Miniproposals

Miniproposal: Date Filed: Title: First Author: Session Leader: Steve Horne (shots 1-32)

Engineering Operator Run Comment

Cont. Restart of Machine.

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Entered: Jul 7 2004 03:49:09:993PM

Author: To Be Determined

Run report, 940518 Horne (SL), Granetz (Phys op.) A somewhat frustrating day. Some minor progress in MP 41, but most of the day was spent fighting breakdown problems. Breakdown seems not to be reproducible. The mflux and vacuum field plots imply that there's a bit of a gradient problem, but in general there seemed not to be a good correlation between the perceived goodness of the fields and the plasma breakdown. We need to take a close look at coil current signals to see if offsets are drifting over time. Possibly we're still dirty -- remember we broke vacuum Monday. Engineering systems worked well; the only shot lost due to power supply problems was 22; it has difficulty following when the plasma does a vertical thrash. The combination of the large fraction of duds and fizzles, the fizzle detector, improved TF cooling, and a change in the TF cooling requirements conspired to give us 32 shots today -- average of 15 min/shot. 1 fill at 1.2e-5, load last shot from yesterday. fizzle. Fields aren't obviously bad. 2 fill to 1.0e-5. Fizzle. 3 bz looks a bit too negative. Try +30 A (2 mt) ef4 tweak. Worked. Lost at 400 msec or so due perhaps to too much gas? drop puff programming for next shot. 4 dud. 5 dud . For next shot, fill to 1.2e-5. Also reduce P and D gain on slow Z after 100 msec. Integral gain is working; see shot 4 drop ef4 by 15 a 6 fizzle. There's a suspicous wiggle in bz just before the plasma dies. It's coincident with turnon of rcur proportional gain (10 msec). Reduce gain there. 7 Fizzle. 40 kA current, lost it at about 22 msec. Again, rcur/bz seems implicated. Drop derivative gain on rcur from 2 to 1 from 20 msec to 40 msec. 8 400 msec plasma. increase fill from 1.0e-5 to 1.1 e-5 9 fizzle 10 dud. for 11, raise br offset by 1 millitesla. Objective is to reduce imbalance in oh2 coils. 11 fizzle; null looks better but perhaps 1 millitesla was too much. 12 back off 1/2 millitesla on br, also make ef4 15 A more negative. 13 dud. another 15 A neg on ef4 14 fizzle. Late breakdown. Mfil says flux pattern has changed. Might be the ef4 tweak, or effect of turning f25 back on in tree. 15 80 ka fizzle. plasma seems to be going up. 16 40 ka fizzle. radial field coming in at around 10 msec. 17 fizzle, tweak br 18 fizzle. 19 Fizzle. Null looks good, but we seem to be loosing the plasma radially out. ef3 is maxed out; so try 15 A more + on ef4. 20 That did it. Plasma lasts to 500+ milliseconds. 21 Repeat, and it did. 520 msec. 22 22 vertical thrash at .04-.06 killed oh2l. 23 another good one. 24 try turning on gap control by adding a point at (.35,2) Turn on argon puff Fizzle. Try to repeat. 25 Fizzle again. Turn off the argon., increase br by .001. (-1.e-3) 26 Better breakdown, but plasm goes screaming up -- drop to -2.5e-3 27 worked. raise gain on clearin (.35,3) 28 dud! 29 fizzle. Try .5 mt more + br. 30 dud, with nice null. Make br more neg by 1 mt. 31 Plasma! try to improve x-point by making ef1l current more negative. 32 dud.

Physics Operator Summaries

Entered: Jul 7 2004 04:36:51:397PM

Author: To Be Determined

From: PHYSOP::MARMAR Date: 19-MAY-1994 22:06:13 Description: Physics Operator's summary, 940519 The run started reasonably well, with shot 2 reaching .8 MA. After shot 3, it was discovered that 3 SCR's and fuses on EF2L were blown. It took about 1.5 hours to fix this. Looking back at old data, Vinny concluded that we had been running with these blown fuses almost since start up at the end of April. Shots 4 through 8 were all plasmas, with generally increasing length, and various tweaks were made in the attempt to improve plasma shape and x-point control. The next decent shot was 14, and later we noticed that the plasma vertical position had moved down compared to that on shot 8, with no requested change in the ZCUR programming. This persisted for the remainder of the run, and Steve Wolfe is tentatively blaming more bad Bp coils (see his session leader summary). Perhaps we should consider thumb screws. The only other decent shots in the rest of the run were 21 and 23. Shot 20 was lost because the engineering tree had been turned off, so no power supplies pulsed. Scorecard 9 plasmas (>.5 MA) 3 early deaths (.1 MA < I < .5 MA) 7 fizzles 3 duds 1 no power supplies

Session Leader Comments

Physics Operator Comments
May 18 1994 09:03:14:040AM	940518001	Robert Granetz	Startup where we left off yesterday. Pre-fill at 1.2e-5.
May 18 1994 09:16:54:110AM	940518001	Robert Granetz	Fizzle. Breakdown at 6-7 ms. Next shot, reduce pre-fill from 1.2 to 1.0e-5.
May 18 1994 09:43:41:600AM	940518002	Robert Granetz	Fizzle. Late breakdown. Bz looks to be about -2 mT. Next shot, increase I_ef4 by 30 amps.
May 18 1994 09:56:37:700AM	940518003	Robert Granetz	Plasma. Ipmax=0.58 MA, duration=0.37 s. Density was rather high, and I think that it caused the plasma to start disrupting at 0.33 s. Next shot, reduce pulse gas a little.
May 18 1994 09:56:57:380AM	940518003	Robert Granetz	Null really doesn't look too good.
May 18 1994 10:09:34:660AM	940518004	Robert Granetz	Dud. Null is not too good, but it looks just like the previous (plasma) shot. Next shot, increase pre-fill from 1.0 to 1.2e-5
May 18 1994 10:14:39:990AM	940518004	Robert Granetz	Also for next shot, reduce P&D gains for slow ZCUR starting at 0.1 s.
May 18 1994 10:31:30:930AM	940518005	Robert Granetz	Dud. Next shot, reduce EF4 by 15 amps (1 mT). We don't understand why the last two shots were duds.
May 18 1994 10:50:55:050AM	940518006	Robert Granetz	Fizzle. I still don't think the null looks good enough. For next shot, reduce initial RCUR P gain from 2 to 1 (starting at 0.010 s).
May 18 1994 11:06:04:700AM	940518007	Robert Granetz	Fizzle. Next shot, reduce pre-fill from 1.2 to 1.0e-5, and reduce the early RCUR D gain from 2 to 1.
May 18 1994 11:25:10:120AM	940518008	Robert Granetz	Plasma. Ip=0.6 MA, duration=0.45 s. Next shot, increase pre-fill from 1.0 to 1.1e-5. No other changes.
May 18 1994 11:41:17:490AM	940518009	Robert Granetz	Fizzle. We can't seem to get two shots in a row. Next shot, no changes.
May 18 1994 12:01:27:850PM	940518010	Robert Granetz	Dud. Null looks worse. With 1 mWb contour spacing, there doesn't even seem to be a null. Next shot, increase BR by 1 mT.
May 18 1994 12:40:54:170PM	940518011	Robert Granetz	Dud. Next shot, reduce BZ by 1 mT and increase pre-fill from 1.1 to 1.4e-5.
May 18 1994 12:27:32:890PM	940518012	Robert Granetz	Fizzle. Bz is better. There is a bit of BR though. For the next shot, reduce the BR demand by 0.5 mT.
May 18 1994 12:43:17:960PM	940518013	Robert Granetz	Dud. Next shot, reduce EF4 by 15 amps again (change BZ by -1 mT).
May 18 1994 12:57:53:630PM	940518014	Robert Granetz	Fizzle. Breakdown at 10-11 ms. Next shot, reduce pre-fill from 1.4 to 1.2e-5.
May 18 1994 01:15:28:350PM	940518015	Robert Granetz	Fizzle, but a little bit of current. There is a teeny-weeny signal on the x-ray array 2, and it's very high up in the vessel. Next shot, no changes.
May 18 1994 01:36:46:590PM	940518016	Robert Granetz	Fizzle, very similar to the previous shot. Apparently the null has a significant BR after breakdown, and this is throwing the plasma upwards. Next shot, reduce BR P gain from 5 to 2.5 after 0.010 s.
May 18 1994 01:54:52:140PM	940518017	Robert Granetz	Fizzle, late. Next shot, decrease BR by 1 mT using minus_one_volt coeff.
May 18 1994 02:08:51:480PM	940518018	Robert Granetz	Fizzle, with a bit of current. Next shot, increase RCUR gain from 1 to 2 after t=0.010 s.
May 18 1994 02:26:52:420PM	940518019	Robert Granetz	Fizzle. Null is looking better, but early position control is not good enough. Next shot, increase EF4 by 15 amps (BZ by 1 mT).
May 18 1994 02:41:57:070PM	940518020	Robert Granetz	Plasma! Finally! Ip=0.6 MA, duration=0.53 s. Vertical position control is awful. Density is good. Next shot, absolutely no changes.
May 18 1994 02:57:24:430PM	940518021	Robert Granetz	Plasma. Two in a row!. Still ends in an upward VDE, just like the last shot. Next shot, no changes. Try for three in a row.
May 18 1994 03:12:26:540PM	940518022	Robert Granetz	Plasma, but short. Ipmax=0.47 MA, duration=0.16 s. OH2's faulted, because feedback drove it crazy. For some reason, the plasma vertical position was oscillating early on much more than the previous two shots. Next shot, still no changes.
May 18 1994 03:21:22:670PM	940518023	Robert Granetz	Plasma. Ip=0.6 MA, duration=0.47 s. Downward VDE. Density has been slowly dropping over the last 4 shots. Next shot, increase pulse gas. Also turn on John Rice's valve for argon puffing.
May 18 1994 03:32:19:010PM	940518024	Robert Granetz	Fizzle. Null is slightly different. Next shot, no changes.
May 18 1994 03:43:06:190PM	940518025	Robert Granetz	Fizzle. Next shot, increase BR by 1 mT using minus_one_volt coeff.
May 18 1994 03:54:01:780PM	940518026	Robert Granetz	Fizzle. Proto-plasma goes shooting up. Obviously the +1 mT change to BR was the wrong thing to do. Next shot, lower BR by 1.5 mT from what it was on this shot (i.e. decrease minus_one_volt coefficient from -1.0e-3 to -2.5e-3. During the preceding series of four good plasma shots, this coeff was at -2.0e-3.
May 18 1994 04:07:55:520PM	940518027	Robert Granetz	Plasma. Ip=0.6 MA, duration=0.483 s. Density is still a little low. Next shot, raise initial CLEARIN P gain from 2 to 3 at 0.35 s.
May 18 1994 04:19:19:160PM	940518028	Robert Granetz	Dud. Next shot, change EF1 current demand by zeroing it out after 0.2 s.
May 18 1994 04:31:38:250PM	940518029	Robert Granetz	Fizzle. Proto-plasma went shooting upwards. Next shot, raise BR demand by +0.5 mT using the minus_one_volt coefficient.
May 18 1994 04:42:11:900PM	940518030	Robert Granetz	Dud. Ian's MFLUX routine gives a fairly good looking null, so perhaps it has a problem. Next shot, lower BR by 1 mT (change minus_one_volt coeff from -2.e-3 to -3.e-3).
May 18 1994 04:57:30:200PM	940518031	Robert Granetz	Plasma. Ip=0.6 MA, duration=0.41 s. Upward VDE. Next shot, program EF1L demand to keep its current higher, which will pull more on the plasma, and turned off all feedback gain on RXL.
May 18 1994 05:06:53:910PM	940518032	Robert Granetz	Dud

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	09:02:27:860AM	Plasma	Ok
2	09:23:59:590AM	Plasma	Bad
3	09:39:43:950AM	Plasma	Ok	580 ka
4	09:57:26:240AM	Plasma	Bad
5	10:15:52:960AM	Plasma	Ok
6	10:32:35:840AM	Plasma	Ok
7	10:51:14:970AM	Plasma	Ok
8	11:06:54:120AM	Plasma	Ok	600 ka
9	11:25:09:250AM	Plasma	Ok	no faults
10	11:36:38:310AM	Plasma	Ok	No plasma
11	12:02:27:280PM	Plasma	Ok	no faults
12	12:16:49:140PM	Plasma	Ok	no faults
13	12:28:42:880PM	Plasma	Ok	no faults
14	12:43:57:830PM	Plasma	Ok	no faults
15	12:59:12:420PM	Plasma	Ok	no faults
16	01:16:39:270PM	Plasma	Ok	Dud. No faults.
17	01:38:13:520PM	Plasma	Ok	Dud. No Faults. TF inversion much faster now.
18	01:56:07:320PM	Plasma	Ok	180 kA plasma. Disruption caused OH2L comm. fault.
19	02:09:56:820PM	Plasma	Ok	Fizzle early. No faults.
20	02:27:51:540PM	Plasma	Ok	600 kA plasma disruption at 0.5 second. No faults.
21	02:43:41:800PM	Plasma	Ok	Nasty VDE 600 kA. OH2L commutation fault after; reason still not clear.
22	02:58:18:030PM	Plasma	Ok	Disruption during current rise at 475 kA
23	03:09:30:280PM	Plasma	Ok	580 kA disruption at .47 sec. No faults.
24	03:22:47:830PM	Plasma	Ok	Late breakdown; fizzle
25	03:32:55:910PM	Plasma	Ok	Argon induced fizzle?
26	03:44:32:940PM	Plasma	Ok	Another fizzle
27	03:54:38:490PM	Plasma	Ok	Another 600 kA disruption at 0.5 sec
28	04:09:14:870PM	Plasma	Ok	Dud. the power system works flawlessly when there is no plasma!
29	04:22:21:210PM	Plasma	Ok	Dud. This is boring. Going to increase TF inversion voltage for next s
30	04:33:49:450PM	Plasma	Ok	Another Dud. Saved 0.15 seconds during TF inversion.
31	04:44:59:270PM	Plasma	Ok	600 kA disruption at 0.4 seconds
32	04:56:53:880PM	Plasma	Ok	dud!