Miniproposals

Miniproposal: 306 Date Filed: 6/19/2002 Title: Density Limits, Transport and Turbulence First Author: Martin Greenwald Session Leader: Martin Greenwald (shots 1-35)

Engineering Operator Run Comment

MP#306 Density Limits, Transport and Turbulence; SL:Greenwald; PO:Wolfe;

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Entered: Jul 7 2004 03:49:29:693PM

Author: To Be Determined

Run Plan for 1020807 mp 306 Density Limits Session Leader - Greenwald Phys Op - Wolfe ============================ Plan: These experiments will extend our previous work in two ways. First, we will run low power Ohmic discharges on the outer limiter. This will allow unambiguous determination of the last-closed flux surface and allow maximum penetration of the scanning probe into the closed flux surface region. The hope is to follow the evolution of the profiles as the collapse is approached. This proposal follows the approach of previous runs where, after establishing a high density discharge, the plasma current is ramped down to reach the density limit at a predictable and reproducible time. (See for example, 1000606014) We will also be monitoring edge turbulence with the new 28 frame PSI camera. ============================== Results: An OK day. The scenario we outlined resulted in strong marfes appearing at about 70% of the limit followed by a sharp density rise and a disruption. We proceeded with this shot and successively scanned the A-side probe, reaching past the last closed flux surface by almost 3 cm. ISat was bursty for all cases, though the correlation time shortened from 30 usec to about 10 usec as the probe was inserted more deeply. At the same time, we got some good data with the psi camera. Blobs were generated from within the lcfs. The DNB was successfully and reliably operated in modulated mode, using three 14 ms pulses separated by 4-ms off-periods. MSE observed high background light during the high-density phase of the discharges, in fact one channel looking across R=82.5 cm often saturated. This may be due to light originating from contact between the plasma and outlying structure. We took the opportunity to take MSE measurements early in the discharge startup when the q-profile was still evolving and when the density was still modest. Good MSE data was acquired on 8-9 discharges, over times 0.05 - 0.60 sec, during which the engineering qedge as deduced by EFIT ranged from 7+ to about 4.7. This data will be useful as benchmark/calibration for the MSE shakedown. ================================ Shots suitable for further analysis include: shot probe depth (mm) 1020807010 1 mm 1020807011 0 1020807012 0 1020807014 -1 1020807015 0 1020807017 0 1020807019 -5.5 1020807021 -6 1020807027 -12.5 1020807028 -16.5 1020807029 -23 1020807030 -23 1020807033 -26 1020807035 -28.5

Physics Operator Summaries

Entered: Jul 7 2004 04:37:15:583PM

Author: To Be Determined

Physics Operator Summary for Wed. 7-AUG-2002 MP#306 Density Limits, Transport and Turbulence SL: Greenwald PO: Wolfe EO: Dekow,Fitzgerald Plan: ----- Run to begin at 9:00 AM ECDC from 06:00-07:00 at 2e-4 T D2. Power systems setup the same as for Shot 1020802029 Gas setup: B-Top 6 psi D2 Hybrid Enabled (PG4) B-side lower 1 psi Ar Hybrid Enabled (PG1) B-side upper 10 psi D2 Hybrid DISABLED (PG2, for ECDC) C-Side 30 psi D2 Hybrid Enabled (PG3) J-Bottom 30 psi D2, Hybrid DISABLED Enable the following gate valves, assuming no vacuum problems: ECE,VUV,DNB RF Setup: None, this will be the ohmic part of the MP. Run Plan: Set up an outboard limited shot at 800kA, ramping down at about 1kA/msec while raising the density to ~2e20 nl_04. Perhaps use J-bottom piezo to augment the main gas puff. Tune up inner gap and triangularity to avoid excess heating on the BN tiles on the J-antenna, and to allow the inner wall probe to sample the SOL. Once the shot is established, leave it alone while probes are scanned, etc., as called for in the MP. These experiments will extend our previous work in two ways. First, we will run low power Ohmic discharges on the outer limiter. This will allow unambiguous determination of the last-closed flux surface and allow maximum penetration of the scanning probe into the closed flux surface region. The hope is to follow the evolution of the profiles as the collapse is approached. This proposal follows the approach of previous runs where, after establishing a high density discharge, the plasma current is ramped down to reach the density limit at a predictable and reproducible time. (See for example, 1000606014) We will also be monitoring edge turbulence with the new 28 frame PSI camera. Results: -------- Started by modifying the PCS programming from a SN 800kA shot as follows: Load seg 2 from seg 3 (0.8MA SN) and increase rcur from .661 to .675 increase zcur from -.015 to 0. ramp ip from [.7,-8e5] to [1.5,0.] ramp nl_04 to 2.5e20 @1.5s pg3 waveform starting with 70V, 40V maint level in seg1 keep zcur at 0 through .1 sec (to match seg2) ic_ef4 from -1305 to -1315 The run began on time (actually slightly early, 8:55). Outboard limited shots were established, but the disruptions were coming at ~60-70% of the nominal limit. Changing the density and current ramp rates did not seem to help. The problem seemed to be a marfe that moved from the top to the bottom. Small equilibrium tweaks to zxl and zxu and the inner gap did not change this behavior. Neither did turning off the segment 2 argon, although we didn't turn it back on again. Shot 17 was an attempt to go to a smaller dsep, i.e. nearly diverted, but I went the wrong way on rcur and actually only succeeded in increasing the inner gap to ~4cm. On shot 19 we did go to a more nearly diverted case, still limiting on the outside with dsep ~-2mm. The marfe behavior changed some, but the disruption still came at about the same time, and same n/I. Martin decided to keep this equilibrium and do the probe scans, and that's where we stayed the rest of the day. The disruptions were coming just before 1sec. Brian got a good set of deep probe scans with the ASP, well inside the LCFS. He also got the inner wall scanning probe running and had some interesting data from it. For instance, on shot 28 it records plasma signal at 0.6 sec, not much of anything at 0.9. EFIT indicates the earlier time has the inner wall connected to the outboard around the top, but at the later time the outboard flux goes tp upper and lower divertor and the inner wall is in a private flux region. The startup was fairly reliable early in the day, although shots that bounced (high Ip/Bz, outward moving) did have runaway problems. Shots 5 and 9 were duds caused by power supply and gas faults, respectively. Later in the run, the startup became problematic. Low pre-puffs and/or more negative Bz, gave runaways and bounces, while higher pressure gave fizzles. Then we started getting bouncy fizzles. And then snuff (low current/Bz) fizzles. Shots 22-26 were a string of five straight fizzles; changes to gas, rcur offset and waveform, and five minutes of ecdc didn't help. Finally a 15A EF4 tweak got the plasma back for four shots (27-30). The first three seemed to evolve toward slower current rise, then 30 moved back out and bounced, and we got a double bounce fizzle on #31. Raising the gas got the startup back on 32-33, but 34 fizzled again, this time losing it inward (snuff). Dropping the gas for 35 was successful, but again with a bounce and runaways. We definitely need to do better on the startup, but it isn't clear to me how. I also turned down the argon in the prepuff from 20msec to 10msec at shot#11, and turned it off completely on shot#26. Since we got no duds today (except for PS and gas faults) it isn't at all clear whether the argon seeding was helping anything. NOTE: Prior to the run, Dave Terry fixed the ZMETER AFOL (again). The fix entailed reversing the polarity of the input leads, so that the AFOL will no longer invert. This means that old nl_zmeter programming will need to be inverted for future use. I put an entry in the hybrid config_id to flag this for future runs. Cooling was very good today, resulting in 35 total shots. Scorecard: ---------- Duds 2 (1 power supply fault in pre-charge, 1 no gas) Fizzles 10 Plasmas 23 ======= == Shots 35

Session Leader Comments
Aug 7 2002 08:50:56:417AM		Martin Greenwald	Run Plan MP 306 Density Limits, Transport and Turbulence SL: Greenwald PO: Wolfe These experiments will extend our previous work in two ways. First, we will run low power Ohmic discharges on the outer limiter. This will allow unambiguous determination of the last-closed flux surface and allow maximum penetration of the scanning probe into the closed flux surface region. The hope is to follow the evolution of the profiles as the collapse is approached. This proposal follows the approach of previous runs where, after establishing a high density discharge, the plasma current is ramped down to reach the density limit at a predictable and reproducible time. (See for example, 1000606014) We will also be monitoring edge turbulence with the new 28 frame PSI camera.
Aug 7 2002 09:13:47:060AM	1020807001	Martin Greenwald	Disrupts during flattop
Aug 7 2002 09:24:26:927AM	1020807002	Martin Greenwald	Gets into rampdown phase, but still disrupts early lots of radiation - near 100%
Aug 7 2002 09:40:17:720AM	1020807003	Martin Greenwald	disrupts at 1 second radiation slightly lower
Aug 7 2002 09:55:17:670AM	1020807004	Martin Greenwald	Similar result, disrupts a 1 second radiation about 50% for most of flattop - goes up toward the end of shot
Aug 7 2002 10:01:36:033AM	1020807005	Martin Greenwald	power supply problem - no plasma
Aug 7 2002 10:28:50:727AM	1020807006	Martin Greenwald	similar shot - disrupts early
Aug 7 2002 10:36:39:843AM	1020807007	Martin Greenwald	similar shot He puff added for fast camera
Aug 7 2002 11:07:01:287AM	1020807009	Martin Greenwald	fizzle
Aug 7 2002 11:17:13:577AM	1020807010	Martin Greenwald	slightly higher density, but disrupts a bit earlier still 65-70% of limit
Aug 7 2002 11:38:26:093AM	1020807012	Martin Greenwald	similar to previous shot triggering problem on fast camera
Aug 7 2002 12:12:53:947PM	1020807014	Martin Greenwald	Similar shot
Aug 7 2002 12:36:43:810PM	1020807016	Martin Greenwald	fizzle
Aug 7 2002 01:07:13:723PM	1020807018	Martin Greenwald	fizzle
Aug 7 2002 01:45:22:627PM	1020807019	Martin Greenwald	went to marginally diverted discharge probe about to separatrix
Aug 7 2002 01:45:45:567PM	1020807020	Martin Greenwald	fizzle
Aug 7 2002 02:06:32:223PM	1020807021	Martin Greenwald	Similar shot probe in about 2 mm
Aug 7 2002 02:06:12:360PM	1020807022	Martin Greenwald	fizzle
Aug 7 2002 02:18:39:920PM	1020807023	Martin Greenwald	fizzle
Aug 7 2002 02:32:06:710PM	1020807024	Martin Greenwald	fizzle
Aug 7 2002 02:49:36:897PM	1020807025	Martin Greenwald	fizzle
Aug 7 2002 03:02:22:983PM	1020807026	Martin Greenwald	fizzle
Aug 7 2002 03:21:31:547PM	1020807027	Martin Greenwald	shot is back probe in a bit farther still bursty fluctuations
Aug 7 2002 03:42:44:827PM	1020807028	Martin Greenwald	Similar shot probe in another 5 mm
Aug 7 2002 03:55:53:197PM	1020807029	Martin Greenwald	similar shot - but disrupted about 100 msec earlier probe in another 7mm
Aug 7 2002 04:10:44:303PM	1020807030	Martin Greenwald	similar shot probe caught plasma just before crash in about 5 mm
Aug 7 2002 04:16:03:993PM	1020807031	Martin Greenwald	fizzle
Aug 7 2002 04:30:11:253PM	1020807032	Martin Greenwald	disrupted early no probe data
Aug 7 2002 04:52:44:663PM	1020807033	Martin Greenwald	similar shot - last through probe plunge in another 5 mm from last good shot
Aug 7 2002 04:53:36:163PM	1020807034	Martin Greenwald	fizzle
Aug 7 2002 05:06:47:737PM	1020807035	Martin Greenwald	similar shot lasts long enough for probes and camera data

Physics Operator Comments
Aug 7 2002 08:06:47:650AM		Steve Wolfe	setting up for outboard limited shot: Load seg 2 from seg 3 (0.8MA SN) and increase rcur from .661 to .675 increase zcur from -.015 to 0. ramp ip from [.7,-8e5] to [1.5,0.] ramp nl_04 to 2.5e20 @1.5s pg3 waveform starting with 70V, 40V maint level in seg1 keep zcur at 0 through .1 sec (to match seg2) ic_ef4 from -1305 to -1310 load/clean coldstart All matrices respond Hloader log ok.
Aug 7 2002 08:12:57:640AM		Steve Wolfe	Change ic_ef4 to -1315
Aug 7 2002 09:10:02:010AM	1020807001	Steve Wolfe	shot#1: plasma to 0.8sec density rise is probably too fast, it reached 1.5e20 by 0.75sec current was down to 700kA, as programmed, although there were some bites out of it earlier Plasma was outer limited. Equilibrium looks good, ghlim very bright. next shot: reduce pre-program gas in seg2, slow down nl4 program.
Aug 7 2002 09:22:56:207AM	1020807002	Steve Wolfe	shot#2: reduce pre-program gas in seg2, slow down nl4 program. plasma to 0.9sec. Ip down to 570kA, nl4 only to 1.1 next shot: reduce the Ar puff in seq2. tweak early Ip program to be higher at 0.1sec, raise gain on Ip from 2.6 to 6 at .35s change ip rampdown to 0.2MA at 1.5sec
Aug 7 2002 09:38:56:033AM	1020807003	Steve Wolfe	shot#3: reduce the Ar puff in seq2. tweak early Ip program to be higher at 0.1sec, raise gain on Ip from 2.6 to 6 at .35s change ip rampdown to 0.2MA at 1.5sec plasma to 1sec. Ip to 550kA, nl4 to ~1.2 inner gap is 3.5cm next shot: clearin from .005 to 0., leave rcur where it is.
Aug 7 2002 09:56:46:640AM	1020807004	Steve Wolfe	shot#4: clearin from .005 to 0., leave rcur where it is. plasma to 1sec, marfe disappears between .45 and .8 inner gap is still over 3cm. Of course, it's the gap to the x-point surface. next shot: clearin in to -.01, rcur in by .005
Aug 7 2002 10:04:59:380AM	1020807005	Steve Wolfe	shot#5: clearin in to -.01, rcur in by .005 dud. PF breaker pulled during charge. OH2L
Aug 7 2002 10:28:42:383AM	1020807006	Steve Wolfe	shot#6: Modified config_id on hybrid tree to indicate change in zmeter input polarity. Out of PCS and back in, reloaded shot to update config_id in PCS. still clearin in to -.01, rcur in by .005 plasma to 1.0 + inner gap down to 2.5cm next shot: rxl in by 1cm NINJA will puff He next shot.
Aug 7 2002 10:40:37:737AM	1020807007	Steve Wolfe	shot#7: rxl in by 1cm NINJA will puff He from 0.7sec plasma to 0.96sec, rxl moved as programmed. Still disrupting below the greenwald limit. This one bounced in startup, some hards. next shot: increase nl4 program slope. move ip gain step up to 6 from .35 to .25 sec increase pre-puff from 20msec to 22msec
Aug 7 2002 10:55:41:373AM	1020807008	Steve Wolfe	shot#8: increase nl4 program slope. move ip gain step up to 6 from .35 to .25 sec increase pre-puff from 20msec to 22msec plasma to 0.9sec. next shot: minor clearin tweak at .275, increase ip to flattop at .3 instead of .4
Aug 7 2002 11:03:47:020AM	1020807009	Steve Wolfe	shot#9: minor clearin tweak at .275, increase ip to flattop at .3 instead of .4 dud. No gas, didn't get HV on the piezo's. Both oh2's swung positive after t=0.3sec; this is probably because I raised the current gain in seg2 next shot: try again, hope for gas
Aug 7 2002 11:19:23:330AM	1020807010	Steve Wolfe	shot#10: try again, hope for gas plasma to .94 sec, lots of hards, it bounced again. next shot: increase the pg4 to 24msec; reduce seg1 argon from 20msec ->10msec reduce seg2 pg3 pre-program early
Aug 7 2002 11:33:24:323AM	1020807011	Steve Wolfe	shot#11: increase the pg4 to 24msec; reduce seg1 argon from 20msec ->10msec reduce seg2 pg3 pre-program early Still with the hards. Still bounces Plasma to .97sec next shot: ip gainstep to .15sec pg4 in seg1 to 26msec
Aug 7 2002 11:52:26:540AM	1020807012	Steve Wolfe	shot#12: ip gainstep to .15sec pg4 in seg1 to 26msec plasma to 0.98sec No bounce this time, and no hards. next shot: no changes
Aug 7 2002 12:07:13:630PM	1020807013	Steve Wolfe	shot#13: no changes plasma to .90sec next shot: move zxu up by 1cm, zxl down by 1cm to try to affect the marfe.
Aug 7 2002 12:18:32:750PM	1020807014	Steve Wolfe	shot#14: move zxu up by 1cm, zxl down by 1cm to try to affect the marfe. plasma to .97, marfe looks the same. next shot: Turn the seg2 argon off. go the other way on zxl and zxu
Aug 7 2002 12:32:02:760PM	1020807015	Steve Wolfe	shot#15: move zxu down by 2cm, zxl up by 2cm to try to affect the marfe. plasma .98sec. Looks the same. next shot: try to get dsep closer to zero clearin from -.01 to .005 rcur from .67 to .6775
Aug 7 2002 12:41:42:737PM	1020807016	Steve Wolfe	shot#16: try to get dsep closer to zero clearin from -.01 to .005 rcur from .67 to .6775 fizzle. Bz is a bit positive, current rise a bit slow. next shot: drop pg4 to 24msec
Aug 7 2002 12:57:34:060PM	1020807017	Steve Wolfe	shot#17: drop pg4 from 26->24msec try to get dsep closer to zero clearin from -.01 to .005 rcur from .67 to .6775 plasma to 0.95, hards again. bounce again. Should have left the gas alone marfe looks similar. That didn't do what I wanted. Inner gap got bigger, still outboard limited. I shouldn't have moved the rcur in that direction. next shot: copy SN traces for rcur, clearin, zxl, zxu move rcur out by .0075
Aug 7 2002 01:05:54:880PM	1020807018	Steve Wolfe	shot#18: pg4 from 24->26msec copy SN traces for rcur, clearin, zxl, zxu move rcur out by .0075 fizzle. slow current rise? next shot: pg4 to 25msec (do next to nothing)
Aug 7 2002 01:33:47:633PM	1020807019	Steve Wolfe	shot#19: pg4 to 25msec copy SN traces for rcur, clearin, zxl, zxu move rcur out by .0075 plasma to .94sec or so, marfe is different. right gap is 0, left gap ~2cm, dsep -4mm to 4mm next shot: no changes
Aug 7 2002 01:40:54:610PM	1020807020	Steve Wolfe	shot#20: no changes fizzle! Looks like it's moving in. next shot: Try changing the rcur offset in seg1 from 3000->2000
Aug 7 2002 01:53:57:557PM	1020807021	Steve Wolfe	shot#21: Try changing the rcur offset in seg1 from 3000->2000 plasma to .95 or so, bounced in startup. Got hards again. cadf05 had camac error, so I can't see the phys_aout response. next shot: rcur offset to 2500
Aug 7 2002 02:09:39:647PM	1020807022	Steve Wolfe	shot#22: rcur offset to 2500 fizzle. Looks like it snuffs again. next shot: change seq1 rcur waveform from .63 to .64 (I'm getting into Ian mode)
Aug 7 2002 02:21:00:780PM	1020807023	Steve Wolfe	shot#23: change seq1 rcur waveform from .63 to .64 (I'm getting into Ian mode) fizzle. slightly late breakdown, never got going. next shot: reload seg1 from shot11
Aug 7 2002 02:35:05:400PM	1020807024	Steve Wolfe	shot#24: reload seg1 from shot11 (reduced argon and 24msec pg4, before the rcur tweaks) fizzle again. double bounce: first on on R, secont on Z. Now it looks like the current rise is too fast. ECDC for 5minutes and go again.
Aug 7 2002 02:52:46:977PM	1020807025	Steve Wolfe	shot#25: No pcs change, run ecdc for 5minutes before shot. Started init at 2e-6 on IGT fizzle. Four in a row! This one snuffs with too low a current. Turn off the argon in seg1
Aug 7 2002 03:06:05:800PM	1020807026	Steve Wolfe	shot#26: Turn off the argon in seg1 And another fizzle! Five in a row. Looks like it snuffed again. Bz looks very positive. next shot: ef4 to -1330.
Aug 7 2002 03:26:47:800PM	1020807027	Steve Wolfe	shot#27: ef4 to -1330. plasma to .97sec. Finally! And no bounce. The rcur looks like it came close though. next shot: no changes (and hope)
Aug 7 2002 03:33:57:273PM	1020807028	Steve Wolfe	shot#28: no changes (and hope) plasma, same length. When it goes, it's reproducible. slightly slower on startup, and not quite as far out. next shot: no changes (I'm tempted to drop the gas, but I won't)
Aug 7 2002 03:46:52:440PM	1020807029	Steve Wolfe	shot#29: no changes (I'm tempted to drop the gas, but I won't) plasma, looks the same. Startup also looks the same, no changes
Aug 7 2002 04:06:03:330PM	1020807030	Steve Wolfe	shot#30: no changes plasma to .93. This one had a bouncy startup and hards again. I could add some gas, but I'll wait one more to see if it gets better or worse
Aug 7 2002 04:18:06:040PM	1020807031	Steve Wolfe	shot#31: no changes and it fizzled. double bouncer. next shot: raise pre-puff to 26msec
Aug 7 2002 04:37:26:447PM	1020807032	Steve Wolfe	shot#32: raise pre-puff to 26msec plasma only to .84sec. Still bouncing and lots of hards. Looks like a small injection at .805 tips it over the edge maybe from J next shot: push pre-puff up to 28msec
Aug 7 2002 04:42:15:733PM	1020807033	Steve Wolfe	shot#33: push pre-puff up to 28msec plasma to 0.99sec; no bounce no hards. next shot: no changes.
Aug 7 2002 04:53:49:570PM	1020807034	Steve Wolfe	shot#34: no changes. fizzled next shot: drop pre-puff back to 25msec
Aug 7 2002 05:06:54:937PM	1020807035	Steve Wolfe	shot#35: drop pre-puff back to 25msec plasma to 0.97sec, it bounced but made it. Lots of hards again. next shot: Will be tomorrow!

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	08:57:27:720AM	Plasma	Ok
2	09:11:50:230AM	Plasma	Ok
3	09:24:58:270AM	Plasma	Ok
4	09:39:46:470AM	Plasma	Ok
5	09:57:51:390AM	Plasma	Bad	OH2L Fault at commutation switch close
6	10:17:59:000AM	Plasma	Ok
7	10:30:13:167AM	Plasma	Ok
8	10:43:02:347AM	Plasma	Ok
9	10:56:34:997AM	Plasma	Bad	No Gas due to no high voltage enabled
10	11:08:57:383AM	Plasma	Ok
11	11:21:13:767AM	Plasma	Ok
12	11:37:57:483AM	Plasma	Ok
13	11:51:07:663AM	Plasma	Ok
14	12:08:30:633PM	Plasma	Ok
15	12:22:24:690PM	Plasma	Ok
16	12:34:46:667PM	Plasma	Ok
17	12:47:22:163PM	Plasma	Ok
18	01:01:19:047PM	Plasma	Ok
19	01:14:43:527PM	Plasma	Ok
20	01:34:29:573PM	Plasma	Ok
21	01:47:13:100PM	Plasma	Ok
22	02:04:16:563PM	Plasma	Ok
23	02:16:57:857PM	Plasma	Ok
24	02:29:46:083PM	Plasma	Ok
25	02:47:47:347PM	Plasma	Ok
26	03:00:49:513PM	Plasma	Ok
27	03:13:32:337PM	Plasma	Ok
28	03:27:32:067PM	Plasma	Ok
29	03:43:46:343PM	Plasma	Ok
30	03:59:48:620PM	Plasma	Ok
31	04:13:01:240PM	Plasma	Ok
32	04:25:53:750PM	Plasma	Ok
33	04:38:44:497PM	Plasma	Ok
34	04:51:08:067PM	Plasma	Ok
35	05:03:34:623PM	Plasma	Ok