Multi-machine Modelling of Divertor Geometry Effects Alberto Loarte EFDA CSU -Garching Acknowledgements: K. Borrass, D. Coster, J. Gafert, C. Maggi, R. Monk, L. Horton, R.Schneider (IPP), A.Kukushkin (ITER), G. Matthews, M. Groth, (JET), G.Porter (LLNL), N. Asakura (JAERI), D. Stotler (PPPL), B. Lipschultz, B. La Bombard (MIT), R. Pitts (CRPP) - -
Outline of the talk. Introduction 2. Divertor Geometry and Detachment (Power Losses + Density Limits) 3. Divertor Geometry and Pumping (Hydrogen + Impurities) 4. Divertor Geometry and Plasma Flows (Hydrogen + Impurities) 5. Conclusions - 2 -
. Introduction Divertor Geometry affects Recycling Sources Divertor Density and Temperature Flow Patterns in Divertor Divertor Radiative Losses Power Recicled Particles 2-D Plasma Edge Simulation Codes used to analyse Effects of Divertor Geometry - 3 -
Divertor Geometry affects Transport in the Divertor itself DIII-D X-point Scans (D. Hill APS 92) Low X-point High X-point qdiv,out (MW/m2) qparallel,mp (MW/m 2 ) 3 2 8 6 4 2.2 (b) (c) (d) 4.8MW 6.8MW Ptot,div/Pbeam P..5..5.2.25.3.35 X-point height (m) - 4 -
Can account for ~ 2 Power Flux Reduction near Separatrix ASDEX-U Div I Æ Div II (ASDEX-U, R. Schneider, B2-Eirene, IAEA 98) Divertor energy balance: For each flux bundle Div I Div II Private flux Energy flux into outer divertor Private flux MW Energy flux onto outer divertor Integrals between target plate and divertor entrance Energy losses from neutrals Contribution of the divergence of radial energy fluxes.8.8 Separatrix Separatrix Private Private flux flux.4.4 MW 2 3 Flux bundle label (B2 grid) 2 Flux bundle label (B2 grid) Decrease depends on anomalous Transport (ASDEX-U, D. Coster, B2-Eirene, EPS 98) fetx_max_outer/psep (/m^2).6.5.4.3.2. 4MW transport variation (base: D=, Chi=2).25.5 transport coefficient multiplier - 5-3e9 4e9
2. Divertor Geometry and Detachment (Power Losses + Density Limits) Effects of Divertor Geometry on Detachment seen in Alcator C-mod, JET, ASDEX-U, JT-6U, TCV Modelling (B2-Eirene, EDGE2D-NIMBUS & UEDGE) reproduces Features in Experiment Vertical Divertor promotes Detachment close to Separatrix Parallel Current (A/m 2 ) 2.5 6 (Alcator C-mod, F. Wising, UEDGE, PSI 96) 2 6.5 6 6 5 5 Parallel Current Attached_Modelled Parallel Current Attached_Exp Parallel Current Detached_Modelled Parallel Current Detached_Exp.5. Radial Distance ρ (m) - 6 -
Easier Separatrix Detachment does not lead to Lower Density Limit (Æ total Detachment) (ASDEX-U, R. Schneider, B2-Eirene, PSI 98) distance along outer target plate [m].78.76.74.72.7.68 #862 ASDEX-U Div I & Div II Divertor T e SOL PFR Div I.5 2. 2.5..5 2. time [sec] Div II #967 2.5 3. 3.5 time [sec] ASDEX-U Div I & Div II calculated Ion Flux.35.3.25.2.5 2 5 5 T_e, div [ev] separatrix particle flux density (2MW) integrated particle flux (2MW) 22-2 - ( m s ) 8 4 Inner DivII Inner DivI Outer DivI Outer DivII 2 3 4 9 separatrix density ( m -3 ) 23 - ( s ).5..5 Outer DivI Inner DivI Outer DivII Inner DivII 2 3 4 9 separatrix density ( m -3 ) - 7 -
Most pronounced Geometry Effects seen in JET-Mk IIA Horizontal ÅÆ Vertical Discharges (JET, K. Borrass, B2-Eirene, EPTSW 98) JET-Mk IIA Vertical Modelling Experiment Modelling <n e > I [A/m 2 sat ] 4 3 2 () 58 6 62 64 66 time [s] (7) (9) (8) Separatrix () (9) Plasma Recombination Sink (7) (8) - 8 -
Divertor Geometry promotes Recombination away from Separatrix for JET-Mk IIA horizontal Discharges (JET, K. Borrass, B2-Eirene, EPS 97) Corner Effect JET- Mk IIA Horizontal Plasma Recombination Sink JET- Mk IIA Vertical Plasma Recombination Sink - 9 -
Detachment characterised by DOD or Degree of Detachment (A. Loarte, Nucl. Fusion 98) High recycling Scaling : scal Id Id DOD... = DOD >> Æ Detachment measured I D.O.D. (Inner) d scal (JET, R. Monk, EPS 97) JET Pulses Nos. 36862, Distance from Separatrix cm.6cm 6.cm.9cm Early Onset of Detachment 3754 Divertor MARFE = C n Horizontal e 2 D.O.D. (Inner) Distance from Separatrix cm 2.cm 4.9cm Vertical Divertor MARFE 2. 2.2 2.4 2.6 2.8 3. 3.2 3.4 3.6 Line Average Density ( 9 m 3 ) - -
ASDEX-U Divertor Geometry promotes Recombination for Div II with respect to Div I at lower Densities (ASDEX-U, R. Schneider, B2-Eirene, PSI 98) Photons [s- cm-2 sr-] 3e+8 2e+8 H α emission (standard ohmic shot) Div II Div I e+8 e+ -45-55 -65 Deflection [degree] deflection angle -75 divertor II scanning mirror - -
Septum in JET-Gas Box isolates in/out Divertor Gas Puffing changes Detachment Asymmetry (JET, C. Maggi, EPS 99) JG99.96/2c Inner divertor Fuelling Outer divertor Fuelling C/SFE/LT Pulse No: 45535, 4495 (MW) ( 22 s - ) ( 22 s - ) ( 9 m -3 ) 3 2 3 2 4 2 4 2 P IN <n e >, Central Inner fuelling Outer divertor fuelling Inner divertor fuelling I out Outer fuelling I in I out I in TOT P RAD 6 7 8 9 2 2 22 23 Time (s) JG99.96/5c - 2 -
Modelling is consistent with Septum Effect (JET, C. Maggi, EDGE2D-NIMBUS, EPS 99) - 3 -
Large Difference in Divertor Power Deposition between Div I and Div II in ASDEX-U power density /MW/m 2 5 4 3 2 (ASDEX-U, Kaufmann, IAEA 98) 2 MW, DIV II 5 MW, DIV I.5 MW, DIV I.5..5 length across targetplate / m q = 4, MA.5 MW, DIV II..5 Prad,divertor/Pheat + Divertor I + Divertor II...5e2.e2 Line averaged electron density (m-3) Change Div I Æ Div II in Agreement with Modelling (ASDEX-U, R. Schneider, B2-Eirene, PSI 98) Experiment Modelling P_heat =.5 MW NBI, Ý = 5e9 /m**3 2 MW input power, density ramp power density / MW/m 2 power density / MW/m 2.2.8.4..2.8.4. Div I Div II outer divertor..4.8 distance from separatrix / m Div I Div II inner divertor..4.8 distance from separatrix / m power density / MW/m 2 power density / MW/m 2.2.8.4..2.8.4. outer divertor Div I..4.8 distance from separatrix / m Div II inner divertor Div I Div II..4.8 distance from separatrix / m - 4 -
-.4 z[m] -.62 -.85 -.8 -.3 EFDA ASDEX-U Divertor Geometry creates optimum Conditions for extended Region of Radiation (ASDEX-U, R. Schneider, B2-Eirene, PSI 98) 5 Experiment B2-Eirene Radiation losses [MW/m¾] 2.4 9.8 7.2 4.6 2. Bolometer line integral [MW/m 2 ] 2.. -.2 6 7 Private Flux8 2 'yaa' 'yab' Measurements 'yac' 'yae' 'yaf' 'yag' B2-Eirene -...2 Distance Separatrix-Bolometer [m] - 5 -
Exact Change in Radiation Div I Æ Div II due mainly to access to Low Divertor T e regimes (ASDEX-U, D. Coster, B2-Eirene, EPS 99) Div. sep. Te (ev) Pure Plasma Simulations DivI 2MW outer Outer target DivII 2MW outer DivI 8MW outer DivII 8MW outer.. 2e+9 4e+9 6e+9 8e+9 e+2 midplane separatrix density Div. sep. Te (ev) DivI 2MW inner Inner target DivII 2MW inner DivI 8MW inner DivII 8MW inner.. 2e+9 4e+9 6e+9 8e+9 e+2 midplane separatrix density - 6 -
Radiation Fraction in ASDEX-U Div II (ASDEX-U, D. Coster, B2-Eirene, EPS 99) Fraction of power reaching target.8.6.4 %.2 3% 5%. Target temperature T=eV T=3eV.8 T=5eV.6.4.2... C chemical sputter yield - 7 -
Carbon Flow Pattern induced by Geometry increases Radiation Efficiency (ASDEX-U, R. Schneider, B2-Eirene, IAEA 98) Carbon mass flow: flow reversal close to separatrix forward flow in the outer SOL - 8 -
3. Divertor Geometry and Pumping (Hydrogen + Impurities) DIII-D Experiments Æ Strong Dependence of Hydrogen Pumping on Divertor Geometry (DIII-D, R. Maingi, PSI 94) Normalised Neutral Pressure @ Pump.9.8.7.6.5.4.3.2 Experiment Horizontal Target Bias Ring -5 5 5 Distance from separatrix to Bias Ring (cm) - 9 -
Strong dependence not seen in other Experiments (JET-Mk I, JET-Mk IIA, C-mod, ASDEX-U) - 2 -
Physics Basis for Difference found with 2-D Codes (Ballistic Flux ÅÆ Diffusive Flux) (A. Loarte, EPS 97).8 Separatrix SOL Distance separatrix slot x Pumping baffle Geometrical Probability PGeo.7.6.5.4.3.2 Slot width =.5 λ ion Slot width =. λ ion Slot width = 2. λ ion Slot width = 4. λ ion Private flux region Slot width Pumping slot JG.42/3c. 2 4 6 8 Distance to pumping slot (normalised to λ ion ) JG.42/4c Mod 3. 2.5 (DIII-D, R. Maingi, Nucl. Fusion 99) Variable Ion Current Model Fixed Ion Current Model Plenum Pressure (mtorr) 2..5..5..5.55.6.65.7.75 Outer Strike Point Location - 2 -
Diffusive Flux leads to weak Strike Point Position Pumping Dependence (A. Loarte, EDGE2D-NIMBUS, EPS 97) Normalised Neutral Pressure @ Pump.8.6.4.2 Bias Ring DIII-D_Press+EDGE2D Horizontal Target Experiment_Low_ne Experiment_High_ne EDGE2D-NIMBUS -5 5 5 Distance from Separatrix to Bias Ring (cm). JET MkII.2.4.6 Z (m).8 2. 2.2 2.4 2. 2.2 2.4 2.6 2.8 3. 3.2 3.4 R (m) JG97.23/2c - 22 -
He, Ne do not Charge-Exchange so Pumping Subject to stronger Effects of Divertor Geometry (ASDEX-U, D. Coster, B2-Eirene, PSI 96) additional pumping baffle pump Distance along Target (m) In ASDEX-U Div I Ne is more compressed than He Compression factor C i. Experiments Ne Helium B2-Eirene modelling Ne He 22 23 Neutral gas flux density [m -2s -] - 23 -
Change from Div I to Div II expected from ballistic Transport of He and Ne to the Pump (ASDEX-U, R. Schneider, B2-Eirene, IAEA 98) Neon Helium Improved helium exhaust, worse compression of neon (ASDEX-U, H. - S. Bosch, B2-Eirene, He Workshop 99) Deuterium Compression He & Ne Pumping - 24 -
Ballistic Effects for He also seen in JET Mk II He enrichment (JET MK II, M. Groth, He Workshop 99).2..8.6.4.2 L-mode Horizontal Corner Vertical Vertical H-mode Horizontal Vertical Vertical 2 3 4 Subdivertor pressure ( 3 mbar) Enrichment decreases with increasing n e + Detachment (JET Mk II Gas Box, H. Guo, EDGE2D-NIMBUS, Nucl. Fusion 2) Edge Enrichment Enrichment Enrichment L-mode : He Ne H-mode : He Ne Ar Core.. 2 3 4 Subdivertor Pressure ( -3 mbar) Ne He Edge Code..5..5 2. Separatrix density ( -9-3 m ) JG99.438/7c - 25 -
For ITER He enrichment increase with Detachment Helium Concentration Separatrix De-enrichment due to Thermal Forces Going to Partial Detachment helps He Pumping (Increased Neutral He Penetration to PFR) - 26 -
Detachment helps for He Pumping in ITER-FEAT Helium Concentration @ Plasma Edge. n_he_core_scaled_full_pumping Detachment Outer Divertor Detachment Inner Divertor. 3 3.2 3.4 3.6 3.8 4 Separatrix Density ( 9 m -3 ) Detachment increases D Pressure & He Enrichment P_PFR_Deuterium_Full_Pumping RC/RTO-ITER He-Enrichment Deuterium Neutral Pressure (Pa) 5 4 3 2 Detachment Inner Divertor Detachment Outer Divertor.5.4.3.2 Helium Enrichment @ Plasma Edge 3 3.2 3.4 3.6 3.8 4 Separatrix Density ( 9 m -3 ). - 27 -
Experiments + Modelling for DIII-D show different D and He Behaviour (DIII-D, A. Loarte, B2-Eirene, PET99) For D standard Detachment Picture Electron Temperature Electron Density Strong Recombination but Neutrals don't reach Main Plasma Ionisation/Recombination Neutral Density - 28 -
For He Main Bulk Plasma collapses before Recombination can be achieved (He penetration through X-point) Electron Temperature Helium Atomic Density Ionisation/Recombination Momentum Losses - 29 -
4. Divertor Geometry and Plasma Flows (Hydrogen + Impurities) Divertor Particle Flows are determined by Sources (Divertor Geometry) but also by Drifts High Density Operation can lead to Flow Reversal in the Divertor Vertical Divertors are more prone to Flow Reversal due to Recycling Pattern (Alcator C-mod, A. Loarte, B2-Eirene, PSI 98) - 3 -
Momentum Transport across the SOL between normal and reversed Flow Regions leads to Separatrix overpressure at Divertor "Death Ray" in Alcator C-mod (A. Loarte, EDGE2D-NIMBUS, PSI 96, Stotler, B2-Eirene, PSI 98) Total Pressure (Pa) 3 2 Upstream Divertor Total Pressure (Pa) 6 4 2 "Death Ray" Upstream Divertor - -5 5 5 2 ρ(mm) - -5 5 5 2 ρ(mm) p tot (Pa) - -2-3 -4-5 DR No DR Momentum Sources (N). -.2 -.4 -.6 DR S No DR S p p S n S n -.8 2 3 4 5 ρ (mm) - 3 -
Flow reversal occurs also in Horizontal Divertors (DIII-D, J. Boedo, G. Porter, UEDGE, APS 99).5.5. Private Region SOL Core..5.5 Mach. Mach. -.5 -.5 -. -. -.5. 2. 4. 6. 8.. 2. 4. 6. Z(cm) -.5. 2. 4. 6. 8.. 2. 4. 6. Z(cm) 942, Attached 75 Outer (fl8-nog) Leg -.6 Z [m] -. -.4..4.8 R [m] - 32 -
Flow reversal disappears at Detachment in Horizontal Divertors (DIII-D, J. Boedo, G. Porter, UEDGE, APS 99) 6. 6 4. 6 Private Region SOL Core 6. 6 4. 6 SOL Core @ 2 Vflow(cm/s) 2. 6. -2. 6 Vflow(cm/s) 2. 6. -2. 6-4. 6-4. 6-6. 6. 2. 4. 6. 8.. 2. 4. 6. Z(cm) -6. 6. 2. 4. 6. 8.. 2. 4. 6. Z(cm) - 33 -
Impurity Flow Reversal can occur without Deuterium Flow reversal Thermal Force wins over Friction (DIII-D, G. Porter, UEDGE, APS 98) Mach number.4.2.8.6.4.2 2 -.4..28.57.85 R (m) -.2.5..5.2.25 Z [m] 3 experiment UEDGE 4 5 6 7 Z [m] 8 9 -.6 -. -.4..4.8 R [m] - Mach number CII flow, attached 4 4 Attached phase -.6 u (C ) [m/s] -4 4-8 4 Experiment UEDGE Z [m] -. -.4..4.8 R [m] -.2 5.4.5.6.7-2.e+5.e+ 2.e+5 Parallel velocity [m/s] - 34 -
In ASDEX-U Div II Impurity Flow Reversal occurs @ Detachment in Agreement with Predictions (ASDEX-U, J. Gafert, D. Coster, R. Schneider, B2-Eirene, PSI 98) -.6 -.8 CIII-emissivity photons [ m 3 sr s ] -.6 -.8 Parallel velocity v (km/s) +5 +4 Measured PU4 Modelled PU4 z (m) -. POU POU2 POU3 POU4 -. POU POU2 POU3 POU4 +3 +2 + - + 2 Km/s + 5 Km/s - 24 Km/s -7 Km/s -.2 -.2-2 -3 C 2+ C 2+ -4-5 -.4.4.5.6.7.8.9 R (m) -.4.4.5.6.7.8.9 Carbon mass flow: flow reversal close to separatrix forward flow in the outer SOL - 35 -
sat / Xp Xp EFDA @ Detachment after ionisation Front in Divertor M (ASDEX-U,D. Coster, B2-Eirene, EPTSW 98) Plasma particle source ( m -3 s -) Mach number.5..6. -.5 -. -.5 e24 e23 e22 e2 -e2-e22 ionization recombination Well documented in JT-6U (JT-6U, N. Asakura, Nucl. Fusion 99) (A m -2 ) 6 5 ne = 3.2x 9 m -3 Xp MARFE large radiation loss j Xp sat (up) 4 ne = 2.6x 9 m -3 ne =.8x 9 m -3 Xp MARFE Attached Detached (down) j sat (up) T e (ev) 2 Detached Mach ~ j Xp. - 36 - flow towards target. 4 8 2 Distance from separatrix (mm) Large Mach
5. Conclusions Predicted Effects of Divertor Geometry on Detachment seen in most Experiments Strong Dependence of Separatrix Detachment on Geometry but weak Geometry Effect on L-mode Density Limit ( Total Detachment ) Divertor Geometry can enhance Radiation Losses for Carbon (ASDEX-U Div I Æ Div II) Divertor Geometry Effect on Hydrogen and Impurity Pumping understood in Terms of ballistic and diffusive Fluxes to pumping Plenum Divertor Geometry Effect on Hydrogen and Impurity Flows from 2-D Codes seen in Experiment: Hydrogen and Impurity Flow Reversal and Momentum Transport ("Death Rays") Modelling with Drifts needed for better understanding Divertor Geometry Effect on Midplane n e Profiles unclear ÅÆ SOL Flows+Main Chamber Recycling - 37 -