| Miniproposals | ||||||||||||
|
| Operators | |
| Session leader(s): | Robert Granetz,Ian Hutchinson |
| Physics operator(s): | Steve Wolfe |
| Engineering operator(s): | Bill Byford,Gary Dekow,Bill Parkin |
| Engineering Operator Run Comment |
| MP#331 Locked Mode Studies with Non-axisymmetric Coils; SL: Hutch/Granet |
| Session Leader Plans |
| Physics Operators Plans |
| Session Leader Summaries |
| Physics Operator Summaries |
| Entered: Jul 7 2004 04:37:17:707PM |
| Author: To Be Determined |
| Physics Operator Summary for Wednesday, 1030507 MP#331 Initial Locked Mode Studies with Non-axisymmetric coil set SL: Hutch (in absentia?), Granetz, (wolfe) PO: Wolfe EO: Dekow,Byord,Parkin Setup and Plan: --------------- ECDC overnight in D2 at 2e-4 Torr, stopping by 07:00. Bake at 60 C overnight as well, if possible. Run begins at 9:00 and ends at 17:00. Power systems the same as on 1030425011 (1 MA, 5.4 T, lower single null) C-Coil: patch panel set up as [+Dtop,-Dbot,-Jtop,+Jbot], NOT enabled for first shot Gas setup: fill B-Top with 6 psi D2 Hybrid enabled (PG4) fill B-side lower with 1 psi Ar Hybrid enabled (PG1) fill B-side upper with 10 psi D2 Hybrid DISABLED (PG2) (for ECDC) fill C-side with 30 psi D2 Hybrid enabled (PG3) Enable the following gate valves, assuming no vacuum problems: ECE, DNB, VUV ------------------------------------------------------------------------------- RF Setup: All 4 transmitters in Remote (but NOT in Run On) by 08:00. ------------------------------------------------------------------------------- Run Plan: -------- The idea is to determine the threshold current in the non-axisymmetric coils required to induce a locked mode in a 1MA moderate density plasma, as a function of mode spectrum and helicity. Magnetics (and other) data will be used to try to characterize the phase of such induced perturbations, and if possible the island size, relative to the orientation and amplitude of the vacuum perturbation. Assuming this part of the experiment is successful, we will then try to apply the information gained to optimize suppression of a naturally occuring locked mode at lower density. I. Reproduce shot 1030425011, a 1MA, nl04=6e19 discharge, to see if locked mode still appears. If not, don't waste time looking for it (yet).[1 shot] II. Increase density to nl04=1×10^20 m-2 or sufficient to remove locked mode. [1-3 shots] IIIa With patch panel set to +Dtop -Dbot -Jtop +Jbot, program a ramp from zero to 3700A (it probably won't get over 2300) between t=0.6sec and t=1.5sec, then back down at EoF. If a locked mode appears, take additional shots at slower ramp rates to refine threshold value, perhaps ramp back down to determine hysteresis. If no locked mode appears, try reducing nl04 to 7×10^19 m-2 and repeat; if this results in a locked mode, turn off C-coil for a shot and make sure locked mode goes away, then proceed as above. [2-6 shots] IIIb Reverse current, i.e. -Dtop +Dbot +Jtop -Jbot and establish current threshold for onset of locked mode (if any) as in IIIa [1-3 shots] IV Change patch panel taps to +Dtop +Dbot -Jtop -Jbot, corresponding to the black solid curve in Figure 2 (symmetric spectrum with large m=1, reduced m=2). As in case III, establish a threshold, if any, with positive current and then reverse the current sense to -Dtop -Dbot +Jtop +Jbot. [2-4 shots] V Change patch panel to +Btop -Dbot -Gtop +Jbot, corresponding to the blue triangle case in Fig. 2. This is supposed to have the highest amplitude of m=1 and m=2 in the correct helicity. Repeat the threshold determination for this configuration and with reverse current -Btop +Dbot +Gtop -Jbot. [2-4 shots] VI Change patch panel to +Btop +Dbot -Gtop -Jbot, corresponding to the reversed helicity case (green diamonds). This is expected to have the least effective field pattern for causing a locked mode, and hence the highest threshold, unless I got the signs wrong. Again establish a threshold current, if any, for positive and negative polarity (-Btop -Dbot +Gtop +Jbot). [2-4 shots] VIIa Reduce the density until a locked mode spontaneously appears with no current in the non-axisymmetric coils. [1-3 shots] VIIb Based on results from steps III - VI, and if possible on estimates of the phase of the mode in VIIa , choose a patch panel configuration and current level and attempt to suppress the onset of the mode. It may be best to employ a constant current level, or perhaps a slow ramp, varying the current shot by shot. [2-5 shots] VIII If time permits, explore thresholds for excitation and suppression with additional tap configurations, such as using only two coils in series to sample more toroidal phase options at reduced total amplitude. Results ------- The run went pretty well. All of steps I-VIIB were completed. We were successful in generating locked modes at 1MA, nl04=1e20 in only one configuration and one polarity, namely the quadrupole of IIIB (-Dtop +Dbot +Jtop -Jbot). The threshold current in this configuration is between 1300A and 1800A, based on shots 8,10,12,and 14; shot #24 was a repeat of the conditions of shot 8, and showed reproducibility. We were also successful in suppressing a natural locked mode at 1MA, nl04=6e19, using maximum current of 2300A with the opposite polarity, +Dtop -Dbot -Jtop +Jbot (shots #25,26,27). This was the same polarity used on 1030425. The fact that we were unable to generate a locked mode with maximum current in the other four-coil series combinations indicates that we are marginal with respect to current. It also implies that the existing error field is quite large. We should see about getting a higher voltage supply on the coils for this purpose, and possibly a higher current as well. It is especially surprising that the non-symmetric (helical) configurations were unsuccessful in producing a locked mode, since they have essentially the same 2/1 amplitude as the quadrupoles and nearly twice the 1/1; of course, the toroidal phase is different, so it may be that they were always bucking one component or the other of the existing error field. The most sensitive magnetic signature of the locked mode seems to be the BP26DE-JK difference, which shows a very sharp feature with amplitude of ~10mT; however, this pair also showed some erratic behavior, possibly a baseline drift. It was not clear that the sign of the locked mode perturbation on a given pair was the same as the sign of the response to the coil in the absence of the mode; this needs to be investigated. John was looking at rotation and, especially on the natural locked mode #25, saw the rotation go from counter to zero, and on the suppressed case #27 it continued spinning. I believe that FRECE was taking data, but I didn't get a chance to look to see if an island was visible in the profile. Certainly both the natural and induced locked modes took substantial bites out of the stored energy and the density, as well as suppressing (or changing) the sawteeth and increasing the Halpha. Other Issues: --------- We went through a long period of alternating fizzles between shots 4-17, probably because I had done too much playing at getting rid of the hard x-rays. Shots 18-27 all ran successfully and only the last had hards; it's not clear why, but a startup from that sequence may be successful again. Among other things, I had tweaked the EF3_volt and pulse gas in seg1; I did one IC_EF4 tweak but put it back after a fizzle. PCS behaved a little funny on initial setup this morning. I imported seg2 from 1030425011 and changed the programming of the Acoil and on load/clean I got warnings about Seg2 wires P11 and P12 having been renamed since they were drawn; these wires were "zero_11" and "zero_12" and had not been renamed. I had to actually touch/move a point in each to get rid of the warning. Factor screens also popped up for IC_EF4 and Ip, which had not been changed. Accepting these once was sufficient to not see them again. I sent a message to jas about it; seems probably harmless. There was some confusion about how the patch panel changes needed to be done. When I specified the reverse polarity configurations the engineers were actually changing all the connections to match, rather than just swapping the power supply leads. This does the right thing, but takes a lot longer. I should probably have specified the reversed polarity quadrupole, e.g., as -Jbot +Jtop +Dtop -Dbot instead of -Dbot +Dbot +Jtop -Jbot. Scorecard --------- Tests 0 Duds 0 Fizzles 7 Plasmas 20 ------- -- Shots 27 |
| Session Leader Comments |
| Physics Operator Comments | |||
| May 7 2003 07:41:57:830AM | Steve Wolfe | Setting up for non-axi coil run Import seg2 from 1030425011 (1MA nl04 program for 4e19). This shot had a locked mode at a density of 6e19. Acoil current programmed for 3700 (but wasn't run on that shot) Startup is from yesterday: Br=.0015, EF4=-1320, PG4=24msec, ef3volt=2650,2600,2550,2250 Change Acoil program: 0.5 sec 0 0.60 700A 1.3sec 3700A 1.5 700 1.7 0 seg2 pg3 programming looks strange, but I'll leave it for the first shot. Do a load. Got a message about icef4 factor? and Ip factor. Don't know why these are coming up. I'm also getting a message anout waveform in wire :p11 seg2 not being the same as when it was last drawn? This seems to be a zero'd wire. What's going on? Also :p12 whichi is also zero. | |
| May 7 2003 07:52:55:077AM | Steve Wolfe | Try taking a test shot to make sure the pcs tree doesn't have a problem. test shot #22 looks ok. | |
| May 7 2003 08:00:57:353AM | Steve Wolfe | Do another load, just to see. I get the warnings on p11 and p12 of seg 2 again. It says the name has changed since it was last drawn. Looks like a bug; the zero button was used long ago on this. Just to make sure it goes away, I'll move a point on the waveforms for zero_11 and zero_12 (and zero_10 just to be sure) Load it one more time. No warnings this time. Have to ask jas to look at the code that generates this message; hybrid test shot #22 will have the mixed times, if they are real. | |
| May 7 2003 08:47:22:613AM | Steve Wolfe | For shot#1, raise the Ar puff from 40 to 50msec | |
| May 7 2003 09:10:11:257AM | 1030507001 | Steve Wolfe | shot#1: Hung in init. Josh cleared it after 3minutes plasma, lots of hards. full length. No locked mode, unless it's ther the whole time, density was low (<4e19 until the tci crapped out). Zmeter signal looks dead. next shot: raise nl04 to 8e19 |
| May 7 2003 09:31:13:867AM | 1030507002 | Steve Wolfe | shot#2: raise nl04 to 8e19, draw pg3 to ~35V through shot. icef4 from -1320 to -1310 in seg 1 plasma. Still lots of hards, looks like a locked mode around 1.39, judging from the ECE, small break in slope on BP diff signals at 1.31, need to raise nl a litle more. next shot: nl4 to 1e20, raise ef3 voltage |
| May 7 2003 09:47:27:533AM | 1030507003 | Steve Wolfe | shot#3: nl4 to 1e20, raise ef3 voltage to 2750, 2650 FRECE pci is hanging in INIT every shot; twf is looking into it, I hope plasma, no hards. No locked mode. Next shot: Turn on the C-coil (+Dtop -Dbot -Jtop +Jbot) ramp |
| May 7 2003 10:08:05:800AM | 1030507004 | Steve Wolfe | shot#4: Turn on the C-coil (+Dtop -Dbot -Jtop +Jbot) ramp from .6 to 1.3sec fizzle. may have overdone the EF3 The pickup on the ccoil shunt trace seems to be correlated with the chopper current; maybe the EFC TMX supply? next shot: drop ef4 back to -1320A |
| May 7 2003 10:16:26:970AM | 1030507005 | Steve Wolfe | shot#5:drop ef4 back to -1320A Turn on the C-coil (+Dtop -Dbot -Jtop +Jbot) ramp from .6 to 1.3sec plasma, no hards, no locked mode. Only got up to 2100A it looks like; shunt trace is noisy, still has -300A offset Next shot: move on to IIIb (reverse polarity) |
| May 7 2003 10:41:39:130AM | 1030507006 | Steve Wolfe | shot#6: move on to IIIb (-Dtop +Dbot +Jtop -Jbot) I'll boost the top program value to 4000A plasma, no hards. Maybe something coming in very late in the pulse, around 1.36 Lose sawteeth in rampdown. Next shot: Bring C-coil current up early and hold it flat. |
| May 7 2003 10:51:56:587AM | 1030507007 | Steve Wolfe | shot#7: Bring C-coil current up to 4000 (demand) at 0.7 sec, flat to 1.3 I think I'm just looking at the ramp-down on shot #6, not a plasma effect. If this doesn't show anything, I'll move on. fizzle! Try dropping the EF3V back some |
| May 7 2003 11:07:20:263AM | 1030507008 | Steve Wolfe | shot#8: EF3volt down 50V at each point I raised, except down 200 at 20msec C-coil current up to 4000 (demand) at 0.7 sec, flat to 1.3 plasma, with hards. Disrupted at 1.6sec with growing vertcal instab. Definite locked mode from 1.07sec on, as seen on ECE next shot: Start looking for the threshold; I'm going to try stepping it down instead of ramping, since it may be time-dependent in the shot. |
| May 7 2003 11:18:23:110AM | 1030507009 | Steve Wolfe | shot#9: Acoil demand to 3000A EF3V at 20msec from 2450 to 2550 Fizzle! Put the EF3V back I guess. |
| May 7 2003 11:35:02:853AM | 1030507010 | Steve Wolfe | shot#10: Acoil demand to 3000A EF3V at 20msec from 2550 to 2450; while I'm at it, raise Ar puff to 60ms plasma, no hards. locked mode at .82sec, acoil current around 1800A next shot: drop Acoil current by 1000A (demand=2000) |
| May 7 2003 11:42:47:490AM | 1030507011 | Steve Wolfe | shot#11: drop Acoil current by 1000A (demand=2000) fizzle again. What happened to thr robust startup? next shot: drop the gas 2ms |
| May 7 2003 12:01:38:710PM | 1030507012 | Steve Wolfe | shot#12: pg4 to 22msec; drop Acoil current by 1000A (demand=2000) plasma, no hards, no locked mode. I think we've bracketed the threshold between 1800A and 1300A (this control is very non-linear!) next shot: Take one more shot and ramp up and down, look for thresh and hysteresis |
| May 7 2003 12:11:48:270PM | 1030507013 | Steve Wolfe | shot#13: ramp from 2000->3000 (at 1.0s)->2000 (at 1.3s) fizzle. Andy points out that the base pressure has been relatively high (1.5e-6 on all the fizzles) next shot: drop the fill another 2msec |
| May 7 2003 12:26:49:090PM | 1030507014 | Steve Wolfe | shot#14: pg4 to 20msec; ramp from 2000->3000 (at 1.0s)->2000 (at 1.3s) plasma, hards again. looks like a locked node late, around 1.25sec, could be the rampdown. This one is inconclusive, but I'm going to move on to next tap setting |
| May 7 2003 12:50:43:847PM | 1030507015 | Steve Wolfe | shot#15: Tap setting to +Dtop +Dbot -Jtop -Jbot, program to 4000 from .7 to 1.3sec fizzle again,pressure was nice and low. In fact, it bounced and didn't recover; that's a first (for today anyway). next shot: raise pg4 2ms |
| May 7 2003 01:03:07:833PM | 1030507016 | Steve Wolfe | shot#16: pg4 back to 22ms. Tap setting to +Dtop +Dbot -Jtop -Jbot, program to 4000 from .7 to 1.3sec plasma, hards, no locked mode. next shot: reverse polarity to -Dtop -Dbot +Jtop +Jbot |
| May 7 2003 01:29:01:547PM | 1030507017 | Steve Wolfe | shot#17: reverse polarity to -Dtop -Dbot +Jtop +Jbot fizzle. Moved inside! Time to undo the EF3V some more. |
| May 7 2003 01:55:31:117PM | 1030507018 | Steve Wolfe | shot#18:EF3V at 20msec from 2450 to 2350. reverse polarity to -Dtop -Dbot +Jtop +Jbot Plasma, no hards. I'm not positive I applied the above change to ef3V No locked mode. Something funny on bp26de-jk around 1.25sec, but I don't see anything else. No sign of anything on ECE signals. That means that the "dipole" phasing is ineffective in either polarity (at these currents) This would make sense if the m=1 is not effective, since this configuration has ~half the m=2 (but twice the m=1) of the quadrupole, and half current was below threshold for quadrupole in the best polarity. next shot: move on to step V. |
| May 7 2003 02:13:37:017PM | 1030507019 | Steve Wolfe | shot#19:move on to step V: +Btop -Dbot -Gtop +Jbot, helical connection. Note that we will experience spurios pickup signals on the upper Bp loops whose cables pass under the Btop coil for these settings. I'll reload the previous pcs to make sure the EF3V change was applied, although it looks like it was from the startup. plasma, no hards. No locked mode. next shot: reverse polarity |
| May 7 2003 02:58:10:460PM | 1030507020 | Steve Wolfe | shot#20:reverse polarity -Btop +Dbot +Gtop -Jbot There may have been some confusion on the way these have been being hooked up. Somehow, Bill didn't just reverse the power supply leads for this change. It does look like the present setup is correct as stated. The way it is presently set, the + supply goes to -Btop, the +Btop goes to +Dbot, -Dbot goes to +Gtop, -Gtop goes to -Jbot and +Jbot goes to the -supply. Plasma, no hards. No locked mode. Next shot: Actually reverse the supply leads, leaving the intermediates the same. |
| May 7 2003 03:08:14:557PM | 1030507021 | Steve Wolfe | shot#21: Actually reverse the supply leads, leaving the intermediates the same; this is almost certainly going to be the same as 19, but I want to be sure. Plasma, no hards, no locked mode. Maybe I got the helicity wrong. Connection is -supply to -Btop,+Btop to +Dtop, -Dtop to +Gtop, -Gtop to -Jbot, +Jtop to +supply which amounts to +Btop -Dbot -Gtop +Jbot next shot: Move on to #VI, +Btop +Dtop -Gtop -Jbot |
| May 7 2003 03:29:39:477PM | 1030507022 | Steve Wolfe | shot#22: Move on to #VI, +Btop +Dtop -Gtop -Jbot The engineering control panel widget is apparently screwing up the entries for the patch panel. I'll have to go back and put them in by hand and send another message to jas about this. plasma, no hards, no locked mode. the 26 de-jk seems to be ramping from t=0 to the end of the shot, must have a bad connection. next shot: reverse the polarity |
| May 7 2003 03:45:51:910PM | 1030507023 | Steve Wolfe | shot#23: reverse the polarity: -Btop -Dtop +Gtop +Jbot plasma, no hards, no locked mode. We haven't succeeded in making one since shot #10 (maybe 14), with the reverse pol quadrupole. next shot: Go back to condition of #8 -Dtop +Dbot +Jtop -Jbot |
| May 7 2003 04:08:59:800PM | 1030507024 | Steve Wolfe | shot#24: Go back to condition of #8 -Dtop +Dbot +Jtop -Jbot, see if it still works. plasma, no hards, disrupted at 1.55sec locked mode at 1sec. It's back. next shot: turn off C-coil, lower the density |
| May 7 2003 04:24:59:057PM | 1030507025 | Steve Wolfe | shot#25: turn off C-coil, nl4 to 6e19, drop the gas 2msec because of the disruption on 24 p[lasma disrupted at 1.5s, locked mode at 1sec Looks amazingly similar H/D is up a bit (.12 to .1), presumably from the disruption next shot: turn on the acoil in revesed polarity from 24 |
| May 7 2003 04:46:51:483PM | 1030507026 | Steve Wolfe | shot#26: acoil back on, +Dtop -Dbot -Jtop +Jbot plasma, no hards, interferometer went berserk, but feedback ok No locked mode, except maybe at the end It may have worked! Next shot: Extend acoil pulse to 1.5sec |
| May 7 2003 05:00:18:043PM | 1030507027 | Steve Wolfe | shot#27: Extend acoil pulse to 1.5sec, still +Dtop -Dbot -Jtop +Jbot plasma, hards are back, disrupted at 1.76sec, no locked mode. I think it works. |
| Engineering Operator Comments | ||||
| Shot | Time | Type | Status | Comment |
| 1 | 08:59:28:210AM | Plasma | Ok | |
| 2 | 09:22:06:590AM | Plasma | Ok | |
| 3 | 09:41:18:790AM | Plasma | Ok | |
| 4 | 09:56:26:327AM | Plasma | Ok | |
| 5 | 10:09:10:003AM | Plasma | Ok | |
| 6 | 10:35:35:653AM | Plasma | Ok | |
| 7 | 10:48:43:830AM | Plasma | Ok | |
| 8 | 11:01:27:943AM | Plasma | Ok | |
| 9 | 11:14:24:637AM | Plasma | Ok | |
| 10 | 11:27:21:547AM | Plasma | Ok | |
| 11 | 11:40:06:863AM | Plasma | Ok | |
| 12 | 11:53:18:757AM | Plasma | Ok | |
| 13 | 12:07:30:373PM | Plasma | Ok | |
| 14 | 12:22:12:507PM | Plasma | Ok | |
| 15 | 12:44:02:960PM | Plasma | Ok | |
| 16 | 12:56:51:450PM | Plasma | Ok | |
| 17 | 01:24:31:803PM | Plasma | Ok | |
| 18 | 01:37:11:650PM | Plasma | Ok | |
| 19 | 02:07:58:947PM | Plasma | Ok | |
| 20 | 02:42:04:233PM | Plasma | Ok | |
| 21 | 02:59:57:433PM | Plasma | Ok | |
| 22 | 03:23:29:140PM | Plasma | Ok | |
| 23 | 03:38:55:523PM | Plasma | Ok | |
| 24 | 04:04:36:667PM | Plasma | Ok | |
| 25 | 04:17:53:290PM | Plasma | Ok | |
| 26 | 04:39:46:137PM | Plasma | Ok | |
| 27 | 04:56:07:490PM | Plasma | Ok | |