Alcator C-Mod Run 931005 Information

Back to runs

Miniproposals
Miniproposal:
Date Filed:
Title:
First Author:
Session Leader:Ian Hutchinson (shots 1-29)

Operators
Session leader(s):Ian Hutchinson
Physics operator(s):Steve Wolfe
Engineering operator(s):Steve Fairfax

Engineering Operator Run Comment
Power Supply/Coil Tests--Reverse current plasma

Session Leader Plans

Physics Operators Plans

Session Leader Summaries
Entered: Jul 7 2004 03:49:09:040PM
Author: To Be Determined
Run of 5 Oct 93

Hutchinson, Wolfe, Fairfax.

Reversed field and current operation.

This run was intended to establish reversed field and current operation.
The objective was well achieved. Breakdown was obtained on the second attempt
and full plasma on the ninth. We probably could have got plasma earlier but we
decided to implement radial feedback earlier in the current rise, ending up at
10ms versus 20ms previously. This required us to null out the offsets in the
hybrid on this line substantially better than before. The pay off was that we
obtained current rises that did not suffer the bouncing on the outer limiter.
Indeed, the subsequent fizzles were all clearly caused by late breakdowns.
Thus the current ramp is now more controlled, although there is a remaining
problem of _breakdown_ reproducibility, that has not usually been our problem.

The plasmas were run with the toroidal field ramping down from 0.1s. This
gives access to lower q at the end of the shot. The ramp is such that the peak
field is 4.8T and it reaches 2.9 T at 1 second, 60% of the peak.

Plasma durations were extended to take the flat-top to one second. No obvious
problems were encountered, except that we have not quite tuned up the shaping
at the end to avoid disruptions. The elongation should be reduced more to
avoid these.

Many plasmas started up in what I would interpret as sharkfin mode. The
density rises rapidly, but the plasma appears detached. There is then a
transition, sometimes a disruption, around 0.2s to a more expanded plasma.
This is generally accompanied by a density drop. Nevertheless, neL of 6e19
were sustained during the flattop. We were not pushing on this.

The marfe, when present, appeared to be predominantly below the midplane. This
would seem to confirm that the grad-B drift (now down) is an important factor
in determining its location. This may well be related to the observed
dependence of the in/out power balance on the grad-B drift in single null
divertors.

Detailed notes:

Computers suffering major problems due to messed-up restore of crashed disk.
Struggling with PCS because of unmatched versions etc.
Shots 1-7 power supply tests under PLC control.
New PID board in hybrid.
p_0=1.2e-5 torr. Deuterium.

shot result
8 dud Big OH2 imbalance. Bz about -10mT. BR about +11 mT.
Compared the obtained signals with those of 931001021. Biggest difs:
EF2u/l -1748/-1843 (vs 1726/1723) and changing fast at commutation.
EF4 -1120 vs 1091, a 30 A difference.
OH2 -19508/-23630 vs 21214/21728. Average is within 200A of correct.
Set BR offset to 0 from 9.
9 (fiz) Breakdown about 50 ms. BR=-2mT, Bz=-14mT @ t=0.
EF2 -1755/-1838, EF4 1128.
OH2 -20609/-22530.
Set EF4 from -1165 to 1065. Br offset to -3.
10 fiz Breakdown at 12ms. Bounce on outside at 30kA.. EF2s still -1755/-1833
got Bz -6 mT.
Set EF2U current demand to -1700 from -1750. Set Ef4 to -1000A.
11 fiz Breakdown at 5.5ms. Bounced several times at about 40kA.
got EF2 -1759/-1795
Set EF2U to -1660 RCUR P to 0ms @ 2. OH1 to 130V from 160V.
12 fiz Bz does not swing at all. It looks as if the feedback has overwhelmed
the programming.
Set V OH1 to 100V at 5ms.
13 fiz Bz still does not swing. Slight Br problem.
Set P on rcur to 10ms. Raise Br 0.5 mT.
14 fiz 30ms breakdown as the feedback pulls Bz back down!
Set TF invert time to .1s for faster turn-around and field ramp.
Set minus_one_volt*(-7000) in rcur predictor.
15 fiz RCUR offset in the wrong direction. Slams out at 20ms. 50kA.
Set +10000*minus_one_volt in rcur.
16 pla Rather high kappa. VDE at 0.550s. Good current rise. Ip=660kA.
Set zcur2 P=0.5 from 1. zcur3 p=9. from 6. Dropout to .76-.765.
17 pla VDE at 380ms.
Set zcur2 P=1. Set clearin P=1 (from 0.2). EF2U to -3000 form -2500.
18 pla VDE at 500.
Set OH2 V=0 at 50ms from 5V.
R demand to .68 from .67.
19 pla 1 second duration.
Set ef2u to -3500
Gas to 40V at .2-.3s.
20 PSD EF2U only 1300. Caused by the scalebit in PCS which was fiddled with
when we made the previous change.
Clearin gain to P=4. OH2V to -5V at 50ms.
21 PSD Same again.
Reload shot 19
Set clearin gain to P=4, OH2v to -5V at 50ms.
22 Dud Not at all clear why this one did not go. Fields look ok.
Set p_o=1.3e-5 from 1.2.
23 Pla 1 second flattop. VDE at 1.01s. Might be the surrent limit in OH2.
OH2 V to -10V at 50ms. Reduced EF1 currents at .4s. Clearin to -2cm
at 1.0s.
24 fiz Late breakdown at 8ms.
25 Pla 1s VDE. Small Disruption at .2s. Discovered P and D turn-offs at 1s.
OOPS! Early (20&30ms)disruptions due to high zmeter, but we survive.
Set PIDS to 1.5s or more.
26 Pla Lasted till 1.1s.
Reduced the gas.
27 Pla Disruption at 63ms, about 100kA. VERY low density:1-2e19. Moly.
Nonthermal ECE.
Reloaded the gas from 26 and tweaked at 150ms.
28 Pla Similar to 26
Ask for .8MA plasma current. Help OH2V up at the end to lower
elongation.
29 fiz Late breakdown at 12ms.

Physics Operator Summaries
Entered: Jul 7 2004 04:36:50:410PM
Author: To Be Determined
Physics Operator Summary for 931005
Session Leader: IHH
Phys. Op.: smw
Eng. Op. : sf

The purpose of this run was to establish operation with negative toroidal
fields and plasma (and PF) current. This was accomplished, despite some
significant difficulties. A secondary goal was to run reduced field, pointing
towards an attempt at achieving ohmic H-mode in a future run. This was
accomplished to the extent of ramping the TF down during the shots, with
inversion at 0.1 sec. Values of q95 in the range of 3, and EFIT li values <1.2
were obtained.

The bus connections had been reversed over the weekend and on Monday, 10/4.
Tests on Monday had to be aborted due to computer problems, so the run began
with power system tests, starting around 7AM. The first 7 shots were power
tests under PLC control. The first plasma attempt was #8, a weak late fizzle
was obtained on #9, and the first successful plasma was #16. The remainder of
the run was spent tuning and extending the plasma. The pulse length was
extended to end the flattop at 1sec on #23, and the longer plasma was tweaked
up over the remainder of the run. An unsuccessful (fizzle) attempt to raise
the plasma current to (-)800kA was made on #29, the last shot of the day.

Owing to the loss of files following the disk crash on 10/4, most of the
preparatory work for this run was lost and had to be re-done at the beginning
of the day. The tree changes were as follows:

1)The HYBRID.CONFIG.OUTPUTS:OUTPUT_nn:GAIN values corresponding to OH1,OH2U&L,
EF1U&L,EF2U&L,EF3,EF4, and CHOPPER, were changed to minus their previous
values. CONFIG_ID was modified to announce the change.
2) A PCS file was prepared (931004999) based on 931001021 in which all the S
and V waveforms coresponding to PF currents and voltages, and also the \S_IP
waveform, were inverted; this was accomplished by PASTEing from a pair of
SCOPEs called PCS_MINUS_{S,V}.DAT in CMOD$MODELS:[HYBRID.SCOPES], and was done
on Monday. The Scopes were lost in the crash and rebuilt this morning.
3) The draw limits in the PCS_931004999 tree were changed as follows:
\S_IP:LIMITS = [-3.e6,3.e6] (was [0.,1.e6])
\V_EF2_*:LIMITS = [-1050,1050]
\V_OH2_*:LIMITS = [-250,250]
4) The resulting PCS file was loaded using PCS, after "touching" all the
EDIT_Ms to insure the new \HYBRID.CONFIG.OUTPUT:*:GAIN values were used.

A significant amount of tweaking had to be done to get the null right using the
reversed fields. In particular:
The offset (MINUS_1_VOLT coefficient) on BR_0 was changed from +9e-3
to -3e-3

The EF4 current had to be reduced in magnitude from -1165 to -1000A

The OH1 voltage at 5msec was changed from +160V to +100V

Additional tweaking included adding an offset (MINUS_1_VOLT coef.) to the RCUR
predictor of +10000 and moving the on-time of this PID to 10msec to help
control the position during early current rise. The IC_EF2U waveform was
adjusted to balance the EF2 currents; this required a total change of 90A,
from -1750 to -1660. Also, the V_OH2 and IC_EF1 waveformns were tweaked to
improve the shape and restrict elongation to stable range. For the long shots,
it was necessary to extend the ZCUR PID times past 1.0 sec to >1.5sec.

Problems:
---------
The CMOD$MODELS disk crashed on 10/4/93. The restores from incremental backups
(nominally vintage 9/29) were done incorrectly, resulting in some obsolete and
inconsistent trees and software being present. These problems were dealt with
during the course of the run, as they became noticed. The PCS software (PROs
in [HYBRID.OCS.IDL]) had several problems attributed to the restoration of
obsolete or inconsistent versions. It is believed, not with extreme
confidence, that the significant lossages have been remedied as of the end of
the run. However, some problem files may remain.

The EF4 bus rogowskis were damaged (leads ripped off) during the bus
reconfiguration. These were fixed during the power tests, but some change in
pickup may have resulted.

The BITBUS INIT actions failed on the first PLC power test; this seemed to
have been remedied by a coldstart before the second shot.

The PID card in the Hybrid was damaged on Monday, following installation of
new circuitry to allow "INTEGRAL" feedback on the even-numbered wires. A spare
(without the new circuitry) was installed on Monday and used throughout the
run without problems. The new (spare) card may have had some different offset
or noise properties which could have contributed to the necessity to tweak
offsets.

FFIT was unable to cope with the reversed field, and gave erroneous values for
kappa (among other things) leading to problems with the Zeff calculation. EFIT
seemed to work correctly, although the number of time points must be increased
to accomodate the longer shots.

The radial field at 0:10msec seemed less reproducible than in recent runs,
leading to some duds and fizzles after a nominal startup configuration had
been reached.

Score Card:
-----------
Tests: 7
Duds: 4 (2 due to software)
Fizzles:10 (including one very short runaway shot, #27)
Plasmas: 8
--
Total 29

Session Leader Comments

Physics Operator Comments

Engineering Operator Comments
ShotTimeTypeStatusComment
107:09:15:790AMTestOkCoil reg. & power supply test ( EF4 Reg. not working).
208:05:52:650AMTestOkEF4 reg Now ok.
308:13:55:370AMTestOkEF2U only.
408:19:28:320AMTestOkEF2 L test O.K.
508:27:53:580AMTestOkPower supplies OK.
608:35:20:960AMTestOkNo faults.
708:44:02:770AMTestOkPower supplies O.K. EF1 U&L 6KA.
810:18:48:390AMPlasmaOkNo Faults. No plasma either.
911:18:05:760AMPlasmaOkFlash. All power supplies OK
1011:37:58:400AMPlasmaOkBetter Flash.
1111:56:28:020AMPlasmaOkFizzle
1212:25:45:270PMPlasmaOkFizzle
1312:42:28:360PMPlasmaOkFizzle
1401:02:05:780PMPlasmaOkFizzle; move TF invert to 0.1 seconds to speed up shots
1501:19:47:610PMPlasmaOkFizzle
1601:34:35:430PMPlasmaOkPlasma
1702:00:46:050PMPlasmaOkDisruption at 640 kA, 0.4 sec
1802:18:33:430PMPlasmaOk650 kA disruption at 0.5 sec
1902:31:59:340PMPlasmaOkNice 1-sec, 650 kA Plasma. Disrupts at 200 kA
2002:50:22:560PMTestOkDud . Hyvbrid computer/PCS bug caused wrong EF2 currents
2103:01:26:140PMPlasmaOkDud. Same PCS Bug.
2203:19:52:310PMPlasmaOkDud. Don't know why. This was a reload of shot 19.
2303:35:29:280PMPlasmaOkPlasma. Longer; disrupted at 1 sec, 600 kA
2403:56:57:450PMPlasmaOkFizzle. Stressed TF core in shear a bit. Lowered OH1 negative current
2504:08:42:170PMPlasmaOkDisrupted at 1 second;
2604:23:25:210PMPlasmaOkPlasma made it further, disrupted at 1.1 sec, 450 kA
2704:37:58:090PMPlasmaOkDeath by injection at 56 msec
2804:52:10:220PMPlasmaOkDisruption at 560 kA
2905:04:09:770PMPlasmaOkFizzle