Miniproposals

Miniproposal: 054 Date Filed: 4/5/1994 Title: ICRF Antenna Conditioning (PRF> 100kW) First Author: Yuichi Takase Session Leader: Steve Golovato (shots 1-26)

Engineering Operator Run Comment

more rf, higher density

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Entered: Jul 7 2004 03:49:10:600PM

Author: To Be Determined

Run Summary - 940602 -------------------- SL - Golovato PO - Granetz EO - Various people, mostly Fairfax Run Goals - (1) Look at higher density to improve loading with larger outer gap. (2) Move closer to having an x-point, see if H-mode returns. (3) Vary hydrogen concentration. We started with the last shot from yesterday and got a dud. A slight adjustment in BZ and gas got a disruptive plasma on the second shot and good plasmas the rest of the run (except shot 15 which had mutliple power supply failures). RF ran from shot 3 on, from 400-760ms, with two power steps, 150 kW, about 1-1.2 MW, and 1.4-1.8 MW. The power varies for the same setting depending on the tuning, which in turn depends on the loading. We initially got plasmas which limited on the inner nose, and saw transient H-modes, much as were seen last week. Puffing more gas to raise the density resulted in fatter plasmas. We then moved the plasma inward to increase the outer gap to 2-2.5 cm, as we will want for good diverted plasmas. The loading clearly decresed with outer gap and also appeared to increase with density. By adjusting the tuning to match at lower loading, we were able to put high power, 1.6-1.7 MW, into limited plasmas with appropriate outer gaps. Thus goals (1) and (2) were achieved. We were informed by the OMA team that we had plenty of hydrogen, even without puffing. It was presumably coming from the walls where it ended up after yesterday's run. It was as high as 15-25% and came down to a steady-state of 10%. There was no sense puffing more since there was already plenty, so goal (3) was achieved. We proceeded to try to develop a real diverted discharge. This involved a lot of pcs manipulation and guesses about what might work. The approach was not to try to put the x-point in the right place all at once but to first make one with symmetric demands, then if successful, move it down. Read the physics operator summary and logbook for this saga. We were close on the last shot, 26, which was highly elongated (kappa=1.76) and diverted after 720ms. It disrupted at 800 ms. The x-point was very close to where we want it in the lower divertor. Shot 26 also had good rf tuning such that the tuning improved as the loading went down, keeping the rf on at hihg power for the full pulse length. There was a long (perhaps debatable) H-mode from 600-760 ms, the length of the high power rf pulse (1.5 --> 1.3 MW). There was one rf interrupt right at the beginning, during which it fell out of H-mode. We made some adjustments to improve the equilibrium but we lost the last shot to a daq3 crash. Nevertheless we made a lot of progress both in learning to couple high power with less loading and in producing diverted discharges.

Physics Operator Summaries

Entered: Jul 7 2004 04:36:51:660PM

Author: To Be Determined

Physics operator's summary for run 940602. SL: S. Golovato PO: R. Granetz EC: D. Gwinn/S. Fairfax As mentioned in the session leader's summary, this was a very good run, with only one dud (001), and one shot lost to a power supply fault (015). In fact, the few problems we did have were mostly related to computer crashes. After shot 015, many of the workstations in the control area crashed and had to be rebooted, resulting in a 40 minute delay. Although the next pulse was a good plasma shot, DAQ2 apparently had not fully recovered from the reboot and as a result, much of the camac on the DAQ2 highway didn't get initialized, so significant data is missing on that shot (including part of the magnetics, such as Ip, fluxes, etc --> no EFIT). Just prior to starting up for shot 027, there was another crash (DAQ3 this time), which effectively terminated the run. There was also a brief power outage in the cell shortly after the run ended, which crashed TORVAC and other pc's. The RF goals of this run were accomplished with several hours to spare (see session leader's report), and this allowed us some time for discharge development, i.e. attempting to get a diverted plasma. The plasmas for the first part of the run were limited on the inner wall (and sometimes on the "bridge" of the inner divertor nose), and had an elongation of only 1.45. Knowing that higher elongation increases the chances of a disruption, our first step was to reduce the plasma current from 0.8 MA to 0.6 MA on shot 020. Next, we turned on the RXL feedback, which attempted to control the x-point radial position using only EF2L, resulting in too much imbalance in the EF2's and other coils (021). So EF2U was added into the RXL controller with symmetric weight (022). The hybrid tried to do the right thing, but the EF2's were driven close to their rail, and the x-point was still too far in. So then the OH2's were added into RXL symmetrically, but with the opposite polarity, so they would push the x-point outward to help the EF2's. It took us a few shots (023, 024, 025) to finally tune the M-matrix coefficient to an acceptable value. Shots 023 and 024 disrupted. The elongation on these shots was still not high enough to give a reasonable diverted plasma. Lacking a suitable predictor/controller for kappa, presently the best way to increase the elongation is through current programming of the EF1's. We did this for shot 025, resulting in a noticeable improvement. It was observed that the EF2 control of RXL was fighting its respective current programming, so for shot 026 we turned off the current programming feedback after t=0.5 s. This resulted in a very highly elongated (kappa=1.75) plasma that became diverted just before going vertically unstable. For the next shot, several tweaks were made, including slightly reducing the demanded currents in the EF1's, which would hopefully result in slightly less elongation. Due to a computer crash, the run was forced to end before we could try this shot. We feel very confident that the PCS programming would have produced a good diverted plasma, so we saved it as shot 940602927 (instead of 027) in the pcs$data area. This might be a good starting point for future divertor discharge development.

Session Leader Comments
Jun 2 1994 09:01:32:350AM		Steve Golovato	Run 940602 Continuation of yesterday's run - Look at higher density to improve loading with larger outer gap. Move closer to having an x-point, see if H-mode returns. Vary hydrogen concentration. Starting with last shot from yesterday but with rf.
Jun 2 1994 04:58:20:730PM		Steve Golovato	940602027 Plasma. no shot, daq3 crashed. run over. This would have been a great shot.
Jun 2 1994 09:07:14:810AM	940602001	Steve Golovato	Run 940602 940602001 Dud. No rf.
Jun 2 1994 09:22:50:750AM	940602002	Steve Golovato	Run 940602 940602002 Plasma but disruptive. No rf, first shot rf in manual, this shot, hv not on. Next shot should go.
Jun 2 1994 09:46:58:920AM	940602003	Steve Golovato	Run 940602 940602003 PLasma. disrupted at 880ms. Hards. RF ran.
Jun 2 1994 10:03:24:760AM	940602004	Steve Golovato	940602004 Plasma, no disruptions. PLasma pulling off of inner and outer limiters, limiting on inner nose (as last week). Loading dropa, rf interrupts above 1.2 MW. Saw brief h-modes as last week under these conditons.
Jun 2 1994 10:16:52:080AM	940602005	Steve Golovato	940602005 Plasma. RF ran better. loading dropping starting at about 500ms as equilibrium evolves. raising gas to raise density, try to maintain loading with plasma off of outer limiter.
Jun 2 1994 10:30:10:840AM	940602006	Steve Golovato	940602006 Plasma. rf ran without interrupts. density stayed up, plasma stayed fat and on outer limiter, rf ran until transmitter faulted at 1.75MW. Reduced power a little for next shot. Loading flat at 10 ohms.
Jun 2 1994 10:54:55:440AM	940602007	Steve Golovato	940602007 Plasma. rf ran longer with power programmed slightly lower, but loading went up as density rose during rf, power was actually higher.
Jun 2 1994 11:13:50:080AM	940602008	Steve Golovato	940602008 Plasma. Tweeked power programming down again. Running ok with outer gap of about 1.5cm.
Jun 2 1994 11:27:49:010AM	940602009	Steve Golovato	940602009 Plasma. Moved puff for tom luke out to 850ms from 780ms. rf interrupted at 710ms at 1.7MW. Density rise with rf raises loading, power.
Jun 2 1994 11:49:43:800AM	940602010	Steve Golovato	940602010 Plasma. Reduced gas after 600ms a little, moving rcur to .665 from .67 between 500 and 600ms. Loading drops from 11.5 ti 6.5 ohms, power at 1.4MW. Plasma goes in and out of h-mode as loading drops further and power drops..
Jun 2 1994 11:58:16:430AM	940602011	Steve Golovato	940602011 Plasma. Try tuning to match better after plasma moves in. Tuning better, tune some more.
Jun 2 1994 12:25:17:560PM	940602012	Steve Golovato	940602012 Plasma. Tune more to try to match lower loading.
Jun 2 1994 12:35:32:850PM	940602013	Steve Golovato	940602013 Plasma. Power droppiong.
Jun 2 1994 12:49:47:060PM	940602014	Steve Golovato	940602014 Plasma. Power has dropped to about 1.3MW, tweeking it up a little. Tweeking up gas also. Laoding went up, as did power.
Jun 2 1994 12:54:03:200PM	940602014	Steve Golovato	940602014 Plasma. Power has dropped to about 1.3MW, tweeking it up a little. Tweeking up gas also. Loading went up, as did power, to 1.6MW.
Jun 2 1994 01:02:34:910PM	940602015	Steve Golovato	940602015 No plasma, power supply failure. Lower gas late to avoid disruption during rampdown, raise it earlier to bring up density sooner.
Jun 2 1994 02:12:53:120PM	940602016	Steve Golovato	940602016 Plasma. Had 40 minute delay because of computer failure. No magnetics or xtomo data. Short periods of elm-free h-mode at 700 and 740ms. 1.6MW. Interrupts continually during first, low power rf because of detuning.
Jun 2 1994 02:20:53:410PM	940602017	Steve Golovato	940602017 Plasma. Many rf interrupts, tuning went too far.
Jun 2 1994 02:35:22:900PM	940602018	Steve Golovato	940602018 Plasma. Tune back a little. Raise plasma 0.5cm. Plasma moved up, also reduced outer gap from 2.5 to 1.5cm, raising loading to 10 ohms.
Jun 2 1994 02:56:16:500PM	940602019	Steve Golovato	940602019 Plasma. Moving plasma in by 0.5cm more. Outer gap to 2cm. Loading lower.
Jun 2 1994 03:11:18:610PM	940602020	Steve Golovato	940602020 Plasma. Lowered current to 600ka before starting to make x-point. RF interrupting at highest power 1.7MW.
Jun 2 1994 03:30:06:330PM	940602021	Steve Golovato	940602021 Plasma. Turning on x-point R-position feedback from .5-.9s. Also moving plasma out 1cm. No x-point but plasma moved out 1cm. rf interrupts after .65s
Jun 2 1994 03:50:52:450PM	940602022	Steve Golovato	940602022 Plasma. Increasing rxl prop gain and adding ef2u as well as ef2l. Keeping things symmetric even if we get double null. Almost got lower xpoint, on nose. Adding oh2's symmetrically to push as well as pulling with ef2's.
Jun 2 1994 03:55:17:060PM	940602022	Steve Golovato	940602022 Plasma. Increasing rxl prop gain and adding ef2u as well as ef2l. Keeping things symmetric even if we get double null. Almost got lower xpoint, on nose. Adding oh2's symmetrically to push as well as pulling with ef2's.
Jun 2 1994 04:03:56:440PM	940602023	Steve Golovato	940602023 Plasma. Disrupted early, not diverted, rf off early. Not very successful yet.
Jun 2 1994 04:22:01:970PM	940602024	Steve Golovato	940602024 Plasma. oh2's too strong. rf a little better. disrupted early.
Jun 2 1994 04:51:57:620PM	940602025	Steve Golovato	940602025 Plasma. disrutped early, no xpoint in sight.
Jun 2 1994 04:54:37:650PM	940602026	Steve Golovato	940602026 Plasma. Long but ambiguous h-mode 620-760ms. diverts at 740, disrupts at 800ms, very elongated, kappa=1.76.

Physics Operator Comments
Jun 2 1994 08:58:11:840AM	940602001	Robert Granetz	Start with last shot from yesterday. Once we get a plasma, the first goal is to increase the plasma density to improve coupling to the RF antenna, and if successful, move the plasma farther off both the inner and outer limiter surfaces.
Jun 2 1994 09:10:29:940AM	940602001	Robert Granetz	Dud. A touch too much vertical field. RF assist didn't go, either. Next shot, 15 amps more negative current in EF4. Hopefully the RF will go also.
Jun 2 1994 09:27:25:940AM	940602002	Robert Granetz	Plasma. Disrupted upwards at 0.765 s, Ip=0.80 MA. Big spike up to 0.97 MA at the disruption. Halo_u=100 kA. Very little in the halo_l or the divertor. Plasma was disruptive for much of the shot. RF did not fire. Next shot, lower pre-fill from 1.1 to 1.0e-5 torr. And get the RF on.
Jun 2 1994 09:43:21:840AM	940602003	Robert Granetz	Plasma. Disrupted downward during the rampdown. Hard x-rays early on. RF fired on this one. Picrf=1.6 MW, in three separate pulses. Next shot, raise pre-fill to 1.1 from 1.0e-5.
Jun 2 1994 09:58:45:480AM	940602004	Robert Granetz	Good plasma. No disruption. Ip=0.80 MA, duration=1.15 s. RF go to about 1.2 MW and then crapped out. Plasma just barely pulls off of both inner and outer midplane surfaces, and is limited on the inner divertor nose bridge around 0.55 s. Next shot, no changes, except enable argon for John Rice.
Jun 2 1994 10:02:11:150AM	940602004	Robert Granetz	There were brief ICRF H-modes during this shot (around 0.65 s).
Jun 2 1994 10:12:15:310AM	940602005	Robert Granetz	Good plasma. No disruption. Ip=0.80 MA, duration=1.14 s. RF had several interruptions, but not as bad as last shot. Picrf=1.2-1.3 MW. Next shot, raise pulse gas programming from 0.4 s onward. Jim Terry reports that hydrogen is increasing during each pulse, and from pulse-to-pulse. I just verified that the relevant valve is NOT enabled, so maybe it's coming from the walls.
Jun 2 1994 10:26:34:910AM	940602006	Robert Granetz	Good plasma. No disruption. Ip=0.81 MA, duration=1.13 s. RF worked much better on this shot, because the density came up, as programmed. Picrf=1.7 MW. No changes for next shot.
Jun 2 1994 10:45:50:920AM	940602007	Robert Granetz	Good plasma. No disruption. Ip=0.81 MA, duration=1.10 s. Picrf=1.75 MW. Next shot, no changes.
Jun 2 1994 11:11:11:590AM	940602008	Robert Granetz	Good plasma. No disruption. Ip=0.81 MA, duration=1.12 s. Picrf=1.6 MW. Density reaches nl=1.0e20 at 0.75 s. More RF H-modes. Next shot, move the 20 ms gas puff later in the shot (from 0.78 s to 0.85 s). This pulse is for Tom Luke.
Jun 2 1994 11:28:25:950AM	940602009	Robert Granetz	Good plasma. A bit disruptive starting at 0.94 s. Ip=0.81 MA, duration=1.05 s Density is too high during the late rampdown. Picrf=1.65 MW. Next shot, lower gas late in shot, and try to get a better density flattop throughout the shot. Also, program RCUR to move the plasma 0.5 cm inward during most of the RF pulse duration (from 0.67 to 0.665 m).
Jun 2 1994 11:45:56:930AM	940602010	Robert Granetz	Good plasma. No disruptions. Ip=0.81 MA, duration=1.12 s. Picrf=1.4-1.5 MW. We got the change in R-centroid, and it affected the RF loading, as expected. Next shot, increase slightly the late 20 ms gas pulse for Tom Luke, and increase the argon gas puff pulse by stretching it out by an extra 1 ms.
Jun 2 1994 11:57:53:980AM	940602011	Robert Granetz	Good plasma. No disruptions. Ip=0.81 MA, duration=1.12 s. Picrf=1.4 MW. Next shot, tweak up Tom Luke's 20 ms gas puff by 1 volt.
Jun 2 1994 12:17:44:270PM	940602012	Robert Granetz	Good plasma. No disruptions. Ip=0.81 MA, duration=1.12 s. Picrf=1.3-1.4 MW. Next shot, make another 1 volt increase on Tom Luke's 20 ms gas puff
Jun 2 1994 12:39:00:730PM	940602013	Robert Granetz	Good plasma. No disruptions. Ip=0.81 MA, duration=1.12 s. Picrf=1.3 MW. Next shot, increase gas puff after 0.4 s.
Jun 2 1994 12:49:14:810PM	940602014	Robert Granetz	Good plasma. Somewhat disruptive during the late current rampdown. Ip=0.81 MA, duration=1.01 s. Too much density late in the shot. Picrf=1.55 MW. Next shot, reduce late gas puffing, and increase early gas puffing.
Jun 2 1994 01:04:15:030PM	940602015	Robert Granetz	No plasma. OH1, OH2's, and EF1's all faulted early and never actually came on. Apparently the PF breaker pulled out before the shot.
Jun 2 1994 01:17:15:310PM	940602015	Robert Granetz	We're having computer and communication problems.
Jun 2 1994 01:54:05:920PM	940602015	Robert Granetz	Multiple workstations have to be rebooted. We'll be down for a while.
Jun 2 1994 02:08:12:280PM	940602016	Robert Granetz	Good plasma. No disruptions. Some of the magnetics and xtomo digitisers failed to initialize, so we have no Ip and no flux data, etc. Picrh=1.6 MW. There apparently were a couple of extremely brief (5 ms) ELM-free periods during the RF pulse. No changes for next shot.
Jun 2 1994 02:23:32:400PM	940602017	Robert Granetz	Good plasma. Minor disruptions late in the rampdown. Ip=0.81 MA, duration=1.02 s. RF was only on for a very brief pulse (.02 s, 1.8 MW) due to being mis-tuned. For next shot, program in a shift of +0.5 cm in ZCUR, starting at 0.5 s. This is in response to a request by Yuichi, since EFIT claims the plasma is 5 mm down. (This does not agree with the x-rays, which claim Z-centroid=-1 mm.)
Jun 2 1994 02:35:51:550PM	940602018	Robert Granetz	Good plasma. Minor disruptions after t=0.94 s, late in the rampdown. Ip=0.81 MA, duration=1.07 s. Could be a sharkfin at 0.94 s. We got the programmed change in ZCUR. Next shot, move RCUR further in by 5 mm starting at 0.5-0.6 s.
Jun 2 1994 02:55:18:620PM	940602019	Robert Granetz	Good plasma. Minor disruptions late in rampdown. Ip=0.81 MA, duration=1.02 s. Picrf=1.6 MW. For next shot, reduce Ip to 0.6 MA, in preparation for trying to get a divertor plasma.
Jun 2 1994 03:12:24:810PM	940602020	Robert Granetz	Good plasma. No disruptions. Ip=0.64 MA, duration=1.11 s. Next shot, try for divertor. Put gain of 2 on RXL, from t=0.5 to 0.9 s, and remove inward 1 cm step in RCUR (the demand is now to stay at 0.67 m).
Jun 2 1994 03:30:46:110PM	940602021	Robert Granetz	Good plasma. No disruptions. Ip=0.63 MA, duration=1.10 s. Picrf=1.5-1.6 MW. Divertor not formed, however. Next shot, increase gain on RXL from 2 to 5, and include EF2U in the RXL controller, with symmetric M-matrix value. This is to reduce the problem of creating an upper x-point near the inner wall, which is what we did on this shot.
Jun 2 1994 03:45:23:180PM	940602022	Robert Granetz	Good plasma, No disruption. Ip=0.63 MA, duration=1.09 s. Came close to getting a diverted plasma, but not quite, and EF2's are nearly maxed out in current. Next shot, include OH2's in the RXL controller with the opposite sign for M, so that they will push the x-points.
Jun 2 1994 04:03:14:290PM	940602023	Robert Granetz	Plasma. Disrupted at 0.6 s (slow dI/dt). Ip=0.60 MA, duration=0.64 s. Shape was not too far from circular (kappa=1.2). On next shot, reduce the M-values for the OH2's in the RXL controllers from -2.e-3 to -0.5e-3 (i.e. a factor of four).
Jun 2 1994 04:22:19:400PM	940602024	Robert Granetz	Plasma. Disruptive, starting at 0.66 s, with final disruption at 0.72 s. Ip=0.60 MA. Still didn't form a divertor, and in fact, the inclusion of the OH2's doesn't seem to have helped us. Density starts climbing after 0.5 s, similar to what is seen in a detached plasma. Next shot, reduce M-values of OH2's in the RXL controller another factor of 10, and change EF1U&L current programming to help increase plasma elongation.
Jun 2 1994 04:31:29:970PM	940602025	Robert Granetz	Good plasma. No disruptions. Ip=0.63 MA, duration=1.14 s. Picrf=1.5 MW. Next shot, turn off feedback on EF2 current programming after 0.5 s (which is about the time that RXL feedback comes on).
Jun 2 1994 04:53:08:920PM	940602026	Robert Granetz	Plasma. Upward disruption at 0.80 s. Ip=0.63 MA. Picrf=1.5 MW We very nearly got a diverted plasma, but the elongation went too high (1.75) and the plasma disrupted. For the next shot, in the RXL controller, double the OH2 M-value and reduce the EF2's M-value by 20%. Also change EF1U&L programming so that they don't pull as much.
Jun 2 1994 04:58:24:240PM	940602026	Robert Granetz	DAQ3 crashed before we could get the next shot off. End of run.
Jun 2 1994 05:02:03:480PM	940602026	Robert Granetz	In order to save the PCS changes that we made for the next shot (which couldn't be taken because of the computer crash), I saved it as shot #940602927 in the pcs$data directory.

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	08:57:21:220AM	Plasma	Ok
2	09:11:50:530AM	Plasma	Ok	disruption 780 ka
3	09:28:18:180AM	Plasma	Ok	800 ka
4	09:44:12:220AM	Plasma	Ok	800 ka
5	09:59:02:470AM	Plasma	Ok	800 ka
6	10:14:55:020AM	Plasma	Ok	800 ka
7	10:37:19:860AM	Plasma	Ok	plasma 800 ka
8	10:54:09:540AM	Plasma	Ok	800ka
9	11:13:21:970AM	Plasma	Ok	oh1 com fail
10	11:30:10:690AM	Plasma	Ok	800ka
11	11:46:32:190AM	Plasma	Ok	no faults
12	12:06:18:840PM	Plasma	Ok	800 kA; plasma current terminated at less than 100 kA
13	12:25:22:650PM	Plasma	Ok	No faults
14	12:40:43:670PM	Plasma	Ok	800 ka; minor disruption at 400 kA during ramp-down but recovered.
15	12:56:54:140PM	Plasma	Bad	PF breaker opened before shot; reason unclear, no OH1, OH2s, EF1s
16	01:50:58:980PM	Plasma	Ok	40-minute delay for DAQ3 reboot; this plasma OK but no magnetics data
17	02:09:54:930PM	Plasma	Ok	No faults
18	02:23:04:860PM	Plasma	Ok	ramped down to 67 kA from 820
19	02:42:28:870PM	Plasma	Ok	1.6 MW RF this shot
20	02:58:03:660PM	Plasma	Ok	600 kA this shot; preparing to bring up diverter
21	03:15:39:930PM	Plasma	Ok	600 kA; down to 40 kA at end; not sure we got diverter this shot
22	03:32:23:340PM	Plasma	Ok	divertor on nose tiles. Checking probes
23	03:52:00:320PM	Plasma	Ok	very round plasma. Interesting 600 kA disruption.
24	04:06:20:960PM	Plasma	Ok	disrupted from 400 kA
25	04:23:27:250PM	Plasma	Ok	plasma finished; no disruption
26	04:36:58:870PM	Plasma	Ok	600 kA disruption; EF2U and EF2L in current limit 0.6-0.8 sec