K. Itami

Fast plasma shutdown by killer pellet injection in JT-60U with reduced heat flux on the divertor plate and avoiding runaway electron generation

A killer pellet is an impurity pellet that is injected into a tokamak plasma in order to terminate a discharge without causing serious damage to the tokamak machine. In JT-60U neon ice pellets have been injected into OH and NB heated plasmas and fast plasma shutdowns have been demonstrated without large vertical displacement. The heat pulse on the divertor plate has been greatly reduced by killer pellet injection (KPI), but a low-power heat flux tail with a long time duration is observed. The total energy on the divertor plate increases with longer heat flux tail, so it has been reduced by shortening the tail. Runaway electron (RE) generation has been observed just after KPI and/or in the later phase of the plasma current quench. However, RE generation has been avoided when large magnetic perturbations are excited. These experimental results clearly show that KPI is a credible fast shutdown method avoiding large vertical displacement, reducing heat flux on the divertor plate, and avoiding (or minimizing) RE generation.

The impact of ELMs on the ITER divertor

Abstract Edge-Localized-Modes (ELMs) are expected to present a significant transient flux of energy and particles to the ITER divertor. The threshold for ablation of the graphite target will be reached if the ELM transient exceeds Q/t1/2xa0∼xa045 MJ m−2 s−1/2 where Q is the ELM deposition energy density and t the ELM deposition time. The ablation parameter in ITER can be determined by scaling four factors from present experiments: the ELM energy loss from the core plasma, the fraction of ELM energy deposited on the divertor target, the area of the ELM profile onto the target, and finally the time for the ELM deposition. Review of the ELM energy loss of Type I ELM data suggests an ITER ELM energy loss of 2–6% of the stored energy or 25–80 MJ. The fraction of heating power crossing the separatrix due to ELMs is nearly constant (20–40%) resulting in an inverse relationship between ELM amplitude and frequency. Measurements on DIII-D and ASDEX-Upgrade indicate that 50–80% of the ELM energy is deposited on the target. There is currently no evidence for a large fraction of the ELM energy being dissipated through radiation. Profiles of the ELM heat flux are typically 1–2 times the width of steady heat flux between ELMs, with the ELM amplitude usually larger on the inboard target. The ELM deposition time varies from about 0.1 ms in JET to as high as 1.0 ms in ASDEX-Upgrade and DIII-D. The ELM deposition time for ITER will depend upon the level of conductive versus convective transport determined by the ratio of energy to particles released by the ELM. Preliminary analysis suggests that large Type I ELMs for low recycling H-mode may exceed the ablation parameter by a factor of 5. Promising regimes with smaller ELMS have been found at other edge operational regimes, including high density with gas puffing, use of RF heating and operation with Type III ELMs.

Heat and particle transport of SOL and divertor plasmas in the W shaped divertor on JT-60U

The geometry effects of the W shaped divertor on the divertor plasma were investigated quantitatively. The ion flux was increased near the divertor strike point, which is effective for reducing the local electron temperature and decreasing the onset e of divertor detachment. The plasma profile and parallel plasma flow in the scrape-off layer were systematically measured using reciprocating Mach probes installed at the midplane and the divertor X point. For the ion ∇B drift direction towards the divertor, `flow reversal was observed at the midplane. A quantitative evaluation of the parallel plasma flow suggesting that the flow is produced in a torus to keep the pressure constant along the field lines was consistent with the measurements.

Measurement of the chemical sputtering yields of CH4/CD4 and C2Hx/C2Dx at the carbon divertor plates of JT-60U

The chemical sputtering yields of CH4/CD4 and C2Hx/C2Dx have been measured at the divertor plates of JT-60U. Spectroscopic measurements for CH/CD and C2 spectral bands are applied to estimate the CH4/CD4 and the C2Hx/C2Dx flux. At the surface temperatures of 380, 440 and 560 K, the CH4 yield is, respectively, ~0.8%, 1-2% and 2-3%, the C2Hx yield 1-2%, 3-4% and 4-5%, and the total sputtering yield by hydrogen ions 3-4%, ~8% and ~10%. With increasing ion flux to the divertor plates (Γion), the sputtering yields (Y) decrease, i.e. Y∝Γion(-0.05 to -0.40). With increasing electron temperature (Te), the sputtering yields increase, i.e. Y∝Te0.5. It is concluded from the result of regression analysis of Y∝Te0.5 that the negative dependence of the yields on the ion flux is attributed to the incident ion energies to the carbon plates. The ratio of the sputtering yields by deuterium ions and hydrogen ions is estimated to be ≥1.5 based on the ion flux measurement by Hα/Dα intensity. The C2Hx/C2Dx sputtering yield accounts for ~80% of the total number of sputtered carbon atoms.

Role of divertor geometry on detachment and core plasma performance in JT60U

Experimental results related to the divertor geometry such as divertor plasma detachment, neutral transport and plasma energy confinement, were compared in the open and W-shaped divertors. The ion flux near the outer strike point was larger than in the open divertor, and the electron temperature at the target, T e div , was reduced. Divertor detachment and x-point MARFEs occurred at n e 10-20% lower than that for the open divertor. Although the leakage of neutrals from the divertor to the main chamber decreased, a neutral source in the main chamber due to an interaction of the outer scrape-off layer (SOL) plasma to the baffle plates became dominant above the baffle. Degradation in the enhancement factor of the energy confinement was observed similarly in the open and W-shaped divertors. The neutral density inside the separatrix was estimated to be a factor of 2-3 smaller, which did not affect the energy confinement.

Impurity and particle recycling reduction by boronization in JT-60U

Abstract In JT-60U boronization using decaborane was carried out. Boronization reduced the oxygen impurity in OH discharges and shortened the wall conditioning after the vacuum vessel vent and consequently enabled JT-60U to produce clean plasmas easily except for NB heated plasmas. After boronization, particle recycling was reduced drastically in OH and NB discharges. High confinement plasmas were obtained including high β p mode and H-mode discharges. In the latest boronization part of divertor plates were replaced with B 4 C coated tiles with a B 4 C thickness ∼ 300 μm. After introducing B 4 C divertor tiles, an explosive generation of boron particles from the tiles was observed. By the combined effects of boronization with decaborane and boron generation from B 4 C tiles, oxygen impurity was so low that oxygen line signals were reduced to noise levels after the latest boronization. On the other hand, boron burst from the B 4 C tiles restricted the operation of JT-60U.

SOL plasma profiles under radiative and detached divertor conditions in JT-60U

Abstract Radial profiles of electron density ne,mid, temperature Te,mid, and ion temperature Ti,mid in the scrape-off layer (SOL) were investigated under radiative and detached divertor conditions in L-mode discharges. Since the ratio of Ti,mid/Te,mid is about 3, the ion pressure is dominant at the midplane, and plays an important role in the pressure balance between the midplane and the divertor targets. Effect of the connection length on the decay lengths of ne,mid and Te,mid, λT|staggered|e and λn|staggered|e, was determined in two SOL regions. At the same time, λT|staggered|i, was compared to λTe and λn|staggered|e, λn|staggered|e, λTe, and λT|staggered|i increase with the connection length. During the X-point MARFE, λTe, λn|staggered|e, and λT|staggered|i increase substantially with a reduction in Te,mid, Ti,mid and ne,mid at the plasma edge and in the first SOL, due to the penetration of the maximum radiation region into the main plasma near the X-point.

Transport analysis of divertor plasma in JT-60U

A simple divertor model has been developed in order to analyze the divertor transport based on experimental data. The fluid equations are solved numerically in one dimension along the magnetic field lines. In contrast to the conventional divertor code, the boundary conditions are given by experimentally determined plasma parameters at the divertor plate, which are measured by Langmuir probes. The neutral particle transport is simulated in two dimensions by a Monte Carlo code. The interaction between the plasma and the neutral particles is solved self-consistently using an iterative procedure; the plasma parameters converge in a few iterations. The little computational time needed facilitates a systematic analysis of the divertor transport. By applying this model to the JT-60U divertor plasmas, the particle confinement time in the main plasma and the heat diffusivity in the scrape-off layer are evaluated.

High radiation and high density experiments in JT-60U

In order to obtain improved confinement plasmas with high radiation at high density, Ar gas was injected into ELMy H mode plasmas in JT-60U. A confinement improvement of HH98(y,2) ≈ 1 was obtained with a high radiation loss power fraction (~80%) at an electron density of ~0.65nGW. The HH factor was about 50% higher than that in plasmas without Ar injection.

Field reversal effects on divertor plasmas under radiative and detached conditions in JT-60U

Reversal effects of the toroidal field Bt on the principal divertor plasma parameters were investigated under radiative and detached divertor conditions in L mode discharges. The ion flux to the inboard separatrix strike point decreased before a MARFE occurred, irrespective of the ion Del B drift direction. The local electron temperature, Te, div, decreased to around 10 eV. The maximum fraction of power radiated in the divertor was comparable between the two directions of Bt. With the power flowing into the two divertor fans being slightly larger on the outboard than on the inboard, a nearly symmetric in-out heat load was observed for the ion Del B drift away from the target. This was due to the outboard enhanced asymmetries in the particle flux and radiation loss distributions. From the viewpoint of in-out symmetry in the target heat load and Te, div, operation with the ion Del B drift away from the target plate is desirable as long as the attached divertor condition is maintained. On the contrary, during the MARFE, although deterioration of the energy confinement as well as the increase in the fuelling efficiency were comparable, for the ion Del B drift towards the target the plasma did not detach completely, and the strong in-out asymmetry in the particle recycling was relaxed to a relatively symmetric distribution. From the viewpoint of particle exhaust to the divertor, operation with the ion Del B drift towards the target is favourable