Alcator C-Mod Run 970520 Information

Back to runs

Miniproposals
Miniproposal:170a
Date Filed: 5/13/1997
Title:Restart Operation: May 1997
First Author:Robert Granetz
Session Leader: Unknown (shots 1-31)

Operators
Session leader(s): Unknown
Physics operator(s):Steve Wolfe
Engineering operator(s):Vinny Bertolino,Frank Silva,Dave Terry

Engineering Operator Run Comment
MP 170 Restart/Wolfe/Bertolino,Silva,Terry, D

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Physics Operator Summaries
Entered: Jul 7 2004 04:36:57:253PM
Author: To Be Determined
Physop Summary for Tuesday 970520

MP 170A: restart tokamak
PO: wolfe
EO: silva, bertolino

Run plan:
=========
We are still trying to get reproducible full length plasmas.

We will start from the best shot from Monday, 970519032, but with the CLEARIN
adjusted to increase the vertical stability a bit. This is an 800kA,
1.2 sec(EOF), SNB shot.
The main emphasis will be to try to improve the robustness of the
startup. Since this is probably a function of cleanliness, there isn't much I
can do about it except bang away, but I'll try anyway.

We may again want to try ECDC between shots, although it didn't seem very
successful the one time Earl tried it on Monday. I may even ask for glow
between shots, depending on how Jim's tests work out Monday night.

If things seem to be going reasonably well, we will want to try hybrid
control on the TF and ramp-down to 4T from a 5.4T startup. However, we will
start the day with normal PLC TF control.

Results:
-------

The run got off to a late but promising start, went through a long frustrating
middle period and then ended on a high note, getting good plasmas on 4 of the
last 6 shots. We learned several things, at least some of which may be false.

First, the plasma report: Shots 2,23,26,27,29 and 31 were 800kA plasmas, of
which all but 23 had flat density traces and made it into ramp-down. All had
essentially the same programming, based on 970519032 with CLEARIN set to
-.005m (not -.01). The best of these is probably 31, which lasted until 1.5
sec, possibly due to changes in the ramp-down gains; the others, except 23,
all went to about 1.38. Shot 23 was dirty and disruptive, lasting to only
650msec and getting to very high density.

It is worth noting that all of the successful plasmas also had the x-pt
Thomson laser firing into the dump, and that 28 and 30 did not, and were
duds. Moreover, for most (maybe all) of the long string of duds in the middle
of the run, the laser was not firing. There is a strong correlation for making
the case that the laser-produced plasma in the beam dump area, which is
visible on the TV before commutation, may be producing electrons and/or UV to
help pre-ionize. There is also a correlation with base pressure being below
1.1e-6, but that is less robust.

The start of the run was delayed until nearly 11:30 because of a problem at
the alternator. Irby made use of the time by running glow discharge cleaning
for a total of about 2 and a half hours, at relatively low current. GDC seems
to be working reasonably well, but needs lots of conditioning time to get the
current up without arcing. The plan is to run it all night tonight.

Shot 1 had a huge fluorocarbon burst, essentially doubling the pressure at
t=0; it was a dud. Many of the first attempts had large teflon bursts, with #2
being an exception. In an attempt to clarify the the dependence of the CF
bursts on the poloidal voltage, we ran the TF at a PLC setting of 1200V,
instead of 1600V, on shots 8 and 9. The early bursts and the 31,50, and 81
peaks on the RGA essentially vanished. An oscillation of several cycles was
noted on the early part of the TF voltage trace; apparently there is
insufficient damping in the voltage regulator circuit and the way it normally
works is to run up to the available voltage, which is less than the set point
of 1600V. Once the voltage sags a bit, the oscillation vanishes. Shot 10 we
put the tf voltage back to 1600V, and the 31,50,81 peaks came back strongly on
the RGA; DAQA crashed, taking much of the CAMAC data with it, so we couldn't
check the pressure signal, but we got a fizzle. Shot 11 had the TF set to
1500V, which gave essentially the same voltage trace as 1600 and produced a
big teflon burst (1.1e-5). On the following shots we got smaller bursts at
1400V and 1300V settings, and back at 1200V we ran shots 14-19 with small or
non-existent teflon. However, when we set the PLC back to 1600V on shot 20 the
bursts did not reappear, so what looked like a good correlation and a possible
fix did not hold up. It may be worthwhile to try reducing the TF voltage again
if the teflon recurs tomorrow though; it did seem to do something.

After a long series of duds (3-7) following the first plasma, we stopped and
ran ECDC for 20 minutes. This did not seem to help much. We had a few fizzles,
but no more plasmas until 23.

I did a lot of fiddling with fields and fills, but while I made the null look
better (and worse) the plasma was not impressed. One thing that may be worth
revisiting is that I found I could balance the EF2's better, and make some
improvement in the shape of the null, by putting an offset of 50A in the EF2U
predictor. This seemed to be consistent with the fact that the Phys_A_out was
showing zero error for each when they actually had a difference current of
about 50A. However, this tweak was thrown away when I reloaded shot 2 into
shot 23.

Power systems were well-behaved, with only one shot lost (as if) to an early
PF crowbar. This was on shot 12, and it was initiated by a spurious COM fault
on OH2L, just like one shot yesterday.

CAMAC was throwing MEDOFL errors all day, but Tom eventually found a patched
driver from before we did the last system upgrade that seemed to
work. However, putting the RF crate on gave more errors, so it's not fully
fixed yet. The new highway driver unit from JORWAY is overdue.

On the last shot I tried turning on the Argon for John, but it didn't give him
any signal even with 47msec of pulse width at 100V. He will try exercising the
valve tonight.

The 2pi bolo is apparently reading high. Shot 23 claimed to be radiating
5MW. Regular bolos are seeing more normal levels.

Summary
-------
All in all, a difficult day, but encouraging in that we got some good shots at
the end. The x-point laser effect may turn out to be a superstition, but it
can't hurt. The plasmas seemed happy with 17msec gas puffs. I got duds at all
fills, but in some cases higher pressure gave fizzles.

We may be closing in on reproducible full length plasmas. The between shot
ECDC seems to be not helpful. I didn't get to try the hybrid TF control, but
the tf voltage effect (?) on the teflon bursts is interesting.

Scorecard:
----------
Duds 20
fizzles 5
plasmas 6
------- --
shots 31


Session Leader Comments

Physics Operator Comments
May 20 1997 07:18:37:510AMSteve WolfeTuesday 970520

MP 170A: restart tokamak
PO: wolfe
EO: silva, bertolino

coldstart successful at 20-MAY-1997 07:12:32.49
Run plan:
=========
We are still trying to get reproducible full length plasmas.

We will start from the best shot from Monday, 970519032, but with the CLEARIN
adjusted to increase the vertical stability a bit. This is an 800kA,
1.2 sec(EOF), SNB shot.
The main emphasis will be to try to improve the robustness of the
startup. Since this is probably a function of cleanliness, there isn't much I
can do about it except bang away, but I'll try anyway.

We may again want to try ECDC between shots, although it didn't seem very
successful the one time Earl tried it on Monday. I may even ask for glow
between shots, depending on how Jim's tests work out Monday night.

If things seem to be going reasonably well, we will want to try hybrid
control on the TF and ramp-down to 4T from a 5.4T startup. However, we will
start the day with normal PLC TF control.
May 20 1997 07:51:33:440AMSteve WolfeLoad_from 970519032
Set CLEARIN to -.005 in seg 2
Load at 20-May-1997 07:51:17.00
May 20 1997 08:49:44:640AMSteve WolfeWe are having a delay while the alternator crew replaces an actuator that burned out
this morning. Should be ready at about 10AM.

GDC is on- it's hard to tell from the TV picture, but the RGA sees it.
May 20 1997 09:47:54:680AMSteve WolfeThe latest word from the alternator is that we should be on the air at about 10:45
May 20 1997 10:25:43:120AMSteve WolfeThe later latest word is another hour for the alternator, i.e. 11:45AM.
Meanwhile, GDC is proceeding.
May 20 1997 11:28:06:080AM970520001Steve WolfeShot 1: PCS from 970519032 with CLEARIN adjust.
Pre-puff still at 17msec
GDC before operation for about 2.5 hours
GDC off at 10:50AM
11:22 Base pressure=4.5e-7
Got a burst of 28 when Thomson valve was re-opened, came back down ok

Dud.
Huge Teflon burst. Ratiomatic F-side was up to 3.3e-5 at t=0, about half CxFy.

May 20 1997 03:20:56:860PM970520002Steve WolfeShot 2: try that again.
PCS from 970519032 with CLEARIN adjust.
Pre-puff still at 17msec
base=4.0e-7

plasma, lasts 1.36sec
some kind of injection at .77sec,
left gap ~1.3cm

bd looks good:
bd at 5.4msec
140kA @30msec, ECE has signal on 8 of 9 channels.
CII peaks at .01mW/cm2/ster
Ratiomatic at 1.7e-5
May 20 1997 11:51:59:460AM970520003Steve WolfeShot 3: No changes
base=1.2e-6

dud. had a Teflon burst, but smaller.
P0 = 1.75-1.8e-5
May 20 1997 11:58:38:250AM970520004Steve WolfeShot 4: No changes
base= 9.5e-7

dud.
P0=1.5e-5 (maybe there was Teflon on 2?)

For next shot prepuff to 20msec from 17 (and move it t0 -.52 : -.50)
May 20 1997 12:08:23:260PM970520005Steve WolfeShot 5: prepuff to 20msec from 17 (and move it t0 -.52 : -.50)
base = 7.4e-7

another dud?
Imbalance on EF4 now at near zero, was at -60
null looks fair.
Ratiomatic= 2.0e-5
very small teflon burst.

Next shot raise Pre-puff to 25msec
May 20 1997 12:19:32:360PM970520006Steve WolfeShot 6: raise Pre-puff to 25msec
base=6.0e-7

dud.
More teflon: 7.2e-6 @-.52sec, -1.3e-6@-11.sec
bz is coming in pretty late.

No changes
May 20 1997 12:25:01:770PM970520007Steve WolfeShot 7: no changes
base=5.5e-7

dud - another big teflon burst, up to 1.3e-5 at -.5sec.

ECDC for 20minutes
May 20 1997 01:03:13:080PM970520008Steve WolfeShot 8:
ECDC for 20minutes
Drop pre-puff back to 16msec
Drop TF voltage to 1200V (from 1600) start at -1.2sec
base=1.7e-6

fizzle. late bd (20msec!)
No Teflon burst (or very small)
ratiomatic=1.5e-5
Unfortunately, the TF voltage rings for the first 250msec; not good
May 20 1997 01:14:56:160PM970520009Steve WolfeShot 9: repeat, same tf setting; maybe it's not too bad.
base=7.4e-7
dud,
but no Teflon. The TF voltage seems to matter.
Ratiomatic at 1.4e-5

May 20 1997 01:34:08:960PM970520010Steve WolfeShot 10: tf back to normal setting,
pre-puff to 20msec
base=6.2e-7
fizzle. Got a burst.
DAQA crashed so I'm missing the data;
May 20 1997 01:42:20:220PM970520011Steve WolfeShot 11: tf to 1500V, start=-1.0sec
Vinnie thinks the reason the ringing stops is the max voltage sags to the voltage limit.
base=5e-7,
no other changes.

fizzle, late bd (11msec)
teflon burst - 1.1e-5 at -.5sec
May 20 1997 01:51:08:820PM970520012Steve WolfeShot 12: tf to 1400V, start=-1.0sec
base=5.2e-7

dud - pf breaker pulled early
oh2l?
got a teflon burst. delta = 2.5e-6
this one looks bigger on the RGA than 11, so it must have had a big post pulse too.

May 20 1997 02:05:22:580PM970520013Steve WolfeShot 13: tf to 1300V, start=-1.0sec
base=4.3e-7

dud
burst, but it must be after (on the invert). Pressure at -.5=1.25e-6,
compared to 1/13e-6 at -1.7sec
ratiomatic at 1.7e-5 at t=0
May 20 1997 02:16:19:270PM970520014Steve WolfeShot 14: tf to 1200V, start=-1.0sec
base=4.1e-7

dud
still a burst on the RGA, but no early burst, or hardly any (1.4-1.1=.3e-6 delta)
May 20 1997 02:27:55:780PM970520015Steve WolfeShot 15: +15Amps on EF4
leave tf at 1200V.
Tom is trying to fix the CAMAC MEDOFL errors; the last driver he tried made it worse.
Note: pre-pulse is at 19msec, not 20.
base=3.7e-7

dud.
no teflon on RGA, or on ratiomatic (1.1e-6 across the board.)
null looks a little worse

May 20 1997 02:37:43:110PM970520016Steve WolfeShot 16: Put a +50 A offset on ic_ef2u
base=3.7e-7

dud,
tiny burst on RGA, none on ratiomatic
the offset tweak seems to improve the balance on the ef2's, but obviously no help on the
breakdown.

May 20 1997 02:45:32:540PM970520017Steve WolfeShot 17: put back the 15 amps on ef4 (to -1375)
base=3.6e-7

dud.
no teflon.
ratiomatic=1.65e-5

try raising the pressure again
May 20 1997 02:53:56:410PM970520018Steve WolfeShot 18: raise pre-fill to 22msec

at least I got a flash.
fizzle - bd at 7msec
tiny little teflon on RGA, none on ratiomatic,
May 20 1997 03:02:59:300PM970520019Steve WolfeShot 19: ICOH1 to -20650 from -20750, ICef4 to -1360 from -1360

dud, No teflon
null looks like to much Bz early, but the index moved the right way (at least the way I meant it to)
May 20 1997 03:18:43:310PM970520020Steve WolfeShot 20: ICef4 back to -1375
tf voltage back to 1600V for this shot.

dud,
Tiny TEFLON, none on ratiomatic - so much for that theory.
Leave the voltage high then, see if it comes back.
Null doesn't look bad. maybe a tiny Br
May 20 1997 03:25:48:140PM970520021Steve WolfeShot 21: increase pre-puff to 24msec
dud
no teflon.

Try a Br tweak
May 20 1997 03:32:47:290PM970520022Steve WolfeShot 22: Br from .0065 to .0060
If this doesn't work, I'll reload shot 2.

No MEDOFL errors after shot 21 (twf, et al. have been playing musical drivers, and finally hit on
one that seems to work).

dud. I don't know where it went, but I'll revert to shot 2.
May 20 1997 03:49:19:100PM970520023Steve WolfeShot 23: reload of shot 2, short early pre-puff and all.

plasma -

bd at 5.2 msec, some hesitation but over 100kA at 30msec, 500kA at 100msec
dies at .65sec, big density burst at 0.4 sec, density rises to 2.65e20 nl4. This may be a
density limit disruption,

Joe says there was a locked mode.

Note that the Boris went down to fix something in the cell before the shot, and the edge
thomson laser was firing, as it was on #2. Another superstition is born!
May 20 1997 03:54:48:280PM970520024Steve WolfeShot 24: No changes.
base= 9.8e-7

dud. Here we go again.
tiny teflon, nothing to speak of on ratiomatic.
May 20 1997 04:06:08:480PM970520025Steve WolfeShot 25: increase pre-puff to 20msec, move it to -.500
base=6.8e-7

fizzle. bd at 5.5msec
some positive radial field
two bounces, it was trying
May 20 1997 04:16:47:470PM970520026Steve WolfeShot 26: decrease pre-puff back to 17msec
base=6e-7

plasma- lasts to 1.38sec in rampdown.
nice flat density, small h-alpha burst around 220msec, near when it diverts, also
ef4 crossover.

Null looks horrible, but breakdown is nice and early and current rise is pretty good.

May 20 1997 04:31:47:850PM970520027Steve WolfeShot 27: No changes
base=1.1e-6.

plasma: 2 in a row!
disrupts in rampdown at almost exactly the same time.
bd at 5.0msec, reasonable current rise
May 20 1997 04:36:03:590PM970520028Steve WolfeShot 28: No changes
base=1.1e-6

dud. Note that the x-pt Thomson laser did not fire this time.
May 20 1997 04:47:14:790PM970520029Steve WolfeShot 29: No changes
base=7.8e-7

plasma. Laser was firing again.
the pulse length is identical! Something happens then.

May 20 1997 04:54:18:230PM970520030Steve WolfeShot 30: Turn on Ar pulse for JR, 47msec at 0.3sec
No other changes. There are lots of gain drops at that time.
base=1.4e-6

dud. No laser; they're still batting 1000


May 20 1997 05:30:15:410PM970520031Steve WolfeShot 31: Kill late gain switches on RXU, ZXU, RXL, ZXL
base=1.0e-6

plasma:
made it to 1.5sec; takinmg aout the gains seems to help.
This one had a shark fin at the end.

Engineering Operator Comments
ShotTimeTypeStatusComment
111:23:10:640AMPlasmaOk
211:35:37:620AMPlasmaOk
311:46:20:120AMPlasmaOk
411:53:25:460AMPlasmaOk
512:02:06:330PMPlasmaOk
612:12:19:510PMPlasmaOk
712:20:37:500PMPlasmaOk
812:54:22:290PMPlasmaOk
901:09:44:460PMPlasmaOk
1001:17:42:510PMPlasmaOk
1101:36:25:710PMPlasmaOk
1201:44:05:910PMPlasmaBadoh2 fault
1301:58:24:930PMPlasmaOk
1402:07:36:980PMPlasmaOk
1502:22:30:160PMPlasmaOk
1602:31:54:420PMPlasmaOk
1702:40:43:240PMPlasmaOk
1802:48:43:500PMPlasmaOk
1902:57:55:940PMPlasmaOk
2003:07:13:040PMPlasmaOk
2103:23:20:410PMPlasmaOk
2203:30:04:060PMPlasmaOk
2303:39:09:730PMPlasmaOk
2403:50:26:850PMPlasmaOk
2504:00:47:510PMPlasmaOk
2604:09:34:450PMPlasmaOk
2704:19:21:480PMPlasmaOk
2804:32:37:690PMPlasmaOk
2904:42:51:830PMPlasmaOk
3004:51:28:800PMPlasmaOk
3104:58:35:780PMPlasmaOk