Predictions for EAST Divertor Performance

5 downloads 56197 Views 621KB Size Report
During the last decade, some expected benefits of a closed divertor have been confirmed ... The schematic EAST SN and Connected Double Null (CDN) divertorĀ ...
JP0655031

Predictions for EAST Divertor Performance S. Zhu ''*, R. Hiwatari2, A. Hatayama3, Y. Tomita 4, 1

Institute of Plasma Physics, Chinese Academy of Sciences, Hefei, P.R.China 2 Central Research Institute of Electric Power Industry, Tokyo, Japan 3 Faculty of Science and Technology, Keio University, Yokohama, Japan 4 National Institute for Fusion Science, Toki, Japan

Abstract A detailed study of the divertor performance in EAST has been performed for both its double null (DN) and single null (SN) configurations. The results of application of the SOLPS (B2-Eirene) code package to the analysis of the EAST divertor are summarized. In this work, we concentrate on the effects of increased geometrical closure and of magnetic topology variation on the scrape-off layer (SOL) and divertor plasma behavior. The results of numerical predictions for the EAST divertor operational window are also described in this paper. A simple Core-SOL-Divertor (C-S-D) model was applied to investigate the possibility of extending the plasma operational space of the low hybrid current drive (LHCD) experiments for EAST. 1. Introduction The scientific mission of EAST is to explore the reactor relevant regimes with long pulse lengths and high plasma core confinement, and to develop and verify solutions for power exhaust and particle control in steady state. To accomplish this aim, both its toroidal and poloidal coils are superconducting magnets and the plasma current (1 MA) will be sustained over long periods of time T pulse = 60 - 1000s by LHCD. EAST is designed to have shaped plasma cross-sections and can operate in DN and SN divertor configurations. The EAST divertor should be designed to accommodate the heat load due to the combined heating power of 7.5MW in long duration discharges. During the last decade, some expected benefits of a closed divertor have been confirmed by the experiments conducted on most existing divertor tokamaks (Alcator C-mod, ASDEX-U, DIII-D, JET and JT-60U, etc) [1]. These experimental devices have modified their divertors to increase the "closure", i.e. to decrease the fraction of recycled neutrals escaping from the divertor region. The experimentally proved effects of the divertor geometry have been considered in the design process of the EAST divertor. In order to increase the degree of closure, the divertor structure has to be designed to minimize its conductance for neutral leakage from the divertor region into the main chamber. So the divertor structure in EAST is deep and consists of inner

157

and outer vertical target plates, tightly fitting side baffles and a dome baffle in the private flux region, which is close to the divertor concept developed by the ITER JCT [2] The purpose of this work is to predict the details of the SOL and divertor plasma in the different divertor topologies and to assess the effects of the divertor geometry considered. An assessment of the divertor operational windows in EAST is also described. Two-dimensional numerical calculations with the SOLPS code package [3-5] have been performed. The results of numerical predictions for the divertor are to be described in this paper. A simple Core-SOL-Divertor (C-S-D) model has been applied to investigate the possibility of extending the plasma operational space of the LHCD experiments for EAST. 2. Simulation Model Tokamak plasma performance generally improves with increased shaping of the plasma cross section. The poloidal field coil system of EAST has the capability to accommodate SN and DN configurations. The major parameters of these configurations are listed in Table 1 and Table 2 separately. Here for the DN configuration, the magnetic connection length is about 31m, and the divertor depth, i.e. the distance between the X-point and the outer target along the separatrix is about 0.30 m. The SOLPS code package has been used in the simulations. It couples a multi-fluid plasma code B2 with a Monte-Carlo neutral code Eirene and is capable of taking into account realistic EAST divertor and SOL geometry. The schematic EAST SN and Connected Double Null (CDN) divertor geometries and the computational meshes used in the present study are shown in Fig. 1. The targets and baffles fit the magnetic geometry tightly to minimize the neutral back flow into the main chamber. The shape of the outer side baffle follows the magnetic flux surface which is at a distance of about 3.5 cm away from the separatrix, measured at the outer midplane. The magnetic equilibriums are generated using EFIT code and are the basis of the numerical grid generation for the SOLPS. The computational domain for the divertor predictive studies covers the whole SOL and divertors. A small region of the plasma core periphery and the private flux region are also included. The whole computational domain is resolved into 120 poloidal divisions and 24 radial divisions. Only hydrogen is used as plasma species in the present work and C is generated self-consistently by sputtering. For the standard operating scenarios (without impurity injection), we assume that 80% of the total power flows into the entire SOL. The anomalous perpendicular transport model used in the present study is constant in space, with the thermal diffusivities x iĀ± = x eJ_. Ultra-long discharges will be achieved on the superconducting tokamak EAST, so the wall pumping effect is ignored and the recycling coefficient R is set to 1.0 for the divertor plasma facing components and the vacuum vessel wall. 3. Results of the SOLPS prediction

158

3.1 Effect of the vertical targets Vertical targets are adopted in the EAST divertor. The neutrals produced at the target plates are preferentially reflected towards the separatrix and hence ionization is enhanced near the vicinity of the separatrix. Figure 2 shows the 2-D distribution of (a) the neutral density and (b) the ionization source (rf" ions m'V 1 ). Since the power is mainly conducted through the region close to the separatrix, this vertical geometry effect is beneficial to improving the power exhaust. As a result, the peak heat flux is reduced and the profile is broader, in comparison with our previous modelling with divertor target plates normal to flux surfaces [6]. The broader heat flux, the more peaked electron density and "inverted" temperature profiles across the lower outboard divertor target are shown in Fig. 3. 3.2 Effect of divertor topology The poloidal field coil system of EAST allows it to run in SN or DN magnetic configurations for more flexibility in experiments. A single null divertor is a configuration with only one active X-point and the outer separatrix far away or even outside the vacuum vessel. If both X-points are active, i.e. both separatrices coincide, a connected double null (CDN) configuration is created. The heat flux sharing by the divertors will be strongly affected by the variation in the magnetic topology of the divertor. To assess this effect, we have performed calculations for several operating points. At each point, the density at the core-edge interface (CEI) and the power flux carried by electrons and ions across CEI are specified. A comparison of the electron temperature and peak heat flux between SN and CDN is shown in Table 3. For all cases, the peak heat loads in SN are much higher than that in CDN at about the same operating points. As the configuration transitions from SN to CDN divertor, there exists a configuration of disconnected double null (DDN). In DDN, if the distance A sep between both separatrices at the outer midplane is comparable to the SOL width of the parallel heat flux, then a significant part of the heat flux can still flow along the outer separatrix to the second divertor. Figure 4 shows the contours of the electron temperature and the total heat flux for a DDN. Sharp change in divertor load due to variation in the magnetic topology of the divertor has been observed in DIII-D and also in our modeling of EAST. Therefore, a sophisticated control over the separatrix distance A sep at the outer midplane is essential. 3.3 Divertor operational regimes To reduce the power load and erosion of the divertor target plates is the main issue in the design of the EAST divertor. Operating in the high recycling or detachment regime can effectively decrease the heat flux flowing to the target and

159

make the electron temperature low at the target plate. Divertor regimes are very sensitive to the midplane separatrix electron density. For higher densities, the divertor has easier access to the high recycling or detachment regime. According to the Greenwald limit, EAST will be able to run safely with the line average densities ne up to 1.0 X1020 m'3 in Ohmic discharges. But the efficiency of LHCD requires operation at a much lower density. To explore the possibility of achieving various operational regimes and to study the divertor behavior in these scenarios, a density scan has been carried out for CDN. At a quite low density of about ne, sep = 0.7 X 1019 m"3, only the low recycling regime can be attained. In this regime, due to the high parallel heat conduction, the electron temperature shows little drop along the field lines and hence is high at the target. With low densities, recycling losses can be ignored, so the peak heat flux at the target is higher than 5 MW/m2, which exceeds the engineering constraint. As the midplane separatrix density is increased, significant gradients of plasma profiles along the field-line can be observed, indicating the accession to the high recycling regime. To reduce the peak heat flux to lower than 3.5 MW/m2, the midplane density ne, Sep should be increased further to about 1.4X 1019 m~3. In this regime, due to the strong ionization sources from recycling neutrals, a plasma with high density and therefore low temperature is formed close to the plate, which reduces target sputtering and makes the Zeff an ideal value of 1.4. Our modelling indicates that, for EAST, the transition to power detachment occurs at the line average densities ne ~ 7.8 X1019 m'3, that is about 80% of the Greenwald limit and is much higher than the density limit posed by the LHCD efficiency. Consequently, additional approach such as gas puffing or impurity seeding should be adopted to attempt detachment. But the modelling results at this high density reveal the effect of the divertor geometry on detachment behavior. In this case, the electron temperature is low enough (< 4 eV) throughout most of the inner target and even lower than 2 eV at the separatrix. Although the separatrix temperatures at the outer target become very small below 2 eV, the outer SOL remains at high temperature > 10 eV and thus keeps attached. This may suggest that detachment in EAST starts from the separatrix due to the effect of the vertical divertor geometry (Fig. 5). The result also indicates that complete power detachment is attained for the inner divertor, whereas partial detachment is attained for the outer divertor. 3.4 Extension of the operational space Consistency between the edge plasma and the core plasma operation is an important issue for the design of fusion devices. A simple Core-SOL-Divertor (C-S-D) model was developed to investigate qualitatively the overall features of the operational space for the integrated core and edge plasma [7]. This model was applied to assess the possibility of extending the plasma operational space of the LHCD experiments for EAST [8]. By using the C-S-D model, it is revealed that gas puffing

160

is an effective method to extend the operational space toward both lower &p and higher Qin region, where