1. ISPDB_CERC and materials for NC discussion from LHD
M.Yokoyama (NIFS)
Special thanks to colleagues in CERC activity and in NIFS

2. Initiated in 2005 as the kick-off topic of broad ISPDB activity
ISPDB_CERC
Initiated in 2005 as the kick-off topic of broad ISPDB activity
CHS, LHD, TJ-II and W7-AS can provide sufficient information
Electron-root feature (specific in helical systems) is the common underline physics
NC transport codes : applied to estimate electron heat diffusivity (strong linkage to the international collaboration)
DCOM : LHD, MOCA : TJ-II, DKES : W7-AS
Analytical formulae : CHS and LHD (now with GSRAKE)
Presentations and Documentations
15th Stellarator Workshop (Madrid, 2005)
Review Talk : “Internal Transport Barrier Physics in Helical Systems”
[First documentation] “Common Features of Core Electron-Root Confinement in Helical Devices” Fusion Science and Technology 50(2006) Commonality

3. Commonality and Difference : more or less summarized
ISPDB_CERC
21st IAEA FEC (Chengdu, 2006)
Oral: “Core Electron-Root Confinement (CERC) in Helical Plasmas”, also mentioned in EX/Summary talk
Will be accepted in Nuclear Fusion soon : Difference
eff (CERC appearance becomes easier for higher eff)
O1 vs X2
Commonality and Difference : more or less summarized
Interesting topic related to the effects of rational surface on CERC establishment (e.g, TJ-II and LHD)
Appropriate analysis has yet been available  further materials in ISPDB sense may not be available for a while…
Registration of CERC discharges on “real” database

4. Example : LHD CERC 32940-2s [K.Ida et al., PRL 91(2003)085003]

5. Example : U files (ITPA database format) of LHD CERC 32940-2s
Created by H.Funaba
Established TASK – ITPA DB linkage can be utilized (TASK is the core of the 3D-integrated code development in Japan)

6. Example : Te-profile data === based on LHD equilibirum database
Te()
selected from VMEC euqilibrium database  R --  transformation

7. Procedure:
Te(R) is projected onto Te() utilizing VMEC equilibrium database (in this case, by utilzing sequential data from “Rax=3.75m vacuum“ case: different P profiles, beta values)
symmetry of Te() w.r.t. =0
Then, lhd-r375q100b016a2020.flx is selected
 Corresponding configuration files (VMEC input file, Boozer) are available

8. Waveform (with documentation)

9. Search capability is also required (to pick up shots of interest)
U file (ITPA) or EFDA format (or both with interface)
How to deal with equilibirum issue ?
Same format in all ISPDB topics is favorable : Any Request or Opinions ?
…..
registration of shots onto the real database should be started (even it is step by step)

10. i reduction with electron-root at the periphral region (NBI)
Materials for NC discussion from LHD Examples of transport analysis so far
CERC (ECH)
i reduction with electron-root at the periphral region (NBI)
Dimensionally similar discharges (Rax, elongation)
Configuration (Rax) scan of ECH low-collisional plasmas
exp >(>) NC in these examples: which has led us (NIFS) to consider the existence of “anomalous (turbulent)” transport
The GKV simulations on the ITG turbulence in helical systems : T.H.Watanabe and H.Sugama (IAEA FEC 2006)
stronger instability in the inward-shifted configuration
because of the larger residual zonal flows, however, the resultant transport is found in comparable magnitude to the standard configuration

11. 1-1 LHD CERC 32940-2s [K.Ida et al., PRL 91(2003)085003]
Fairly good agreement between GSRAKE and DCOM
Also quite good agreement with CXRS
Same profiles were used for GSRAKE and DCOM calculations

12. 1-2 Fluxes estimated by GSRAKE and DCOM differ each other
NC e (with ambipolar Er) is about one-order smaller than NC e (Er=0)
DCOM and GSRAKE give different ambipolar particle flux and heat fluxes (e ) even for similar Er value (e.g., around =0.5)  What is the reason ?
Benchmarking of NC codes based on real experiment (beyond the mono-energetic coefficient) might be worthwhile.  Discussion
NC e (both GSRAKE and DCOM) is about one-order smaller than e,exp around =0.2

13. 2-1 i reduction with electron-root at the periphral region (NBI) [K
2-1 i reduction with electron-root at the periphral region (NBI) [K.Ida et al., PRL 86 (2001)5297]
Density scan  ion root to electron root
Reduction of i with the Er (electron-root)
i,NC can not explain i,exp for large Er  anomalous contribution

14. 3-1 Dimensionally similar discharges (H.Yamada ICPP2006) -- Elongation
Central heating of ECH
Operational parameters
are the same except for
elongation
 Same r*, n*, b,….
PECH = 0.93MW for k=0.8
0.35MW for k=1.0
Power
deposition
zone

15. 3-2 Dimensionally similar discharges (H.Yamada ICPP2006) -- Elongation
Heat transport in k=1 is smaller than k=0.8 with a factor of 2-3
e,exp >(>) e,NC for both cases
Trend is close to neoclassical prediction
 Configuration effect on anomalous transport is correlated (or happens
to coincide) with nature of neoclassical transport !?

16. Configuration scan (Centrally focused ECH)
4-1 Effect of NC-optimization on Electron Heat Transport in Low-Collisional LHD Plasmas, S.Murakami et al., FST 51(2007) 112.
Configuration scan (Centrally focused ECH)
Rax=3.45, 3.53, 3.6, 3.75, 3.9 m
ρ=2/3
Clear Dependence of fren on Ɛeff observed  NC optimization works !

17. But, reduces once the electron-root is realized
4-2 Exp. vs. NC (DCOM) Electron Heat Flux (Ti=Te/2 from Ti(0)
Total exp. flux = 0.88 MW (ECH Power)
Rax=3.6m
Rax=3.53m
But, reduces once the electron-root is realized
Rax=3.75m
Rax=3.9m
NC heat flux becomes more than half of Exp. value at the highest Te (ion-root solution)

18. Materials for NC discussion from LHD
Summary
ISPDB_CERC has been successfully promoted : presentations and documentations  Discussion of real “Database“
Materials for NC discussion from LHD
NC code benchmarking for experimental shots
NC optimization (reduction of eff) works for global confinement improvement (ISS04)
But, NC prediction can not explain radial heat transport ((Q)exp >(>) (Q)NC)
 Interpretation : existence of anomalous (turbulent) transport (e.g., ITG turbulence study)
We will see examples from W7-AS in the following talk.
 NC discussion

21. 1-3 Proposal of NC code benchmarking on experimental situation
If we plot thermodiffusion part, DCOM and GSRAKE give almost same values  Flux dependence on Er is different ?

24. 3-1 Dimensionally similar discharges (H.Yamada ICPP2006) -- Rax
Comparison of cases with Rax=3.6m and 3.75m in LHD
Almost equivalent Te and ne profiles.
 1.2&1.73MW 65% : larger power for Rax=3.75m
1. Loss of high energetic particle in the slowing down process is almost the same.
11.0% in 3.6m  10.4% in 3.75m
2. A large difference with a factor of 2 exists in neoclassical ion heat conduction loss.

25. 3-2 Dimensionally similar discharges (H.Yamada ICPP2006) -- Rax
Comparison of cases with Rax=3.6m and 3.75m in LHD
Almost equivalent Te and ne profiles.
 1.2&1.73MW 65% : larger power for Rax=3.75m
1. Loss of high energetic particle in the slowing down process is almost the same.
11.0% in 3.6m  10.4% in 3.75m
2. A large difference with a factor of 2 exists in neoclassical ion heat conduction loss.
GSRAKE
Neoclassical transport does not explain this difference

26. Neoclassical optimization is pronounced in deep collisionless regime
nb* ~ 0.03 (much lower than the reactor condition)
For the case with Rax=3.6m
 Neoclassical heat conduction loss :
still less than 25 % of the heating power
 If the same condition is assumed for the case
with Rax= 3.75m
 N.C. conduction loss exceeds the heating power
 never happened  not realized

27. ρ=0.75: dominated by anomalous transport
4-3 Exp. vs. NC (DCOM) Electron Heat Flux (Ti=Te/2 from Ti(0)
Rax=3.6m
Rax=3.53m
Rax=3.75m
Rax=3.9m
(ρ=0.65)
Anomalous transport dominates at the edge region (especially in the NC-optimized configuration)
ρ=0.75: dominated by anomalous transport

28. Effect of NC-optimization on Electron Heat Transport in Low-Collisional LHD Plasmas, S.Murakami et al., FST 51(2007) 112.
Experimental e,eff
ρ=0.5
Rax=3.53m
Rax=3.6m
Rax=3.75m
Rax=3.9m
A lower e,eff observed in smaller-Ɛeff configuration (e.g., 3.53m)

29. Effect of NC-optimization on Electron Heat Transport in Low-Collisional LHD Plasmas, S.Murakami et al., FST 51(2007) 112.
Experimental e,eff
ρ=0.75
Rax=3.53m
Rax=3.6m
Rax=3.75m
Rax=3.9m
(ρ=0.65)
A lower e,eff observed in smaller-Ɛeff configuration (e.g., 3.53m)

30. cf., Te in various Rax ρ=0.5 ρ=0.75 Rax=3.53m Rax=3.6m Rax=3.53m
Effect of NC-optimization on Electron Heat Transport in Low-Collisional LHD Plasmas, S.Murakami et al., FST 51(2007) 112.
cf., Te in various Rax
ρ=0.5
ρ=0.75
Rax=3.53m
Rax=3.6m
Rax=3.53m
Rax=3.6m
Rax=3.75m
Rax=3.9m
Rax=3.75m
Rax=3.9m
(ρ=0.65)
Electron root feature

31. H.Funaba ISCDB NC