1. 14 Oct. 2009, S. Masuzaki 1/18 Edge Heat Transport in the Helical Divertor Configuration in LHD S. Masuzaki, M. Kobayashi, T. Murase, T. Morisaki, N. Ohyabu, H. Yamada, A. Komori and LHD Experiment group National Institute for Fusion Science

2. 14 Oct. 2009, S. Masuzaki 2/18 MOTIVATION Understanding of heat and particle transport in edge region is essential for divertor design in fusion reactor. –Complex field line structure in which stochastic region, islands and laminar region coexist exists in edge region in heliotron-type devices, stellarators and tokamaks with RMP. –Different mechanism which determines divertor heat and particle flux profiles from poloidal divertor tokamaks may exist. In this study, we focus on profiles of heat and particle flux on helical divertor in LHD heliotron.

3. 14 Oct. 2009, S. Masuzaki 3/18 OUTLINE Edge magnetic field line structure in LHD Relation between the structure and profiles of particle and heat load on divertor Transport change with T e rise Summary

4. 14 Oct. 2009, S. Masuzaki 4/18 Edge field line structure in LHD RRR R  = 0º  = 9º  = 18º R ax =3.6m R ax =3.75m Helical divertor SOL has three dimensional structure. In the HD SOL, the stochastic region, islands and laminar layer co-exist. The fine structure in the HD SOL varies with the operational magnetic structure.

5. 14 Oct. 2009, S. Masuzaki 5/18 Edge field line structure in LHD R R  = 0º R ax =3.6m R ax =3.75m L c (m)

6. 14 Oct. 2009, S. Masuzaki 6/18 Divertor leg Torus center － trace CW － trace CCW Z R ax =3.5m R ax =3.6m Outer of this line, field lines interrupted by divertor plates or first wall within about ten meter divertor leg Z (m) R ax =3.9m In divertor region, poloidal field component is larger than toroidal one

7. 14 Oct. 2009, S. Masuzaki 7/18 Divertor Plate Array Divertor plate arrays and diagnostics Langmuir probes and thermocouples are embedded in divertor plates An IR camera observes inboard divertor plates 1,700 water cooled graphite tiles IR camera Divertor plates Private region

8. 14 Oct. 2009, S. Masuzaki 8/18 Particle Flux Profiles on a Divetor Plate  =0° 36° 72° Outboard Inboard Outboard Bottom Top Divertor plate array Probe array Inward shifted Rax Outward shifted Rax

9. 14 Oct. 2009, S. Masuzaki 9/18 Heat Flux Profiles on a Divertor Plate Heat flux profiles were reconstructed by using temperature rise profiles at the beginning of NB injection. Semi-infinite assumption was applied neglecting three dimensional heat diffusion.

10. 14 Oct. 2009, S. Masuzaki 10/18 Heat and Particle Deposition Profiles on divertor Toroidal angle,  (°) R ax =3.75m R ax =3.6m 252015105 0 red blue red blue Particle deposition patterns on the HD. (  =0) (Calculation results of field lines tracing) inboard outboard bottom top Particle number (cal.) Normalized temperature rise Toroidal angle (°) Profile of normalized temperature rise measured by thermocouples, and particle deposition obtained by field line tracing.  =0° 36° 72° Outboard Inboard Outboard Top

11. 14 Oct. 2009, S. Masuzaki 11/18 Change of heat flux profile during discharge (Rax=3.60m) Such change of heat flux profile has been observed frequently. Particle flux profile is also changed.

12. 14 Oct. 2009, S. Masuzaki 12/18 Application of the 3D edge transport codes : EMC3-EIRENE Physics  standard fluid equations of density, momentum, energy of ion & electron  simplified fluid model for impurities (not for present analysis)  kinetic model for neutral gas Geometry  fully 3D for plasma, divertor plates, baffles and wall  ergodic or non-ergodic B-field configurations Numerics  Monte Carlo technique on local field-aligned vectors, piecewise parallel integration for isolation of the small  from the large II-transport (  /II~10 -8 )  new Reversible Field Line Mapping (RFLM) technique, Finite flux tube coordinates for B-field line interpolation Coupled self-consistently

13. 14 Oct. 2009, S. Masuzaki 13/18 Reproduce of the profile change using EMC3-EIRENE code Blue profile in experiment was reproduced by calculation. Red profile was not well reproduced. But increasing of diffusion coefficient flatten the profile in calculation. Heat load decreases with increase diffusion coefficient. experiment calculation

14. 14 Oct. 2009, S. Masuzaki 14/18 At 10.5U probe position Heat load (and particle load) increases with increase of diffusion coefficient at this position. Heat and particle transfer from “long” flux tube to laminar region is enhanced. calculation The ratio of total I sat at 10.5U probe to that at 6I probe increase during the profile change (yellow hatched). Consistent with calculation assuming larger D and . experiment

15. 14 Oct. 2009, S. Masuzaki 15/18 In the case of Rax=3.75m Profiles of heat and particle flux are not largely changed by plasma conditions. In experiment, ratio of peak heat flux on the divertor to heating power is large for relatively low Te discharges (blue line). In calculation, the ratio increase with decreasing of D and . Divertor heat flux vs. heating power N e and T e at LCFS are shown for each symbol Exp. Cal.

16. 14 Oct. 2009, S. Masuzaki 16/18 Less collision looks to enhance the heat transfer from long flux tube to laminar region Normalized heat flux vs. collision mean free path (10 16 T e 2 /n e ) around the last closed flux surface. Decreasing of the normalized heat flux suggests that heat transfer from “long” flux tube to laminar region is enhanced.

17. 14 Oct. 2009, S. Masuzaki 17/18 Do diffusion coefficients increase with increasing of T e ? TeTe I sat @6I I sat @10.5U TeTe TeTe flattening of heat flux profile@6I P div @6I Heating power P div @6I Heating power Calculations with increasing D and  Observations in discharges with Rax=3.75m Observations in discharges with Rax=3.60m consistent

18. 14 Oct. 2009, S. Masuzaki 18/18 Summary Profiles of heat and particle load on divertor plates are roughly determined by magnetic field line structure. –Though the edge field line structure is complex, the profiles can be roughly predicted even in the complex e. Heat and particle transfer from “long” flux tube to laminar region look to be enhanced with T e rise rather than increasing collisionality.

19. 14 Oct. 2009, S. Masuzaki 19/18 Closed Helical Divertor will be installed (2/10 sections) in 2010 Present divertor Closed divertor

20. 14 Oct. 2009, S. Masuzaki 20/18

21. 14 Oct. 2009, S. Masuzaki 21/18 Edge T e profiles D and  were estimated by fitting of calculation to experimental data T e,LCFS ~200eV n e,LCFS ~1-1.5E19m -3 D = 0.125 - 0.25 m 2 /s  = 0.375 - 0.5 m 2 /s T e,LCFS ~400eV n e,LCFS ~1-1.5E19m -3 D = 0.5 – 1.0 m 2 /s  = 1.5 – 3 m 2 /s T e,LCFS ~200eV n e,LCFS ~6E19m -3 D = 0.25 - 0.5 m2/s  = 0.75 – 1.5 m2/s These results suggest that the cross-field transport coefficients have temperature dependence, and they also depend on density weakly.