Advanced Density Profile Reflectometry; the State-of-the-Art and Measurement Prospects for ITER

Advanced Density Profile Reflectometry; the State-of-the-Art and Measurement Prospects for ITER by E.J. Doyle With W.A. Peebles, L. Zeng, P.-A. Gourdain, T.L. Rhodes, S. Kubota and G. Wang Dept. of Electrical Engineering and PSTI University of California Los Angeles Los Angeles, California 90095 Presented at the 48th Annual Meeting of the Division of Plasma Physics Philadelphia, Pennsylvania October 30 through November 3, 2006 DIII D NATIONAL FUSION FACILITY SAN DIEGO EJD/APS06

Outline Introduction - why use reflectometry, basic principles Technology has transformed ability to make high quality density profile measurements via reflectometry High performance solid-state sources and frequency multipliers Improved data acquisition and data analysis capabilities New set of measured density profiles from DIII-D and NSTX, illustrating current performance levels Achieved spatial and temporal resolution are in-line with ITER targets Challenges and issues for basic feasibility of reflectometer density profile measurements on ITER e.g. cyclotron absorption, cutoff downshift/flattening, refraction Core measurements on ITER may be more feasible than previously thought Summary 11/14/06 2

Why profile reflectometry? - High resolution measurements, and reactor compatible For physics studies on current devices, reflectometry offers unique combination of high temporal and spatial resolution Systems on DIII-D, NSTX, AUG, TS, etc. DIII-D and NSTX results presented here Second major driver is desire to use microwave diagnostics on ITER, where optical diagnostics face serious challenges Reflectometry is reactor compatible in terms of: Tolerance of mirror imperfections Radiation resistance Mechanical robustness Modest access requirements US is currently responsible for the main (low field side) profile reflectometer on ITER ITER 11/14/06 3

Profile reflectometry is a radar-like technique to measure the electron density profile Measure time/phase delay to plasma cutoff layer, using radar techniques Two cutoffs available: O-mode (EIIB), f pe X-mode (E B), f r Two requirements for profile measurements: Probing frequency must cover density profile to be measured Cutoff frequency profile must have finite gradient For fusion plasmas, cutoff frequencies typically lie in range from 10-200 GHz Frequency (GHz) Density (1019 m-3) 100 80 60 40 20 0 6 5 4 3 2 1 0 f pe (O-mode) 1.8 1.9 2 2.1 2.2 2.3 R (m) 11/14/06 4 f ce f r (X-mode) Cutoff frequencies Density Profile Detection and analysis

Solid-state microwave sources now provide fast, full-band frequency sweep capability on DIII-D Use continuous broadband frequency swept techniques (FM-CW radar) Solid-state sources provide fast, full-band, high power capability Fast-tuning Active x4 Directional Launch Oscillator Coupler 10dB Plasma Varactor Driving Circuit 8-12.5 GHz 12.5-18 GHz Solid-state 50-72 GHz source, fullband sweep time ~10 µs, with 16-18 dbm, 45-65 mw power Freq. Multiplier x 4 33-50 GHz 50-72 GHz Arbitrary Function Generator Power Amplifier WG Delay Line Mixer Frequency Doubler Receive IF signal, frequency time delay difference Active Frequency Doubler 12.5-18 GHz Tunable Source 11/14/06 5

Solid-state sources exist for higher frequency measurements, as required for ITER Tore Supra and JET currently have solidstate profile systems to 155/160 GHz Maximum frequency required on ITER may be ~200 GHz (less than previously thought) Solid-state sources with >30 mw power currently exist to 200 GHz Virginia Diode Output Power [mw] 40 35 30 25 20 15 JPL 30 mw 10 10% output bandwidth 5 175 180 185 190 195 200 205 210 215 Frequency (GHz) 11/14/06 6

Automatic, between-shot profile analysis enabled by multi-processor cluster on DIII-D Flexible digital signal processing similar to conventional radar, but with addition of profile inversion and other steps required for plasmas Digital complex demodulation used to extract phase/time delay Range resolution filtering and phase delay smoothing Profile inversion from phase data 10 processors (photo) provide between-shot profile analysis Analysis of profiles measurements every 5 ms throughout a 5 s discharge takes ~4-5 min. Rapid growth in this technology, which can be expected to continue over next decade 11/14/06 7

Outline Introduction - why use reflectometry, basic principles Technology has transformed ability to make high quality density profile measurements via reflectometry High performance solid-state sources and frequency multipliers Improved data acquisition and data analysis capabilities New set of measured density profiles from DIII-D and NSTX, illustrating current performance levels Achieved spatial and temporal resolution are in-line with ITER targets Challenges and issues for basic feasibility of reflectometer density profile measurements on ITER e.g. cyclotron absorption, cutoff downshift/flattening, refraction Core measurements on ITER may be more feasible than previously thought Summary 11/14/06 8

Example of fully-automatic DIII-D profile analysis with 1 ms time resolution, for duration of >3 s Example of ECH heated plasma with transient L-H transitions at NBI blips R (m) (a.u.) 2.0 2.1 2.2 2.3 4 3 126857 L->H H->L Density profile contours (x10 19 m -3 ) D α 2 ECH 1 NBI 0 1.0 2.0 3.0 4.0 5.0 Time (s) L->H H->L ne=2.7 ne=2.4 ne=2.2 ne=1.9 ne=1.6 ne=0.8 ne=0.03 3.0 2.5 2.0 1.5 1.0 Reflectometer profile Thomson Scattering n e (10 19 m -3 ) 2.512 s 126857 0.5 Separatrix ρ=0.6 radius 0.0 2.05 2.10 2.15 2.20 2.25 2.30 R (m) 11/14/06 9

Very high time resolution is available, e.g. 25 µs (40 times previous example) Profiles with 25 µs time resolution are available continuously for up to 2.5 s, limited only by data acquisition system 10 µs time resolution demonstrated Standard DIII-D TS repetition rate is 12.5 ms Example of profile data across L-H transition (a.u.) R (m) 3 2 1 2.0 2.1 2.2 2.3 2.0 2.1 D α ne= 4.0x10 19 ne= 3.5x10 19 ne= 3.0x10 19 ne= 1.0x10 19 Separatrix L-H transition ne= 4.5x10 19 ne= 0.2x10 19 ne=0.05x10 19 118865 Reflectometer data Thomson data ρ = 0.4 ρ = 0.8 R (m) 2.2 2.3 1680 1700 1720 1740 1760 Time (ms) 11/14/06 10

Excellent agreement found between reflectometer and other diagnostics in edge Excellent agreement between three independent measurements Reflectometry, Thomson scattering (TS), reciprocating Langmuir probe measurements 120350-4150ms (EFIT03) 10 19 Electron density (m -3 ) 10 18 10 17 Separatrix Thomson Scattering Langmuir Probe Reflectometer -4.0-2.0 0. 2.0 4.0 6.0 8.0 10. R-R sep (cm) 11/14/06 11

Reflectometer measurements have high precision Direct comparison of reflectometer and Thomson data over same spatial and temporal range Reflectometer with 5 ms and TS with 12.5 ms repetition rate Less jitter on reflectometer data Reflectometer data are from automatic between-shot analysis 3 2 1 0 3 0 4 3 2 1 Reflectometer data (x10 19 m -3 ) 123175 Thomson scattering data (x10 19 m -3 ) Line average density (x10 19 m -2 ) R=2.17 m R=2.192 m R=2.223 m R=2.25 m R=2.27 m R=2.17 m R=2.192 m R=2.223 m R=2.25 m R=2.27 m CD 4 puff 2 1.0 2.0 3.0 4.0 5.0 Time (s) 11/14/06 12

Core density profile modulation by sawteeth directly observed on DIII-D Te (kev) R (m) 2.6 2.4 2.2 2.0 126599 1.7 Density contours (x10 19 m -3 ) 1.8 Core 1.9 2.0 Core ECE Sawteeth start Magnetic axis Inversion radius 5.9 Density Steep, ITB-like profiles in L-mode plasma relax when sawteeth start 5 ms time resolution Also observed on Tore Supra, Sabot, EPS 2006 2.1 2.2 2.3 0.0 Separatrix Edge 1750 1800 1850 1900 1950 Time (ms) 11/14/06 13

Core profile modulation by fast ion driven coherent modes directly observed on NSTX ~10 khz energetic particle driven mode, with larger profile modulation in core than in edge 25 µs time resolution Potential to provide mode structure and displacement measurements Density Contours Electron Density (x10 19 m -3 ) 3.0 2.0 1.0 Reflectometry MPTS 0 20 40 60 80 100 120 140 160 Major Radius (cm) Magnetic probe

DIII-D and NSTX data provide proof-of-principle demonstration of ITER performance targets Current DIII-D performance levels over density range of 0-6.4x10 19 m -3 are: Spatial resolution: ~0.4 cm at edge (twice rms jitter) (ITER target 0.5 cm) ~2 cm in core (ITER target a/30~6.7 cm) Time resolution: 25 µs typical, 10 µs demonstrated (ITER target is 10 ms for core and edge) Has proved capable of resolving edge H-mode pedestal, evolution through ELMs, ITBs, etc. Measurements are important for ITER, which seeks proof-ofprinciple demonstration of target performance levels on current devices 11/14/06 15

OUTLINE Introduction - why use reflectometry, basic principles Technology has transformed ability to make high quality density profile measurements via reflectometry High performance solid-state sources and frequency multipliers Improved data acquisition and data analysis capabilities New set of measured density profiles from DIII-D and NSTX, illustrating current performance levels Achieved spatial and temporal resolution are in-line with ITER targets Challenges and issues for basic feasibility of reflectometer density profile measurements on ITER e.g. cyclotron absorption, cutoff downshift/flattening, refraction Core measurements on ITER may be more feasible than previously thought Summary 11/14/06 16

Reflectometry is intended to provide multiple plasma control and physics capabilities on ITER Reflectometry on ITER will address a wide range of needs (Vayakis et al, NF 2006), using multiple systems, e.g. Density profile Plasma position control Plasma rotation H-mode pedestal physics MHD and EPMs Main profile reflectometer design calls for core and edge profile measurements with: Density range of 0-3x10 20 m -3 >2.5 times Greenwald density (Propose to reduce to ~0-1.8x10 20 m -3 ) O-mode measurements from 15-150 GHz X-mode from 75-250 GHz (reduce to 200 GHz) 6 waveguide/antenna pairs Position control and HFS reflectometers Main profile reflectometer Possible divertor system 11/14/06 17

ITER presents new challenges and issues for feasibility of reflectometer profile measurements In plasma core, measurement access to cutoffs is affected by multiple relativistic effects: Downshift and flattening of cutoffs Absorption at downshifted and broadened cyclotron layers 3-D refraction affects Conclusions from this study: In the pedestal, relativistic effects are smaller than in core: Pedestal measurements should be similar to measurements on current tokamaks In general edge measurements easier to make than core, e.g. lower frequencies, etc. Refraction and time/phase delay variation due to cutoff curvature may be greatest limits on core measurements Antenna arrangement needs to account for refraction - propose linear array configuration Propose to reduce target measurement range to ~0-1.8x10 20 m -3 11/14/06 18

Access to cutoffs on ITER is affected by multiple relativistic effects Calculations for ITER base case of ELMy H-mode operation (Scenario 2) Relativistic downshift/flat density profile makes cutoff profiles flat to hollow Cyclotron layers are also downshifted, leading to potential absorption problems Te (kev) Frequency (GHz) 20 15 10 5 250 200 150 100 f pe_rel f pe T e fce_rel 2f ce_rel f r_rel n e 2f ce f r f ce 20 15 10 5 ne (10 19 m -3 ) 3-4 Log(α) m -1 50 6.5 7.0 7.5 8.0 R (m) 11/14/06 19

However, density peaking currently predicted for ITER will allow core measurements Latest predictions for ITER (Weisen, IAEA 2006) based on AUG/JET results predict n e (0)/<n e > 1.35 for ITER Even modest density peaking makes both O- and X-mode measurements viable Creates gradient in cutoff frequency Example for modified ITER reference case, keeping constant pressure and line density ne (10 19 m -3 ) Frequency (GHz) 12 8 4 0 180 160 140 90 70 50 Original flat density profile Peaked density profile, n e (0.2)/<n e >=1.35 X-mode cutoff frequencies (relativistic) Magnetic axis O-mode cutoff frequencies (relativistic) 6.5 7.0 7.5 8.0 R (m) 11/14/06 20

Cyclotron absorption may not be significant limitation on ITER, except close to plasma center Analytic approximation for cyclotron absorption (Batchelor, 1984) benchmarked to GENRAY relativistic calculation (for X-mode) ITER reference case plasma Cyclotron absorption double pass loss only becomes significant close to plasma center Power (a.u.) 1.0 0.8 0.6 0.4 176 GHz Cutoff 172 GHz Cutoff 0.2 Magnetic 190 GHz Axis ρ =1 0.0 6.0 6.5 7.0 7.5 8.0 R (m) Genray Analytic code Magnetic Axis 0 6.0 6.5 7.0 7.5 8.0 11/14/06 21 Power Attenuation (db) 30 20 10 ITER Reference case, flat density ITER "Reference case," peaked density ITER Steady-state scenario R (m) ρ =1

3-D relativistic ray tracing for ITER using GENRAY (Harvey, CompX) shows large refraction on ITER Substantial toroidal and poloidal offsets, even from edge 165 GHz Vertical X-mode launch Toroidal Launch antenna Poloidal flux surfaces Vertical direction Radial direction Toroidal direction Launch antenna Equatorial plane 166 GHz X-mode launch Toroidal Vertical Return beam (1/e 2 radius) Launch antenna Return beam (1/e 2 radius) 11/14/06 22

Tilted poloidal antenna array required to cope with refraction offsets for various shapes/positions X-mode refraction shown (substantially larger for O-mode) Solution is to rearrange ITER antennas in linear poloidal array Also advocated by Kramer et al., NF 2006, from 2-D fullwave calculations Refraction and variable time/phase delay from cutoff curvature may be greatest limits on core measurements Return offset from launch antenna (cm) 30 20 10 0-10 -20-30 Return offset from launch antenna for different vertical plama offsets 171 GHz X-mode - Non-relativistic 167 GHz X-mode - Relativistic -30-20 -10 0 10 20 30 Antenna/plasma-midplane vertical separation (cm) 11/14/06 23 167 GHz X-mode - Non-relativistic

Reflectometry can determine T e and n e profiles on ITER if density is measured using two different cutoffs Need to know T e to determine n e profile on ITER. However, due to different T e sensitivities, can determine T e and n e if profile measurements made with two cutoffs. Perfect reconstruction possible with ideal data. Step 3: Te profile from X-mode inversion is now used as input for O-mode density profile and process is iterated ne (10 19 m -3 ) T e (kev) 12 10 8 6 4 2 0 25 20 15 10 1st profile using Te=0 O-mode profile Actual ne profile Actual Te profile Step 1: assume Te=0 and invert O-mode phase to get density profile Step 2: X-mode phase - now a function of ne and Te- is inverted to extract Te profile using O-mode density profile as input 5 Te from X-mode phase after 1st iteration 11/14/06 0 6.5 7.0 7.5 R (m) 8.0 24

0 Edge pedestal profiles can be measured by reflectometry on ITER Edge measurements on ITER will not have problem with cutoff flattening or absorption T e and refraction corrections will be needed Pedestal measurements will be critical on ITER ELMs, wall interaction, performance, stability, H-mode operation R ( m ) Dα ( a.u. ) 1.90 2.00 2.10 2.20 n e =4.5 ELM 2.30 2 Example of pedestal 1 measurements on DIII-D, with 2.40 n e =0.1 Separatrix 25 µs time resolution through Precursor three ELM cycles with high 2.50 spatial resolution 3120 3140 3160 3180 Time (ms) 11/14/06 25 118219 34 8 7 6 5 8 7 6 5 4 3 21 4.5 ne (1019 m-3)

Summary Technological advances have transformed performance of reflectometer systems for measurement of electron density profiles Broadband, solid-state fast-sweep mm-wave sources Improved data acquisition/analysis with use of parallel processing for between-shot profile analysis Spatial and temporal resolution of current DIII-D measurements provides proof-of-principle demonstration of ITER target values Increases confidence in achieving targets on ITER Edge pedestal measurements should be possible on ITER. Core measurements should also be possible if density profile is peaked Refraction and time/phase delay variation due to cutoff curvature may be greatest limits on core measurements Propose linear array configuration to account for refraction Other issues need further study, e.g. turbulence effects, waveguides, etc. 11/14/06 26