1. SOFT 2006 Warsaw 1 M. Kaufmann Supported by H. Bolt, R. Dux, A. Kallenbach and R. Neu Tungsten as First Wall Material in Fusion Devices

2. SOFT 2006 Warsaw 2 Tungsten as First Wall Material in Fusion Devices 1.Introduction 2.Plasma Wall Interaction with Tungsten 3.Edge and Core Transport 4.Technological Developments 5.Summary

3. SOFT 2006 Warsaw 3 Introduction: PLT with tungsten limiter (1975) Consequence of accumulation and central radiation! Since then most tokamaks and stellarators have used graphite as first wall material. V. Arunasalam et al., Proc. 8th Conf. EPS, Prague 1977

4. SOFT 2006 Warsaw 4 Tokamaks with High-Z-surfaces Limiter tokamaks: FTU (ENEA) Textor (FZJ) Divertor tokamaks: Alcator C-Mod (MIT) ASDEX Upgrade (IPP) future: JET ITER M.L. Apicella et al., Nucl. Fusion 37 A. Pospieszczyk et al., J. Nucl. Mater. 290-293 B. Lipschultz et al., Nucl. Fusion 41 R. Neu et al., Plasma Phys. Control. Fusion 38 J. Pamela, this conference G. Janeschitz,J. Nucl. Mat. 290-293

5. SOFT 2006 Warsaw 5 FTU (ENEA Frascati) until 1994: poloidal limiter (steel, TZM, W) now: toroidal limiter TZM M.L. Apicella, et al., J. Nucl. Mater. 313-316

6. SOFT 2006 Warsaw 6 Alcator C-Mod (MIT) Divertor configuration with a complete set of Mo-tiles B. Lipschultz et al., Phys. Plasmas 13

7. SOFT 2006 Warsaw 7 ASDEX Upgrade (IPP Garching) Stepwise approach: remaining parts will be covered with tungsten in the 2007 campaign! R. Neu et al., Nucl. Fusion 45

8. SOFT 2006 Warsaw 8 Graphite versus Tungsten positive negative graphite: low central radiation high erosion radiation in boundary tritium co-deposition forgives overload destruction by neutrons tungsten: low erosion high central radiation low tritium co-deposition accumulation in centre resistant to neutrons critical with overload radioactive, however, short decay time

9. SOFT 2006 Warsaw 9 Graphite versus Tungsten tungsten: low erosion high central radiation low tritium co-deposition accumulation in centre resistant to neutrons critical with overload Test in linear machines of limited relevance!

10. SOFT 2006 Warsaw 10 Graphite versus Tungsten positive negative graphite: low central radiation high erosion radiation in boundary tritium co-deposition forgives overload destruction by neutrons tungsten: low erosion high central radiation no tritium co-deposition accumulation in centre resistant to neutrons critical with overload JET/ITER-generation

11. SOFT 2006 Warsaw 11 Graphite versus Tungsten positive negative graphite: low central radiation high erosion radiation in boundary tritium co-deposition forgives overload destruction by neutrons tungsten: low erosion high central radiation no tritium co-deposition accumulation in centre resistant to neutrons critical with overload DEMO-generation

12. SOFT 2006 Warsaw 12 Tungsten: Erosion versus Radiation W-erosion much lower than graphite! (R.T) C chem (800K)

13. SOFT 2006 Warsaw 13 Tungsten: Erosion versus Radiation But central W-radiation much higher! LZLZ

14. SOFT 2006 Warsaw 14 Ignition Condition: Tungsten vs. Carbon

15. SOFT 2006 Warsaw 15 Gain Experience: Diagnostic W-lines at low temperature to determine influx (Textor) W-lines at high temperature to determine core concentration (AUG) graphite: extensive experience tungsten: limited experience A. Thoma et al., Plasma Phys. Control. Fusion 39 A. Pospieszczyk et al., to be published

16. SOFT 2006 Warsaw 16 Tungsten as First Wall Material in Fusion Devices 1.Introduction 2.Plasma Wall Interaction with Tungsten 3.Edge and Core Transport 4.Technological Developments 5.Summary

17. SOFT 2006 Warsaw 17 Plasma Wall Interaction Low erosion + no formation like hydro-carbons  low hydrogen retention (0.1 …1% instead of 40…100%) W: high mass, low velocity of eroded particles  ionization length << gyro radius  90% prompt redeposition W C D. Naujoks et al., Nucl. Fusion 36 R. Causey, J. Nucl. Mater. 300 J. Roth, M. Mayer, J. Nucl. Mater. 313-316

18. SOFT 2006 Warsaw 18 Erosion on Target Plates/Limiter V. Philipps et al., PPCF 42

19. SOFT 2006 Warsaw 19 Typical ITER reference H-mode pressure profile forms steep edge pedestal: Sources for W-Erosion: ELMs ELMs produce main chamber erosion and target plate erosion. In both cases sputtering by low Z-components dominant. Pedestal breaks down during ELMs! n r A. Herrmann et al., accepted for publ. in J. Nucl. Mater

20. SOFT 2006 Warsaw 20 Sources for W-Erosion: NBI Fast particles losses from neutral beam injection can be identified as a tungsten source on limiters. Increase during ELMs. 3+8 Quantitative agreement with calculations  Extrapolation to ITER: no problem! R.Dux, to be published R. Dux et al., accepted for publ. in J. Nucl. Mater

21. SOFT 2006 Warsaw 21 Sources for W-Erosion: ICRH Localized boronization by ECRH helps to identify zone of Mo-erosion. Alcator C-Mod: In ICRH heated plasmas without boronization: high radiation by molybdenum. Strongly reduced by boronization. However, effect lasts only for 10s total pulse duration. B. Lipschultz et al., Phys. Plasmas 13

22. SOFT 2006 Warsaw 22 Sources for W-Erosion: ICRH - small zone on top of divertor responsible for Mo-erosion. - field lines map back to antenna. - sheath potential 100-400eV Conclusions:

23. SOFT 2006 Warsaw 23 Sources for W-Erosion: ICRH Can one reduce the sheath potential? Lots of open questions! Faraday screen parallel to field lines: small effect Is tungsten ITER/reactor compatible? ICRH reactor compatible? Vl.V. Bobkov et al., accepted for publ. in J. Nucl.

24. SOFT 2006 Warsaw 24 Replacement of Carbon as Radiator Carbon radiates in the plasma boundary. It reduces therefore the load to the target plates considerably. It is highly self-regulating! Replacement by a noble gas such as Argon or Neon seems necessary: Robust feed back method is needed! controlled argon seeding Control by thermo currents through divertor plates: A. Kallenbach et al., J. Nucl. Mater. 337-339

25. SOFT 2006 Warsaw 25 Tungsten as First Wall Material in Fusion Devices 1.Introduction 2.Plasma Wall Interaction with Tungsten 3.Edge and Core Transport 4.Technological Developments 5.Summary

26. SOFT 2006 Warsaw 26 Neoclassical Transport Neoclassical transport by Coulomb collisions including drift motion leads to two fluxes. Strong peaking of tungsten concentration in case of peaked density profiles ( small) is expected. diffusion: inward drift:

27. SOFT 2006 Warsaw 27 Transport in the H-Mode Pedestal Steep density profile  strong inward drift! n r ELMs wash tungsten out! High ELM frequency is required anyhow to reduce load to target plates! P. Lang et al., Nucl. Fusion 45

28. SOFT 2006 Warsaw 28 Influence of Anomalous Transport A peaked density profile without strong anomalous transport leads to strong tungsten accumulation. Central heating overcompensates neoclassical inward drift by anomalous transport! A. Kallenbach et al., Plasma Phys. Control. Fusion 47

29. SOFT 2006 Warsaw 29 Influence of Anomalous Transport Anomalous transport induced by central heating can easily overcompensate neoclassical inward drift : Recent theoretical work: no turbulent transport mechanisms for strong high Z-ions inward drift! In summary, one expects with a high probability no peaked W concentration profiles in a burning device! C. Angioni, A.G. Peeters, Phys. Rev. Let. 96

30. SOFT 2006 Warsaw 30 W-concentration W-concentration strongly depending on discharge conditions! Erosion and transport determine concentration. AUG

31. SOFT 2006 Warsaw 31 Tungsten as First Wall Material in Fusion Devices 1.Introduction 2.Plasma Wall Interaction with Tungsten 3.Edge and Core Transport 4.Technological Developments Tungsten Coatings Massive Tungsten 5.Summary

32. SOFT 2006 Warsaw 32 W-Coatings on Graphite In present day devices with low particle fluencies W-coating on graphite is used - because of lower eddy and halo currents. - because of lower weight. Different techniques are available, e.g.: - physical vapor deposition (PVD) - chemical vapor deposition (CVD) - plasma spray (PS) H. Maier et al., accepted for publ. in J. Nucl. Mater.

33. SOFT 2006 Warsaw 33 W-Coatings on Graphite: JET In JET the ‘ITER like wall project’ is under preparation. The first wall will be partly covered with tungsten. green: Be red: W-Coating blue: massive W (probably) highly loaded areas: 200µ sheath by PS others: PVD Highly loaded areas can be later replaced by uncoated graphite!

34. SOFT 2006 Warsaw 34 Massive W-Structures: JET High particle fluencies (ITER, DEMO): massive W-structures are necessary. They are ‘castellated’ - because of eddy currents (JET) - because of different thermal expansion (ITER, DEMO). FZJ

35. SOFT 2006 Warsaw 35 The ITER reference design test at FZJ

36. SOFT 2006 Warsaw 36 DEMO positive negative graphite: low central radiation high erosion radiation in boundary tritium co-deposition forgives overload destruction by neutrons tungsten: low erosion high central radiation no tritium co-deposition accumulation in centre resistant to neutrons critical with overload DEMO-generation Is ITER DEMO-relevant? Can the first wall be exchanged?

37. SOFT 2006 Warsaw 37 Developments for DEMO Ductile to brittle transition temperature (DBTT) high. Problem e.g. in W-steel-connections He-cooled divertor (FZK): Nuclear loads increase DBTT. Development of W-alloys can reduce that problem.

38. SOFT 2006 Warsaw 38 Developments for DEMO Surfaces with reduced load: A few mm tungsten sheets on EUROFER by PS or CVD IPP, Petten. FZJ

39. SOFT 2006 Warsaw 39 DEMO: Safety Issues SEIF Study, EFDA-S-RF-1, April 2001 Loss of coolant and intense air ingress: formation of radioactive WO 3 - compounds with high evaporation rate which can leave hot vessel. Tungsten : WSi 0.82 Cr 0.45 : Oxidation rate (mg cm -2 s -1 ) WSi 0.82 : F. Koch, H. Bolt, subm. to Physica Scripta

40. SOFT 2006 Warsaw 40 Summary In a fusion reactor, low-Z as a first wall material (graphite, Be) will have to be replaced by tungsten. So far, plasma experiments have demonstrated that in most scenarios the tungsten erosion of the surfaces and its concentration in the central plasma can be kept sufficiently low. In certain scenarios with high edge temperatures this may, however, not be the case. In addition, the high erosion in the neighbourhood of an ICRH antenna needs particular attention. As an intermediate solution, the coating of graphite with tungsten is an available technology. Technological solutions for the highly loaded divertor targets in a fusion reactor are under development. The relatively high ductile to brittle transition temperature, however, poses specific problems.

41. SOFT 2006 Warsaw 41 Summary Altogether tungsten as the first wall material looks promising. However, several open questions still remain to be solved.

42. SOFT 2006 Warsaw 42 Reserve

43. SOFT 2006 Warsaw 43 Sources for W-Erosion: ELMs Erosion on target plates:

44. SOFT 2006 Warsaw 44 Sources for W-Erosion: ICRH ASDEX Upgrade: Localized measurement on ICRH-antenna Fast (< 1ms) and localized increase  increase due to sheath rectified E-fields

45. SOFT 2006 Warsaw 45 Transport in the H-Mode Pedestal Argon seeding has to be well controlled!

46. SOFT 2006 Warsaw 46 Tungsten has 200 times larger conductivity than graphite, therefore eddy and halo currents larger. Tungsten has 8.5 times larger mass density than graphite. In case of low particle fluencies often W-coating on graphite are used. Different techniques are available, e.g.: - physical vapor deposition (PVD) - chemical vapor deposition (CVD) - plasma spray (PS)

47. SOFT 2006 Warsaw 47 Plasma Wall Interaction Blistering:

48. SOFT 2006 Warsaw 48