From: WOLFE@PSFC.MIT.EDU
Reply-To: WOLFE@PSFC.MIT.EDU
Subject: Alcator C-Mod Weekly Highlights
Newsgroups: sci.physics.plasma
Organization: MIT

Alcator C-Mod Weekly Highlights
June 12, 2000

Operations
----------

Plasma operations continued at Alcator C-Mod last week. Four run days were
scheduled and completed. A total of 67 plasma shots were produced with a
reliability of 87%. Principal experiments included studies of ICRF-induced
rotation, evolution of peaked density profile H-modes, and an investigation of
density limits.

Plasma operations will continue this week.

Physics and Analysis
--------------------

The physics of density limits was the subject of experiments on Tuesday and
Friday of last week. Earlier work suggested that the tokamak density limit can
be explained fully only by consideration of transport effects. While there is
little dispute about the final stages of a density limit collapse - edge
cooling via radiation then shrinkage of the current channel leading to MHD
instability and disruption - theories based on edge power balance physics
alone can not explain important aspects of density limit phenomenology. In
particular, the power balance theories predict a density limit with strong
dependence on input power and on impurity content. Neither of these
dependences is generally observed. These classes of theories also have
difficulty explaining the robustness of the density limit with respect to
plasma configuration, divertor topology, or operational regime. The work by
Drake and Rogers suggests a specific turbulence mechanism for this phenomenon.
The simulations show a boundary, above which fluctuations increase
dramatically from nominal ``L'' mode levels. This boundary has somewhat
complex parametric dependence but has much in common with the empirical limit.

Experiments to test these ideas were carried out by reducing the plasma
current while increasing the density, until a density-limit disruption
occurred. The edge turbulence was monitored using the A-side scanning probe,
which was inserted at three times during each shot, including one plunge just
prior to the disruption, when the plasma was essentially at the empirical
density limit. There was a substantial increase in the potential and density
fluctuations for the final scan. The fluctuation character also seemed to
change: the ISat fluctuations were dominated by lower frequency features.
These observations generally support the notion of the density limit as a
transport collapse (as compared to a radiative collapse). Unfortunately the
probe was damaged during an earlier run, so we were unable to determine the
ExB convected flux; a replacement probe is being installed this week. We do
have good Lyman alpha data. Detailed analysis of this data is in
progress. Future experiments will extend the observations to limited
plasmas, vary the probe scans to better determine the temporal evolution of
the fluctuations (and transport), and run at different toroidal fields to vary
beta for comparison with the Rogers and Drake theory.


In discharges with off-axis ICRF heating, where the hydrogen minority resonance
is located approximately 10 cm to the high field side of the magnetic axis, the
formation of core particle transport barriers develops while the plasma remains
in H-mode. Coincident with the internal transport barrier formation, the
toroidal rotation of the plasma, which is usually in the co-current direction,
slows, and eventually reverses direction. As the density continues to peak,
sawteeth become smaller, and eventually stop. The central electron density can
reach values as high as 6e20 /m^3; central Z_eff also rises during this time.
Typical density profiles, inferred from Thomson scattering and visible
continuum imaging can be seen in
www.psfc.mit.edu/people/marmar/den_profs_1000523016.jpg
The figure shows two profiles: the flat profile typical of H-mode, taken early
in the H-mode phase of the discharge; the peaked profile characteristic of the
double barrier phase of the discharge.

Last week we continued our investigations of these phenomena. We were able to
reproduce plasmas routinely with impressive core density barriers at 4.3-4.5T
and with 2.8 MW of off-axis ICRF heating. The core rotation was again
observed to reverse sign. By scanning the toroidal field in small steps, we
were able to establish a critical field of 4.5 T for the onset of the density
barrier mode and the reversal of the toroidal rotation. It is worth noting
that even at 4.6T with the ICRF resonance located 8.4 cm to the high field
side (HFS), the rotation remained co-current for the entire pulse. This result
would seem to contradict the theory of Perkins et al., where the rotation is
predicted to be counter-current with an HFS ICRF resonance.

The phase contrast imaging (PCI) diagnostic has measured electron density
fluctuations associated with the RF wave propagation for the first time. The
intensity of the laser beam (which passes through the plasma vertically at E
port) is modulated at a frequency within a few hundred kHz of the RF source
(80.5 MHz for D port, the only source seen so far). This is done by splitting
the beam, upshifting one half and downshifting the other in frequency, and
recombining before passing through the plasma. With this technique, strong
signal (about 5 times higher than background noise) has been seen for several
low field (4.5 T), high density (5 - 6 10^20 m^-3 ne0), peaked density profile
shots. The RF PCI signal was at 300.85 kHz, and there was no significant
signal near that frequency observed in the magnetic fluctuations data, ruling
out other (non-RF) 300 kHz plasma fluctuations. In addition, the RF PCI signal
was present on one shot, but absent (signal dropped by a factor of 10) from
the next, identical, shot for which half the PCI beam was blocked, indicating
that the signal was not an artifact of RF pickup in the circuitry. The
measured radial wavenumber, in the range from 1 to 2 cm^-1 is in good
agreement with that expected from the RF fast magnetosonic wave dispersion
relation.


ICRF Systems
-------------

The D and E port antennas were operated this week in support of physics
experiments on density limits, and ICRF-induced rotation and core barrier
formation, with power levels of up to 2.8MW. For the rotation and ITB
experiments, at 4.1-5.4 T, both antennas ran reliably; however, D-port had
difficulty with 5.5-6 T. Although the tune appears good, we consistently had
fast high reflection events that trip the system. One obvious difference
between discharges below and above 5.4 T is the H resonance location. Below
5.4 T, the H resonance is on the high field side of the tokamak and above it
is on the low field side. However, we have in the past run easily up to 6.7 T
with H minority, so this is somewhat of a mystery. D and E port antenna
operation at high field will be the subject of its own mini-proposal to
investigate this problem.

RF support of the density limit experiments was limited by excessive impurity
injections from E-port antenna. An upper section of the Faraday screen is
becoming hot. This appears only when E-port is energized but does not appear
to be an arc in the antenna box. Further tests will be conducted in the
coming week.

A telephone workshop was held on 6/8 at the request of DoE to elicit input
from the U.S. RF community to the low observed heating efficiency of the PPPL
J-Port ICRF antenna. R. Dagazian and T. V. George from DoE as well as RF
physicists from GA, ORNL, MIT and PPPL participated. The basic antenna
design, observed behavior, present thoughts on the missing power, further
experiments and our ideas for improvements during the planned July vacuum
opening were presented by the MIT-PPPL C-Mod ICRF group. There is agreement
among the experts that our plans are basically sound, and useful suggestions
for additional measurements were received.


Lower Hybrid System
--------------------

During initial testing at a pulse width of 80 microseconds, the klystron
S/N:111 produced 250 kW (full power) at a beam voltage of -45 kV (full beam
voltage) with no faults. Subsequently, the same klystron was tested at full
power to a pulse width of 10 milliseconds. As a side note, edits to the
Allen-Bradley PLC ladder-logic software and user-interface screens were
completed. The klystron testing was accomplished totally under PLC control.


DNB System
----------

Parameter scans for the DNB were completed in deuterium. Predictions for the
penetration of the beam indicate that D2 penetration is less than H2, as
expected. Since we plan to run the beam in H2 to improve the penetration, it
was decided to request time to determine the modification of the H-to-D ratio
which could adversely affect the RF deposition. These experiments will be run
during the current week. A need for additional capacitance between the
gradient grid and the accelerator grid was identified to prevent arcing at
higher extraction voltages.



Engineering
------------

A status review meeting on the new Inner Divertor/Inner Vessel Wall design was
held on 6/8. The inner divertor "nose" will be flattened in order to achieve
higher plasma triangularity. The inner divertor tiles will be mounted on stiff
sandwiched plates with total thickness of 1.5". These plates form a
stiffening ring ("girdle") for the inner vessel wall behind the inner
divertor, which will reduce the disruption-induced stress in the inner vessel
wall, permitting operation up to 2.5MA. Vacuum preparation and installation is
planned for the next major C-Mod opening. Several issues remain to be
resolved, including thermal stresses, halo current paths, and tile mounting
details.

Travel and Visitors
-------------------

Rejean Boivin presented a talk to the W7-AS and ASDEX groups on "High
Resolution Bolometry on Alcator C-Mod - Applications to H-mode pedestals and
neutrals."

Bob Childs attended the Board of Directors Meeting for the AVS held at
Jefferson Labs, Newport News, VA last weekend and then stayed for a three day
AVS sponsored workshop on Exttreme High Vacuum and Surface Conditioning.

Gary Taylor (PPPL) spent three days at C-Mod, continuing to make improvements
to the electron temperature diagnostic GPC2. He also discussed recent TORIC
improvements with Paul Bonoli.

Kevin Fournier from LLNL visited last week, for analysis of Kr and Mo
x-ray spectra.

Prof. Rex Gandy of Univ. Idaho visited to discuss progress and plans for the
heterodyne radiometer, which he is working on in collaboration with
Univ. Texas and Auburn Univ.