Newsgroups: sci.physics.plasma
From WOLFE@CMOD2.PFC.MIT.EDU Wed Apr 5 09:03:06 1995
From: WOLFE@CMOD2.PFC.MIT.EDU
Organization: MIT
Subject: Alcator C-MOD Weekly Highlights

Alcator C-MOD Weekly Highlights
April 3, 1995

Plasma operations are continuing on Alcator C-MOD. Four runs were scheduled
and completed last week, including a full run carried out from the Remote
Experimental Site in Livermore. The principal experiments were concerned with
dissipative divertor physics, mitigation of VDE's, high beta_p operation with
current rampdown, and the effects of Lithium wall-conditioning on H-modes. A
total of 127 plasmas were produced over the four run days.

On Tuesday, following a brief morning session in which a new plasma
shape-control technique was tested, control of the run was transferred to a
team of MIT and LLNL scientists operating from the Remote Experimental Site in
Livermore, CA. A full physics run, addressing issues of divertor detachment in
ICRF-heated L- and H-mode plasmas, was then conducted from the RES. Twenty-one
shots were run remotely on Tuesday, and another seven shots during an
additional two-hour remote session on Wednesday to wrap up the physics
program. MIT participants at Livermore were Steve Horne, who served as
Physics Operator, Brian LaBombard, who acted as Session Leader and controlled
the Fast Scanning Probe diagnostic, Yuichi Takase, who operated the ICRF
heating, and Josh Stillerman, who dealt with computer and network issues.
Livermore personnel included Tom Casper, Bill Meyer, Jeff Moller,Joe
Borrachio, Tony Hain, John Perkins (ITER), Gary Porter, Ray Jong, and a
number of others who came by to observe during the day. The PCS Plasma Control
Software, SCOPE data displays and IDL data analysis and display tools were
used over the ESNET link successfully. Communications tools included video and
audio links (MBONE and NV) over the INTERNET; the video information included a
real-time plasma TV display. Several IRC (Internet Relay Chat) channels
provided additional communication links: between Physics Operator (at LLNL)
and Engineering Operator (at MIT); between the remote RF operator and local RF
engineering and Physics staff; an open channel between the remote crew and
participants in the MIT Control Room; and additional remote collaborators at
U. of Md. and Johns Hopkins University. The demonstration of remote operation
was quite successful, with many important issues addressed and lessons
learned. A full report is in preparation.

The physics results from the Tuesday and Wednesday remote operation were
significant in their own right. Density and RF power scans were carried out to
investigate divertor detachment and re-attachment phenomena. Divertor
detachment was retained under RF heating at the 0.7 MW power level with no
impurity puffing. At higher RF power (up to 1.3 MW) partial detachment, with
divertor temperatures in the 3-6 eV range, was observed during H-mode.

After control had been transferred back to MIT on Wednesday, an experiment to
investigate the possibility of mitigating the effects of VDE disruptions by
intentionally inducing a rapid thermal quench was carried out. This
experiment, which was inspired by an ITER request, employed lithium
pellets doped with up to 300 micrograms of gold to rapidly radiate the thermal
energy of a plasma in which a VDE had been artificially produced (by turning
off the fast vertical position feedback). The idea is to cause an abnormally
fast thermal and current quench, hopefully terminating the plasma before it
can move very far off the midplane. The Au doped pellets were successful in
radiating away the thermal energy while the plasma was still within a few cm
of the midplane, but the plasma continued to drift down and the current quench
and halo currents were not significantly affected. Additional experiments with
larger quantities of high-Z dopant are being planned.

Current rampdown experiments to achieve high-beta_p discharges (MP#089) were
carried out on Thursday. This proposal was inspired by TFTR results, which
indicated that confinement in the post-rampdown phase remained high for a
period of order the skin time, permitting high-beta_p, high-q operation. The
C-MOD experiments were proposed in collaboration with Jay Kesner (MIT Theory
Group) and Mike Mauel (Columbia University), both of whom were also involved
in the TFTR work. The current was ramped from an initial 800kA down to a
second plateau at 400kA in less than 100msec, holding the shape approximately
constant (standard lower single null). ICRF (up to 1.5MW) was injected both
before and after the rampdown. Beta_p was observed to peak shortly after the
completion of the ramp, and decay thereafter. The highest beta_p values were
obtained with the fastest rampdown time, with ICRF heating before, during and
after the ramp. Confinement immediately following the ramp was apparently
enhanced relative to steady-state L-mode values, but was not as high as
in the high-current phase. Maximum beta_p obtained was ~0.85. Detailed
analysis of these experiments is proceeding.

On Friday we conducted a test of the role of lithium-pellet wall conditioning
in determining H-mode thresholds and performance. These experiments were
motivated by previous indications on C-MOD that Li pellets might have had a
beneficial effect in achieving H-modes, and by experience on TFTR indicating
lithium wall conditioning had important effects. Multiple lithium pellets were
injected into a series of about 20 discharges over the first half of the run.
Li levels in the plasma were monitored spectroscopically. The second half of
the run was used to look for improvements in ohmic and RF H-mode behavior at
various densities and currents up to 1.0 MA. No such improvements, in either
threshold, performance, or ELM behavior, was observed, apparently indicating
that Li conditioning is of less importance in a metal wall machine than for
carbon walls.

C-MOD is about to begin a brief campaign with reversed toroidal field and
plasma current, such that the ion grad-B drift is away from the closed
divertor structure. Bus reconfiguration for these runs was completed on
Saturday, and power tests carried out on Monday, April 3.

We have taken delivery of 5 new DEC alpha computers. One of these is an
AlphaServer 1000 4/200 which couples a fast cpu to fast I/O busses. This machine
will become our primary data acquisition engine and support SCSI based serial
highway drivers as well as an array of fast disks. (The data will be "striped"
onto multiple platters to increase throughput.) Three new AlphaStation 200
4/233's will be used for compute intensive data analysis, running the EFIT MHD
equilibrium code; the density profile reconstruction code; and edge probe
analysis. (Rapid completion of these codes are essential for efficient machine
operation.) The final new Alpha, a somewhat slower model 200 4/166, will be
used for porting of the MDSplus code to unix (and possibly to NT), and for
importing and maintaining new versions of the TRANSP code.

Prof. Miklos Porkolab took part in the "Fusion Forum" in Washington DC on
Tuesday, March 28. Prof. Ian Hutchinson, Prof. Porkolab, and Dr. Earl Marmar
attended the Field Work Proposal Presentations at D.o.E. in Germantown, Md.,
on Wednesday, where Prof. Hutchinson presented the Alcator C-MOD plans for the
next two years.