R Sartori

1
Impact of rotation and momentum input on the L-H
transition

• R Sartori
• Thank you to the people who contributed

2

• Outline of the talk
• L-H threshold and
• Changes in rotation by changing NBI torque
  (DIII-D)
• Changes in rotation due to changes in SOL flow
  (Alcator C-Mod)
• Changes in rotation due to TF ripple (JET)
• Changes in rotation due to changes in momentum
  input NBI vs no momentum input ICH and ECH
  (Asdex Upgrade and JET)
• Changes in rotation due to RMP

3

• DIII-D L-H threshold depends on applied beam
  torque and toroidal rotation

P Gohil, H-mode workshop 2007

• The L-H threshold power is reduced by a factor of
  3 going from prevalent co-injection to prevalent
  counter in USN (ion grad-B drift away from
  X-point)
• Smaller reduction of PLH in LSN ? 80
• No difference between USN and LSN for counter and
  balanced injection
• Reduction in edge toroidal rotation is
  correlated with the observed reduction in L-H
  threshold power

4

• DIII-D L-H threshold depends on applied beam
  torque and toroidal rotation

The L-H transition can be induced by reduction of
applied torque at constant power
5
The experiment

• DIII-D L-H threshold dependence on
  torque/rotation is independent of ion mass

P Gohil, NF, to be published

• The increase of PLH with torque is seen also in
  Hydrogen plasmas
• In Hydrogen a factor of 2 larger threshold
  observed in the range of torque explored

6
Description of the Experiment

• DIII-D Changes in rotation/torque correlate with
  edge turbulence dynamics

GR McKee, IAEA 2008 D J Schlossberg, Phys. Of
Plasmas, 2009

• Changes in fluctuation level and flow pattern
  between co-injection and balanced injection
• Turbulence flow reversal in balanced injection ?
  poloidal flow shear

• spectra obtained in 100 ms before L-H. Phase
  shift of poloidally displaced channels Z1.2cm.

• Correlation between changes in injected torque
  and poloidal group velocity of turbulence near
  edge

7
Description of the Experiment

• DIII-D Increased Er shear at lower rotation

• Interplay between pressure gradient and v X B
  terms in Er
• lower contribution of toroidal rotation term ?
  increase in Er shear at the edge ? de-correlation
  and suppression of turbulence .

8

• Alcator C-Mod Changes in L-H threshold
  correlated with changes in SOL flows/rotation

From H Hubbard, EPS 2004

• Changes in core rotation correlated with changes
  of SOL flows (largest in inner SOL)

• PLH two times higher in USN (ion ?B drift away
  from X-point )
• Strong parallel flows are driven in the inner
  SOL, co-current for LSN, counter current for USN
• At the L-H threshold flows and rotation near a
  constant value, independent of configuration ?
  strong suggestion that SOL flows are causing
  differences in PLH with configuration

9
Comparison of density and temperature and
toroidal rotation profiles

• JET Variation of toroidal rotation due to TF
  ripple changes

1.5 ripple L-modes with low level of beam power
P De Vries et al, NF 48 2008

• Lower rotation with increasing ripple. Edge
  rotation becomes negative at high ripple
• ASCOT calculation ? that dominant mechanism for
  reduction of rotation is banana orbit diffusion ?
  induces a radial return current ? JXB torque in
  counter current direction. For ripple of 0.5
  this torque was found to be 20-30 of that
  supplied by the NB system

10
Comparison of power and torque density

• JET L-H threshold is not affected by large
  variation in toroidal rotation due to TF ripple

Y Andrew et al, PPCF 50 2008

• 0.08 ripple ? v? in co-current direction
• 1 ripple ? v? in counter-current direction
• Similar Te and Ti the transition
• v? unaffected by ripple

11
Global and pedestal confinement is similar

• Asdex Upgrade L-H threshold is independent of
  heating method ?momentum input/rotation

F Ryter et al, NF 49 2009

• No difference in L-H power threshold and in the
  density dependence of PLH between ICH and NBI,
  for both D and He plasmas

12

• JET L-H threshold is independent of heating
  method ?momentum input/rotation

Data from 1MA/1T to 3.8MA/3.8T
LH scaling from Righi et al NF 39, 1999
Andrew et al, PPCF 48 2006

• JET results shows that L-H access is independent
  of heating method, as seen previously in JET for
  all H isotopes
• New data confirms the results and shows typical
  changes from 6-8 kHz (NBIICH) down to 2.5KHz
  (ICH) at 3T with similar threshold

13
Effect of RMPs on the L-H transition Summary
from Carlos Hidalgo- H-mode workshop talk
JET No effect observed (2009 experimental
campaign)
MAST On MAST there is observation of the delay of
the L-H transition time if the n1 coils are
applied before the L-H transition but not with
n3.
DIII-D In Helium plasmas PL-H increased by gt50
in the presence of n3 RMP perturbations.  
NSTX PL-H increases from 1.4 to 2.6 MW with
higher n3 current (65 increase for Pheat/ne)
14
Summary and future work

• NXTS and JET results on L-H threshold with RMP

• PLH unaffected by application of RMP? if effect
  exists it depends on the strenght of the
  perturbation
• 30 decrease in rotation

• PLH increases from 1.4 to 2.6 MW with higher n3
  current

15

• Summary

• Evidence from DIII-D of strong effect of beam
  torque on L-H threshold
• The effect persist when mass (and X-point
  height) are changed
• For counter-injection, it reduces differences
  between favourable and unfavourable grad-B
  direction
• The effect is less strong in favourable grad-B
  direction (why?), therefore effect on threshold
  scaling not as large as largest reported
  variations
• When rotation is changed by changing heating
  methods and ripple the L-H threshold is
  unchanged, with RMP it increases or is unvaried ?
  suggest weaker link with rotation in DIII-D
  results

16
Comparison with same density and total power
JET data LSN configuration and reversed Bt and
Ip

• Reversed Bt direction? ion ?B drift away from
  X-point and counter injection
• Forward Bt direction ?ion ?B drift towards
  X-point and co injection

• No difference between L-H power threshold with
  ion ?B drift direction in the JET data.