Current status of FIReTIP and its applications

1. NSTX Supported by Current status of FIReTIP and its applications College W&M Colorado Sch Mines Columbia U CompX General Atomics INEL Johns Hopkins U LANL LLNL Lodestar MIT Nova Photonics New York U Old Dominion U ORNL PPPL PSI Princeton U Purdue U SNL Think Tank, Inc. UC Davis UC Irvine UCLA UCSD U Colorado U Illinois U Maryland U Rochester U Washington U Wisconsin June-Woo Juhn (Seoul National U, Korea), K.C. Lee (UC Davis), R. Kaita (PPPL) Culham Sci Ctr U St. Andrews York U Chubu U Fukui U Hiroshima U Hyogo U Kyoto U Kyushu U Kyushu Tokai U NIFS Niigata U U Tokyo JAEA Hebrew U Ioffe Inst RRC Kurchatov Inst TRINITI KBSI KAIST POSTECH ASIPP ENEA, Frascati CEA, Cadarache IPP, Jülich IPP, Garching ASCR, Czech Rep U Quebec Oct.30th, 2010

2. Outline • FIReTIP status • Basic information about FIReTIP • Fringe jump error correction • 3-point density profile by the inversion • FIReTIP applications • Early density distribution • Density fluctuations • ELM behavior

3. Far Infrared Tangential Interferometer/Polarimeter (FIReTIP) is a powerful diagnostic for many applications. Light • Interferometer is the one of the refractive index measurements specialized to plasma density. • A line-integrated density is obtained from phase information of the laser beam. • Methyl alcohol (CH3OH) laser that emits 118.8μm far infrared ( = 2.52 THz ) beam which is favorite for interferometers in most mid-size tokamaks. • A heterodyne interferometer provides direct phase difference between the probing beam and the reference beam • This configure requires two laser beams with slightly different frequencies. • Laser beam amplitude doesn’t matter, provided sufficient signal to noise ratio for detector and electronics. • Frequency shift from Stark-effect enables high intermediate frequency (IF) as about 5MHz which is larger than twice that of common Methyl alcohol lasers. • High bandwidth of signal up to 4MHz is possible because both of high IF from Stark-effect and recently improved electronics. Plasma AIR

4. Electronics in FIReTIP has been recently upgraded leading to better performance especially for fast bandwidth. Simple block diagram of Electronics* Phase Comparator *W-C Tsai, UC Davis Probing Beam 1 Fringe Counter SBD Mixer Mixer 2 I I-Q Mixer SBD Mixer Mixer Local Oscillator Q SBD : Schottky barrier diode mixer Mixer : Frequency Up-converter Reference Beam Local Oscillator Basic information about the FIReTIP output signals

5. Both density-converted signals from the fringe counter and the I-Q mixer are in good agreement with their own features. Raw signals from each output Voltage[V] Switching in turn Density from raw signals x1019 #136094 FC1 FC IQ FC2 Line-averaged electron density[m-3] I Time [sec] Q Arctangent(Q/I) Time [sec] Phase difference φ

6. Line-integrated density up to 3 channels in mid-plane is being measured. Layout of the FIReTIP beam channels. (2) FC signals (for faster analysis with fewer data points) will be available soon including CH.1. (1) CH.3, 5 and 6 of IQ signals are currently on operation after July 2010.

7. Fringe Jump Error has been a serious problem in FIReTIP likely to most other interferometers. x1019 Thomson scattering There were hundreds of discharges suffering from fringe jumps (FJs). The number of fringe jumps in those discharges are from zero to thousands,provided FIReTIP system including laser power is acceptable. Apparently, the whole waveform seems to become more absurd as the number of FJ increases. Even in some discharges with only several fringe jumps the difference between data from FIReTIP and Thomson scattering diagnostics is quite distinct so we are not able to use the original data. FIReTIP 18 FJs #135103 Indicators of fringe jumps when they appeared. Line-averaged electron density[m-3] 368 FJs #136105 Time [sec]

8. Most Fringe Jump Errors appear in FIReTIP have common characteristics. It is easily characterized with common parameters, which is due to the recently upgraded fast electronic system by W.C. Tsai, UC Davis 1. They take place in a few microseconds usually, and up to a couple of tens microseconds. 2. Most of them have typical magnitudes, i.e. one-fringe (2π)-equivalent voltages or densities. For Channel 1 Obviously not a plasma activity

9. Fringe Jump Errors were suppressed by post-processing of stored data. Fringe jump corrected data compared with those of Thomson scattering diagnostics By applying the algorithm, corrected FIReTIP data is compared with Thomson Scattering data which is reproduced along the beam path of FIReTIPCH.#1. (a) An example of full waveforms of each case: The corrected data show very good agreement with those of Thomson scattering (b) Averages and deviations of the discrepancies from most shots after July 2009: They are distinctivelyreduced after correction .

10. This algorithm works even in cases of up to a few hundreds of fringe jumps event. Fringe jump corrected data compared with those of Thomson scattering diagnostic : FI means FIReTIP data, TS Thomson scattering diagnostic data respectively. 368 Fringe jumps were reported and suppressed. Even in this case of huge errors, corrected data has quite reasonable value compared with data without correction in dashed blue line. The correction algorithm has high opportunity to make the data reliable in most situations. Time [sec] This algorithm has been routinely working since FY2010 and it is archiving the corrected FIReTIP data automatically in most cases without distinct problems.

11. FIReTIP data is being archived though PCS digitizer for real-time density control. x1019 CH.3 Fringe counter (FC) data is being used after the fringe jump error correction. 5kHz digitizer is fast enough to control gas injection valves which operate at least in every 2ms. PCS algorithm for density control has to be prepared. x1019 x1019 #139833~138939

12. Interferometer basically provides only line-integrated density information. • In many cases, density profiles are of interest than line-integrated ones. • It can never be determined where the specific density variations take places, if one channel of interferometer data are only obtained. • In order to have spatial resolution, It is required to combine the FIReTIP result with other diagnostics, OR to measure several channels together then reconstruct the profile. #141325 CH.3 CH.5 CH.6

13. Linear interpolation between several points of measurement shows rough profile but good trends following the TS data. Major Radius R [m] Conventional slice & stack concept Interferometer laser beam channels Density [arb. unit] Density [arb. unit] Reconstructed one Reconstructed one Original model profile Original model profile Major Radius R [m] Major Radius R [m]

14. Linear interpolation between several points of measurement provides reasonable profile. Reconstructed density profile [m-3] x1020 Pre-determined density profile [m-3] Calculated interferometer signals on acutal channels of FIReTIP [m-2] x1019 Major Radius R [m] Major Radius R [m]

15. Linear interpolation between several points of measurement shows rough profile but good trends following the TS data. Reconstructed density profile [m-3] x1020 x1020 Pre-determined density profile [m-3] Calculated interferometer signals on each channel of FIReTIP [m-2] Major Radius R [m] Major Radius R [m] x1019

16. It is getting better as channel numbers increase. • Although it is not practical in NSTX, the reconstruction with virtual 13 channels provides much better profile information, as expected, in entire cases that were tried before. Reconstructed density profile [m-3] Major Radius R [m] • However, only 3~4 channels are practically available at the moment without huge sacrifice of reliability of measurements on each channel.

17. Only 3 points of measurements can reconstruct rough profiles but their trends looks good following the TS data. #135695 • The MPTS diagnostic is able to provides much better density profile trace for the global plasma waveform. • Instead, FIReTIP is capable of seeing much faster behavior than MPTS does.

18. One example is the early electron distribution . • As a well-known feature, the electrons are generated mainly at the HFS region during plasma ramp-up. • The reason is though to be the combination of higher connection length and the higher inductive field strength in HFS of tokamak. #135695

19. One example is the early electron distribution .

20. FIReTIP can roughly localize the density fluctuation #135161 Pay attention to this edge.

21. Termination of plasma is visible. #135163

22. FIReTIP can provide a particle point of view for plasma behaviors, for example, ELMs. • Dα signals has distinct relation to the electron density reduction in ELM events. #132403 #132401 #132407 '\bayc_dalf_fscope'

23. Even ELM activity is still slow enough for FIReTIP. Dα #132403 • ELMs are thought to take place in LFS of tokamaks. (Kamiya 2007) This time

24. Another shot is also indicating the density loss at the HFS of NSTX rather than LFS. #132401

25. 3-D view of density profile shows ELM events clearly LFS HFS Time

26. 4 channels makes the ELM behavior more clear. #132403 Layout of the FIReTIP beam channels. CH. 1 CH.3 CH. 5 CH. 6 Dα We have few discharges with 4

27. A Filament rotation is measured in FIReTIP. #135163 Time [ms]

28. Filament peak points are measured and compared in each channel. CH.1 #135163 Δt31 Δt13 CH.3 Δt53 CH.5 Δt35 Δt55

29. Line-integrated density up to 3 channels in mid-plane is being obtained. Layout of the FIReTIP beam channels. Calculated path lenth distribution according to traveling radius R • But measured peaks do not have similar relation from the calculated ones which means the filaments have complex behavior unlikely to assumed simple travel. L1 L5 L4 L2 L3 R L3 φ L4 L2 L1 L5 R [m] L R Travel length L=Rφ φ

30. Summary • Recently upgraded electronics from UC Davis enables the fast signal processing and this leads to the easy characterization and suppression of fringe jumps in FIReTIP. • About 1MHz and 4MHz signal bandwidth from FC and IQ were achieved. • Fringe jump errors are automatically corrected after almost every shot in FIReTIP normal operation. • The correction algorithm is also applicable to the FIReTIP signals digitized by PCS for real-time density control and they are being archived together. • 1st step of density profile has been carried out via conventional method with linear interpolation that compensates the limited number of channels. • Comparing with MPTS data, the trend from FIReTIP data in each shot seems reasonable considering just 3 channels are available at the moment. • The fast density profile provides a little bit of local information of fast variations such as early density distribution, density fluctuations and ELMs.