Alcator C-Mod Run 1000329 Information

Back to runs

Miniproposals
Miniproposal:
Date Filed:
Title:
First Author:
Session Leader:Steve Wolfe (shots 1-11)

Operators
Session leader(s):Steve Wolfe
Physics operator(s):Robert Granetz
Engineering operator(s):Bill Cochran,Gary Dekow,Frank Silva

Engineering Operator Run Comment
z

Session Leader Plans

Physics Operators Plans

Session Leader Summaries
Entered: Jul 7 2004 03:49:22:757PM
Author: To Be Determined
Session Leader's summary for Wednesday, 3/29/2000

Reverse field Startup
SL: Wolfe
PO: Granetz
EO: Silva,Cochran,Dekow

The plan was to continue the startup and tokamak conditioning discharges from
yesterday.

Shot 2 was the best shot so far, with 800kA to 1.3sec, terminated when the
OH's ran into the (reduced) current limit of 22kA. However, the F28 flux loop
(decant loop), which had been observed to be drooping following yesterday's
run, broke at the disruption.

Re-doing all the PCS predictors, including a new re-write of FLUXIHH and
generating a new MHD_STATIC tree for the RECON_VAC predictors, updating the
hybrid model file, plus modifying the processed flux expressions in the
MAGNETICS tree, took over four hours. The process was not helped by a plague
of computer crashes.

Meanwhile, the leak at f-port is still there, and there may be another one
someplace else. There is a lot of ice around F-port, probably contributing to
the problem.

Once we got back on the air, we had plasmas on shots #4 (runaways), 7, and 10.
Shot 10 is probably the best place to start tomorrow.

Physics Operator Summaries
Entered: Jul 7 2004 04:37:10:160PM
Author: To Be Determined
Physics operator summary for run 1000329:

MP ? -- Obtain and improve plasma initiation with reversed Bt

SL: S. Wolfe
PO: R. Granetz
EO: F. Silva/B. Cochran/G. Dekow

Summary: We had a 1.3 second long plasma on shot 02, but it terminated in
a disruption which broke the F28 flux loop across H-port. Everything
went downhill from there. The F27 loop, which is redundant with
F28, is working fine, but patching it into the hybrid estimator
software took a long time. Plus, the new estimators give slightly
different results, so we essentially had to start from scratch again,
at least as far as getting a robust null. Lump in a few annoying
computer crashes, a few PCS oddities that we unexpectedly tripped
on, and we essentially ended up making little or no progress.

Scorecard: 4 plasmas (only 1 reached flattop, shot 02)
7 fizzles
0 duds (0 due to lack of one or more power supplies)
0 test
0 aborted
----------
11 total

-------------------------------------------------------------------------------

Engineering setup for Wednesday, 1000329:

Run will begin at 09:00 and end at 17:00

Overnight ECDC in D2, stopping 1 hour prior to the run.

Power system setup as on 1000328014.

Note: Fizzle detector should be on from t=0.5 to 1.7 s (for reasons having
to do with the OH2L coax)

Gas setup - fill B-top with 6 psi of D2 Hybrid enabled
fill B-side lower with 1 psi of Ar Hybrid DISABLED
leave B-side upper as is Hybrid DISABLED
fill C-side with 30 psi of D2 Hybrid enabled
leave J-bottom as is Hybrid DISABLED

The following gate valves should be enabled, assuming no vacuum problems:
ECE, VUV

-------------------------------------------------------------------------------

Physics operator's setup for Wednesday, 1000329.

MP ? -- Obtain and explore initiation with reversed Bt

Session leader: S. Wolfe
Physics operator: R. Granetz

Plans: More tweaking of breakdown and rampup. Clean up machine. RF will
probably not be available.

-------------------------------------------------------------------------------

Note: Overnight ECDC in D2 was requested, but we actually got
about 3.5 hours total. ECDC was stopped 1 hour before run.

Hybrid coldstart done by Steve Wolfe; no errors found

Loaded 1000328014.

We did not swap F27 for F28 in either the hybrid estimators nor the
processed fluxes.

The power engineers have added a fizzle delay to all the OH PLCs.
Today we will:

. set the fizzle detector start back to .08 seconds
. set all the OH PLC fizzle delays to 0.500 seconds
. set the TF PLC fizzle delay to 0.600 seconds

-------------------------------------------------------------------------------

Shot 01 -- Fizzle. Fill pressure is a touch high. Null is okay, but
Bz needs to come up faster.

Next shot: reduce pre-fill puff from 23 to 20 ms, and increase
magnitude of early EF3 voltage from -2500 to -2800 V.

Shot 02 -- Plasma. Ip=0.8 MA, hard disruption at t=1.32 s. Probably ran
into the recently decreased PLC current limits on the OH supplies.

Note: F28 broke at the disruption. It is now electrically open.
F29 is still okay.

There will be a long break while Steve reprograms the hybrid to use F27
instead of F28. I will change the processed flux accordingly.
During the break, we have ECDC on as much as possible.

Next shot: increase OH1 PLC limit from 22 to 27 kA, and the
OH2's from 22 to 26 kA. (I'm referring to the limits on the
reverse bridge.) Swap F27 for F28 in the hybrid
estimators and the processed fluxes.

We've had several hours of computer crashes, flux estimator code work,
PCS estimator updating, PCS bug recovery, more computer crashes, etc.
It's now 13:45 and we're finally ready to go again.

Shot 03 -- Fizzle. Br0 looked a little high, but the null looks very
different, and I don't know whether it's due to the new flux
loops in the analysis or in the hybrid or both.

Next shot: reduce Br0 by 1 mT.

Shot 04 -- Plasma. Disrupted at t=0.3 s. Br0 doesn't seem to have changed.
The startup hit something and almost didn't make it. Looks like
there were lots of runaways.

Next shot: increase early gas fueling.

Note: there might have been a bitbus problem, which means that the preceding
changes for shot 04 might not have actually gotten to the hybrid. I just
did a coldstart, successfully.

Shot 05 -- Fizzle. Rather weak. Br0 looks good. Breakdown traces look very
similar to shot 02.

Next shot: retry with no changes

Shot 06 -- Fizzle. Rather weak again. It seemed to like having Br0 slightly
positive (at least according to our latest calculations).

Next shot: increase Br0 back up to 0.000 mT.

Shot 07 -- Plasma. Died at t=0.21 s, after several disruptions. Excellent
current rise. Can't see anything obviously wrong, except for
lots of dirt.

Next shot: repeat with no changes.

Shot 08 -- Fizzle.

Next shot: repeat with no changes.

Shot 09 -- Fizzle. A little too much Bz.

Next shot: add 15 amps to IC_EF4 in segment 1.

Shot 10 -- Plasma. Hard disruption at t=0.38 s. The recon_vac:v_loop is
broken, and has been broken since we swapped F27 for F28 after
shot 02. This shot was much better behaved than the previous
short plasmas.

Next shot: no changes.

Shot 11 -- Fizzle.

End of run.

Session Leader Comments
Mar 29 2000 09:44:42:857AM1000329001Steve Wolfeshot#1: fizzle. f28 still looks ok.
Bz0 looks too positive at t=0. CII~.015
fill looks high ~.09mtorr
Mar 29 2000 10:01:21:530AM1000329002Steve Wolfeshot#2: plasma, to 1.3sec.
Ran into the OH current rails, which had been set to 22kA (had been -30 I think)
Loop voltage is 1.7V.

F28 seems to have broken at the end of that shot.
We will need to re-do all the predictors, etc.
Mar 29 2000 02:29:36:170PM1000329003Steve WolfeShot 3: After several hours of re-coding, reloading PCS, and computer
crashes, not to mention ECDC, ready to start up again.

The only waveform changes are to remove the tf waggle. Also, the current
limits on the Oh''s should be back to -26kA.

Got a fizzle. Ian''s R and Z traces look funny. Otherwise not bad.
PhysAOut looks ok.
Mar 29 2000 02:36:23:903PM1000329003Steve WolfeShot 3: Ian''s FLUXIHH things in the analysis tree use the unprocessed magnetics
data, and have not been updated to not use F28, so they are broken.
Mar 29 2000 02:49:07:340PM1000329004Steve WolfeShot 4: Tweak the Br and try again.
plasma to about .3 sec, lots of injections.
lots of runaways.
Mar 29 2000 03:36:21:483PM1000329006Steve WolfeAt the request of the RF group, turned RF tree off (set noinclude)
in CMOD_MANAGE for shot 7.
Mar 29 2000 03:53:57:687PM1000329007Steve Wolfeshot#7: Br tweak- plasma to .2, gets in trouble at .14
Very resistive, lots of injections.
Initial current rise is very fast, and looks smooth.
The F28 signal at the hybrid is hard saturated >+10V after .15sec
Mar 29 2000 04:47:39:577PM1000329010Steve Wolfeshot#10: plasma to .38 sec, slight hesitation in current rise.
Cleaner than 7, around 2-2.5V loop.

Bob points out the recon_vac data is showing +/- 1e37 on vloop;
also br, bz, dbz, etc. It's been that way since shot 3, i.e. after
the change for flux loop replacement. Looks like an IDL v5.2 bug?

Physics Operator Comments
Mar 29 2000 08:04:19:763AMSteve WolfeBob will not be in until 8:45 or so.
Running a coldstart. 08:02:32.83
All matrices respond
Mar 29 2000 09:16:31:310AMRobert GranetzEngineering setup for Wednesday, 1000329:

Run will begin at 09:00 and end at 17:00

Overnight ECDC in D2, stopping 1 hour prior to the run.

Power system setup as on 1000328014.

Note: Fizzle detector should be on from t=0.5 to 1.7 s (for reasons having
to do with the OH2L coax)

Gas setup - fill B-top with 6 psi of D2 Hybrid enabled
fill B-side lower with 1 psi of Ar Hybrid DISABLED
leave B-side upper as is Hybrid DISABLED
fill C-side with 30 psi of D2 Hybrid enabled
leave J-bottom as is Hybrid DISABLED

The following gate valves should be enabled, assuming no vacuum problems:
ECE, VUV
Mar 29 2000 09:16:48:483AMRobert GranetzPhysics operator''s setup for Wednesday, 1000329.

MP ? -- Obtain and explore initiation with reversed Bt

Session leader: S. Wolfe
Physics operator: R. Granetz

Plans: More tweaking of breakdown and rampup. Clean up machine. RF will
probably not be available.
Mar 29 2000 09:26:58:577AMRobert Granetz #4 29-MAR-2000 09:25:09.85 NEWMAIL
From: NEL::IRBY
To: GRANETZ, WOLFE
CC: IRBY
Subj: Fizzle Changes

We have added a fizzle delay to all the OH PLCs.
Today we will:

. set the fizzle detector start back to .08 seconds
. set all the OH PLC fizzle delays to 0.500 seconds
. set the TF PLC fizzle delay to 0.600 seconds

jim
Mar 29 2000 09:32:53:950AMRobert Granetz Note: Overnight ECDC in D2 was requested, but we actually got
about 3.5 hours total. ECDC was stopped 1 hour before run.

Hybrid coldstart done by Steve Wolfe; no errors found

Loaded 1000328014.

We did not swap F27 for F28 in either the hybrid estimators nor the
processed fluxes.
Mar 29 2000 09:49:05:403AM1000329001Robert GranetzShot 01 -- Fizzle. Fill pressure is a touch high. Null is okay, but
Bz needs to come up faster.

Next shot: reduce pre-fill puff from 23 to 20 ms, and increase
magnitude of early EF3 voltage from -2500 to -2800 V.
Mar 29 2000 01:28:23:653PM1000329002Steve WolfeAll the codes have been updated.
Bob went through and recalculated all predictors and did a LOAD/CLEAN
but it gave a controller out-of-range error on SEG1 RCUR; it turns out
the controller was right on the edge, and since he had to lower the
predictor factor it pushed it over; note that seg1 has hand-made controllers.

Before we could complete updating the controller, SPANNA crashed.
While it is re-booting, I''m going to try.

The trick is going to be to give it a time-constant of 10. in each row,
then change the common factor to 9.0 on the time_constant window. OK
everything and LOAD/CLEAN
Mar 29 2000 01:44:59:857PM1000329002Robert GranetzWe''ve had several hours of computer crashes, flux estimator code work,
PCS estimator updating, PCS bug recovery, more computer crashes, etc.
It''s now 13:45 and we''re finally ready to go again.
Mar 29 2000 01:48:40:153PM1000329002Steve WolfeDoing a new coldstart, since computers have been crashing (and
there is a new tree there anyway).

OK, it coldstarted ok.
I couldn''t look at the log file because my quota got drained by
a 5.5Mbyte mail message.
Mar 29 2000 03:11:32:530PM1000329002Robert GranetzShot 02 -- Plasma. Ip=0.8 MA, hard disruption at t=1.32 s. Probably ran
into the recently decreased PLC current limits on the OH supplies.

Note: F28 broke at the disruption. It is now electrically open.
F29 is still okay.

There will be a long break while Steve reprograms the hybrid to use F27
instead of F28. I will change the processed flux accordingly.
During the break, we have have ECDC on as much as possible.

Next shot: increase OH1 PLC limit from 22 to 27 kA, and the
OH2's from 22 to 26 kA. (I'm referring to the limits on the
reverse bridge.) Swap F27 for F28 in the hybrid
estimators and the processed fluxes.
Mar 29 2000 02:41:24:250PM1000329003Robert GranetzShot 03 -- Fizzle. Br0 looked a little high, but the null looks very
different, and I don''t know whether it''s due to the new flux
loops in the analysis or in the hybrid or both.

Next shot: reduce Br0 by 1 mT.
Mar 29 2000 03:04:38:653PM1000329004Robert GranetzShot 04 -- Plasma. Disrupted at t=0.3 s. Br0 doesn''t seem to have changed.
The startup hit something and almost didn''t make it. Looks like
there were lots of runaways.

Next shot: increase early gas fueling.
Mar 29 2000 03:04:45:763PM1000329004Robert GranetzNote: there might have been a bitbus problem, which means that the preceding
changes for shot 04 might not have actually gotten to the hybrid. I just
did a coldstart, successfully.
Mar 29 2000 03:22:24:513PM1000329005Robert GranetzShot 05 -- Fizzle. Rather weak. Br0 looks good. Breakdown traces look very
similar to shot 02.

Next shot: retry with no changes
Mar 29 2000 03:30:48:327PM1000329006Robert GranetzShot 06 -- Fizzle. Rather weak again. It seemed to like having Br0 slightly
positive (at least according to our latest calculations).

Next shot: increase Br0 back up to 0.000 T.
Mar 29 2000 03:52:45:280PM1000329007Robert GranetzShot 07 -- Plasma. Died at t=0.21 s, after several disruptions. Excellent
current rise. Can't see anything obviously wrong, except for
lots of dirt.

Next shot: repeat with no changes.
Mar 29 2000 04:04:17:187PM1000329008Robert GranetzShot 08 -- Fizzle.

Next shot: repeat with no changes.
Mar 29 2000 04:28:26:937PM1000329009Robert GranetzShot 09 -- Fizzle. A little too much Bz.

Next shot: add 15 amps to IC_EF4 in segment 1.
Mar 29 2000 04:47:51:577PM1000329010Robert GranetzShot 10 -- Plasma. Hard disruption at t=0.38 s. The recon_vac:v_loop is
broken, and has been broken since we swapped F27 for F28 after
shot 02. This shot was much better behaved than the previous
short plasmas.

Next shot: no changes.
Mar 29 2000 05:00:15:310PM1000329011Robert GranetzShot 11 -- Fizzle.

End of run.

Engineering Operator Comments
ShotTimeTypeStatusComment
109:37:13:293AMPlasmaOkno faults/plasma
209:51:16:670AMPlasmaOkno faults/plasma
302:20:28:687PMPlasmaOkFizzle
402:43:35:280PMPlasmaOk
503:11:04:530PMPlasmaOkNo Plasma
603:25:59:793PMPlasmaOkNo Plasma
703:43:07:280PMPlasmaOkSupplies Okay
803:58:02:373PMPlasmaOkSupplies okay
904:22:11:700PMPlasmaOkNo Plasma
1004:36:30:187PMPlasmaOk
1104:58:43:013PMPlasmaOk