Alcator C-Mod Run 941222 Information

Back to runs

Miniproposals
Miniproposal:
Date Filed:
Title:
First Author:
Session Leader:Steve Golovato (shots 1-33)

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

Engineering Operator Run Comment
icrf heating

Session Leader Plans

Physics Operators Plans

Session Leader Summaries
Entered: Jul 7 2004 03:49:11:740PM
Author: To Be Determined
run summary - 941222
---------------------

SL: golovato
PO: horne
EO: bertolino/daigle

Goals- Tune E-port antenna at high density (where we can get match
easier) and increase power, condition antenna. Bring up D-port to
high power simultaneously, get as much total power as possible.
Compare performance of the two antennas, look for any interactions
between them.

We accomplished some of the above goals but were hampered by two problems with
the rf systems. First, the E-port tuners are slightly less than one half
wavelength, leaving a gap in the tuning ranges. This is a particular problem
for the stub, whose admittance goes like the cotangent of its length. Most
of the variation of the impedance occurs over a short length near half wavelength
intervals. As luck would have it, that's where our gap is. We will fix this
by adding a short length of coax to the stub (in fact it has already been done)
allowing us to tune to good matches over a wider range of plasmas. For this
run we chose high density plasmas and/or small outer gaps to get the high loading
we needed to get adequate, but not great, matches. The second problem was that
transmitter 2, which powers the E-port antenna, was crowbarring every shot.
The power is supposed to interrupt and retry rather than trip the power supply.
The retries provide conditioning of the antenna, the lack of them slows down the process.
Crowbarring is also bad for the tube and power supply.

Nevertheless, we made some progress during the run. We strarted with a slot divertor
plasma, shot 941220020, at high density (nl_04=1.5e20). Horne adjusted this plasma
to have the standard divertor strike points to keep it off the inner nose.
We tried this for the first 11 shots. We had trouble getting power up on E-port
and thought high gas puffing required to hold the density might be affecting the
power handling. So we switched to another configuration, shot 941216024, which
was inner wall limited, had a 0.5 cm outer gap, but more modest nl=1.e20. It took
several shots to make the switch. We increased the outer gap to 1cm and it worked from
then on. We ran the antennas at different times, E from .35 to .8 sec and D
from .8 to 1 sec unitl shot 30, where we started to try overlapping the pulses.
D port ran reliably at 1.7-1.8 MW and even ran at 2 MW for 50ms on shot 28. At the
highest power, the transmitter was tripping from screen overcurrent, the usual
limation at high power. Te and Ti were doubled simultaneously with a 30 kJ increase
in stored energy, on shot 22 for example. E port was more erratic but did get up to
high power on the last few shots. The best shot was the last, where D and E overlapped for
60 ms and reached 2.7MW of total input rf power with over 1 MW from E-port.

We have some problems with the plasma apparently striking the E-port antenna protection
tiles (on the F-port side) near the midplane, causing them to glow noticably every shot.
We tried hard to maintain the outer gap, particularly during crossover, to eliminate this
but were only successful in reducing the effect. This may play some role in limiting
the power to E-port thus far.

Lithium pellets were also applied during the run, medium pellets on shots 23 and 32 and small pellets
on shots 24-30. They were all fired at 450ms, before the rf. There were apparently some interesting
and new effects seen. Snakes were generated, even with the small pellets, indicating penetration
to q=1.

The session leader shot-to-shot summary can be found in the logbook under the
session_leader topic.

The shot summary was: 31 plasmas (3 disrupted very early)
2 fizzles

Thanks to the engineering group for giving us the last shot after 5:00pm to finish out the year


in style with a record total rf power. Happy Holidays!!!

Physics Operator Summaries
Entered: Jul 7 2004 04:36:52:583PM
Author: To Be Determined




run 941222 Golovato, Horne

Rf run. Loaded shot set up for slot divertors, includes reference
switch on segment 2.

Plan to work on rampdown as piggyback. Try to develop a strategy based
on zero drawn IP, very low gain on IP. If gain is constant, should
get exponential rampdown. Consider gain steps.

From wirehead point of view, run has two parts.

Shots 1-11 were diverted. 12-33 limited.

EF1 L had relatively consistent problems in inversion.
Segment 1 programming was adjusted to match up the ef1 currents across the joint.
Some shots had imbalanced ef1 currents at breakdown
but there was only one fizzle today (shot 5).

The diverted shots had rampdown problems, possibly due to the
rather high programmed density; more likely due to the loss
of control of the EF2 coils when their current goes to zero.
I tried 3-segment operation on shot 4 but had not properly programmed it.

The E-port antenna was apparently unhappy with these plasmas so we switched to
a diverted plasma on shot 12. This was successful after a couple
of tweaks.

Starting shot 19 I tried the 3'rd segment again, programmed
properly this time. Over the next several shots evolved to a ramp-down sequence
with zero drawn IP -- object was to make a sequence which could be switched in
at any time, independent of plasma current. This seemed to
work -- shots 23-29 and 33 had this programming.

For 31 and 32, switch back to 2-segment operation, using 1 and 3.
This worked. I think we've verified PCS operation in nearly all of the
more exotic configurations; since we are running 3 segments, Josh says he will
create segment 4 for our startup next year.

Steve G and Yuichi will post an RF report; the bottom line is
2.8 MW on shot 33. (for rather short time -- dt or so).

Scorecard -- 32 plasmas, one fizzle.

Shot log:


1 load 941220020. Copy segment 2 to segment 3.
Set segment 3 start time to 1.25 sec, not on yet though.
Reduce strike_psi proportional gain to .8, integral gain off (no reset).
Nasty shot.

2 Reduce programmed density. Fizzle detector to 1.6
Good shot.

3 increase density. x-point isn't perfect -- inner strike point is
on the nose. Fix next shot.
Died early -- too much gas? No -- EF1L breaker pulled out.


4 strike_psi gain reduced to .5, programmed to -.001 (was -.003)
Increase density to 1.5e20
Turn on segnment 3!!.
Plasma exploded at the segment switch.
EF1 L breaker.

5 Turn off segment switch. I had not corrected segment 3 to match
segment 2 -- that may have been responsible for the disruption.
Other possibility is bug w 3 segments.
Raise density earlier. Raise zxl and rxl gain, remove gain notch from
zxl
Plasma fizzled. EF1 L breaker again. It's pulling out in inversion;
so backing off it's gain shouldn't affect the startup much (I hope).

Scatterplot shows br is terrible. I've no idea why it's worked at all.

6 repeat
Plasma. TV shows disruption.
We loose plasma when ef2 coils go to zero current.
EF1L power supply gain has changed.

7 Repeat, plasma disrupted just after segment switch due to EF1 L
pulling out.

8 change ef1 segment 1 programming to match up better to segment 2.
want's to be -3000 , was -2000, make it -2500.

Plasma ran. Disruption on rampdown at .25 MA.

9. Increase integral gain on ef1l only in segment 1 from 3 to 4.

Plasma. We loose it when ef2 currents go to zero. Try moving x-points
to larger radius. Should try lowering zxl a half cm or so -- elongation
isn't too high so we should be able to stand it.

10 Plasma. Still trouble with ef2
Drop zxl to -.001 for next shot.


11 Init failure, try again...
Plasma.

12 RF wants a limited plasma now. Try getting it by
adjusting programmed waveforms and gains, so we keep segment 2 reference switch.
Cut in rcur, zcur, clearin, xpoint locations from pcs s scope shot 941216024
strike point gain to zero.
nl_04 programmed to 1.e20.
Fizzle. Possibly ef1 imbalance again, or dirty -- CII light high.
Try ef1l integral gain to 4.5

13 short, dirty, disruptive, nasty, foul,...
Just load the shot.

14 Plasma, came to a bad end. 800 ka disruption.
Rice claims credit. We deleted his second and third puff.
efit looks good.
Startup is different from shot 941216024 ; ef1L in particular.


15 Better but not good. Rice now claims that
RF tossed moly into plasma.

Adjust ef1 currents in segment 1 to match across the segment jump.
reduce rcur at end of shot from .666 to .663 to increase outer gap.
rcur=number_of_beast might be cause of disruption. Especially since it's
so close to Christmas.

16 Much nicer shot, rampdown to < 70 kA.
Program lower xpoint a bit lower during rampdown.

17 Like last shot.

Small tweak to clearin and rcur late.
18 Nominal.
Add segment 3 at 1.4 sec (copy of segment 2)
Change zxl programming early to prevent "accidental" diversion.

19 Nice shot. No sign of glitch at segment switch.
segment 3 start at 1.3
aout error changes by 40000 at rampdown.
This x pid p gain of 3.6 = nice rampdown.
Draw IP in segment 3 to zero; set PID gain to 40/200 * 3.6 = 150/200 = 1.
Try 1.6 or it will take forever.

20 Nice plasma, rampdown in segment 3 worked !!
Try moving segment 3 to 1.2 sec. Error will go from 300kA to 500kA.

Nice rampdown again. We seem to be able to tolerate this, but
the PID output changes by 1 V, compared to .25 volts during our usual
rampdown.
Set start time at 1.1; error goes to 700 ka. Reduc gain to .5 there, and
step it up later.
Between .250 and .4-.45 or so, as plasma recovers from crossover,
it is bumping the antenna. You can see the antenna glow. Program rcur in by 1/2 cm
during this period.

21 Nice again. The gain changes in IP in segment 3 result in an almost linear rampdown.
Try switch at 1.0.
"Medium" Lithium pellet at .45 sec

22 No problems. Pellet causes small glitch in plasma current.

Discovered that current in segment 3 wasn't properly drawn to zero.
Went to zero at 1.2 sec.
Draw to zero for next shot.
want 40 / 800 * 3.6 =160/800 = .2 Set .3

23 Worked fine.

24 No change, no problem.
Plasma still gently kissing e-port limiter.

25 Tweak rcur another 2 mm in as in shot 20.
Plasma.

26 Hard x-rays creeping up early. Raise fill to 1.6, program density up a
little faster.
Lithium Pellets still going in at .45
Plasma.
still touching antenna at .2 or so. Adjust clearin gain early.

27 Plasma
No pellet this time.

28 Pellet back on.
Fiddle with rcur.

Good shot, outer gap seems better controlled.

29 no change.
Good shot. Bright spot still there. Boring. For
excitement, try running segments 1 and 3.
Copy segment 2 to 3.
New rampdown scenario can be recovered from segment 3, this shot. (29)

30 Good shot. Segments 1 and 3 did right thing, apparently.

31 No change. Good shot.

32 Prepare to put my rampdown back, but don't load it yet.
Good shot.

33 Load the rampdown.
Good shot; rampdown worked.





Session Leader Comments
Dec 22 1994 08:53:50:640AMSteve Golovatorun 941222 - last run of 1994

Goals- tun eE-port antenna at high density (where we can get match
easier, and increase power, condition antenna. Bring up D-port to
high power simultaneously, get as much total power as possible.
Compare performance of the two antennas, look for any interactions
between them.

Dec 22 1994 08:56:45:190AMSteve Golovatorun 941222 - last run of 1994

Goals- Tune E-port antenna at high density (where we can get match
easier) and increase power, condition antenna. Bring up D-port to
high power simultaneously, get as much total power as possible.
Compare performance of the two antennas, look for any interactions
between them.

Dec 22 1994 08:59:49:480AM941222001Steve Golovato941222001 - loading 941219020 which matched antenna well, had slot
divetor and good rampdown, nl programmed to 1.5e20. tweaked a little
to move x-point away from nose.

early disruption, too much gas too soon for first shot.

Dec 22 1994 09:01:46:590AM941222002Steve Golovato941222002 - good shot, density lowered both early and late.
no rf data, mistuned anyway since density lower than we tuned to,

Dec 22 1994 09:06:15:060AM941222003Steve Golovato941222003 - good shot, but disrupted at 850ms. programmed at 1.2e20.
lost ef1-l at 800ms. both antennas reasonably well matched.

Dec 22 1994 09:10:23:860AM941222003Steve Golovato941222003 - good shot, but disrupted at 850ms. programmed at 1.2e20.
lost ef1-l at 800ms. both antennas reasonably well matched.

Dec 22 1994 09:31:03:900AM941222004Steve Golovato941222004 - raised density programming to 1.5e20. good shot,density
reaches programmed value. 1.2MW on D-port, tuning not bad but needs
tweaking.

Dec 22 1994 09:30:40:600AM941222005Steve Golovato941222004 - raised density programming to 1.5e20. good shot,density
reaches programmed value. 1.2MW on D-port, tuning not bad but needs
tweaking.

Dec 22 1994 09:35:15:400AM941222005Steve Golovato941222005 - brought denity earlier, raised powerand tuned D-port,
separated D and E port pulses in time. D at 0.825-1.sec, E at
.350-.825 sec.

fizzle and ef1 power supply tripping out, to be adjusted.

Dec 22 1994 10:01:37:010AM941222006Steve Golovato941222006 - good shot, density came up slowly. no efit data for
som ereason. D-port interrupting, tuning bad.

Dec 22 1994 10:14:40:580AM941222007Steve Golovato941222007 - early disruption, like shot 1. ef1 behaving differently
from shot 941220020.

Dec 22 1994 10:27:06:940AM941222008Steve Golovato941222008 - good shot. 1.2MW. E-port crowbarring. density looks good.
heating with D-port looks good.

Dec 22 1994 11:01:42:390AM941222010Steve Golovato941222010 - density gets up to 15e20 quicker, not like last shot
although nothing was changed. 1.3MW on D.

Dec 22 1994 11:10:24:700AM941222010Steve Golovato941222010 - density gets up to 15e20 quicker, not like last shot
although nothing was changed. 1.3MW on D.

Dec 22 1994 11:34:41:650AM941222011Steve Golovato941222011 - good shot, 1.45MW on D. E tripping at lower and lower power
maybe high gas load at high density makes it difficult to run it.
E port ran better on 941216 with innner wall limited plasmas centerred
in z with a 1 cm outer gap.

Dec 22 1994 11:36:31:860AM941222012Steve Golovato941222012 - fizzle.

Dec 22 1994 11:51:33:670AM941222013Steve Golovato941222013 - early disruption.

Dec 22 1994 12:08:25:750PM941222014Steve Golovato941222014 - reloaded shot 941216024 entirely, not just some traces.
disrupted at 830ms. E port ran well, match not bad. Disrupted before
D port pulse.

Dec 22 1994 12:28:03:480PM941222015Steve Golovato941222015- same shot, same disruption. E-port to .5 MW, no D-port rf.
try increasing outer gap from .5 to 1 cm..

Dec 22 1994 12:47:20:960PM941222016Steve Golovato941222016 - good shot, .6 kw on E, 2 MW on D.

Dec 22 1994 12:55:57:830PM941222017Steve Golovato941222017 - 1.8MW on D, .5MW on E, crowbars each shot. outer gap grows
starting at 850ms. extend clearin and rcur programming out to 950ms
from 850ms for next shot.

Dec 22 1994 01:18:13:110PM941222018Steve Golovato941222018 - good shot, D port interrupting with plasma held at 1cm,
needs to be tuned. E port crowbars at 600kw.

Dec 22 1994 01:31:57:660PM941222019Steve Golovato941222019 - D-port at 1.3MW. E-port crowbars early when plasma is on
outer limiter.

Dec 22 1994 01:56:12:320PM941222020Steve Golovato941222020 - good shot, E still crowbars.

Dec 22 1994 01:58:12:030PM941222021Steve Golovato941222021 - good shot. plasma moves out to outer limiter between
300-450ms during crossover and final current rise, cuases heating
of antenna tiles on E-port. this may limit powere. we ill try to
eliminate this excursion.

Dec 22 1994 02:24:04:820PM941222022Steve Golovato941222022 - good shot. add Li pellet at 450ms on next shot.

Dec 22 1994 02:25:47:870PM941222022Steve Golovato941222022 - heating very good. del(Ti)=1200eV to 2300eV,
del(Te)=1800eV to 3800eV. del(Wp)=30kJ.

Dec 22 1994 02:45:05:730PM941222023Steve Golovato941222023 - 1.7MW on D, short E pulse. Li pellet at 450ms.

Dec 22 1994 02:56:35:540PM941222024Steve Golovato941222024 - 1.8MW on D, .65 on E. Plasma still hitting E-port limiter,
moving plasma in another 5mm 200-450ms.

Dec 22 1994 03:25:18:260PM941222026Steve Golovato941222026 - still problems with E-port crowbar. turning off Li pellet
for next shot.

Dec 22 1994 03:40:52:970PM941222027Steve Golovato941222027 - good shot. no Li pellet.

Dec 22 1994 03:55:33:800PM941222028Steve Golovato941222028 - 2MW on D for 75ms, then transmitter tripped.

Dec 22 1994 03:56:17:890PM941222028Steve Golovato941222028 - Li pellet back on.

Dec 22 1994 04:18:23:730PM941222029Steve Golovato941222029 - try overlapping D and E on next shot.

Dec 22 1994 04:23:43:270PM941222030Steve Golovato941222030 - D-port tripped when E went on, screen current trip.

Dec 22 1994 04:42:43:710PM941222031Steve Golovato941222031 - both ran together for about 8ms at 2.5MW total power.
then E-port tripped followed by D port.

Dec 22 1994 04:50:35:470PM941222032Steve Golovato941222032- medium Li pellet this shot. both tripped as usual.

Dec 22 1994 05:20:13:310PM941222033Steve Golovato941222033- no Li pellet. 50ms overlap getsa to 2.8MW total.

Physics Operator Comments

Engineering Operator Comments
ShotTimeTypeStatusComment
108:33:04:280AMPlasmaOkno faults
208:45:00:390AMPlasmaOkno faults
308:58:14:460AMPlasmaBadef1l comm fault
409:12:11:930AMPlasmaOkdisruption induced comm faults ef1l, oh1
509:27:16:850AMPlasmaOkfizzle
609:41:18:650AMPlasmaOkno faults
709:57:22:890AMPlasmaBadef1l failed at segment change in hybrid program
810:18:23:060AMPlasmaOkno faults
910:36:24:990AMPlasmaOkno faults
1010:50:53:550AMPlasmaOkno faults
1111:05:18:580AMPlasmaOkno faults
1211:34:09:910AMPlasmaOkno faults
1311:48:10:130AMPlasmaOkno faults
1412:00:40:740PMPlasmaOkno faults
1512:12:58:300PMPlasmaOkno faults
1612:29:28:490PMPlasmaOkno faults
1712:44:17:220PMPlasmaOkNo faults.
1812:58:28:590PMPlasmaOkno faults plasma no disruption
1901:13:20:120PMPlasmaOkNo Faults
2001:28:08:510PMPlasmaOkNo Faults
2101:41:32:020PMPlasmaOkNo Faults.
2202:02:24:470PMPlasmaOkNo faults.
2302:22:19:300PMPlasmaOkNo Faults.
2402:39:46:260PMPlasmaOkNo faults.
2502:57:02:770PMPlasmaOkNo Faults
2603:13:32:870PMPlasmaOkNo Faults.
2703:30:01:290PMPlasmaOkNo faults.
2803:44:37:070PMPlasmaOkNo Faults.
2903:58:21:470PMPlasmaOkNo Faults
3004:13:57:100PMPlasmaOkNo Faults.
3104:29:45:150PMPlasmaOkNo Faults.
3204:44:08:750PMPlasmaOkNo Faults.
3305:02:07:590PMPlasmaOkNO faults.