Poloidal Transport Asymmetries, Edge Plasma Flows and Toroidal Rotation in B. LaBombard, J.E. Rice, A.E. Hubbard, J.W. Hughes, M. Greenwald, J. Irby, Y. Lin, B. Lipschultz, E.S. Marmar, K. Marr, C.S. Pitcher, N. Smick, J.L. Terry, S.M. Wolfe, S.J. Wukitch, Team Presented at the 6th International Conference on Plasma-Surface Interactions in Controlled Fusion Devices May 24-28, 24 Portland Maine, USA
Motivation: Strong parallel flows (M // ~.5) have been seen in the SOLs of many tokamaks, far from material surfaces......yet, the underlying physics has not been fully resolved Questions: Do these flows impact SOL impurity transport & screening from core?...balance of material erosion/deposition in divertor legs (e.g. JET)? Are SOL flows just a passive, local response? Or, do they couple to flows in the confined plasma in any significant way? Focus of this talk: SOL plasma flow experiments in - plasma flow pattern - underlying drive mechanisms
Key Results: A remarkable interplay between ballooning-like transport, parallel plasma flows and toroidal rotation A cross field transport-driven plasma circulation loop is evident in ballooning-like transport x-point topology sets // flow direction promotes main-chamber impurity migration toward inner divertor near-sonic // flows SOL flows set flow boundary conditions for confined plasma x-point dependent toroidal rotation of core! => x-point dependent toroidal rotation of SOL Surprising result: Topology-dependent SOL flow boundary condition may explain sensitivity of L-H power threshold on upper/lower x-point topology!
Outline of Talk Diagnostics: Profiles & Parallel Flows in High and Low-Field Side SOLs Cross-Field Flow Information Flux-Tube Particle Balance Analysis X-point Dependent Flow Boundary Conditions Imposed on Confined Plasma Inner Scanning Probe Outer Scanning Probe Connection to L-H Power Thresholds & Topology Inner Wall Probe: N. Smick, P-56 Inner Wall Doppler: K. Marr, P2-42
5 5 5 Rh o (m) 5 5 5 5 Rh o (m) 5 5 5 Scrape-off Layer Profiles Reveal Transport Asymmetries and Topology-Dependent Near-Sonic Parallel Plasma Flows LSN ev 2 m -3... Density Electron Temperature 347:75 RED - Inner SOL BLUE - Outer SOL Facing Up Facing Down USN ev 2 m -3... Density Electron Temperature 34:995.. Parallel Mach Number Bx B -. Parallel Mach Number Bx B -. 5 5 r (mm) 5 5 r (mm) Independent of Topology Lower average Te, higher average density on high-field side (inner) SOL Lowest Te are detected by inner probe when facing inner strike-point Dependent on Topology Inner SOL: Near-sonic parallel flow is co-current directed (+) in LSN, counter-current directed (-) in USN Outer SOL: Stronger co-current flow in LSN, weaker in USN
5 5.....4.3.2.. 5 25-25 -5 5 5 Near-Sonic Inner SOL Flows are Connected to Cross-field Transport Asymmetries Inner SOL Outer SOL Electron Pressure 2 ev m -3.3.2. RMS Jsat/<Jsat> 5 25 Bx B km s - 25-25 -5 5 Toroidal Projection of Parallel Velocity 5 r (mm) km s - -25
5 5.....4.3.2.. 5 25-25 -5 5 5 Near-Sonic Inner SOL Flows are Connected to Cross-field Transport Asymmetries Inner SOL Electron Pressure Outer SOL 2 ev m -3 Inner SOL plasma 'disappears' in Double Null! L nt reduced by factor of 4!.3.2. RMS Jsat/<Jsat> 5 25 Bx B km s - 25-25 -5 5 Toroidal Projection of Parallel Velocity 5 r (mm) km s - -25
5 5.....4.3.2.. 5 25-25 -5 5 5 Near-Sonic Inner SOL Flows are Connected to Cross-field Transport Asymmetries.3.2. 5 Inner SOL Electron Pressure Outer SOL 2 ev m -3 RMS Jsat/<Jsat> 25 Inner SOL plasma 'disappears' in Double Null! L nt reduced by factor of 4! Fluctuation levels persistently lower on inner SOL (See: J. Terry, O-9) Consistent with low ^ transport in inner SOL Bx B km s - 25-25 -5 5 Toroidal Projection of Parallel Velocity 5 r (mm) km s - -25
5 5.....4.3.2.. 5 25-25 -5 5 5 Near-Sonic Inner SOL Flows are Connected to Cross-field Transport Asymmetries Bx B.3.2. km s - 5 25-25 -5 Inner SOL Electron Pressure 5 Outer SOL 5 r (mm) 2 ev m -3 RMS Jsat/<Jsat> Toroidal Projection of Parallel Velocity 25 km s - -25 Inner SOL plasma 'disappears' in Double Null! L nt reduced by factor of 4! Fluctuation levels persistently lower on inner SOL (See: J. Terry, O-9) Consistent with low ^ transport in inner SOL Inner SOL // flows are always directed from outer to inner SOL in upper and lower-null, but ~stagnant in double-null
5 5.....4.3.2.. 5 25-25 -5 5 5 Near-Sonic Inner SOL Flows are Connected to Cross-field Transport Asymmetries Bx B.3.2. km s - 5 25-25 -5 Inner SOL Electron Pressure 5 Outer SOL 5 r (mm) 2 ev m -3 RMS Jsat/<Jsat> Toroidal Projection of Parallel Velocity 25 km s - -25 Inner SOL plasma 'disappears' in Double Null! L nt reduced by factor of 4! Fluctuation levels persistently lower on inner SOL (See: J. Terry, O-9) Consistent with low ^ transport in inner SOL Inner SOL // flows are always directed from outer to inner SOL in upper and lower-null, but ~stagnant in double-null Plasma exists on inner SOL because it flows along field lines from outer SOL!
5 5.....4.3.2.. 5 25-25 -5 5 5 Near-Sonic Inner SOL Flows are Connected to Cross-field Transport Asymmetries Bx B.3.2. km s - 5 25-25 -5 Inner SOL Electron Pressure 5 Outer SOL 5 r (mm) 2 ev m -3 RMS Jsat/<Jsat> Toroidal Projection of Parallel Velocity 25 km s - -25 Inner SOL plasma 'disappears' in Double Null! L nt reduced by factor of 4! Fluctuation levels persistently lower on inner SOL (See: J. Terry, O-9) Consistent with low ^ transport in inner SOL Outer SOL flows weaker, co-current, appear modulated by topology (Location is near poloidal flow stagnation point) Inner SOL // flows are always directed from outer to inner SOL in upper and lower-null, but ~stagnant in double-null Plasma exists on inner SOL because it flows along field lines from outer SOL!
Strong Parallel Flow or Strong Toroidal Rotation? Near-sonic parallel flows, V //, on inner SOL appear to be transport-driven, i.e., driven by ballooning-like cross-field transport asymmetries: V//f cocurrent countercurrent V//f Ip B T => link between Inner SOL flow direction/magnitude and magnetic topology However, total plasma flow vector, V, includes cross-field fluid drift, V^: Mach probe measurement V // V V^ => Need to examine cross-field flow information B
Inner SOL Plasma Flows: Total Plasma Flow Vector is Closely Aligned with Field Lines Data from C + "plumes" at inner midplane location C + light, CH 4 Puff Camera Vertical (cm) Vertical (cm) 6-6 -2 6-6 -2 55 nm Magnetic Field Line Direction of Flow Upper Null Discharge - 2 Toroidal Distance (cm) Lower Null Discharge Magnetic Field Line Direction of Flow - 2 Toroidal Distance (cm) Plasma flow direction depends on Upper/Lower Null topology, identical to that seen by Inner Mach Probe Impurity dispersal pattern is closely aligned with magnetic field line => Inner SOL flows dominated by parallel flow component => Direct evidence of wall-source impurity migration toward inner divertor leg D. Jablonski, et al., J. Nucl Mater. 24-243 (997) 782.
5 Outer SOL Plasma Flows: Largely Toroidal Rotation and/or Pfirsch-Schlüter Flow Outer probe data from matched discharges with normal and reversed Ip & B T Parallel Mach numbers reverse direction when Ip & B T reverse.5 -.5 Parallel Mach Number n/n G :.8 Normal B.24.3.36.8.36.3.24 Reversed B 5 r (mm) Similar reduction in flow as normalized density is increased => Consistent with parallel flow arising from co-current toroidal rotation and/or Pfirsch-Schlüter contributions Case of pure toroidal rotation... Vf= E r /B q Normal B: Reversed B: E r /B B V // E r /B Vf Vf V //...V // reverses with B B
Outline of Talk Profiles & Parallel Flows in High and Low-Field Side SOLs Diagnostics: Cross-Field Flow Information Flux-Tube Particle Balance Analysis Inner Scanning Probe Outer Scanning Probe X-point Topology-Dependent Flow Boundary Conditions Imposed on Confined Plasma Connection to L-H Power Thresholds & Topology Vertical Scanning Probe
Data can be Mapped to a "Flux-Tube Coordinate", S, Revealing Transport-Driven Component of Parallel Flow Definition of flux-tube coordinate, S Lower Null Upper Null.5 S.75.25.75.25 S Ip B T.5
Data can be Mapped to a "Flux-Tube Coordinate", S, Revealing Transport-Driven Component of Parallel Flow Definition of flux-tube coordinate, S Lower Null.5.75.25 S Ip B T Upper Null S.75.25.5 Data from matched Lower-Null and Upper-Null discharges 2 ev m -3 Mach# r = 4 mm Electron Pressure, 2 nte Mach Number, M // 2 ev m -3 2 nte(+ M // 2 /2) Outer SOL Inner SOL..25.5.75. Normalized distance along field line, S
Data can be Mapped to a "Flux-Tube Coordinate", S, Revealing Transport-Driven Component of Parallel Flow Definition of flux-tube coordinate, S Lower Null.5.75.25 S Ip B T Lower nte on Inner SOL Upper Null S.75.25.5 Data from matched Lower-Null and Upper-Null discharges 2 ev m -3 Mach# r = 4 mm Electron Pressure, 2 nte Mach Number, M // 2 ev m -3 2 nte(+ M // 2 /2) Outer SOL Inner SOL..25.5.75. Normalized distance along field line, S
Data can be Mapped to a "Flux-Tube Coordinate", S, Revealing Transport-Driven Component of Parallel Flow Definition of flux-tube coordinate, S.75 Lower Null.5 S.25 Ip B T.75 Upper Null S.5 Lower nte on Inner SOL.25 Transport-driven parallel flow from Outer to Inner SOL Toroidal rotation, Pfirsch-Schlüter flows,......appear as offsets to average of + Data from matched Lower-Null and Upper-Null discharges 2 ev m -3 Mach# 2 ev m -3 2 r = 4 mm Electron Pressure, nte Mach Number, M // 2 nte(+ M // 2 /2) ~ transport-driven // flow component Outer SOL Inner SOL..25.5.75. Normalized distance along field line, S
Data can be Mapped to a "Flux-Tube Coordinate", S, Revealing Transport-Driven Component of Parallel Flow Definition of flux-tube coordinate, S.75 Lower Null.5 S.25 Ip B T.75 Upper Null S.5 Lower nte on Inner SOL.25 Transport-driven parallel flow from Outer to Inner SOL Toroidal rotation, Pfirsch-Schlüter flows,......appear as offsets to average of + Thermal + flow energy ~constant Data from matched Lower-Null and Upper-Null discharges 2 ev m -3 Mach# 2 ev m -3 2 r = 4 mm Electron Pressure, nte Mach Number, M // 2 nte(+ M // 2 /2) ~ transport-driven // flow component Outer SOL Inner SOL..25.5.75. Normalized distance along field line, S
Data can be Mapped to a "Flux-Tube Coordinate", S, Revealing Transport-Driven Component of Parallel Flow Definition of flux-tube coordinate, S.75 Lower Null.5 S.25 Ip B T.75 Upper Null S.5 Lower nte on Inner SOL.25 Transport-driven parallel flow from Outer to Inner SOL Toroidal rotation, Pfirsch-Schlüter flows,......appear as offsets to average of + Thermal + flow energy ~constant Net flow of plasma out ends of flux tube requires a net particle source into (^) flux tube Data from matched Lower-Null and Upper-Null discharges 2 ev m -3 Mach# 2 ev m -3 2 r = 4 mm Electron Pressure, nte Mach Number, M // 2 nte(+ M // 2 /2) ~ transport-driven // flow component Outer SOL Inner SOL..25.5.75. Normalized distance along field line, S
..5.... Flux-Tube Particle Balance Analysis: Net Particle Source Profiles Exhibit Ballooning Asymmetry Parallel Mach Number Outer SOL r = 2 mm r = 6 mm Inner SOL.5 Flux-tube coordinate, S Probe Data: Upper Null Lower Null Model: Parallel Mach Number Net Particle Source...conserving particles & // momentum
..5.... Flux-Tube Particle Balance Analysis: Net Particle Source Profiles Exhibit Ballooning Asymmetry Parallel Mach Number Outer SOL r = 2 mm r = 6 mm Inner SOL.5 Flux-tube coordinate, S Probe Data: Upper Null Lower Null Model: Parallel Mach Number Net Particle Source...conserving particles & // momentum A transport-driven plasma circulation loop is implied! Cross-field transport overpopulates flux tubes on the low-field side, driving parallel flow The resultant circulation loop likely closes via cross-field fueling near the inner divertor region
..5.... Flux-Tube Particle Balance Analysis: Net Particle Source Profiles Exhibit Ballooning Asymmetry Parallel Mach Number Outer SOL r = 2 mm r = 6 mm Inner SOL.5 Flux-tube coordinate, S Probe Data: Upper Null Lower Null Model: Parallel Mach Number Net Particle Source...conserving particles & // momentum A transport-driven plasma circulation loop is implied! Cross-field transport overpopulates flux tubes on the low-field side, driving parallel flow The resultant circulation loop likely closes via cross-field fueling near the inner divertor region Note: A separate set of radial fluxes associated with main-chamber recycling and ionization can be superimposed
Outline of Talk Profiles & Parallel Flows in High and Low-Field Side SOLs Diagnostics: Ar 7+ X-ray Doppler Cross-Field Flow Information Flux-Tube Particle Balance Analysis Inner Scanning Probe Outer Scanning Probe Topology-Dependent Flow Boundary Conditions Imposed on Confined Plasma Connection to L-H Power Thresholds & Topology Vertical Scanning Probe
-5 - -5 5 5-5 - -5 5 5 X-point Topology Sets Magnitude and Direction of Transport- Driven SOL Flows => Core Plasma Rotation is Affected Ip B T Bx B Toroidal Projection of Parallel Velocity (km s-) 3-3 -6 2 Upper Null Double Null Lower Null Inner Probe r = 2 mm Outer Probe r = mm Toroidal Velocity (km s-) - -2-3 -4 Core Ar7+ Doppler -5 - -5 5 5 Distance Between Primary and Secondary Separatrix (mm)
-5 - -5 5 5-5 - -5 5 5 X-point Topology Sets Magnitude and Direction of Transport- Driven SOL Flows => Core Plasma Rotation is Affected Toroidal Projection of Parallel Velocity (km s-) 3-3 -6 2 Ip B T Upper Null Double Null Lower Null Bx B Inner Probe r = 2 mm Outer Probe r = mm Toroidal projections of flows near separatrix shift toward counter-current in sequence: lower => double => upper-null Toroidal Velocity (km s-) - -2-3 -4 Core Ar7+ Doppler -5 - -5 5 5 Distance Between Primary and Secondary Separatrix (mm)
-5 - -5 5 5-5 - -5 5 5 X-point Topology Sets Magnitude and Direction of Transport- Driven SOL Flows => Core Plasma Rotation is Affected Toroidal Projection of Parallel Velocity (km s-) 3-3 -6 2 Ip B T Upper Null Double Null Lower Null Bx B Inner Probe r = 2 mm Outer Probe r = mm Toroidal projections of flows near separatrix shift toward counter-current in sequence: lower => double => upper-null Central plasma toroidal rotation correspondingly shifts more toward counter-current direction Toroidal Velocity (km s-) - -2-3 -4 Core Ar7+ Doppler -5 - -5 5 5 Distance Between Primary and Secondary Separatrix (mm)
-5 - -5 5 5-5 - -5 5 5 X-point Topology Sets Magnitude and Direction of Transport- Driven SOL Flows => Core Plasma Rotation is Affected Toroidal Projection of Parallel Velocity (km s-) 3-3 -6 2 Ip B T Upper Null Double Null Lower Null Bx B Inner Probe r = 2 mm Outer Probe r = mm DV 5 km/s 2 km/s Toroidal projections of flows near separatrix shift toward counter-current in sequence: lower => double => upper-null Central plasma toroidal rotation correspondingly shifts more toward counter-current direction Toroidal Velocity (km s-) - -2-3 -4 Core Ar7+ Doppler -5 - -5 5 5 Distance Between Primary and Secondary Separatrix (mm) 8 km/s Toroidal velocity change is largest on inner SOL => suggests inner SOL flow is responsible for change in rotation of confined plasma! \Transport-driven SOL flows impose boundary conditions on confined plasma
If Transport-Driven SOL Flow/Rotation Paradigm is Correct, Radial Electric Fields in SOL Should Depend on X-point Topology ^ transport-driven parallel SOL flows V//f Ip B T V//f Ballooning-like transport leads to a helical flow component in the SOL with net volume-averaged toroidal momentum: co-current for lower null, counter-current for upper null
If Transport-Driven SOL Flow/Rotation Paradigm is Correct, Radial Electric Fields in SOL Should Depend on X-point Topology ^ transport-driven parallel SOL flows V//f V//f Ip B T Influence on plasma rotation Ballooning-like transport leads to a helical flow component in the SOL with net volume-averaged toroidal momentum: co-current for lower null, counter-current for upper null Being free to rotate only in the toroidal direction, the confined plasma acquires a corresponding co-current or counter-current rotation increment DVf DVf Ip B T
If Transport-Driven SOL Flow/Rotation Paradigm is Correct, Radial Electric Fields in SOL Should Depend on X-point Topology ^ transport-driven parallel SOL flows V//f V//f Ip B T Influence on plasma rotation Ballooning-like transport leads to a helical flow component in the SOL with net volume-averaged toroidal momentum: co-current for lower null, counter-current for upper null Being free to rotate only in the toroidal direction, the confined plasma acquires a corresponding co-current or counter-current rotation increment DVf stronger Er DVf weaker Er Via momentum coupling across separatrix, a topology-dependent toroidal rotation component, Er/B q, should appear in the SOL DErxB q Ip B T DErxB q => Stronger Er in SOL for lower null => Weaker Er in SOL for upper null
5 5 Plasma Potentials Near Separatrix Systematically Increase in the Sequence: Upper, Double, Lower-Null Plasma potential profiles estimated from sheath potential drop Caution: Accuracy of potential profile shape is uncertain! Estimated Plasma Potential (volts) 6 4 2 7 5 3 7 5 3 Inner Probe Outer Probe Vertical Probe 5 r (mm) Upper Null Double Null Lower Null More positive Er in SOL near separatrix in Lower-Null DEr/B q ~ 5 km/s, a significant fraction of measured change in parallel flow => Consistent with an increased co-current plasma rotation in lower-null, arising from transport-driven SOL plasma flows!
Connection Between Transport-Driven SOL Flows and Topology-Dependence of L-H Power Threshold? Transport-driven SOL flows lead to topology-dependent toroidal plasma rotation (and Er) near separatrix LSN DVf stronger Er DErxB q USN DVf weaker Er DErxB q Bx B Ip B T SOL widths are unchanged; Toroidal rotation fi near wall. => Implies toroidal velocity shear (ErxB shear) near separatrix is: stronger Bx B weaker Ip B T L-H transition is thought to involve velocity shear suppression of plasma turbulence => May explain why the L-H power threshold is lower when Bx B is pointing toward the x-point!
L-H Transition Coincides with Plasma Rotation Attaining Roughly the Same Value, Independent of Topology 2 m -3 MW 2 3 2 ev4 2 Line Averaged Density ICRF Power Electron Temperature y=.95 TS: ECE: Bx B Input power level to attain L H depends on x-point topology Ohmic+ICRF => no momentum input kev m - km s - 4-4 Max p e/ n e from TS in region.95 < y < Ar 7+ Toroidal Velocity -.2 -... L-H transition time (s)
L-H Transition Coincides with Plasma Rotation Attaining Roughly the Same Value, Independent of Topology 2 m -3 MW 2 3 2 ev4 2 Line Averaged Density ICRF Power Electron Temperature y=.95 TS: ECE: Bx B Input power level to attain L H depends on x-point topology Ohmic+ICRF => no momentum input kev m - km s - 4-4 Max p e/ n e from TS in region.95 < y < Ar 7+ Toroidal Velocity -.2 -... L-H transition time (s) Edge Te and electron pressure gradients at L H transition also different
L-H Transition Coincides with Plasma Rotation Attaining Roughly the Same Value, Independent of Topology 2 m -3 MW 2 3 2 ev4 2 Line Averaged Density ICRF Power Electron Temperature y=.95 TS: ECE: Bx B Input power level to attain L H depends on x-point topology Ohmic+ICRF => no momentum input kev m - km s - 4-4 Max p e/ n e from TS in region.95 < y < Ar 7+ Toroidal Velocity -.2 -... L-H transition time (s) Plasma rotation during ohmic phase starts out counter-current in USN... Edge Te and electron pressure gradients at L H transition also different...but ramps toward co-current as pressure gradients build up ---- SOL flow boundary condition!
L-H Transition Coincides with Plasma Rotation Attaining Roughly the Same Value, Independent of Topology 2 m -3 MW 2 3 2 ev4 2 Line Averaged Density ICRF Power Electron Temperature y=.95 TS: ECE: Bx B Input power level to attain L H depends on x-point topology Ohmic+ICRF => no momentum input kev m - km s - 4-4 Max p e/ n e from TS in region.95 < y < Ar 7+ Toroidal Velocity -.2 -... L-H transition time (s) Plasma rotation during ohmic phase starts out counter-current in USN... Edge Te and electron pressure gradients at L H transition also different...but ramps toward co-current as pressure gradients build up similar rotation at the L H transition! ---- SOL flow boundary condition! => Potential explanation for x-point topology dependence of L-H power threshold
Summary A cross field transport-driven plasma circulation loop is evident in ballooning-like transport x-point topology sets // flow direction promotes main-chamber impurity migration toward inner divertor near sonic // flows SOL flows set toroidal rotation boundary conditions for confined plasma x-point topology and toroidal rotation near separatrix are linked! stronger V//f DVf Er V//f DVf DErxB q Ip B T Bx B weaker Er DErxB q
Summary Possible explanation for the x-point dependence of L-H power threshold SOL flows + topology influence flow shear near separatrix DVf stronger co-current Vf and Vf shear DVf weaker co-current Vf and Vf shear Ip B T Bx B L-H threshold studies with different x-point topologies support hypothesis L-H transition is coincident with toroidal rotation achieving similar level, independent of x-point topology SOL flows impede co-current rotation with upper x-point Correspondingly, more input power (which promotes co-rotation) is required