† Corresponding author. E-mail:
It is very important to improve the penetration depth and fueling efficiency of supersonic molecular beam injection (SMBI) especially for the next generation fusion devices such as ITER. Two components, a fast component (FC) and a slow component (SC), have been observed in the HL-2A SMBI experiments for several years, and the FC can penetrate much more deeply than the common SMBIs which draws a great deal of attention for a better fueling method. It is the first time to the FC and SC of SMBI have been simulated and interpreted in theory and simulation in this paper with the trans-neut module of the BOUT++ code. The simulation results of the FC and SC are clear and distinguishable in the same way as the observation in experiment. For the major mechanism of the FC and SC, it is found that although the difference in the injection velocity has some effect on the penetration depth difference between the FC and SC, it is mainly caused by the self-blocking effect of the first ionized SMB. We also discuss the influence of the initial plasma density on the FC and SC, and the variation of the SC penetration depth with its injection velocity.
It is important to improve the penetration depth and efficiency of plasma fuelling for controlling the plasma density and studying plasma physics especially in the next generation fusion devices such as ITER. There are three major fueling methods: gas puffing (GP),[1] pellet injection (PI),[2] and supersonic molecular beam injection (SMBI).[3] GP is the simplest and most traditional method with the lowest fuelling efficiency and the shallowest penetration depth. PI can penetrate into the core plasma but costs a lot for its complex structure and technology. SMBI costs less than PI because of its simpler structure and technology, and its fuelling efficiency is about three to four times as high as that of GP.[4] The key technique difference between SMBI and GP is the Laval nozzle of SMBI. It leads to a directed group fluid velocity of molecules during SMBI, which is very different from the individual isotropic thermal speed of molecules during GP. An improved SMBI with clusters, which are like micro-pellets, can be injected more deeply.[5] Besides being used in the plasma fuelling, SMBI can also be applied to study plasma impurity transport,[6–8] L–H transition and confinement improvement,[9–11] edge localized mode (ELM) mitigation,[12–18] and nonlocal heat transport.[19–22]
Since the fuelling efficiency depends on the penetration depth, some experiments have been carried out to improve the SMBI penetration depth on the HL-2A tokamak.[4,23,24] The SMBI fuelling depth can be measured with electron-cyclotron emission, soft x-ray diagnostics, Dα array, and CCD camera on the HL-1M and HL-2A,[24–27] fast TV camera and charge exchange recombination spectroscopic in JT-60U.[28,29] The penetration characteristics of SMBI studied by Dα array and CCD camera on HL-2A showed that the SMBI from the low field side consists of a slow component (SC) and a fast component (FC). The FC can penetrate more deeply than the common SMBI, and it is about 4–5 cm deeper than the SC.[23] It is the observation of FC that greatly draws our attention to develop a new method of SMBI for deeper penetration and higher efficiency. Although the FC and SC have been studied experimentally for many years, there is little research in theory and simulation to interpret it. In order to study the penetration mechanism and characteristic of the FC and SC, a module (named trans-neut) of the BOUT++ [30] code is used to simulate their penetration processes. Trans-neut is a module developed for studying the time-dependent three-dimensional (3D) neutrals and plasmas transport dynamics during fueling of SMBI or GP in real tokamak geometries.[31] It can be used to simulate, understand, and predict SMBI experiments, and its simulation results are in good agreement with the experimental data.[32] With the module, the injection efficiency and depth have been compared between tokamak fuellings of GP and SMBI,[33] and the molecular penetration depth variation with the SMBI fluxes has also been studied.[34] A new one-dimensional (1D) code named TPSMBI has also been developed recently based on a six-field SMBI model in the cylindrical coordinate and it has been benchmarked well with the trans-neut module of the BOUT++ code.[35]
In this paper, the Dα intensity is calculated by using the data in the atomic data and analysis structure (ADAS) database of the photon emissivity coefficient of Dα, which is related to the plasma density and electron temperature.[36] The simulation results will be compared with the experiments by calculating the Dα intensity. We will further study the major mechanism of the FC and SC, the influence of the initial plasma density on the FC and SC, and the influence of the injection velocity on the SC for the underlying physics. The remainder of this paper is organized as follows. The physical model is briefly described in Section
A seven-field fluid model, which couples plasma density
Equations (
In order to do the simulation and comparison with HL-2A experiments, a real HL-2A tokamak geometry of shot #9940 at 221 ms just before the first SMBI pulse is used in the simulation according to the first observation of FC and SC in Ref. [23]. Unfortunately, because the experimental data of plasma density and temperature profiles have been lost, the numerical initial profiles of plasma density and temperatures are selected to be characteristic linear profiles based on the measurement of line-averaged electron density and electron temperature in the core. The plasma density and electron temperature profiles are shown in Fig.
Numerical implementation of the boundary conditions is expressed in Ref. [31], including the radial flux-driven boundary condition, the particle recycling boundary condition, the SMBI localized boundary condition, and the sheath boundary condition. The field-aligned coordinates (x,y,z) are applied in the simulation, which are related to the usual flux coordinates (
The penetration depth of SMBI is measured by Dα array in experiment and it is defined as the distance from the plasma edge to the maximum Dα intensity region. In the simulation, the spatio-temporal evolution of the Dα intensity is calculated as follows:[1]
Dα is the photon emitted when a deuterium atom jumps from the excited state of
To calculate the photon emissivity coefficient q, we first add together the coefficients of electron-collision excitation and recombination radiation to get a total value, then the cubic spline interpolation is carried out in the space of plasma density
Both FC and SC are widely observed in lots of shots on the HL-2A tokamak. Figure
In the experiments, the velocities of FC and SC are estimated by the distance from the nozzle to the plasma edge divided by the time intervals between the two peaks of the Dα signals and the SMBI triggering signal, respectively. According to the experimental results of shot #9940, we choose 1200 m/s and 800 m/s as the initial FC and SC injection velocities, respectively. A 1 ms SMBI pulse with a fast component and a slow component injects at the low field side of the HL-2A tokamak. Since the time interval between FC and SC is about 0.15 ms in the experiment, we make the SC injection 0.15 ms later than the FC in the simulation. The simulation results are shown in Fig.
Figure
In order to study the effect of the initial plasma density
For a straightforward verification of the self-blocking effect on the penetration depth, the FC and SC are firstly supposed to be injected with the same velocity and then the FC is injected slower than the SC in the simulations.
In Ref. [23], the analysis of experimental results shows that the difference in the penetration depth between FC and SC is mainly due to the self-blocking effect. In order to verify it, we use the same injection velocity 1200 m/s for the FC and SC in the simulation. The simulation results are shown in Fig.
Since the proportion of the FC is small and its advantage in the penetration depth diminishes gradually with the increase of the plasma density, the penetration depth of SMBI is mainly determined by the SC. Therefore, in order to promote the penetration effect of SMBI, the penetration depth of the SC needs to be improved first. Because the penetration depth of the SC remains almost invariant with different plasma densities, we try to simulate the influence of the injection velocity on the SC penetration depth. The simulation results are shown in Fig.
It is very important to improve the penetration depth and the fueling efficiency of supersonic molecular beam injection (SMBI) especially for the next generation fusion devices such as ITER. Two components, a fast component and a slow component, are observed in the HL-2A SMBI experiments, and the FC can penetrate much deeper than the common SMBI does. It is the observation of the FC that greatly draws our attention to develop a new method of SMBI with deeper penetration and higher efficiency. In this paper, we have first studied and interpreted the FC and SC of SMBI on the HL-2A tokamak with the trans-neut module of the BOUT++ code. The FC and SC are observed and realized in the simulation, which are qualitatively consistent with the experimental results. The principal simulation results are summarized as follows. (i) The penetration depth difference between the FC and SC is mainly caused by the self-blocking effect of the first ionized atoms on the following ones due to the highly increased plasma density, though the injection speed may also influence the injection depths of the FC and SC. (ii) The influence of the initial plasma density on the FC and SC has been discussed. The penetration depth of the FC becomes shallower with the increase of the plasma density due to the larger ionization rate, but the penetration depth of the SC keeps almost invariant. (iii) The penetration depth of the SC is gradually improved with the increase of the injection velocity, but it is still much shallower than the FC. We should find a way to reduce the self-blocking effect by developing a new technology to make a little-transient-dense SMBI pulse (like a pellet) on tens of microseconds before the blocking effect turns on.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] | |
[31] | |
[32] | |
[33] | |
[34] | |
[35] | |
[36] | |
[37] |