Miniproposals

Miniproposal: 053 Date Filed: 4/8/1994 Title: Study of Closed Loop Response of Elongated Plasmas. First Author: Gerry Tinios Session Leader: Gerry Tinios (shots 1-18)

Engineering Operator Run Comment

closed loop response

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Entered: Jul 7 2004 03:49:11:257PM

Author: To Be Determined

Session leader's summary for run 940624 SL: G. Tinios PO: S. Horne EO: V. Bertolino/J. Daigle This run is dedicated to MP#53, "Closed Loop Response of Elongated Plasmas". The object was to intorduce square steps in ZCUR programming and watch the plasma relax to a new equilibrium state after each step. We planned to do this at various feedback gains. We started with shot 940623028, which proved to be very robust yesterday. For fear of disruptions, we reduced the plasma current to 600 kA. We then introduced four 1-cm steps spaced 60 msec apart. The first such shot ran well, so we became adventurous and increased the plasma current to 800 kA. This reduced the background Z-oscillations, as was expected, but it also resulted in a bad downward disruption at the last downward step, so the rest of the run was carried out at 600 kA. The base shot seemed not to be always reproducible, sometimes disrupting before we got to the ZCUR steps. This seems to be related to the upper X-point, which exists until about .6 sec and may introduce impurities. We did not get to experiment with the gains much, but to the extent that we did we saw the following: -- Derivative gain on the fast ZCUR wire seems to reduce OH-related oscillations. -- Proportional gain on the slow ZCUR wire also seems to enhance OH-related oscillations as was expected. -- Proportional gain on the fast ZCUR wire seems to enhance the oscillations. Score: Shots: 18 Plasma: 17 Fizzles: 0 Duds: 1 Upward disruptions: 2 Downward disruptions: 9 Here are the details: 01 Fiducial shot (940603015) --> upward disruption (30 kA halo) 02 Reload 940623028 with current reduced to 600 kA , prefill=1.2e-5 Density feedback on. --> Nice rampdown. 03 Introduce 60msec/1cm steps into ZCUR programming. --> Nice rampdown. Steps are clearly visible. 04 Same shot as the previous one, but increase Ip to 800kA. It hoped that this will give a cleaner Z trace. --> Bad downward disruption (150 kA) on the last downward step. 05 Change Ip back to 600 kA. Eliminate the last downward step. Raise the gain on wire 7 to 4. --> Good rampdown. Overshoot in Z- steps only marginally reduced. 06 Raise the initial flattop Zcur programming to -1.7 cm so we can squeeze in one more downward step. Drop the derivative gain on wire 3 to 5. --> Downward disruption (80 kA). Looks like we need more (not less) derivative gain. Z-oscillations are considerable. 07 Wire 7: drop proportional gain to 2, raise derivative gain to 2 and later to 5. --> downward disruption (70 kA) Z-oscillations are reduced, fast-Z derivative gain seems to help in this respect. 08 Wire 2: drop p-gain and d-gain to 0.5 and raise i-gain to 4. Wire 7: d-gain set at 5 all through the shot after .5 sec Go to the ZCUR programming as in shot 05, flattop at -2.7 before starting the steps. --> Downward disruption (80 kA). We seem to have bad slow vertical control during lowering of the plasma. 09 Reload shot 5 and reduce p-gain on wire 2 to 0.5 after 0.7 sec. --> Good rampdown. Oscillations seem to die out after 0.7 sec. 10 Wire 2: reduce d-gain to 0.5 after 0.7 sec. --> Upward disruption (40 kA). Lost it on a downward going step. 11 Wire 2: increase i-gain to 5 after 0.7 sec. --> Downward disruption (60 kA). We lose it at the same time as the previous shot. 12 Wire 2: bring d-gain back to 1. Wire 7: raise p-gain to 8 at 0.7 sec. --> Downward disruption (70 kA). Lost it at the first upward step. 13 Reload shot 09. -->Downward disruption (70 kA). Lost it before reaching flattop. 14 Turn ZXL wire at 0.5 sec hoping it will get rid of the upper x-point earlier. -->Better, nice rampdown. 15 Drop p-gain on wire 7 to 0 at 0.7 sec. -->Lost it at 0.56 16 Reload shot 05. --> Downward diruption (70 kA). Lost it at 0.55 sec. This shot is not reproducible. 17 Start CLEARIN control earlier. --> Ran to the end. Chopper demand is maxxed out. 18 Start CLEARIN control even earlier. Drop p-gain on wire 7 to 0 at 0.7sec. -->Dud.

Physics Operator Summaries

Entered: Jul 7 2004 04:36:51:990PM

Author: To Be Determined

Physics Op report, 940624 Tinios, Horne, Bertolino, Daigle. mp 53 closed loop response . The objective was to program small vertical steps in plasma position, and observe the response of the plasma at different vertical gains. Most of the run was at 600 ka, in deference to our worries about disruption currents. The single shot we tried at 800 kA ended at a 150 kA disruption, so we retreated back to 600 ka and removed the last downward step from the Z programming.. We did have several disruptions at ip=600 ka, but typical currents were 70 kA or less. There is a minor problem in the development of these discharges arount .5-.6 sec, when an incipient upper x-point appears. The run ended early when a problem with oh2l developed. 1 fiducial shot ok (disrupted up,30 kA on halo rogowski). 2 load 940623028, drop IP to 600 ka. nl feedback should be on at 5.e19. 3 Gerry's steps. 1 cm excursions up, no gain changes. Beautiful steps 4 Raise ip to 800 ka. tf crowbar to 1.1 sec to save cooling. Fizzle detector off time at 1.5 (why was it set to 1.2?) Died at ~ 1 second. 150 ka disruption on last down step. 5 drop ip to 600 ka. Remove last downward glitch. Raise fast zcur P on wire 7 from 2 to 4 (add this to wire 3 for full chopper gain. Complete rampdown. Nice steps. 6 Add a step earlier to diagnose the downward vertical overshoot. Drop derivative gain (wire 3) on second step from 9 to 5. Lost on second down step, 70 ka disruption. Possibly due to reduced deriv gain. 7 drop P wire 7 back to 2. Rais D to 2 at .5,5 at .78 or so Died at .69 8 went back to steps as in 5. Raise D on wire 7 to 5 at .5 Drop P on wire 2 from 1 to .5 Drop D on 2 from 1 to .5 Raise I on wire 2 from 3 to 4. Crash and burn. Dropping the P gain this early was a bad idea; the oh2's can't fight the current programming and the plasma was lost up. 9 Reload shot 5, but drop wire 2 P from 1 to .5 at .7 sec. Objective is to deal with the quivering. Full shot. 10 drop slow D gain at .7 to .5 Set integral gain at 2 on nl feedback, unshort integrator at .5 sec. died around .8 sec. 11 raise I gain on wire 2 from 3 to 5 at .7. Drop p gain on nl, raise I gain from 2 to 4. Died around .8 sec. 12 slow Z (wire 2) D back to 1. wire 7 P from 4 to 8 at .7 Lost at first glitch. 13 Remove slow Z I change at .7 sec Remove step in fast Z at 700. Drop NL P from 4 to 2. Died due to crud at .58 14 Turn on zxl at .5 instead of .6, low gain. Full shot, nice rampdown. 15 turned off p gain on wire 7 at .7 Died at .5+. 16 Reload shot 5. Disruption near .5 17 turn clearin on earlier. Got the rampdown!. 18 nl Integral gain back on at 5, reset at .4 sec, p gain at .2 clearin on even earlier, p gain on 7 to zero at .7 Dud, oh2, ef2 seem unhappy during precharge.

Session Leader Comments
Jun 24 1994 02:50:18:570PM		Gerry Tinios	SL: G. Tinios PO: S. Horne EO: V. Bertolino/J. Daigle This run is dedicated to MP#53, "Closed Loop Response of Elongated Plasmas".
Jun 24 1994 02:51:37:470PM	940624001	Gerry Tinios	Fiducial shot (940603015) --> upward disruption (30 kA halo)
Jun 24 1994 02:53:09:010PM	940624002	Gerry Tinios	Reload 940623028 with current reduced to 600 kA , prefill=1.2e-5 Density feedback on. --> Nice rampdown.
Jun 24 1994 02:53:53:930PM	940624003	Gerry Tinios	Introduce 60msec/1cm steps into ZCUR programming. --> Nice rampdown. Steps are clearly visible.
Jun 24 1994 02:54:28:220PM	940624004	Gerry Tinios	Same shot as the previous one, but increase Ip to 800kA. It hoped that this will give a cleaner Z trace. --> Bad downward disruption (150 kA) on the last downward step.
Jun 24 1994 02:55:29:380PM	940624005	Gerry Tinios	Change Ip back to 600 kA. Eliminate the last downward step. Raise the gain on wire 7 to 4. --> Good rampdown. Overshoot in Z- steps only marginally reduced.
Jun 24 1994 02:56:49:780PM	940624006	Gerry Tinios	Raise the initial flattop Zcur programming to -1.7 cm so we can squeeze in one more downward step. Drop the derivative gain on wire 3 to 5. --> Downward disruption (80 kA). Looks like we need more (not less) derivative gain. Z-oscillations are considerable.
Jun 24 1994 02:57:31:070PM	940624007	Gerry Tinios	Wire 7: drop proportional gain to 2, raise derivative gain to 2 and later to 5. --> downward disruption (70 kA) Z-oscillations are reduced, fast-Z derivative gain seems to help in this respect.
Jun 24 1994 02:58:34:030PM	940624008	Gerry Tinios	Wire 2: drop p-gain and d-gain to 0.5 and raise i-gain to 4. Wire 7: d-gain set at 5 all through the shot after .5 sec Go to the ZCUR programming as in shot 05, flattop at -2.7 before starting the steps. --> Downward disruption (80 kA). We seem to have bad slow vertical control during lowering of the plasma.
Jun 24 1994 02:58:59:760PM	940624009	Gerry Tinios	Reload shot 5 and reduce p-gain on wire 2 to 0.5 after 0.7 sec. --> Good rampdown. Oscillations seem to die out after 0.7 sec.
Jun 24 1994 02:59:26:700PM	940624010	Gerry Tinios	Wire 2: reduce d-gain to 0.5 after 0.7 sec. --> Upward disruption (40 kA). Lost it on a downward going step.
Jun 24 1994 03:01:15:370PM	940624011	Gerry Tinios	Wire 2: increase i-gain to 5 after 0.7 sec. --> Downward disruption (60 kA). We lose it at the same time as the previous shot.
Jun 24 1994 03:01:42:810PM	940624012	Gerry Tinios	Wire 2: bring d-gain back to 1. Wire 7: raise p-gain to 8 at 0.7 sec. --> Downward disruption (70 kA). Lost it at the first upward step.
Jun 24 1994 03:06:17:200PM	940624013	Gerry Tinios	Reload shot 09. -->Downward disruption (70 kA). Lost it before reaching flattop.
Jun 24 1994 03:07:21:960PM	940624014	Gerry Tinios	Turn ZXL wire at 0.5 sec hoping it will get rid of the upper x-point earlier. -->Better, nice rampdown.
Jun 24 1994 03:07:37:380PM	940624015	Gerry Tinios	Drop p-gain on wire 7 to 0 at 0.7 sec. -->Lost it at 0.56
Jun 24 1994 03:08:29:920PM	940624016	Gerry Tinios	Reload shot 05. --> Downward diruption (70 kA). Lost it at 0.55 sec. This shot is not reproducible.
Jun 24 1994 03:09:14:710PM	940624017	Gerry Tinios	Start CLEARIN control earlier. --> Ran to the end. Chopper demand is maxxed out.
Jun 24 1994 03:09:45:760PM	940624018	Gerry Tinios	Start CLEARIN control even earlier. Drop p-gain on wire 7 to 0 at 0.7sec. -->Dud.

Physics Operator Comments

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	08:33:55:800AM	Plasma	Ok	no faults
2	08:47:23:700AM	Plasma	Ok	no fault
3	09:05:54:590AM	Plasma	Ok	no faults
4	09:25:51:080AM	Plasma	Ok	comm faults after disruption ef1 and oh1
5	09:44:39:390AM	Plasma	Ok	no faults
6	10:07:34:000AM	Plasma	Ok	ef1l comm fault after disruption
7	10:32:47:190AM	Plasma	Ok	no faults
8	10:54:08:400AM	Plasma	Ok	no faults
9	11:07:27:600AM	Plasma	Ok	no faults
10	11:40:42:310AM	Plasma	Ok	no faults
11	11:55:43:240AM	Plasma	Ok	no faults
12	12:12:53:160PM	Plasma	Ok	no faults
13	12:28:38:790PM	Plasma	Ok	oh2u comm fault after disruption
14	12:46:18:320PM	Plasma	Ok	No faults.
15	01:06:48:850PM	Plasma	Ok	No faults.
16	01:19:25:700PM	Plasma	Ok	OH2 U Comm Fault.
17	01:31:04:110PM	Plasma	Ok	No Faults.
18	01:53:33:020PM	Plasma	Bad	OH2L opened at -0.265 seconds; run over