Miniproposals

Miniproposal: Date Filed: Title: First Author: Session Leader: Steve Golovato (shots 1-27)

Engineering Operator Run Comment

RF Conditioning

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Entered: Jul 7 2004 03:49:09:007PM

Author: To Be Determined

Run Summary - 930928 - Session Leader SL - Golovato PO - Goetz EO - Bertolino Pre_run Comments ---------------- Goals - (1) Produce reproducible deuterium discharges. (2) Bring up ICRF power slowly to avoid disruptions and the accompanying fizzles. (3) Add hydrogen to get some core absorption, which hopefully reduces edge effects which lead to impurity influx, disruptions. (4) Match plasma shape to antenna, move antenna and limiter closer, scan density, plasma radial position, H minority. Starting with last shot of last friday, 930928030, except moving plasma to -2cm from -3cm, added more late gas to raise density (avoid shark fins), and adding rf. rf Limiter at 92cm, antenna at 93cm. 1 -- Dud. Setting up rf for several shorter pulses instead of one long one, to hopefully reduce risk of disruptions, also to see effects of turnon and turnoff more clearly. 2 -- rf still in single pulse, next shot will have three ~100ms pulses between 300-750ms. Fizzled. 3 -- Fizzled. 4 -- rf in three pulses for this shot. Dudded. 5 -- Fizzled. 6 -- Fizzled, but almost, like shots 2 and 3. 7 -- Plasma, 400ka. Small plasma, to outside, no marfe, density ramped up to disruption at 375ms. rf interrupted during plasma, ran without plasma. Tuning off. Lowered power of first rf pulse so it won't interrupt with plasma and we can get tuning measurement. 8 -- Similar to 7, disrupted vertically at at 700ms. n=3e20 at 500ms. 9 -- Fizzled. Standard fizzle after disruption. 10 -- Fizzled. Too much too much. 11 -- Almost, disrupted up and in at 120ka, 53ms. 12 -- First good plasma. Density lower, marfed. 13 -- Pretty good but disrupted at 900ms, finned to death (quoth goetz). 14 -- Good one. rf tuning progressing slowly. 15 -- Fizzled. No apparent reason. Just try again. 16 -- Fizzled. Better do something. 17 -- Good plasma, no disruptions. Lowered fill to get it. Also kept current up, positive current ramp (or at least flat) avoids shark fins (or so the lore has it). 18 -- Good One. 19 -- Fiddling and diddling with hybrid gains to cure z_cur shakes during breakdown. Good plasma. Worked somewhat. 20 -- Good again. Tuning progressing. When the tuning gets better, we'll add some H and start raising power more seriously. So far we've only raised it a little. 21 -- Adding 10ms, 100v H puff at 400ms. Tuning pretty good, up to 100 kw. Still playing with breakdown wiggles. Added gas after 200ms to up density, reduce hards which are just beginning to appear. Density has been falling over last few shots. Plasma disrupted, first at 550ms, lost at 600ms. 22 -- H puff widened to 12ms. Added more gas to raise density. Low density and higher rf power lead to shark fin, disruption. Making more, shorter rf pulses to avoid disrupting, continuing to increase power. Tuning now good. Dud. Dud. Dud. Should have raised the fill like Goetz wanted. 23 -- Raise the fill. Ran but disrupted again at 600 ms. 24 -- Raised the fill again, increased Ip to 600ka and lowered rf power. Ran but disrupted again at 600 ms. 25 -- Adding second H puff at 500ms, raised fill, lowered later rf pulses to avoid disruption. Also raised Ip more and earlier. Fizzled. Got several rf plasmas. Too much fill. 26 -- Lowered fill. Moved limiter and antenna in one cm each. Maybe getting closer to fast wave cutoff layer will reduce edge interaction which leads to impurity influx. Fizzled. 27 -- Lowered fill again. Fizzled! Aghhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh!!! Post-run Comments ----------------- Shot accounting: duds 3 fizzles 8 engr failures 0 almosts 4 plasma 12 We struggled with the disruption-fizzle syndrome most of the day. This is exacerbated by running with rf since the rf often causes disruptions by raising the radiated power until the plasm detaches. This effect looks like a sharkfin on the standard sharkfin diagnostics, that is, the signal rises then falls abruptly at the disruption. Whether it is an official sharkfin or just looks like one I leave to the ichthyologists in the group. A rule of thumb (according to Earl) is more than 500 kw radiated with 500 ka (rouhgly the criterion for detachment) causes a disruption. This was roughly true for this run. We tried to raise the current near the end of the run to raise this limit, but we were stuck in a fizzle rut and never got plasmas at higher current. The other way we tried to avoid rf-driven disruptions was to run the rf in several shorter pulses (50-150 ms each) to limit the radiated power rise and to allow recovery between rf pulses. This worked somewhat, but there was some ratcheting, typically leading to disruptions around 600 ms (see shots 21, 23, 24). As long as the power stayed below about 100 kw, we avoided disruptions. Pulsing the rf also helped pick out rf-induced effects on diagnostic signals. We were ale to see titanium influxes caused by rf, presumably from the TiC coated antenna shield and/or tiles. This was seen on the McPherson. We also tried puffing hydrogen, to increase core absorption, and moving the antenna closer to the plasma, to improve coupling to the fast wave. These were done late in the run and we didn't get good plasma shots to see if they helped the impurity influx problem. In making good plasmas, we seem to have problems in two regions of the shot. First is the disruption-fizzle syndrome, which is a breakdown current ramp problem. Hopefully this will be improved or cured this Friday by discharge cleaning after disruptions. I believe this is standard procedure on many experiments, JT-60U does a half hour of discharge cleaning after a disruption. The other problem is that adding gas later in the shot to avoid the marfe and get higher densities often leads to detachment and disruption (see shots 7, 8, 11). Getting the current up can help this. The key question for the rf is what was different last month when we put in 250 kw without significant impurity influx. The antenna and limiter are at the same positions but we are running fatter plasmas which nearly limit on the outboard limiter. This may be one problem, an unconditioned antenna may be another. We probably should go back to these plasmas, which are farther from the antenna but match its shape better. Since these plasmas are more elongated, we risk vertical disruptions. Since we will be running such plasmas on Friday, maybe the nice guys in charge of the run will let us put rf into some of the shots. The worst thing that can happen is a disruption, which they can practice recovering from. One thing to remember is that the fixed antennas and limiters to be installed for next year are even closer to the plasma, even after being moved back 2 cm. We will have to learn how to run with the limiter and antenna closer, and the plasma more elongated. There is a strong tendency on non-rf runs to move the antenna and limiter back and make less elongated plasmas to get more good shots. This may hurt us later since we are avoiding learning to run the strongly rf-heated plasmas on which all of our future experiments are contingent.

Physics Operator Summaries

Entered: Jul 7 2004 04:36:50:380PM

Author: To Be Determined

Physics operator summary for 930928 EO: Bertolino PO: Goetz SL: Golovato The goal of Run 930928 was to obtain reliable breakdowns with deuterium fill and fueling gas and to condition the Rf antenna. The plan for the day was to start with the last discharge from Run 930924 and fiddle to the RF needs. and fueling pulse, and get an acceptable current rise. The run started with a reload of shot 930924030, which was the last shot of the previous run (dedicated to deuterium breakdown studies). Changes to this shot included extending the IC_EF4 waveform to .8 s and drawing ZCUR at -.02 m. The beginning of the run was plagued with < 50 kA plasmas (not fizzles). These plasmas were bouncing off the outer limiter (up to four times). This behavior was due to the radial position predictor calculating that the plasma was inside (while UV imaging shows it centered->outside) and moving it out. This problem was never solved but the effects were softened by raising the fill pressure and programming the plasma to be inside and then moving out after the plasma current had risen. Also, early in the run there were no MARFE's and the density rise was enormous leading to disruptions. By shot 012, shark fins were occurring and radiative disruptions were the norm. From shot 016 to 021 good (although dirty) plasmas were obtained. This was accomplished by lowering the fill to 1.1e-5 gauge. Throughout this series of shots the density was falling and repeated attempts to raise it were unsuccessful. The RF injected power was also increased during these shots with a lot of Mo being injected. The fill needed to be increased to 1.3e-5 gauge after a dud on shot 022. Two more RF-induced-shark-fin disruptions followed. These were followed by three fizzles in which the field structure was more than adequate for a good plasma. Once again the operators are left with the nagging question of whether the problem is fields, fill, or cleanliness. Notes: Some fine-tuning was needed to the V_OH1 to smooth out the Bz ramp. Some twiddling of the offset on Br_0 was done but once again it ended up at the nominal value of +.009. RCUR needs to be fixed during the startup phase. Engineering today: Reliable operation at 5.3T. Problems with OH1 and EF1's reset after Type I faults (Bertolino and Daigle investigating). Scorecard: 27 shots = 12 plasmas + 4 early deaths + 8 fizzles + 3 duds Shot summary for Run 930928 Reload of 930924030 Changes: IC_EF4 -1000V extended to .8s S_ZCUR = -.02 1 P_0=1.0e-5 dud 2 P_0=1.2e-5 bad plasma 3 move PULSE_GAS_3 3 msec later (now @ -.018) bad plasma 4 IC_EF4 to 1110A (was 1160) dud 5 IC_EF4 to 1135A M_ZCUR(3) on @ .9 S_RCUR =.67 @ 0.; .68 @ .1 fizzle 6 IC_EF4 to 1150A Br_0 offset to .009 PULSE_GAS_3 waveform 3 msec later (now @ -.015) bad plasma 7 stairstep RCUR gain 2@.2; 4@.4; 6@.6 plasma 8 move PULSE_GAS_3 6msec earlier (now @ -.021) P_ZCUR(2) = 4 V_OH1 = -165V @ .005 plasma 9 lower early fueling gas raise EF1 comm. resistance to .1 Ohm fizzle 10 no changes fizzle 11 lower first two points on PULSE_GAS_3 bad plasma 12 P_ZCUR(2) = 4 turn off P_BR_0 @ .026 plasma 13 P_BR_0 on until .036 P_ZCUR(3) off @ .038 more gas @ .1; less gas after .6 plasma 14 stair-step turn-on of P_ZCUR(2) lower gas @ .1 plasma 15 less late gas recalculate ZCUR(2) with FILAMENT/nocoils ramp Ip to .525MA @ .8s fizzle 16 Br_0 offset now 9.5e-3 fizzle 17 P_0=1.1e-5 Br_0=9.0e-3 PULSE_GAS_3 moved 2msec later(-.019) plasma 18 less gas after .6 plasma 19 P_ZCUR(2)=6 @ .09 S_RCUR = .66 @ 0. plasma 20 S_RCUR=.64 @ 0. more gas @ .2 plasma 21 more fuel after .2 enable PULSE_GAS_2 - pulse 10msec@100V D_ZCUR(2)=2 P_RCUR stair-steps earlier plasma 22 D_ZCUR(2)=1 more gas after .2 dud 23 P_0=1.2e-5 Br_0 offset = .0095 plasma 24 P-0=1.3e-5 S_IP=625kA plasma 25 P_0=1.4e-5 S_IP=725kA second H2 puff @ .5 fizzle 26 P_0=1.3e-5 fizzle 27 P_0=1.1e-5 fizzle

Session Leader Comments
Sep 29 1993 10:38:30:890AM		Steve Golovato	Pre-run Comments ----------------- Goals - (1) Produce reproducible deuterium discharges. (2) Bring up ICRF power slowly to avoid disruptions and the accompanying fizzles. (3) Add hydrogen to get some core absorption, which hopefully reduces edge effects which lead to impurity influx, disruptions. (4) Match plasma shape to antenna, move antenna and limiter closer, scan density, plasma radial position, H minority. Starting with last shot of last friday, 930928030, except moving plasma to -2cm from -3cm, added more late gas to raise density (avoid shark fins), and adding rf. rf Limiter at 92cm, antenna at 93cm.
Sep 29 1993 10:20:30:700AM	930928001	Steve Golovato	Dud. Setting up rf for several shorter pulses instead of one long one, to hopefully reduce risk of disruptions, also to see effects of turnon and turnoff more clearly.
Sep 29 1993 10:21:08:820AM	930928002	Steve Golovato	rf still in single pulse, next shot will have three ~100ms pulses between 300-750ms. Fizzled.
Sep 29 1993 10:22:23:970AM	930928003	Steve Golovato	Fizzled.
Sep 29 1993 10:23:35:980AM	930928004	Steve Golovato	rf in three pulses for this shot. Dudded.
Sep 29 1993 10:24:15:630AM	930928005	Steve Golovato	Fizzled.
Sep 29 1993 10:26:02:240AM	930928006	Steve Golovato	Fizzled, but almost, like shots 2 and 3. Goetz claims these shots aren't real fizzles because the cuurent ramps but disrupts very early. However, for all practical purposes, they are the moral equivalent of fizzles.
Sep 29 1993 10:26:28:130AM	930928007	Steve Golovato	Plasma, 400ka. Small plasma, to outside, no marfe, density ramped up to disruption at 375ms. rf interrupted during plasma, ran without plasma. Tuning off. Lowered power of first rf pulse so it won't interrupt with plasma and we can get tuning measurement.
Sep 29 1993 10:26:53:830AM	930928008	Steve Golovato	Similar to 7, disrupted vertically at at 700ms. n=3e20 at 500ms.
Sep 29 1993 10:27:16:900AM	930928009	Steve Golovato	Fizzled. Standard fizzle after disruption.
Sep 29 1993 10:27:37:350AM	930928010	Steve Golovato	Fizzled. Too much too much.
Sep 29 1993 10:28:15:090AM	930928011	Steve Golovato	Almost, disrupted up and in at 120ka, 53ms.
Sep 29 1993 10:28:37:970AM	930928012	Steve Golovato	First good plasma. Density lower, marfed.
Sep 29 1993 10:29:05:310AM	930928013	Steve Golovato	Pretty good but disrupted at 900ms, finned to death (quoth goetz).
Sep 29 1993 10:29:42:090AM	930928014	Steve Golovato	Good one. rf tuning progressing slowly.
Sep 29 1993 10:30:02:980AM	930928015	Steve Golovato	Fizzled. No apparent reason. Just try again.
Sep 29 1993 10:30:24:940AM	930928016	Steve Golovato	Fizzled. Better do something.
Sep 29 1993 10:30:48:970AM	930928017	Steve Golovato	Good plasma, no disruptions. Lowered fill to get it. Also kept current up, positive current ramp (or at least flat) avoids shark fins (or so the lore has it).
Sep 29 1993 10:31:14:010AM	930928018	Steve Golovato	Good One.
Sep 29 1993 10:31:57:550AM	930928019	Steve Golovato	Fiddling and diddling with hybrid gains to cure z_cur shakes during breakdown. Good plasma. Fiddling worked somewhat.
Sep 29 1993 10:32:31:310AM	930928020	Steve Golovato	Good again. Tuning progressing. When the tuning gets better, we'll add some H and start raising power more seriously. So far we've only raised it a little.
Sep 29 1993 12:52:52:430PM	930928021	Steve Golovato	Adding 10ms, 100v H puff at 400ms. Tuning pretty good, up to 100 kw. Still playing with breakdown wiggles. Added gas after 200ms to up density, reduce hards which are just beginning to appear. Density has been falling over last few shots. Plasma disrupted, first at 550ms, lost at 600ms.
Sep 29 1993 10:34:12:900AM	930928022	Steve Golovato	H puff widened to 12ms. Added more gas to raise density. Low density and higher rf power lead to shark fin, disruption. Making more, shorter rf pulses to avoid disrupting, continuing to increase power. Tuning now good. Dud. Dud. Dud. Should have raised the fill like Goetz wanted.
Sep 29 1993 10:34:34:960AM	930928023	Steve Golovato	Raise the fill. Ran but disrupted again at 600 ms.
Sep 29 1993 10:35:17:930AM	930928024	Steve Golovato	Raised the fill again, increased Ip to 600ka and lowered rf power. Ran but disrupted again at 600 ms.
Sep 29 1993 10:36:01:150AM	930928025	Steve Golovato	Adding second H puff at 500ms, raised fill, lowered later rf pulses to avoid disruption. Also raised Ip more and earlier. Fizzled. Got several rf plasmas. Too much fill.
Sep 29 1993 10:36:42:700AM	930928026	Steve Golovato	Lowered fill. Moved limiter and antenna in one cm each. Maybe getting closer to fast wave cutoff layer will reduce edge interaction which leads to impurity influx. Fizzled.
Sep 29 1993 10:37:03:770AM	930928027	Steve Golovato	Lowered fill again. Fizzled! Aghhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh!!!

Physics Operator Comments

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	09:14:13:730AM	Plasma	Ok	oh2u >1scr not conducting late in the shot
2	09:30:51:810AM	Plasma	Ok	no faults
3	09:50:34:560AM	Plasma	Ok	no faults
4	10:12:03:310AM	Plasma	Ok	no faults
5	10:26:54:730AM	Plasma	Ok	no faults
6	10:44:04:510AM	Plasma	Ok	no faults
7	11:02:51:560AM	Plasma	Ok	no faults
8	11:20:24:850AM	Plasma	Ok	no faults
9	11:42:39:500AM	Plasma	Ok	no faults
10	12:00:51:640PM	Plasma	Ok	no faults
11	12:16:17:660PM	Plasma	Bad	oh2u+l commutation faults
12	12:38:26:430PM	Plasma	Ok	no faults
13	01:00:26:930PM	Plasma	Ok	no faults
14	01:15:19:770PM	Plasma	Ok	no faults
15	01:30:47:690PM	Plasma	Ok	no faults
16	01:48:03:080PM	Plasma	Ok	no faults
17	02:01:04:050PM	Plasma	Ok	no faults
18	02:16:58:220PM	Plasma	Ok	no faults
19	02:36:26:570PM	Plasma	Ok	no faults
20	02:51:44:690PM	Plasma	Ok	no faults
21	03:10:16:720PM	Plasma	Ok	bad disruption -- comm faults on all pf supplies except ef1 lower
22	03:43:41:490PM	Plasma	Ok	no faults
23	03:55:47:050PM	Plasma	Ok	no faults
24	04:10:45:710PM	Plasma	Ok	no faults
25	04:25:50:310PM	Plasma	Ok	fizzle with commutation failures on ef1 and oh1
26	04:48:15:140PM	Plasma	Ok	no faults
27	05:03:04:990PM	Plasma	Ok	no faults