anl.gov/Publications/rep-beam.pdf. It's a snapshot of the page taken as our search engine crawled the Web.
The web site itself may have changed. You can check the current page or check for previous versions at the Internet Archive. Yahoo! is not affiliated with the authors of this page or responsible for its content.
Progress toward a microsecond duration, repetitively pulsed, intense-ion beam for active spectroscopic measurements on ITER
Progress toward a microsecond duration, repetitively pulsed, intense-ion
beam for active spectroscopic measurements on ITER
H. A. Davis,
a)
J. C. Olson,
b)
W. A. Reass, Cris W. Barnes, and R. R. Bartsch
Mail Stop E526, Los Alamos National Laboratory, Los Alamos, New Mexico 87545
D. M. Coates, J. W. Hunt, and H. M. Schleinitz
DuPont Central Research and Development, Wilmington, Delaware 19880
R. H. Lovberg
c)
University of California, San Diego, California 92093
J. B. Greenly
Laboratory of Plasma Studies, Cornell University, Ithaca, New York 14853
Presented on 13 May 1996
We describe the design of an intense, pulsed, repetitive, neutral beam based on magnetically
insulated diode technology for injection into ITER for spectroscopic measurements of thermalizing
alpha particle and thermal helium density proles, ion temperature, plasma rotation, and low Z
impurity concentrations throughout the connement region. The beam is being developed to
enhance low signal-to-noise ratios expected with conventional steady-state ion beams because of
severe beam attenuation and intense bremsstrahlung emission. A 5 GW e.g., 100 keV, 50 kA 1
s
duration beam would increase the charge exchange recombination signal by 10
3
compared to a
conventional 5 MW beam. � 1997 American Institute of Physics. S0034-6748 97 52301-X
I. INTRODUCTION
Charge-exchange recombination spectroscopy
1,2
CXRS
is now the primary diagnostic of ion temperature, rotational
velocity, and helium ash concentration
3
in tokamaks. Using
visible spectroscopic views across a neutral beam which pro-
vides the charge exchange source for fully stripped ions in
the plasma, CXRS provides good spatial localization of the
Doppler shifted and broadened light. Present tokamaks are
able to use the same neutral beams used for heating the bulk
plasma for the CXRS source. The 100120 keV energy of
these positive-ion-source neutral beams is well matched to
the peak of the charge exchange cross section. Future large,
high-density tokamaks such as ITER are faced with a serious
problem. Positive-ion-source neutral beams will not pen-
etrate the plasma and are not suitable for core heating. High-
energy negative-ion-source beams planned for heating inter-
act at too high an energy with the plasma ions and too low of
a cross section to produce useful CXRS signals.
4
Steady-
state beams provide a CXRS signal which increases linearly
with plasma density, while the bremsstrahlung background
increases quadratically. In high-density operation of ITER, it
is expected that only in the outer half of the plasma can
CXRS signals be seen by modulating the beam and phase
averaging over a few seconds.
5,6
Thus, ion temperature and
helium ash density in the central core of the plasma would be
unmeasurable.
It has been proposed that intense ion diode technology
could be adapted to make a diagnostic neutral beam source.
7
Instead of steady or slowly modulated beams with less than
1 A/cm
2
beam density, microsecond pulses of a few kA/cm
2
at the beam focus in the reactor would be used. Operated at
100 keV/AMU near the peak of the CXRS cross sections,
these beams would increase signal-to-noise ratios because of
their high intensity coupled with very short gating times on
the detectors to reduce the bremsstrahlung background. Not
only would these beams provide vital measurements in the
plasma core, but would do so with time resolution governed
by the repetition rate of the beam. Such a neutral beam was
developed,
8
and it was demonstrated that effective charge
neutralization could be achieved in a gas cell at current den-
sities of 20 A/cm
2
only a factor of 37 below the required
value depending on the diode diameter , answering the pri-
mary scientic question of the approach.
Many of the remaining key issues in the application of
intense ion-diode technology to diagnostic neutral beams
have largely been solved, albeit not simultaneously: pulse
lengths of longer than 1
s have been achieved with careful
attention to magnetic geometry; more than 90% of the ion
diode beam can be at full energy; and active plasma anodes
see discussion below about plasma anodes have achieved
more than 90% hydrogen for cleanliness.
9
Beam divergences
less than 0.8� FWHM have been demonstrated.
10
This is
within a factor of two of the divergence needed to maintain
the desired current densities in ITER over the required 15 m
source-plasma distance required if the beam source is to be
located outside the biological radiation shielding wall with
adequate room for valves and the neutralizing cell. The light-
ion inertial connement fusion program regularly proposes
such divergences to make intense ion beam implosion
schemes plausible, but some development may be required.
The remaining technical issue, demonstrating repetitive
beam operation, is discussed in the remainder of the article.
Over the past two decades researchers in United States,
Germany, Russia, and Japan, have been investigating the ap-
plication of intense-pulsed-ion beam technology E
130
a
Electronic mail: davis@lanl.gov
b
Current address: Varian Ion Implant Systems, 508 Dory Rd., Gloucester,
MA 01930-2297.
c
Emeritus.
332
Rev. Sci. Instrum. 68 (1), January 1997
0034-6748/97/68(1)/332/4/$10.00
� 1997 American Institute of Physics MeV, I
0.11 MA,
1050 ns
11,12
to inertial conne-
ment fusion defense and energy programs. More recently
over the past decade , using more modest beam parameters
E
0.11.0 MeV, I
0.0050.05 MA,
1001000 ns , re-
search into the processing of materials using this technology
has emerged.
1315
Other applications such as high-ux neu-
tron sources and neutral beam sources for tokamak diagnos-
tics, the subject of this article, also appear promising.
These beams are produced in vacuum magnetically insu-
lated diodes which require a source of ions, an accelerating
voltage, and a magnetic eld transverse to the acceleration
gap to suppress electron ow and enhance the ion ow Fig.
1 . Ion currents typically exceed the vacuum space-charge
limit by 550 times owing to electrons conned in the ac-
celeration region by the applied magnetic eld. The beams
are produced and transported in vacuum of less than
10
4
Torr. Traditionally ions are drawn from the surface of a
polymer anode
16
converted to a plasma by a combination of
high-voltage ashover and electron impact. Polymer anodes
are unacceptable for applications requiring repetitive opera-
tion because of limited lifetime, excessive heat loading, and
high gas production. Also polymer anodes produce excessive
debris, have poor uniformity and reproducibility, and do not
allow the selection of the ion species typically these beams
have a mix of hydrogen and carbon ions . Anodes, which
draw ions from a preformed plasma are being developed at a
number
of
laboratories
to
overcome
the
above
limitations.
9,17,18
Some of these anodes allow the selection of
any gaseous ion species including hydrogen isotopes and he-
lium. Traditional single-shot beam accelerators using Marx
generators and high-voltage pulse lines , incompatible with
repetitive operation, are yielding to new high-average power
beam accelerators. An ion beam system operating at 100 Hz
in 10-shot burst mode, since no active cooling was
available
9
and a 300 keV intense beam source at 0.3 Hz
19
have been demonstrated.
II. DIODE DESIGN
The rst diode used on the Los Alamos Continuous High
Average-power Microsecond Pulser CHAMP accelerator
will be for a variety of applications. It will use a magneti-
cally insulated extraction diode
20,21
with plasma anode in
ballistically focused geometry 45� full focusing angle with
30 cm focal length . Extension to a straight unfocused beam
or longer focal length beams for tokamak diagnostics is
straightforward requiring only different anode and cathode
angles with respect to the system axis. The diode shown in
Fig. 1 operates as follows. The anode consists of a at pulsed
induction coil
22
in an aluminum housing. The high-voltage
coil is formed from four parallel sets of two turn spiral wind-
ings coaxial with the system axis. The coil in focusing ge-
ometry is dished in the form of a cone having a normal to the
surface of 22.5� with respect to the system axis. The plasma
anode is formed by rst radially ducting a puff of gas with a
fast acting valve rise time
100
s , located on axis, over
the coil surface. The valve will be actuated by a metallic
diaphragm driven either by eddy currents or a voice-coil
mechanism. When the gas puff is properly distributed, a fast
rising current pulse 1020 kA,
rise
12
s delivered to
the induction coil breaks the gas down and induces azimuthal
current in the plasma at the coil surface. The j
B
r
force on
the plasma accelerates the plasma to the radial opening in the
aluminum anode housing where it is stagnated against