INTERNAL TRANSPORT BARRIERS WITH

INTERNAL TRANSPORT BARRIERS WITH COUNTER-NEUTRAL BEAM INJECTION IN C.M. GREENFIELD, E.J. SYNAKOWSKI, K.H. BURRELL, M.E. AUSTIN, D.R. BAKER, L.R. BAYLOR, T.A. CASPER, J.C. DeBOO, E.J. DOYLE, D. ERNST, J.R. FERRON, P. GOHIL, A.W. HYATT, G.L. JACKSON, T.C. JERNIGAN, G.R. McKEE, M. MURAKAMI, R.I. PINSKER, M. PORKOLAB, R. PRATER, C.L. RETTIG, G.L. SCHMIDT, T.L. RHODES, B.W. RICE, G.M. STAEBLER, B.W. STALLARD, E.J. STRAIT, M.R. WADE, W.P. WEST, and L. ZENG QWER UCLAUCLA UCLA R E S E A R C H C E N T E R UNIVERSITY OF TEXAS UNIVERSITY OF WISCONSIN M A DIS O N

Internal Transport Barriers with Counter-Neutral Beam Injection in * C.M. Greenfield, E.J. Synakowski, K.H. Burrell, M.E. Austin, # D.R. Baker, L.R. Baylor, T.A. Casper, J.C. DeBoo, E.J. Doyle, D. Ernst, J.R. Ferron, P. Gohil, A.W. Hyatt, G.L. Jackson, T.C. Jernigan, G.R. McKee, M. Murakami, R.I. Pinsker, M. Porkolab, R. Prater, C.L. Rettig, T.L. Rhodes, B.W. Rice, G.M. Staebler, B.W. Stallard, E.J. Strait, M.R. Wade, W.P. West, L. Zeng General Atomics, San Diego, California, USA Princeton Plasma Physics Laboratory, Princeton, New Jersey, USA # University of Texas, Austin, Texas, USA Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA Lawrence Livermore National Laboratory, Livermore, California, USA University of California, Los Angeles, California, USA University of Wisconsin, Madison, Wisconsin, USA Massachusetts Institute of Technology, Cambridge, Massachusetts, USA Internal transport barriers are frequently formed in discharges with weak or negative central magnetic shear and moderate neutral beam (NBI) heating in the tokamak. In this regime, toroidal rotation, driven by co-directed (parallel to the plasma current) NBI, is the primary driver of the sheared E B velocity believed to suppress turbulence and its associated transport. With co-injection, the pressure gradient and toroidal rotation terms of the radial force balance equation ( E r = (Zien i) Pi νθibφ + νφibθ ) are in opposition, so that one must dominate in order to form a large radial electric field. Experiments are underway in to evaluate the potential for extending both the spatial extent and duration of these barriers through the use of counter-nbi. In these plasmas, the pressure gradient and rotation terms of the force balance add to each other, offering the potential for increased E B shear over an extended region of the plasma. At the same time, however, it is expected that the shearing rate ω E B will be smaller in the vicinity of the magnetic axis, suggesting a larger power threshold for establishing the barrier in this case. Experimental results thus far are consistent with this expectation. Transport barriers are difficult to form with moderate neutral beam heating, requiring a pellet trigger for formation, as was observed in TFTR 3 with balanced NBI. It is also expected that counter-nbi will aid efforts to extend the duration of the ITB by supporting an elevated minimum safety factor q min with counter-directed neutral beam current drive in the plasma core. Results of these experiments, both with higher power NBI and pellet assisted ITB formation will be discussed. * Work supported by U.S. Department of Energy under Contracts DE-AC3-99ER53, DE-AC- 7CH373, W-75-ENG-8, DE-AC5-9OR, and Grants DE-FG3-97ER55, DE-FG3-8ER535, and DE-FG3-9ER5373. C.M. Greenfield, et al., Phys. Plasmas, 59 (997). T.S. Hahm, K.H. Burrell, Phys. Plasmas, 8 (995). 3 M.G. Bell, et al., Phys. Plasmas, 7 (997).

OVERVIEW Internal transport barriers (ITB) have been formed with counter-neutral beam injection, with characteristics that contrast with co-injection experience. Barrier formation dynamics are fundamentally different than with co-injection. Alignment of p and rotation terms of radial force balance to maximize E r : E r = (Z i en i ) - p i - v θi B φ + v φi B θ Power threshold appearance may be consistent with expectations based on the familiar E B shear turbulence suppression hypothesis. Resulting ITBs may be a good target for efforts at transport barrier control. Nearly stationary q profiles maintained via counter-nbcd. Broader profiles May be more favorable for MHD stability. Pellet injection has been used to form an ITB with a strong barrier in the electron density and T i T e. Barriers formed both with high field side deuterium and low field side lithium pellets. These have been sustained for up to s. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

OUTLINE ECH preheat as a substitute for early neutral beams can form a target plasma with negative central magnetic shear (NCS). Internal transport barriers (ITB) have been formed with counter-neutral beam injection, with characteristics that contrast with co-injection experience. Power threshold. Nearly stationary q profiles. (Relatively) broad profiles. Pellet-triggered ITB characteristics. has been used to form an ITB with a strong barrier in the electron density and T i T e. Upgrades to pellet injection system for high-field side (HFS) deuterium pellet injection. ITBs with very strong density peaking. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

EARLY ELECTRON CYCLOTRON HEATING SLIGHTLY OFF- AXIS RESULTS IN INCREASED NEGATIVE CENTRAL SHEAR Poor beam ion confinement at low current contradicts the desire for a preheat phase to generate the conditions for ITB formation. Lost fast ions sputter carbon from walls. Large losses result in inefficient heating. Power from a single ECH gyrotron, applied early in the current ramp, produces the desired q profile without introducing fast ions to the plasma. Best results obtained when heating slightly off-axis (.5). On-axis absorption results in higher temperatures but drives instabilities. Little difference in q profile seen when resonance moved off-axis. Evidence that ECCD is dominant effect? MA MW kev. -. -...5..3 I P P ECH T e () q q min qmin 9985 (ECH@3ms) 9983 (Ohmic)....8. time (s) Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 3

ECH PREHEAT CAN PRODUCE A NEGATIVE CENTRAL SHEAR (NCS) TARGET ECH startup shown to be a viable alternative, even with only one gyrotron. Might be even better with multiple gyrotrons with independent aiming. 8 9 (Co-NBI, no RF) 9985 (ECH@.3s) 9983 (Ohmic) Further analysis is needed to understand the roles of current drive and heating. ECH preheat was used in all counterinjection ITB experiments......8. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

ITB OPTIMIZATION WITH COUNTER-NBI Goal of experiment: create, optimize and characterize core transport barriers formed with counter-injected neutral beams alone (no pellets). Used ECH preheat, although one shot without preheat also produced an ITB. ITBs were formed, but required more power than expected with co-injection. At full field, power threshold is around MW. Power threshold <5MW with co-nbi. Not always a clear bifurcation above threshold. Differences in barrier characteristics (compared to co-injected) More evident in density profile. Perhaps less so in ion profiles. Broader, with less steep gradients. Nondisruptive barrier termination. Work on sustainment was started. ITB sustained for a hew hundreds of ms after power stepdown (until neutral beams turned off or stepped up). Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 5

INTERNAL TRANSPORT BARRIER FORMATION WITH COUNTER-INJECTED NEUTRAL BEAMS I P Internal transport barrier experiments with counter-nbi follow a new recipe. ECH at.5 starting at.3s. High power neutral beams applied at.s (I P.9 MA). Later injection tried also, but best results found when NBI applied with elevated q. MHD stability behavior contrasts with co-injection discharges. Minidisruptions dump energy from plasma, can be cured. True disruptions are relatively infrequent. Sign of weaker pressure gradient than co-injected ITB? MA MW 5 s - 5 5 S N P ECH P NBI 998 (7 sources) 9987 ( sources) 9989 (5 sources) 9985 ( sources) 9985 (3 sources).5..5. time (s) Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

ITB FORMATION WITH NEUTRAL BEAM COUNTER- INJECTION REQUIRES HIGH POWER With neutral beam co-injection, core transport barriers format very low power and expand when P NBI 5MW. With counter-injection, clear barriers are formed only when P NBI exceeds MW. Barrier formation is delayed or prevented when heating power is decreased. Formation occurs following a short ELMing or dithering H mode phase. Time shown at right is delay from high power onset to end of ELMing/dithering phase; ITB formation probably begins slightly earlier. Short-lived, more localized barriers are observed on some lower power discharges. seconds.9.8.7..5..3.. Time from high power onset to ITB formation No ITB formed Weak or no ITB High power NBI onset:.s.8s.s.s/no ECH preheat (ELMing H-mode) 8 P NBI (MW) Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 7

FLUCTUATIONS ARE DECREASED DURING ITB FORMATION Fluctuation amplitude drop during H- mode phase indicates ITB formation begins during H-mode phase. Does ITB starve the edge of power and trigger return to L-mode? Jump at H L transition probably due to increased fluctuations near the edge. Increase prior to terminating MHD event may indicate beginning of barrier degradation. Spatially localized data expected to show clearer signature of ITB. FIR scattering integrates over region.3 r/a, includes large contribution from outside barrier. BES, reflectometer data currently being evaluated. a.u. khz.3.. 8 9987 ( sources).. MHD event 9988 (5 sources)...... time (s) ELMing/dithering H-mode 9985 ( sources) MHD event 9985 (3 sources) RMS amplitude Mean frequency FIR scattering: k θ = cm - r/a =.3- source discharge: ITB forms later at.s. 3 source discharge: No fully developed ITB. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 8

HIGHER POWER THRESHOLDS MAY BE EXPECTED FOR ITB FORMATION WITH COUNTER-NBI The Hahm-Burrell formula for the E B shearing rate: can be rewritten as: where ω ω RBθ Er E B = ( ) B ψ RBθ RBθ E B = ( ) f p + frot B n p p f p = ne ψ ψ ne ψ [ ], and f rot = Ω. ψ Choosing the coordinate system so that the sign of f p is the same with either co- or counter-nbi, we find that near the axis (ψ ), f p is positive, and For co-injection: f rot is positive. For counter-injection: f rot is negative. This results in cancellation of the two terms and smaller values of the shearing rate near the axis with counter-injection, perhaps leading to a higher power threshold. Moving away from the axis, the two terms may add, rather than cancel, with counterinjection, thereby allowing larger shearing rates and more transport barrier expansion. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 9

ITB FORMATION WITH COUNTER-INJECTION ITB formation evident in increasing separation of ion temperatures between nearby channels. Formation nearly coincident with return to L mode following ELMing or dithering H mode phase. kev 8.77m.83m.87m.9m.93m.9m.99m.m.m.8m.m.5m T i (CER) ITB terminated by nondisruptive MHD event. Regrowth occurs following MHD event. ELMing or dithering H mode phase repeats prior to barrier regrowth. MW 5 s - 5 5.5..5..5. P NBI S N..8.. t (sec) 998 Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

PROFILE PEAKING IS SEEN IN ALL KINETIC PROFILES DURING ITB FORMATION Profiles from second ITB formation in 7-NBI source discharge Peaking evident in all kinetic channels q profile relatively stationary Central Z eff remains at.5 while edge carbon content is increasing. s - [ 5 ] kev 3. -.5 -. -.5 T e Ω.9.95..5. -......8. kev 8 5 3 T i Z eff.....8. m -3 [ 9 ] 8 5 3 q n e.....8. 998 Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

NEOCLASSICAL ION DIFFUSIVITY RECOVERED WITH COUNTER-NBI χ e χ i Neoclassical levels obtained in ion thermal diffusivity. Some reduction seen in χ e, D e and χ ϕ. Smaller reductions: error analysis is in progress. m /s m /s.. D e 998B.9 998B.. χ ϕ χc-h neo i.....8.......8. 998 Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

IS THE ITB POSITION CONTROLLED BY CANCELLATION OF RADIAL ELECTRIC FIELD TERMS WITH CO-INJECTION? Shearing rate ω E B believed to control transport barrier behavior. In most co-injected discharges, the pressure gradient and rotation terms of the radial electric field cancel at some radius. Does this lock ITB position? qmin > ITB in co-nbi discharge shown. No such cancellation in counter-injected discharges. May allow broader profiles. qmin < ITB in counter-nbi discharge shown. kv/m kv/m 5 5-5 - - - - E r total E p (ITB) 998.9 (ctr) E p 9. (co) E v B E v B E r total q min 5 s - 5 s - -3 5-8 - q (ITB) min - q min (ITB).....8......8. 3 - - 3 ω E B (ITB) ω E B q min Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 3

COUNTER-INJECTION CALCULATED TO SUPPRESS TURBULENCE WITHIN A LARGE RADIUS 5 998. 5 cm/s 3 Transport barrier region ω E B γ max.....8. Calculated linear growth rate γ max for low-k turbulence (ITG modes) negligible at <.5. Shearing rate ω E B exceeds calculated linear growth rate at <.7. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

8 NEUTRAL BEAM COUNTER-CURRENT DRIVE INHIBITS CURRENT PROFILE EVOLUTION DURING THE ITB PHASE 8 Co-NBI (9) Counter-NBI (998).s.s.3s.85s.95s.5s.....8. In a ms interval in the presence of an ITB, the entire q profile significantly decreases due to current diffusion. This is an eventual limitation to ITB performance, as the barrier is always terminated if q min reaches unity. Counter-NBI provides central counter-current drive, which can maintains an elevated q profile. Necessary, but not sufficient, for ITB sustainment......8. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 5

POWER STEPDOWN EXPERIMENTS ARE A PROMISING BEGINNING TO ITB CONTROL EFFORTS An element of the ITB experiment plan was to use neutral beam power modulation to control the ITB position. Although we did not have time for extensive studies, we had a few shots where the power was stepped down from to either or 5 sources. Difference in performance between two - source discharges may point to TFTR-like bifurcation. Highest performance was obtained on a discharge which stepped down from to sources, and back up to 5. Finer control of beam power is needed to exploit this technique as a control mechanism. MA MW 5 s - 5 5 I P S N P ECH 9987 ( sources) 99857 ( sources) 99859 ( 5 sources) 998 ( 5 sources) P NBI.5..5. time (s) Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

PROFILES ARE BROADER WITH COUNTER-NBI Similar discharges Lower current in counter- NBI discharge since data taken earlier in current ramp. Both discharges have absorbed NB power 9- MW and similar stored energies ~MJ. Peak values of profiles lower with counter, but profiles are broader. kev s - [ 5 ] 5 3 3 T e Ω CTR: 9989B.s CO: 873F8.8s.....8. kev 5 5 5 3 Z eff T i.....8. m -3 [ 9 ] 8 8.....8. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 7 q n e

PREHEAT IS NOT A NECESSARY CONDITION FOR ITB FORMATION Identical discharges, except 9988 has no ECH preheat. Initial profiles (dashed) very similar. Broader, flattened region in core of T e profile after end of ECH pulse applied at.5. q profile more reversed with ECH q profiles nearly identical at peak performance. ECH preheat discharge rotates faster, slightly higher T i. s - [ 5 ] kev 3.5. -.5 -. -.5 9988A5 (no ECH) 9989B (ECH preheat) 8 T e Ω -......8. kev 8 5 3 Z eff T i.....8. m -3 [ 9 ] 8 q t=. s t=. s.....8. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 8 n e

PELLETS CAN BE USED TO TRIGGER ITB FORMATION WITH DIFFERENT BARRIER CHARACTERISTICS Recent modifications to the pellet injection system allow deep penetration of deuterium pellets. Both deuterium and lithium pellets have been used to trigger formation of the ITB. Similar to PEP regimes previously observed in other devices. Barriers exhibit high gradients in both the density and temperature profiles. Lower temperatures and higher densities observed. T i / T e.5, much lower than observed with neutral beams alone. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 9

PELLET INJECTION PROGRAM Modifications to injector (previously on JET, 987-99) Three.7 mm guns. Variety of pellet velocities possible. Two independent guide tubes on centerpost (HFS). Can be connected to any of the guns or a gas valve. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

HIGH FIELD SIDE (HFS 5 ) PELLET INJECTION YIELDS DEEPER PARTICLE DEPOSITION THAN LFS INJECTION.7 mm Pellets - HFS 5 vs LFS HFS 5 V+ LFS n e ( m -3 )..5 DIII-D 9879 - measured n e HFS 5 v p = 8 m/s t = 5 ms HFS mid Calculated Penetration LFS v p = 58 m/s t = ms Four positions of pellet injection guide tubes installed on DIII-D......8.. Net deposition is much deeper for the lower velocity HFS 5 pellets. The pellets were injected into the same discharge under the same conditions (ELMing H- mode,.5mw NBI, T e () = 3keV). Net deposition profile measured by Thomson scattering -5 ms after injection. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

n e ( 9 m -3 ) P NBI (MW) H_L89P 5 5 DIII-D 9978 HFS PELLETS DURING CURRENT RISE FORM AN INTERNAL TRANSPORT BARRIER.5..5. Time (s).7 mm pellets injected during current rise from the HFS 5 location produces peaked density profile for ITB studies with T i T e. MW of counter-nbi applied to produce ITB in L mode. ne ( 9 m -3 ) Te, Ti (kev) 5 5 3 T e - TS DIII-D 9978.s n e - TS T i - CER....8.. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

DENSITY AND PRESSURE PROFILES ARE STRONGLY PEAKED ne (m -3 )....8... 99788.9s DIII-D Counter NBI MW 9978 PEP.9s......8. P (Pa) e+5 8e+ e+ e+ e+ 99788.9s.....8. DIII-D Counter NBI MW 9978 PEP.9s Density profile from.7mm pellet is strongly peaked with steep density gradient formed at =. that persists for.7s. Pressure profile is also strongly peaked with a steep gradient at =. with T i T e. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 3

ION THERMAL TRANSPORT IS REDUCED IN THE PELLET TRIGGERED ITB χ e χ i m /s 9978.s (pellet) 99788.s (no pellet) χ i neo.....8.....8. Ion thermal diffusivity is reduced in the core following introduction of the deuterium pellet. Little or no reduction is seen in the electron channel. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

STRONG OFF-AXIS BOOTSTRAP CURRENT AND NEGATIVE CENTRAL SHEAR OBSERVED IN PEP MODE JBS (A/cm ) 8 9978 PEP.9s 99788.9s 99788.....8......8. DIII-D Counter NBI MW q 8 7 5 3 9978 PEP.9s DIII-D Counter NBI MW Bootstrap current calculated with NCLASS shows strong off-axis contribution to current density profile. q profile determined using MSE indicates strong negative magnetic central shear during PEP phase. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 5

TOROIDAL ROTATION PROFILE SHAPE DIFFERS BETWEEN CO- AND COUNTER-INJECTED PEP MODES Toroidal rotation ( rad/s). 8.... DIII-D t =.898s 9987 co-nbi......8.. Notch in co-injected rotation profiles similar to that seen in TFTR supershots. Attributed to neoclassical parallel momentum exchange. 9978 counter-nbi Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99

SUMMARY Internal transport barriers (ITB) have been formed with counter-neutral beam injection, with characteristics that contrast with co-injection experience. Power threshold appearance may be consistent with expectations based on the familiar E B shear turbulence suppression hypothesis. Nearly stationary q profiles produced via counter-nbcd. Broader profiles May be more favorable for MHD stability. First attempts at barrier control are promising. Pellet injection has been used to form an ITB with a strong barrier in the electron density and T i T e. Barriers formed both with high field side deuterium and low field side lithium. These have been sustained for up to s. Greenfield 7th IAEA TCM on H-Mode and Transport Barrier Physics 9/99 7