1. Particle Transport And Density Peaking At Low Collisionality On Alcator C-Mod
49th Annual Meeting of APS - DPP
Orlando, 11/14/2007
M. Greenwald, J.W. Hughes, D. Mikkelsen, J. Terry, Alcator Group
C. Angioni, H. Weisen

2. Particle Transport and Density Profiles
We want to be able to predict density profile
Better fusion performance with moderate density peaking
Effects on stability, divertor operation etc.
Results from ASDEX (Angioni et al., PRL 2003), JET (H. Weisen, et al., NF 2005) show increase in density peaking at low n* for H-mode plasmas.
Central fueling (NBI) can play an important role as well.
Scales to ITER (with weak fueling): ne(0)/<ne> ~
In this talk, we’ll also look at an additional effect: the role of safety factor (or magnetic shear) in the particle transport

3. Lower Density H-Modes Show Modest Profile Peaking
BT = 5.2 T
IP = 1.0 MA
PIN = 3.9 MW
H98 = 1.05
BT = 5.4 T
IP = 0.9 MA
PIN = 2.3 MW
H98 = 1.3

4. C-Mod Data Helps Break Covariance Between nEFF and n/nG Makes Extrapolation To ITER More Certain.
ICRF Only
ITER
ICRF Only
ITER
nEFF= nei/wD ≡ 0.1RZEFF<ne>/<Te>2

5. More Density Peaking With NBI Heating (Fueling)
ICRF Only
NBI + ICRF

6. C-Mod Data Fit Reasonably By Scaling Derived for JET/AUG
ne(0.2)/<ne> = – 0.117log(nEFF) GNBI – 4.03bT

7. This Isn’t The Whole Story: q95 Dependence Observed on C-Mod
C-Mod ICRF Data

8. At Same Collisionality More Peaking Seen As q95 Is Increased
C-Mod ICRF Data

9. At Same Collisionality More Peaking Seen As q95 Is Increased
C-Mod ICRF Data

10. Main Difference In Density Profiles Is Extent Of Peaked Region
Magnetic Axis
Separatrix
R/Ln ~ 3.1
R/Ln ~ 2.2
Time to establish peaking << a/VWare

11. Residuals from JET/AUG scaling vs q95
No NBI data
Apparent linear scaling with q95
ne(0.2)/<ne> = – 0.117log(nEFF) GNBI – 4.03bT
ICRF Data

12. New Scaling Including q95 Dependence
Weak linear q dependence added to previous scaling
Best fit to C-Mod data
(JET and AUG not included)
ne(0.2)/<ne> = 1.1 – 0.117log(nEFF) GNBI – 4.03bT q95

13. q95 Dependence Not As Evident In JET & AUG Data
When important dependences are factored out, q95 (or li) have small statistical significance for JET/AUG scaling. (Angioni, Weisen, et al, NF 2007).
We haven’t closed the loop with this new C-Mod data.
ICRF + NBI Data

14. Initial Simulation Work Begun
GYRO simulations have been performed using profile data from C-Mod shots (see Mikkelsen poster NP )
Density profile is adjusted to achieve zero particle flux
With R/LTi > R/LTe (as measured), null flux is seen at the observed density gradient
(Raising gradient by 20% produces outward flux)
Physics seems to be reduction of ITG instability-drive, enabling TEM-driven pinch with kqri > 0.5
In this analysis, collisionality dependence is through relaxation of LTi from reduced ion-electron coupling – raising nEFF by factor of 2 in simulations did not remove pinch.
Differences/similarities with AUG and JET work need to be explored.

15. Summary
Collisionality is the leading dependency for density profile peaking in H-modes.
C-Mod data helps break covariance between nEFF and n/nG (Greenwald et al., Nucl. Fus. L26, 2007)
In C-Mod, secondary effect found: weak q scaling
Commonly seen in L-Mode (see Baker DIII-D, Weisen TCV)
This dependence is not as evident in JET and AUG data
Peaking/particle transport is beginning to be understood
Initial simulations suggest that interplay of ITG and TEM turbulence is the key
More simulations/analysis required
Experiments to look for changes in fluctuations are planned