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
                  July 23, 2001

Plasma operations continued at Alcator C-Mod last week. Five run days were
scheduled and four and a half were completed. A total of 64 plasma shots were
produced with a startup reliability of 80%. Runtime was devoted to ITB and
Double Barrier studies, non-dimensional similarity experiments to be
coodinated with ASDEX-Upgrade, and turbulence imaging; half a run was also
devoted to MSE calibration shots requiring DNB pulses into gas with controlled
toroidal and vertical fields, but no plasma.

Plasma operations are scheduled to continue this week. We are also considering
extending the campaign into next week, which had been designated as a
contingency option.

Physics
-------

The first three run days last week were in support of MP#289A, "Double Barrier
Studies". This MP is a multi-run experiment to investigate the physics of the
internal barrier formation previously reported in EDA H-mode plasmas with
off-axis ICRF heating. While previous results were reproduced, and some
additional data on the evolution and possibility of control of the core
barrier by on-axis heating and field ramping were obtained, these experiments
were limited by the high hydrogen fraction and by control circuit problems
withe the 70MHz ICRF system. Additional experiments in support of this
MiniProposal are planned for this week.

Thursday's run was in support of MP287, "ASDEX-Upgrade Similarity
Experiments". The goal was to make high density, high power EDA H-modes in a
shape identical to that in ASDEX-Upgrade 'Type II ELM' discharges.  This
regime, which was only recently obtained on AUG, appears rather similar to
that obtained on C-Mod when power is increased into EDA H-modes.  At reduced
powers, AUG sometimes sees a 'quasi-coherent' mode.  We would like to see if
the same behavior is observed on the two machines with the same dimensionless
parameters.  We will obtain edge profiles both at threshold and in the steady
H-mode period.  AUG will then try to match the dimensionless parameters
(nu*,beta, rho*) of these discharges.  The shape development part of the run
went quite smoothly, and we did get shots which were a very close match to
ASDEX Upgrade shapes. Compared to our usual operation, these plasmas are quite
low triangularity and very close to being double null. As expected, it was
more difficult to get steady EDA H-modes in the lower triangularity
ASDEX-Upgrade shape, and the reduced absorbed power because of the high H/D
ratio was an additional impediment. More run time with more favorable machine
conditions would benefit this experiment.

The first half of the 7/20/01 run was for edge turbulence imaging measurements
using the Princeton Scientific Instruments Ultra-Fast Camera (PSI-3), which
can take 12 successive frames up to 1 million frames/sec.  Discharge
conditions were set to match a baseline case for which edge profiles are
available and theoretical simulations are being done by Hallatschek and
Rogers.  The camera view is essentially tangential to the field near the
outboard midplane at the plasma edge, and illumination is provided by a gas
puff, either of D2 or He, from a capillary located at a gap in an outboard
protection limiter assembly. Excellent images were obtained showing the
evolution and motion of turbulent structures ("blobs") in the poloidal
plane. Images obtained with D2 puffs and a D-alpha filter showed structures
near the separatrix and extending into the SOL, while images formed with a
helium puff and a He-II (4686A) filter showed emission further inside the
LCFS, with smaller features and apparent rotation in the opposite sense to
those at larger radius.

In reviewing magnetics fluctuation data it was noted that, under some
conditions with small outer gaps <= 1 cm and at relatively low density, the
poloidal magnetic field pick-up coils located underneath the tiles on
the outboard AB and GH limiters have sufficient signal from the Quasi-Coherent
Mode (QCM) in EDA H-mode to be able to make out the toroidal mode number.  At
least a dozen shots have been found with reasonable signal levels and the
resulting toroidal mode numbers of the QCM range from n = 15 - 19 for ICRF
heated EDA H-modes.  The coils beneath the tiles are typically only 2.5 cm
from the LCFS whereas the coils on the sides of the limiters and on the inner
wall are typically 5 cm from the LCFS.  Because of the high mode numbers of
the QCM, the rapid radial decay of the mode amplitude makes the signal too
small to be observed on the coils that are 5 cm from the LCFS.  These toroidal
mode numbers are somewhat smaller than the calculated n~30 from the radial
decay of the QCM from the A-side scanning magnetic probe measurements in Ohmic
EDA H-mode, but still indicate a rather high toroidal mode number.

Operations
----------

The machine ran well last week, but the H/D ratio during high-power RF
operation was higher than optimum, rising above 0.1 during typical H-modes,
and as high as 0.2 during some shots.  As a result, the ICRF minority heating
efficiency degraded significantly during the pulse, compromising some planned
experiments. Additional glow discharge cleaning (GDC) in helium was carried
out from Friday evening to Sunday afternoon (7/20-7/22), followed by a fresh
boronization on Sunday evening. It is hoped this procedure will reduce the H/D
to acceptable levels for this week's experiments.

Wednesday's run was delayed in order to remove a shunt from the OH2L buss. The
shunt had been observed to be giving high readings during Tuesday's run,
indicating that its resistance had increased by about a factor of two from its
nominal value. The shunt was replaced by a length of busswork, and is being
shipped back to the manufacturer for tests and analysis. The buss shunts are
not presently used for any control or feedback functions, and primarily
provide redundant information with the buss rogowski loops.

For several weeks we have noted signs of damage in snubber resistors in the
bypass circuit of the chopper supply used for fast vertical position control;
we have been replacing these resistors during maintenance, and the problem has
not resulted in any loss of run-time. Analysis by the resistor manufacturer
indicates that the resistors suffered from overheating as a result of too much
power dissipation, but our initial calculations indicate that they were being
operated within their specification. We are now testing a new set of
resistors, with higher power/energy dissipation rating and better temperature
capability, in the snubber circuit.  Comparison of waveforms from this
campaign with earlier run periods does indicate we have been drawing somewhat
higher average currents from this supply than in the past. It is not known
whether this is related to the resistor damage. Last week the programming for
this supply was changed to reduce the average current back to its historical
value, which has the benefit of increasing the dynamic range available for
feedback stabilization of the vertical instability.

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

The J-port antenna operation at 70 MHz made good progress towards reliable
operation.  By the end of the week, 1 MW was coupled to the plasma without
faulting.  The decoupling stub was optimized and the vacuum and plasma
operation became more stable.  A reflected power signal critical to the
accurate prediction of antenna match was found to be malfunctioning; this was
repaired, allowing efficient matching to plasma loads.  The antenna phase
determined from the voltage probes was found to be unreliable and resulted in
false phase balance faults.  From current probes at the same location, the
phase is as expected, 0 deg, and stable.  We concluded that these probes are
located near a voltage minimum at 70 MHz and the voltage phase measurement is
difficult due to the small signal level.  The current probes will be used for
the phase balance fault arc detectors for the remaining 70 MHz operation.


Diagnostic Neutral Beam Systems
-------------------------------

The beam operated at full voltage and current throughout the week.  We
continued with diagnostic development.  The arc power supply was modified to
allow the arc pulse to be increased from a maximum of approximately 300 ms to
a maximum that is at least 700 ms.  This is to facilitate conditioning of the
arc.  On Friday, we completed a beam into gas experiment with various toroidal
and vertical magnetic fields to calibrate the MSE diagnostic.

Lower Hybrid MIE Project
------------------------

The reflected power probes of the lower hybrid launcher have been tested
successfully at PPPL by Stefano Bernabei. At MIT, work continued on the TPS
chassis wiring and timing circuitry.

Inner Divertor Fabrication Project
----------------------------------

The vendor responsible for fabrication of the Rear Girdle Plates has informed
us that they will be unable to meet their previously agreed schedule, and
proposed a new delivery date two and a half months in the future. In view of
the delays already experienced at this vendor, we have decided to transfer the
remaining work on these components to the shop which previously produced and
delivered the C-plates; the latter vendor has agreed to complete the work on a
schedule compatible with our requirements.

In-house work on the probe box and mock-up parts is proceeding.  All short
studs have been welded on the mock-up wall. Diagnostics were mounted on the
wall as well.  A plastic Rear Girdle Plate, Inconel C-Plate, Tiles, Transition
keys were mounted on the Mock-up wall. Two limiter backing plates with tiles
were placed directly above the new Inner Divertor and found to fit perfectly,
confirming that there should be no installation difficulties provided the Rear
Girdls Plates are within specs.  Furthermore, there was very little work
required to align the Girdle Plates with the Limiter Backing plates and
interlock them with a transition key.

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

Stewart Zweben was at MIT for the week, continuing to obtain high-speed, high
sampling rate "movie" images of plasma edge turbulence with the gas puff
imaging diagnostic.

A quarterly review was held on Monday, July 16. Rostom Dagazian attended from
OFES. The viewgraphs from the presentations can be found at

www.psfc.mit.edu/cmod/sciprogram/Q_rev_Jul01/C-Mod_Quarterly_Rev_July_2001.pdf