Effect of divertor nitrogen seeding on the power exhaust channel width in Alcator C-Mod

Transcription

1 Effect of divertor nitrogen seeding on the power exhaust channel width in Alcator C-Mod B. LaBombard, D. Brunner, A.Q. Kuang, W. McCarthy, J.L. Terry and the Alcator Team Presented at the International Conference on Plasma Surface Interactions in Controlled Fusion Devices Princeton University, NJ, USA, June 218 This work was supported by US DoE cooperative agreements DE-SC14264 and DE-FC2-99ER54512 on Alcator C-Mod, a DoE Office of Science user facility.

2 Question: Does upstream power exhaust width depend on conditions in the divertor? radiation, detachment, electrical disconnection? Scrape-off layer power channel widths (λ q ) are projected empirically to be ~.5 mm in power reactors (B p ~ 1.2 T), based on data from attached divertor plasmas with low levels of volumetric dissipation. Theory indicates that electrical connection to the target can play a role in SOL turbulence: Ø Ø parallel currents to target can reduce blob polarization/transport [e.g. O.E. Garcia PoP 26] sheath potentials can impose ExB shear in near SOL regulating turbulence [e.g. F. Halpern NF 217] new database probe-based sensors old database IR thermograpic analysis At high levels of volumetric dissipation and/or partial detached target conditions electrical connection can be weak or broken Does upstream λ q change under these conditions? T. Eich 213, D. Brunner 218 Experiments were performed on C-Mod to examine this question... 2

3 Experiment: Systematically vary divertor dissipation (via N2 seeding) in otherwise identical plasmas; study SOL response. Experiments enabled by real-time heat flux measurements (via surface thermocouples) to feedback control divertor N2 seeding [1] Surface Thermocouples Experiment Produce a series of ohmic L-mode plasmas with B-field constant core plasma conditions Ramped Tiles Systematically lower set point of divertor heat flux, causing divertor to change from sheath-limited to high-recycling to partially detached. Record divertor response with Surface TC array and with a high resolution Rail Langmuir probe array [2] Scanning Mirror Langmuir Probe 21 Rail Langmuir Probes Record upstream profiles with Scanning Mirror Langmuir Probe [3] => examine SOL response and heat flux widths Divertor N2 seeding [1] Brunner, RSI 87 (216); [2] Kuang,RSI (218); [3] LaBombard, PoP 21 (214) Effect of divertor nitrogen seeding on power exhaust channel width in Alcator C-Mod B. LaBombard, PSI 218 3

4 Real-time surface heat flux measurements are used to feedback control divertor N 2 seeding allows divertor heat fluxes to be precisely specified MA MW mm MW m Plasma Current Psol Prad Ptot Ave. Surface Heat Flux Scanning MLP Insertion Density Gas Valve Set Point seconds Divertor surface heat fluxes reduced by factor of ~1, approaching (but not triggering) complete divertor detachment Scanning MLP samples upstream profiles during this evolution m % Scanning Mirror Langmuir Probe Divertor N 2 seeding 4

5 High resolution SOL profiles are deduced from time-averaging MLP data Scanning Mirror Langmuir Probe: Electrode Geometry Langmuir-Mach Probe High heat-flux geometry Raw data consist of, measurements of each parameter from each electrode (NE, SE, SW, NW) Fluctuations in data are not noise! MLP bias system tracks turbulence with high fidelity. 5

6 High resolution SOL profiles are deduced from time-averaging MLP data Scanning Mirror Langmuir Probe: Electrode Geometry Langmuir-Mach Probe High heat-flux geometry Raw data consist of, measurements of each parameter from each electrode (NE, SE, SW, NW) Step 1: average data from four electrodes Step 2: time average over 2 µs 6

7 High resolution SOL profiles are deduced from time-averaging MLP data Scanning Mirror Langmuir Probe: Electrode Geometry Langmuir-Mach Probe Data from spline-fitted curves are shown in subsequent slides High heat-flux geometry Raw data consist of, measurements of each parameter from each electrode (NE, SE, SW, NW) Step 1: average data from four electrodes Step 2: time average over 2 µs Step 3: fit smooth spline curves to data Step 4: shift profiles in rho to satisfy SOL power balance 7

8 What might we expect? As N 2 divertor seeding is increased... reduction in divertor target electron temperature increase in divertor target density increase in divertor plasma collisionality Myra and D Ippolito plasma collisionality parameter, Λ, increases, possibly enhancing cross-field blob transport increase in SOL width decrease in sheath potential decrease in SOL plasma potential decrease in ExB shear Reduction of radially sheared ExB flows allows enhanced cross-field transport increase in SOL width Blob Propagation Model [1] Myra, PoP 13 (26) 1122 Plasma Potential and ExB Shear Layer Model Φ/T e [1] Halpern, NF 57 (217) 341 8

9 What did we find? Look at four L-mode cases at three plasma currents... Ohmic L-mode Toroidal field: 5.4 tesla Greenwald fraction: ~ Sheath-limited divertor conditions prior to N 2 injection Case 1 (216) Case 2 (215) Case 3 (216) Case 4 (216) Plasma Current: 1.1 MA n e ~ 1.7x1 2 P SOL ~ 1.1 MW Max measured q // ~ 3 MW m -2 Plasma Current:.8 MA n e ~ 1.3x1 2 P SOL ~.7 MW Max measured q // ~ 12 MW m -2 Plasma Current:.55 MA n e ~.9x1 2 P SOL ~.4 MW Max measured q // ~ 7 MW m -2 9

10 What did we find? Look at four L-mode cases at three plasma currents... Ohmic L-mode Toroidal field: 5.4 tesla Greenwald fraction: ~ Sheath-limited divertor conditions prior to N 2 injection Case 1 (216) Case 2 (215) Plasma Current: 1.1 MA n e ~ 1.7x1 2 P SOL ~ 1.1 MW Max measured q // ~ 3 MW m -2 Examine 1.1 MA data first: 21 shot/time slices Case 3 (216) Case 4 (216) Plasma Current:.8 MA n e ~ 1.3x1 2 P SOL ~.7 MW Max measured q // ~ 12 MW m -2 Plasma Current:.55 MA n e ~.9x1 2 P SOL ~.4 MW Max measured q // ~ 7 MW m -2 1

11 1.1 MA datasets: Divertor N 2 seeding reduced divertor surface heat fluxes by a factor of ~1 with core plasma relatively unperturbed Line-Averaged Density 1 2 m -3 Psol (MW) MA : STC ave q > : SSEP < 1 mm 1.1 MA : 25 < STC ave q < : SSEP < 1 mm 1.1 MA : STC ave q < 25 : SSEP < 1 mm 1.1 MA : STC ave q > 1.1 MA : 25 < STC ave q < 1.1 MA : STC ave q < 25 Power into Scrape-off Layer Line average density held ~constant Power into SOL ~constant Rail probe data ρ = 2 mm 5 Rail probe data Parallel Heat Flux Density on Divertor Surface (MW m -2, averaged over multiple Surface TCs) Divertor conditions near strike point change from sheath-limited to highrecycling to near detached (~ 5 ev) 11

12 1.1 MA dataset: Divertor Upstream Scrape-off Layer Midplane Electron Temperature Jgnd (A mm -2 ) Net Current Density to Divertor Surface 1 1. Midplane Density - normalized to core line-averaged q (MW m -2 ) Parallel Heat Flux at Divertor Surface Surface Thermocouple Data Density/NeBar.1 12

13 1.1 MA dataset: Jgnd (A mm -2 ) Divertor Divertor conditions change dramatically with N 2 seeding Net Current Density to Divertor Surface Divertor Response Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Upstream Scrape-off Layer 1 1. Midplane Electron Temperature Midplane Density - normalized to core line-averaged q (MW m -2 ) Parallel Heat Flux at Divertor Surface Surface Thermocouple Data Density/NeBar.1 13

14 1.1 MA dataset: Lambda Myra Divertor Divertor conditions change dramatically with N 2 seeding Divertor Collisionality Parameter, Λ div - Myra Divertor Response Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Divertor collisionality (Λ Myra[1]) increases factor of ~ near strike Upstream Scrape-off Layer 1 1. Midplane Electron Temperature Midplane Density - normalized to core line-averaged q (MW m -2 ) Parallel Heat Flux at Divertor Surface Surface Thermocouple Data [1] Myra, PoP 13 (26) Density/NeBar.1 14

15 1.1 MA dataset: Jgnd (A mm -2 ) Divertor Upstream T e, n e profiles steepen slightly near LCFS and reduce in far SOL in response to divertor N 2 seeding Net Current Density to Divertor Surface Divertor Response Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Divertor collisionality (Λ Myra[1]) increases factor of ~ near strike Upstream SOL Response Upstream Scrape-off Layer 1 1. Note log scale Midplane Electron Temperature Midplane Density - normalized to core line-averaged q (MW m -2 ) Parallel Heat Flux at Divertor Surface Surface Thermocouple Data T e and n e reduced in far SOL(!) Near SOL width becomes slightly narrower (!) with increased divertor dissipation (N 2 ) [1] Myra, PoP 13 (26) Density/NeBar.1 Note log scale 15

16 1.1 MA dataset: Jgnd (A mm -2 ) Divertor Upstream T e, n e profiles steepen slightly near LCFS and reduce in far SOL in response to divertor N 2 seeding Net Current Density to Divertor Surface Divertor Response Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Divertor collisionality (Λ Myra[1]) increases factor of ~ near strike Upstream SOL Response LTe (mm) Upstream Scrape-off Layer 1 1 Midplane Electron Temperature Gradient Scale Length Midplane Density Gradient Scale Length q (MW m -2 ) Parallel Heat Flux at Divertor Surface Surface Thermocouple Data T e and n e reduced in far SOL(!) Near SOL width becomes slightly narrower (!) with increased divertor dissipation (N 2 ) [1] Myra, PoP 13 (26) LNe (mm)

17 Question: Drop of T e & n e in far SOL and narrowing of Near SOL with N 2 Reproducible? => yes, also seen in 215 investigation Upstream Temperature and Density Profiles Midplane Electron Temperature Midplane Electron Temperature Midplane Density - normalized to core line-averaged Density/NeBar 1. Midplane Density - normalized to core line-averaged Density/NeBar

18 Question: Drop of T e & n e in far SOL and narrowing of Near SOL with N 2 -- What caused it? RMS Te/<Te> Upstream Temperature and Density Fluctuation Profiles Te Fluctuation Amplitude (RMS Te/<Te>) 1.1 MA, 216 Density Fluctuation Amplitude (RMS n/<n>) RMS Te/<Te> Te Fluctuation Amplitude (RMS Te/<Te>) 1.1 MA, 215 Density Fluctuation Amplitude (RMS n/<n>) Something specific to the Unseeded cases systematically had factor of ~2 higher fluctuation levels at LCFS! RMS n/<n> RMS n/<n> But, is this effect caused by the change in divertor conditions?... Correlated with: Reduction in plasma fluctuations (and transport) 18

19 Next, look at.8 MA dataset... Ohmic L-mode Toroidal field: 5.4 tesla Greenwald fraction: ~ Sheath-limited divertor conditions prior to N 2 injection Case 1 (216) Case 2 (215) Case 3 (216) Case 4 (216) Plasma Current: 1.1 MA n e ~ 1.7x1 2 P SOL ~ 1.1 MW Max measured q // ~ 3 MW m -2 Plasma Current:.8 MA n e ~ 1.3x1 2 P SOL ~.7 MW Max measured q // ~ 12 MW m -2 Plasma Current:.55 MA n e ~.9x1 2 P SOL ~.4 MW Max measured q // ~ 7 MW m -2.8 MA dataset: 16 shot/time slices 19

20 .8 MA dataset: Divertor N 2 seeding reduced divertor surface heat fluxes by a factor of ~1 with core plasma relatively unperturbed Line-Averaged Density Same as 1.1 MA case 1 2 m -3 Psol (MW) MA : STC ave q > 6.8 MA : 25 < STC ave q < 6.8 MA : STC ave q < 25 Power into Scrape-off Layer Line average density held ~constant Power into SOL ~constant Rail probe data ρ = 2 mm Divertor conditions near strike point change from sheath-limited to highrecycling to near detached (~ 5 ev) Parallel Heat Flux Density on Divertor Surface (MW m -2, averaged over multiple Surface TCs) 2

21 .8 MA dataset: Divertor conditions change dramatically with N 2 seeding Jgnd (A mm -2 ) q (MW m -2 ) Divertor Net Current Density to Divertor Surface Parallel Heat Flux at Divertor Surface Divertor Response Same as 1.1 MA Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment - Surface Thermocouple Data

22 .8 MA dataset: Jgnd (A mm -2 ) q (MW m -2 ) Divertor Upstream T e, n e profiles ~flatten slightly near LCFS; No change in far scrape-off layer with divertor N 2 seeding Net Current Density to Divertor Surface Parallel Heat Flux at Divertor Surface Divertor Response Same as 1.1 MA Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Upstream SOL Response Different from 1.1 MA T e and n e unchanged in far SOL A hint that near SOL width becomes less narrow with increased divertor dissipation (N 2 ) Density/NeBar Upstream Scrape-off Layer Midplane Electron Temperature Midplane Density - normalized to core line-averaged - Surface Thermocouple Data

23 .8 MA dataset: Jgnd (A mm -2 ) q (MW m -2 ) Divertor Upstream T e, n e profiles ~flatten slightly near LCFS; No change in far scrape-off layer with divertor N 2 seeding Net Current Density to Divertor Surface Parallel Heat Flux at Divertor Surface Divertor Response Same as 1.1 MA Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Upstream SOL Response Different from 1.1 MA T e and n e unchanged in far SOL A hint that near SOL width becomes less narrow with increased divertor dissipation (N 2 ) LTe (mm) LNe (mm) Upstream Scrape-off Layer Midplane Electron Temperature Gradient Scale Length Midplane Density Gradient Scale Length - Surface Thermocouple Data

24 .8 MA dataset: Jgnd (A mm -2 ) q (MW m -2 ) Divertor Net Current Density to Divertor Surface Parallel Heat Flux at Divertor Surface Surface Thermocouple Data Upstream T e, n e profiles ~flatten slightly near LCFS; No change in far scrape-off layer with divertor N 2 seeding Divertor Response Same as 1.1 MA Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Upstream SOL Response Different from 1.1 MA T e and n e unchanged in far SOL A hint that near SOL width becomes less narrow with increased divertor dissipation (N 2 ) NO SYSTEMATIC CHANGE in SOL fluctuations with N seeding 2 Upstream Scrape-off Layer RMS Te/<Te> RMS n/<n> Te Fluctuation Amplitude (RMS Te/<Te>) Density Fluctuation Amplitude (RMS n/<n>)

25 Next, look at.55 MA dataset... Ohmic L-mode Toroidal field: 5.4 tesla Greenwald fraction: ~ Sheath-limited divertor conditions prior to N 2 injection Case 1 (216) Case 2 (215) Case 3 (216) Case 4 (216) Plasma Current: 1.1 MA n e ~ 1.7x1 2 P SOL ~ 1.1 MW Max measured q // ~ 3 MW m -2 Plasma Current:.8 MA n e ~ 1.3x1 2 P SOL ~.7 MW Max measured q // ~ 12 MW m -2 Plasma Current:.55 MA n e ~.9x1 2 P SOL ~.4 MW Max measured q // ~ 7 MW m MA dataset: 32 shot/time slices 25

26 .55 MA dataset: Divertor N 2 seeding reduced divertor surface heat fluxes by a factor of ~1 with core plasma relatively unperturbed Line-Averaged Density Same as 1.1 and.8 MA cases 1 2 m MA : STC ave q > 3.55 MA : 1 < STC ave q < 3.55 MA : STC ave q < 1 Line average density held ~constant Psol (MW) Power into Scrape-off Layer Rail probe data ρ = 2 mm Power into SOL ~constant Divertor conditions near strike point change from sheath-limited to highrecycling to near detached (~ 5 ev) Parallel Heat Flux Density on Divertor Surface (MW m -2, averaged over multiple Surface TCs) 26

27 .55 MA dataset: Divertor conditions change dramatically with N 2 seeding q (MW m -2 ) Jgnd (A mm -2 ) Divertor Net Current Density to Divertor Surface Parallel Heat Flux at Divertor Surface Divertor Response Same as 1.1 &.8 MA Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Surface Thermocouple Data

28 .55 MA dataset: Jgnd (A mm -2 ) q (MW m -2 ) Divertor Net Current Density to Divertor Surface Parallel Heat Flux at Divertor Surface Upstream T e, n e profiles are not affected by divertor N 2 seeding (within experimental uncertainties) Surface Thermocouple Data Divertor Response Same as 1.1 &.8 MA Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Upstream SOL Response Similar to.8 MA case T e and n e unchanged in far SOL => Profiles are unchanged, within statistical uncertainties NO CHANGE in SOL fluctuations with N seeding Upstream Scrape-off Layer 2 Density/NeBar Midplane Electron Temperature Midplane Density - normalized to core line-averaged

29 .55 MA dataset: Jgnd (A mm -2 ) q (MW m -2 ) Divertor Net Current Density to Divertor Surface Parallel Heat Flux at Divertor Surface Upstream T e, n e profiles are not affected by divertor N 2 seeding (within experimental uncertainties) Surface Thermocouple Data Divertor Response Same as 1.1 &.8 MA Parallel heat fluxes on divertor surface reduced by factor of ~1 Net current densities on divertor surface reduced by factor of ~1 Divertor Te approaches ~5 ev; attains partial detachment Upstream SOL Response Similar to.8 MA case T e and n e unchanged in far SOL => Profiles are unchanged, within statistical uncertainties NO CHANGE in SOL fluctuations with N seeding Upstream Scrape-off Layer 2 LTe (mm) LNe (mm) Midplane Electron Temperature Gradient Scale Length Midplane Density Gradient Scale Length

30 .55 MA dataset: But - Upstream Plasma Potential Profile is strongly affected by divertor N 2 seeding Phi (V) Lambda Myra Upstream Plasma Potential > 3 Divertor Collisionality Parameter, Λ div - Myra Plasma potential at LCFS drops by ~ V, roughly consistent drop in divertor T e and corresponding sheath potential Divertor collisionality in near SOL increases by two orders of magnitude with N 2 seeding

31 .55 MA dataset: But - Upstream Plasma Potential Profile is strongly affected by divertor N 2 seeding Phi (V) Lambda Myra Upstream Plasma Potential > 3 Divertor Collisionality Parameter, Λ div - Myra Plasma potential at LCFS drops by ~ V, roughly consistent drop in divertor T e and corresponding sheath potential Divertor collisionality in near SOL increases by two orders of magnitude with N 2 seeding Message: Near SOL density and temperature profiles (~heat flux widths) are robustly insensitive to: Time-averaged potential, ExB flows and their shear Divertor conditions (e.g. collisionality)

32 Summary Upstream T e, n e profiles near the LCFS (~heat flux widths) are robustly insensitive to divertor plasma conditions Factor of ~1 reduction in divertor target plate heat fluxes (with core plasma unchanged) Divertor target conditions varying from sheath-limited to high-recycling, approaching detachment; divertor collisionality (Λ div Myra) changing by factor of ~. DC current densities to the target plate reduced by over an order of magnitude Upstream T e, n e profiles near the LCFS (~heat flux widths) are insensitive to plasma potential profile (and ExB shear details) Upstream T e, n e profiles are sensitive to plasma fluctuations (e.g. shoulder formation in unseeded 1.1 MA cases) Mechanism that generates fluctuations seen in 1.1 MA unseeded cases is unknown message: beware of confounding influences when doing controlled experiments 32

33 Why do these results matter? Practical Indicates that divertor dissipation does not reduce peak heat flux densities entering into the divertor volume via an increase in λ q Theoretical Indicates that divertor plasma conditions, including divertor sheath boundary conditions, and plasma potential profiles (~equilibrium ExB shear) do not play a defining role in the physics of cross-field transport in the near SOL region contrary to some notions Implication for Advanced Divertors Indicates that: (1) upstream λ q will be largely unaffected by divertor details length of divertor leg, flux expansion, embedded x-points,... (2) divertor should be designed to accommodate empirical upstream λ q 33