Alcator C-Mod Particle Transport in the Alcator C-Mod Scrape-off Layer B. LaBombard, R.L. Boivin, B. Carreras, M. Greenwald, J. Hughes, B. Lipschultz, D. Mossessian, C.S. Pitcher, J.L. Terry, S.J. Zweben, Alcator C-Mod Group Presented at the Meeting on Plasma Turbulence and Transport in Edge/SOL Regions 4-5 May, 2, Fairbanks, Alaska
Motivation: Cross-field particle transport (Γ ) in SOL and its dependence on discharge conditions impacts tokamak operation & design Γ determines level of plasma/wall interaction in main-chamber neutral pressures (<=> confinement) impurity sources => impacts divertor design Low Transport High Transport SOL Core Plasma Core Plasma ions neutrals Divertor C-Mod Q: What will be operating regime in a reactor? Heat convection across separatrix and its scaling may play role in tokamak density limit Goals: Characterize and understand Γ : - empirical scalings - underlying turbulence - relationship to tokamak density limit
Outline of Talk Main-Chamber Recycling Diagnostics: Effective Cross- Field Particle Diffusitives (Deff) & Scalings Cross-Field Heat Convection Horizontal Scanning Probe Midplane Dα Fluctuations Behavior Near Discharge Density Limit 6 5 2 4 Vertical Scanning Probe 8 3 6 Divertor Probes
Scrape-off Layer Density Profiles Exhibit a "Two-Exponential" Decay 2 SOL Density Profiles n e /n G =.43 [Ohmic L-Mode] Limiter Shadow (m-3) 9 8 Near SOL n e /n G =.6 5 5 2 25 Separatrix Far SOL Distance into SOL (mm) Near SOL: steep decay, λ n ~ 2 to 8 mm Far SOL: shallow decay, λ n ~ 8 to > mm At low ne, density at limiter edge is less than ~/ of separatrix density Density at limiter edge increases sharply with increasing ne => Always some level of main-chamber (limiter) recycling
Main Chamber Ionization and Fluxes in the Far SOL Are Large Compared to Flows Towards Div./Baffle 23 Main Chamber Ionization (Dα) Ion Flux (D + /s) 22 Flux on Divertor (Div. Probes) Flow Towards Divertor/Baffle (Mach Probe)..2.5 2 ne/n G Recycling in Main Chamber SOL is primarily balance by fluxes onto main-chamber walls Poloidal flows to divertor/baffle are weak Main-Chamber Recycling Regime (MCR) persists over wide parameter range M.V. Umansky, et al. Phys. Plasmas 5, 3373 (998).
A New View of Particle Transport Processes in SOL Old Paradigm: SOL density decays "exponentially" because... plasma drains along field lines towards divertor/baffle SOL Profiles Density Limiter Shadow Ionization FluxTowards Divertor/Baffle Dominates Separatrix SOL drained by parallel flow New Paradigm: SOL density decays "exponentially" because... cross-field transport velocity increases across SOL, maintaining cross-field flux towards wall SOL Profiles Density Limiter Shadow Flux Towards Limiter/Wall Dominates Velocity Ionization Separatrix
Outline of Talk Main-Chamber Recycling Diagnostics: Effective Cross- Field Particle Diffusitives (Deff) & Scalings Cross-Field Heat Convection Fluctuations Edge Thomson Scattering Tangential- Viewing Ly α Array Horizontal Scanning Probe Behavior Near Discharge Density Limit
-5 5 5 2-5 5 5 2-5 5 5 2 (m (m -3 s - ) (m -2 s - ) -3 )2 (m 2 s - ) In MCR Regime, Cross-Field Diffusion Coefficient Profiles (Deff) can be Inferred Directly from Profile Measurements 9 23 22 2 2.. n Sion Γ Deff. -5 5 5 2 Separatrix Distance into SOL (mm) Limiter Shadow - Density n e /n G.43.27.9 - Ionization Source (from Lyα profiles) - Cross-Field Flux Profile - Effective Diffusion Coefficient: Deff = -Γ / n Persistent trend of Deff increasing by ~ or more with distance from separatrix => variation in Deff reflects variation in n Deff increases with discharge density => Γ gets larger, n gets smaller Method benchmarked against UEDGE modeling Deff (χ eff ) increasing seen before: ASDEX, JT-6, JET,...
Magnitude of Deff in Near SOL is Correlated with Collisionality in Near SOL. Regression Analysis of D eff, 2 mm into SOL µte -3.8 n. Ip -.3 BT.8 64 Ohmic L-Mode Datapoints:.8 < ne < 2.5x 2 m -3.6 < Ip <. MA 4 < B T < 6 tesla.4 < ne/n G <.47 (m2 s-)... µ(lei/l) -.7 Multiple Correlation Coefficient =.8 4 Parameter Regression: => Suggests (B T /Ip), q, or L dependence (m2 s-). Multiple Correlation Coefficient =.75.... (m2 s-) Parameter Regression: => Statistics point to (lei/l) as most relevant parameter => Deff correlates with local collisionality: Deff ~ (lei /L) -.7 Trend: ne/n G => lei/l => Deff near sep.
5 5 2 5 5 2 Cross-Field Heat Convection to Limiter/Wall Competes with Parallel Conduction Losses to Divertor at Moderate ne/n G Finite Te on open field lines => power conducted to divertor: Qdiv T 7/2 /L δρ' ρ Cross-field particle fluxes (Γ ) => power convected: Qconv ~ 5 Te Γ A sep 5 Electron Temperature (ev). Power Through Flux Surface ρ (MW)... ne/n G =.9 Qdiv ne/n G =.9 ne/n G =.43 ne/n G =.43 Qconv Qconv Qdiv 5 5 2 ρ (mm) Limiter Distance into SOL Shadow Qconv > Qdiv only in far sol Qconv > Qdiv over entire SOL At low density, heat losses in Near SOL are dominated by parallel conduction to Divertor At moderate density, cross-field heat convection to Limiter/Wall exceeds conduction losses to Divertor/Baffle over entire SOL
Cross-Field Convection Increases with ne/n G, Affecting SOL Power Balance (MW)..8.6.4 Power into SOL [Pin - Prad] Power Conducted to Divertor [Qdiv ] Power Convected to Main-Chamber Limiter.2 Power Convected Across Sep. [Qconv ]..2.3.4.5 n e /n G Outer Divertor Detached At low density, parallel conduction to Divertor dominates SOL power balance At moderate density, cross-field heat convection to Limiter/Wall becomes important => Cross-field convection losses to main-chamber wall may precipitate divertor detachment
A New View of Heat Transport Processes in SOL Old Paradigm: (in absence of a "radiating mantle") Parallel e - conduction to divertor dominates heat losses in Near SOL region Te at separatrix (Tsep) is a weak function of λ T e and SOL power (P sol): Psol Tsep (Psol / λ T e )2/7 Modified Paradigm: Parallel e - conduction and cross-field heat convection contribute to heat losses in Near SOL region (CX is typically a minor player) Psol At low collisionality, parallel conduction regulates Tsep: Tsep (Psol / λ T e )2/7 At high collisionality, heat convection becomes large, Tsep is reduced and is no longer "regulated" by this law!
Outline of Talk Main-Chamber Recycling Diagnostics: Effective Cross- Field Particle Diffusitives (Deff) & Scalings Cross-Field Heat Convection Limiter-Shadow Particle Flux Probe Horizontal Scanning Probe Fluctuations Turbulence Imaging Behavior Near Discharge Density Limit Divertor Probe Triplet
Midplane and Divertor Probe Geometry Midplane Probe 3.8mm Flux Surface Probe Body West North South East Axis of Pyramid Close-up View of Probe Elements Along Axis of Pyramid B-field Direction Four Langmuir Probe Elements:.5 mm dia. tungsten wires, cut to match pyramidal surface Divertor Probe Triplet P P P2 Normal View.5 o 6.6mm B-field Direction 3mm dia. Grazing view Normal view probe triplet B-field Direction P P P2 Grazing View
..5 Time (milliseconds)..5..5 Time (milliseconds)..5..5 Time (milliseconds)..5..5 Time (milliseconds)..5 2 2 2 2 2 Fluctuations Exhibit Different Character in Near and Far SOL Regions ( 2 m -3 ). Near SOL SOL Density Profile Far SOL.2 5 5 2 Distance into SOL (mm) Isat/<Isat> "Snapshots" τ ac ~.2µs 2.2µs 5.5µs 3µs µs..5..5 Time (ms) Limiter Shadow 25 Auto-Correlation Functions Isat/<Isat> (µs) Near SOL (steep n profile): -> moderate amplitude, "random" fluctuations Far SOL (flatter n profile): -> large amplitude, intermittent Isat "bursts" => Consistent with Deff with distance into SOL
..5 Time (milliseconds)..5..5 Time (milliseconds)..5..5 Time (milliseconds)..5..5 Time (milliseconds)..5 - - - - - Fluctuations Exhibit Different Character in Near and Far SOL Regions (ev) 6 4 2 Near SOL SOL Te Profile Far SOL 5 5 2 Distance into SOL (mm) (Vf-<Vf>)/Te "Snapshots" τ ac ~.5µs 3 µs 5 µs 3µs 23µs..5..5 Time (ms) Limiter Shadow 25 Auto-Correlation Functions Vf (µs) Near SOL (steep Te profile): -> moderate amplitude, "random" fluctuations Far SOL (flatter Te profile): -> large amplitude, intermittent Vf "bursts" => Consistent with Deff with distance into SOL
Fluctuation PDFs: ~Gaussian Near Sep., Increasingly Skewed with Distance Into SOL.6.4.2 ( 2 m -3 ) (ev)..2 6 4 2 Near SOL SOL Profiles Density Electron Temperature Far SOL 5 5 2 Distance into SOL (mm) Near SOL (steep Te & n profiles): PDFs have ~Gaussian statistics Limiter Shadow Probability Distribution Functions Floating Potential Skewness: -.9 -.8 -.38 -.77 -.34 Flatness: 2.63 2.55 2.77 3.8 6.95 Ion Current Skewness: -.5.35.483.7 2.4. -4-2 2 4-4 -2 2 4 (Vf-<Vf>)/σ V f (Isat-<Isat>)/σ I sat Far SOL (flatter Te & n profile): Isat PDF has positive tail, Vf negative tail => Consistent with n and Te 'blobs' in Far SOL 25 Flatness: 3.6 3.36 2.95 4.22 2.5
2-D Turbulence Imaging: Intermittent, ~ cm Scale "Blobs" of Emission Extend into Far SOL Near SOL Separatrix Density Far SOL Limiter Shadow Turbulence Imaging: 2-D Images of Da emission, looking along field lines at a "plume" from D2 gas nozzle ~ mm spatial resolution 2 ms exposure times 7 ms between exposures 2 cm ~ cm scale blobs intermittently occupy Far SOL zone, and extend to Limiter Shadow Intermittent "Blobs" (2-D imaging) and "Bursts" (probes) are consistent with large density and temperature (?) perturbations rapidly transporting particles and energy to Limiter/Walls
787.2 787.4 787.6 787.8 msec 788. 788.2 788.4 787.2 787.4 787.6 787.8 msec 788. 788.2 788.4 787.2 787.4 787.6 787.8 msec 788. 788.2 788.4 Isat < Isat > Vf Vf2 ~ Γ < n > Time-History of Fluctuation-Driven Flux Shows Large-Amplitude, Bursty Behavior 3. 2... -2-4 -6-8 -2-4 -6-8 4 2-2 -4 [V] [V] ~ ~ Data from Midplane Probe, 7 mm outside LCFS ~ Particle flux estimate neglects T e ne Velocity [m s - ].2.4.6.8..2 Time [ms] Time-averaged transport velocity: < Γ > < n > ~ 2 m s- Bursts in transport velocity exceed 2 m s - ~ What is influence of T e on these estimates?
~~ <ne> -Derived Cross-Field Flux Shows Similar Trend with as Particle Balance-Derived Flux ne Mean Flux ( 2 m -2 s - ) Estimates of Cross-Field Particle Fluxes (Γ) 7 mm Outside the Separatrix 2 5 5 ~ ~ <ne> (midplane), divided by 8 Particle Balance.5..5 2. 2.5 3. Line-averaged density( 2 m -3 ) ~ ~ ~ <ne> - inferred Γ does not account for possible T e Γ from both methods show similar trend, nonlinearly increasing with ne ~ ~ Magnitude of Γ inferred from Midplane Probe <ne> is a factor of ~8 times larger than that derived from particle balance
PDF of Fluctuation-Driven Particle Flux Exhibits Power-Law Tail, Independent of ne Probability * <Flux> Probability * <Flux> Γ Inferred from Midplane Probe, ~7 mm outside LCFS - -2-3 -4-5 -2-3 -4-5 ne:.8.6.45.3.87-2 3 4 5 6 Flux/<Flux> ne:.8.6.45.3.87 Flux/<Flux> Similar P(Γ ) for all densities => change in <Γ > not associated with change in P(Γ ) Always has SOC-Like behavior: Positive Γ events with Γ greater than 4.6 <Γ > account for 5% of total particle transport. These events happen ~5% of time. Analysis by B. A. Carreras, V. E. Lynch, submitted to PoP.
Fluctuations at Fixed Location in Far SOL Depend on ne/n G Floating Potential Auto-Power Spectra Midplane Probe - -2-3 - -2-3 n e /n G =.4 n e /n G =.22 n e /n G =.3 - f Location maps to 5-8 mm outside LCFS Divertor Probe -2 f 578:Nebar=.87 573:Nebar=.3 579:Nebar=.8-4 -5 2 3 Location maps to 9 mm outside LCFS 4 Hz Breakpoint frequency in Vf power spectra evolve similarly at midplane or divertor locations 5 Different high frequency roll-off 6
Long Time-Range Correlations of Plasma Fluctuations Increase with ne 4 Divertor Probe Data R/S 3 2 - H = H =.5 mesoscale range High density Low density 2 3 4 5 Time lag (µs) Rescaled adjusted range statistic (R/S) indicates increase in long range correlation with ne H.75.7.65.6 999 2 Similar increase in H seen on Midplane and Divertor probes.55 Midplane Divertor.5.8.2.4.6.8 2 Line-averaged density( 2 m -3 ) Analysis by B. A. Carreras, V. E. Lynch, submitted to PoP.
Outline of Talk Main-Chamber Recycling Diagnostics: Effective Cross- Field Particle Diffusitives (Deff) & Scalings Cross-Field Heat Convection Limiter-Shadow Particle Flux Probe Horizontal Scanning Probe Fluctuations Behavior Near Discharge Density Limit
..2.4 seconds.6.8...2.4 seconds.6.8. (MA) Collisionality at the Separatrix and Heat Convection to Limiter/Wall Increases as Discharge Density Limit is Approached (MW) Diverted Discharge with Ramping ne/n G. 5 Plasma Current.5 2..5 Input Power ne/n G Line Averaged Density Radiated Power λei/l....2.4.6.8. (seconds) Horizontal Scanning Probe records profiles at three times 4 3 2 ( 2 m -3 ).8.4 Heat Convection to Limiter/Wall based on Limiter Particle Flux Probe λei/l, mm into SOL As density limit is approached: λei/l near separatrix drops dramatically Radiation and Convection to Limiter/Wall are comparable and mostly account for input power Near density limit: Radiation + Convection to Walls ~ Input Power
-2 2 4 6 8 2 2 4 6 8 2-2 2 4 6 8 2 Near Density Limit: Large Amplitude, Long- Correlation Time Fluctuations Envelop Entire SOL and Cross Separatrix n / n G at separatrix does not increase!.4. SOL Profiles Density / n G ne/n G.82.4.34 (ev) 6 4 2-2 Electron Temperature (µs) -2 Auto-Correlation Times (Vf data).3. -2 2 4 6 8 2 Separatrix Distance into SOL (mm) RMS Isat/<Isat> Near density limit: SOL n & Te profiles become flat, Tsep low ~ 25 ev! Fluctuations characteristic of "Far SOL" now occur everywhere, even across the separatrix => Consistent with large Convection Losses
Summary C-Mod SOL Core Plasma Divertor SOL density profiles exhibit a "twoexponential" decay: Near and Far SOL Yet, Main-chamber plasma exhausts primarily onto Limiter/Wall surfaces! - Why? - => New Particle Transport Paradigm: Density "decays exponentially" because... cross-field transport (Deff ) increases rapidly with distance into SOL... not because parallel flows "drain" SOL plasma Particle (Deff) and heat convection near separatrix increases with collisionality: Deff ~ (λei /L) -.7 => Heat Transport Paradigm Modified: At moderate collisionality (ne/n G ~.5), crossfield heat convection exceeds conduction losses Tsep no longer regulated by "conduction law": Tsep (Psol / λ T e )2/7
Summary (page 2) Fluctuation behavior supports picture of particle & energy transport increasing with distance into SOL Near SOL: (steep density profile) low amplitude "random" fluctuations Far SOL: (flat density profile) large amplitude intermittent "bursts" in Isat and ~ cm "blobs" in Dα, extending into Limiter Shadow ~ ~ <ne> -derived cross-field flux (Γ) supports results inferred from particle balance: Γ similarly increases with ne Γ is factor of ~8 times larger than particle balance => supporting MCR conclusion Far SOL turbulence has some SOC-like characteristics (~generic to edge plasmas) PDF(Γ) has power-law tail (independent of ne ) Long-range time correlations are increasingly seen as ne is increased
Summary (page 3) => New Insight on Density Limit Physics: As density limit is approached, λei/l near separatrix drops and transport across the SOL increases dramatically Heat Convection to Limiter/Wall becomes large fraction of input power "Bursty" fluctuations (large transport) occur over entire SOL and begins to attack plasma on closed flux surfaces ~ at limit: Radiation + Convection to Wall ~ Input Power Rapid increase of Heat Convection as edge plasma cools may play role in thermal instability leading to disruption