Published papers

SONIC divertor code facilitates simultaneous calculations of seeding impurity (Ar) and fusion product (He) transport. D2 and Ar gases were injected at the outer midplane and upper location of the outer divertor, respectively. He exhausts in the detached divertor have been investigated for a reference series of JA DEMO with an increasing deuterium gas puff rate (50–200 Pa m3s−1) where Psep ∼ 250 MW and the radiation fraction is 0.8. The outer divertor heat load (qtotdiv) peaked in the attached region on the partially detached target, and the peak-qtotdiv was reduced to lower than 10 MW m−2 by increasing the D2 puff rate while maintaining a relatively low nesep (2 × 1019 – 2.5 × 1019m−3). Neutral and gas pressures in the divertor (PD0div, PD2div) were evaluated at exhaust slots of the dome and in the sub-divertor. The total pressure (PDtotdiv) in the D-puff scan was increased from 1.4 to 3 Pa at the outer exhaust slots, and the outer peak-qtotdiv was proportional to (PDtotdiv)−0.7. Plasma flow reversals with M// = 0.2 – 0.3 were generated near the detach–attach boundary above the outer divertor target. The influence of the flow reversal on the plasma and impurity transport was determined. Convective heat flux was reversed, which significantly affected the ion heat flux profile in the upstream SOL. The flow reversal reduced the impurity shielding and radiation loss in the outer divertor. Inner-enhanced in–out asymmetry of Ar concentration (cAr) was large near the separatrix in SOL, suggesting that Ar ions were transported from the inner SOL to the confined plasma. In the confined plasma (edge), both cAredge and cHeedge were reduced and poloidally uniform, where cAredge was reduced to the marginal level and cHeedge was below the design reference by system code assumption (cArmain: 0.6%, cHemain: 7%). Consequently, these SONIC results are generally consistent with the JA DEMO higher-κ design, and larger D-puff (and smaller ГAr,puff) operation is preferable for the large power exhaust in the divertor and the impurity control in the edge plasma.

A multi-ion-temperature model for hydrogen plasmas has been newly developed and implemented in the linear divertor plasma simulation code LINDA-NU to investigate detachment mechanisms involving atomic and molecular processes. In this model, H+, H2+, and H3+ ions are treated as separate fluids with independent temperatures, enabling a detailed assessment of ion-species-dependent thermal and momentum dynamics in detached plasmas. Plasma detachment transitions were studied by scanning atomic and molecular neutral densities. An increase in atomic hydrogen density leads to electron–ion recombination-dominated detachment, where charge exchange with atomic neutrals governs momentum loss. In contrast, increasing molecular hydrogen density results in molecular-activated recombination (MAR)-dominated detachment, in which molecular ions become quantitatively significant and strongly influence particle, momentum, and energy balances. A comparison with a simplified single-ion-temperature model demonstrates that the multi-ion-temperature treatment is essential for accurately describing the MAR-dominated regime, where multi-ion temperatures significantly modify molecular reaction rates, leading to qualitative differences in plasma profiles.

Thermal instability of divertor plasma transport in steady-state operation is a critical issue in reactor design. Identifying the thermally stable regime is crucial for the successful operation of a fusion reactor. This study aims to explore the regime of stable operation for divertor plasma and to investigate how this stability depends on key engineering parameters. For this purpose, a linear stability analysis is applied to equilibrium solutions of the particle, momentum, and energy balance equations obtained from the core-SOL-divertor model, providing a novel application of point-model-based analysis. In this regard, this study specifically addresses plasma heat control via radiation within the divertor volume. While alternative heat exhaust strategies utilizing strong radiation localized near the X-point—such as the X-point radiator regime—are also highly important, they are outside the scope of this work. Under this framework, this study presents two main findings through model calculations for JT-60U and JA-DEMO. First, the divertor plasma exhibits three thermally stable states: low-recycling, high-recycling, and weakly detached, while the intermediate states between them are unstable. Within the context of the localized divertor-volume radiation, the weakly detached state is identified as the expected operational regime in future fusion reactors. Second, as initial examples demonstrating the capability of the stability analysis framework for future extensive parametric surveys, the stability of the weakly detached state is found to be primarily sensitive to the exhaust power; increasing power raises the lower temperature limit and significantly narrows the stable range. Variations in the major radius have little influence on stability, whereas a reduction in the minor radius slightly raises the lower stability limit within the parameter range examined. The proposed stability evaluation framework provides qualitative guidance for selecting reactor design parameters that expand the stable operational window of divertor plasmas.

In this research, we investigate the effect of magnetic flux tube expansion on the divertor plasma parameters by using the fluid code “LINDA”. A comparison between the cylindrical flux tube (without the magnetic flux expansion) and the expansion magnetic flux tube has been undertaken. The aim of the study is to understand the impact of magnetic field expansion on the divertor physics by using the LINDA fluid code. The plasma density (ni) is decreased and parallel velocity is increased (ui||) toward the target plate with the expansion of magnetic field lines near the target plate. The heat and particle fluxes are reduced significantly on the target plate in the case of the expansion mesh configuration. For the case of cylindrical mesh, advection becomes stronger with the decreasing distance from the target plate. In the case of expansion mesh, diffusion is stronger with the decreasing distance from the target plate. These outcomes clearly indicate the effect of the magnetic field structure on the divertor plasma parameters.

Divertor heat loads are one of the most significant issues affecting fusion reactors. Atomic processes play a crucial role in reduction of the divertor heat load. Notably, elastic scattering between ions and neutral particles can be characterized as large-angle scattering. A large fraction of ion energy is transferred to neutral particles, and the ion direction can be significantly changed by a single large-angle scattering event. In abundant neutral particle regions such as divertor plasmas, the large-angle elastic scattering results in additional ion transport perpendicular to magnetic field lines. Effect of the additional ion transport is expected to be significant at low magnetic field strength and long Larmor radii, such as in a case of advanced divertors (e.g., Super-X and Snowflake divertors). In this study, we investigated the effect of the large-angle elastic scattering at low magnetic field strength and long divertor legs with reference to advanced divertor configurations using an orbital calculation. The large-angle elastic scattering transport is seen to cause a spread in density profiles and a reduction of heat flux. The results of this study show that for the short (long) leg divertor configuration like JT-60U (advanced divertor), the peak heat flux is reduced by around 15% (21%) when the magnetic field strength is 0.5 T in comparison to the model that assumes no guiding center movement due to the elastic scattering. It is also shown that the assumption of isotropic elastic scattering with neutral particles leads to excessive suppression of ion flows.

For JA DEMO, one of the grand challenges is efficiently handling the extreme power loads onto plasma-facing components, particularly the divertor targets. Impurity seeding is an essential strategy for increasing the radiation losses and thereby reducing the heat load on the targets. A critical remaining issue is the trade-off between the retention of radiative impurities in the divertor region to maximize the radiation loss, and their leakage to the upstream region, which degrades the core plasma confinement. To address this issue, based on the current baseline divertor geometry of JA DEMO, the SOLPS-ITER code is applied for the first time to evaluate the divertor power exhaust and Ar impurity distribution across various Ar seeding levels in detail. The simulation results show that increasing the Ar seeding level is highly beneficial for reducing the heat load on the targets. Moreover, in the near SOL there is a strong flow reversal of the main D ions and Ar impurity ions from the lower outer divertor target toward the upstream. This flow reversal promotes the transport of Ar ions out of the lower outer divertor region, leading to significant core contamination. As a result, such leakage severely constrains the viability of using high seeding levels for heat-load control in the current JA DEMO divertor geometry and plasma operating scenario.

Plasma boundary codes are widely used to interpret and predict SOL/divertor plasmas. The comparison of results obtained by these codes is one of the options to improve the SOL/divertor plasma modelling by defining possible limitations and enhancing the fidelity of future simulations. The aim of this study is to validate the transport models of SOL/divertor transport codes SOLPS-ITER and SONIC by terms-to-terms comparison, from the viewpoint of physical explanation rather than the numerical viewpoint. In this study, the comparison of the results between SOLPS-ITER and SONIC on pure-D plasma conditions with attached and partially detached plasmas in JT-60SA has been performed as a first step of the benchmarking activity. The calculation conditions such as computational grids, boundary conditions, and atomic/molecule database, are set to be as close as possible. In summary, the comparison can be concluded that no substantial difference can be seen although some differences in the plasma parameters have been observed. The main difference can be seen as the higher ratio of pressure drop in SOLPS-ITER than SONIC in front of the divertors in the partially detached case. It was found that the difference can be understood by the following process. The initial difference originated from the electron temperature and ion temperature in the vicinity of divertor plates. The electron temperature in SONIC near outer strike point is slightly higher than that of SOLPS-ITER. The difference in the electron temperature results in broader D+ particle source by ionization of neutrals and higher D+ ion density. Higher D+ ion and neutral densities in SOLPS-ITER leads to a larger momentum loss by charge exchange and thus the plasma loses its pressure. This result suggests that the small difference in the initial plasma parameters can be propagated by the interaction between neutrals, which affects the estimation of degree of detachment.

Understanding hydrogen trapping in tungsten is crucial for accurate modeling of plasma-wall interactions in fusion devices. In this study, we developed a deep learning model to predict hydrogen trapping sites, which serve as essential input parameters for kinetic Monte Carlo (kMC) simulations. The model employs a U-Net–based convolutional neural network that directly maps three-dimensional potential energy distributions to trapping site positions. Training data were generated from atomistic calculations using the embedded atom method (EAM) potential, and ground-truth trapping sites were systematically identified by force relaxation. The trained model achieved an F1 score of approximately 0.76, with most predicted sites coinciding with the true minima within ± 1 voxels (1 voxel = 0.1 Å per side). Visual comparisons confirmed the ability of the model to capture both global and local features of the potential energy landscape. In terms of efficiency, the proposed approach reduced prediction time by more than three orders of magnitude compared with conventional force-relaxation searches, enabling predictions in less than one second on a GPU. These results demonstrate that deep learning provides an accurate and computationally efficient method for identifying trapping sites. Future extensions include incorporating more complex defect structures and integrating the model into molecular dynamics-kMC hybrid frameworks for large-scale plasma-wall interaction simulations.

The so-called “ion-ion” plasma, which consists primarily of negative and positive ions, has been observed near the extraction aperture in negative hydrogen ion (H−) sources with a substantial amount of surface-produced negative ions, according to several experiments. In this study, to elucidate the characteristics of ion-ion plasmas with substantial surface H− production, the relationship between surface-produced negative ion density and electric potential is investigated using three-dimensional Particle-In-Cell simulations with the KEIO-BFX code. The results reveal that the plasma in a negative ion source can be classified into four distinct regions based on the relationship between H− density and electric potential: Region (I), where H− follows the Boltzmann relation; Region (II), where H+ follows the Boltzmann relation; Region (III), a transition region between the plasma region and the beam acceleration region; and Region (IV), the beam acceleration region. In negative ion sources with surface-produced H− ions, plotting the H− density against the electric potential offers a useful method for identifying the plasma meniscus, which is crucial for optimizing extracted beam quality and ensuring proper operation of beam extraction systems.

Handling the heat load on the divertor target is a critical issue in the design of fusion reactors. Estimating the heat load, supported by computationally efficient simulations with varying design parameters, provides valuable input to high-level design discussions and the development of design guidelines. This study aims to estimate the spatial distribution of the heat load on the target at a low computational cost. The Core-SOL-Divertor (CSD) model is employed to provide the integrated heat load on the target. The CSD output forms the basis of an analytical approach to solving the heat transport equation, which yields the spatial distribution of the heat load on the target. The resulting heat load distribution shows good agreement with SOLPS-ITER. The predicted distribution and its dependence on design parameters provide useful guidance for the preliminary design of fusion reactors.

In divertor plasmas, ion–atom collisions critically influence plasma transport and heat flux to divertor targets. Fluid neutral models are widely used in divertor simulations due to their low computational cost and favorable convergence properties. However, source terms representing ion–atom collisions often differ across codes because of variations in the cross section data and in the physical approaches used to model collisions between hydrogen isotope ions and atoms of identical nuclei. This study systematically evaluates these differences and establishes more consistent modeling approaches based on the Boltzmann collision integral (BCI). Using the BCI, we investigate particle transport in velocity space and assess the effect of scattering angles and differential cross sections (DCSs). Several analytical expressions—rarely used in divertor studies—are reviewed. For energy transfer, we derive a new BCI-based expression and validate the analytical expressions by comparing with the direct BCI evaluation. It is also pointed out that the commonly used charge–exchange (CX) cross section data no longer reflect recent experimental and theoretical developments. For deuteron-deuterium atom collisions at T = 10 eV, the CX-based model overestimates the friction force by more than 30% compared with the BCI-based model. For energy transfer between deuterons at T = 20 eV and deuterium atoms at T = 10 eV, the CX-based model overestimates the transferred energy by approximately 25%. The use of DCSs and the analytical expressions introduced in this study is therefore recommended. Implementing them in place of the conventional CX-based model may modify the predicted plasma profiles and the assessment of the detached divertor.

CSD (Core, SOL and Divertor) models sampling a fusion device with three points (Core, SOL and Divertor) aim at understanding fusion plasmas in a wide range of parameter space. Conventional CSD models estimate interactions among plasmas, neutrals and impurities, which are incorporated as loss factors in basic equations in models, with the use of physical quantities of plasmas at three sampled points. Physical quantities are assumed to be uniform in each of the three regions. Neglecting plasma spatial distributions in a divertor prevents the models from accurately evaluating loss factors at low temperatures. This is because loss factors strongly depend on the location-dependent temperature when the temperature is low. Consequently, conventional CSD models have difficulties in reproduction of the detachment. This research increases sampling points for the calculation of loss factors in a divertor. Loss factors at each point are integrated in the volume, then effective loss factors (ELF) for the divertor are obtained. Conventional loss factors in transport equations are replaced with ELFs for an improved CSD model proposed in this study. As a result, the plasma detachment is successfully reproduced in the CSD model without changing the equations. A CSD model for qualitative analyses with low computational costs has been developed to study plasma detachment.

A detached plasma integrated simulation code with various elastic/inelastic reactions (DISCOVER) was developed for better understanding of detached helium plasma experiments. DISCOVER, which includes a neutral transport code, fluid code, and collisional radiative code, considers elastic and inelastic collisions between recycling neutrals and bulk particles, as well as the transport of metastable atoms [He( 21S) and He( 23S)]. The code was used for simulating helium plasma in the linear plasma device NAGDIS-II. The transition from attached to detached plasma was demonstrated in the simulation by increasing the upstream electron density through an increase in the gas supply from the plasma source while maintaining a fixed downstream gas injection rate. In addition, similar transition was also observed by introducing neutral particles from downstream, while maintaining a fixed upstream electron density. In the detached plasma, the density of He( 23S) near the end target was approximately seven times higher than that in the attached plasma. Furthermore, it was evident that the transport of He( 23S) enhanced the electron energy rate coefficient and influenced the spatial distribution of the detached plasma.

It is reported from the measurement of a negative ion beamlet emittance that the characteristic three-Gaussian components are observed in the emittance diagram. The negative ion trajectories are calculated using 3D PIC-MCC simulation in order to clarify the origin of this complicated phase space structure and underlying physical mechanism. The characteristic three-Gaussian components can be reproduced in the emittance diagram from the simulation result. It is verified that the Gaussian components are caused by the negative ions extracted from the different plasma meniscus region, that is, the central region or the region near the edges of the meniscus.

The dynamics of the screening effect of Ar impurity by the injection of additional Ne has been studied through time-dependent analysis with the integrated divertor code SONIC. In the preceding study (Yamoto et al 2020 Plasma Phys. Control. Fusion 62 045006), the predictive simulation of JT-60SA plasma by SONIC has shown that the injection of additional Ne into Ar-seeded plasma results in lower Ar density and radiation power in the SOL and core edge than in the Ar-only seeded case. The results have demonstrated that the mixed impurity seeding of Ar and Ne may be advantageous for maintaining a high core plasma performance with a low divertor heat load. It was found that the friction force induced by the high D+ flow in the SOL towards the inner divertor (ID) region in the Ar + Ne seeded case pushes Ar impurities to the ID. However, the dynamics of D+ flow acceleration cannot be interpreted in the previous study because SONIC was a steady state code. In this study, we have developed the time-dependent version of SONIC and applied it to the transient analysis of the injection of additional Ne into Ar-seeded plasma in JT-60SA. When additional Ne is injected, Ne ions stay in the ID plasma near the X-point. As a result, the Ne radiation power increases near the X-point. The electron pressure then decreases due to the radiation cooling and the D+ flow is accelerated by the electron pressure gradient. The ion pressure also decreases due to the convection by the accelerated D+ flow by electron pressure gradient. The resulting ion pressure gradient further accelerates the D+ flow velocity towards the ID. The results suggest that both the high-performance core plasma and the low divertor heat load can be achieved by the Ar + Ne mixed impurity seeding.

In general, a quality of negative ion beam optics is determined by the shape of the plasma meniscus, which is an ion emitting surface. In this study, the key parameters to control the plasma meniscus and relevant physical structure in the electronegative plasma including the surface produced H- ions is investigated by using PIC-MCC simulation. The region from the source plasma up to the accelerator is modeled for a single aperture, and thus, the plasma meniscus can be obtained self-consistently. Parameter survey for the net flux of surface produced H- ion is conducted by varing the numbers of the surface produced H- ion super-particles per timestep in order to investigate the effcet of the electronegativity, that is, the ratio of the negative ion density to the electron density. It is shown that the effective distance d eff, between the plasma meniscus and the extraction grid depends on the electronegativity as well as the plasma density. Especially, the effective distance d eff, decreases with the increase of the electronegativity under the constant plasma density, which means that the shape of the plasma meniscus becomes flat or convex rather than concave. Moreover, it is also verified that the electronegativity affects the H- ion trajectories extracted from near the edge of plasma meniscus through the shape of plasma meniscus.

Collisional-radiative models are commonly used to analyse atomic and molecular processes in low temperature plasmas by determining the distribution functions of excited states as functions of various plasma parameters. This paper outlines the improvements to a zero-dimensional collisional-radiative model, developed at Keio University, for purposes relevant to the analysis of emission measurements in low temperature hydrogen plasmas. The extension of this 0D model centres on the inclusion of additional molecular species and reactions, a calculation of emission intensity in order to directly compare with experimental work, and the addition of a simple wall model to allow for a deeper understanding of recycling of atoms and molecules in low temperature plasmas close to plasma facing components in fusion machines. Results from the improved model are then compared with both an existing CR model for benchmarking and experimental emission spectroscopy data from an inductively-coupled plasma device at the University of Liverpool. These results show how the developments to the model have increased the relevance to experimental plasmas, such as those in the power exhaust regions of fusion machines, with the ratio of the Hβ/Hα line intensities outputted from the improved CR model mirroring more closely the emission measured experimentally in the ICP device.

To realize the development of a long plasma source with a uniform electron density distribution in the axial direction, the spatial distribution of plasma under a multi-cusp magnetic field was analyzed using a KEIO-MARC code. Considering a cylindrical plasma source with an axial length of 3000 mm and a cross-sectional diameter of 100 mm, in which the filament electrode was the electron source, the electron density distribution was calculated using the residual magnetic flux density, Bres, and the number of permanent magnets installed at different locations surrounding the device, Nmag, as design parameters. The results show that both Bres and Nmag improved the uniformity of the electron density distribution in the axial direction. The maximum axial electron density decreased with increasing Nmag and increased with increasing Bres. These trends can be explained by considering the nature of the multi-cusp field, where particles are mainly confined to the field-free region (FFR) near the center of the plasma column, and the loss of particles due to radial particle transport. The use of multiple filaments at intervals shorter than the plasma decay length dramatically improved axial uniformity. To further improve axial uniformity, the filament length and FFR must be properly set so that electrons are emitted inside the FFR.

The latest version of the SONIC code with the extended thermal force model that is capable to treat collisionality dependence of the force has been applied to the interpretative simulation of L-mode discharge of JT-60 U to investigate the impact of kinetic correction of the thermal force on the scrape-off layer (SOL) plasma properties. As a first attempt, we have selected an L-mode attached discharge #39090 (Ip = 1.6 MA, Bt = 3.1 T, and PNBI = 4.5 MW), where (Formula presented.) = 2.08 × 1019 m−3, (Formula presented.) = 0.392 with deuterium plasmas. It has been found that the peak values of SOL impurity density are reduced by a factor of 2, with the kinetic correction, because the thermal force is reduced by the low collisionality of SOL plasma between the X-point and the (inner and outer) midplane. On the other hand, the profiles near the divertor plates are unaffected by the kinetic correction due to the highly collisional condition. The kinetic correction is found more pronounced for the higher charge states impurity due to the charge dependence of the parallel impurity force balance. Overall effects due to the kinetic correction on the total impurity radiation and the divertor heat load are confirmed to be a few % in the present discharge condition of JT-60U.

© 2020 The Author(s) Handling of a large thermal power exhausted from the confined plasma is one of the most important issues for ITER and DEMO. A conventional divertor, which has the closed geometry similar to that of ITER and longer leg of 1.6 m, was proposed for the Japanese (JA) DEMO reactor (Rp/ap = 8.5/2.42 m). A radiative cooling scenario of Ar impurity seeding and the divertor performance have been demonstrated by SONIC simulation, in order to evaluate the power exhaust in JA-DEMO 2014 (primary design with Psep ~ 283 MW) and JA-DEMO with higher plasma elongation (a revised design with Psep ~ 235 MW). The divertor operation with the peak qtarget ≤ 10 MWm−2 was determined in the low nesep of 2–3 × 1019 m−3 under the severe conditions of reducing radiation loss fraction, i.e. f*raddiv = (Pradsol + Praddiv)/Psep, and diffusion coefficients (χ and D). The divertor geometry and reference key parameters (f*raddiv ~ 0.8, χ = 1 m2/s and D = 0.3 m2/s) were so far consistent with the power exhaust concepts in the nesep range, and the revised JA-DEMO design has advantages of wider nesep range and enough margin for the divertor operation. For either severe assumption of f*raddiv ~ 0.7 or χ and D to the half value, higher nesep operation was required for the primary design in order to control the peak qtarget ≤ 10 MWm−2, i.e. the operation window was reduced. Applying the two severe assumptions, the divertor operation was difficult in the low nesep range for the both designs.