UEDGE Modeling of the Effect of Changes in the Private Flux Wall in DIII-D on Divertor Performance N.S. Wolf, G.D. Porter, M.E. Rensink, T.D. Rognlien Lawrence Livermore National Lab, And the DIII-D team at General Atomics Presented at the APS-DPP Meeting, Nov. 17, 1999 APS-DPP 1999 1
MOTIVATION: determine the effect of changes in the DIII-D upper divertor 1.5 Shot 9244, 375 ms PUMP PUMP 1.5 NEW RDP Shot 9244, 375 ms PUMP.5 Thomson.5 Thomson Z (m) CER Z (m) CER -.5 -.5 PUMP Divertor Thomson Divertor Thomson Open Divertor Lower Single Null [used for validation] -1.5 1. 1.5 2. 2.5 R (m) RDP or Closed Divertor Upper Single Null -1.5 1. 1.5 2. 2.5 R (m) Private Flux Region Dome In the new RDP 2
Upper divertor (RDP) now has a dome in the private flux region and a pumped inner leg The new RDP New Inner cryopump Existing outer cryopump New Private Flux Region Baffle 3
METHOD: simulate the SOL and divertor plasma with the UEDGE code! UEDGE solves equations for ion continuity, energy and momentum on 2-D domain, simulating the plasma from: -- 96% flux surface to limiter surface in radial direction, -- plate to plate in poloidal direction! Hydrogenic ion fluid (Braginskii equations)! Navier Stokes fluid equations for neutrals! Collisional-radiative hydrogenic rates! Turbulence-driven perpendicular (radial) transport! Carbon source is physical and chemical sputtering from walls, using Univ. Toronto model [Davis & Haasz, J. Nucl. Mater. 241-243, 37,1997]! Force balance equations for 6 ionization states of carbon -- diffusive neutral carbon fluid -- non-equilibrium coronal impurity radiation 4
VALIDATION: simple radial transport model agrees with midplane n e and T e profiles! UEDGE uses constant turbulencedriven radial diffusion coefficients: D =.1 m 2 /s χ i = χ e =.4 m 2 /s! Radial profiles match experimental results for the lower single null plasmas shown! These and many other results indicate that UEDGE models the DIII-D experiment well 5
SUMMARY of our SIMULATIONS for the new RDP! 1) Core carbon density (and hence Z eff ) increases with increasing carbon yield! 2) Core carbon density increases with core heating power! 3) The radiation fraction is much less than that for similar lower single null plasmas! 4) Inner and outer pumping is not independent. 6
Specific UEDGE assumptions for new RDP divertor modification simulations! Density at 96% surface specified; n e (core) = 2.7x1 19 m -3! Radial power across 96% surface in ions and electrons specified! Particles removed in two ways: 1) neutral albedo at the plate, A P =.98, simulates cryopumping at baffle entrances 2) neutral albedo at walls A w =.97 (wall pumping)! Spatially constant particle and thermal perpendicular diffusivity! Carbon sputtered from walls (as fraction of Davis & Haasz model) 7
We examine the effect of changing carbon yield on core impurity density in new RDP! WHY? Experiments on DIII-D show a nearly constant core impurity concentration as the lower divertor apparent yield has decreased in seven years of operation [see D. Whyte s poster this session]! UEDGE Modeling Results:! Lower Divertor: Impurity density remains nearly constant as the carbon yield is lowered {agrees with experiment} [see G. Porter, et al. and also P. West, et al. at this session]! Upper Divertor (New RDP+PF Dome): Impurity density decreases as carbon yield is lowered {new prediction} 8
KEY RESULT: Core impurities insensitive to sputtering with lower divertor, but not RDP RDP with PF Dome & Lower Single Null Divertor Impurity Density at Midplane [m -3 ] 2.5 1 18 2 1 18 1.5 1 18 1 1 18 5 1 17 RDP RDP+Dome: nc=2.7e19 LSN: nc=4.5e19 RDP+Dome: nc=4.5e19 RDP+Dome: nc=3.5e19 LSN.2.4.6.8 1 1.2 Sputtering Yield Multiplier [Y/Y(Haasz)] 11/1/99 9
Core impurity density is linear with impurity source strength with new RDP Core Impurity Density (midplane) [m-3] 2 1 18 1.5 1 18 1 1 18 5 1 17 Imp_density Imp_source.2.4.6.8 1 6 5 4 3 2 1 Total Impurity Source Current [A] Sputtering Yield Multiplier [Y/Y(Haasz)] 1
Is core impurity level much influenced by parallel transport in the new RDP?.6 C 4+ C 5+ C 6+! Carbon is sputtered from ion and neutral impact on all walls:! Y/Y(Haasz) =.18! C 4+ flows freely from inner divertor and is attenuated from outer divertor! There is some C 4+ accumulation due to local parallel potential wells! Carbon enters the core as C 4+ via radial diffusion, and exits as C 6+! The potential integrates the parallel net force, which is a balance between drag and T i forces RADIAL FLUX ni(c 4+ ) Γ r[α/ m 2 ].4.2 -.2 X-POINT X-POINT -.4 1 2 3 4 5 6 INNER DIVERTOR 1 1 18 8 1 17 6 1 17 4 1 17 2 1 17 ni(c 4+ ) C 6+ C 4+ out of core into core Distance from ISP [m] Y/Y(Haasz) =.18 Potential H18q3 2-2 -4-6 -8-1 -12 Impurity Potential (C 4+ ) -14 1 2 3 4 5 6 Distance from ISP [m] H18q3 11
Potential is not changed by higher sputtering yield in new RDP! Carbon is sputtered from ion and neutral impact on all walls:! Y/Y(Haasz) =.76! As before, carbon enters core as C 4+ via diffusion, and exits as C 6+ (at much larger values than for low yield)! But the parallel potential is not much changed from lower sputtering yield value, implying transport is not changed.! Therefore the higher core carbon density is dominated by the higher carbon source current, not parallel transport forces. RADIAL FLUX ni(c 4+ ) [m -3 ] Γ r [Α/ m 2 ] 2 1.5 1.5 -.5-1 C 4+ C 5+ C 6+ X-POINT X-POINT -1.5 1 2 3 4 5 6 1 1 18 8 1 17 6 1 17 4 1 17 2 1 17 ni(c 4+ ) C 6+ C 4+ out of core into core Distance from ISP [m] Y/Y(Haasz) =.76 H76q3-1 -12-14 1 2 3 4 5 6 Distance from ISP [m] Potential 2-2 -4-6 -8 H76q3 Impurity Potential (C 4+ ) 12
Core impurity density increases with the core heating power in the new RDP RDP with PF Dome: n c = 2.73E19, Sputt. Mult. =.5 2.5 1 18 Imp_density Imp_source 8 Impurity Density [m -3n] 2 1 18 1.5 1 18 1 1 18 5 1 17 Z eff =2.86 Z eff =2.7 7 6 5 4 3 2 1 Impurity Source [A] 2 4 6 8 1 12 14 16 18 Core Power [MW] 11/6/99 13
Radiation fraction remains below 4% for the new RDP.5! Radiation fraction increases with carbon yield (at fixed core power, P c = 9 MW)! Radiation fraction is higher in Radiated Power/Core Power.4.3.2.1 LSN (usually above 7%).2.4.6.8 1 1.2 Sputtering Yield Multiplier [Y/Y(Haasz)].5 Radiated Power/Core Power.4.3.2.1 Radiation fraction remains between 25% & 35% as core power increases (at fixed yield =.5) 2 4 6 8 1 12 14 16 18 Core Power [MW] 14
Inner and outer pumping is not independent in the new RDP! Core density fixed: n c = 2.73x1 19! Core Power fixed: P c = 9 MW! Carbon yield multiplier =.25 4 RDP vs PF Dome Pumping (Fixed Albedo of Inner Pump =.98) Ipump_in Ipump_out Ipump_wall! Wall pumping is insignificant! The inner pump exhaust decreases as outer pumping eficiency (albedo) increases Pumping Current [A] 3 2 1! For equal albedos, the outer pump is more effective in particle removal 1.99.98 Albedo of Outer Pump.97.96 15
Core carbon densit y may be more dependent on wall and plasma conditions in new RDP! 1) Core carbon density (and hence Z eff ) increases with increasing carbon yield, indicating that core carbon can be influenced by wall conditioning.! 2) Core carbon density increases with core heating power, indicating there may be a problem with high power operation - at low density.! 3) Radiation fraction is below 4% (less than that for the lower single null plasma with similar parameters)! 4) Inner and outer pumping is not independent. 16