Cross-Field Plasma Transport and Main Chamber Recycling in Diverted Plasmas on Alcator C-Mod B. LaBombard, M. Umansky, R.L. Boivin, J.A. Goetz, J. Hughes, B. Lipschultz, D. Mossessian, C.S. Pitcher, J.L.Terry, Alcator C-Mod Group Laboratory for Laser Energetics, Rochester, NY, USA
Motivation: Q: What will be the operating regime for SOL & Divertor particle handling in a reactor? Ideal SOL/Divertor C-Mod SOL/Divertor SOL Core Plasma Core Plasma ions neutrals Divertor Part I C-Mod SOL/Divertor is far from ideal: "Main-Chamber Recycling" dominates SOL flows Why? => cross-field particle transport, Γ Part II Need to characterize and understand Γ : => Γ can be directly measured in MCR regime
I. Main Chamber Recycling Regime Diagnostics: U Upper Chamber Neutral Pressure Outline: 6 5 1 2 4 Scanning Langmuir Mach Probe 8 6 3 1 10 Divertor Probes Divertor Bypass Flaps Midplane Dα M Midplane Neutral Pressure First-Principles Particle Balance Estimates Conditions for Main-Chamber Recycling -> Simplified SOL Model Implications for SOL Particle/Heat Transport -> 2-D UEDGE Simulations
High Neutral Pressures Surround Core Plasma, Independent of Divertor Bypass Neutral Pressure (mtorr) 100.0 10.0 1.0 0.1 0.01 Ohmic L-Mode Density Scan, I p =0.8 MA, B T = 5.3 T Divertor Upper Chamber Midplane Divertor Bypass Open 1.0 1.4 1.8 2.2 Line-Averaged Electron Density (10 20 m -3 ) 10.0 1.0 0.1 0.01 (pascals) Neutral pressures in Upper Chamber are higher than at Midplane! => implies large neutral fluxes (Γ w ) attack core plasma directly from "wall" surfaces: Γ w (s -1 ) 8x10 22 Pw (mtorr)
Main Chamber Ionization and Neutral Fluxes Are Much Larger than Plasma Flux Towards Divertor 10 23 Ohmic L-Mode Density Scan, I p =0.8 MA, B T = 5.3 T Main Chamber Ionization (Dα) Ion Flux [D + /s] 10 22 Flux on Div. (probes) 10 21 Flow Towards Div. (Mach Probe) 10 21 10 22 10 23 Estimate of Neutral Flux from Wall [D 0 /s] (midplane pressure) Recycling in Main Chamber is comparable to or exceeds Divertor! Yet flow between volumes is small Regime persists over all divertor regimes: sheath-limited, conduction-limited or detached
Conditions for Main-Chamber Recycling Simplified SOL, Part 1 A critical flux-surface averaged neutral density symmetry plane Core Plasma SOL Γ Γ // Main Chamber Wall B = n e Divertor Surface Plasma Continuity Γ = n 0 n e k ion Flux Surface Average x Γ n e [ n 0 k ion C s] x 2 L => Γ increases with x when n 0 n 0 crit n 0 crit ~ Cs 2 k ion L
Simplified SOL, Part 2 A critical flux-surface averaged plasma flux Core Plasma symmetry plane Γ 0 SOL Γ 0// Main Chamber Wall Divertor Surface = n 0 Mass balance requires: Γ 0 = Γ Neutral Flux requires a minimum Neutral Density: n 0 ~ Γ 0 / v th0 (v th0 ~ set by CX) => Γ crit above which n 0 must exceed n 0 crit If Γ crit is exceeded at x=x0 then Γ increases with x for x > x0 => Main-Chamber Recycling => Independent of separatrix-wall distance! => Main-Chamber Recycling depends more on particle transport than divertor/wall geometry x
Main Chamber Ionization (Dα) in C-Mod Exceeds Upper Bound Estimate of Γ crit in All Discharges Estimate with T0 Ti Te, free-streaming neutrals Γ crit 2x10 20 T(eV) (m -2 s -1 ) L (m) Typical Number: Γ crit A sep ~ 5x10 21 (s -1 ) 10 23 Ohmic L-Mode Density Scan, I p =0.8 MA, B T = 5.3 T Main Chamber Ionization (Dα) Ion Flux [D + /s] 10 22 10 21 10 21 10 22 10 23 Estimate of Neutral Flux from Wall [D 0 /s] (midplane pressure)
0 5 10 15 0 5 10 15 m 2 s -1 m -2 s -1 10 20 m -3 Effective Particle Diffusion Coefficient in SOL Must Increase with Distance from Separatrix 1.5 1.0 0.5 22 10 10 21 10 20 10 1 0.1 0.01 Results from UEDGE Simulations Density, n Measurements UEDGE Profiles: High Pmid Low Pmid Plasma Flux, Diffusion Coefficient, 0 5 10 15 Distance from separatrix at midplane, ρ (mm) Comments: Γ D eff "Exponentially" decreasing density profile is always measured Yet, plasma flux (Γ ) increases with distance into SOL! if Γ = -D eff n, then D eff must increase strongly with distance into SOL! "Exponentially" decreasing density profile is NOT evidence that flow to divertor dominates D eff increasing with ρ seen in other experiments: ASDEX, ASDEX-U 3 He transport (see Nachtrieb, P-3.37) χ eff increasing with ρ has been seen: JT-60, C-Mod, JET
Electron Convection and Charge Exchange Dominate SOL Heat Transport at High Density Cross-Field Power (MW) 0.8 0.6 0.4 0.2 0.0 0.4 0.2 UEDGE Results High Pmid Total Low P mid Total Electron Convection q = 5/2 Te Γ Charge Exchange q = - n χ cx T i Anomalous Plasma Conduction q = - n χ a (T i +Te) 0.0-5 0 5 ρ (mm) 10 15 Electron Convection dominates heat transport at high density - even through separatrix! Charge Exchange ~ always a player in far SOL Plasma Conduction ~ only important near separatrix at low densities
II. Cross-Field Particle Transport Experiments => D eff Profiles & Scalings Diagnostics: Outline: Edge Thomson Scattering Tangential- Viewing Ly α Array Horizontal Scanning Probe Γ from Integral of Ionization Profile => D eff Profiles D eff Scales with Collisionality Implication: => Convection & CX set SOL power balance at high density
-5 0 5 10 15-5 0 5 10 15-5 0 5 10 15 In MCR Regime, Cross-Field Diffusion Coefficient Profiles (D eff ) can be Inferred Directly from Local Profile Measurements (m 2 s -1 ) (10 20 m -2 s -1 ) (10 23 m -3 s -1 ) (10 23 m -3 s -1 ) 0.8 0.4 0.0 4 0-4 10 1 1.0 0.1.01 Sion Γ Typical Ohmic L-Mode Shot#990429012:550-5 0 5 10 15 Divertor/Wall Flux Ratio: -. Γ Deff = -Γ / n ρ (mm) = 1.0 = 0.5 ~ measurement = 0.0 - Ionization Source Profile - Sensitivity Study: Possible. Γ Profiles - Cross-Field Flux Profile: Γ = Source δρ - Effective Diffusion Coefficient: Deff = -Γ / n => Clear trend of Deff increasing by 10 or more with distance from separatrix => Virtually identical to UEDGE results
Analysis of 75 Profiles Yields Similar Trend - 65 Ohmic L-mode profiles (density scan) - 10 ICRF-heated H-mode profiles 1.00 Deff ( -Γ / n) (m 2 s -1 ) 0.10 0.01 L-mode EDA H-mode Elm-Free H-mode 0 5 10 15 ρ (mm) H-mode discharges have lowest D eff near separatrix => BUT: Scatter in Deff at fixed ρ is outside measurement error! -> "Scatter" should reflect dependence of Deff on plasma conditions...
Deff (m 2 s -1 ) Regression Analysis: Correlation of D eff at ρ=2 mm with Local Te and n 0.10 55 Ohmic L-Mode Datapoints 1.0 < ne < 2.2x10 20 m -3 0.7 < Ip < 0.8 MA B T = 5.3T Multiple Correlation Coefficient = 0.85 0.01 0.01 Te -3.5 n 1.7 0.10 Local Particle Balance UEDGE Correlation accounts for much of the "scatter" in D eff measurements UEDGE-derived points scale similarly => Suggests that Deff scales with Collisionality: Deff ~ (λei /L) -1.7
MID_Lambda_SD () 0.1 Convected Heat Flux Through Separatrix Increases with Plasma Density (m 2 s -1 ) Data from ohmic L-mode density scan 0.10 0.01 D eff ~ (λ ei /L) -2 D eff at ρ= 1 mm Fraction 0.1 Power fraction convected into SOL: 5 Te Γ A sep Psol 0.01 0.03 0.10 0.30 High λ ei /L core density n/n G ~0.35 0.1 Low As collisionality increases, heat convection and charge exchange across separatrix becomes more important in SOL power balance positive feedback => Suggests a direct link between scaling of anomalous particle transport and existence of a discharge density limit => further studies needed!
C-Mod SOL/Divertor SOL Summary Core Plasma Divertor C-Mod SOL/Divertor is far from ideal "Main-Chamber Recycling" dominates SOL flows Why? MCR regime is determined more by cross-field plasma transport than divertor/wall geometry => critical Γ, exceeded in C-Mod Effective cross-field plasma diffusivity (D eff ) rapidly grows with distance from separatrix Electron convection and charge exchange dominate heat transport at high densities D eff near separatrix correlates with local λ ei /L in ohmic L-mode discharges: Deff ~ (λei /L) -1.7 This scaling suggests the existence of a discharge density limit (~ collisionality limit) set by heat convection and charge exchange losses