Step 1. Hirex Sr absolute calibration (locked mode shot)
Step 2. Resurrect modes in TAE band gap from July, August in the summer of 2010
Step 3. Check for Alfven scaling of modes
Step 4. See whether unmodulated ICRF power makes mode amplitude increase
Step 5. See whether modulated ICRF at mode frequency can couple power to modes, increasing their amplitudes
Step 6. Increase density beyond observed stability threshold of ~0.6x10^{20} m^{-2} and see whether modes are suppressed
Step 7. Repeat Step 4, but now with otherwise stable modes instead of unstable modes.
Step 8. Repeat Step 5, but now with stable modes
Step 5 mod. => rather than sweeping modulation frequency, held constant at 600 kHz. Removes complication on modulation end, and is still appropriate for experiment because modes spin up in frequency strongly, themselves, and 600 kHz should pass right through the sweep across several harmonics. Ran into this late in the day.
Shot-by-Shot:
SHOT: Goal: Result:
20 Step 1 Disruption
21 Step 1 Success
22 Step 2 Modes are not clear - lower density (may apply to Step 6)
23 Step 2 Success
24 Step 3 Success in input - Needs analysis to check for Alfven scaling
25 Step 3 Disruption
26 Step 4 J antenna trippy
27 Step 4 Shutdown
28 Step 4 Got power ramp, got into H-mode. Needs analysis.
29 Step 5 No power into D antenna
30 Step 5 D trips - a few restarts visible in mod. freq. sweep on mag.
31 Troubleshoot Disruption
32 Troubleshoot Got power into D antenna - ~12.5 % depth mod.
33 Step 5 mod. Disruption
34 Step 5 mod. Disruption
Scorecard:
15 shots, 9 about full length, 5 disruptions, 1 shutdown
Many cell accesses to troubleshoot current signal from D antenna, power supply problems with CNPA, also some polarimeter fixes.
PCI was not available due to broken shutter. CNPA was down for a while, but came back in time for Shot 28, which was the shot it was really needed for (hooray).
To Do:
-Investigate whether modes exhibit Alfven scaling - Shot 1101123024
-Investigate whether frequency spacing of modes corresponds to rotation frequency, thanks to Hirex Sr calibration (will need to wait some time for data to be available, according to Matt Reinke) - Shot 1101123023,024
-Investigate whether unmodulated ICRF made modes appear more strongly - Shot 1101123028
-May want to repeat Step 2 tomorrow with I_P and B_T ramping down instead of up so that V_A scaling will run counter to natural spin-up of modes.
-Look on FRCECE to see whether modes can be localized radially.
-Some 900+kHz activity early in Shot 32 - might be worth looking into.
-Check out amplitude modulation in Shot 32.
-Make sure that we digitize current signal out of ICRF rack.
Tomorrow: after locked-mode shot, pick up where we left off today. If modulation doesn't work, switch to Part II.
MP616
There were no major hang-ups this morning, and so we were able to get through
the bulk of the run plan. We:
. . . loaded a near DN shot in which we could vary SSEP and determined the
power threshold in the base LSN case (Shots 2--8)
. . . applied ramps to ZCUR in order to slowly shift the magnetic balance
from LSN to DN/USN and observed a ~30% reduction in the power threshold.
(Shots 11--16)
. . . began with USN and ramped down toward DN/LSN, and started to map out
the other side of the P_L-H "well" (Shot 17--18)
There were some injections that may complicate interpretation, but the
preliminary results suggest I can back out a quantitative description
of the P_L-H suppression.
A graphic can be made of VERY PRELIMINARY results:
First panel below is P_loss vs. SSEP, where P_loss=P_oh+0.9*P_rf-dW/dt. Second
panel is P_loss/P_th vs. SSEP where P_th is calculated from the Y. Martin
scaling law for H-mode power threshold. Third panel is P_net vs. SSEP where
P_net=P_loss-P_rad,main. Here the diamonds represent points on the standard EFIT
timebase taken from the shots 1101123002 through 1101123018. [Color mapping to
shot # for diamonds is not particularly meaningful: BLACK--2,13; RED--3,14;
BLUE--4,15; VIOLET--5,16; ORANGE--6,17; MAGENTA--7,18; CYAN--8; YELLOW--11;
GREEN--12]These data are taken between 0.6s and either 1.5s or the L-H
transition time, whichever comes first. Many shots didn't have a transition at
all. The big black circles represent the APPROXIMATE conditions at the first
L-H transition in each shot where H-mode occurred.
To follow up, here is my more formal summary for the C-Mod weekly report:
The morning session on Tuesday was used to explore the apparent reduction in
H-mode power threshold in nearly double null (DN) plasma discharges. Such a
reduction is prominent (approximately a factor of two) in NSTX and MAST, but was
estimated to be more modest (~30%) with a conventional machine aspect ratio
(ASDEX Upgrade). To examine this phenomenon on C-Mod, we operated in near DN
configurations and programmed slow ramps to the plasma vertical position in
order to shift magnetic balance from lower single null (LSN) to upper single
null (USN) and vice versa. Input power was kept fairly steady throughout the
balance scans, at levels found to be below the threshold power in LSN plasmas.
Preliminary results indicate that the reduction in power threshold at DN is
about 30%, as suggested by ASDEX Upgrade results. Indications are that this
reduction occurs over an extremely narrow window (<1mm) in SSEP, the midplane
distance between the primary and secondary separatrices as reconstructed from
magnetics. L-H transition data are being analyzed, particularly with regard to
edge conditions. This experiment supports ITPA joint experiments PEP-6,
"Pedestal structure and ELM stability in DN" and PEP-26, "Critical edge
parameters for achieving L-H transition".