Alcator C-Mod Run 940519 Information

Back to runs

Miniproposals
Miniproposal:
Date Filed:
Title:
First Author:
Session Leader:Robert Granetz (shots 1-8)
Session Leader:Steve Wolfe (shots 9-23)

Operators
Session leader(s):Robert Granetz,Steve Wolfe
Physics operator(s):Earl Marmar
Engineering operator(s):Joe Daigle

Engineering Operator Run Comment
closed loop, icrf tuning

Session Leader Plans

Physics Operators Plans

Session Leader Summaries
Entered: Jul 7 2004 03:49:10:007PM
Author: To Be Determined
Part II of Run Summary for 940519
(Bob Granetz will write up part I, containing most of the good news)

SL: Wolfe
PO: Marmar
EO: Daigle

Miniproposal 41 - Restart C-MOD

Granetz turned over the run to me at around 13:30, after shot 8. Things had
been going well, about six good shots in a row; somehow he managed to sabotage
things before I even sat down. Shot 9, which had already been set up and
differed from shot 8 only by changes after 0.6 sec, fizzled.

Following this inauspicious beginning, we floundered for a few shots, some of
which were actually duds, the first of theday. Shot 14 ran until 530msec, then
was lost down. This shot, like earlier shots, was programmed to be at
ZCUR=0.0, but in fact bothe xrays and EFIT claimed it was at -2cm. The Hybrid
Aout claimed it was centered. Apparently, there was a problem with the ZCUR
predictor.

Ignoring the ZCUR problem for the moment, since we couldn't find any bad
magnetics by inspection, we tried getting rid of some of the IC controllers
which were fighting the shape controllers. Specifically, we turned off the
gain on IC_EF1L and IC_EF2L after .2sec, when ZXL and RXL turned on. RXL was
running inside and did not appear well-controlled.

The next several shots were failures of one flavor or another. For shot 21 I
decided to bite the bullet and work on the null. I made the EF1 currents 300
amps more positive to try to reverse the Bz gradient and provide a stable
index during the current rise. Amazingly, this worked, and shot 21 ran through
700msec. The null was only there briefly, because we did not compensate for
the Bz increase, but it broke down and the current rise was good. On shot 22,
we attempted to compensate the Bz, but may have overdone it; the shot fizzled.
Shot 23 was a reload of shot 21, and ran well until 530msec. Both shots 21 and
23 had the same ZCUR problem as shot 14, namely the hybrid thought they were
centered when the xrays and EFIT thought they were low by at least 2cm.

After the run, the problem with ZCUR predictor on the later shots was
investigated. It turns out that BP17_DE went bad sometime between shots 8 and
14, possibly right at the end of shot 8. This failure is insidious, in that on
shot 14 at least the signal comes back to zero at the end of the shot, but is
wrong (relative to BP17_JK, aka the processed flux) by up to .15 Tesla during
flattop. This was found by direct comparison of all of the BP inputs to the
hybrid, with BTOR_PICKUP*\BTOR subtracted, to the corresponding JK loops.
Since BP17_DE was installed last year as a replacement for BP17_BC, it will be
necessary to replace it with the JK loop (or the GH, provided we reconnect it
to CAMAC).

Essentially no progress on the miniproposal was made during the second half of
the run, although turning off the superfluous IC_EF1L and IC_EF2L controls is
logically a good thing, and seemed to do no harm. The null with the EF1's more
positive is a step in the direction of more stable breakdowns and current
rises, I believe, but it needs some additional tweaking up. The Bz is
presently too positive, and the index may actually be a bit too positive
during the early rise phase.

The plasma is still fat and riding the inner limiter until after 0.5 sec.
CLEARIN does not seem to be working correctly when turned on early. RXL is too
far inside and did not seem well controlled. ZXL is close to where it is being
commanded, which is somewhat too high for good matching to the divertor. The
elongations get quite high with the plasma centered if we allow the xpoint
height to be correct. It may be that xpoint control will improve when BP17 is
replaced with a working signal, but the RXL control didn't seem to be working
earlier in the run either.

Entered: Jul 7 2004 03:49:10:180PM
Author: To Be Determined
Part I of Run Summary for 940519

SL: Granetz/Wolfe
PO: Marmar
EO: Daigle

Miniproposal 41 - Restart C-MOD

The first half of this run was more successful than prior runs this
week. We had 8 shots, 6 of which had plasma currents of more than 0.8 MA.
Most of these lasted to at least 0.7 s, with one actually getting part way
down the current rampdown before disrupting.

There sometimes is a position control problem early in the plasma
formation (25-50 ms). It kills the proto-plasma on shots which have a
relatively slower current rise. So far the only way around this is to
have a clean enough machine so that dIp/dt is not slowed down. Obviously not
very satisfactory.

There was a problem with blown SCR's and a failed gate-driver board in
EF2L, which was discovered after the first couple of shots, but which
probably had been bad since April.

The first two plasmas that we got both disrupted due to
excessive elongation. This probably was caused by CLEARIN feedback on
the OH2's, and is not understood. By delaying CLEARIN feedback until
after 0.6 s, we managed to get longer shots. The best
shots (6,7,8) were diverted after 0.5 s and were centered on the midplane,
which is required for matching to the antennas. The programmed position of the
x-point was pushed outward in major radius on successive shots, and this
seems to be working, for the most part, although RXL is still not as
far out as we'd like.

During the second half of the run, it was discovered that 2 Bp coils
became flaky. These both are connected at D-bottom port, which is the
other flange that was re-installed during my absence last March. It is
very likely that those connectors were also not properly screwed on, just
like at B-bottom.

Physics Operator Summaries

Session Leader Comments
May 19 1994 06:26:39:980PMSteve WolfeWe have another bad bp-coil hooked to the hybrid, and this time it's a DE loop.
BP17_DE went bad sometime during the run today. Furthermore, you can't tell from
looking at the reset; it is bad during the shot and comes back to zero. I found
it by direct comparison with BP17_JK, after subtracting the BTOR_PICKUPs from
each.

On shot 14, which ended up 2cm low despite the fact that ZCUR said it was
centered, the BP17_DE (corrected by BTOR_PICKUP) goes to -.6T during flattop,
while BP17_JK (similarly corrected, identical to processed BP_16) goes to -.75,
a discrepancy of 0.15 Tesla. On this shot both signals come back to zero, within
10mT.

On shot 8, which actually was centered, the two signals overlay perfectly during
the shot, both getting to about -.75T, but the DE loop has about a .1Tesla
offset at the end of the shot.

We (Bob) need to fix this before tomorrow's run. I suppose we'll have to use the
JK loop, unless we want to reconnect the GH loop to CAMAC. Note that BP17_xx is
on the outer shelf, right next to the crandulated flux loop. Maybe there's a
connection, maybe not, but we clearly need this coil to help make up for the
lack of F17.

We are still losing magnetics at an alarming rate. This one is presumably on a
different port, and may not be related to connector problems at all, but it
could be. Perhaps Bob's thumbs should be re-lengthened.

May 19 1994 09:28:33:180AM940519001Robert GranetzFizzle, but current started to rise. Ipmax=0.16 MA, duration=.056 s.
Lost vertical control, downward VDE. Rate of current rise was not as fast
as yesterday's shot. I think that we do not have adequate feedback control
of ZCUR early on. This is not a problem if we have adequate rate of current
rise.

Next shot, repeat. Also lower the pulse gas a little after 50 ms. Also
change pre-fill from 1.2 to 1.1e-5.

May 19 1994 10:05:09:480AM940519002Robert GranetzPlasma. Higher current. Ip=0.83 MA, duration=0.43 s. Downward VDE. No
halo signals. Elongation got to 1.6 at 0.4 s, so we don't have much chance
of handling vertical position in this case. We have to reduce the maximum
elongation. The x-point is 5 cm too far inside, and a couple of cm low.
Voltage demand to EF2L is behaving properly, but the supply is not responding
well.

Next shot, raise programmed ZCUR to 0.0 cm (rather than -2 cm), RXL feedback
gain is turned on earlier, and ZXL gain is enabled.

May 19 1994 10:37:36:340AM940519003Robert GranetzFizzle with some current. Same as shot 001. Lose plasma downward. dIp/dt
was slower than the previous good shot.

Problem with EF2L fuses. Will take about an hour to fix. Meanwhile, during
a cell access, I found that we had patched the halo signals into Joe Snipes's
Aurora camac, rather than the TRAQ's. This has now been remedied for the
next shot.

May 19 1994 12:09:07:560PM940519003Robert GranetzFor next shot, RXL and ZXL and CLEARIN gains now turn on at 0.2 s (which is
how it was on the shot from last October.

May 19 1994 12:23:32:310PM940519004Robert GranetzPlasma, but not as long. Ipmax=0.73 MA, duration=0.32 s. Downward VDE.
Things start to decrapitate starting at 0.20 s, which is when we're turning
on CLEARIN, RXL, and ZXL. The problem with EF2L is still there also.

Consensus is that CLEARIN is screwing up the OH2's.

May 19 1994 12:32:21:070PM940519004Robert GranetzNext shot, delay CLEARIN until 0.6 s. Engineers replaced bad gate driver
board in EF2L.

May 19 1994 12:54:05:060PM940519005Robert GranetzPlasma. Ip=0.85 MA, duration=0.71. Plasma position control was good until
late, when kappa went above 1.6. Plasma was vertically centered, which is
what the RF group needs. Clearance to inner and outer divertors was good after
0.5 s.

Next shot, add offset to RXL to move it further out, move ZXL farther up
by 5 mm, correct EF1U current programming to undo change we made yesterday,
and increase pulse gas.

May 19 1994 01:11:07:870PM940519006Robert GranetzPlasma. Ip=0.85 MA, duration=0.74 s. RXL moved out, but not enough.
Looks like plasma disrupted first, and then the remnants went down.

Next shot, increase offset on RXL again, resulting in about another cm farther
out.

May 19 1994 01:20:23:580PM940519007Robert GranetzPlasma, getting better too. Ip=0.85 MA, duration=0.79 s. Current ramped
down to 0.6 MA before it disrupted. Density is dropping alot late in the
shot.

Next shot, more gas puffing.

May 19 1994 01:29:25:050PM940519008Robert GranetzPlasma. Ip=0.84 MA, duration=0.75 s. I'm going to hand the reins over to
Steve Wolfe.

May 19 1994 01:49:35:760PM940519009Steve Wolfe940519009
----------
Wolfe replaces Granetz as SL

Raised controller on Fast ZCUR by a factor of 2, starting at 0.6sec

Result:
fizzle. Thanks, Bob!

May 19 1994 01:58:58:180PM940519010Steve Wolfe940519010
----------
Lower P to 1.0e-5
No other changes

Result:
dud!

May 19 1994 02:09:37:720PM940519011Steve Wolfe940519011
----------
Change Br_0 from -3.0mT to -2.5mT


Result:
Another dud. Pressure was <1.0e-5 when the shot went

May 19 1994 02:28:40:430PM940519012Steve Wolfe940519012
----------
Reload shot 8, just in case something got changed inadvertantly.

Result:
-------
Early breakdown (5.5msec)
Lost it at 65msec, screaming down.
Looks like earlier today.

May 19 1994 02:46:20:190PM940519013Steve Wolfe940519013
----------
Lower the puff after 400msec, and try again.
This worked before!!

Result:
-------
dud.
What the hell is going on.

May 19 1994 03:02:45:990PM940519014Steve Wolfe940519014
----------
Raise fill to 1.3e-5 and try it again.

Result:
-------
Plasma. Lasted till 530msec

May 19 1994 03:18:50:910PM940519015Steve Wolfe940519015
----------
Turn IC_EF1L control to zero at .2sec
Turn IC_EF2L control to zero at .2sec


Result:
-------
Fizzle

May 19 1994 03:22:40:770PM940519016Steve Wolfe940519016
----------
Lowered pressure to 1.2e-5


Result:
-------
Screamer down.

May 19 1994 03:35:45:040PM940519017Steve Wolfe940519017
----------
Change Br0 offset from -3 to -2mT


Result:
-------
Screamer down
Lasted till 65msec, 220ka.

May 19 1994 03:41:45:950PM940519017Steve Wolfe940519017
----------
Change Br0 offset from -3 to -2mT


Result:
-------
Screamer down
Lasted till 65msec, 220ka.

May 19 1994 03:52:25:430PM940519018Steve Wolfe940519018
----------
Leave Br_0 control on until 59msec


Result:
-------
fizzle. Early breakdown (5.5msec)

May 19 1994 04:01:47:730PM940519018Steve Wolfe940519019
----------
lower p to 1.1e-5

Result:
-------
fizzle - looks like the last one.

May 19 1994 04:01:57:500PM940519019Steve Wolfe940519019
----------
lower p to 1.1e-5

Result:
-------
fizzle - looks like the last one.

May 19 1994 04:18:03:220PM940519020Steve Wolfe940519020
----------
Change IC_EF1L and IC_EF1U from -4800 to -4500 at commutation.
This should give about 20mT/m more negative Bz'.

Result:
-------
No power supplies? \ENGINEERING::TOP.POWER_SYSTEM was turned off in the
tree!!!!!

May 19 1994 04:34:23:410PM940519021Steve Wolfe940519021
----------
Change IC_EF1L and IC_EF1U from -4800 to -4500 at commutation.
This should give about 30mT/m more negative Bz'.

Result:
-------
Plasma, lasts till 700msec
Early vertical thrash looks (and feels) MUCH BETTER NOW...
Breaks down at 6msec
The index looks good.
Now for the bad news- there is no null after 3msec.
RXL is still way in.

May 19 1994 04:51:21:070PM940519022Steve Wolfe940519022
----------
Change EF4 from -1020 to -1050A
Turn on Argon puff again

Result:
-------
fizzle- breaks down at 7msec.

May 19 1994 06:19:25:830PM940519023Steve Wolfe940519023
----------
Reload 21

Result:
-------
PLasma lasts till 500msec. Dies down.

Physics Operator Comments

Engineering Operator Comments
ShotTimeTypeStatusComment
109:07:58:520AMPlasmaOkoh2l + ef1 comm fault
209:27:29:310AMPlasmaOkno faults
310:05:53:970AMPlasmaBadef2l two fuses blown
412:06:32:340PMPlasmaBadef2l still has a problem
512:34:32:890PMPlasmaOkno faults
612:53:57:140PMPlasmaOkoh1 comm fault after plasma
701:08:40:700PMPlasmaOkno faults
801:23:58:720PMPlasmaOk.84 meg amp no faults
901:39:42:170PMPlasmaBad
1001:50:28:940PMPlasmaBadDUD no faults
1102:03:33:100PMPlasmaBadDud no faults.
1202:21:14:590PMTestOkplasma dis oh2l comm fail
1302:34:28:200PMTestBaddud power sys. O K.
1402:49:44:430PMPlasmaOkgood shot no faults
1503:08:21:780PMPlasmaBadDUD NO faults
1603:19:08:480PMPlasmaBaddud No faults
1703:30:15:130PMPlasmaOkcomm fault oh2l & ef1U plasma disr. early
1803:43:38:720PMPlasmaBaddud no faults
1903:56:37:830PMPlasmaOk
2004:08:12:400PMPlasmaBadNO shot P.S. off
2104:20:52:520PMPlasmaOkgood shot comm fault OH2 L
2204:45:14:620PMPlasmaBadDUD NO faults.
2304:54:48:020PMPlasmaOkGood shot NO faults.