Date: Mon, 26 Jan 1998 18:24:38 -0500
From: WOLFE@CMOD.PFC.MIT.EDU
Reply-To: WOLFE@PSFC.MIT.EDU
Message-ID: <980126182438.42e00113@CMOD.PFC.MIT.EDU>
Subject: Alcator C-Mod Weekly Highlights
Newsgroups: sci.physics.plasma
Organization: MIT Plasma Fusion Center


Alcator C-Mod Weekly Highlights
Jan 26, 1998

Plasma operations continued at Alcator C-Mod last week. Four run days were
scheduled and completed, with nearly a hundred plasmas successfully
produced. Overall machine performance was good. Startup reliability, always
low after a boronization, rose from less than 50% on Tuesday to over 95% on
Friday. This week's runs included some of the highest performance plasmas of
the campaign, with stored energy (W_mhd) over 200kJ. Thursday and Friday had
the 2nd and 4th highest daily neutron production of any C-Mod run, reflecting
a large number of long, quasi-steady, high quality H-mode plasmas.

Over the previous Friday night and Saturday morning a fresh boronization was
carried out. 137 psi of diborane was used, for an average coverage of about
2000 Angstroms. This boronization processed the most diborane ever used during
a single C-Mod boronization. In addition, we scanned the ECDC resonance out
past the Faraday shield midplane location to just behind the current strap.
Previous boronizations during this run campaign have been limited to just up
to the Faraday shield.The first part of the run on Tuesday was devoted to
standard (fiducial) discharges to evaluate the impact of the new boronization,
and to complete the wall conditioning by running tokamak discharges.

As expected from previous experience, startup reliability was poor
during this phase. The walls were clearly pumping quite strongly, as evidenced
by a lower than normal pressure rise from the initial puff, and the current
rise was slower than normal. Once a series of successful startups was
obtained, the fiducial shots evolved quickly, indicating improving wall
conditions. EDA H-modes were readily obtained at 1MW of RF power. Moly levels
at input powers up to 2MW of RF were significantly reduced from the previous
week. After eight or nine successful discharges, we obtained early L-H
transitions and the stored energy and H-factors were similar to the best
obtained in the previous weeks.

The second part of Tuesday's run was an investigation (MP#173B) of the relative
efficiency of inboard and outboard gas fueling into limiter plasmas. This
experiment was inspired by ASDEX results indicating that pellet fueling from
the inboard side was more effective than with outside launched pellets. It was
reasoned that if a deep source was not essential, similar physics might obtain
for a gas puff directly into a limited plasma at the plasma-limiter contact
point. An experiment was carried out using inside and outside limited plasmas,
with deuterium fueling by gas capillaries in the respective limiter
structures. We can conclude that gas fueling into the high-field side does NOT
show any enhanced fueling efficiency compared to the low-field side. The
actual density rises were small, and the absolute fueling efficiencies may be
affected by marfe activity, but the comparative null result is clear.

Wednesday's run investigated the effect of the prototype cryopump on impurity
transport (MP #177). Ohmic plasmas with density ramps to divertor detachment
and EDA H-mode plasmas were investigated. Argon or neon gas was puffed into
these types of plasmas with the cryopump on or off. When on, the single
module was determined to have deuterium pumping speeds in the range of
700-1000 l/sec. The core impurity ion densities were monitored with x-ray and
VUV spectrometers and the divertor impurity neutral densities with an RGA.

In Ohmic plasmas, the core argon and the edge neon were "pumping out", i.e.,
the time history of the signal was decaying away. Comparing the core argon
density with and without the cryopump operational indicates that the pump had
no effect on the decay rate. This indicates that something else, most likely
the wall, is doing the pumping. Neon line brightnesses from charge states in
the edge plasma (temperature < 150eV) do show a decrease in the equilibrium
value with the cryopump operational. However, the initial "pumpout" is nearly
identical with and without the cryopump, also indicating a strong wall sink.
The density in these Ohmic plasmas was ramped to divertor detachment. After
divertor detachment, the cryopump is observed to make a difference in core
argon density. The penetration factor for puffed argon entering the core is
lower after detachment with the cryopump on than it is after detachment with
the cryopump off. Measurements of the neutral particle density in the
divertor indicate that the density is lower by approximately 20% with the pump
operational.

In EDA H-modes, the "pump out" effect is still evident on the core argon
density and there is no difference between plasmas with and without the
cryopump operational. Core neon brightness measurements show no "pump out"
behavior, although the SNR is small. There is also no discernible difference
in the core neon behavior with and without the cryopump. However, there is a
slight decrease, <20%, in both the divertor neutral neon and argon with the
cryopump operational. These results are in contrast to those obtained with
the non-recycling impurity gas nitrogen on 970617. During that run, the edge
nitrogen brightness was reduced by a factor of two with the cryopump
operational. Further analysis of these data continues.

MP 192 Dynamic q scans in Enhanced D alpha H-mode was performed Thursday, with
long (~1 sec) EDA H-modes during plasma current and toroidal field ramps to
vary the edge safety factor. Confinement clearly increased with increasing
plasma current and the best H-modes had H factors up to nearly 1.8. Stored
energies up to 200 kJ were obtained with less than 3 MW of ICRF. It is more
difficult to obtain EDA at higher current, but by starting in EDA at 0.8 MA,
the EDA can be maintained up to at least 1.2 MA. Indeed, the last shot
remains in EDA H-mode up to nearly 1.4 MA. The variation in q had little
effect on the enhanced D alpha nature of the H-mode and preliminary analysis
shows little increased MHD activity at rational edge q values, though this
still needs to be analyzed in detail. This suggests that if MHD activity is
responsible for the enhancement in D alpha emission, it must be broadband
multiple mode numbers not linked to particular rational q values.

During these ramped field and current discharges, the edge xray pedestal width
was found to scale faster than linear with 1/Ip, and had no dependence on
Btor. These data were obtained during steady state enhanced D-alpha H-modes
by ramping plasma current and/or toroidal field up and down. Interpretation
of these results is complicated by the fact that the observed quantity is the
soft
xray emissivity profile, not T, n, or p directly.

Friday's run was the latest devoted to the multi-day experiment MP#166, "High
q_parallel Dissipative Divertor". his proposal is an attempt to achieve
simultaneously a good confinement H-mode with low Zeff and a detached
divertor. The goal is to determine the threshold amount of impurity gas
necessary to detach the divertor. Also, operation just above and just below
the detachment threshold will be investigated.

The run was successful in producing enhanced Dalpha (EDA) H-modes with
'steady-state' H-factors of about 1.6. Up to 3MW of power was coupled to the
plasma. SOL power was up to 2.5MW, corresponding to an estimated q_parallel >
0.5GW per square meter. Divertor detachment was achieved with a combination
of deuterium, helium, and nitrogen gas puffing. The first half of the run
consisted of attempts to achieve a good EDA H-mode and to increase the
divertor radiation. The deuterium gas puff during the H-mode was determined
to be important in keeping the molybdenum levels low and the neutron rate
high, and seemed to eliminate the H->L transitions. The puffing of helium and
nitrogen in the divertor led to "death-ray" formation, high divertor radiated
power, and eventually divertor detachment. It was later determined that the
helium puffing was not necessary for detachment as nitrogen puffing alone was
sufficient, although the amount of nitrogen needed was increased.

Impurity gas feedback was also accomplished during this run. Feedback was
done on an edge bolometer channel in the core array. This channel was
identified earlier as a likely candidate to indicate when detachment was
ocurring. The waveforms drawn were trapezoidal in shape (in detachment the
waveform is triangular). Essentially this means we were trying to clip the
bolometer signal off by using feedback on the impurity gas, in this case,
nitrogen. The day started with hand drawn waveforms for the gas voltage.
When detachment was observed, the feedback was switched on with a drawn
waveform close to that of the previous shot. The feedback worked well, trying
to keep the bolometer waveform close to that requested. With the helium puff
turned off, the feedback requested extra nitrogen to match the drawn waveform
and reached detachment.

The goal of simultaneously achieving good confinement, low Zeff, and a
detached divertor was not quite reached. On the best shot, H-factors of 1.6
decreased to 1.4 after the divertor detached, but did not decrease any further
as detachment progressed. Zeff was high all day, never much less than 1.7
after the EDA H-mode was fully established. With impurity puffing, the
typical change in Zeff was about 0.5. There is now a starting point for
H-mode detachment studies. Comparison of plasma and divertor performance
above and below the detachment threshold can now proceed (and other
experiments requiring a detached H-mode).

Also during Friday's run, reconstructions (Boswell-Terry) of the fish-eye view
in D-gamma light show the regions where recombination is occuring. There is a
region above the outer strike point that moves up the plate as the detachment
proceeds. When the impurity gas flow is stopped, the region jumps back to
just above the strike point. More of the delta(stored energy)
vs. delta(toroidal rotation) space was filled in by Rice.

Progress continued on development of the DNB for C-Mod. Assembly tests of
the high capacitance section of the HV transmission line allowed completion
of the mechanical design. Design of controls for the Accelerator supply
continued. In-situ conditioning of the oil for the supply achieved a
breakdown voltage of 40 kV and thus has shown continued improvement. The
thermocouple system to be used for measurement of the beam footprint was
tested in its final configuration. This included test of thermocouples,
amplifiers and CAMAC data acquisition as well as setting up of the
acquisiton modules in the DNB tree, and writing of test programs for data
analysis.

Gary Taylor and Bob Cutler of PPPL visited last week. Good signals are now
being obtained on the new 19-channel polychromator. Some further optimization
for the fields used on C-Mod is planned.

The LANL IR camera was pulled off of C-Mod last week, and brought back to
LANL for some repair and improvements. A broken calibration wire deep within
the camera, which had been preventing us from doing remote-controlled internal
hot/cold calibrations, was fixed. In addition, the camera was sent out to a
prospective vendor for testing of a 1.2 Gigabit/second 12-bit digital
interface over a fiber optic link, into a PCI card for a PC. The digital
fiber link would allow direct 12-bit resolution capture into the PC, which
would improve the system's dynamic range from the present standard 8-bit
video, to 12-bits digital. We expect to have the unit back at C-Mod next week.

Miklos Porkolab and Steve Witch attended 2nd European RF Conference in
Brussels. Porkolab presented an invited talk, "Recent Progress in ICRF
Physics", and Wukitch presented a poster on "D-He3 heating in Alcator C-Mod"

Martin Greenwald spent 3 days in Washington DC attending a meeting of the
ESnet Steering Committee, where he represents the interests of the FES
program. Highlights include finalization of the ESnet Program Plan and a
white paper on advanced network applications; proposed changes in the MICs
network research program; NGI initiatives; progress in network performance
monitoring; and discussion of an ESnet supported "telepresence" toolkit.

Earl Marmar attended FESAC meeting in Germantown (Jan. 22).

This week is a scheduled maintenance week at Alcator C-Mod. No runs are
scheduled, although power system tests will be carried out. A review of the
C-Mod five-year plan, and a meeting of the C-Mod Physics Advisory Committee,
will take place at MIT this week.