed changes cover the new RF building, transmission line layout and launchers. In addition, several outstanding issues are still under investigation, which involve the functional aspects of the EC system in heating.

The main purpose of the ITER ECH upper launcher is to control magnetohydrodynamic activity, in particular neoclassical tearing modes, in the plasma. The mm-wave optical system is optimized to insure that the eight RF beams are all focused to a small beam width at the resonance location. The current design uses two miter bends per beam and a focusing mirror for each group of four beams to orientate the beams into onto a single steering mirror for each group of four beams. The toroidal and poloidal launch angles have been determined by numerical calculations to ensure nearly coincident deposition over the entire steering range. The system is being tested with a low RF power beam to ensure that the optical system directs and properly focuses each beam. The fourth mirror is steered using a frictionless and backlash-free mechanical system which is pneumatically actuated using helium to provide rapid and accurate angular positioning of the mirror. A first prototype of the steering mechanism has been constructed to demonstrate the manufacturability, the actuation principle and to develop an adequate control strategy to meet ITER requirements.

A new transmission line has been designed for the planned upgrades for the 28GHz ECRH system on the HSX stellarator. In order to create a reliable and cost effective system, the new line utilizes quasioptical design techniques and oversized straight waveguide sections similar to those used in the existing transmission line. The first stage of the line converts the native TE₀₂ gyrotron output mode to a free space Gaussian beam. The two methods considered are a Vlasov mode converter and a TE₀₂–to–TE₀₁–to–TE₁₁–to–HE₁₁ conversion sequence using waveguide mode converters retained from a previous ECRH system. A matching optics unit couples the beam to a smooth, circular ₁₁ + TM₁₁ dual-mode waveguide sections for traversing the experimental hall, and a grooved polarizer rotates the beam polarization for O–mode or X–mode ECRH. Transmission line bends are implemented with ellipsoidal mirrors, with the final bend acting as a rotatable switch to select between HSX and a dummy load. A steerable mirror in boxport B allows independent selection of the heating surface in the plane of the boxport.

A new multi-frequency quasi-optical mode converter design for four frequency/mode pairs for a step-tunable gyrotron is discussed. The design frequency/mode pairs are 110 GHz/TE_22,6, 124.5 GHz/TE_24,7, 127.5 GHz/TE_25,7 and 107.5 GH/TE_21,6. The design is for an 88 mm diameter variable-spacing double-disk vacuum window. All four design frequencies have an output propagation direction angle from the window axis of 0.4° or less and a complex coupling to an ideal Gaussian beam greater than 99.45% over the window aperture.

The measurement of the power injected by the electron cyclotron heating (ECH) system in the DIII-D tokamak is a critical requirement for analysis of experiments, for tuning the gyrotrons for maximum power and efficiency, for tracking long-term operational trends and for providing a warning of problems with the system. The ECH system at General Atomics consists of six 110 GHz, 1 MW class gyrotrons. The rf power generated by each gyrotron is determined from calorimetry, using the relevant temperature and flow measurements from the cooling circuits of cavity, matching optics unit and dummy loads. The rf pulse length and time dependence are measured using an rf monitor at the first miter bend in the transmission line. The direct measurement of the efficiencies of four of the transmission lines was performed using a high power, small dummy load (SDL) placed alternately in 2 positions of each DIII-D waveguide line, at accessible points close the beginning and the end of each line. Total losses in the transmission lines range from 21.2% to 30.7%. Experimental results are compared to theoretical predictions of the performance of the components and waveguide lines.

In the context of the completion of the ITER ECRH&CD Upper Launcher detailed design phase, we conclude the beam truncation study on the focusing mirror and we perform an evaluation of the effect of the propagation through the shielding block structure on the beam propagation. For this purpose, we used a simplified version of the blanket module with different propagation paths and varying distance between beam center and wall.

The evolution of beams has been calculated with the electromagnetic code GRASP®, a numerical tool in which the detailed characteristics of the reflectors surfaces and the surrounding structures can be introduced and the resulting field propagation in vacuum can be computed including all the relevant effects. The description of the beam resulting from Physical Optics calculations takes into account the relevant causes of deformation and non gaussianity, such as aberration and beam truncation and diffraction from structure edge.

In the electromagnetic simulation, we first consider the extreme case of a beam running close to one corner all along the blanket length: in this worst case the beam is affected by two faces of the shield module. Intermediate cases (beam propagating close to one side or entering close to the edge and then deviating) are also considered.

The installation of 84 GHz Electron Cyclotron Heating (ECH) system has been completed in Korea Superconducting Tokamak Advanced Research (KSTAR). The 84 GHz EC-wave is transmitted from the gyrotron to the antenna system through evacuated HE₁₁-mode circular corrugated waveguides with inner diameter of 31.75 mm and miter bends with flat mirror. The L-box chamber is designed and installed to match the output beam of TEM00 mode from the gyrotron into the 31.75-mm circular corrugated waveguide by a large ellipsoidal mirror inside the chamber. The ellipsoidal mirror can be adjusted in two perpendicular axes and moved in or out for the minor misalignment. The elliptical polarization of EC-wave required for the specific mode coupling at the plasma edge is obtained by means of subsequent two grooved mirrors inside the miter bend. The torus vacuum window, which is half-wavelength thick diamond disk, is installed for the vacuum isolation of the transmission line vacuum from the KSTAR vacuum vessel. For the RF power measurement, two dummy loads are installed near the gyrotron and near the KSTAR tokamak, respectively. The dummy load near the KSTAR tokamak will be used for the transmission efficiency measurement of the transmission line system. This paper describes the evacuation test result of the transmission line from the L-box to the torus diamond window, the transmission efficiency measurement of 84-GHz millimeter wave, and the measurement of elliptical polarization parameters as a function of the grooved mirror angles. Also, the requirement of the polarization control for the EC-wave injection and mode coupling is discussed.

Ultra low mm-wave loss CVD diamond windows with loss tangent tan δ < 2 × 10^-5 have been manufactured. The status of the window design is reviewed. The fabrication of the prototypical window unit is qualified for the selected diamond grades by a dielectric characterisation carried out by low power measurement (FZK). The high power RF performance is quantified by measurements (IR imaging during high power microwave loading) at the 170 GHz/1 MW gyrotron of JAEA and by investigating arcing processes. For this purpose, window assemblies (steel/copper/diamond) adapted to dedicated waveguides are used.

A 28 GHz electron cyclotron heating (ECH) and electron Bernstein wave heating (EBWH) system has been proposed for installation on the National Spherical Torus Experiment (NSTX). A 350 kW gyrotron connected to a fixed horn antenna is proposed for ECH-assisted solenoid-free plasma startup. Modeling predicts strong first pass on-axis EC absorption, even for low electron temperature, T_e ~ 20 eV, Coaxial Helicity Injection (CHI) startup plasmas. ECH will heat the CHI plasma to T_e ~ 300 eV, providing a suitable target plasma for 30 MHz high-harmonic fast wave heating. A second gyrotron and steered O-X-B mirror launcher is proposed for EBWH experiments. Radiometric measurements of thermal EBW emission detected via B-X-O coupling on NSTX support implementation of the proposed system. 80% B-X-O coupling efficiency was measured in L-mode plasmas and 60% B-X-O coupling efficiency was recently measured in H-mode plasmas conditioned with evaporated lithium. Modeling predicts local on-axis EBW heating and current drive using 28 GHz power in β ~ 20% NSTX plasmas should be possible, with current drive efficiencies ~40 kA/MW.

The EU is working towards providing 2 MW, coaxial-cavity, CW, 170 GHz gyrotrons for ITER. Their design is based on results from an experimental pre-prototype tube in operation at FZK for several years, having a pulse length of several milliseconds. The first industrial prototype tube is designed for CW operation, but, in a first phase, will be tested out to 1s at the European Gyrotron Test Facility in Lausanne, Switzerland as part of a phased testing/development program (1 s, 60 s, CW). It is known that RF beam profile shaping, stray radiation handling, and collector cooling at these high power levels are three issues for the gyrotron, The gyrotron, magnet and body power supply have been delivered and successfully installed at the test stand, hosted by the CRPP. The main high voltage power supply delivery is delayed, so one of the power supplies dedicated to 3 of 9 gyrotrons in the TCV EC system is being used as a backup power source (all 3 TCV power sources can be interfaced with the test stand). Cathode conditioning began in November 2007 followed by collector conditioning in December. Parasitic low frequency oscillations have not hindered operation, and the tests have progressed to conditioning out to 0.14 s pulses by March 2008. During this period, the perfomance concerning microwave generation has been characterised and the RF beam profile has been measured at several planes to allow reconstruction of the phase and amplitude profile at the gyrotron window and to provide the necessary information permitting proper alignment of the compact RF loads prior to pulse extension. The power will be measured, according to the pulse length, using either a very-short pulse (<0.01 s) load on loan from FZK, or short-pulse (<0.2 s) or long-pulse (CW), spherical, calorimetric loads developped as part of this program by CNR. This paper presents the preliminary results of these operations.

A coaxial cavity gyrotron, similar to the EU 2 MW tube for ITER, is under investigation at FZK at reduced magnetic field, which limits the expected RF power to 1.5 MW. While in former experiments only 1.1 MW was obtained, now, after the electron gun was changed, parasitic low frequency oscillations were removed and an RF power of 1.3 MW was reached. The power now seems to be limited by another parasitic high frequency oscillation at 160 GHz, which appears simultaneously to the desired working mode.

A second problem of the pre-prototype tube, insufficient Gaussian mode content of the output wave of only 76%, was investigated by verifying the employed tools. The results indicate a problem at the third, phase correcting mirror, while the simulations appear to be acceptable. Finally, to further improve the quasi-optical mode converter a different approach to launcher design using arbitrary wall deformations is underway.

A 10 MW ECRH system is currently under construction for the stellarator W7-X which will be built up and operated by IPP in Greifswald, Germany. The present status of the complete system is reported in [1]. The RF power will be provided by 10 gyrotrons. A European collaboration has been established to develop and build the 10 gyrotrons each with an output power of 1 MW for continuous wave (CW) operation [2]. Nine gyrotrons are being manufactured by Thales Electron Devices (TED), Vélizy, France, one gyrotron was produced by CPI, Palo Alto, CA and passed the acceptance tests at IPP. The acceptance tests of the TED gyrotrons are performed at the test stand at FZK and on site at IPP. The first series tube yielded a total output power of 0.98 MW, with an efficiency of 31 % (without a single stage depressed collector) in short pulse operation and of 0.92 MW in pulses of 1800 s (efficiency of almost 45 % at a depression voltage of 29 kV) [3], The Gaussian mode output power was 0.91 MW. The RF power, measured in a calorimetric load at the end of a 25 m long quasi-optical transmission line with seven mirrors, was 0.87 MW.

In this contribution typical results of the next series gyrotrons will be reported.

Fast poloidal magnetic field changes during plasma disruptions in ITER generate forces and torque moments on the in-vessel components. In numerical electromagnetic simulations the eddy currents and the resulting mechanical loads have been calculated for the reference upward VDE disruption with a fast linear current quench. In structural analyses the loads are applied to a 3D finite element model of the upper launcher structure based on the actual launcher design including the upper port. The analysis covers the launcher deformation in comparison to the spacing to neighbouring components and the classification of the occurring stress peaks.

The performance requirement of 1 (possibly 2) MW cw at 170 GHz for ITER Electron Cyclotron Heating and Current Drive transmission line components is much more demanding than the 1 MW 10 s performance, generally at 110 GHz, that has been demonstrated on present devices. The high ITER heat loads will require enhanced cooling and, for some components, new or modified designs. In addition to thermal management issues, the components must be designed to have very low losses in order to meet the ITER transmission line efficiency requirements. Testing at representative ITER conditions of some components has been initiated at the JAEA 170 GHz gyrotron test stand at Naka, Japan. Plans for testing additional components, both at JAEA and in the US will be presented. The US ITER Project Office (USIPO) plans to test a complete prototype ITER transmission line in order to validate the designs for use on ITER, and GA is providing some components to the USIPO for initial tests.

Recent activities of JAEA 170 GHz gyrotron development and EC H&CD technology are introduced. The basic criteria of ITER were satisfied using a TE_31,8 mode gyrotron. In the next step, a TE_31,12 mode gyrotron was fabricated and the long pulse experiment is planned in 2009. In parallel, feasibility studies of power modulation and dual frequency gyrotrons have progressed. Power from the operating gyrotron is being used for the development of transmission line components and the ITER launcher.

The following sections are included:

• LIST OF PARTICIPANTS

• AUTHOR INDEX