Alcator C-Mod Run 1050329 Information

Back to runs

Miniproposals
Miniproposal:416
Date Filed: 2/24/2005
Title:Compensating input offsets in DPCS
First Author:Steve Wolfe
Session Leader:Steve Wolfe (shots 1-10)
Miniproposal:417
Date Filed: 2/24/2005
Title:Plasma performance, boronization and ICRF heating effectiveness with high-Z antenna guards and all-m
First Author:Yijun Lin
Session Leader:Yijun Lin (shots 3-25)

Operators
Session leader(s):Steve Wolfe,Yijun Lin
Physics operator(s):Jim Irby
Engineering operator(s):Bill Cochran,Gary Dekow,Bill Parkin

Engineering Operator Run Comment
MP #417

Session Leader Plans
Entered: Mar 28 2005 02:57:23:037PM
Author: Steve Wolfe
Plan for MP#416 "Compensating input offsets in DPCS" 'piggyback'

This experiment is intended to verify the correct operation of the 
realtime baseline subtraction algorithm baseline_sub, which is contained in
/usr/local/cmod/codes/dpcs/baseline_sub_init.pro. For each shot, this
routine performs a baseline average of the dpcs input signals, based 
on averaging time intervals specified in the tree in hybrid::top.dpcs_config,
and, beginning at a time designated by the node 
dpcs::top.loadables.procedures.input_pros:switch_times
subtracts these baseline values from the inputs before doing the control
calculations. For times earlier than the averaging intervals the routine 
can optionally subtract a pre-determined fixed offset, also given under
hybrid::top.dpcs_config
This mechanism should provide robust correction of drifts in the input
signal baselines, and may ultimately improve reproducibility. In the short
term, implementation of this technique will require respecification of some
"hand-tuned" offsets in the observer algorithms specified in the PCS interface.

The algorithm has been extensively tested offline, and appears to be working 
correctly. This satisfies step 0 of the shot sequence in the MP. The remainder
of the shot sequence will be modified somewhat from the MP, because I have
decided to take the advice of the EPC and go straight to the more complicated
"live" routine, rather than separately test the scheme that just subtracts
fixed offsets.

1. Begin with the software in place, with switch_times set to +2.9 sec. This
will allow us to see the effect of the subtraction at a time well after the 
actual shot. This should verify proper operations, and the jump in AOUT values
will provide feedback on the magnitude of the changes that will be required to
 the predictor offset values. These should be compared to the values inferred 
from offline testing. 

[1 shot]

2. Move the switch_time from 2.9 to 0.10, the start of the second segment
in the PCS setup. This will test the subtraction in the equilibrium phase
without impacting the startup. Change the second segment offset values as
indicated in step 1. Offline testing indicates that the appropriate changes
will be to remove the offsets on CLEARIN and IP, i.e.
   CLEARIN offset from 12700->0
   IP offset from -70000 -> 0
and leave everything else alone. If we're really paranoid, we could first add a 
predictor switch at 2.9sec (or 2.95) incorporating these changes, 
leaving the switch_time=2.9, and see if the Aout jumps at the switch time go away. 

Compare position and equilibrium results, and if necessary tweak further. 
[1-2 shots]

3. Move the switch_time to -2.1sec, before the startup. Since this is before
the specified averaging intervals for the bus currents and the magnetics
(both set for [-2.,-1.9], just after the integrator gates), we will see 
the effect of first the fixed offsets and then the calculated baseline values. 
The key values are the Br0 offset and the RCUR offset. Fortunately, the 
indications are that the change in RCUR is pretty negligible, perhaps 
500A-m; however, the Br0 may need about a 2mT adjustment. I think the correct
direction is positive, e.g. change Br0 offset from -.0025 to -.0005. We 
may lose a shot at this point if we don't get it right. 

Observe the breakdown and current rise, and adjust accordingly.
[2-3 shots]

4. Move switch_time to -3.45 seconds. This should have no effect except that
it will move the subtraction to the beginning of the data window. Add 
a comment to hybrid::top.config:config_id describing the change in DPCS 
operation and necessary actions. Physop should exit PCS and restart it to pick
up the message. Leave dpcs set up to use the subtraction from this point on. 

>> If at any point in the above sequence things are not going according to 
plan, or if it is severly impacting the run, then punt on this MP and
remove the offset subtraction routine from node dpcs::top.loadables.procedures:input_pros  
until we're ready to try again. 



Entered: Mar 28 2005 03:42:30:007PM
Author: Yijun Lin
Run plan for 1050329

MP417 (day 1 of pre-boronization)
"Plasma performance, boronization and ICRF heating effectiveness with
high-Z antenna guards and all metal surface"

Purpose:
The run will be part of the Level 1 JOULE target of FY 2005. This run
and the next run would complete the pre-boronization part of MP417.

Goals:
We will study the plasma behavior at different RF power, density, impurity
puff, etc, and compare to results obtained from previous campaigns.
After these two runs, we will be able to figure out the
answers to the following questions:

1. Do we get better plasma performance after removing the BN tiles and
the thick boron layer?
2. Is there any change in RF heating efficiency?
3. How do the radiated power level and profile vary with the RF power,
impurity puff and plasma confinement modes?

Gas setup:
Setup for usual plasma operation plus Nitrogen, Ar and Ne puff.

---------------------------------------
Shot Sequence:

Use shot 1040212030 configuration for the fiducial plasma.
LSN, 5.4 T, 1 MA, nl = 0.9e20 m^-2.
In most shots, an RF notch should be added near the end of the RF
pulse to have dW/dt information. The dip of the notch will be around 1 MW,
but not too large to cause H-L transition. The length of the notch is about 20 ms.

------1050329---

2 shots: Antenna comparison in H-mode.
B = 5.2 T, 1 MA, nl04 = 0.9e20 m^-2.
A shot with D+E (2 MW, 0.6-0.9 sec) and J (2 MW, 1.0-1.3 sec), the other with
D+E (2 MW, 1.0-1.3 sec) and J (2 MW, .6-.9 sec).

4 shots: Performance at RF power scan. RF power at 1.5, 2, 3, 4 MW.
Same power from each transmitter. Fiducial plasma. (Also determine the
max RF power that can be delivered reliably. If 4 MW is not reachable,
do 3.5 MW).

6 shots: L-H threshold.
Ramping up RF power from 0.5 to 3 MW.
Density scan nl04 = 0.8, 1.0, and 1.2e20 m^-2.
Also current scan at 0.8 MA and 1.0 MA.

4 shots: Performance vs density.
density scan nl04 = 0.7, 0.8, 1.0, and 1.1e20 m^-2 at the max RF power.

2 shots: Heating hysteresis.
RF notch (1 MW dip, 20 ms length) at 40 ms and 80 ms after the RF power on.
Max RF power. Fiducial plasma.

5 shots: Density scan at L-mode inner wall limited plasmas. Check the scaling with
density in L mode confinement.
nl04 = 0.7, 0.8, 0.9, 1.0 and 1.1 e20 m^{-2} at max RF power (~4 MW).

Perhaps we should stop here for this run day and leave the rest
for 1050330. If we do have time, continue the following.

--------1050330-------------

5 shots: Argon puff (calibration of bolometry and estimate of RF efficiency).
Puff Argon to ohmic plasma until 100\% radiated power,
then repeat a shot with 3 MW RF to 100\% radiated power.

5 shots: Repeat above with Neon puff.

5 shots: Nitrogen puff. 3 MW RF power. Start with small puff
until 100\% radiated power.

3 shots: No puff. Repeat at 3 MW to see whether the N effect lasts.

5 shots: ECE cutoff for Thomson calibration.

Physics Operators Plans
Entered: Mar 29 2005 07:35:03:010AM
Author: Jim Irby
Engineering setup for Tuesday, 29 Mar 2005

MP #417 Plasma performance, boronization and ICRF heating effectiveness
with high-Z antenna guards (Lin)

and

MP #416 Compensating input offsets in DPCS (Wolfe)


Session leader: Lin, Wolfe
Physics operator: Irby

Overnight ECDC in D2, bake at 60 C.

Run begins at 09:00 and ends at 17:00.

Power systems as on 1040212030

A-coils in standard configuration (lock mode suppression):
+Dtop -Dbot -Jtop +Jbot (same as 1050316)

Gas Setup:
fill B-Top with 6 psi D2 Hybrid enabled (PG4)
fill B-side lower with 2 psi Ar Hybrid enabled (PG1)
leave B-side upper with He3 Hybrid disabled
fill B-main (C-side) with 30 psi D2 Hybrid enabled (PG3)
leave J-bot Hybrid disabled (PG2)

Enable the following gate valves:
ECE, Z-bolo shutters

Session Leader Summaries
Entered: Mar 29 2005 03:52:45:397PM
Author: Steve Wolfe
Session Leader summary for MP#416 piggyback on 1050329
-------------------------------------------------------

The first part of the plan was accomplished, but implementation of
baseline subtraction for the entire shot was not successful today.

Baseline subtraction was successfully implemented for segment 2,
with the "correct" value for the predictor offsets on CLEARIN and
IP both at 0.0 (formerly 12700A-m and -70000A). This was done as of shot
1050329007. The corrected offset for SSEP (formerly -2500) was not
determined, but can probably be investigated offline, and is probably
also zero. A disruption on shot 6 was found not to be associated
with the control system changes, and was probably a lower hybrid effect.

The attempt to move the subtraction to the startup segment
(shots 8 and 9) resulted in two duds. The Br0 offset apparently
needed to be changed by going more positive by .0015, but the
change to IC_EF4U and RCUR were not successfully made. I tried
making IC_EF4U more positve by 10A (shot 8: -1335->-1325) and then
more negative by 10A (shot9: -1335->-1345) but the Bz remained too
positive. Not wanting to further delay the main run, I punted on the
early subtraction and put the switch_time back at the beginning of
segment 2 (+0.1sec) for shot#10, which was successful.

I have added an entry to hybrid::top.config.config_id calling for removal
of offsets in the second segment. This must be done to avoid a serious
error in CLEARIN (and a 70kA error in Ip).

We can/should revisit the final implementation of the baseline subtraction in
a later run. After giving up for the day, I discovered that, contrary
to my expectation, the RCUR proportional gain in segment 1 actually
turns on at t=0, not after breakdown as I thought. Therefore, if the change
in the RCUR offset (positive by 500A-m) was incorrect, this could have
been responsible for the duds. This, however does not explain why Bz
before t=0 was not responding to changes in IC_EF4U program.
programming. I would expect no more than 4 or 5 additional shots will be
required to complete the MP, but I will try to get these done in a run that
hasn't already had several hours of delays.

Looking back at the baseline subtraction in segment 1,
I find that in fact the other coils have a bigger baseline
offset issue than IC_EF4U. The jumps in AOUT (taken from shot 9,
considering the change from no subtraction to live average
subtraction, i.e. before -2.1 compared to after -1.9) are approx:
OH1 -440A
OH2 -430A
EF1U -100A
EF1L -70A
EF2U -13A
EF2L -16A
EF4U -14A
So we were compensating the wrong thing.
For the next try, we should re-program at least the OH1 and OH2 currents
to be what we want, namely something like
IC_OH1 = -19.6e3 instead of -20.0e3
IC_OH2U= -27.1e3 instead of -27.5e3
as well as Br0 offset more positive by 1.5e-3 and
IC_EF4U more POSITIVE by 10 or 15A. The EF2 change is probably
too small to bother with. CHanging the EF1's from -6000A to -5900A
wouldn't hurt, but probably won't help either.
Entered: Mar 29 2005 05:10:48:490PM
Author: Yijun Lin
Summary:
Due to numerous operational problems, we only have 8 useful plasmas.
We did finish the first two parts of the run plan: antenna comparison
and RF power scan. The RF appears to be heating, but the H-modes only triggered by the sawtooth heat pulses. The plasma went into H-mode after a sawtooth, then had a H-L transition after the edge was cooled down (by excessive radiation?). H factor has been around 1 for almost all useful shots. H/D ratio was creeping up during
the run and up to > 0.1 in shot 16. All in all, we have had some plasmas that
can be compared with future post-boronization shots.

Physics Operator Summaries
Entered: Mar 29 2005 06:13:23:723PM
Author: Jim Irby
Not a good day. Many duds followed shot 16, our last plasma shot. We
are still not sure what caused the problem. We tried both wire and
gas solutions with no positive results. Impurity levels as determined
by the before/after RGA data were worse than on 1050318 and increased
from shot 13 to shot 16. Increases were seen in mass 18, 44, 49. All
twelve klystrons operated into vacuum and plasma for the first time
today. About 250 kW of LH power was run on shot 6, with numerous
reflected power trips. On shot 7 the power was reduced, but we were
still able to see the grill light up on wide2. The coupler seemed to
condition during the run (less light on video or h-alpha).

Luckily we began cycle tests at 8:15 this morning. A broken fiber to
the reflectometer rack (gka502) caused troubles. A major problem
also arose with cmodws34 which eventually had to be rebooted. In spite
of an early start these problems delayed our first plasma until 10:30.
We had another couple of data system problems during the day. We lost
the data for shot 19, no store cycle started following the shot.

Score:

cycle tests 01,02,03,04,19
duds 08,09,18,20,21,22,23,24,25
fizzles 12,17
plasmas 05,06,07,10,11,13,14,15,16

Session Leader Comments
Mar 29 2005 07:57:15:510AMSteve Wolfe
DPCS is set up for the first shot of the MP#416 piggyback:
\dpcs::top.loadables.procedures:input_pros = ["baseline_sub"]
\dpcs::top.loadables.procedures:input_pros:switch_times = [2.9]

Fixed offsets in 
\hybrid::top.dpcs_config.inputs:input_*:offset
and averaging intervals in
\hybrid::top.dpcs_config.inputs:input_*:offset:base_intrvl
are as previously set (see DPCS logbook entries from Jan06 and Mar18).

/usr/local/cmod/codes/dpcs/baseline_sub_init.pro is dated Mar 28 08:17
and is the version tested yesterday

/usr/local/cmod/codes/dpcs/dpcs_startup.pro is identical to
/usr/local/cmod/codes/dpcs2/dpcs_startup.pro-Mar28

/usr/local/cmod/logs/dpcs.log indicates
Tue Mar 29 05:00:10 EST 2005, server process starting
and looks like a normal startup. 

I have signed in as SL (and had the signin widget appear in 
this logbook entry window, forcing me to retype all the above).

I think we're ready to start the piggyback

Mar 29 2005 08:23:43:340AM1050329001Steve Wolfe
Shot#1: Cycle test- it's not in the run plan but is a good
idea. Hung in Init, no action showing on the monitor. 
Tom's working on it.


Mar 29 2005 08:30:00:370AM1050329001Steve Wolfe
Shot#1: Cycle test- it's not in the run plan but is a good
idea. Hung in Init, no action showing on the monitor. 
Tom's working on it.

Ran through the sequence manually, very short wait in check.
Log file shows init finished (20.3sec) but no real_time or store. 
This should definitely tickle the new cleanup and garbage collection
features in dpcs_init next time through. 


Mar 29 2005 08:52:34:900AM1050329001Steve Wolfe
Shot#2: Cycle test. Had to reboot the PC on Ratiomatic and NINJA 
before this cycle. 

Camac server lost errors in init. CAMAC2 and CAMAC3 servers
at least are reporting errors.  


Mar 29 2005 09:07:28:667AM1050329002Steve Wolfe
Shot#2: Cycle test. Had to reboot the PC on Ratiomatic and NINJA 
before this cycle. 

Camac server lost errors in init. CAMAC2 and CAMAC3 servers
at least are reporting errors.  

DPCS timed out? That shouldn't have happened, it was a very fast 
check? Maybe it actually didn't get through dry run until after
the trigger?

No, I think somehow that was the result for shot 1. It must have
run store as well, because there were no heap_free messages in 
the log for shot2.

Shot 2 seems to have run through on dpcs. I see the effect of the
 baseline sub on the aout traces at 2.9 seconds.
Baseline outputs are there in
hybrid::top.hardware.dpcs.signals:a_in:baselines
So DPCS seems to be ok on shot#2, even if all the CAMAC things
were dead. 


Mar 29 2005 10:03:51:807AM1050329003Yijun LinData cycle.
Mar 29 2005 10:08:31:043AM1050329003Steve Wolfe
Shot#3: still trying for a successful test cycle. 
Got gas. Looks like .08mTorr from the pre-puff. 
May have started with the gate valve still closed 
from the previous aborted INIT? 

dpcs.log file looks ok, 20.45sec in dpcs_init,
No errors showing up. Outputs still look like they 
do on test shots. 

I had Jim get rid of the late points in the Acoil waveform
at 10sec and 6sec, and this time dpcs is back to 85000 
samples (from 135000).

There seems to be some problem with the gate valve
PLC logic, but it's not my problem.

Mar 29 2005 10:23:35:167AM1050329004Yijun Lindata cycle
Mar 29 2005 10:26:13:230AM1050329004Steve Wolfe
Shot#4: Still working on a dead CAMAC crate. 
Cycle test, got gas again. Gate valves seem to be
doing the right thing. 

Mar 29 2005 10:39:34:793AM1050329005Steve Wolfe
Shot#5: Try for plasma 
Plasma full length. 
Next shot: Move switch_time to 0.1 sec
Advise Jim to change offsets on CLEARIN and Ip to zero.

Mar 29 2005 10:46:35:433AM1050329005Yijun Linshot 05: load shot 1040212030 (lower B to 5.2 T for antenna comparison).
Plasma: Bt 5.2 T, 1 MA, nl04 = 0.9e20m^-2.
Shape: LSN, Kappa = 1.65, delta_u = 0.28, delta_l = 0.5, outer gap = 1 cm.
RF power:
D+E: 2 MW. t = [0.6, 0.9] sec
J: 2 MW, t = [1, 1.3] sec
result:
E antenna was faulting.
H/D at 0.05 - 0.06 level.

Mar 29 2005 11:47:55:007AM1050329006Yijun Linshot 06: (no Argon)
Plasma: Bt 5.2 T, 1 MA, nl04 = 0.9e20m^-2.
Shape: LSN, Kappa = 1.65, delta_u = 0.28, delta_l = 0.5, outer gap = 1 cm.
result:
Early disruption at about 0.5 sec (LH was to blame?)
Mar 29 2005 11:59:02:630AM1050329006Steve Wolfe
Shot#6: Move switch_time to 0.1 sec so we'll 
do subtraction in seg2. Physop to remove offsets
in seg2 (CLEARIN and Ip are the only ones). Also 
put an ssep with zero gain and -2500 offset on wire 11.
Unfortunately, putting the switch at seg switch time isn't
going to let me see what the jump is. 

Plasma disrupts at .5sec, 
Ip looks ok. Equilibrium looks ok. 
There is a rise on the bolo at .507, followed by injections.
It's right during the LH pulse. Could they actually be doing
something? 
Aout signals all seem happy. 
CLEARIN aout signal is continuous at segment switch, which 
would be fine except that in seg1 that wire claims to be 
zxl! Could be a coincidence, but the real zxl on wire 6 looks
different. Furthermore, the zxl supposedly on wire 4 in seg1 has
a later version date. This is probably a red herring. Signals
seem consistent with shot 5. 
   

Mar 29 2005 12:07:59:287PM1050329007Steve Wolfe
Shot#7: Try again, previous may have just been an injection. 
Plasma to 1.7sec, something is lighting up on the rhs of wide2
Gaps look good. I think the subtraction is working. 

The baseline data look ok in the tree, the log file looks good. 
I think we are ready to do the startup. 
 

Mar 29 2005 12:09:10:973PM1050329007Yijun Linshot 07: (LH was asked to drop power)
Plasma: Bt 5.2 T, 1 MA, nl04 = 0.9e20m^-2.
Shape: LSN, Kappa = 1.65, delta_u = 0.28, delta_l = 0.5, outer gap = 1 cm.
RF power:
D+E: 2 MW. t = [0.6, 0.9] sec
J: 2 MW, t = [1, 1.3] sec
result:
ran through.
Not able to stay in H mode. Multiple LHL transitions.
next: swap the order of the antenna.
Mar 29 2005 12:11:05:160PM1050329007Steve Wolfe
Shot#8: Change switch_time to -2.1
Recommend changing offset on 
  Br0 more positive by 1mT
  RCUR more positive by 500
Waveform on IC_EF4U more positive by 10A

Mar 29 2005 12:34:15:787PM1050329008Steve Wolfe
Shot#8: Change switch_time to -2.1
Recommend changing offset on 
  Br0 more positive by 1mT
  RCUR more positive by 500
Waveform on IC_EF4U more positive by 10A

Dud. Br0 looks more negative by .0005T, 
and Bz looks more positive.

Next shot: another .5mT more positive on Br0
and reverse the change on IC_ef4u

Mar 29 2005 12:39:23:363PM1050329008Yijun Lindud
Mar 29 2005 12:44:52:287PM1050329009Steve Wolfe
Shot#9: switch_time still at -2.1sec
 another .5mT more positive on Br0
 and reverse the change on IC_ef4u 
Another dud. Br changed as expected, Bz looks unchanged?

Punt on the startup, back to 0.1sec for the 
switch_time, Jim reloads seg1 from shot7

Mar 29 2005 12:47:22:007PM1050329009Yijun Lindud
Mar 29 2005 01:00:00:660PM1050329010Steve Wolfe
Shot#10:
Punt on the startup, back to 0.1sec for the 
switch_time, Jim reloads seg1 from shot7

Plasma: Must have been the baselines. Br0 is just
like on shot 9 and Bz is back where it should be. 
I don't understand why Bz didn't go more negative 
than it did on shot 9. However, I don't want to interfere
any further with the run. 

Leave the switch time at 0.1sec and put the message in 
the hybrid config_id. Come back to the startup in a later 
run. 

Mar 29 2005 01:01:04:363PM1050329010Yijun Linshot 10:
Plasma: Bt 5.2 T, 1 MA, nl04 = 0.9e20m^-2.
RF power:
D+E: 2 MW. t = [1, 1.3] sec
J: 2 MW, t = [.6, .0] sec
result:
ran through.
Not able to stay in H mode. Multiple LHL transitions.
(The shots got so far should be enough to compare with boronized in future).

Next: start rf power scan. Ran 1.5 constant RF.
Mar 29 2005 02:01:05:130PM1050329011Yijun Linshot 11: (Performance at RF power scan, 1.5 MW)
Plasma: Bt 5.4 T, 1 MA, nl04 = 0.9e20m^-2.
RF power: 0.375 X 4 = 1.5 MW, [0.6, 1.35] sec. Notch on J port at 1.2 sec
result: Multiple L-H transitions.
Woh = 49 kJ, Wmax = 80 kJ, P_rad main = 1.3 MW,
@Wmax, Prf = 1.64 MW, dW/Prf = 25 kJ/MW, H89 = .92
@notch, efficiency = .7 (large uncertainty)
H/D from 0.05 to 0.09 during the RF pulse.
next: 2 MW RF.
Mar 29 2005 01:31:49:053PM1050329012Yijun Linfizzle
Mar 29 2005 02:01:59:460PM1050329013Yijun Linshot 13: (Performance at RF power scan, 2 MW)
Plasma: Bt 5.4 T, 1 MA, nl04 = 0.9e20m^-2.
RF power: 0.5 X 4 = 2 MW, [0.6, 1.35] sec. Notch on J port at 1.2 sec
result: H-modes, Multiple L-H transitions.
Woh = 53 kJ, Wmax = 97 kJ, P_rad main = 1.6 MW,
@Wmax, Prf = 2.1 MW, dW/Prf = 21 kJ/MW, H89 = 1
@notch, efficiency ~ .7 (large uncertainty)
next: 3 MW RF.

Compared with pre-boronization
shot 1040211009, 2.2 MW, steady H-mode, Wmax = 110 kJ.

Why not getting steady H-modes?

Next: 3 MW
Mar 29 2005 02:18:44:443PM1050329014Yijun Linshot 14: (Performance at RF power scan, 3 MW)
Plasma: Bt 5.4 T, 1 MA, nl04 = 0.9e20m^-2.
RF power: 0.75 X 4 = 3 MW, [0.6, 1.35] sec. Notch on J port at 1.2 sec
result: H-modes, Multiple L-H transitions.
Woh = 46 kJ, Wmax = 104 kJ, P_rad main = 2 MW,
@Wmax, Prf = 3.1 MW, dW/Prf = 16 kJ/MW, H89 = 1
@notch, efficiency ~.7

next: 4 MW RF.
Mar 29 2005 02:33:20:740PM1050329015Yijun Linshot 15: (Performance at RF power scan, 4 MW)
Plasma: Bt 5.4 T, 1 MA, nl04 = 0.9e20m^-2.
RF power: 1.0 X 4 = 4 MW, [0.6, 1.35] sec. Notch on J port at 1.2 sec
result: H-modes, Multiple L-H transitions.
Woh = 46 kJ, Wmax = 120 kJ, P_rad main = 2.5 MW,
@Wmax, Prf = 4 MW, H89 = 1
@notch, efficiency .7
H/D went up to > 0.1 later in the shot
Next: reduce RF power to 3.5 MW. repeat.
Mar 29 2005 02:46:00:130PM1050329016Yijun Linshot 16: (Performance at RF power scan, 3.5 MW)
Plasma: Bt 5.4 T, 1 MA, nl04 = 0.9e20m^-2.
RF power: 3.5 MW (D 1 MW, E, 0.8 MW, J 1.8 MW),
[0.6, 1.35] sec. Notch on J port at 1.2 sec
result: H-modes, Multiple L-H transitions.
Woh = 50 kJ, Wmax = 120 kJ, P_rad main = 2.3 MW,
@Wmax, Prf = 3.6 MW, H89 = 1
@notch, efficiency .8

Next: L-H threshold study. Ramping RF power. 0.8 MA, 0.8e20 nl.


Mar 29 2005 03:07:16:397PM1050329017Yijun Linfizzle
Mar 29 2005 03:11:11:803PM1050329018Yijun Lindud
Mar 29 2005 03:55:06:223PM1050329019Yijun Linno power shot cycle
Mar 29 2005 03:55:18:443PM1050329020Yijun Lindud
Mar 29 2005 04:09:09:147PM1050329021Yijun Lindud

Mar 29 2005 04:19:52:520PM1050329022Yijun Lindud
Mar 29 2005 04:40:00:130PM1050329023Yijun Lindud
Mar 29 2005 04:57:19:927PM1050329024Yijun Lindud. 5 in a row.
Mar 29 2005 05:08:34:427PM1050329025Yijun Lindud

Physics Operator Comments
Mar 29 2005 08:44:38:510AM1050329001Jim IrbyLoaded From 1040212030

Imported seg 1 from 1050318035

We will attempt running all twelve klystrons into plasma and vacuum today. Will put in a short
pulse at 0.5 s

Loaded -70000.00 into Ip predictor.

This shot was a cycle test
Mar 29 2005 08:58:59:340AM1050329002Jim IrbySerious CAMAC problems continue: gka5** highway or server down. Josh onsite...

Mar 29 2005 10:06:29:307AM1050329003Jim IrbyRemoved late points on A-coil control line. Points as late as 10 seconds were included causing lots
of extra data to be stored and calculations done by DPCS.

GKA502 (reflectometer) was the problem with the highway.

Gatesvalves are set to toggle and remained closed for 300 s

Enabled VUV gatevalve

Another cycle test --- seems to have worked --- got some gas

data is rolling in.
Mar 29 2005 10:31:48:823AM1050329004Jim IrbyPrepuff level very high on shot 3 (.08 mTorr vs .05 last week).
Prepuff from 28 to 22 ms.


Mar 29 2005 10:48:17:917AM1050329005Jim Irbyprepuff to 20 ms
TF from 150 to 144.5 kA

plasma into rampdown

TF Rogowski in trouble
TF shunt okay

Mar 29 2005 11:50:05:833AM1050329006Jim IrbyDropped argon puff from 30 to 0 ms. Ian sees doubling of radiation from puff

zeroed IP and clearin offsets in prep for DCPS tests


stopped gatevalve toggles --- problem with VUV gatevalve

plasma for 0.524 s

TF Rowgowski if back in operation

Asked LH to reduce power. They had about 240 kW at 0.5 s going in with high reflected power on
most channels
Mar 29 2005 12:10:55:740PM1050329007Jim IrbyYuri and Dave have made the reflectometer rack operational again (gka502)

requested LH power dropped by factor of two or so

trying again --- Steve could find nothing wrong with DPCS --- if this one disrupts we will
go back to the original offsets in Ip/clearin etc.

plasma into rampdown

looks like the refectometry crate is back online. J221 triggers are being generated. Still have
a problem with the firewire cameras.

DCPS looks good.

Plasma plume on LH grill seen on wide2

Mar 29 2005 12:35:51:613PM1050329008Jim IrbyBr0 increased by 1 mT from -0.002 to -0.001

RCUR from 2000 to 2500

EC_EF4U from -1335 to -1325 A

LH delayed to 1.4 s

dud!

EF4 and Br0 need a tweak. Pressure looks okay
Mar 29 2005 12:47:08:960PM1050329009Jim IrbyEF4 from -1325 to -1345 A
Br0 from -0.001 to -.0005 T

dud ---
Mar 29 2005 01:10:48:973PM1050329010Jim Irbyload seg 1 from 1050329007 --- get back a plasma

plasma into rampdown

no h/(h+d)
Mar 29 2005 01:28:13:927PM1050329011Jim IrbyBt back to 5.4 T (150 kA)

plasma into rampdown

h/(h+d) at 6.5% at 1s

Mar 29 2005 01:41:41:317PM1050329012Jim IrbyNo DPCS changes

fizzle --- prepuff a little higher than on shot 11. Fields look very good


Mar 29 2005 02:41:08:553PM1050329015Jim IrbyPlasma into rampdown

h/(h+d) is 10% at 1s and rising

Mar 29 2005 02:44:42:723PM1050329016Jim Irbyplasma into rampdown

h/(h+d) high
Mar 29 2005 02:59:48:490PM1050329017Jim IrbyIp to 800 from 1000 kA
nl04 request from 1e20 to 0.8e20 /m^2

fizzle --- why?? fields and gas look okay



Mar 29 2005 03:11:42:317PM1050329018Jim Irbypg3 early rise from 85 to 95 %

dud --- prepuff low compared to shot 16
Mar 29 2005 03:56:57:460PM1050329019Jim Irbyno power cycle test following action monitor problem
Mar 29 2005 04:03:23:757PM1050329020Jim Irbyef4 from -1335 to -1340 A
ip request 800 kA
nl04 request 0.8e20/m^2

dud
Mar 29 2005 04:12:34:397PM1050329021Jim Irbyreduced gas by 2 ms

dud --- field okay, gas reproduced earlier plasma shots, but...
Mar 29 2005 04:35:07:130PM1050329022Jim Irbyincreased prepuff by 4 ms (-.522 to -.526)

data sysem problem --- data lost --- probable dud
Mar 29 2005 04:51:05:443PM1050329023Jim Irbyincreased prepuff by another 2 ms (-.528)

dud
Mar 29 2005 05:04:27:007PM1050329024Jim Irbyreload 1050329016

gv1 closed

dud!!!
Mar 29 2005 05:11:10:863PM1050329025Jim Irbyincreased prepuff to 35 ms total --- try for at least a fizzle

gv1 open again

dud

Engineering Operator Comments
ShotTimeTypeStatusComment
108:16:02:747AMTestBadShot cycle test
208:49:33:243AMTestBadShot cycle test
309:58:14:760AMTestOkShot Cycle Test
410:21:57:357AMTestOkShot Cycle test
510:34:41:620AMPlasmaOkPlasma,Except no TF Rogowski
611:31:13:870AMPlasmaOk
712:00:48:600PMPlasmaOk
812:25:46:723PMPlasmaOk
912:38:30:583PMPlasmaOk
1012:53:47:020PMPlasmaOk
1101:12:17:287PMPlasmaOk
1201:27:22:257PMPlasmaOk
1301:50:52:677PMPlasmaOk
1402:04:11:537PMPlasmaOk
1502:16:51:553PMPlasmaOk
1602:35:46:270PMPlasmaOk
1702:52:44:100PMPlasmaOk
1803:05:17:990PMPlasmaOk
2003:51:46:943PMPlasmaOk
2104:04:46:677PMPlasmaOk
2204:17:20:880PMPlasmaOk
2304:36:16:193PMPlasmaOk
2404:54:15:020PMPlasmaOk
2505:06:45:850PMPlasmaOk