.
T
1
.
?
11111
.
.
1
.
'
.
5
.
-
.
.
1
:
:
T
CA
S
.
I OFI
ORNL P
1429
V
-
..
.
-
1
no
...
-
-
-
.
:
.
O
ry
.
.
!
?..
tu
0
.
.
.., .,
:.
...!
»
...
-
. .
.
.
.
.
.
. .
.
.
.
...
.
.
:
..
00
P
.
home i
.
..
1.25
1.1.4
1.6
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS -1963
..
f
27 ر-م- ا (2) ه
CONF-650935-6
.
.
.
.
CN 21/112
-.
JUL 20 1965
-
...
-
-
-
The Oak Ridge Multiple-Pass Injection Experiment, DCX-2-
--
-
Viimiin..
Ini
-
-
.
.
• Bell
N. H. Lazar :*
R. A. Gibbons*
J. F. Lyon
G. G. Kelley
R. F. Stratton
Oak Ridge National laboratory
Oak Ridge, Tennessee
U.S.A.
:-
ir
1.
Introduction
sind
wir
:::8::
e
The two. DCX devices at Oak Ridge produce plasma of very high proton energy
by dissociation of 600 kev molecular ions in a magnetic mirror containment :
m. geometry. [1], [2] The densities in both experiments were expected to be suffi-
ciently high so that the ion loss rate would be dominated by scattering as in a :
controlled fusion device. These densities have not been reached as a result of
the presence of strong microinstabilities which occur at the lower densities 89
far obtained. In DCX-2, the maximum density has been at least 8 x 10% ions/cm'.
The effect of these microinstabilities is to spread the plasma energy both above
and below the injected energy and also to cause radial and axial spread of the
p].asma. Other high energy injection devices have been plagued, in addition, by i
low frequency hydromagnetic instabilities. [3],[47,57 These lower frequency
phenomena have been shown to be controll.ed by proper shaping of the containing
magnetic geometry (minimum B) or by, applying suitable electric fields between
the plasma and the walls. [3],[57,767,777 Such complications have, so far, been
unnecessary lw the DCX devices since the low frequency phenomena are not presente i
ine
n
Both DCX devices produce plasma of higher energy than any of the other high
energy injection experiments. The mean energy in DCX-2 is sometimes greater
than 800 kev. In DCX-1 it is about 300 kev. All of the other experiments of
this type have much lower injection_energies and produce plasma of less than
100 kev mean energy. [3], [4], [5], [87
2. Description of DCX-2
weiterende med malinamikshat 031.6 hów.ke:.
The principle of injection and the design parameters of DCX-2 were described
in the Salzburg Conference. [2] The device in its present form is shown in
Fig. 1. The overall length is 8.2 meters and the central region between the
mirrors where the trapping occurs is 2.65 meters. The central tank 1.6 approxi-
mately 1 meter in dianeter. The beam is injected near one end of the central
region at a place where the field is slightly depressed. The injection angle
? of about 7.50 is chosen so that the first turn of the 600 kev molecular jon
beam mi sses the injection duct. This duct shields the beam from the main field
and is compensated to cancel approximately the disturbance it creates in the
field. Over a region of about 70 cm either side of the midplane the field may
Research sponsored by the U. S. Atomic Energy Commission under contract with
the Union Carbide Corporation.
',!1111"11 MON
* Deceased
PATENT CLEARANCE OBTAINED. RELEASE I
.. THE PUBLIC IS APPROVED. PROCEDURES
ARE ON FILE IN THE RECEIVING SECTION,
.
.
..
.....
............
...............
.
.....
.
.-
....
....
...
..
.
.
.
..
jo o o o o o o o o o o o o o o.
..
-
-20
.
.
..
.
...
..
.
.
---
I be made flat to within about # 2 gauss. It has been found, however, as will !
be discussed below, that the maximum plasma density occurs when the field at
the midplane 18 slightly depressed. The magnetic field in the mirror throats
is 39 KG with a mirror ratio of 3.3. All field currents can be held constant
during an experiment to about 0.3%. The beam 18 operated steadily or it may
be pulsed ac any convenient repetition period -- usually about i per sec with
50% duty cycle.
--
-
:
i
i
--
.....
-
-.
-..-...
--
-
..
..
The accelerator tube shown in Fig. I makes use of cast epoxy skirts to
provide a sufficiently long external voltage breakdown path. 19 The internal
electrode shapes are designed to eliminate trapping regions for electrons pro-
duced in crossed electric and magnetic fields which under some conditions can
cause steady current drains, and to provide rapiü cleanup. Although molecular
ion currents of about 100 ma have been produced at 600 kev through a duct con-'
figuration like that in DCX-2, the maximum DCX-2 current has been limited to
55 ma apparently because of incomplete shielding of the stray magnetic field.
.. .. ..
-------
..
...
-
- -
..
The diagnostic instruments which have provided the rost information con-
cerning the plasma in DCX-2 are listed below.
--
----
-
-
-
-
-
-
.
:.:.
.. -
.
-
-
. .
-
-
...
-
...
.
.
.
-.
.
.
-
. . .
...
.
.
.
-
.
.
.
- -
-
-
2) Foil-covered Faraday cups for measurement of the current of energetic
neutrals from the dissociation of the molecular ion beam and from .
charge exchange. The attenuation of the foils normally used (0.254
nickel) begins at 150 kev and is 100% at about 50 kev. Gated integra- .
ting amplifiers are used to determine the total charge collected by
these detectors after the molecular ion beam has been turned off.
Callimated silicon barrier detectors. These detectors produce pulses
· proportional to the energy of the incident particles and are used with
a multichannel pulse-height analyzer to measure the energy distribu- .
tion of charge exchange neutrals leaving the plasina. They are collimated
typically by means of a su slit and 20u pinhole suitably spaced to i
accept a fan of particles 40° high by 0.15° wide. This geometry is
indicated schematically in Fig. 2.
Radial electric probes and magnetic loops are used with narrow and
wide band rf spectrometers for measurement of the rf in the plasma
region. One of the spectrometers can sweep & range of 2 Gc from
O to 42 Gc. A second sweeps up to 3 wc from 0 to 25 mc and a third
10 to 100 mic at a sweep frequency of 10 kc.
Various calorimeters. The power in the beam and plasma striking the
injector is routinely monitored. The remainder of the accelerator
beam power is measured by separate water circuits inside the beam
duct. A probe may be inserted near the injector to measure the in:
jected beam power.
Gridded probes for measurement of cold plasma characteristics. Two
or three grids in front of a collector are used in either plane or
hemi spherical geometry. Voltages are applied to the grids to separate
the ions and electrons, and the collector voltage 18 swept to measure
the energy of either ions or electrons.
-
...
-
-
-
.
..- -
tana
San
'mon
C
Experimental Results
The experiments with DCX-2 may be divided into two main classes:
(1) Dissociation of the molecular ion beam by the beckground gas and secondary :
11.41.,IIICATION
0:0
O
o
o
o
o
o
o
o
o
u
-
-
a
..
.
.....
....
....
men
-
-
.
.
.
.
. .
. .
.
.
.
-
.
.
-
-
-
..
.
-
-
-
.
.
.
.
plasina. Roughly 40% of the protons produced are in orbits which will not strike
the injector when the orbit precesses about the axis of the machine.
(2) Enhanced dissociation of the niolecular ion beam in a dense high vacuum
arc. The arc is situated along flux lines which pass throv.gh the molecular ion
helix diametrically opposite the injection point. All protons formed in the
Arc must be scattered radially more than ane larmor radius to strike the
injector. Both lithium and hydrogen arcs have been used for this purpose.
Most of the experiments performed in DCX-2 in fact have been in the presence of
a lithium arc.
Lithium Arc Experiments. The highest densities yet observed in DCX-2 have
been produced with lithium arc dissociation. The arc operetes between eleca
trodes spaced approximately 6 meters placed outside the mirror coils. Lithium
16 stored in a stainles8 steel boiler which is heated electrically to tempera-
tures between 000-700°C. A large diameter tube connects the boiler with the
tantalum anode which has a tungsten plug to terminate the arc. The temperature:'
of the anode 18 independently controlled. A filamentary tungsten cathode is
used so that the arc may be turned on and off easily. The effective cathode
cross-section is 0.45 x 1.0 cm. The arc normally is operated at approximately
20 amperes and a voltage of 180 volts 2.5 cm from the field axis opposite the
injector. The injector 18 positioned so that the 26.5 cm diameter inolecular
ion orbits pass through the expected position of the main arc column (Fig. 2).
From probe and spectroscopic evidence the arc density has been estimated to be
5 x 1015 ions/cm). A large radial transport of lithium ions is observed which
is found to increase when the hot plasma is present.
The total number of energetic particles in the plasma can be determined from
the time integral of the charge exchange flux after the beam is shut off. If
losses other than charge exchange occur after the beam is turned off the ..
measurement is an underestimate of the trapped particle number. To obtain &
density the mean radius of the plasma is taken as 15 cm. The results must
include a correction for low energy particles which are absorbed in the nickel
foil. The correction amounts approximately to a factor of two but cannot be
determined exactly because of incomplete knowledge of the energy spectruat
low energies. An oscillograph trace taken during a typical beam turnoff
experiment is shown in Fig. 3. Here the integral of the charge exchange flux
is plotted as a function of time. The best estimates of plasma density yield
values of 8 x 109 energetic ions per cm trapped in the central volume of the
main magnetic field with an injected H2 beam current of ~ 40 ma. Experiments
have also been carried out with the arc 7.5 cm from the axis and on the axis
itself with quantitatively somewhat different results but qualitatively
similar behavior. The mean lifetime of the particles in the plasma cannot easily be ::
evaluated from the time dependence of either the current or the current integral
since the current decay is not a simple exponential. The trapped particles
have a very wide energy distribution and the density of particles which charge
exchange with them changes with time after the beam is turned off. There is
also evidence of continuing energy mixing. All of these effects influence
the shape of the decay curve.
A second major source of informacion concerning the plasma ions is derived
from the collimated surface barrier silicon detectors. Data typically are
taken by scanning across the plasma volume for a period of several minutes while
the..beam is being pulsed on and off continuously at a one second repetition .
rate. The pulse-height distribution thus obtained, when corrected for the
charge exchange rate as a function of energy, yields the energy distribution
of the contained particles. A typical spectrum with high plasma density 18
.
.
.-..
,
,
-
-
.more
.
.
...
...
-
-
-
.
.
-.-
-
--
--...
1. Vio 0 0
0 0
0
0
0
0 0
0
0
0
0
0
-4.
shown in Fig. 4. A scan in angle made with the slit perpendicular to the
field axis would show only two peaks if the trapped particles retained the
angle at which they were injected. Figure 5 shows this behavior at low beam
currents. When this beam current is raised above several milliamperes,
however, a prominent peak appears at zero degrees. The peak shown in Fig. 5b, :
observed with ~ 40 ma injected H2 beam current corresponds to particles with
a spread in angle of less then + 2°. Such particles are reflected by the very :
slight rise in the magnetic field at + 70 cza from the midplane. The intensity
of this peak and in general the trapped. density is maximized by varying the
shape of the field slightly. A field dip of only no 20 gause over the central
140 cm produces the maximum density. Variations of only a few gau88 from this
field shape result in a significantly smaller trapped density.
From the angle corresponding to the maxima of the side peaks observed 12
scans with selected total energy, the axial energy of the particles 18 de-
termined. (Fig. 6). It is apparent that the average axial energy is almost
unchanged from the injected axial energy, in spite of the very large change
in total energy.
Energy distributions were observed for the particles in the central
peak and similar data were taken for the side peaks. The results, corrected
for the charge-exchange cross sections, are shown in Fig. 7. It is clear
that the highest energy particles are contained in the plasma group with the
nearly zero axial energy.
An indication of the radial distribution of the plasma 18 obtained from
angular scans with the slit aligned parallel to the field axis. Such ar
angular distribution for the group with the higher axial energy was obtained
using a bar placed to intercept particles with pitch angles of less than +3º.
The particles with the larger &xial energy are restricted to a radius of less
than ~ 15 cm. This limit is determined by the position of the injector and the
fact that the particles precess about the machine and therefore are intercepted
by the injector as they spread radially.
From the intensity of the observed current in the central peak and the
longer lifetime of these more energetic particles, it is apparent that most
of the density in the plasma is associated with this group. Typic 1 angular
scans with the collimator slit parallel to the field axds for the higher
energy particles indicate they are spread in radius to ~ 30 cm with a mean
value of . 15 cm. Despite their large radial extent, these particles do not
strike the injector since they are reflected by the slight field increase in-
side the injector position.
!
.
CX
Quantitative estimates of the average lifetime of the charge-exchanging
particles may be determined by averaging the energy-dependent lifetime over
the observed energy distribution; i.e., the mean lifetime for charge-exchange, 1
is ir = I (E) (nov )- dE/ So Iex (E) DE where no 18 the density of
charge-exchange centers, Icx(E) is the charge-exchange current per unit energy
interval, Ony is the charge-exchange cross section at a given energy, and v
the velocity of the energetic particle. Measurements of Ocx are available
only to 1 Mev (19 and have been extrapolated to higher energy. The assump-
tion is made that the charge-exchanging centers have cross sections equal to
the nitrogen cross section. The values of 7cx determined in this way may be
multiplied by the total charge exchange current to yield a density estimate
during equilibrium. The values so obtained, assuming no to be the neutral
Ooo o o o o o o o
o
o
o
particle density obtained by pressure measurement, are a factor of 5 to 10
higher than the values obtained from the time integral of the neutral flux
after the beam is turned off. The disagreement between these two measurements
of plasma density is considered to be the result primarl.ly of a high density
of cold plasma in the volume, which increases the charge-exchanging density.
The cold plasma density may be estimated from the dissociation current
observed to the energy analyzer from positions outside the arc in the regioa
of the cold plasma. Using the nitrogen dissociation cross section one obtains
a density of ~ 5 x 1010 ions/cm). The distribution and density of these cold
ions also may be estimated from the saturation current to a gridded probe
inside the plasma chamber tut outside the region where the energetic ions are
trapped. An order of magnitude agreement is obtained. Retarding potential
measurements using these gridded probes show the eleciron energy and plasma
falling from 250 ev near the axds to less than 20 ev at 15 cm.
Figure 8 shows a series of measurements taken with the collimator slit
of the energy analyzer aligned parallel to the field axis wher specific
energy ranges were electronically selected. Assuming the particles prece88
uniformly about the axis of the field, one would expect to find a distribution
of current as a functio of detector angle symmetric about a point displaced
one larmor radius from the axis. In fact, the observed distribution is
asymetric. We believe the asymmetry to be due to non-uniform distribution
of the cold ions.
The peaks observed in the parallel slit scans produced by dissociation
of the direct and reflected moledular ion beams in the cold plasma and
neutral gas also are used to estimate the percentage dissociation of the
beam. The results vary from ~ 50% to greater than 90% dissociation depending
on the arc conditions. The power to the water-cooled target on the injector,
however, is typically. 60-90% of the total beam power. The additional power
appears to be due to the loss to the injector of protons from the plasma group
with the large axial momentum. From the observed charge exchange flux, in
some cases as much as 3/4 of the injected beam current appears to have been
lost by this mechanism. At plasma densities of 4 x 109 ions/cm) witîl an
injected beam of 25 ma, approximately half of the particles were accounted
for by charge exchange flux. For beam currents below 25 me, the energetic
ion density decreases approximately linearly with injected current. As the
current is raised above this value there is evidence that the density does i
not increase proportionately. Insufficient data concerning the nature of the
arc and the change in composition of the background plasma at the higher :.
currents makes it uncertain, however, that a density limit determined by
plasma instabilities has been reached.
.
..
.
.
. .
.
.
.
-
-
...
.
.
.
Radiofrequency measurements have been made using both electric field
sensitive antennas and magnetic loops. These probes show evidence of
electric and magnetic fields at the ion cyclotron frequency and many of its
harmonics. For example in Fig. 9a 18 shown the frequency spectrum to 2 Gc
observed with a radial electric probe, and harmonicB as high as the 100th
-
...
.
related both to the energetic ion plasma density and the density of cold
plasma. Tie radiofrequency persists after the beam is shut off for times i
. . 88 long as several seconds. Measurement of the energy spectrum during this
time shows, a continuing production of low energy ions. This experiment 18
Tin
lil,!:"111
-6
additional eridence that the energy spread of the plasma particles 18 related
to the presence of the rf.
--
Another oda characteristic of the radiofrequency's signals is found in
detailed studies of the structure of the individual harmonic peaks. The
width of the peaks is quite large, of the order of 700 kc at high densities,
and the shape and frequency of the peaks change when the beam is turned off.
Figure 3. shows the spectrum in the vicinity of the ion cyclotron frequency
as a function of time. It is seen that the frequency varies from slightly
below the ion cyclotron frequency up to that value but apparently never ::
exceeds the ion cyclotron frequency.
.
.
--
.
-.
.
-
.
..
- -
-
-.
-
-
-
- -
-
-
...
-
Careful examination of the low frequency end of the spectrum shows no
evidence of low frequency oscillations of the plasma as 18 seen in other
high energy Injection machines 37,247,157. Unexplained fluctuations in the
current to the neutral purticle detectors have been noticed. These fluctua-
tions are related to variations in the low density, large axial energy plasma
and are not seen in the charge-exchange flux from the plasma with low axdal
energy. The fluctuations are found to be unrelated to energetic charged
particles escaping to the walls. In fact, although pulses from energetic
charged particles are seen with gridded pro?:rs outside the trapped plasma, these
are independent of plasme, density and do not appear to represent a signifi-
cant 1088. Axial correlation of these gridded probe signals 18 absent.
This fact and the lack of low frequency electric oscillations indicate that
flute-like disturbances are not present.
e..
...
!
-
-
-
.
-
-...
-
-.
Hydrogen Arc Experiments. The parallel slit scans using the energy
sensitive detector looking at the plasma produced with lithium arc dissocia-
tion show that the 1088 by charge-exchange is at least as much from the arc
as it is from the background gas and cold ions. A hydrogen arc has been
used to eliminate this 1068. Unfortunately the lowest background pressure
attained thus far with this arc however, is quite high ( 3 x 10-6mm Hg in
DCX-2) because of plasma flutes · [11] These arc flutes have been identified by
probe correlation measurements.
-
by
An evaluation of the usefulness of the arc as a dissociation medium is
complicated by the possible influence of several effects. The percentage
dissociation of the molecular hydrogen ion beam and the observed trapped
plasma density increase with increased gas feed to the arc. It is not
clear, however, that the increase in observed density is due only to increased
dissociation since the partial pressure of hydrogen in the plasma region
also is increased. As will be explained below an increase in hydrogen ::.
pressure in the same pressure range with gas breakup also increases the
density of the trapped plasma.
....
.
....
....
..
The perpendicular slit scans of the plasma produced by hydrogen arc
breakup show that the group of particles with low axial energy is present
also with the hydrogen arc. The parallel slit scans, however, as expected,
show no peak in the direction of the arc indicating that it does not contri-
bute appreciably to charge-exchange 1088. Integrale of the charge-exchange
flux give deneities of 1 to 2 x 109 ions/cm at a gage pressure of
5 x 10-6 mm Hg. The rf seen with the hydrogen arc 18 qualitatively the same
as with the lithium arc. The arc is formed between a long hollow copper
anode and a hollow tungsten cathode [12],[137. Under the condition of ;
lowest gas flow, 1t operates at about 80 amperes at 200 volta'withi'ss gas
flow of 0.25 atm cm of hydrogen per second.
.
.
....
...
iiiiitill: VOJ
o oca ao o o o o o o o o o
4.
Gas Dissociation
Experiments with gas dissociations at base pressure (~ 2 x 10-7 mm Hg)
show significantly lower density than is obtained with the arcs :
(w 1 x 10° ions/cm”). As expected, the charge-exchange current increases
and the plasma density remains approximately cons'tant as the pressure 18
incressed by introducing N, g88. The decreased lifetime is offset by an
increased trapping rate although an exact balance is not expected due to
variations in energy mixing. When hydrogen gas is added, hoyever, the
plasia dens ty increases, reaching a maximum value of 5 x 10° 1ons/ cm at
a gage pressure of 1.8 x 10-) mm Hg. The energy of the trapped plasma
spreads and the mean energy i 1808. Figure 10 shows the variation of
density and the mean energies of this plasma as a function of gage pressure.
Perpendicular slic scans did not show a central peak, however, during this
experiment, although a small central peak usually is observed at the opti-
mum field setting at base pressure. Electric probe measurements show an rf
spectrum differing somewhat from those obtained with arc breakup; see
Fig. 9b. The plasma potential at base pressures with gas dissociation is
much higher then in the arc experiments. The potential varies rapidly in
the region of several hundred volte positive relative to the walls, the
electron energy is observed to vary between 50 and 100 ev; the ion energies
between 45 and 130 ev. At the highest densities at elevated pressure, ion
and electron energies are considerably higher.
5.
Discussion of Results
The plasma in DCX-2 18 quite complex, consisting of usually two groups
of fast ions and a relatively higher density of colder ion components. In
every case the highest densities are associated with a broad spread of
energy and a mean energy many times the injected energy. With src breakup
the particles having the widest energy spread have a very small axdal energy,
implying heating by electric fields very nearly perpendicular to the axis.
With gas breakup, however, under the condition of a high background pressure
of hydrogen, the broad energy distributions can be obtained without the
appearance of this separate perpendicularly oriented group. The lifetime 18
long in both cases in spite of high gas pressure because of the accumula-
tion of very high eiergy ions.
.
-
.
-
..
.
c
The pressure in the plasma region is not reduced by plasma pumping as
much as had been expected. Apparently the cold hydrogen ions, and to a
lesser extent heavier ions, are being heated by interaction with the fast
ions. If the cold ions are heated more quickly in the perpendicular
direction than they can scatter axially, they are prevented from flowing
out of the plasma region along the flux lines through the mirrors. Eventually
they are lost by charge-exchange and returned as neutral gas particles.
.m
-
.-.
-
-.-.-.
-.
-.-.
.
.
.
e
The mechanism by which the group of particles of zero average axial
energy 18 produced is not understood. Several mechanisms have been suggested.
One possibility is a form of streaming interaction among the counterstreaming
particles having the initial axdal energy. A second possibility is that cold
hydrogen ions are resonantly heated to high energy by essentially perpendicular
fields at a sufficient rate to overcome charge-exchange 108808. The difficulty
in choosing between the various possible models lies in the imprecision of
measurement of plasma properties. These difficulties must be resolved by
improved measurement techniques.
lin! Ils ...,6
0
0
?
.
.
Ô
o
o
o
o
o
o
o
o
o
o
d
-8-
The authors gratefully acknowledge the technical assistance of
J. S. Culver, S. M. Decamp, J. C. Ezell, A. J. Wyrick, E. . Moore,
T. F. Rayburn, and C. W. Wright. In addition we take pleasure in acknowledge
ing the work of J. E. Francis in the development of the filamentary 13thium
arc and the improvements in the hydrogen arce vsed in DCX-2 and the major
contributions of 0. B. Morgan and R. C. Davis to the developmerit of the
injector. We wuld like to tiank Mozelle Rankin for computational assistance
in many phases of the design of experiments. We would also like to expre 88
our appreciation for helphu decussions with T. K. Fowler, J. L. Dunlap, and
1. Postma. We would like to thank A. H. Snell for continued encouragement
and 88818tance.
.
. .
.
.
. . .
.
. .
.,-,.
ivoi!
!!:,:
1.
BARNETT, SELJ, LUKE, SHIPLEY, and SIMON, Proceedings of Second UN
International Conference on Peaceful Uses of Atomic Energy 31, 298
(1958).
BELI., KELLEY, LAZAR, and MACKIN, Nuclear Fusion (Supplement) Part 1,
251 (1962). Full details of the DCX-2 experiments are found in the
Thermionuclear Division Semiannual Reports ORNL 3239,3315, 3392, 472,
3566, 3652, 3760, and 3836.
BOGDANOV, GOLOVIN, KACHERYAEV, and PANOV, Nuclear Fusion (Supplement)
Part 1, 215, (1962).
DAMM, FOOPE, FUICH, and POST, Phys. Rev. Letters 10, 323 (1963).
KUO, MURPHY, PETRAVIC, and SWEETMAN, Phys. Fluids 7, 988 (1964).
5
BAIBORODOV, TOFFE, PETROV, and SOBOLEV, Atomnaya Energya 14, 143 (1963)
(Translation: Soviet Physics, Atomic Energy 14, 459 (1964).
8
DAM, FUOTE, FUTCH, GORDON and POST, Phys. Rev. Lett. 13, 464 (1964).
G. GOURDON and F. PREVOT, Nuclear_Fusion (Supplement) Part 1, 265, 1962.
G. G. KELIEY, 0. B. MORGAN, Minutes, 1964 Linear Accelerator Conference,
p. 456,' MURA : 714,UC-28, TID-4500.
10
BARNETT, RAY, and THOMPSON, Atomic and Molecular Collision Cross-Sections
of Interest in Controlled Thermonuclear Research, ORNL-3113 (Rev)(1964).
R. A. GIBBONS, N. H. LAZAR, Phys. Rev. Letters 13, 703 (1964.
12 R. A. GIBBONS, and R. J. MACIIN, Proceedings Fifth International Conference
on Ionization Phenomena in Gases, Munich 1961, Vol. II, 1769, North
Holland Publishing Company.
13. FRANCIS, ELL, LAZAR, GIBBONS, ORNL Report 3760, D.: 52.
ooooooooooooooooo
-10-
Figure Captions
Fig. 1
Longitudinal section through DCX-2. Rings of detector ports are
located in the spaces between the" coilB..
Fig. 2.
Transverse section through DCX-2 showing possible locations for
various detectors. The energy analyzer is shown with the slit per-
pendicular to the axis.
Fig. 3,
RF spectrum and foil neutral detector integrale vs. time after beam
turn-off. Sweep sneed 0.1 sec/cm. Upper irace, integral at center
of machine with 3.3 x 10-10 coul.omb/cm. Lower trace, integral of
the sum of five detectors along the machine axis with 10° coulomb/ cm.
Trace is intensity modulated by the amplitude of the rf.
Fig. 4.
Energy spectrum of charge-exchange neutral particles from the DCX-2
plasma and the corresponding energy distribution of the trapped
plasma as seen from the center of the machine. Mean energy of trapped
.protons is 847 kev. Symbols at 10° indicate no counts.
Fig. 5.
Pitch angle distribution of charge-exchange neutrals as seen from the
center of the machine. Results at low beam current (0.3.ma) are
shown in (a) and at high beam current (40 ma) in (0).
Fig. 6.
Axdal energy of the side peak maxima for different total ion energies.
Fig. 7.
Energy distribution of the plasma components with high and low axial
· ezzergy.
(a) che low axdal energy component.
Th? higher axial energy component. The circles and crosses
distinquish between the distributions seen looking toward and
away from the injector respectively. The detector was in the
mid-plane of DCX-2. Symbols at 10° indicate no counts. ;**
Fig. 8.
Energy-resolved parallel slit scans of the DCX-2 plasma as seen from ·
the center of the machine. The detector is in the position shown in
Fig. 2. The peaks seen at 27° and 31° are due to 300 kev dissocia-
tion neutrals from the beam and are separated because of precession
of the molecular ion orbit during reflection. These peaks are also
seen in the scans not including 300 kev because of scattering in the
slit, resolution in the detector and che occurance of random sum
lines. The peak near zero degrees is from charge exchange in the
Lithium arc.
Fig. 9.
Frequency spectrum from 100 mc to 2 GC observed with a radial
electric probe near the mid-plane in DCX-2. (8) Hydrogen erc
dissociation. (b) Gas dissociation.
F18. 20. Density and mean energy of the trapped plasma produced by gas
dissociation as a function of pressure. The pressure was raised
by adding hydrogen.

aug
LEGEND
1. ELECTROMAGNET COILS
2. MAIN VACUUM TANK
3. COOLED LINER
4. DIFFUSION PUMP AND BAFFLE
5. ION SOURCE AND ACCELERATOR
6. INJECTION SNOUT
7. PLASMA
8. ARC ELECTRODES
9. ACCELERATOR X-RAY SHIELD
10. DIAGNOSTIC PROBE (ONE OF MANY
11. TITANIUM EVAPORATOR
12. BEAM TARGET
.
7
.
-
IT
.
Mgo 1
. Longitudinal section through DCX-2. Rings of detector ports are
located the spaces between the coils..
..
.

·
FOIL
DETECTOR
INJECTOR SNOUT
BEAM TARGET
2.5 cm
POSITION OF ARC
al, ...
--600 kev H2+
ION ORBIT
26.5 cm DIA.
.
GRIDDED
PROBE
. .
.
.
91.5 cm I.D LINERS
.
.
.
-1.6 cm x 5 u SLIT
204 PINHOLE
SILICON
BARRIER
DETECTOR
-AIR LOCK PORTS
FOR PROBE
INSERTION
.
Fig. 2. 'Transverse section through DCX-2 showing possib). Locations for
various detectors. The energy analyzer is also with the slit pero
pendicular to the ado.
PHOTO 69968
- 18.5 MC
.
1+ ION CYCLOTRON FREQUENCY
- 18.0 MC
.
.

.....:
:
.. F1g. :3. RF spectrum and foil neutral detector integrals vs. time after bean
turn-off. Sweep speed 0.1 sec/cm. Upper trace, 'integral at center ...
of machine with 3.3 x 10-10 coulomb/cm. Lower trace, integral of
...the sum of five detectors along the machine axde with 10° coulomb/ caine :
Trace 18 intensity modulated by the amplitude of the ri.
ORNL-DWG 64-11750

4/16/64 B-35
0500 noon
TRAPPED PROTON
DENSITY
DENSITY OR COUNTING RATE PER ENERGY INTERVAL
(arbitary scale)
CHARGE-EXCHANGE
NEUTRAL CURRENT
100 Ollilllllll
0 0.4 0.8 1.2 1.6 2.0 2.4 2.8
ENERGY (Mev)
*
Fig. Hon
;
Energy spectrum of charge-exchange neutral particles from the DCX-2
plasma and the corresponding energy distribution of the trapped
.pleame, as seen from the center of the machine. Mean energy of trapped
proteins is 8447 kev. Symbols at 10° indicate no counts.
.
.

ORNL-DWG 65-6702
(a)
(b)
Fig.-5.
nicio
Pitch angle distribution of charge-exchange neutrals as seen from the 1::
:center of the machine. Results at low beam curreat 10.3 me) are
shown in (a) and at high bean current (40 m2) 10 (b).
...licoid
.....
iodo.

...
1
ORNL.-DWG 64-9097)
3.0
;
AXIAL ENERGY (kev)
7/15/64
III
200
500
500
TOTAL ENERGY ( kev)
50

100
1000
1000
.
...
-
.
Fig. 6.
Axdal energy of the side peak maxima for differebi total ion energies.
-.-.
-
..
.
.
.
.
ORM OUG 68-703
CENTRAL PEAK SCAN
11/2/64 B-6b, 70
ego
11/2/64 B-78
00000 at
DENSITY PER ENERGY INTERVAL
(arbitrary scale)
CENTRAL PEAK
MAXIMUM
பாட்டபட்ட
. DENSITY PER ENERGY INTERVAL
(arbitrary scale)
To 0000
100 LIIIIIIIIIIIIII Lorelor
* Loodan-kod dooooooooooo
O 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
ENERGY (Mev)

100L
O
0.2
moet obenem
0.4 0.6 0.8 1.0 1.2
· ENERGY (Mev)
1.4
Fig. 7.
Energy distribution of the plasma components with high and low axial :
energy.
(a) · The low axdal energy component...
(6) The higher axdal energy component. The circles and crosses
distinquish between the distributions seen looking toward and
away from the injector respectively. The detector was in the
mid-plane of DCX-2. Symbols at 10° indicate no counts.

COUNTS (arbitrary units)
•
..
·
.
UP 20 10
DETECTION ANGLE (deg)
0 10 DOWN
150-200 kev
100-150 kev!
C
300-350 kev| DETECTION ANGLE ( deg)
UP 20 10 0 10 DOWN
200-250 key F1.2 -9.4 Mev +
1600-700 kev h
450-500 kev
ORNL-DWG 64-4900

Fig. 8
Energy-resolved parallel slit scans of the DCX-2 plasma as seen from the center
of the machine. The detector is in the position shown in Fig. 2. The peaks
seen at 27° and 31° are due to 300 kev dissociation neutrals from the beam and
are separated because of precession of the molecular ion orbit during reflection.
These peaks are also seen in the scans not including 300 kev because of scattering
ita the slit, resolution in the detector and the occurance of random sum lines.
mha wanlo nanm nonn dorreed to pum charop exchange in the lithium arc.
USTUP HII!hange in the lithium arc.
PHOTO 69967
2000
T
о
мс
1000
ó 'Mc' '1000
(a)
'I '2000

(b)
.
.
100 m tief beomaduseth mesos
Fig. 9.. · Frequency spectrum from 100 BC to 2 Gc observed with a radial ;:
:::
electric probe near the mid-plane in DCX-2. (a) Hydrogen arc: .
dissociation. (b) Gas dissociation.
;.........

ORNL-DWG 65-6701

T1000
DENSITY N+ (IONS/cm3)
AVERAGE ION ENERGY Ē (kev)
1071
10°7
10-6
10-4
105
GAGE PRESSURE (mm Hg)
Mg. 10. Density and bean energy of the trapped plasma produced by gas
...!! dlouocation as a function of prospire. The pressure was raised
iwi by adding hydrogen...
..
.
.
.
.
.
.
. .
..
.
11/9/65
DATE FILMED
END
-
-
...
*
.-
*
--
-
-
-
-
.
.-
:
.
.
.
v
-