Alcator C-Mod Run 940503 Information

Back to runs

Miniproposals
Miniproposal:
Date Filed:
Title:
First Author:
Session Leader:Steve Wolfe (shots 1-16)

Operators
Session leader(s):Steve Wolfe
Physics operator(s):Steve Fairfax
Engineering operator(s):Joe Daigle,Vinny Bertolino

Engineering Operator Run Comment
fix power supplies

Session Leader Plans

Physics Operators Plans

Session Leader Summaries
Entered: Jul 7 2004 03:49:09:727PM
Author: To Be Determined
Session leader's report on run of 940503

SL: Wolfe
PO: Fairfax
EO: Bertolino, Daigle

Miniproposal #041 - CMOD Startup

The plan was to first verify that the problem with OH2L not following its
demand signal was fixed, then proceed to get a null and tune it upo to get
plasma. As for the past several runs, the ultimate model was 931020003, with
all the predictors modified to account for the dead F17 loop, and with the
current ramp flattopped at 400kA.

The good news is that the OH2L supply does indeed seem to be fixed. The power
supply seemed to be following its demand signal well, except that the maximum
voltage is still less than the supply rating. The bad news is, although we
got fairly reliable breakdowns, we did not get the current to take off.

We got an acceptable null on shot#2 (#1 had a fault early), and got a
breakdown at 11msec. Tuning of Br got the breakdown earlier (8msec) on shot
#3. The null looked fair. The next few shots were screwed up by a bad flux
loop signal from F25. We actually prepared a shot with all predictors having
F25 weight set to zero and adding half the F25 weight to the weights of F24A
and F26. However, Granetz found a possible cause for the problem (a stray
strand of wire at the integrator input) and so we took shot #6 without loading
the hoked up predictors, and this worked. (The hokey shot was saved in
[Fairfax.mds]PCS940503006.*;)

Subsequent tuning seemed to favor a more positive Bz (30amps on EF4). The
breakdown moved as early as 5.4msec, but did not take off. Dropping the
pressure as low as 1.0e-5 Torr seemed to improve the burnout, but still did
not give a current takeoff. We also tried (shot 12 and thereafter) putting the
offset on RCUR predictor to 10000 (the value it used to have); this seemed to
make the evolution of the fields better to my eye, but the plasma got somewhat
shorter. We also had one more flux loop problem, on F23 on shot 13. F25 was
still somewhat suspect.

I think we don't yet have the post-null evolution tuned up right. Also, it
appears possible that lowering th pressure further could be helpful. The
ratiomatic is reading higher than the torus gage; it is possible the torus
gage is reading too low compared to last year.

The problems with intermittant flux loop signals are troubling. After the run,
Bob interchanged the wiring on F25 and F24 (F24 is not used, F24A is) in case
the problem was in the integrator or connectors. F25 is known to be shorted to
the vessel, but this problem may be external. F23 is not shorted, and the
evidence points to a flakey integrator.

The power supplies behaved well. One shot (#1) was lost to an early fault
indication, apparently spurious. There were a couple more problems after the
plasma which did not impact the run.

Despite our good intentions, the run did not start until 10:00. The alternator
took longer than expected and the machine was not cold until 9:40. Apparently
Monday's "cooldown" was done in bakeout mode, which did not do anything to the
structure temperature.


Scorecard:
Duds 5
Fizzles 11
Good 0
----------
Total 16

Physics Operator Summaries
Entered: Jul 7 2004 04:36:51:003PM
Author: To Be Determined
Physics Operator Summary May 5, 1994

Miniproposal 41

Session Leader: Wolfe
Physics Operator: Fairfax
Eng. Operator: Bertolino
State Operator: Pierson
Cryo Operator: Sorci
Power Systems: Paranay

Delayed 1 hour by alternator. Procedure error in warming yesterday.

Start with 940429008 loaded into hybrid

Shot 1 - OH2U commutation failure Early

Shot 2 Flash. Breakdown at 11 msec.
Changed Br offset by -1 mT
Delayed 15 minutes by warm structure; procedure error in cooling
yesterday.

Shot 3 Flash. Breakdown at 8 msec.
Br is better; responded exactly as programmed.
Increased EF4 current by 30 amps (about 1 mT) from -1012.5 to -982.5 A

Shot 4 No flash. Br and Bz off. Not clear what is wrong.
reload shot 3 and try again

Shot 5 No flash. OH2 currents quite different from shot 3, but supplies
appear to be working OK. Check of flux loops shows 25 is reading only
about 75% of it's value on shot 3.

Steve Wolfe tried to fake the F25 signal as the average of F24a and
F26 signals. Bob Granetz and I went into the cell and checked the
F25 resistance. F25 is shorted to the machine. We found a small
wire touching the connector body on the cable leading to the
integrator input.

Next shot will use the matrix from shot 3; Steve Wolfe's patches are
finished but not yet loaded.

20 minutes delay while recooling coils due to warm structure.

Shot 6 Flash. Flux 25 is good. Saved Steve's patch in [fairfax.hybrid] as
shot 940503006 (even though it's not.)

Changed EF4 current precharge from -1012.5 to -952.5 amps (about 2 mT)

NOTE: "Compose Character" is not the ALT key on PHYSOP as it was
on XHARPO. The "Alt Function" key is ALT.

Shot 7 No flash. Bz slightly positive.
Changed EF4 current precharge from -952.5 amps to -982.5 (about 1 mT)

Shot 8 Flash. Breakdown at 6.5 msec; better.
Set Br another mT lower. Offset now -2 mT.

Shot 9 Flash. Breakdown at 5.5 msec.
Changed pulse gas program back to -10 msec from -50 msec.

Shot 10 Flash. Basically same breakdown.
Changed fill pressure from 1.4 to 1.2e-5

Shot 11 Flash.
Added +10000 offset to RCUR

Shot 12 Flash. Lowered gas fill to 1.1e-5

Shot 13 no flash. Flux loop 23 now screwed up.
Horne doesn't have to shuts up yet.
No changes for next shot.

Shot 14 Flash. Flux loops seem OK.
Artur says diverter guage shows gas puff at 4 msec; he's wrong.(it's 40)
Delay pulsed gas by another 10 msec anyway.

Shot 15 Flash. Marginally worse breakdown.
Copied OH2U voltage prog to OH2L (there was a 20 v difference after
commutation.) Dropped fill pressure to 1.0e-5.

Shot 16 Flash. Deliberate crowbar of TF for Gwinn.


Session Leader Comments
May 3 1994 08:40:23:010AMSteve WolfeRun 940503
Miniproposal #041 : Startup C-MOD

SL: Wolfe
PO: Fairfax
EO: Bertolino

P before ECDC = 5.8e-8
Discharge cleaned in D2 for 3.5 hours, starting at 04:30.

Hybrid program:start from 940429008, before last Friday's "adjustments",
which were basically trying to compensate for a problem in OH2L.

Plan:
Get plasma. If first try indicates OH2L problem is still there,
go to PLan B -- fix the supply first, then get plasma.

Once breakdown and current rise are established, adjust radial position
inward to clear limiter. Then extend pulse length.

May 3 1994 09:32:46:610AMSteve WolfeWe will be delayed until alternator and cooling come up. All coils are
showing green except TF. The TF vertical arms aer lagging. The cylinder
is quite warm; yesterday's cooling did nothing noticeable to structure
or dome temperatures. The cooling was done in BAKEOUT mode which just
keeps the coils safe, does not really cool.

May 3 1994 10:10:56:960AM940503001Steve WolfeFirst try for plasma. P=1.4e-5T, program loaded from 9404290008

OH2U faulted at -.82 sec.

May 3 1994 10:40:10:830AM940503002Steve WolfeGot a flash.

OH2L seems ok - follows its demand.
Breakdown at 11msec.
Br = 1mT , Bz= -5mT

For next shot change Br offset to -.001mT, don't worry about Bz for now.

May 3 1994 11:13:21:060AM940503003Steve WolfeGot a flash.
PF breaker pulled out late.

Breakdown at 8.1msec
Null doesn't look too horrible. Some Br after breakdown, and Bz is -5mT.

For next shot, make EF4 30A more positive.

May 3 1994 11:41:54:410AM940503004Steve WolfeNo flash.
Big radial field, where the hell did that come from?

There is a 300 amp oscillation on the OH2's. Also the OH2L is now more
negative than OH2U by 2000A; it was the other way around on shots 2&3!

For next shot, reload shot 3

May 3 1994 11:49:11:800AM940503005Steve WolfeNo flash.
Big radial field, where the hell did that come from?
Similar to 4

May 3 1994 11:59:18:540AM940503005Steve WolfeNo flash.
Big radial field, where the hell did that come from?
Similar to 4

Flux loop 25 (F25) is reading low! It's died since shot 3.
It's reading 1.07 instead of 1.27
Can try to fix in software, Bob should look at it after the run.

May 3 1994 01:07:45:010PM940503005Steve WolfeQuick and dirty fixup for F25 dead in predictors. Replace F25 by
0.5*(F24A+F26). Do thins manually in the CUSTOM predictors as follows:

RCUR F25=-21998.9, F24A=-8859.39->-19858.8
F26=-35558.6 -> -46558

ZCUR F25=-38522, F24A= -38949.3 -> -58225.3
F26= -4044.16 -> -23320.2

ZXL F25=-42468.5 F24A= -26993.3 -> -48167.6
F26 = -18424.9 -> -39658.25

RXL F25=-53095.9 F24A= -39502.0 -> -66049.95
F26 = -50611.2 -> -77158.95

Clearin F25=18042.0 F24A= 48204.0 -> 57225.0
F26= -68957.7 -> -59936.7

BR_0 F25=9.359961e-2 F24A= 3.830151e-2 -> 8.51013e-2
F26= .104361 -> .151161

Needed to change factor on BR_0 from 40. to 25.
Also changed RXL factor from 7e-5 to 5e-5.

Set up these changes, but don't load for next shot. Granetz found a
possible bad connection, so we'll see if it works now.

May 3 1994 01:32:02:200PM940503006Steve WolfeGot a flash back. Did not use the previous hacks, Bob may have fixed
it. Flux loop looks ok, maybe. Changes saved in
[FAIRFAX.HYBRID] PCS_940503006*

Breakdown at 7.8msec

For next shot, try again to make EF4 more positive by 60Amps.

May 3 1994 02:03:17:980PM940503007Steve WolfeNo flash
MFLUX says no null. Vacuum says Bz=0 from 0 to 6 msec

Try setting EF4 more negative by 30 Amps.
There may still be a problem on F25. It looks like it didn't come back
to zero this time.

May 3 1994 02:17:13:050PM940503008Steve WolfeGot the flash back.
Breakdown at 6msec

MFLUX looks like it's having a positive BR at 10-15 msec

Try changing Br_0 offset from -.001 to -0.002 T

May 3 1994 02:37:59:960PM940503009Steve WolfeFlash is back
Breakdown at 5.4msec

Delay pulse gas 3 back to -10msec (from where Joe had moved it last
week)

May 3 1994 03:01:30:180PM940503010Steve WolfeFlash is back
Breakdown at 5.4msec

Light lasts longer, but no more current.

For next shot change P from 1.4e-5 to 1.2e-5

May 3 1994 03:15:30:980PM940503011Steve WolfeFlash is back
Breakdown at 5.7msec

Light lasts longer, but no more current.

No obvious change.

For next shot change offset on RCUR from 0. to 10000.

May 3 1994 03:22:17:570PM940503011Steve WolfeFlash is back
Breakdown at 5.7msec

Light lasts longer, but no more current.

No obvious change.

For next shot change offset on RCUR from 0. to 10000.

May 3 1994 03:43:40:980PM940503012Steve WolfeFlash is different. Shorter, and maybe a msec or two later.

Lower fill to 1.1e-5

May 3 1994 03:50:09:270PM940503012Steve WolfeFlash is different. Shorter, and maybe a msec or two later.

Lower fill to 1.1e-5

May 3 1994 03:58:35:790PM940503013Steve WolfeLost the breakdown.

Integrator problem on flux loop 23 ????
Maybe something different on 25

Try again.

May 3 1994 04:15:21:820PM940503014Steve WolfeGot the breakdown back.

Try delaying the gas another 10msec (to 0.)

May 3 1994 04:17:22:850PM940503014Steve WolfeGot the breakdown back.

Try delaying the gas another 10msec (to 0.)

May 3 1994 04:32:01:940PM940503015Steve WolfeTry delaying the gas another 10msec (to 0.)

No obvious difference.

Try removing the differnece in the OH2 V programming
(copy upper into lower.)

May 3 1994 04:33:53:000PM940503015Steve WolfeTry delaying the gas another 10msec (to 0.)

No obvious difference.

Try removing the differnece in the OH2 V programming
(copy upper into lower.)

Also lower P to 1.0e-5

May 3 1994 04:56:46:760PM940503016Steve WolfeBurnout is better, but still no takeoff.

This was the last shot of the day. Intentional E-stop hit to give
Gwinn some data on alternator shaft vibration.

Physics Operator Comments

Engineering Operator Comments
ShotTimeTypeStatusComment
110:05:49:850AMPlasmaBadoh2u comm fault
210:29:21:070AMPlasmaOkno faults
310:57:11:920AMPlasmaOkoh2u comm fault after shot
411:16:33:160AMPlasmaOkno faults
511:43:41:590AMPlasmaOkno faults
601:19:58:050PMPlasmaOkno faults
701:53:30:520PMPlasmaOkno faults
802:08:01:390PMPlasmaOkno faults
902:28:37:560PMPlasmaOkno faults
1002:49:59:970PMPlasmaOkno faults
1103:09:15:160PMPlasmaOkno faults
1203:27:55:440PMPlasmaOkno faults
1303:48:00:740PMPlasmaOkno faults
1404:07:01:970PMPlasmaOkno faults
1504:25:19:000PMPlasmaOkno faults
1604:45:28:170PMPlasmaOke-stop at end of the shot to check alernator rpm data