Alcator C-Mod Run 1050722 Information

Back to runs

Miniproposals
Miniproposal:424
Date Filed: 5/12/2005
Title:Initial tests of disruption mitigation using a gas jet
First Author:Robert Granetz
Session Leader:Robert Granetz (shots 1-26)

Operators
Session leader(s):Robert Granetz
Physics operator(s):Steve Wolfe
Engineering operator(s):Bill Cochran,Bill Byford

Engineering Operator Run Comment
MP #424 -- Gas jet disruption mitigation

Session Leader Plans
Entered: Jul 21 2005 06:09:07:650PM
Author: Robert Granetz
Same as yesterday (1050721).

Physics Operators Plans
Entered: Jul 21 2005 05:17:25:663PM
Author: Steve Wolfe
MP #424 -- Gas jet disruption mitigation

Session leaders: R. Granetz, D. Whyte
Physics operator: S. Wolfe

NO ECDC 
Run begins at 09:00 and ends at 17:00.

Power systems as on 1050721001 (1.0 MA, 5.2 T, SNL)

A-coil configuration: normal [+Dtop -Dbot -Jtop +Jbot] and enabled.

Gas Setup:
   fill B-Top            with  6 psi D2  Hybrid enabled  (PG4)
   fill B-side lower     with  (1 psi Ar + 2psi Ne) Hybrid Enabled (PG1)
  leave B-side upper     with 15 psi He3 Hybrid disabled (PG2)
   fill B-main (C-side)  with 30 psi D2  Hybrid enabled  (PG3)
No NINJA for this run.

Gas jet should have high-pressure ( >70 bar) neon bottle installed, WITH
HIGH-PRESSURE HELIUM AND ARGON BOTTLES READY TO BE SWITCHED IN DURING THE
RUN.

Torvac setup: unnecessary diagnostic gate valves (ECE) should be kept
closed on all gas jet shots.

Enable the following, assuming no vacuum problems:
Z-bolo shutters, VUV (yes, enable the VUV gate valve- per JT)

Necessary diagnostics:
  IR camera
  Fast PSI camera on B-horizontal, starting at t=0.800 s, Ne+0, He+0, Ar+0
    filters
  Thomson scattering, with adjustable timing between the two lasers
  Bolometry
  AXUV array with reduced gain
  Halo current Rogowskis
  SX arrays with fast sampling window (t=0.795 to 0.820 s)
  Mirnov Bp fluctuations

-------------------------------------------------------------------------------
ICRF Setup: All transmitters in run on at 08:30, but begin the run with
            ohmic discharges.

-------------------------------------------------------------------------------
PCS: Load from 1050721009 (which includes changes after shot 1, 
     which never got used on Thursday).
----------------------------------------------------------------------
Run plan: Start with neon at 70 bar, 2 ms.  Get IR divertor images, AXUV
          profiles, global bolometry, fast visible camera images, as well
          as disruption quench data (halo currents, timescales).  Then turn
          on RF, He3 puff, to get higher Wth and pressure -- same drill.

          About halfway through the run, switch to helium at 70 bar, 2 ms.
          Get VUV data to look at self-absorption issues.

          Save the last hour of the run for argon.  Compare quench times
          with helium and neon. 



Session Leader Summaries
Entered: Jul 22 2005 05:20:12:840PM
Author: Robert Granetz
Session leader summary for Friday 22 July 2005:

MP #424 -- Gas jet disruption mitigation

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

The gas jet system worked well after some initial problems at the start of
the run. Except for early in the run, disruption recovery was very good.
We used neon gas jets for the first half of the day, then switched to argon
jets for the remainder of the day. The timing of the gas-jet-induced
disruptions was very reproducible, resulting in a good sets of Thomson
profiles for both gases. The argon current quenches consistently started 1
ms later than the neon quenches, and the quench duration was marginally
shorter. Halo current with argon were also somewhat less than neon. The
TS temperature profiles for argon indicate that the gas barely gets into
the plasma at all, whereas the neon gets in to about r/a=0.85. Fast
visible images of the neon and argon jets are quite different than helium
(from 1050603). We obtained one good set of IR images of the divertor
region on shot 11, and this will be analyzed to determine the thermal
energy deposition on the tile surfaces. The reduced gains of the AXUV
array worked well, and none of the signals saturated. The radiated power
profiles for neon and argon were quite different. We had 2+ MW of RF power
(mostly 2 MW from J-port at 50 MHz), but only got a maximum Wth of 100 kJ.
Higher Wth and kinetic pressures will be the goal of the next gas jet
mitigation run.

Physics Operator Summaries
Entered: Jul 22 2005 05:37:07:917PM
Author: Steve Wolfe
The machine ran well today, especially considering 
the number of (intentional) disruptions.
We completed the Neon and Argon gas jet experiments, 
and got most of the data wanted. The IR camera was 
problematic, but collected images on one shot.
JT did not turn on the voltage on the McPherson. 
We did not go back to helium jets. 

The run got off to a shaky start with problems with 
the GV1 valve on the gas jet system. This was apparently
not staying open long enough. After several shots and 
a couple of manual tests, this was remedied by increasing
the time it was programmed to stay on, although this had 
been working yesterday and on June 3. At any rate, after
the change it worked reliably for the remainder of the 
run. 

The GPC1 dewar froze up during filling, and had the tube stuck in the
neck until after the run. It may be ok now. 

Startups were pretty good today. There were two duds early on, 
possibly due to gas effects (or maybe not). There was one very
late breakdown (15ms) that ran fine. There was also a one TF malfunction
resulting in a dud (#24), and one ack failure on CHECK->PULSE resulting in 
no shot (#12). 

At shot#16, Yijun requested a reduction in the outer gap to improve
the RF coupling. This was the only segment 2 PCS change associated
with the run all day. 

I kept busy by testing and somewhat optimizing the new Acoil control
based on feedback control of the total (calculated) 2/1 error field
projected onto the Acoil orientation. This was quite successful, 
and I recommend that it become standard for normal runs. The setup
(shot 1050722023 and after) is: 

Wire 9 (non-Ip reference) was changed to "B21_PROJ" with the 
algorithm "nonaxi_smw". The factor value is set to 10000.
Target waveform is zero; note that as long as this stays zeroed,
it doesn't matter whether the wire is Ip referenced or not.
PID Gains are P=1.0 @0.2sec, D=0.0, I=4 @0.2sec, and the integrator C 
gate turns on at 0.3sec, which is also when the coil is enabled; it 
is important not to turn on the integrator before the coil is 
enabled to avoid a thrash at turn on.  
The controller has a coefficient of 3.e6 on ACOIL_CUR and a "time
constant"=1.0.
The feedforward waveform on the ACOIL_CUR (V10) has been set to zero,
so the coil is completely controlled by feedback. The resulting
(calculated) error field is kept below 0.5G for almost the entire 
shot using these settings. 

The other options in /usr/local/cmod/codes/pcs/user_algs/nonaxi_smw.pro 
have not yet been live-tested, but should work, probably with
similar gains and controllers. Note that the predictor needs to be 
recalculated for non-standard Acoil configurations.

Scorecard:
-----------
      Tests    0
      Duds     4 (1 power supply, 1 Ackfail)
      fizzles  0
      plasmas 22
      ------- --
      Shots   26
     

Session Leader Comments
Jul 22 2005 08:48:06:140AMRobert GranetzReady to go with neon jet and ICRF (J-port only) on the first shot.
Jul 22 2005 09:12:35:437AM1050722001Robert GranetzNeon gas jet fired, but it doesn't look like anything got into the machine.
It's almost as if GV1 didn't open.

I manually operated GV1 after the shot and it indicates that it is working.

Next shot: run GV2 and GV1 in manual override mode.
Jul 22 2005 09:39:13:593AM1050722002Robert GranetzAgain, the neon gas jet fired, and GV1 and GV2 had the correct state indications,
but nothing got into the machine.

After the shot I ran a manual gas jet cycle and it got 300 mTorr of neon into the
torus. The valves again gave the correct state indications.

Next shot: try one more time.
Jul 22 2005 10:54:46:220AM1050722004Robert Granetz
Shot 04 -- Good shot.  Neon gas jet worked.  I_halo=149 kA.  Current quench time
is about 2 ms.  Current quench started at about t=0.805, which is about 4 ms
earlier than neon shots from yesterday and 03 June.

THIS SHOT WORKED BECAUSE WE RAN THE GAS JET GATE VALVES IN MANUAL MODE, I.E. NOT
UNDER PLC/STATE CONTROL  (JOE AND YURI HAD TO MODIFY THE LOGIC JUST NOW TO ALLOW
THAT.)

Plenum gas: neon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 62
Plenum change in  pressure (bar):  8

Next shot: TS will move timing back by about 4 ms.

Jul 22 2005 10:57:13:343AM1050722005Robert GranetzFor this shot we went back to automatic mode (i.e. under PLC and state control).
The gas jet fired, but again nothing got into the machine. This time I think I
saw GV1 start to close just before the plasma began. Joe and Yuri are looking
at the PLC logic to see if that could be the problem.

Next shot: automatic mode, but set the GV1 close time to be 10 seconds (instead
of 7 seconds)
Jul 22 2005 11:52:21:157AM1050722007Robert Granetz
Shot 07 -- Good shot.  Neon gas jet worked.  I_halo= 136 kA.
Current quench starts at t=0.805, which is the same as shot 4.  Maybe this is because
the neon plenum pressure is higher than on 1050603.  Current quench duration is about
2 ms again.

Automatic mode with GV1 shut time = 10.0 s seems to be working reliably.

JT is adjusting the timing of the fast PSI camera, and will need at least one more
shot to pin it down.  We're getting some good Ne+0 frames even now.

The IR camera may also get images on the next shot.

Plenum gas: neon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 62
Plenum change in  pressure (bar):  8

Jul 22 2005 11:47:04:907AM1050722008Robert Granetz
Shot  08 -- Good shot.  Neon gas jet worked.  I_halo= 132 kA.
Current quench starts at t=0.805, even though there was slightly lower pressure in
the plenum (I was too busy to refill it.)

We saw real-time IR images on the video display, but it didn't get recorded.  JT is
working on that.

Plenum gas: neon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 62
Plenum post-pulse pressure (bar): 55
Plenum change in  pressure (bar):  7

Jul 22 2005 11:55:22:907AM1050722008Robert GranetzStarting with this shot, Joe and Yuri changed to TS PLC logic to allow the TS
gate valves to remain open up to 500 mTorr. This will help speed up the
shot cycle, since they no longer have to access the cell after
every shot to re-open their gate valves.
Jul 22 2005 12:10:59:673PM1050722009Robert Granetz
Shot 09 -- Good shot.  Neon gas jet worked.  I_halo=136 kA.
Current quench starts at t=0.805 s.  Neon plenum has not been refilled
for the last two shots, but the start of the quench is still at the same
time.

IR video look good, but we still can't get the data to record.

Plenum gas: neon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 55
Plenum post-pulse pressure (bar): 48
Plenum change in  pressure (bar):  7

Jul 22 2005 12:32:04:733PM1050722010Robert Granetz
Shot 10 -- Good shot.  Neon gas jet worked.  I_halo= 121 kA.
Current quench starts at t=0.805.
Current quench duration looks a little faster than 2 ms on this shot, and the
halo current is a little less than it has been.

IR video shows a glowing tungsten pin.  JT and Glen are getting close to having
the triggering and data acquisition working.  TS is getting good profiles at
a number of different times.

Plenum gas: neon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 63
Plenum change in  pressure (bar):  7

Jul 22 2005 12:50:38:127PM1050722011Robert Granetz
Shot 11 -- Good shot.  Neon gas jet worked.  I_halo=125 kA.
Current quench starts at t=0.805 s.
Current quench duration is about 2 ms, maybe slightly faster.

GOT THE IR IMAGES RECORDED THIS TIME !!  Leading edge of outboard
limiter clearly gets hot, and a tungsten pin gets really hot.  The
pin's cooling timescale is minutes.

Plenum gas: neon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 63
Plenum change in  pressure (bar):  7

Jul 22 2005 01:07:53:313PM1050722012Robert GranetzAction failure - no MDSplus cycle, no shot
Jul 22 2005 01:08:39:190PM1050722012Robert GranetzThe IR camera has failed and will not be on for the next shot. After the next
shot we'll switch to argon.
Jul 22 2005 01:23:44:297PM1050722013Robert Granetz
Shot 13 -- Good shot.  Neon gas jet worked.  I_halo= 142 kA.
Current quench starts at t=0.805
Current quench duration is about 2ms, maybe slightly slower.

No IR camera on this shot.

Plenum gas: neon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 63
Plenum change in  pressure (bar):  7

End of neon for today.  We'll have an extended cell access to install
the argon bottle.

Jul 22 2005 02:00:40:420PM1050722014Robert Granetz
Shot  -- Good shot.  Argon gas jet worked.  I_halo=161 kA.
Current quench starts at t=0.806 ms, about 1 ms later than with neon.
Current quench duration is about the same as neon, certainly not any faster.

IR camera is still not working.

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

Jul 22 2005 02:21:57:683PM1050722015Robert Granetz
Shot 15 -- Good shot.  Argon gas jet worked.  I_halo=138 kA.
Current quench starts at t=0.806 s
Current quench duration is about 2 ms

We're getting the fast PSI images with an Ar+0 filter.  AXUV array is not saturating.

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

Jul 22 2005 02:38:46:510PM1050722016Robert Granetz
Shot 16 -- Good shot.  Argon gas jet worked.  I_halo=105 kA.
Current quench starts at t=0.806 s.
Current quench duration is about 2 ms, maybe a bit faster.

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

Jul 22 2005 02:58:11:493PM1050722017Robert Granetz
Shot 17 -- Good shot.  Argon gas jet worked.  I_halo=130  kA.
Current quench starts at t=0.806 s.
Current quench duration is about 2 ms, maybe a bit faster.

Valerie reports that the TS profiles seem to indicate that the argon
is not getting in past the very edge of the plasma.  We're getting
fast PSI images with Ar+0 filter.  IR camera is still not working.

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

Jul 22 2005 03:07:16:853PM1050722018Robert Granetz
Shot 18 -- Good shot.  Argon gas jet worked.  I_halo=109 kA.
Current quench starts at t=0.806 s.
Current quench duration is slightly less than 2 ms.

Still no IR camera.

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

Jul 22 2005 03:22:38:417PM1050722019Robert Granetz
Shot 19 -- Good shot.  Argon gas jet worked.  I_halo=111 kA.
Current quench starts at t=0.806 s.
Current quench duration is about 1.7 ms

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

Jul 22 2005 03:37:07:463PM1050722020Robert Granetz
Shot 20 -- Good shot.  Argon gas jet worked.  I_halo=115 kA.
Current quench starts at t=0.8065 (slightly later than previous shots today).
Current quench duration is about 1.5 ms.

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

Jul 22 2005 03:49:30:730PM1050722021Robert Granetz
Shot 21 -- Good shot.  Argon gas jet worked.  I_halo=105 kA.
Current quench starts at t=0.806 s.
Current quench duration is about 1.7 ms

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

TS needs one more profile for this parameter set.

Jul 22 2005 04:04:47:073PM1050722022Robert Granetz
Shot  -- Good shot.  Argon gas jet worked.  I_halo=100 kA.
Current quench starts at t=0.806 s.
Current quench duration is about 1.7 ms

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

The TS has a complete set of Te profiles for this parameter set.

Next shot: increase gas jet pulse to 4 ms (from 2 ms); argon

Jul 22 2005 04:16:49:590PM1050722023Robert Granetz
Shot 23 -- Good shot.  Argon gas jet worked.  I_halo=100 kA.
Current quench starts at t=0.806 s.
Current quench duration is about 1.8 ms

Plenum gas: argon
Pulse length (ms): 4.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 63
Plenum change in  pressure (bar):  7

Doubling the pulse length only increased the amount of argon marginally (7 bar, rather
than 6 bar).

So for the next shot, go back to 2 ms and instead raise the RF power as much as possible
(50 MHz, from J-port only).

Jul 22 2005 04:22:18:027PM1050722024Robert GranetzNo TF, no plasma.
Jul 22 2005 04:42:58:040PM1050722025Robert Granetz
Shot 25 -- Good shot.  Argon gas jet worked.  I_halo=104 kA.
Current quench starts at t=0.806 s.
Current quench duration is about 2 ms.

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

Not as much RF power as on previous shots.

Jul 22 2005 05:02:54:197PM1050722026Robert Granetz
Shot 26 -- Good shot.  Argon gas jet worked.  I_halo=116 kA.
Current quench starts at t=0.806 s.
Current quench duration is about 2 ms.

Plenum gas: argon
Pulse length (ms): 2.0
Plenum pre -pulse pressure (bar): 70
Plenum post-pulse pressure (bar): 64
Plenum change in  pressure (bar):  6

End of run.

Physics Operator Comments
Jul 22 2005 07:40:17:267AMSteve Wolfe
Setting up (again) for the gas jet run MP#424
Dennis has gone home, so Bob is on his own today. 
PCS setup: 
  It should be all set, but just to be sure, reload from 1050721009,
which had the changes made after shot#1 yesterday:
  ef4 from -1320 to -1310, br0 to from +.00025 to 0
  He3 puff (pg2) on at .3 off at .4
 Note that the run plan for today calls for Ar and Ne in pg1
Change seg2 title to indicate 5.2T
Load at 22-Jul-2005 07:39:31.00
Open tree /home/wolfe/pcs_scratch      -1
Open tree done


Jul 22 2005 08:17:58:890AMSteve Wolfe
Now to play with the new nonaxi predictor routines.
In seg2, change wire 9 from bolofeed to b21_proj
algorithm from custom to "nonaxi_smw"
hit apply 
 ack the p_info warning
 ack the all-zero warning (for clear line)
 Predictor output lines look like:
***********************************************
% Compiled module: NONAXI_SMW.
About to set def back to \HYBRID::TOP
% Compiled module: HYBRID_DIAGS.
About to set def back to \HYBRID::TOP
% Compiled module: REAL_PART.
Open tree /home/wolfe/pcs_scratch      -1
Open tree done
MAKE PREDICTOR
INPUT_NAMES     STRING    = Array[79]
P_NAMES         STRING    = Array[13]
input_names =  F01 F02 F03 F04 F05 F06 F07 F09 F10 F11 F12 F16 F17 F18A F19 F21 F22 F23 F24 F25 F26 F08 F13 F14 F15 F20 F27 F28 F29
BP01_BC BP02_BC BP03_BC BP04_BC BP05_BC BP06_BC BP07_BC BP08_BC BP09_BC BP10_BC BP11_BC BP12_BC BP13_BC BP14_BC BP15_BC BP16_DE
BP17_JK BP18_JK BP19_JK BP20_BC BP21_BC BP22_BC BP23_JK BP24_BC BP25_BC BP26_GH ROG_FG BTOR TF:BUS_R_CUR OH1:BUS_R_CUR
OH2_U:BUS_R_CUR OH2_L:BUS_R_CUR EF1_U:BUS_R_CUR EF1_L:BUS_R_CUR EF2_U:BUS_R_CUR EF2_L:BUS_R_CUR EF3:U_BUS_R_CUR EF3:L_BUS_R_CUR
EF4:U_BUS_R_CUR EF4:L_BUS_R_CUR EFC:U_BUS_R_CUR EFC:L_BUS_R_CUR CHOPPER:CURRENT NL_04 RF_LOAD_E CORRECT_COIL:BUS_R_CUR Z_METER
DIVBOLO1  DIVBOLO2 MINUS_1_VOLT
p_names=  TF:BUS_R_CUR OH1:BUS_R_CUR OH2_U:BUS_R_CUR OH2_L:BUS_R_CUR EF1_U:BUS_R_CUR EF1_L:BUS_R_CUR EF2_U:BUS_R_CUR EF2_L:BUS_R_CUR
EF3:U_BUS_R_CUR EF3:L_BUS_R_CUR EF4:U_BUS_R_CUR EF4:L_BUS_R_CUR CORRECT_COIL:BUS_R_CUR
% Compiled module: INTERSECT.
max p_to_v*weight on input CORRECT_COIL:BUS_R_CUR
max,min of scaled_pred      2.50000     -2.49186
**************************************************
Ack the zero-warning on the line 3 clear
Row2 factor suggestion is 11149.5, with old factor=2
set factor to 1e4 and ok

Note that the version id field did not update on the edit wire
screen? 

Bring up the controller screen: ack the no controller warning
Note that the limit on acoil_cur is 1.25e7 and the 
"time constant" is set at 0. The physical response to acoil 
is ~3e-7 T/amp, which may correspond to .33e7. Try entering
3e6 as a controller value | OK | Apply the controller screen

Bring up the time constant screen; still says 0.0, which doesn't
seem right. Enter 1.0/sec.
It doesn't like that, it complains "all if the rate must be filled in"
There was no value in the individual tc line, put a 1 there as well.

Bring back the edit controller screen. It has a 1 in the time constant
and the physics value is at 3e6. I could play with which way it
goes if t.c. isn't 1, but it doesn't matter. Ok back to the edit wire
screen.
Note that there is a non-zero final value(-.1325) on the I gain after 
2.0sec; zero this point, just for consistency.  

OK the edit wire screen
ack the p_info warning

Note that the wire label is "b21_proj" and has been since I 
did the apply.
Put away the all_p's screen and bring it back up. Wire 9 is
now labeled "B21_PROJ" (the waveform is still zero.)

Bring the edit_wire screen back up. The version field is now filled in
and reads:"21-Jul-2005,total 2/1 non-axisymmetric field at phi=  21.12 degrees, Acoil=+DTOP-DBOT-JTOP+JBOT"

I think I'm ready to test it open loop. The ACOIL_CUR V waveform
is still 2000A, so I should see that on the aout as well as the calculated
intrinsic field. 

Do the load. No complaints
Load at 22-Jul-2005 08:15:56.00
Open tree /home/wolfe/pcs_scratch      -1
Open tree done



Jul 22 2005 09:14:15:780AM1050722001Steve Wolfe
Shot#1: Bob says try RF on the first shot, so enable B-side upper as well
Gas jet set for 0.8sec
plasma to 1.7sec, lots of hards.
Had an h-mode at .7sec
No gas jet????
Breakdown at 5msec, slightly bouncy startup. 
ratiomatic wasn't on.

Leave pg4 where it is, since the walls will load a bit.  

Jul 22 2005 09:23:34:687AM1050722002Steve Wolfe
Shot#2: Leave pg4 where it is, since the walls will load a bit. 
Try again, Bob is trying to force gv1 manually.
plasma full length. No jet?
No hards this time. Short H-mode. 

The fast valce seems to be firing, but gv1 may not be open? 

Jul 22 2005 09:45:58:017AM1050722003Steve Wolfe
Shot#3: The jet worked in manual mode after shot#2. 
Try again with a shot. 
Dud? Did not get gas jet into the machine. 
Fields don't look bad. I don't understand the dud. 

Raise pg4 by 2msec

Jul 22 2005 10:33:36:640AM1050722004Steve Wolfe
Shot#4: Raise pg4 from 21 to 23msec
set gain on wire9 b21_proj to .01 beginning at 0.2sec
 to see if it's working
plasma. Got the gas jet this time, with gv1 opened
manually in check. Disrupted during H-mode.
Startup looks good, early burnout, moving out but no
runaways. Plasma went down.

The b21 demand out of wire9 is visible at max amplitude
of .08V out of PID. It's asking for positive voltage .01V on 
the coil at 0.2sec; that looks like the sign is correct.

next shot: increase gain on b21_proj to .5

Jul 22 2005 10:54:48:563AM1050722005Steve Wolfe
shot#5: increase gain on b21_proj to .5
 No other pcs changes (I'm tempted to raise pg4 again, but won't)
plasma full length. They ran the gas jet on auto this time,
and Bob thought gv1 closed before the plasma. 

I can see the effect of the b21_proj feedback on the output, 
but it is small. I could stand more gain. 

Jul 22 2005 11:08:05:127AM1050722006Steve Wolfe
shot#6: Turn on ftae for Jason: 
set \HYBRID::TOP.DPCS.LOADABLES.PROCEDURES:TEST_PROS
to ["dpcs_ftae"]

dud. got the gas jet, no plasma.
Fill looks high. We're getting a pressure rise around
-0.9sec,to .007mTorr. Could be a gate valve?

next shot: try again, no changes

Jul 22 2005 11:25:19:343AM1050722007Steve Wolfe
shot#7: no changes, trying to get the jet under auto and the 
plasma on the same shot. 
Got it. disrupts on cue.
some hesitation in current rise at 45kA

next shot: raise b21_proj gain to 1.0

Jul 22 2005 11:54:51:843AM1050722008Steve Wolfe
shot#8:  raise b21_proj gain to 1.0
plasma disrupts on cue, during H-mode this time. 

b21_proj looks like it's working, could stand more 
gain. 



Jul 22 2005 12:06:57:377PM1050722009Steve Wolfe
shot#9: No pcs changes
plasma disrupted at 0.8sec (as programmed)
Bob is moaning in the background;he forgot to top off the plenum.

next shot: b21_proj gain to 2.0

Jul 22 2005 12:26:30:483PM1050722010Steve Wolfe
shot#10: b21_proj gain from 1.0 to 2.0
JT and Glen Wurden are trying to get the IR camera to store.

plasma disrupting at 0.8sec.
Acoil is oscillating on this one, apparently the
gain is now too high. Decrease b21_proj gain to 1.5

Jul 22 2005 12:41:52:687PM1050722011Steve Wolfe
shot#11: b21_proj gain from 2.0 to 1.5
plasma disrupts as planned.

acoil still oscillates, smaller amplitude. 
try some D gain

Jul 22 2005 01:01:07:220PM1050722012Steve Wolfe
shot#12: b21_proj P=1.5,D=1.0
E-port may be fixed. 

Ack failure going to pulse; No pulse. 
No gas, no jet, no supplies, no data.

Jul 22 2005 01:38:26:983PM1050722013Steve Wolfe
shot#13: b21_proj P=1.5,D=1.0
E-port may be fixed. 

plasma disrupted at 0.8sec, tv picture looked brighter to 
me, but all signals look normal. 
This will complete the Ne experiments; next will
be argon. 

Acoil current looks essentially the same as 11.
The PID output has more high frequency (700Hz) and 
maybe a little lower low frequency. Turn off D gain,
drop P gain to 1.0

Jul 22 2005 02:02:10:467PM1050722014Steve Wolfe
shot#14:Changeover to argon gas jet
At JR's request, turn off b-side lower (Ar+Ne)
 b21_proj P=1.0,D=0.0

plasma disrupts on cue. Very late breakdown this time 15ms!
Amazing that it made it. 

acoil doesn't oscillate this time. 
Next: increase P gain on B21_PROJ after 0.4sec
No change to breakdown, see what happens.

Jul 22 2005 02:17:07:463PM1050722015Steve Wolfe
shot#15: increase P gain on B21_PROJ to 2.0 after 0.4sec
No change to breakdown, see what happens.

plasma disrupts at 0.8sec, something glowing on wide2
after disruption. 
breakdown back to normal. 
Yijun requests smaller outer gap by 5mm

Acoil oscillates during the high gain phase. Need to drop it
back I guess; it does make the average error smaller though. 

Jul 22 2005 02:31:22:480PM1050722016Steve Wolfe
shot#16: RCUR program out by .0025 at 0.7sec,
and .001 at 1.5sec (not that it makes it that far).
and at 0.4; this means RCUR=.6595 from .4 to 1.5 sec
B21_proj P=1.5 after 0.4sec
 
disrupts at the usual time, still showing something 
glowing after the shot on the outer divertor near gport.

B21_proj error signal looking pretty good: small oscillations, 
gets to small amplitude <3e-5 by 0.8sec

Jul 22 2005 02:49:09:027PM1050722017Steve Wolfe
shot#17: No PCS changes
plasma disrupts at the usual time. 

Try turning off the pre-program on Acoil_cur

Jul 22 2005 03:14:50:057PM1050722019Steve Wolfe
shot#19: B21_PROJ Integral gain to 1.0 at .2sec, C-integrator 
gate start at 0.25sec

plasma disrupts as usual.startup ok. 
B21_proj error decreasing to 5e-5 by 0.8sec,
Acoil current ramping to 1500A. 
Increase Integral gain to 2.0

Jul 22 2005 03:37:22:510PM1050722020Steve Wolfe
shot#20: B21_PROJ Integral gain to 2.0 at .2sec
plasma, disruption, ...
Significant turn-on transient ring on Acoil, looks good 
thereafter. It probably doesn't help that the integrator gate 
turns on before the coil enable. Move the gate to 0.3sec

Jul 22 2005 03:49:28:243PM1050722021Steve Wolfe
shot#21: Integrator C gate from 0.25 to 0.3sec
plasma, disruption, ...
Seems to have eliminated the transient completely. 
Looks good. Try increasing I gain again

Jul 22 2005 04:01:15:807PM1050722022Steve Wolfe
shot#22: B21_proj Integral gain from 2 to 4
And another one. 
B21_proj error goes to <5e-5 by 0.4sec
and stays there. No turn-on transient
Remove P gain step to 1.5 at 0.4sec

Jul 22 2005 04:17:44:603PM1050722023Steve Wolfe
shot#23: B21_proj P gain to 1.0 flat.
Bob will be increasing the argon pulse length on this one
plasma, disruption,...
B21_proj error very similar to 22, maybe a little less 
ripple. 

Zcur error at the disruption is a little larger, close
to 5e4 vs 4e4 amp-meters.

Next shot: Bob asks for more RF power, no PCS change

Jul 22 2005 04:23:00:760PM1050722024Steve Wolfe
shot#24:  Bob asks Yijun for more RF power, no PCS change
No TF. 

Next shot: try again.

Jul 22 2005 04:44:18:667PM1050722025Steve Wolfe
shot#25:  no PCS change
plasma, disruption, ...
Looks like we got lower stored energy on this one. 
Lots of faulting on J and E

Next shot will be the last shot of the day,week.

Jul 22 2005 04:55:06:730PM1050722026Steve Wolfe
shot#26: Trying for higher RF power; no pcs changes
plasma, disruption, sameo sameo. 
Rf still had multiple faults, Wmhd~95kJ.

End of Run

Jul 22 2005 05:01:21:290PM1050722026Steve Wolfe
Following the end of the run, remove the dpcs_ftae procedure from 
\HYBRID::TOP.DPCS.LOADABLES.PROCEDURES:TEST_PROS node. 

Engineering Operator Comments
ShotTimeTypeStatusComment
109:00:55:343AMPlasmaOk
209:14:18:610AMPlasmaOk
309:38:52:797AMPlasmaOk
410:18:16:563AMPlasmaOk
510:43:44:970AMPlasmaOk
610:57:27:360AMPlasmaOk
711:18:16:233AMPlasmaOk
811:42:16:733AMPlasmaOk
912:01:52:483PMPlasmaOk
1012:21:23:797PMPlasmaOk
1112:36:25:860PMPlasmaOk
1212:59:00:047PMPlasmaBadAck Failure
1301:12:20:907PMPlasmaOk
1401:54:42:017PMPlasmaOk
1502:10:17:760PMPlasmaOk
1602:23:38:620PMPlasmaOk
1702:42:17:947PMPlasmaOk
1802:56:35:743PMPlasmaOk
1903:09:04:307PMPlasmaOk
2003:30:38:243PMPlasmaOk
2103:43:01:980PMPlasmaOk
2203:55:25:917PMPlasmaOk
2304:07:58:790PMPlasmaOk
2404:20:19:510PMPlasmaBadTF did not fire. No faults were observed
2504:37:53:650PMPlasmaOk
2604:51:10:290PMPlasmaOk