Mitglied der Helmholtz-Gemeinschaft Application of a multiscale transport model for magnetized plasmas in cylindrical configuration Workshop on Plasma Material Interaction Facilities 13. April 2015 1 Institute 2 Dep. | Christian Salmagne1, Detlev Reiter1, Martine Baelmans2, Wouter Dekeyser2 of Energy and Climate Research - Plasma Physics, Forschungszentrum Jülich GmbH of Mechanical Engineering, K.U.Leuven, Celestijnenlaan 300 A, 3001 Heverlee, Belgium Outline 0. Motivation 1. Using the ITER divertor code B2-EIRENE for PSI-2 2. Simulation of PSI-2 3. Extension of the numerical model 4. Summary & Outlook 13. April 2015 2 0. Motivation Linear plasma device PSI-2 has been transferred from Berlin to FZJ last year. The modeling activities carried out in Berlin are not usable anymore and are rebuild in Jülich, using the ITER divertor code B2-EIRENE. Modeling of PSI-2 creates the possibility of an additional analysis of a plasma that resembles the edge plasma of a Tokamak in important points. That gives the opportunity to verify and improve the Code with another type of experiment. 13. April 2015 3 1. Using the ITER divertor code B2-EIRENE for PSI-2 PSI-2 Jülich Using the B2-EIRENE code for a linear device Governing equations Boundary conditions, grid and used parameters 13. April 2015 4 PSI-2 Jülich Six coils create a magnetic field B < 0.1 T. Plasma column of approx. 2.5 m length and 5 cm radius Densities and temperatures: 1017 m-3 < n < 1020 m-3, Te < 30 eV MFP of electrons indicate that fluid approximation is likely to be valid 13. April 2015 5 Use of B2-EIRENE code for a linear device Plasma source Midplane topol. equiv. Aspect ratio: a/R=∞ Direct use of B2EIRENE (SOLPS) for PSI-2 is possible, but the coordinates have to be adapted linear toroidal radial radial polar toroidal axial poloidal polar (toroidal) coordinates are neglected (symmetry is assumed) Target Target Tokamak MAST 13. April 2015 PSI-2 6 Boundary conditions, grid and used parameters First aim: Reproduction of radial profiles using all existing information about the simulation from Berlin [1] Boundary conditions: Walls perpendicular to the field lines: Sheath conditions Axis of the cylinder: vanishing gradients in Te,TI and n „Vacuum-boundary“ and anode: 1cm decay length in Te,TI and n Parameters: Pumping rate: 3500l/s Neutral influx(D2): 6.32 x 1019 s-1 Anomalous diffusion: Din = 3.0m2/s; Dout = 0.2 m2/s Perpendicular heat conduction: κe,in= 5.0 m2/s; κe,out= 11.0 m2/s Source next to anode at given temperature (Te = 15 eV; TI = 5 eV) 13. April 2015 [1] Kastelewicz, H., & Fussmann, G. (2004). Contributions to Plasma Physics, 44(4), 352-360 7 2. Simulation of PSI-2 Summary of existing results: [1] Kastelewicz, H., & Fussmann, G. (2004). Contributions to Plasma Physics, 44(4), 352-360 [2] Vervecken, L. (2010). Extended Plasma Modeling for the PSI-2 Device. Master thesis. KU Leuven Reproduction of existing numerical and experimental results Dependency on kinetic flux limiter 13. April 2015 8 Summary of existing results Modeling activities in Berlin with former B2-EIRENE Version SOLPS4.0, 1995, Summary can be found in [1] In [2] the model was rebuild, old results could already be partially reproduced. Figures: Radial profiles at two different positions, Coefficients for anomalous transport adapted to fit experiment [1] 13. April 2015 9 Reproducing existing results First results did not match old results FLIM = 0,8 „flux limiter“ was introduced into B2 to compensate kinetic effects Parallel heat conductivity is limited to: with parameter FLIM Different values of FLIM found in old input It is not possible to reconstruct, which value was used in [1] 13. April 2015 10 Dependency on kinetic flux limiter Dependency on the flux limiter indicates the importance of kinetic effects Additional free parameter influencing the parallel transport Experimental values at at least two axial positions needed Values for the flux limiter can be obtained using the comparison with experimental data or a complete kinetic model of PSI-2 13. April 2015 11 3. Extension of the numerical model Extension of the neutral particle model using a collisional radiative model an metastable states Incorporation of parallel electric currents 13. April 2015 12 Refinement Extension of the neutral model Model [1]: neutral model as used in [1] Model I: Collisional radiative model for H2+ and H2 Model II: Vibrationally excited states treated as metastable Particle and heat fluxes on the neutralizer plate strongly depend on the used model Heatflux [W] Particle flux [s-1] Model [1] 274.8 1.21 x 1020 Model I 224.2 1.45 x 1020 Model II 318.9 1.73 x 1020 Plasma density and temperature also change strongly 13. April 2015 13 Extension of the neutral Model: Recombination Reaction rates show that H2+-MAR is the most important recombination channel Most recombination takes place at neutralizer and cathode 3 body recombination and radiative recombination are unimportant in the model 13. April 2015 14 Model [1] Extension of the neutral Model: MAR Ratio Model I / Model II Model I Model II 13. April 2015 H2+-MAR rates also depend on the used model With Model I rates are overestimated in the target chamber and underestimated at the anode Vibrationally excited states have to be modeled as metastable 15 Incorporation of parallel electric currents The plasma potential is not calculated and the potential drop is only important for the heat flux, and thus for the boundary condition for the electron energy. For equal electron and ion temperatures it can be approximated as: Since the variation with the temperatures is small, the potential drop is provided as a constant input parameter 13. April 2015 16 Incorporation of parallel electric currents In “extended B2” [3] currents are incorporated. Then, the potential drop depends on the current and changes to: That also changes the electron energy flux In this version the possibility to set the wall potential for each wall differently exists. That makes it possible to bias the neutralizer wall 13. April 2015 [3] Baelmans, M. (1993). Code Improvements and Applications of a two-dimensional Edge Plasma Model for toroidal Fusion Devices. Katholieke Universiteit Leuven. 17 Incorporation of parallel electric currents: Code verification Normalized current density: Normalized heat flux density: Heat flux and electric current behave exactly as expected when the potential is changed 13. April 2015 18 Incorporation of parallel electric currents When no potential is applied, the direction of the current is depending on the radial position The direction of the electric currents can be influenced by changing the potential at the neutralizer plate Direct influence of strong current densities on the electron temperature can be seen 13. April 2015 19 Incorporation of parallel electric currents Ion temperature and plasma density do not change significantly Electric current on the neutralizer plate changes and reaches a saturation for negative potentials of the neutralizer Heat flux on the wall also changes and has a minimum near the floating potential Minimal heat flux still larger than in case of disabled currents Heatflux not minimal, if total current vanishes 13. April 2015 20 4. Summary & Outlook Summary Numerical model was rebuild and old numerical and experimental results were reproduced using the ITER divertor code B2-EIRENE. A dependency on the kinetic flux limiter was found. The neutral particle model was improved and it was shown that the correct treatment of the vibrationally excited states is crucial in the model. B2-EIRENE can account for parallel electric currents in a linear machine Outlook: Classical drifts and diamagnetic currents will be introduced. Experimental data is needed to compare target biasing effects and to cope with the dependency on the kinetic flux limiter. Neutral particle simulation can be further extended. The model of the reactions at the walls has to be checked. Impurities will be introduced. 13. April 2015 21 Thank you for your attention! 13. April 2015 22 Governing equations Continuity equation: Parallel momentum equation: Radial momentum equation: 13. April 2015 23 Governing equations Electron and ion energy equations: 13. April 2015 24