1. Carine Giroud 1 ITPA Naka 1.10.2007 Impurity transport at JET On-going analysis from recent campaign C. Giroud, C. Angioni, L. Carraro, P. Belo, I. Coffey, V. Naulin, M-E. Puiatti, M. Brix, H. Leggate, K. Lawson, A.D. Whiteford, M. Valisa.

2. Carine Giroud 2 ITPA Naka 1.10.2007 Linear relationship assumed between impurity flux and density gradient Experimental determination of transport coefficients Diffusion coefficient Convection coefficient Vz >0 outwards Extrinsic impurities injected by laser ablation (Ni) or gas injection (Ne, Ar). − D and V determined individually by modelling of time evolution of spectroscopic data − Ni: soft x-ray and VUV − Ne and Ar: soft x-ray and VUV and also from charge exchange spectroscopy

3. Carine Giroud 3 ITPA Naka 1.10.2007 NCLASS: neoclassical coefficients calculated from NCLASS Time trace

4. Carine Giroud 4 ITPA Naka 1.10.2007 Discharges analysed so far A wide dataset of He, Ne, Ar, Ni has been collected: He still on-going analysis. many discharges cannot be analysed due to MHD, changing plasma conditions

5. Carine Giroud 5 ITPA Naka 1.10.2007 Global density peaking used in this talk: – determined Vz and Dz – assume a constant source  nz(r) – definition global impurity density peaking: – also calculated corresponding peaking: with Dmeas, V=Vneo – for C, direct measurement of nz(r) Volume averaged Measured peaking of impurity

6. Carine Giroud 6 ITPA Naka 1.10.2007 – Z dependence – For Argon, assumption V=Vneo with D=Dmeas does not reproduce peaking. – Variation in Argon peaking  More discharges to be analysed. Impurity peaking Discharges without RF

7. Carine Giroud 7 ITPA Naka 1.10.2007 Diffusion coefficients r/a~0.15 r/a~0.6

8. Carine Giroud 8 ITPA Naka 1.10.2007 Transport coefficients: V Z V_ar measured different from neoclassical. r/a~0.15 r/a~0.6 C Ne Ar Ni

9. Carine Giroud 9 ITPA Naka 1.10.2007 Comparison L-mode and H-mode

10. Carine Giroud 10 ITPA Naka 1.10.2007 Comparison H-mode and Hybrid Ip/B Ptot Wdia ne Halpha Peaking varies for Argon: MHD ? Hybrid: black H-mode color (-15.5s)

11. Carine Giroud 11 ITPA Naka 1.10.2007 – Rf effect confirmed but local effect r/a~0.3 – Seem to be dependent on Z ! – Compatible with a neoclassical effect Effect of ICRH preliminary Carraro EPS 2007 Giroud EPS 2007 Belo

12. Carine Giroud 12 ITPA Naka 1.10.2007 Conclusion Work on-going on the analysis of impurity transport experiments from 2006-2007 database. Validation of consistent analysis technique and error analysis. First results: L-mode/H-mode : similar peaking Hybrid/H-mode: same peaking from C to Ar. ICRH effect: core effect and maybe due to neoclassical effect. Ni transport analysis to be extended and He transport analysis started. Beta scaling and eff scaling Systematic comparison with theoretical models.

13. Carine Giroud 13 ITPA Naka 1.10.2007 Two very different Ni profiles ICRH dominant ion Peaked Ni profile ICRH dominant electron Slightly hollow Ni profile [M-E. Puiatti PoP 13 2006] Steady-state profile

14. Carine Giroud 14 ITPA Naka 1.10.2007 Recent development in turbulent transport theory Two main electrostatic micro-instability considered ITG/TEM MicroinstabilityITGTEM Direction of propagation Ion diamagnetic Electron diamagnetic driveR/LTiR/LTe & R/Ln

15. Carine Giroud 15 ITPA Naka 1.10.2007 Recent development in turbulent transport theory Three main mechanisms have been identified Curvature pinch 1 Compressibility of ExB drift velocity Thermodiffusion pinch 2 Compression of the diamagnetic drift velocity Pinch connected to the parallel dynamics of the impurity 3 Parallel compression of parallel velocity fluctuations produced along the field line by the fluctuating electrostatic potential Decreases with increasing q 2 [X. Garbet PoP 12 2005] 2,3 [C. Angioni C PRL. 96 2006] 1 [J. Weiland NF 29 1989] 1 [X. Garbet PRL 91 2003] 1 [M. B. Isichenko PRL 1996] 1 [D.R. Baker PoP 5 1998] 1 [V. Naulin Phys Rev. E 2005] 2 [M. Frojdh NF 32 1992]

16. Carine Giroud 16 ITPA Naka 1.10.2007 Pinch mechanisms in theory of turbulent impurity transport All contribute to the total turbulent pinch

17. Carine Giroud 17 ITPA Naka 1.10.2007 Illustration of complex Z dependence of turbulent transport D and V calculated with the linear version of the gyrokinetic code GS2: only the fastest growing mode is taken in the quasi – linear model, no neoclassical transport included. Trace impurity considered. No general trend in Z of turbulent transport  specific calculation needed for studied discharge GS2 [R/LTi=7, R/LTe=6, Te/Ti=0.88] [C. Angioni]

18. Carine Giroud 18 ITPA Naka 1.10.2007 ICRH dominant ion Peaked Ni profile ICRH dominant electron Slightly hollow Ni profile [M-E. Puiatti PoP 13 2006] Steady-state profile R/Ln=3.9 R/LTe=4 Te/Ti=0.95 R/LTi=6.6 R/Ln=5.2 R/LTe=6. Te/Ti=1. R/LTi=6.6 Different dominant instability for peaked and flat Ni density ITG dominatedTEM dominated GS2

19. Carine Giroud 19 ITPA Naka 1.10.2007 Measured Z dependence of impurity peaking #66134 Neoclassic measure- ment measure- ment Neoclassic r/a =0.15 r/a =0.55 Negative C peaking Peaking lower than neoclassical stronger z dependence in core than at mid-radius Ne, Ar and Ni injected in ELMy H-mode q0>1, 0.1 <neff <0.2 Bt=2.9T, q95=7, 2MW ICRH, 8.6MW NBI

20. Carine Giroud 20 ITPA Naka 1.10.2007 GS2 Anomalous part: -R(V-Vneo)/(D-Dneo) GS2 w/o Thermodiffusion Linear GK calculation reproduces trend of measured Z dependence Discharge ITG dominated R/LTi~5.8, R/LTe~6.3, R/Ln~0.3 and Te/Ti~1.1, *~0.10 GS2 Measurement

21. Carine Giroud 21 ITPA Naka 1.10.2007 Discharges analysed so far A wide dataset of He, Ne, Ar, Ni has been collected: He still on-going analysis. many discharges cannot be analysed due to MHD, changing plasma conditions Ti/Te