Search

A one dimensional model of the sheath between the plasma and the wall in a negative ion source has been developed. The plasma consists of positive ions, electrons and negative ions. The model takes into account the emission of negative ions from the wall into the sheath and thus represents the conditions in a caesiated ion source with surface production of negative ions. At high current densities of the emitted negative ions, the sheath is unable to support the transport of all the negative ions to the plasma and a virtual cathode is formed. This model takes this into account and allows the calculation of the transported negative ions across the sheath with the virtual cathode. The model has been extended to allow the linkage between plasma conditions at the sheath edge and the plasma to be made. Comparisons are made between the results of the model and experimental measurements.

The Superconducting Proton Linac (SPL) is a novel linear accelerator concept currently studied at CERN. As part of this study, a new Cs-free, RF-driven external antenna H{sup -} plasma generator has been developed to withstand an average thermal load of 6 kW. The magnetic configuration of the new plasma generator includes a dodecapole cusp field and a filter field separating the plasma heating and H{sup -} production regions. Ferrites surrounding the RF antenna serve in enhancing the coupling of the RF to the plasma. Due to the space requirements of the plasma chamber cooling circuit, the cusp magnets are pushed outwards compared to Linac4 and the cusp field strength in the plasma region is reduced by 40% when N-S magnetized magnets are used. The cusp field strength and plasma confinement can be improved by replacing the N-S magnets with offset Halbach elements of which each consists of three magnetic sub-elements with different magnetization direction. A design challenge is the dissipation of RF power induced by eddy currents in the cusp and filter magnets which may lead to overheating and demagnetization. In view of this, a copper magnet cage has been developed that shields the cusp magnets from the radiation ofmore » the RF antenna.« less

We have been developing a negative hydrogen ion (H− ion) source for a spatial beam profile monitor of a high intensity positive ion beam as a new diagnostic tool. In case of a high intensity continuous‐wave (CW) deuteron (D+) beam for the International Fusion Materials Irradiation Facility (IFMIF), it is difficult to measure the beam qualities in the severe high radiation environment during about one‐year cyclic operation period. Conventional techniques are next to unusable for diagnostics in the operation period of about eleven months and for maintenance in the one‐month shutdown period. Therefore, we have proposed an active beam probe system by using a negative ion beam and started an experimental study for the proof‐of‐principle (PoP) of the new spatial beam profile monitoring tool. In this paper, we present the status of development of the H− ion source as a probe beam source for the PoP experiment.

The development of a high performance negative ion (NI) source constitutes a crucial step in the construction of Neutral Beam Injector (NBI) of the future fusion reactor ITER. NI source should deliver 40 A of H− (or D−), which is a technical and scientific challenge, and requires a deeper understanding of the underlying physics of the source and its magnetic filter. The present knowledge of the ion extraction mechanism from the negative ion source is limited and concerns magnetized plasma sheaths used to avoid electrons being co‐extracted from the plasma together with the NI. Moreover, due to the asymmetry induced by the ITER crossed magnetic configuration used to filter the electrons, any realistic study of this problem must consider the three spatial dimensions. To address this problem, a 3D Particles‐in‐Cell electrostatic collisional code was developed, specifically designed for this system. Binary collisions between the particles are introduced using Monte Carlo Collision scheme. The complex orthogona...

Surface condition of a plasma grid in a negative hydrogen ion source is controlled so as to maximize the beam current under a discharge operation with introducing Cs into the ion source. Photoelectric current induced by laser beams incident on the plasma grid can produce a signal to monitor the surface condition, but the signal detection can be easily hindered by plasma noise. Reduction in size of a detection electrode embedded in the plasma grid can improve signal‐to‐noise ratio of the photoelectric current from the electrode. To evaluate the feasibility of monitoring surface condition of a plasma gird by utilizing photoelectric effect, a small experimental setup capable of determining quantum yields of a surface in a cesiated plasma environment is being assembled. Some preliminary test results of the apparatus utilizing oxide cathodes are reported.

A three dimensional ion optical code IBSimu is being developed at the University of Jyvaskyla. So far the plasma modelling of the code has been restricted to positive ion extraction systems, but now a negative ion plasma extraction model has been added. The plasma model has been successfully validated with simulations of the Spallation Neutron Source (SNS) ion source extraction both in cylindrical symmetry and in full 3D, also modelling electron beam dumping and ion beam tilt. A filament‐driven multicusp ion source has been installed at the Texas A&M University Cyclotron Institute for production of H− and D− beams as a part of the facility upgrade. The light ion beams, produced by the ion source, are accelerated with the K150 cyclotron for production and reacceleration of rare isotopes. The extraction system for the ion source was designed with IBSimu. The extraction features a water‐cooled puller electrode with a permanent magnet dipole field for dumping the co‐extracted electrons and a decelerating Einz...

Cusp leak width (CLW) is an important parameter used in designing of H‐ ion source as it helps in determining the total power requirement of the source by considering particle loss at the multicusp regions. This parameter has been derived by many workers based on certain assumptions and approximations but it does not take into account the curved nature of magnetic lines of force in the cusp region[1–3]. This statement is vindicated by the fact that above method derives same expression for CLW irrespective of different cusp geometries. Similarly, the final expression of CLW depends on ion acoustic velocity, magnetic field at the cusp and half‐length of magnetic lines of force [2,3]. The last parameter is the only geometrical parameter in the expression. However, it does not define which particular half‐length of magnetic lines of force has been considered, thus, leading to insensitivity to geometrical aspects of the multicusp: planar, cylindrical etc.In the present analytical study, we report on a rigorous...

We describe the characteristic of stable beam injection in a neutral beam injector (NBI) for Large Helical Device (LHD) in high injection power of more than 6 MW. In the NBI, it takes a week after starting Cs seeding to finish the pre‐injection conditioning. The injection starts with the beam power ∼6.2 MW, and the maximum power reaches ∼7 MW. The Cs‐seeding rate affects the beam stability in such high power injections. By optimizing the rate to 0.65 mg/shot, the success ratio, which is defined as a ratio of actual pulse duration to setting one, increases to 85–90% in the power and energy range of more than 6.2 MW and 185 keV, respectively. The weights of Cs adsorbed on several surfaces in the ion sources of the NBI are measured by means of Inductively‐Coupled Plasma Atomic Emission Spectroscopy (ICP‐AES), and averaged surface densities are calculated by dividing with the several surface areas. The seeded Cs of 99.5% is condensed in the plasma generator, and very tiny amount of Cs reaches the surfaces of the accelerator grids. This very low amount of Cs on the grids is interpreted that most of the Cs atom evaporated from the inner walls is ionized during the arc discharges, and repelled to the source plasmas by the electrostatic field for H‐ extraction.

Vacuum ultra violet (VUV) emission spectra from plasmas near walls of different metallic materials were measured to estimate the effect upon the local production rate of vibrational excited hydrogen molecules due to plasma wall interaction. Among Cu, Mo, Ni, Ta and Ti, the intensity of band spectrum around Lyman‐α had become the largest when Cu wall was used while it was the smallest for Ti. The role of particle reflection from the plasma electrode surface upon the H− production by a pure electron volume process is discussed.

An important feature of the ITER project is represented by additional heating via injection of neutral beams from accelerated negative ions. To study and optimise their production, the SPIDER test facility is under construction in Padova, with the aim of testing beam characteristics and to verify the source proper operation. STRIKE (Short-Time Retractable Instrumented Kalorimeter Experiment) is a diagnostic to characterise the SPIDER negative ion beam during short operation (several seconds). During long pulse operations, STRIKE is parked off-beam in the vacuum vessel. The most important measurements are beam uniformity, beamlet divergence and stripping losses. STRIKE is directly exposed to the beam and is formed of 16 tiles, one for each beamlet groups. The measurements are provided by thermal cameras, current sensors, thermocouples and electrostatic sensors. This paper presents the investigation of the influence on the response of STRIKE of: thermal characteristics of the tile material, exposure angle, features of some dedicated diagnostics. The uniformity of the beam will be studied by measurements of the current flowing through each tile and by thermal cameras. Simulations show that it will be possible to verify experimentally whether the beam meets the ITER requirement about the maximum allowed beam nonuniformity (below ±10%). In the simulations also the influence of the beam halo has been included; the effect of off-perveance conditions has been studied. To estimate the beamlet divergence, STRIKE can be moved along the beam direction at two different distances from the accelerator. The optimal positions have been defined taking into account design constraints. The effect of stripping on the comparison between currents and heat loads has been assessed; this will allow to obtain an experimental estimate of stripping. Electrostatic simulations have provided the suitable tile biasing voltage in order to reabsorb secondary particles into the same tile as the one where they were emitted from.

Conference date: 16-19 November 2010 Location: Takayama City, (Japan); International audience; The knowledge of the mean free path (mfp) of the negative ions is important for many purposes, e.g. evaluating the distance from which the ions can be extracted, or understanding the significance of experimental results. We will present the mfp for similar conditions in volume and caesium seeded sources. It appears that the mfp is longer in the caesium seeded source than in the volume source under identical conditions.

An additional beamline, BL5, equipped with four positive ion sources will be installed on Large Helical Device (LHD) in 2010. The performance of an ion source which generates 80 keV deuterium and 60 keV hydrogen beams was investigated. The structure of the ion source is based on that of a BL4 ion source on LHD. The main differences between the ion sources for the BL4 and BL5 are the acceleration voltages and the materials of plasma electrodes: copper and molybdenum, respectively. The molybdenum plasma electrode for BL5 has better performance than the copper plasma electrode of BL4. The integrated performance of the ion source for BL5 reached a value equivalent to approximately 58 A in the beam current of hydrogen positive ion at 60 keV in the beam energy.

The phase space distribution of a strongly focused He(+) ion beam source equipped with concave multiaperture electrodes was measured using a pepper-pot plate and a Kapton foil. The substructure of 301 merging He beamlets was clearly observed on a footprint of pepper-pot hole at the beam waist, where the beam density was 500 mA/cm(2). The position and the width of each beamlet substructure show the effect of interference of beamlets with surrounding one.

A spontaneous rotation in the co-direction is observed in plasmas with an ion internal transport barrier (ITB), where the ion temperature gradient is relatively large (∂T i/∂r ∼ 5 keV m−1 and ) in LHD. Because of the large ion temperature gradients, the magnitude of the spontaneous toroidal flow, , becomes large enough to cancel the toroidal flows driven by tangential injected neutral beams and the net toroidal rotation velocity is almost zero at the outer half of the plasma minor radius even in the plasmas with counter-dominant NB injections. The effect of velocity pinch is excluded even if it exits because of zero rotation velocity. The spontaneous toroidal flow appears in the direction of co-rotation after the formation of the ITB, not during or before the ITB formation. The causality between the change in and ∂T i/∂r observed in this experiment clearly shows that the spontaneous rotation is driven by the ion temperature gradient as the off-diagonal terms of momentum and heat transport.

At the National Institute for Fusion Science (NIFS) large-scaled negative ion sources have been widely used for the Neutral Beam Injectors (NBIs) mounted on the Large Helical Device (LHD), which is the world-largest superconducting helical system. These injectors have achieved outstanding performances in terms of beam energy, negative-ion current and optics, and represent a reference for the development of heating and current drive NBIs for ITER. In the framework of the support activities for the ITER NBIs, the PRIMA test facility, which includes a RF-drive ion source with 100 keV accelerator (SPIDER) and a complete 1 MeV Neutral Beam system (MITICA) is under construction at Consorzio RFX in Padova. An experimental validation of the codes has been undertaken in order to prove the accuracy of the simulations and the soundness of the SPIDER and MITICA design. To this purpose, the whole set of codes have been applied to the LHD NBIs in a joint activity between Consorzio RFX and NIFS, with the goal of comparing and benchmarking the codes with the experimental data. A description of these modeling activities and a discussion of the main results obtained are reported in this paper.

For the validation of the numerical models used for the design of the Neutral Beam Test Facility for ITER in Padua [1], an experimental benchmark against a full-size device has been sought. The LHD BL2 injector [2] has been chosen as a first benchmark, because the BL2 Negative Ion Source and Beam Accelerator are geometrically similar to SPIDER, even though BL2 does not include current bars and ferromagnetic materials. A comprehensive 3D magnetic field model of the LHD BL2 device has been developed based on the same assumptions used for SPIDER. In parallel, a detailed experimental magnetic map of the BL2 device has been obtained using a suitably designed 3D adjustable structure for the fine positioning of the magnetic sensors inside 27 of the 770 beamlet apertures. The calculated values have been compared to the experimental data. The work has confirmed the quality of the numerical model, and has also provided useful information on the magnetic non-uniformities due to the edge effects and to the tolerance on permanent magnet remanence.

International audience; Extraction and acceleration of positive and negative ions from a strong electronegative plasma and from an ion‐ion plasma is investigated in the PEGASES thruster, working with SF6. The plasma is generated in a cylindrical quartz tube terminated by metallic endplates. The electrons are confined by a static magnetic field along the axis of the cylinder. The electron mobility along the field is high and the electrons are determining the sheaths in front of the endplates. The core plasma potential can therefore be controlled by the bias applied to the endplates. An ion‐ion plasma forms at the periphery as a result of electron confinement and ions can freely diffuse along the perpendicular direction or extraction axis. Langmuir probe and RFEA measurements are carried out along this axis. The measured ion energy distributions shows a single peak centered around a potential consistent with the plasma potential and the peak position could be controlled with a positive voltage applied to the endplates. When the endplates are biased negatively, the plasma potential saturates and remained close to 15 V. A beam of negatively charged particles can be observed under certain conditions when the endplates were biased negatively.

This paper reviews recent experimental results from compact helical system (CHS), a heliotron/torsatron type device with a low aspect ratio of 5. Transport phenomena are discussed through analyses of MHD equilibrium, thermal/particle diffusivity and viscosity. It is seen that transport is dominated by anomalous processes which are thought to be due to electrostatic fluctuations. On the basis of these results, optimization of confinement by shifting the vacuum magnetic axis is briefly discussed.

A new multi-cusp H− ion source using a localized virtual magnetic filter of type I in the plasma electrode is investigated. A multi-pole arrangement with a spacing of 10 mm of the magnet bars holds an extraction hole, optimizing the efficient production of high H− current, and at the same time only a small electron component was co-extracted with the H− ions. The local filter arrangement separates the beam electrons at a low energy. It is shown that the coextracted total electron current is determined principally by the integrated magnetic field flux (Gcm) of the local filter with an extraction system at a constant extraction voltage. When the value of the Gcm is increased, the total electron component is reduced, while the H− electrical efficiency had a broad maximum around the optimized value of the Gcm. A thicker plasma electrode should be necessary for sufficient reduction of electron current. In pure hydrogen operation, the achieved current density of H− is 10 mA/cm2. When Cs is seeded in a filter op...

Cesium recovery from the polluted layers in the 1/3 scale hydrogen negative ion source for LHD-NBI system has tested. It was found that the cesium recovery can be produced by additional discharges as from the cesium layer, aged by tungsten and residual gas, so as from the cesium layers, polluted by an occasional water leak. The highest cesium recovery to negative ion production was produced by a xenon arc, while glow discharge and arcing in hydrogen were less effective. The mechanism of recovery is the ejection of cesium from the underlying enriched layer by the arc and its transport to the surface.

A strongly focusing high-intensity He(+) ion source equipped with three concave electrodes has been designed and constructed as the beam source for a high-energy He(0) neutral beam probe system to diagnose fusion-produced alpha particles in thermonuclear fusion plasmas. The reduction of heat load onto the concave extraction electrodes is particularly important for a long pulse operation, as the heat load deforms the electrodes and thus the beam focal length. The effects on the arc efficiency (beam current/arc power) of the ion source due to the discharge filament structure (straight-type and L-shape-type filaments), size (filament diameters of 2 and 1.5 mm), number, and the locations have been studied. Choice of the appropriate filament structure improved the arc efficiency by 17%.

To investigate a Cs behavior, optical diagnostic tools have been installed in the large negative ion source, an arc discharge used at large helical device neutral beam injector. A large Cs sputtering is observed during beam extraction due to the backstreaming H(+) ions. Distribution of Cs(+) light is uniform in the case of a balanced arc discharge, but large increase of Cs(+) light during beam extraction is observed in a nonuniform arc discharge. Controlling of the discharge uniformity is effective to reduce the local heat loading from the backstreaming H(+) ions at the backplate of ion source.

In the large area negative ion source for the LHD negative-ion-(H(-))-based neutral beam system, (I) we used the spectrometer to measure caesium lines in the source plasma during beam shots. (II) With Doppler-shifted measurements, the H(alpha) line at three different locations along the beam as well as the spectrum profile for cases of different plasma grid areas. (III) Caesium deposition monitor with a high speed shutter was tested to measure the weight of the deposited Cs layer. In the observation, cleaner spectra of Doppler-shifted H(alpha) line with only a small level of background light were obtained at a new observation port which viewed the blueshifted light in the drift region after the accelerator of a LHD ion source. Both the amounts of Cs I (852 nm, neutral Cs(0)) and Cs II (522 nm, Cs(+)) in the source plasma light rose sharply when beam acceleration began, and continued rising during a 10 s pulse. It was thought that this was because the cesium was evaporated/sputtered from the source back plate by the back-streaming positive ions. Cs deposition rate to the crystal sensor measured by adjusting the shutter open time was evaluated to be 2.9 nanograms/s cm(2) for preliminary testing. More neutral Cs tended to be evolved in the source after arc discharge. Much Cs could be consumed in a high rate-pulsed operation (such as LHD source).

We have developed a time-of-flight analyzer to measure energy distributions of reflected particles from solid surfaces bombarded by low-energy (1?2 keV) ions. The analyzer yields energy distributions of neutrals which can be compared with the energy distributions of charged particles measured by a magnetic deflection-type momentum analyzer. We have tested the system to measure the angular dependence of energy and intensity for neutrals reflected from a polycrystalline W target. The energies of the reflected neutrals are much smaller than the incident ion energies, suggesting multiple scattering in the target. No angular dependence is observed under the condition that the sum of the incident and reflected angles is constant. The intensity of the reflected neutrals takes the maximum around the mirror angle. We compare these characteristics of neutral particle reflections with those of reflected ions.

The phase diagram for the palladium-silicon system is reinvestigated by thermal analysis using a method taking the second derivative of time-dependent temperature and by metallographic measurement, where the cooling rate is about 140‡ C min−1. The second derivative offers a sensitive and accurate means for thermal analysis with aid of a microcomputer. The phase diagram has no peak for PdSi. At silicon concentrations higher than 50 at. %, the alloy appears to be the eutectic for a composition of PdSi and Si, with the eutectic point at 895‡ C and 50 at. % Si, and at silicon concentrations lower than 50 at. %, the alloy appears to be the eutectic for Pd2Si and PdSi, with the eutectic point at 886‡ C and about 48.5 at. % Si. The phase properties appear to be largely affected by the phase stability of PdSi in company with the departure from thermal equilibrium in the thermal analysis. The addition of small amounts of silver impurity causes the collapse of the PdSi phase and the alloy appears to be a eutectic of constitutions Pd2Si and Si in the concentration range from 33 to 100 at. % silicon.

Large‐scaled hydrogen negative‐ion sources, in which cesium is introduced in the source plasma, have been developed for neutral beam injectors in Large Helical Device, and their operational characteristics are reviewed. For high‐efficient negative ion production, configuration of the magnetic filter field and the cusp magnetic field was optimized, resulting in a high arc efficiency for the negative ion production of 0.23A/kW. With use of a multi‐slotted grounded grid, the gas pressure in the acceleration gap is lowered, leading to reduction of the heat load of the grounded grid. As a result, the voltage holding ability is much improved, and the rated energy of 180 keV is achieved in a short conditioning period of 4 days. The injection power is increased linearly to the 5/2 power of the beam energy and reached 5.7MW with an energy of 184keV, which exceeds the specified value of 180keV–5MW. Beam uniformity has been improved with an individual control of the local arc discharge by adjusting 12‐divided output...

The discharge scenario of high temperature plasma with a helical configuration has significantly progressed. The increase of central ion temperature due to the reduction of wall recycling was clearly observed. The peaking of the ion heating profile and the reduction of charge exchange loss of energetic ions play an important role for further improvement of ion heat transport in the ion internal transport barrier (ITB) core. The ion ITB and electron ITB have been successfully integrated due to the superposition of centrally focused electron cyclotron heating to the ion ITB plasma, and the high temperature regime of the ion temperature comparable to the electron temperature (Ti ~ Te) has been significantly extended. The width of the ion ITB formed with electron ITB is wider than the width of electron ITB. The positive radial electric field was observed in the integrated ITB plasma by a heavy ion beam probe, while the negative radial electric field was observed in ion ITB plasmas. The ion temperature gradient decreases with the increase of the temperature ratio (Te /Ti).

H− ion density in the extraction region of a volume-production-type ion source is measured, using the photodetachment method, and directly compared to H− ion beam extracted from the ion source in order to understand the effective production of H− ions. It is observed that there is qualitative similarity between H− density and extracted H− beam current, as expected. Characteristics of H− ion production under two values of magnetic strength are compared to study the effect of the filter field. It was found that the weaker magnetic field of 750 G cm is not sufficient for the filter field since high energy electrons can come into the extraction region across the magnetic field.

Two large helical device–neutral beam injector (LHD–NBI) ion sources were fabricated and tested in the test stand for producing a beam of 180 keV×40 A with H− ions. They are Cesiated multicusp ion sources with a rectangular discharge chamber and a single stage multihole accelerator. These are scaled up from the 16 A H− ion sources in the National Institute for Fusion Science (NIFS). A plasma source with a high aspect ratio was operated stably with an arc power up to ∼300 kW for 10 s, after balancing of the electron emission from the filaments was made. A satisfactorily dense and uniform plasma without mode flip was produced. Electrons accompanied by H− ions were reduced by an extraction grid with the electron trap, instead of straight holes. The electron beam component caused by the stripping of electrons from H− ions was detected with an array of calorimeters at the bottom of the connecting duct. At the first stage of the test, one of the five segment grids of the accelerator was installed. An H− ion cur...

In negative-ion-based neutral beam injection (NBI) systems for the large helical device (LHD), beams must be transported over 13 m from the H− ion source to the injection port. In order to clarify beam deflection by the electron deflection magnets set in a beam extraction grid (EG) and to control beam transport direction, we analyzed beam trajectories. The physics of the beam deflection was studied with theoretical calculations and the deflection angle was estimated by 3D beam trajectory simulation. The evaluated deflection angle was 10 mrad in the opposite direction of the electron deflection when the maximum magnetic field on the beam axis was 480 G and the beam energy was 83.2 keV. The electrostatic lens effect on the beam deflection at the EG exit was estimated to be larger than the magnetic field effect. This deflection was reduced to 2 mrad by a 1.3 mm displacement of the grounded grid (GG) aperture, a result in agreement with experimental results of a 1/3-scale model for the LHD ion source. The max...

A large hydrogen negative-ion source was constructed as a prototype one for the negative-ion-based neutral beam injection (NBI) system in large helical device. The ion source is designed to produce 180 keV-40 A of the negative ion beam with a current density of 40 mA/cm2, and is a cesium-seeded volume production source characterized by the external magnetic filter. The arc chamber of multicusp bucket is rectangular of 35 cm×145 cm in cross section and 21 cm in depth. The accelerator is a three-grid single-stage one, and the total grid area is 25 cm×125 cm, which is divided into five sections. The ion source was installed on the negative-ion-based NBI test-stand, and using one section of the grid 5.5 A of the negative-ion beam was produced at a gas pressure of 1.8 mTorr corresponding to a current density of 27.5 mA/cm2. The extracted electron current is low of 50% of the negative ion current, and the acceleration efficiency, defined as the negative ion current divided by the acceleration drain current, is ...

This paper presents recent experimental results in the Large Helical Device (LHD), which is the world largest helical system (Rax~3.6 m, aav~0.6 m, Bax~2.75 T) and has capability of steady state operation because of the superconducting magnets. LHD confines the plasma with Tio~2 keV and ne ~0.8times1019 m-3 for 1905 sec using the ICRF heating. In this case, the total plasma heating energy was ~1.3 GJ. This indicates that high energy ions generated by ICRF is well confined. Large magnetic island in the vacuum field due to the error field is vanished in plasma with higher temperature or with higher beta, so the error field is not an issue in helical systems. The obtained averaged beta is more than 4.3 % in LHD. Since the beta is not limited by the MHD events or magnetic island but the beta is limited by the lack of heating power, further improvement of beta can be expected. Internal and external transport barrier have been also observed, and the electron temperature and the ion temperature exceed 10 keV in the improved confinement mode in LHD. Besides the potential of the disruption free steady stated operation, LHD experiment accumulates evidences so that fusion reactor becomes more realistic

Compact cesiated hollow cathodes (CHC) has been developed and studied in two large multicusp H− ion sources at NIFS, aiming the long lifetime cathodes for plasma production in the neutral beam injectors. Results on hollow cathodes and power supplies improvement are described. Basic characteristics of multicusp source discharge, driven by several CHCs were studied. Key issues about simultaneous CHCs operation in high current discharge mode are discussed. Each of advanced CHC delivered stable and quiet income with current up to 70A for 5sec to multicusp source discharge. The hollow cathode operation with no Cs feed was obtained by using 30%Xe+70%H2 gas mixture.

Recent progress in LHD N‐NBI system is described. The system was designed to provide 15 MW of port‐through beam power by three Beam Lines (BL‐1 to BL‐3) with 180 keV hydrogen for 10 sec. However, starting from 3.2 MW injection by two beam lines in 1998, their performance has not yet satisfied all the specifications at the same time. This is because several difficulties were found in large negative ion sources while the specifications were determined based on the R&D results of small ion sources. Hence, continuous efforts have been done to improve the performance of negative ion source during the machine maintenance period between the experimental campaigns of LHD. As a result, the injection power has been increased year by year, and in the latest LHD experimental campaign the total port‐through power of 13.1 MW was achieved, where R&D on the ion source of BL‐1 was most successful. Two improvements have been done on the ion source of BL‐1. One is the optimization of magnetic cusp structure by modifying the...

To understand H − /D − negative ion formation mechanism and behavior of H − /D − ions in volume negative ion sources, axial distributions of H − /D − densities in the source are measured directly by laser photodetachment method. Relationship between H − /D − production and plasma parameter control with magnetic filter (MF) is studied. H − and D − densities have different spatial distributions, respectively, corresponding to those different plasma parameter conditions.

The new beam accelerator with multi‐slot grounded grid (MSGG) has been developed to increase the port‐through power into large helical device (LHD). Using the accelerator, the maximum power of 5.7 MW was achieved at the beam energy of 186 keV in the beam injection to LHD plasma last year. Although the port‐through power increased compared with conventional accelerators with multi‐hole grounded grid (MHGG), the accelerator with the MSGG includes a disadvantage of bi‐focal condition in parallel and perpendicular direction to the long side of the slots. When the beam width in one of those directions gets narrower, the width in another direction becomes wider. This disadvantage includes the loss of beam port‐through power and induces internal damages in neutral beam line. In order to reduce the disadvantage, an experiment has been done using a small‐scaled negative ion source with racetrack‐shaped apertures for the steering grid installed at beam upstream of the MSGG. By applying the racetrack apertures to the accelerator, it is observed that the beam widths in the parallel and perpendicular directions to the slot long side have almost the same focal condition to obtain minimal beam widths.

A large vacuum‐immersed H− ion source has been operated on the negative‐ion‐based neutral beam teststand. The achieved level of the beam and the pulse duration in beam conditioning were limited by a high‐voltage breakdown in the vacuum vessel. A baffle plate at grounded potential for shielding completely from the charged particles was successful. A beam with an energy of up to 102 keV and the H− ion current of 0.55 A were achieved for 0.29 s without the breakdown. The ion current of ∼1.5 A was accelerated in cesium‐seeded operation. The corresponding ion current density was ∼7 mA/cm2. A magnetic filter (as Type I LV magnetic filter) on the plasma electrode was applied. The electron beam component which was extracted from the plasma source together with H− ions was found to be very reduced.

A large hydrogen negative‐ion source with an external magnetic filter has been developed for a neutral beam injection (NBI) system in the Large Helical Device (LHD), and a high‐energy and high‐current H− ion beam has been produced. The ion source is operated at a high arc efficiency of 0.1 A/kW at an operational gas pressure of less than 3.5 mTorr, and produces 47 keV–16.2 A of a H− ion beam from a grid area of 25 cm×50 cm. With two‐stage acceleration, 13.6 A of a H− ion beam has been successfully accelerated to 125 keV. Multibeamlet focusing by the aperture displacement technique has been achieved 11.2 m downstream with a gross divergence angle of 9 mrad. The alternate beamlet deflection by the magnetic field at the extraction grid, which results in beam broadening in the deflection direction, was well compensated also by the aperture displacement technique. These results satisfy the specification of the negative‐ion‐based LHD–NBI system.

A large rf‐driven hydrogen negative‐ion source has been developed for a neutral beam injection (NBI) system in the fusion application. The rf plasma generator is a metal‐walled multicusp bucket source, the dimensions of which are 30 cm×30 cm in cross section and 20 cm in depth. An induction coil antenna is fully immersed in the rf plasma generator, and is used for coupling the rf power into plasma. A strong external magnetic filter is generated in front of the plasma electrode for negative‐ion production, and a filter field strength of more than 1000 G cm is required to reduce the electron temperature in the extraction region as low as 1 eV. Precisely 5.5 mA of a negative‐ion current was obtained from a single aperture, corresponding to 4.1 mA/cm2 of current density. A small amount of cesium was supplied to the rf plasma generator from a cesium oven. As a result, a lowering of the operational gas pressure and improvement of the negative‐ion production efficiency were observed. The optimum bias voltage was...

The helicity-induced Alfv\'en eigenmodes (HAEs) with the toroidal mode number $n=2$ and 3 are observed for the first time in the Large Helical Device plasmas heated by neutral beam injection. The observed mode frequency is about 8 times higher than that of the observed toroidicity-induced Alfv\'en eigenmodes, and is proportional to the Alfv\'en velocity. The modes are excited when the ratio of the beam velocity to the Alfv\'en velocity exceeds about unity. The frequency lies just above the lower bound of the HAE gap in the plasma edge region of $\ensuremath{\rho}g0.7$ ($\ensuremath{\rho}$: normalized minor radius).

Sawtooth oscillations have been observed in current-carrying helical plasmas by using electron-cyclotron-emission diagnostics in the Large Helical Device. The plasma current, which is driven by neutral beam injection, reduces the beta threshold of the sawtooth oscillation. When the central q value is increased due to the plasma current, the core region crashes, and, when it is decreased, the edge region crashes annularly. Observed rapid mixture of the plasma in the limited region suggests that these sawtooth crashes are reconnection phenomena. Unlike previous experiments, no precursor oscillation has been observed.

Angle-resolved energy distribution functions of positive and negative hydrogen ions produced from a rough-finished Si surface under 1 keV proton irradiation have been measured. The corresponding distribution from a crystalline surface and a carbon surface are also measured for comparison. Intensities of positive and negative ions from the rough-finished Si are substantially smaller than those from crystalline Si. The angular distributions of these species are broader for rough surface than the crystalline surface. No significant temperature dependence for positive and negative ion intensities is observed for all samples in the temperature range from 300 to 400 K.

A spectrally selective imaging system has been developed to obtain a distribution of Hα emissions at the extraction region in a hydrogen negative ion source. The diagnostic system consisted of an aspherical lens, optical filters, a fiber image conduit, and a charge coupled device detector was installed on the 1/3-scaled hydrogen negative ion source in the National Institute for Fusion Science. The center of sight line passes beside the plasma grid (PG) surface with the distance of 11 mm, and the viewing angle has coverage 35 mm from the PG surface. Two dimensional Hα distribution in the range up to 20 mm from the PG surface was clearly observed. The reduction area for Hα emission caused by beam extraction was widely distributed in the extraction region near the PG surface.

The electron density was measured at various bias voltages to understand how the plasma grid bias affects the electron near the plasma grid in large-scaled negative ion sources. It was found that the response of the electron to the bias voltage changes depending on negative ion production processes. The electron density remarkably decreases with increasing the bias voltage in the pure-volume plasma. On the other hand, the electron density depends on the bias voltage weakly in the Cs-seeded plasma. In addition, it was observed that the response of the co-extracted electron current to the bias voltage has similar trend to that of the electron density.

It was observed that the vacuum magnetic island produced by an external error magnetic field in the large helical device shrank in the presence of plasma. This was evidenced by the disappearance of flat regions in the electron temperature profile obtained by Thomson scattering. This island behavior depended on the magnetic configuration in which the plasmas were produced.

In the Large Helical Device, stable discharges lasting longer than one minute have been obtained using the complete heating scheme, including ECH. The plasma is sustained with NBI or ICRF of 0.5-1 MW. The central plasma temperature is higher than 1.5 keV with a density of (1-2) x 10(19) m(-3) maintained until the end of the pulse. Full installation of the carbon divertor has contributed to this achievement. This provides a sufficient basis for physics and technology studies for the next campaign. Long pulse operation opens new possibilities in diagnostics and in physics studies. Higher accuracy and reliability are obtained with a diagnostics parameter scan, longer integration of signals or two dimensional measurement. The mechanism of a slow oscillation called 'breathing' is discussed. Hydrogen recycling analysis has been carried out and preliminary results have been obtained. On the basis of these results, the future programme is divided into two categories: (i) physics and technology experiments utilizing long pulse discharges lasting up to 5 minutes, and (ii) extension of the pulse length up to one hour.

Ion-cyclotron heating was applied to the Large Helical Device. When the proton-cyclotron resonance was near the saddle point of the magnetic field-strength plane, strong ion-cyclotron damping occurred. Under these conditions efficient plasma heating was achieved for more than one minute. A high-energy ion tail was observed, and the effective tail temperature was determined by a balance between the wave acceleration and the electron-drag relaxation. There was no apparent sign of particle orbit loss effect in the investigated density range of 0.8-1.3x10(19) m(-3).

Line spectra from intrinsic impurity ions have been monitored during the three kinds of long-pulse discharges (ICH, ECH, NBI). Constant emission from the iron impurity shows no preferential accumulation of iron ion during the long-pulse operations. Stable Doppler ion temperature has been also measured from Fe XX, C V and C III spectra.