| Citation: |
Yuwen YANG, Bin LI, Jianglong WEI, Lizhen LIANG, Yahong XIE, Chundong HU. Physical design of electron dumps for the beamline of CFETR advance neutral beam equipment (CANBE)[J]. Plasma Science and Technology. DOI: 10.1088/2058-6272/adcb18
|
Negative-ion-based neutral beam injection (NNBI) at high energy is still a potential heating and current drive method for future large-scale fusion reactors. One feature of the negative ion source is that the electrons are also accelerated with the negative ions in the accelerator. In addition to the electrons co-extracted from the source plasma, electrons are inevitably generated in the main acceleration stage due to the stripping loss of negative ions or the ionization of background gas. Many of these accelerated electrons can be ejected by the negative ion source, which poses a risk for the downstream components in the beamline of the NNBI system, especially the thermosensitive cryopumps. To prevent the ejected electrons from impacting the cryopumps, electron dumps have been designed for the beamline of the China Fusion Engineering Test Reactor (CFETR) advance neutral beam equipment (CANBE). Using a self-consistent model of the negative ion accelerator, the generation and motion of the ejected electrons were simulated for different negative ion sources to be tested on the CANBE. Based on the orbits of the ejected electrons, a set of fixed and movable electron dumps was designed to be placed in front of the neutralizer inside the CANBE beamline. The effects of the electron dumps were quantitatively analyzed.
Neutral beam injection (NBI) has proven to be an efficient external heating method in many magnetic confinement fusion devices [1–8]. When the required injection beam energy is higher than 150 keV, negative-ion-based neutral beam injection (NNBI) systems are more practical because of their higher neutralization efficiency than positive-ion-based neutral beam injection (PNBI) systems. However, the development and operation of NNBI systems are more complex and challenging than those of PNBI systems, in terms of the unstable and low yield of negative ions, the higher acceleration energy, the longer beam transmission distance, and so on [9–11]. Although high-energy (~ 1 MeV), high-power NBI (dozens of MW) is still desired for future large-scale fusion reactors, the NNBI technology itself slows down its application prospects [12–17]. Therefore, an NNBI test facility has been under construction since 2019 in the framework of the Comprehensive Research fAcility for Fusion Technology (CRAFT), which is an advance research project for different key technologies that contribute to fusion energy. For project acceptance in 2025, the CRAFT NNBI test facility is required to achieve a hydrogen atom beam power of 2 MW at a particle energy of 200‒400 keV over a pulse duration of 100 s.
There are two test stands in the CRAFT NNBI test facility [18]. One is the Hefei Open-facility for Negative-ion Source Research (HONOR), which is dedicated to research and development (R&D) of the negative ion source for NBI with various diagnostic techniques. The other is the CFETR Advance Neutral Beam Equipment (CANBE). The engineering designs of the CFETR have been completed [19]. Thus, the CANBE test stand will function as the CFETR NNBI prototype, allowing researchers to carry out comprehensive research including negative-ion-beam acceleration, negative ion neutralization, neutral beam purification, and neutral beam transport. Considering the lack of NNBI R&D experience in China, single-, dual-, and quad-driver radio frequency (RF) negative ion sources with acceleration voltages of 50 kV, 200 kV, and 400 kV are planned to be developed step by step in the CRAFT NNBI project to reach the acceptance targets progressively [20–22].
One distinctive characteristic of negative ion sources is that the electrons are co-extracted or co-accelerated, accompanying the negative ions. The accelerated electrons are wasteful in NNBI systems. They can not only reduce the acceleration efficiency of the high-voltage power supply but also impact the accelerator or the beamline components at high energy. In general, to prevent the co-extracted plasma electrons from entering the major acceleration stage, electron deflection magnets are embedded inside the second electrode of the accelerator (called an extraction grid, EG). Almost all the co-extracted electrons can be deflected and dumped onto the EG [23]. However, the stripping loss of negative ions or the ionization of background gas can also generate electrons during acceleration. The motions of the lightweight electrons are easily influenced by the electromagnetic field and collisions with solid surfaces in the accelerator. Although those stray electrons can be intentionally intercepted by the other electrodes, high-energy electrons are still ejected by the negative ion source [24].
High-power ejected electrons have been found in both JT-60U and Large Helical Device (LHD) NNBI systems [25–27]. Special electron beam dumps were installed downstream from their negative ion sources to avoid melting of the other components. During the long-pulse beam extraction of the CRAFT single-driver RF negative ion source, degradation of the cryopump and a heat spot on the vacuum vessel were observed due to the heat load caused by the ejected electrons [28]. Thus, ejected electron dumps are necessary in NNBI systems to reduce the damage to the non-cooling components and the cryopumps.
The trajectories and the kinetic energies of the ejected electrons are different for different negative ion sources or under different operational conditions. In addition, the beamline structure downstream from the negative ion source differs among the different devices. For example, the designs of the electron dumps for the ITER Heating Neutral Beam (HNB) and Diagnostic Neutral Beam (DNB) differ due to differences in electron energies and angular spectra [29], and in their updated versions, the structures of the electron dumps have been modified to reduce their influence on gas pumping [30–33].
A picture of the CANBE test stand and a sectional view of the beamline are shown in figure 1. The experimental study is first conducted with the CRAFT negative ion sources working under atmospheric conditions because the technology of high-voltage power transmission in vacuum is immature. Thus, the beam source vacuum vessel is not used (even though it has been manufactured). The internal dimensions of the beamline vacuum vessel are 12 m × 4 m × 4 m. It is composed of the neutralization vessel, the purification vessel, and the diagnostic vessel, which are similar to those of the ITER HNB beamline. These three vessels are vacuum-sealed and connected by O-rings and clamps.
A gas neutralizer is used to neutralize the high-energy negative ion beam [34–36]. The neutralizer is a rectangular structure with a length of 3 m. To reduce the gas inlet for the required gas target thickness, the neutralizer gas cell is divided into two channels. The cross section of each channel is adjustable for different negative ion sources and has a maximum cross section of 1.7 m in height and 0.3 m in width. At present, hydrogen gas is injected from the middle of each channel (i.e. 1.5 m downstream the neutralizer entrance), but the injection position can be easily changed by ±0.5 m.
To remove the high-energy (~ 1 MeV) and large-area (~ 1 m2) residual ion beam, an electrical deflection system is more practical than a magnetic deflection system. Therefore, an electrical residual ion dump (ERID) is adopted in the CANBE to develop this new technique for future NNBI systems [37]. The ERID of CANBE mainly comprises three plates that form two beam channels. The middle plate is connected to a high positive potential to collect the negative ions, while the other two are at ground potential to collect positive ions. The cross section of the two beam channels is fixed at 2×0.31 m2. The current length of the ERID is 3 m.
A V-shaped beam dump is used for beam operation and the measurement of beam power [38, 39]. Two panels forming the array of cooling pipes are installed on a support frame. Their length is also 3 m and the “V” opening angle is 15°, which is based on the estimated beam properties.
Two cryopumps (cryo-sorption type) are installed on the lateral wall of the vacuum vessel [40, 41]. For each cryopump, the length is 8 m and the height is 2.3 m, and the design pumping speed is 4,000,000 L/s. The working temperature of the liquid helium panels is 5 K for gas pumping, and the working temperature of the liquid nitrogen panels is 77 K for precooling and heat shielding. Two gate valves with openings of 2 m in diameter are connected to the entrance duct and exit duct of the vacuum vessel, respectively. The test negative ion sources are directly installed on the gate valves.
In the CRAFT project execution period, a total of four negative ion sources have been manufactured. They are: (i) the single-driver negative ion source, (ii) the dual-driver negative ion source, (iii) the quad-driver negative ion source working in air, and (iv) the quad-driver negative ion source working in vacuum. Among them, the single-driver negative ion source has not been tested on the CANBE; the quad-driver negative ion source working in vacuum will not be tested until 2027, because the development of the high-voltage power transmission into the vacuum vessel is progressing more slowly than expected.
The structure of the dual-driver negative ion source is shown in figure 2(a). It consists of a plasma generator with two RF drivers and a three-electrode accelerator operating at 200 keV. The plasma is produced via an inductively coupled plasma (ICP) mode in two identical drivers, which have a diameter of 24 cm [42]. The two antennas connected in series operate at a maximum RF power of 200 kW (100 kW each) and a frequency of 1 MHz. After production, the generated plasma diffuses and mixes in the expansion chamber. In order to confine the plasma, rows of permanent magnets are arranged around the lateral walls of the expansion chamber. Cs vapor is injected from the top and bottom walls to enhance the production of negative ions. A magnetic filter field (MFF) is formed with the current flowing vertically through the plasma grid (PG). The MFF is used to decrease the electron temperature and electron density in front of the PG, which can reduce the collisional loss of negative ions and the number of co-extracted electrons [43]. A bias plate is placed and insulated between the expansion chamber and the PG to further suppress the co-extracted electrons by a positive bias voltage on the PG [44].
The grid electrodes in the accelerator contain the PG, the EG, and the ground grid (GG). The grids are divided into two segments, and one segment of the grid has two groups (each containing 6×16 apertures). The electrostatic field between the PG and the EG (up to 12 kV within a distance of 7 mm) extracts the negative ions near the PG through the round apertures, which are then accelerated by the electrostatic field between the EG and the GG (up to 200 kV within a distance of 90 mm). These accelerated negative ions are launched from 4×16 slot apertures with the aim of reducing particle bombardment and heat deposition on the GG. The high voltage difference between the different grids is maintained via large rectangular insulating tubes made from fiber-enhanced epoxy. To ensure that the extraction gap and the acceleration gap are maintained correctly, the grid electrodes are held up by supporting frames.
In order to prevent these co-extracted electrons from being accelerated to high energy, co-extracted electron deflection magnets (CEDMs) are embedded periodically with alternating magnetization row by row inside the EG, and the co-extracted electrons are deflected and dumped onto the EG. However, the magnetic field also causes deflection of the H− beamlet because of the imbalanced line integrals of the magnetic forces. Therefore, asymmetric deflection compensation magnets (ADCMs) are applied to correct the deflection [45].
The quad-driver negative ion source (shown in figure 2(b)) comprises a plasma generator with four RF drivers and a four-electrode accelerator operating at 400 keV. In general, the quad-driver negative ion source has a similar physical design to that of the dual-driver negative ion source. The key design parameters of the two negative ion sources are listed in table 1. One major difference is that the quad-driver negative ion source is doubled in size in the horizontal direction, including the number of RF drivers, the width of the expansion chamber, the width of the extraction area, and so on. Another major difference is that the quad-driver negative ion source is equipped with a two-stage accelerator, where the negative ions are accelerated twice. The largest difference in the engineering design is that the grid support frames and the insulating tubes have circular cross sections in order to reduce manufacturing difficulty and enhance their structural strength.
| Design parameter | Dual-driver negative ion source | Quad-driver negative ion source |
| RF power | 2×80 kW | 4×80 kW |
| RF frequency | 1 MHz | 1 MHz |
| RF driver | Φ 24 cm, 16 cm | Φ 24 cm, 16 cm |
| Expansion chamber | 90×50×20 cm3 | 110×90×21 cm3 |
| PG aperture | Φ 14 mm, 16×6×4 | Φ 14 mm, 17×12×4 |
| Beam size | 74.8×30 cm2 | 81.4×60 cm2 |
| Grid electrodes | PG-EG-GG | PG-EG-AG-GG |
| Extraction voltage | 8 kV | 8 kV |
| Acceleration voltage | 200 kV | 2×200 kV |
| Ion current | 15 A (~ 255 A/m2) | 28 A (~ 225 A/m2)* |
| Pulse duration | > 100 s | > 100 s |
| * Limited by the acceleration voltage power supply. | ||
DownLoad:
CSV
The design of the electron dump must be based on the movement and power deposition of ejected electrons. A self-consistent model comprising an electric field, a magnetic field, particle tracing, and a rarefied gas flow has been adopted to simulate the generation, motion, and deposition of electrons in accelerators using different negative ion sources [46, 47].
The design parameters of either negative ion source in table 1 are applied as input conditions. The ratio of co-extracted electrons to ions is assumed to be 1.0. The single negative beamlet optics are calculated with the IBSimu code to determine the emitting surfaces of negative ions for this study [48]. The emitting surfaces are assumed to be the same for different PG apertures, and the negative ions are assumed to be distributed uniformly on the emitting surface. The multi-beamlet optics are simulated using the bidirectional particle–field coupling. The trajectories of the negative ions show little change after several iterations.
The beam–gas interactions are an important source of electrons in the accelerator (called stray electrons). The rarefied gas flow in either accelerator is simulated by the test particle Monte Carlo (TPMC) method. In the model, the filling pressure is 0.3 Pa, and the temperature of the residual gas produced by the plasma generator is 630 K. In addition, the temperatures of the PG, EG, and GG are set to 200 °C, 100 °C, and 100 °C, respectively. All the grid support frames are set to room temperature. The reflections of gas molecules on the surface follow a cosine law, and the thermal accommodation coefficient is assumed to be 1.0. The number of simulated H2 molecules is 5×106. The pumping capture factor is set to 0.3 at the exit of the negative ion source. Figure 3(a) illustrates the simulated gas density in the accelerator of the dual-driver and quad-driver negative ion sources.
The magnetic fields formed by the magnets (i.e. the CEDMs and ADCMs) and the PG current (1.1 kA, a typical operational value for the dual-driver negative ion source) in the accelerator are calculated and shown in figure 3(b). Under the combination of the CEDMs and the ADCMs, the upstream field strength of the EG is enhanced to 640 Gauss, while the downstream field strength is weakened to 340 Gauss. At a PG current of 1.1 kA, the magnetic field in front of the PG is about 12 Gauss, and it decreases to 1.5 Gauss close to the GG.
The beam–gas processes in both accelerators were simulated using the simulation results of the multi-beamlet optics and the rarefied gas flow. The particles’ movement and their collisions are decoupled in the model. This means that during a time step, the particles only move according to Newton’s laws of motion; at the end of the time step, the collisions of the particles with the gas are calculated based on the Monte Carlo (MC) method. The major collision reactions considered in the simulation and their cross sections are illustrated in figure 4 [49]. All the particles, including the products of collisions, are traced during the simulation. Here, the time step is set to 1×10−11 s, and the total simulation time was 1×10−7 s. The position and velocity of the ejected electrons leaving the GG were recorded as input conditions for the design of the electron dump.
The characteristics of the stray electrons in the accelerator of the CRAFT dual-driver and quad-driver negative ion sources are illustrated in figure 5, showing only one of the multi-beamlet rows for better understanding. The stray electrons generated between the PG and EG are almost dumped on the surface of the EG owing to the electron deflection magnetic field. Downstream from the EG, the stray electrons generated at random positions in the accelerator have different initial kinetic energies. These electrons are accelerated by the electrostatic field and deflected by the magnetic field at the same time. Some of the electrons are deposited on the electrode grid, while others are ejected from the GG. Figure 6 illustrates the particle energies and deflection angles of the ejected electrons. The green dots represent the electrons ejected from the left group of apertures, and the blue dots are the electrons ejected from the right group of apertures. High-energy ejected electrons with large deflection angles can cause severe damage to the cryopumps.
A simplified geometric model of the CANBE beamline was built to design the electron dump (shown in figure 7(a)). All the internal components (i.e. the neutralizer, residual ion dump, beam dump, and cryopumps) and the gate valve and connecting ducts were simplified to their characteristic shapes, but their key geometric parameters were kept. The thermal limitation of the cryopump was 1 kW, and the design goal of the electron dump was to minimize the heat load on the cryopump as much as possible, ensuring its stable and normal operation.
The backscattering of ejected electrons on the target was previously implemented, as described in [50]. The backscattering probability, i.e. the probability with which a primary electron collides with a target at an incidence angle of θ1, is simulated by this expression:
| ηb(θ1)=ηb0exp[Ab0(1−cosθ1)], |
where ηb0 is the backscattering probability at normal incidence (θ1=0) and the coefficient Ab0(E0) is obtained by fitting experimental data:
| Ab0=κ(E0)ln(1/ηb0), |
with
| κ=1−exp(−1.83E1/40(keV)), |
where E0 is the energy of the incident electron, and the probability of electron backscattering on a copper plate at normal incidence is approximately 0.3. The value ηb0≅0.3 is applied in this simulation. The backscattered electrons are reemitted at the location where the primary electrons collide with the copper target. The average energy of the backscattered electrons is about 0.7–0.9 times the energy of the primary electrons, depending on the incident electron energy and the material of the target. Materials with high atomic numbers, such as copper, have less energy loss, and here the incident electrons have high energy. Therefore, the average energy of the backscattered electrons is relatively high, about 0.7–0.95 times the energy of the incident electrons. In this model, the energy of the backscattered electrons is 0.8 times the primary electron energy.
The trajectories of the electrons produced by the CRAFT dual-driver and quad-driver negative ion sources in the CANBE beamline are shown in figure 7(b). It can be noted that: (1) the transmission space due to the GG support frame of either source was also considered in the model; (2) the cross section of the neutralizer channels was tailored to different sources. In the beamline, some of the ejected electrons are directed into the neutralizer channels. It can be seen that the high-energy ejected electrons have a large deflection, and most of them are deposited on the cryopumps on both sides of the vacuum vessel. Moreover, a number of the backscattered electrons are generated when the ejected electrons collide with the entrance plate of the neutralizer. The power load distributions imposed on the cryopumps by the electrons in the cases of the CRAFT dual-driver and quad-driver negative ion sources are presented in figure 7(c). The locations of the calculated thermal deposition are basically the middle and rear sections of the cryopumps.
Based on the trajectories of the electrons in the CANBE beamline, a set of electron dumps was designed to intercept the ejected electrons (shown in figure 8(a)). A fixed electron dump was located around the entrance of the neutralizer, and two movable electron dumps were designed at the entrance of the duct in front of the neutralizer. The fixed electron dump is a plate with a length of 2350 mm, a width of 1140 mm, and a thickness of 10 mm, and it has a cutout, the cross-sectional area of which is the same as that of the entrance to the neutralizer. The sizes of the two movable electron dumps are the same; their length, width, and thickness are 2000 mm, 360 mm, and 20 mm, respectively. The two movable electron dumps have central axial symmetry; they are rotated by 45° in the horizontal direction, and they can move horizontally to match the different negative ion sources.
With these electron dumps, most of the electrons are obstructed rather than deposited on the cryopumps. However, a number of backscattered electrons are generated when the ejected electrons collide with the electron dumps, and the result is that the thermal deposition is mainly concentrated at the front of the cryopumps. Due to the difference in acceleration energy, the power load on the cryopumps caused by the electrons in the case of the CRAFT quad-driver negative ion source is higher than that of the CRAFT dual-driver negative ion source.
Three different simulations were calculated, i.e. without an electron dump, with only a fixed electron dump, and with both a fixed dump and movable dumps. A summary of the results is listed in table 2 for the CRAFT dual-driver negative ion source. Without protection, the electron power load on the cryopumps is 36.33 kW, which is a significant threat to operation. A fixed electron dump decreases the heat load on the cryopumps to 6.83 kW, while the peak power density on the cryopumps is also reduced from 6.51 kW/m2 to 4.22 kW/m2. When added to the fixed electron dump, the movable electron dumps diminish the heat load to 1.74 kW and decrease the peak power density to 1.00 kW/m2, with the high heat load located at the forward ends of the cryopumps.
| Component | Estimated heat load (kW) and peak power density (kW/m2) on component | ||
| Without electron dump | Fixed electron dump | Fixed electron dump and movable electron dumps | |
| Cryopumps | 36.33 kW/6.51 kW/m2 | 6.73 kW/4.22 kW/m2 | 1.74 kW/1.00 kW/m2 |
| Fixed electron dump | - | 51.18 kW/175.84 kW/m2 | 38.12 kW/176.64 kW/m2 |
| Movable electron dumps | - | - | 28.79 kW/112.65 kW/m2 |
DownLoad:
CSV
The results for the CRAFT quad-driver negative ion source are reported in table 3. Without protective measures, the cryopumps receive a power deposition of 40.68 kW and a power density of approximately 11.56 kW/m2. The fixed electron dump reduces power load by 9.44 kW. Significantly, in the case of the CRAFT quad-driver negative ion source, the cross-section of the neutralizer channel is enlarged, the size of the fixed electron dump is reduced, and the number of electrons that can be intercepted is decreased. However, the heat load on the cryopumps is greatly reduced (from 31.24 kW to 1.80 kW) due to the effect of the movable electron dumps. The position with the largest thermal distribution is again at the front ends of the cryopumps, which is mainly caused by backscattered electrons.
| Component | Estimated heat load (kW) and peak power density (kW/m2) on component | ||
| Without electron dump | Fixed electron dump | Fixed electron dump and movable electron dumps | |
| Cryopumps | 40.68 kW/11.56 kW/m2 | 31.24 kW/10.90 kW/m2 | 1.80 kW/2.08 kW/m2 |
| Fixed electron dump | - | 41.95 kW/308.07 kW/m2 | 21.43 kW/304.40 kW/m2 |
| Movable electron dumps | - | - | 41.73 kW/265.29 kW/m2 |
DownLoad:
CSV
The peak power density of the movable electron dumps is about 265 kW/m2. The cooling channel is embedded in the electron dumps, and the inner diameter of the water channel is 12 mm. The highest temperature rises of the electron dumps and the cooling water are about 150 °C and 110 °C, respectively (shown in figure 9). In addition, the influence of the electron dumps on gas pumping is also calculated. Figure 10 illustrates the gas density in the beamline vacuum chamber for different volumes of supplemental gas, with and without electron dumps. The supplemental gas for the neutralizer is H2. In the section from the GG to the entrance of the neutralizer, the gas density with electron dumps is higher than that without electron dumps. However, it can be seen that the installation of electron dumps has little impact on the gas pumping in the beamline vacuum when additional gas is supplied to the neutralizer.
An NNBI test facility is under construction as part of the CRAFT framework. In the CRAFT NNBI test facility, there are two test stands, HONOR and CANBE. The former aims to develop the negative ion source, and the latter will be used to carry out comprehensive research. To achieve the target of the CRAFT NNBI test facility, single-, dual-, and quad-driver RF negative ion sources with acceleration voltages of 50 kV, 200 kV, and 400 kV are being developed step by step.
There are stray electrons generated in the accelerator of the negative ion source. These high-energy electrons are ejected from the negative ion source and pose a significant threat to the beamline components, which are sensitive to heat. For the negative ion sources that will be operated on the CANBE test stand in the near future, a self-consistent model comprising an electric field, a magnetic field, particle tracing, and a rarefied gas flow has been adopted to simulate the generation and motion of electrons in the accelerator of different negative ion sources (the CRAFT dual-driver and quad-driver negative ion sources). Using the positions and velocities of the electrons obtained from the former simulation, the trajectories of these electrons are simulated in the CANBE beamline, including the backscattering of electrons on the target. The thermal deposition on the cryopumps is mainly concentrated in the middle and rear sections, which is caused by the high-energy electrons with large deflection.
Based on the trajectories of the electrons, a fixed electron dump was designed for the entrance of the neutralizer, and two movable electron dumps were designed for the duct in front of the neutralizer. The movable electron dumps can move horizontally to match different negative ion sources. A comparison of the simulation results with and without the application of electron dumps shows that the use of the designed electron dumps decreases the heat load on the cryopumps by 34.59 kW in the case of the CRAFT dual-driver negative ion source, while in the case of the CRAFT quad-driver negative ion source, it reduces the heat load by 38.88 kW. In addition, the use of the electron dumps focuses the position of thermal deposition at the forward end of the cryopumps. Moreover, the installation of the electron dumps has little influence on the gas pumping in the beamline vacuum chamber.
This work has been carried out within the framework of the CRAFT NNBI Physics Activity. This work was supported by the Comprehensive Research Facility for Fusion Technology Program of China (No. 2018-000052-73-01-001228).
| [1] |
Eubank H et al 1979 Phys. Rev. Lett. 43 270 doi: 10.1103/PhysRevLett.43.270
|
| [2] |
Wagner F et al 1982 Phys. Rev. Lett. 49 1408 doi: 10.1103/PhysRevLett.49.1408
|
| [3] |
Keilhacker M et al 1999 Nucl. Fusion 39 209 doi: 10.1088/0029-5515/39/2/306
|
| [4] |
Fujita T et al 1999 Nucl. Fusion 39 1627 doi: 10.1088/0029-5515/39/11Y/302
|
| [5] |
Maslov M et al 2023 Nucl. Fusion 63 112002 doi: 10.1088/1741-4326/ace2d8
|
| [6] |
Han H et al 2022 Nature 609 269 doi: 10.1038/s41586-022-05008-1
|
| [7] |
Magee R M et al 2023 Nat. Commun. 14 955 doi: 10.1038/s41467-023-36655-1
|
| [8] |
Ding S et al 2024 Nature 629 555 doi: 10.1038/s41586-024-07313-3
|
| [9] |
Hemsworth R S and Inoue T 2005 IEEE Trans. Plasma Sci. 33 1799 doi: 10.1109/TPS.2005.860090
|
| [10] |
Kojima A et al 2010 Rev. Sci. Instrum. 81 02B112 doi: 10.1063/1.3279398
|
| [11] |
Tsumori K et al 2022 Nucl. Fusion 62 056016 doi: 10.1088/1741-4326/ac2d59
|
| [12] |
Polevoi A R et al 2020 Nucl. Fusion 60 096024 doi: 10.1088/1741-4326/aba335
|
| [13] |
Chen J L et al 2021 Nucl. Fusion 61 046002 doi: 10.1088/1741-4326/abd7b8
|
| [14] |
Vincenzi P et al 2021 Plasma Phys. Control. Fusion 63 065014 doi: 10.1088/1361-6587/abf402
|
| [15] |
Sugiyama S et al 2024 Nucl. Fusion 64 076014 doi: 10.1088/1741-4326/ad49b6
|
| [16] |
Mikkelsen D R et al 2018 Nucl. Fusion 58 036014 doi: 10.1088/1741-4326/aaa4d2
|
| [17] |
Hopf C et al 2021 Nucl. Fusion 61 106032 doi: 10.1088/1741-4326/ac227a
|
| [18] |
Wei J L et al 2025 Plasma Sci. Technol. 27 044001 doi: 10.1088/2058-6272/ad8da7
|
| [19] |
Song Y T et al 2022 Fusion Eng. Des. 183 113247 doi: 10.1016/j.fusengdes.2022.113247
|
| [20] |
Xie Y H et al 2023 J. Instrum. 18 C07017 doi: 10.1088/1748-0221/18/07/C07017
|
| [21] |
Gu Y M et al 2023 Fusion Eng. Des. 192 113603 doi: 10.1016/j.fusengdes.2023.113603
|
| [22] |
Xie Y H et al 2021 Fusion Eng. Des. 167 112377 doi: 10.1016/j.fusengdes.2021.112377
|
| [23] |
Cui Q L et al 2020 Fusion Eng. Des. 161 112056 doi: 10.1016/j.fusengdes.2020.112056
|
| [24] |
Yang Y W et al 2023 Nucl. Eng. Technol. 55 939 doi: 10.1016/j.net.2022.12.002
|
| [25] |
Umeda N et al 2006 Rev. Sci. Instrum. 77 03A529 doi: 10.1063/1.2170088
|
| [26] |
Hanada M et al 2008 IEEE Trans. Plasma Sci. 36 1530 doi: 10.1109/TPS.2008.927289
|
| [27] |
Oka Y et al 2002 Rev. Sci. Instrum. 73 1054 doi: 10.1063/1.1432469
|
| [28] |
Yang Y W et al 2024 Plasma Phys. Control. Fusion 66 055019 doi: 10.1088/1361-6587/ad3c1e
|
| [29] |
de Esch H P L and Singh M J 2010 Fusion Eng. Des. 85 707 doi: 10.1016/j.fusengdes.2010.04.011
|
| [30] |
Agostinetti P et al 2011 Nucl. Fusion 51 063004 doi: 10.1088/0029-5515/51/6/063004
|
| [31] |
Agostinetti P et al 2012 IEEE Trans. Plasma Sci. 40 629 doi: 10.1109/TPS.2011.2179947
|
| [32] |
Veltri P et al 2013 Fusion Eng. Des. 88 1011 doi: 10.1016/j.fusengdes.2013.04.012
|
| [33] |
Dalla Palma M et al 2013 Fusion Eng. Des. 88 1020 doi: 10.1016/j.fusengdes.2013.01.018
|
| [34] |
Zhang Z Y et al 2019 Fusion Eng. Des. 148 111316 doi: 10.1016/j.fusengdes.2019.111316
|
| [35] |
Wei J L et al 2021 Fusion Eng. Des. 169 112482 doi: 10.1016/j.fusengdes.2021.112482
|
| [36] |
Yi W et al 2023 Fusion Eng. Des. 194 113749 doi: 10.1016/j.fusengdes.2023.113749
|
| [37] |
Liang L Z et al 2023 Fusion Eng. Des. 190 113499 doi: 10.1016/j.fusengdes.2023.113499
|
| [38] |
Tao L et al 2022 Fusion Sci. Technol. 78 490 doi: 10.1080/15361055.2022.2050131
|
| [39] |
Tang N et al 2022 Fusion Eng. Des. 176 113018 doi: 10.1016/j.fusengdes.2022.113018
|
| [40] |
Hu C D et al 2023 Fusion Eng. Des. 189 113469 doi: 10.1016/j.fusengdes.2023.113469
|
| [41] |
Wang Q X et al 2025 Nucl. Eng. Technol. 57 103178 doi: 10.1016/j.net.2024.08.047
|
| [42] |
Wang N et al 2023 Plasma Sci. Technol. 25 045601 doi: 10.1088/2058-6272/aca1fb
|
| [43] |
Xie J W et al 2024 Plasma Sci. Technol. 26 045602 doi: 10.1088/2058-6272/ad1043
|
| [44] |
Peng X F et al 2023 Phys. Plasmas 30 103507 doi: 10.1063/5.0156271
|
| [45] |
Agostinetti P et al 2016 Nucl. Fusion 56 016015 doi: 10.1088/0029-5515/56/1/016015
|
| [46] |
Wei J L et al 2019 Rev. Sci. Instrum. 90 113313 doi: 10.1063/1.5128335
|
| [47] |
Wei J L et al 2023 Phys. Plasmas 30 033102 doi: 10.1063/5.0139827
|
| [48] |
Kalvas T et al 2010 Rev. Sci. Instrum. 81 02B703 doi: 10.1063/1.3258608
|
| [49] |
Hunter H T et al 1990 Atomic data for fusion (Oak Ridge: Oak Ridge National Lab.
|
| [50] |
Fubiani G et al 2008 Phys. Rev. ST Accel. Beams 11 014202 doi: 10.1103/PhysRevSTAB.11.014202
|
| [1] | Junwei Zhang, Shifeng Mao, Jin Guo, Jiafeng He, Tianyuan Liu. Calculation of neutral source terms with deep learning to accelerate edge plasma simulations[J]. Plasma Science and Technology. DOI: 10.1088/2058-6272/add1b0 |
| [2] | Borui ZHENG, Jianbo ZHANG, Shaojie QI, Jianghua XU, Yiche LI, Yuanzhong JIN, Dongliang BIAN. Spatiotemporal evolution laws of sector-shaped dielectric-barrier-discharge plasma actuator[J]. Plasma Science and Technology, 2024, 26(10): 105504. DOI: 10.1088/2058-6272/ad5d4f |
| [3] | Hanyu ZHANG, Lina ZHOU, Yueqiang LIU, Guangzhou HAO, Shuo WANG, Xu YANG, Yutian MIAO, Ping DUAN, Long CHEN. Deep learning approaches to recover the plasma current density profile from the safety factor based on Grad–Shafranov solutions across multiple tokamaks[J]. Plasma Science and Technology, 2024, 26(5): 055101. DOI: 10.1088/2058-6272/ad13e3 |
| [4] | Fusheng WANG (王富生), Xiangteng MA (马襄腾), Han CHEN (陈汉), Yao ZHANG (张耀). Evolution simulation of lightning discharge based on a magnetohydrodynamics method[J]. Plasma Science and Technology, 2018, 20(7): 75301-075301. DOI: 10.1088/2058-6272/aab841 |
| [5] | Li ZHANG (张丽), Dezheng YANG (杨德正), Sen WANG (王森), Wenchun WANG (王文春). Spatiotemporal characteristics of nanosecond pulsed discharge in an extremely asymmetric electric field at atmospheric pressure[J]. Plasma Science and Technology, 2017, 19(6): 64006-064006. DOI: 10.1088/2058-6272/aa632d |
| [6] | Kai GAO (高凯), Nasr A M HAFZ, Song LI (李松), Mohammad IRZAIE, Guangyu LI (李光宇), Quratul AIN. Online plasma diagnostics of a laser-produced plasma[J]. Plasma Science and Technology, 2017, 19(1): 15506-015506. DOI: 10.1088/1009-0630/19/1/015506 |
| [7] | ZHANG Jingyang (张镜洋), HAN Le (韩乐), CHANG Haiping (常海萍), LIU Nan (刘楠), XU Tiejun (许铁军). The Corrected Simulation Method of Critical Heat Flux Prediction for Water-Cooled Divertor Based on Euler Homogeneous Model[J]. Plasma Science and Technology, 2016, 18(2): 190-196. DOI: 10.1088/1009-0630/18/2/16 |
| [8] | DING Yonghua (丁永华), JIN Xuesong (金雪松), CHEN Zhenzhen (陈真真), ZHUANG Ge (庄革). Neural Network Prediction of Disruptions Caused by Locked Modes on J-TEXT Tokamak[J]. Plasma Science and Technology, 2013, 15(11): 1154-1159. DOI: 10.1088/1009-0630/15/11/14 |
| [9] | FENG Wu (冯武), LI Xibo (李希波), ZHENG Chunyang (郑春阳). Spatiotemporal Evolution of Stimulated Raman Scattering in the Absolute and Convective Regimes¤[J]. Plasma Science and Technology, 2013, 15(8): 721-726. DOI: 10.1088/1009-0630/15/8/01 |
| [10] | WU Tao (吴涛), WANG Xinbing (王新兵), WANG Shaoyi (王少义). Spectral Efficiency Extreme Ultraviolet Emission from CO2 Laser-Produced Tin Plasma Using a Grazing Incidence Flat-Field Spectrograph[J]. Plasma Science and Technology, 2013, 15(5): 435-438. DOI: 10.1088/1009-0630/15/5/08 |
Supported by: Beijing Renhe Information Technology Co., Ltd. 
| Design parameter | Dual-driver negative ion source | Quad-driver negative ion source |
| RF power | 2×80 kW | 4×80 kW |
| RF frequency | 1 MHz | 1 MHz |
| RF driver | Φ 24 cm, 16 cm | Φ 24 cm, 16 cm |
| Expansion chamber | 90×50×20 cm3 | 110×90×21 cm3 |
| PG aperture | Φ 14 mm, 16×6×4 | Φ 14 mm, 17×12×4 |
| Beam size | 74.8×30 cm2 | 81.4×60 cm2 |
| Grid electrodes | PG-EG-GG | PG-EG-AG-GG |
| Extraction voltage | 8 kV | 8 kV |
| Acceleration voltage | 200 kV | 2×200 kV |
| Ion current | 15 A (~ 255 A/m2) | 28 A (~ 225 A/m2)* |
| Pulse duration | > 100 s | > 100 s |
| * Limited by the acceleration voltage power supply. | ||
DownLoad:
CSV
| Component | Estimated heat load (kW) and peak power density (kW/m2) on component | ||
| Without electron dump | Fixed electron dump | Fixed electron dump and movable electron dumps | |
| Cryopumps | 36.33 kW/6.51 kW/m2 | 6.73 kW/4.22 kW/m2 | 1.74 kW/1.00 kW/m2 |
| Fixed electron dump | - | 51.18 kW/175.84 kW/m2 | 38.12 kW/176.64 kW/m2 |
| Movable electron dumps | - | - | 28.79 kW/112.65 kW/m2 |
DownLoad:
CSV
| Component | Estimated heat load (kW) and peak power density (kW/m2) on component | ||
| Without electron dump | Fixed electron dump | Fixed electron dump and movable electron dumps | |
| Cryopumps | 40.68 kW/11.56 kW/m2 | 31.24 kW/10.90 kW/m2 | 1.80 kW/2.08 kW/m2 |
| Fixed electron dump | - | 41.95 kW/308.07 kW/m2 | 21.43 kW/304.40 kW/m2 |
| Movable electron dumps | - | - | 41.73 kW/265.29 kW/m2 |
DownLoad:
CSV
