T. Oishi

Effective screening of iron impurities in the ergodic layer of the Large Helical Device with a metallic first wall

In the Large Helical Device (LHD) operated with a metallic (stainless steel) first wall, it is found that the iron density, nFe, at the plasma core is fairly low (nFexa0⩽xa0108xa0cm−3) in general neutral beam (NB)-heated discharges, while the iron quickly increases with the appearance of impurity accumulation when a multi-hydrogen ice pellet is injected or the NB input power is largely reduced. Although the highest iron density (nFexa0⩽xa01010xa0cm−3) at the plasma centre in the LHD is observed from such discharges, it suggests a still low iron concentration (nFe/nexa0<xa010−3). Therefore, the edge iron transport in the ergodic layer, which determines the iron influx to the core plasma, is studied to clarify why the iron density in the core plasma is low. A line ratio of Fexa0XV located in the vicinity of the last closed flux surface to Fexa0VIII (or Fexa0IX) located in the ergodic layer decreases with density. The two-dimensional (2D) edge iron emission of Fexa0XVI and Fexa0IX is enhanced in the vicinity of the X-point with a larger number of magnetic field lines directly connected to divertor plates, which suggests that iron ions from the first wall move downstream. The density of edge Fe15+ ions giving the iron influx to the core plasma is analysed with the 2D distribution. The analysis also shows that the iron influx to the core plasma decreases with density. These results clearly indicate that the screening effect developed in the ergodic layer works well for iron ions coming from the first wall. A three-dimensional edge transport simulation with EMC3-EIRENE can also predict an effective impurity screening for heavy impurities compared to light impurities.

EMC3-EIRENE modelling of edge impurity transport in the stochastic layer of the large helical device compared with extreme ultraviolet emission measurements

The transport properties and line emissions of carbon impurity in the stochastic layer of the Large Helical Device have been investigated with the 3D edge transport code EMC3-EIRENE. A parameter study has been performed to examine the sensitivity of the simulation results on each transport term in the impurity transport model and the impurity source characteristics, i.e. the source amount and the location. The modelling has revealed that in order to reproduce the experimental results of the emission distribution, the impurity perpendicular transport coefficient (D imp) and the first wall source play important roles, while changes to the ion thermal and the friction forces are rather irrelevant. The detailed study of flux tube tracing and magnetic field structure in the edge stochastic layer, in relation to impurity transport, has shown that the deeper penetration of impurity into the higher plasma density region due to the enhanced D imp and the first wall source is responsible for the change of emission pattern as well as the intensity. The analysis indicates that D imp might be larger than that of background plasma by a few factors and also that there probably exists a substantial amount of first wall impurity source.

A study of tungsten spectra using large helical device and compact electron beam ion trap in NIFS

Tungsten spectra have been observed from Large Helical Device (LHD) and Compact electron Beam Ion Trap (CoBIT) in wavelength ranges of visible to EUV. The EUV spectra with unresolved transition array (UTA), e.g., 6g-4f, 5g-4f, 5f-4d and 5p-4d transitions for W+24-+33, measured from LHD plasmas are compared with those measured from CoBIT with monoenergetic electron beam (≤2keV). The tungsten spectra from LHD are well analyzed based on the knowledge from CoBIT tungsten spectra. The C-R model code has been developed to explain the UTA spectra in details. Radial profiles of EUV spectra from highly ionized tungsten ions have been measured and analyzed by impurity transport simulation code with ADPAK atomic database code to examine the ionization balance determined by ionization and recombination rate coefficients. As the first trial, analysis of the tungsten density in LHD plasmas is attempted from radial profile of Zn-like WXLV (W44+) 4p-4s transition at 60.9A based on the emission rate coefficient calculated...

Performance improvement of two-dimensional EUV spectroscopy based on high-frame-rate CCD and signal normalization method in Large Helical Device

In the Large Helical Device (LHD), two-dimensional (2D) distribution of edge impurity line emissions from the ergodic layer with a three-dimensional (3D) magnetic field structure has been observed in the wavelength range of 30–650 A by horizontally scanning a space-resolved extreme ultraviolet (EUV) spectrometer during a steady phase in electron cyclotron heating (ECH) and ion cyclotron range of heating (ICRF) discharges with long pulse duration (τd ≥ 10 s). The 2D distribution of impurity line emissions can be measured at the upper or lower half in the full vertical range of LHD plasmas by a single horizontal scan. The performance of 2D EUV spectroscopy, however, was not sufficient for discussing the impurity transport in the ergodic layer from the viewpoints of the quality of 2D images and the requirement of long pulse discharges. The 2D EUV spectrometer is therefore modified by installing a high-frame-rate charge-coupled detector (CCD) and increasing its horizontal scanning speed. As a result, the quality of 2D images is significantly improved with increased horizontal spatial resolution, and then the 2D measurement is also possible for neutral beam injection (NBI) discharges with short pulse duration (τd ~ 3 s). In addition, temporal and shot-by-shot intensity variations in the edge EUV emission are significantly reduced by applying a signal normalization method using another EUV spectrometer with high time resolution. A 0.5 µm poly(ethylene terephthalate) (PET) filter is also installed in front of an entrance slit of the space-resolved EUV spectrometer. The spike noise originating from high-energy neutral particles, which are enhanced in low-density NBI discharges, has been successfully reduced. Typical examples of the 2D distribution of impurity line emissions with improved quality are presented for several impurity species.

Coaxial pellets for metallic impurity injection on the large helical device.

Two coaxial pellets with tungsten inserted into graphite carbon and polyethylene (PE) tubes are compared for tungsten spectroscopic study in the Large Helical Device. The tungsten pellet with carbon tube causes plasma collapse, while that with PE tube smoothly ablates without collapse. The deposition profile of the pellets is analyzed with a help of pellet ablation spectroscopy. It is found that the tungsten pellet with carbon tube can significantly penetrate into the core plasma and leads to the plasma collapse. A tungsten spectrum with radial profile is successfully observed when the tungsten pellet with PE tube is used.

Observation of W IV–W VII line emissions in wavelength range of 495–1475 Å in the large helical device

Vacuum ultraviolet spectra of line emissions from tungsten ions at lower ionization stages have been measured in the large helical device (LHD) using a high-resolution 3 m normal incidence spectrometer in the wavelength range of 495–1475 A. Tungsten was introduced in the LHD plasma by injecting a coaxial tungsten impurity pellet. Many tungsten lines of W IV–W VII were successfully observed in low-temperature plasmas just after the tungsten pellet injection. It is found that some W VI lines are emitted with extremely high intensity and entirely isolated from other intrinsic impurity lines, in particular, W VI at 605.926 A (5d–6p), 639.683 A (5d–6p), 677.722 A (5d–6p), 1168.151 A (6s–6p) and 1467.959 A (6s–6p). The result strongly suggests that those lines may be useful for the spectroscopic study in ITER and other magnetic fusion devices with tungsten materials as the plasma facing component. The ion temperature was also measured from Doppler broadening of W V and W VI lines. The result indicates that the measured ion temperature is clearly higher than the ionization energy of such ions. The reason is discussed with regarding to the pellet injection.

Vertical profiles and two-dimensional distributions of carbon line emissions from C2+−C5+ ions in attached and RMP-assisted detached plasmas of large helical device

In Large Helical Device (LHD), the detached plasma is obtained without external impurity gas feed by supplying an m/nu2009=u20091/1 resonant magnetic perturbation (RMP) field to a plasma with an outwardly shifted plasma axis position of Raxu2009=u20093.90u2009m where the magnetic resonance exists in the stochastic magnetic field layer outside the last closed flux surface. The plasma detachment is triggered by the appearance of an m/nu2009=u20091/1 island when the density, increased using hydrogen gas feed, exceeds a threshold density. The behavior of intrinsically existing impurities, in particular, carbon originating in the graphite divertor plates, is one of the important key issues to clarify the characteristic features of the RMP-assisted plasma detachment although the particle flux still remains on some divertor plates even in the detachment phase of the discharge. For this purpose, vertical profiles and two-dimensional (2-D) distributions of edge carbon emissions of CIII to CVI have been measured at extreme ultraviolet wavelen...

Up-down asymmetry measurement of tungsten distribution in large helical device using two extreme ultraviolet (EUV) spectrometers

Two space-resolved extreme ultraviolet spectrometers working in wavelength ranges of 10-130 Å and 30-500 Å have been utilized to observe the full vertical profile of tungsten line emissions by simultaneously measuring upper- and lower-half plasmas of LHD, respectively. The radial profile of local emissivity is reconstructed from the measured vertical profile in the overlapped wavelength range of 30-130 Å and the up-down asymmetry is examined against the local emissivity profiles of WXXVIII in the unresolved transition array spectrum. The result shows a nearly symmetric profile, suggesting a good availability in the present diagnostic method for the impurity asymmetry study.

Line spectrum and ion temperature measurements from tungsten ions at low ionization stages in large helical device based on vacuum ultraviolet spectroscopy in wavelength range of 500-2200 Å.

Vacuum ultraviolet spectra of emissions released from tungsten ions at lower ionization stages were measured in the Large Helical Device (LHD) in the wavelength range of 500-2200 Å using a 3 m normal incidence spectrometer. Tungsten ions were distributed in the LHD plasma by injecting a pellet consisting of a small piece of tungsten metal and polyethylene tube. Many lines having different wavelengths from intrinsic impurity ions were observed just after the tungsten pellet injection. Doppler broadening of a tungsten candidate line was successfully measured and the ion temperature was obtained.

Effect of neutron and γ-ray on charge-coupled device for vacuum/extreme ultraviolet spectroscopy in deuterium discharges of large helical device

A charge-coupled device (CCD) is widely used as a detector of vacuum spectrometers in fusion devices. Recently, a deuterium plasma experiment has been initiated in a Large Helical Device (LHD). Totally 3.7 × 1018 neutrons have been yielded with energies of 2.45 MeV (D-D) and 14.1 MeV (D-T) during the deuterium experiment over four months. Meanwhile, γ-rays are radiated from plasma facing components and laboratory structural materials in a wide energy range, i.e., 0.01-12.0 MeV, through the neutron capture. It is well known that these neutrons and γ-rays bring serious problems to the CCD system. Then, several CCDs of vacuum ultraviolet/extreme ultraviolet/X-ray spectrometers installed at different locations on LHD for measurements of spectra and spatial profiles of impurity emission lines are examined to study the effect of neutrons and γ-rays. An additional CCD placed in a special shielding box made of 10 cm thick polyethylene contained 10% boron and 1.5 cm thick lead is also used for the detailed analysis. As a result, it is found that the CCD has no damage in the present neutron yield of LHD, while the background noise integrated for all pixels of CCD largely increases, i.e., 1-3 × 108 counts/s. The data analysis of CCD in the shielding box shows that the background noise caused by the γ-ray is smaller than that caused by the neutron, i.e., 41% from γ-rays and 59% from neutrons. It is also found that the noise can be partly removed by an accumulation of CCD frames or software programming.