Alcator C-Mod Run 930929 Information

Back to runs

Miniproposals
Miniproposal:010
Date Filed: 7/13/1993
Title:Transport in Ohmic Plasmas
First Author:Martin Greenwald
Session Leader:Robert Granetz (shots 1-31)

Operators
Session leader(s):Robert Granetz
Physics operator(s):Steve Fairfax
Engineering operator(s):Joe Daigle,Vinny Bertolino

Engineering Operator Run Comment
ohmic transport

Session Leader Plans

Physics Operators Plans

Session Leader Summaries
Entered: Jul 7 2004 03:49:09:023PM
Author: To Be Determined
Session leader summary for run 930929

EO: Daigle
PO: Fairfax
SL: Granetz

The object of this run was to explore low field operation. In terms of
meeting its objective, this run was quite successful. We started by
initiating plasmas at the nominal field of 5.3 tesla and ramping down the
TF during the discharges, which were about 1 second long. Shots 6 to 11
are good examples, with the latter ones having B ramping down to 3.8 T
during the current flattop (at 0.5 MA), and getting down to B=2.7 T at the
end of the shot. Starting with shot 12, we changed our modus operandi and
began trying to initiate plasmas at lower fields. Shot 12 broke down and
ran at 4.5 T, and was a very good shot. The field ramped down to 2.8 T
during the flattop. Following a series of fizzles and duds, we decided to
try breaking down at 3.5 T. After a half-dozen tries, during which the
vacuum null was being tweaked/improved, we got several good
plasmas at this lower field. See shots 22, 23, and 24. All had Ip=0.5 MA
and lasted for 1.0 s. The TF was programmed to ramp down during shots
23 and 24, and on the latter one, it got down to B=2.18 T during the
flattop. This resulted in significantly broader Te profiles, and had
q*=2.95 and q95=3.2 at t=0.8 s. No problems were observed with MHD
instabilities at these lower q's. For the last six shots of the day (26-31),
we tried to get breakdown at B=2.5 tesla, but we were unsuccessful due to field
null problems. I think that if we had had another hour, we probably would
have been able to tweak things enough to get a breakdown. All-in-all, I was
very pleased with what we accomplished today.

Scorecard: 31 shots = 11 plasmas + 5 fizzles + 1 TF test + 14 duds
(The duds were mostly due to power supplies not being able
to respond properly to the change in alternator loading caused
by the lower TF demands.)

Physics Operator Summaries
Entered: Jul 7 2004 04:36:50:397PM
Author: To Be Determined
Physics Operator's Summary

September 29, 1993

Session Leader: Bob Granetz
Physics Operator: Steve Fairfax

Engineering Operator: Joe Daigle
Power Systems: Vinnie Bertolino
Aux. Systems: Gerry Gerolamo
State: Sam Pierson


The purpose of today's run is to achieve low Q by progressively lowering the
TF. We started by loading shot 930928017. After getting a target plasma, we
began inverting the TF supply during the plasma. This reduced the TF at
roughly 2 Tesla/sec. Most of the run objectives were met by using this trick.
Shots 6-12 were good and shots 22-24 set progressive records for q*.
Shots 26-31 were spent in a vain attempt to get a breakdown at 2.5 Tesla.

Tuning of dBz/dR and Bz was relatively easy by changing EF1 and
EF4 currentl levels respectively. Increaseing EF1 pre-charge by 300 amps
while decreasig EF4 by 60 amps lowers dBz/dR by 10 mT/m without changing Bz.

Scorecard: 1 test (TF only for Bob G.)
11 plasma I > 50 kA shots
5 fizzles
13 duds (including a shot where turning the PCS OH2 gain too
high caused the power supplies to oscillate and trip.
1 engineering faiures when OH1 was left off.
--
31 shots
2 shots lost due to problems with hybrid tree or server

Shot 1 was a TF only. The TF integrator did not reset, but the pulse was OK.

Shot 2 was a fizzle with a trip in the OH2 power supply. The trip was probably
spurious; the conduction monitors in these supplies are still being
debugged for 4-quadrant operations. The fizzle did not show any
thrashing.

Shot 3 was a repeat; no changes in settings. Delayed gas start
from -19 msec to -16 msec. Started TF inversion at .5 seconds to
reduce recool time between shots. The alternator delivered 122 MJ
on this shot. The OH2U tripped again.

Shot 4 was an ugly disruptive shot. The PF breaker tripped. The alternator
delivered 100 MJ. The OH1 blew fuses. Reduced pulse gas 3 program
using stretch y from 54 volts peak to 50 volts peak.

Shot 5 was a dud because OH1 had been left disconnected from the AC line.
The alternator delivered 96 MJ. The OH2 supplies worked OK.

Shot 6 Nice shot -- Ip=.5MA; successful rampdown TF rampdown @ .5s = 4T @ 1s
fast current rise resulting in bounce on limiter

Shot 7 Ip=.48MA; TF rampdown @ .1s = 2.8T @ 1s
glowing spot on RF limiter. Two bounces on limiter during
startup. Density 3e19 nel. Oscillation in
z position starting around .8 seconds, during ip
rampdown. Raise D on slow z_cur from 1 to 4 at .75 sec. for
shot 8. Stretch PULSE_GAS_3 back to 55V for shot 8

Shot 8 Ip=.49MA; TF rampdown @ .1s = 2.8T; MARFE @ .25 chunks the plasma
current; VDE @ .77; colder and higher density this shot
D_ZCUR(2)=0 after .75

Shot 9 Ip=.5MA;
density disruption during current rise (shrink P_G_3 to 52V)
growing z oscillation - disruption @ .77 (turn off dropout in ZCUR
gains for shot 10)

Shot 10 Ip=.49MA; successful rampdown
density disruptions on current rampup (lower first two P_G_3 points
for shot 11)
z-oscillations growing throughout the shot - tied to OH2 ( drop wire 2
P gain from 4 to 1 at .76 sec. for shot 11)

Shot 11 Nice shot TF ramping from 4.8 to 4.0 during shot
set TF current limit to 130 kA for next shot

Shot 12 TF ramps from 4.5 to 2.8 during plasma current flattop
some hard x-rays. Increase gas after 75 msec from 32 to 35 volts.

Shot 13 Thrashing Fizzle. lowered 1volt gain in Br_0 from 9e-3 to 7e-3

Shot 14 Dud. Radial field better but vertical field off.
Set OH1 current limit to -10 kA to reduce TF core shearing.
Raised EF1 pre-charge currents from 4800 to 5100 amps to lower
dBz/dR by 1e-2

Shot 15 Fizzle. Very late breakdown. Bz is too high but dBz/dR is just as
we asked. Lowered EF4 pre-charge from 1150 to 1070 amps.

Shot 16 Fizzle. We figured out that the real problem was sagging in OH2
currents. Reloaded shot 12 and increased OH2 gain from 4 to 8.
Set TF current to 100 kA.

Shot 17 Dud. OH2 still sagging because TF flat-top is early.
Increased gas after 75 and 117 msec from 33 to 36 volts, at 207 msec
from 31 to 34 volts. Increased OH2 gain to 10. Set OH1 current program
to peak earlier.

Shot 18 Dud. Increased OH1 precharge to 20 kA at -.5 sec, then down to 16.6 kA
by commutation. This should help the OH2's get up to full current.

Shot 19 Dud. Increased M gain on OH2_Cur (set to max), Br.
Br gains on both OH2 and EF2 increased 50%.

Shot 20 OH2 supplies oscillated and tripped off early
Reloaded shot 19, set OH2 power supply inversion gains to less
sensitivity for AC input voltage, and set OH2 currents to peak
earlier.

Shot 21 OH2 currents OK. Br OK. Bz too high, and dBz/dR marginal.
Lowered EF4 by 50 amps.

Shot 22 Nice plasma. Set TF invert to 0.5 seconds, left everything else alone.

Shot 23 Nice plasma. TF ramps from 3.5 to 2.8 Tesla during pulse
removed decrease in Z_cur slow control at 0.75 seconds
Set TF invert at .25 seconds

Shot 24 Nice plasma. Some hard x-rays. Added 1 volt to gas programming
from 75 msec to 210 msec. Waited a few minutes for repairs when
hybrid tree was found slightly broken.

Shot 25 Dud. Set TF to 72 kA (2.5 Tesla), with inversion at +1 second.

Shot 26 Dud. No suprise. EF3 breaker tripped after shot. Bz is 4 mT too
negative. Added 70 amps to EF4 before shot.

Shot 27 Dud. Raised EF1 currents by 600 amps to lower dBz/dR by 20 mTesla.
Lowered EF4 110 amps to keep |Bz| constant. Lowered OH1 peak current
demand from 20 kA to 18 kA to reduce gratuitous TF shear.
Lost 11 minutes when hybrid server showed access violation.
New software interlock caught the error. Restarted bitbus server,
cold start hybrid, and did another load into model.

Shot 28 Dud. Did not get EF1 currents that we asked for. Decreased power
supply inversion endstop compensation gain.

Shot 29 Dud. EF1 commutation failure. Reloaded shot 27, changed OH1 peak
demand to 18 kA, increased endstop compensation gain.

Shot 30 Dud. EF1 commutation failure. Gradient Bz too high.
Raised EF1 300 amps. Lowered EF4 60 amps.

Shot 31 dud. Vacuum field traces aren't that bad, but flux plots show no null.



Session Leader Comments
Sep 29 1993 09:14:30:590AM930929001Robert GranetzTF only test at 5.3 tesla, for use in diagnosing the diamagnetic coils.
Unfortunately the TF bus rogowski did not get an integrator reset, so it
didn't register a valid signal. The magnetics got BTOR correctly, so I don't
know if we need to repeat this TF only test. It would be nice if we had a
fully documented TF shot, but I'd rather not delay plasma operation.

Next shot is for plasma.

Sep 29 1993 09:27:36:480AM930929002Robert GranetzFizzle. Fields look okay. Repeat with no changes.

Sep 29 1993 09:45:52:680AM930929003Robert GranetzAnother fizzle. Fields look good again. Try delaying the pulse gas by
3 msec for the next shot. Also, the TF inversion will start at 0.5 s instead
of 1.0 s, in order to get the recool time down, and to start the lower BTOR
operation if we get plasma.

Sep 29 1993 10:14:37:210AM930929004Robert GranetzPlasma, but short and ugly. Ended with disruption due to a fault in the
PF supplies. However, the shot was very disruptive throughout, due to
extremely high density (nl=1.e20 by t=0.3 s). For the next shot, reduce the
pulse gas voltage programming everywhere by nearly 10%.

Sep 29 1993 10:17:32:280AM930929004Robert GranetzApparently we need less pulse gas puffing at the beginning of a run to get the
same density as a shot from the middle or end of the previous day's run. This
may be due to pre-run ECDC loading up the walls with gas.

Sep 29 1993 10:21:53:410AM930929005Robert GranetzNo breakdown. OH1 didn't fire. The breaker was inadvertently left open
after fixing a blown fuse after the previous shot. Repeat for the next
shot.

Sep 29 1993 10:51:27:230AM930929006Robert GranetzNice shot. Ip=0.50 MA, duration=1.02 s. TF started ramping down at 0.5 s,
and by 1.0 s it was at 4 tesla. The current rise was much faster than previous
shots (due to the decreased gas puffing?), and this caused the current to go
screaming to the outer limiter at 0.019 s, with a restrike starting at 0.020 s.
For the moment, I'm not going to do anything about it.

For the next shot, change the TF inversion time to t=0.1 s. No other
changes.

Sep 29 1993 11:13:18:800AM930929007Robert GranetzGood shot. Ip=0.48 MA, duration=1.02 s again. The TF starting ramping
down at 0.1 s and reached a value of 2.8 T at 1.0 s. There are very bad
problems with the fast current rise, which is causing the plasma to hit
the outer limiter (twice on this shot). The plasma is also riding on the
outer limiter throughout the shot. The RF limiter is not designed to be
the primary limiter, so for the next shot I've directed Yuichi to pull
the limiter and antenna back by a cm (from 0.92 to 0.93 m). Also, in order
to get the current rise better matched with the Bz programming, I can either
fiddle with the PF programming, or increase the pulse gas (to slow down
dI/dt). I'll try the latter since it's simpler. For the next shot, increase
the pulse gas programming back up by about 10%. It will now be the same as
shot 930928017, which was the template for the beginning of today's run.
Leave the TF the same.

Sep 29 1993 11:29:57:860AM930929008Robert GranetzDisrupted at 0.77 s, due to problems with vertical stability. The kappa has
been getting up to 1.6 on these low field shots, maybe because the q is getting
lower. There was a marfe from 0.25 to 0.33 s, which seems to cause the dI/dt
to go to zero for that time. For the next shot, change the derivative gain on
the EFC, and try to recover from the disruption.

Sep 29 1993 11:44:26:700AM930929009Robert GranetzNearly identical to the last shot. Fast disruption (1 msec) at 0.77 s. The
plasma was disruptive early on, so for the next shot, lower the gas programming
by 5% (from 55 volts max to 52.5 volts. Also, get rid of the short dropout
programmed into the ZCUR gain. (This was for position control response
tests for Gerry Tinios's thesis. It seems to kick the vertical position
control "over the edge".) As with the previous few shots, the TF has been
ramping down from 5.3 at 0.1 s to 2.8 T at 1.0 s.

By the way, we will have to do something with the high kappa problem after the
next shot.

Sep 29 1993 12:05:49:280PM930929010Robert GranetzGood shot. No disruption. Ip=0.49 MA, duration=1.0 s. Plasma was disruptive
during the first 0.2 s, so for the next shot, reduce the pulse gas a little
during the first 50 ms only. The kappa is reaching 1.6 again, resulting in
some Z oscillations which grow toward the end of the shot, but this time it
survives, possibly because I removed the dropout pulse in the gain. TF ramp
was the same as previous shots. For the next shot, enable the TF PLC current
limit at 140 kA (requires setting the limit at 144 kA on the PLC). There
were multiple moly injections. Also, reduce the gain on the slow timescale
ZCUR wire in an attempt to reduce the oscillations (gain reduced after 0.75 s).


Sep 29 1993 12:18:53:840PM930929011Robert GranetzGreat shot. Relatively moly and disruption-free. Duration=1.02 s. TF was
limited to 140 kA, but this barely made any change (as expected). The ZCUR
control was much more stable this time, so the change that Steve Horne made
to the slow timescale ZCUR feedback gain seems to have worked. For the
next shot, limit the TF to 130 kA. This should reduce the field by 0.33 T,
but may screw up the breakdown.

Sep 29 1993 12:28:59:750PM930929012Robert GranetzAnother very good shot. It broke down at 4.5 Tesla, which is what the PLC
limit corresponds to. At 1.0 s, the TF was down to 2.4 T. This shot had
some hard x-rays starting at 0.070 s, and the breakdown hit the outer limiter.
For the next shot, increase the pulse gas slightly, starting at 0.070 s. No
other changes.

P.S. The breakdown fields may have been affected by the lower BTOR, as
evidenced by the slightly later breakdown time. We've had a report that
the field configuration is not good at breakdown.

Sep 29 1993 12:40:46:430PM930929013Robert GranetzFizzle. The Br is at about +1.0 to +1.5 mT, and we would like it to be
nominally at -0.5 mT, so for the next shot, program the Br offset down by
2.0 mT. No other changes.

Sep 29 1993 01:00:04:470PM930929014Robert GranetzNo breakdown at all. Br is okay, but the null looks really crummy. For
the next shot, reduce the dBz/dr at breakdown by raising the EF1U&L by
300 amps each. Effectively, this compensates for the reduced single turn
TF current in the corners. No other changes.

Sep 29 1993 01:12:32:860PM930929015Robert GranetzFizzle at least. Bz is 4 mT too high (consistent with the change we made
in the EF1's). For the next shot, drop the EF4's by 70 amps (4-5 mT less Bz).


Sep 29 1993 01:29:21:680PM930929016Robert GranetzFizzle. After some consultation, it appears that the field null changes have
been due to the fact that the TF has been getting into regulation earlier
(at -0.1 s, since we've lowered the TF), and this causes changes in the
OH source voltages. Therefore, we've decided to take the plunge off the
cliff and ask for 3.5 T breakdowns (I_TF=100 kA), and increase the feedback
gain on the OH2's. The TF will go into regulation sooner, allowing more
time for the OH2 feedback to respond (especially coupled with the increased
gain). Then we'll start tweaking. The TF inversion has been moved out
to 1.0 s.

Sep 29 1993 01:31:03:670PM930929016Robert GranetzP.S. We re-loaded shot 012 for the next shot, but the TF current limit will
be set at 100 kA (3.5 T).

Sep 29 1993 01:56:35:610PM930929017Robert GranetzNo breakdown. The OH2 available max voltage drops to only about 50 volts
(from about 150 volts) when the TF goes into regulation. Thus the OH2's
can't get up to the requested currents by t=0. Part of their problem is
due to fighting induced current from the OH1, which is also trying to rise
at this time. For the next shot, program the OH1 to come up earlier, so that
it has flattopped by the time the OH2's have to ramp up.

Sep 29 1993 01:58:52:830PM930929018Robert GranetzNo breakdown. The OH2's still didn't make it to the requested current by
t=0. For the next shot, Steve Fairfax is going to try a hairbrained idea:
get the OH1 up to 4 kA more than we want, then ramp it down to the desired
level a little before t=0. This will induce positive currents in the OH2's,
which should help them reach their limit.

Sep 29 1993 02:09:39:360PM930929019Robert GranetzNo breakdown. The OH2's got closer to the demand current, but still lag by
about 900 amps. The feedback gain has already been max'ed out, but the
M value in the controller has only been 20% of the maximum. For the next
shot, increase the controller M value to its maximum.

Sep 29 1993 02:14:32:160PM930929019Robert GranetzAlso for the next shot, the M values in the Br controller were increased
by about 50%.

Sep 29 1993 02:21:55:800PM930929020Robert GranetzPower supplies faulted very early. It looks like the OH2's were unstable.
For the next shot, don't bring up the M*Gain until later in the rampup. Also
increase the voltage pre-programming on the OH2's.

Sep 29 1993 02:51:33:630PM930929021Robert GranetzNo breakdown. OH2's got to the demand currents. Bz is about 2-3 mT too
high. For the next shot, reduce EF4 by 50 amps.

Sep 29 1993 03:00:58:190PM930929022Robert GranetzGot it! An 0.5 MA plasma for 1.0 seconds at BTOR=3.5 Tesla! No disruptions.
There was some kind of impurity injection or something at 0.78 s. Also there
were some oscillations in ZCUR after 0.8 s. For the next shot, start the
TF inversion at 0.5 s, so that BTOR ramps down to 2.5 T by 1.0 s.

Sep 29 1993 03:17:46:020PM930929023Robert GranetzAnother good shot. Ip=0.48 MA, duration=1.0 s. No injections, no disruptions.
A really nice shot. Zeff=2.0. For the next shot, invert the TF at 0.25 s.

Sep 29 1993 03:34:45:870PM930929024Robert GranetzAnother good shot. Ip=0.48 MA, duration=0.98 s. BTOR ramped down starting
at 0.25 s (where B=3.5 T), to B=2.24 T at 0.8 s, and 1.84 T at the end of
the shot. There was just a hint of hard x-rays on this shot. John Rice wants
a repeat, so we'll do it. In order to avoid a possible runaway problem, we
have increased the pulse gas voltage by only 1 volt, starting at 70 ms.
Amanda and John Rice report a much broader current profile. Steve Wolfe
reports that at t=0.8 s, q*=2.95 and q95=3.2.

Sep 29 1993 03:36:47:830PM930929025Robert GranetzNo breakdown. Reason is unknown. We're not going to deal with it though,
since we want to go to 2.5 T breakdown and flattop. That should cause
more problems with breakdown anyway.

Sep 29 1993 03:52:35:230PM930929026Robert GranetzNo breakdown. No surprise. Bz was too negative by about 4 mT. For the next
shot add 70 amps to EF4.

Sep 29 1993 04:08:40:630PM930929027Robert GranetzNo breakdown. Bz is now okay, but dBz/dR is too high by about 2e-2 T/m.
To fix this, raise EF1's by 600 amps. This increases Bz as well, so reduce
EF4 by 110 amps also for the next shot.

Sep 29 1993 04:38:13:830PM930929028Robert GranetzNo breakdown. The EF1's did not give us the extra 600 amps that we demanded.
However, the EF4's did drop the requested 110 amps, so we ended up with about
8 mT too negative a Bz. For the next shot, the engineers went out and
adjusted something on the EF1's.

Sep 29 1993 04:51:25:590PM930929029Robert GranetzNo breakdown. EF1's made it to the demand current, but now there's too
negative a gradient. Re-load shot 027 and try again.

Sep 29 1993 05:00:30:750PM930929030Robert GranetzNo breakdown. dBz/dr is too high by about 1e-2 T/m. Increase EF1's by
300 amps, and compensate the increased Bz by decreasing EF4 by 60 amps for
the next shot.

P.S. The next shot will be the last one of the run.

Sep 29 1993 05:12:29:410PM930929031Robert GranetzNo breakdown. EF1's responded okay. There was about 1 mT too much Bz.

End of run.

Physics Operator Comments

Engineering Operator Comments
ShotTimeTypeStatusComment
109:02:40:580AMTestOktf only bus rog. did not work.
209:17:21:230AMPlasmaBadOH2L > than one scr not conducting fault.
309:34:07:680AMPlasmaBadOH@L > than one scr not conducting.
409:50:23:590AMPlasmaBadOh1 both oh2's &ef1 faulted comm.fail on all .
510:17:27:530AMPlasmaBadOH1 Breaker not in
610:32:24:160AMPlasmaOkGood shot
710:50:15:470AMPlasmaOkGood shot 480ka plasma shot.
811:11:49:700AMPlasmaOkComm fault OH1 & EF1 lower.
911:27:10:310AMPlasmaOkno faults
1011:43:50:510AMPlasmaOkno faults
1112:02:51:640PMPlasmaOkno faults
1212:17:40:910PMPlasmaOkno faults
1312:30:57:610PMPlasmaOkNo faults fizzle
1412:43:29:020PMPlasmaOk
1501:00:38:860PMPlasmaOkNO faults.
1601:14:34:130PMPlasmaOkNo faults Fizzle.
1701:27:39:420PMPlasmaOkNo faults dud
1801:49:37:610PMPlasmaOkNo faults Dud
1902:02:09:520PMPlasmaOkNo faults Dud.
2002:13:28:160PMPlasmaBadThe OH2 power supplies single phased
2102:33:59:170PMPlasmaOkNo faults Dud.
2202:48:05:450PMPlasmaOkPlasma shot very good
2303:02:05:820PMPlasmaOkNo faults good shot
2403:15:33:010PMPlasmaOkNO faults good shot.
2503:31:25:640PMPlasmaOkNo faults Dud.
2603:44:24:830PMPlasmaOkNo faults Dud.
2703:57:48:310PMPlasmaOkNO faults Dud.
2804:09:41:060PMPlasmaOkno faults
2904:36:47:550PMPlasmaBadCommutation fault EF1 Lower.
3004:50:13:100PMPlasmaBadEf1 L comm fault in inv. rev supply.
3105:02:30:310PMPlasmaOkNO faults Dud.