.
KU
.
1.
49
UNCLASSIFIEDI
ORNL
'
N
.
..
.1
2
WAR
UN
41.
.
P.
2
EN
.
3
.
4
456
www
.....................................................
ORNL-P-456
EINF-2017/
JCT 81964
STUDIES DI THE HALO Q ME DEUTERIUM ARC AND A PIG -TYPE DISCHARGE*
N. H. Lazar
C. W. Blue
R. A. Gibbons
0, D. Matlock
_
Oak Ridge National Laboratory
Oak Ridge, Tennessee
August 20, 1964
.
-LEGAL NOTICE

of way before us, or where in contation, cum swall of the Council or the Valsad
The mport me popared muscount of Govonul sponsored wort. Neither the Unimed
ham, ter te mbulon, mor my person kung on behalf of the Counselons
A. Xan my nerasty or reprimaution, eixound or implied, mul rompact both acry.
macy, completene, or woulun ollo Information containdu do report, or that we were
of way tulonution, parto, med, or proces. d cloud in to report may not latring
peintuu omand : or
D. Aamuwa w Idabellum w roupact to the ww. of, or lor damages resulting frow we
no way Wormatien, men method, or pcom dielowed in a report.
Ao wood in the whore, "perma athy on behalf of the counlukom" including my m.
mergo or contractor of Commalaston, or employee of much contracbr, to the out that
mot a g e or contractor of the Countertoo, or raplayne of much coolriclor proper
terminal, or provides me te, any information pureut to Memploy or contract
wth Cuestoa, or ws onploynal mu much contractor.
Mr
with
o
to, way mornin
much contractor
,
(To be published in the proceedings of "International Symposium on Diffusion
of Plasma Across a Magnetic Field," held at the Hotel "Kaiserin Elisabeth."
Feldaf ing/Starnberger See, June 29 to July 3, 1964. Publication will be
informal and distributed to conferees only.)
.
:
I
*Research sponsored by the U. S. Atomic Energy Commission under contract with
the Union Carbide Corporation.
:
V
22
.
STUDIES IN THE HALO OF THE DEUTERIUM ARC AND A PIG-TYPE DISCHARGE*
N. H. Lazar
C. W. Blue
R. A. Gibbons
0. D. Matlock
Oak Ridge National Laboratory
Oak Ridge, Tennessee
.-.-
.-
S
ZE
+
For some time we have been interested in the source of the relatively
high pressure observed in the operation of a deuterium arc.? The arc
column has been studied by Mackin and Gibbons and, as a result of many
combined diagnostic techniques, le observed to have the following properties:
Te 0.15 Farc
T4 = 16 ev
A = 1.6 cm2
n = 1-2 x 1014
The temperature of the electrons (in ev) appears to vary with arc current
(given in amperes) as shown. A 18 the arc cross-sectional area, determined
optically. Geometrically, the arc is a fairly thin cylinder with a diameter of
approximately 2 cm. The density 18 quoted for an input deuterium gas feed rate
of 3.0 cc/sec and 18 expected to be proportional to gas feed rate. All the gas
fed into the anode is observed to appear at the cathode, where it is pumped by
& 20-in. diffusion promp. The facility in which our measurements were carried out,
the Gas Arc Facility (GAF), 18 the same as the one previously used. Figure 1.
shows a schematic layout of the experiment. The magnetic coils, arranged to
produce a 2:1 mirror ratio, produce a field in the midplane of up to 9000 gau88.
The arc has been run at fields as low as ~ 1.5 kgauss. Base pressures are
~1000 mm Hg in the central region with cylindrical baffles ~2-7/8 in. Inside
diameter, 9 in. long in the mirror throats. These were chosen, experimentally,
to produce the minimm pressure in the central region with the arc running;
p~ 2 x 10-5 mm Hg as indicated on an ion gauge for the deuterium arc conditions
.
.
.
yi
.
RO
1
as given above.
3.79
....
*Research sponsored by the U. S. Atomic Energy Commission under contract with
the Union Carbide Corporation.
.
.
.
....
X
.
"
.
.
2
.
2
11
1
.
.
.
"
Il
.
3
U
In order to determine if the source of gas may be as a result of lon
diffusion across field lines, the radial distribution of lon density was
measured in the halo of the arc column as a function of magnetic field. The
effective area of the tungsten wire probes were 2.5 x 10-2 cm“; 0.010 in.
diameter ~ 1/8 in. long. The wire was shielded with boron nitride surrounded
by a grounded cylinder beyond this area back to the insulator so as to define
the geometry. Langmuir curves were taken to establish the current beyond lon
saturation. The results are shown in Fig. 2, where the current at each point
18 'multiplied by (r/T.)1/2 for comparison with Simon's theory of diffusion
in cyll.ndrical geometry.? The e-folding length 18 been to be nearly independent
of B and to be equal approximately to 1 cm. This length may be related to a
diffusion coefficient for transport perpendicular to the direction of the field
assuming a model for the longitudinal transport. In view of the neutral
pressure range, the perpendicular transport should be governed by ion-electron
scattering. Taking any of the current simple theories, there would be either
a 1/B or 1/B² dependence of the diffusion coefficient. The observed dependence
18 independent of B.
In view of this difficulty in comparing the e-folding Jength with a theory, .
It was decided to look for time variations in the radial diffusion. Figure 3 .
shows the time dependence of the floating potential as seen on a probe and the
1.4,
observed correlation of two probes displaced a distance spam in z. Figures
4 and 5 indicate the correlation for lon current and electron currents to two.
.
probes, as seen on a dual beam oscilloscope triggered from the upper trace
signal. The correlations are so exact, that when the two probes are exactly
on the same field lines, the two traces may be superimposed on the oscillo-
scope screen. The correlations between probes displaced in r or o are not
as good. Mgure-6 shows such a correlation; from pulse to pulse the
.
.
.
..
.
S
Tri.
exact shapes of the signals seen on the two probes differ somewhat.
however, by taking many pulses, an "average" time delay may be deter-
mined, so that an average radial or azimuthal velocity of the plasma can
be defined. This correlation becomes poorer as the spacing between probes
increases. For example, if the probes are displaced 90° (at r = 2 cm),
the two pulbes appear randomly related. When the probes are close, the
direction of rotation of the plasma can easily be established by triggering
the scope on the socond trace; only the tails of the first are then vieible.
The sense of rotation is in the direction of an ion moving in its Larmor
orbit.
These pulses, then, correspond to a narrow plasma (~0.4 cm wide)
moving radially and azimuthally correlated all the way from the midplane,
through the mirrors, to the anode; geometrically, a "Plute."
The radial and azimuthal velocity has been measured from the observed
"average" correlation between two probes, as a function of distance from the
arc center in a field of 7 kilogauss in the midplane. These results are shown
in F188. 7 and 8. It may be seen that outside r = 2 cm, vx/v0 = constant ~ 1.2,
but inside this radial distance, v, increases rapidly. 'It appears the flute
starts radially then slows down and moves in a spiral toward the walls.
Both the maximum amplitude and the frequency of flutes are observed
to decrease as a function of radius. To determine if the ions were simply
running out the ends of the flutes along field lines, 81x-inch diameter plates ..
were placed about each baffle at the two mirror coils. These end plates
were also divided into quadrants to pewnit correlation measurements. Figure 9
shows the correlation observed between the ion current to a small wire probe
and the current to one of the quadrants. The decay of the current to a
quadrant may be analyzed assuming the ion temperature is unmodified from its
.
.
?"
AN
R32
.
.
IT
value in the arc and the ions diffuse along field lines like a gas. It
appears, then, that the flutes are dissipated as the long run out their
ends along field lines as they spiral outward. Taking the average currents
observed on these end plates after recombination as the influx of gas into
the central region, and using the known arc pumping speed, it appears that
we now may explain the neutral pressure about the arc. This current
represents about 4% of the total ion current in the arc.
We still have difficulty in reconciling the observed radial dependence
of the average current to a probe with the behavior of the flutes. Although
the higher observed average current with lower fields is consistent with the
end plate current behavior, we have not, as yet, been able to establish &
quantitative relation of the probe current with the frequency and amplitude
of the flutes. There seems to be additional background plasma, perhaps
resulting from tearing up of the flutes as they move radially in addition
to the lons diffusing axially, which has not been accounted for in the
model.
restauranter
So far, we have established that the primary source of gas arises from the
cross-field transport of some of the arc plasma and the subsequent recombi-
nation of the lons on the end walls. We have attempted to determine the
dependence of this transport on arc parameters, with only moderate success.
Figures 10 and 11 show the variation of the total currents to the end plate
(presumably total transport) as a function of arc current (electron temperature)
for different magnetic fields and gas feed rates. There seem to be critical
conditions at certain values of electron current density at which the transport
increases by a factor of 2-3; however, it is anomalously high for all the
conditions represented by the curves.
d
iz
It may be worth remarking, at this point, that we have not yet
established the cause of fluting. The excellent correlation along z
suggests the axial wavelength is very long - much longer than the arc
length. This seems to rule out, helical-type instabilities. It is
apparent that the flutes are of high n-order, and for the field gradient
in which the plasma exists, we would expect it to be stabilished by the
finite Larmor radius of the lone. Nevertheless, the plasma column 18
observed to rotate with a velocity, v. 2 x 105 cm/sec at the surface
(much faster than expected from the field gradient). This is not
as fast as expected for the diamagnetic drift, which is supposed to be
stabilized by the finite Larmor radius--and so it is difficult to under-
stand how this can be the source of the instability.
-161.
i
s
iseviidiini
A second series of experiments have been performed with a somewhat
less dense arc produced in a PIG geometry. I would like to describe these
experiments because of the possible interest in such a plasma source and,
perhaps, in the gridded probe technique which has been developed; primarily,
through the work of P. R. Bell on the DCX-2 experiment.
ild
mbinados
o
The source was originally developed for acceleration of a high current
of hydrogen lons to energies of ~ 30 kev and, in fact, ~250 ma can be ob-
tained in d.c. operation. Figure 12 shows a sketch of the source geometry--
generally an electron reflecting electrode 18 used at the far end of the
colum of plasma which diffuses through the anode hole. This electrode,
however, 18 not essential for stable arc operation and may be put at anode
potential with oniy slight changes in the arc characteristics. This emphasizes
the point that other mechanisms besides simple oscillation of the electrons
18 critical for the arc properties observed.
23-
.
C
O
.
.
The hollow, cylindrical cathode 18 made of tungsten and 18 heated
to incandescence by ions flowing out of the plasme.. Accurate density
determinations have not been made, but estimates range from 10hd ions/cc
to over 10 lons/cc in the arc column, based on preliminary probe measure-
ments and geometrical estimates..
We have used a gridded probe to determine the ion energy outside the
main plasma core in a uniform magneic field. The probe and circuit used
18 shown in Fig. 13. The outside grid and entire assembly 18 at wall
potential. The middle grid (G2) 18 biased either positive or negative to
repel either all the lons or all the electrons. The collector potential
is then varied so that the lon or electron energy may be determined and
also any directed energy resulting, for example, from a plasma potential
different from wall potential. .
The results of some of these experiments are shown in the next few
figures. The observed lon distribution is shown in Fig. 14a (with the
electrons repelled). The mean energy of the lons, taken as the 1/e point
of the distribution for no especially profound reason, is 350 ev. Mean
energies approaching 500 ev have been observed on occasion. The electron
distribution shown in Fig. 1lib 18 typical--the mean energy is only 14 ev and
the distribution (although only the tail of the total distribution in the
plasma) 18 nearly exponential. The only parameter which appears to drasti,
cally affect the ion energy 18 the gas feed rate into the source. Figure 15; .
... shows the variation of observed energy with this feed rate.
.
.
.
. .
-
"
.
'
.
.
VI
.
.
in
.
REFERENCES
1. R. A. Gibbons and R. J. Mackin, Jr., Proceedings of Fifth International
Conference on Ionization Phenomena in Gases, Munich, 1961 (North Holland
Publishing Company, Amsterdam, p. 1769).
2. ORNL-3564, Thermomuclear Division Seniannual Progress Report for Period
Ending October 31, 1963. R. A. Gibbons, N. H. Lazar, and T. F. Rayburn,
p. 63.
7
,
.
.
.
13
WW T
YT
T
-
TT
m.
...
.
.
.
.
..
..
.
.
.
LIST OF FIGURES
1. Gas Arc Facility. Spacing Between Mirror and Centers is 27 Inches.
Anode Show 18 No Longer Used.
2. Radia. Density Profile from Deuterium Arc as Function of Coil Current.
Magnetic Center in Midplane is 1.75 Ing.
3. Floating Potential Seen on Two 0.010-Inch Tungsten Wire Probes.
Dual Beam Oscilloscope Triggered on Rise of Upper Trace.
4. Correlation of Ion Current Between Two Probes Arranged on Same Field
Line Spaced 11-1/2 Inches.
5. Correlation of Electron Current Between Two Probes.
6. Ion Current Correlation Between Two Probes Separated Radially by
0.26 cm. Trace Triggered on Upper Trace.
7. Radial Flute Velocity as a Function of Radial Position from Center of Arc.
8. Azimuthal Flute Velocity as a Function of Radial Position from Center of Arc.
9. Dual Trace Triggered from Lower Signal Showing Ion Current to an End Plate
Following the Observation of the Presence of a Flute in the Midplane.
10. Total Ion Current to End Plate for Various Magnetic Fields in the Midplane
as a Function of Arc Current.
11. Total Ion Current to End Plate for Various Gas Feeds as a Function of Arc
Current.
vision
12. Section Through PIG-Type Plasma Source.
13. (a) Schematic Diagram of Gridded Probes Used for Energy Measurements
.
in PIG Geometry Plasma Source.
(b) Tungsten Wire Probe.
14. (a) Repulsion Curve for Ions, Electrons Suppressed, Seen for a Gas Flow
of 0.063.cc/sec.
(b) Electron · Repulsion Curve.
15. "Average" Ion Energy Determined from Repulsion Curves as a Function of
Gas Input.
ni
SY
L
11
"
ws.
20
UNCLASSIFIED
ORNL-LR-DWG. 48989

.
-.
rende
HO XX
9
FACILITY FOR GAS ARC STUDIES
.
SINA
.
.
:
:
.
.
.
SAN
CV:
Z
A.
LT
*
!
::..
UNCLASSIFIED
ORNL-DWG 63-6058

0.92 cm e FOLDING LENGTHS
. 1.01 cm
0.88 cm
-0.90 cm
. 110 amp ARCS
3.0 cc/sec Da
>
'I 24,
2
alg = 4000 amp
ale : 2000 omp
olg = 1000 amp
• 1 : 600 amp
:
*
*
*
.
2 3 4
R (cm)
DC. Ion Current to Langmuir. Probe.
-
-
i
.
..
.
P.
WI
.
2
M
YA
.
NO
A
12:
A
I L
itaire V
l
w
10
mi
West House Silit more
bilans han inte kanten der
.
inimene ei ole to
in their own food, aktiviti
van een tegen
motor post for bonds
,
NRW
in
.
.

9/14/63
Floating potential correlation
5.0 v/cm
Probes in-line
1.46. cm. spacing
.
"""....
..
X
.
.
-
-
n
i
...
"
"
1.
IN
PUN
.
O
TY
.
7
.
.
.
.
2
.
TY
.
ULA
NEMA.
1
PA
.
..
ADAN
IN
44
INYA.
.
mine,
inn ....
.
.......... wwwmore
..
....
...
..

..
L
8/9/63
Spacing 11 1/2 in. + 1 cm
Horizontal out 5/16
(MR 2:1 expect horizontal to 1.050
if originally at 3/4 in.)
41
.
.
-
28
-
-
-
ri
-
-
-
-
A
-
..:-
.
-*
SARA
SI
!
.
P
.
.
..
....
......... ...
...
... ... ...
...
..
TA
..
,
*
.
--
- ,
.
.
.
..

10/7/63
Sync - upper +
Otherwise same as #44
Upper - ion current ,
Lower - electron current ...
Sweep - 10. usec/cm
4000 AIR.
150 A 105 v 3.0 cc/sec
Deuterium arc
-.
•
TUNYAN
1
,
is:
,73
M
GRA
N
D
21
MAN
.
.
..
. .
4
. ....
..
. .
...
..
.
.
.
.
.....
.
.
..
....

che il
l
imite del
Minister is comincino
Mico
.
.
V
.
i
.
i
.
::..:
heit
-,
· 10/10/63
18 2500 A
1136 A
VA 126 V
6 V . .
Sweep 0.5 usec/cm
lon current - 20 v bias.
0.2 v/ cm across 10 12
Probes in-line
~0.26 cm radial separation
: 1.46 cm Z separation
.....
predliwiera contamin
.
.
..
hy
.
10.
411
NA
UNCLASSIFIED
ORNL-DWG 64-4874

.
Bo = 7.0 kilogauss
PARC = 150 amp
VARC = 115 volts
Qo. = 3.0 atm cc /sec
P= 1.5 x 10-6 torr
ARC CROSS SECTION : 1.6 cm2
RADIAL FLUTE VELOCITY ( 105 cm/sec )
OD ARC CORE
A
!
.
2
5
3 4
RADIUS (cm)
E
.
7
C.
...
SA
. +
..
.
ET
UNCLASSIFIED
ORNL-DWG 64-473

AT
AZIMUTHAL FLUTE VELOCITY ( 105 cm
PARC = 150 amp :
Bo = 7 kilogauss
OO
RADIUS (cm)
...1
.
...

...
.
-
-.-
;
*
1
9/14/63
Ion currents
Upper - top QPK 2.0 v/cm at 10 12.
Lower - probe in line 0. I v/cm at 10 12
Sweep 20. usec/cm
4000 A
150 A 105 v 3.0 cc/sec D,
1 .........
.
:
th
1
-.
.
0
7
LI
T
.
VA
.
.
..
UNCLASSIFIED
ORNL-DWG 64-513
01.
07

5.3
6.1
lo D2 INPUT 3.0 ATM CC/SEC
NOS. -BO IN KILOGAUSS
E at ARC "AREA" 1.6 CM 2
-7.0
.
..
3,4 4.4
END PLATE ION CURRENT-AMPERES
80
200
100 120 140 160 180
END. PLATE LION VS IARC
ARC CURRENT-AMPERES
MIL
CT
UNCLASSIFIDO
ORNL-DWG 64-5430

NOS. D2 INPUT IN A
Bo = 7.0 KILOGAUSS
ARC "AREA" 1.6 CM2
END PLATE ION CURRENT-AMPERES
5.0
3.5
3.0
2.54
20-
.241.52
60
80
180
200
100 120 140 160
ARC CURRENT-AMPERES
END PLATE ION VS IARC
srt
.
WIT'S
4m
PLN
"
UNCLASSIFIED
ORNL-DWG 64-4596

1000000
0 0 0 0 0
"nninn
VUVUU
-
-
---
-
innnnnn
Juuuuu
rran
TILL
1. COPPER ANODE
2. INSULATOR
3. TUNGSTEN CATHODE
4. CATHODE HOLDER
5. TUNGSTEN INSERT
6. VACUUM FLANGE
7. GAS SUPPLY LINE
8. "O"RING VACUUM SEAL
9. COOLING WATER LINES
PIG ION SOURCE
figit
LIS
SU
N
.
UNCLASSIFIED
ORNL-DWG 64-4595
SCOPE VERT.
INPUT
VARIAC
TYPE W5
145 VACS|| 115 VAC

10000
100 OR
OOV
Kn
.0047
SCOPE HORZ.
INPUT
i GRIDDED
PROBE
BIAS
SUPPLY
COLLECTOR
FIG.
12" COPPER TUBE
.010" TUNGSTEN WIRÉREMOVABLE
SCREEN

VACUUM SEAL & INSULATOR
FIG. 16
UNCLASSIFIED
PHOTO 65380A

TA
tv,
guany more than
:..
fo
1
min
I = 80
V = 345 v
Q=0.063 cc/sec
.
nia
----
-
-
-890 v
-
+890 v
----

RY
--
--
de bois..
.
..
Home
ZO..1'.
-
-3.71g
I = 7.8 0.
V = 145 v
Q = 0.14 cc/sec
Ilo
'
..
;
-42 v
+42 v
UNCLASSIFIED
ORNL-DWG 64-4594:
400

300
MEAN ION ENERGY
VERSUS H2 FEED
RATE TO SOURCE
4-9-64
Ēi (ev)
0.1
0.2 0.3
Q (cc/sec)
0.4
0.5
0.6
Yig s.
DATE FILMED
12/ 9 /64








:
- LEGAL NOTICE

This report was prepared as an account of Government sponsored work. Neither the United
States, nor the Commission, nor any person acting on behalf of the Commission:
A. Makes any warranty or representation, expressed or implied, with respect to the accu-
racy, completeness, or usefulness of the information contained in this report, or that the use
of any information, apparatus, method, or process disclosed in this report may not infringe
privately owned rights; or
B. Assumes any liabilities with respect to the use of, or for damages resulting from the
use of any information, apparatus, method, or process disclosed in this report.
As used in the above, "person acting on behalf of the Commission” includes any em-
ployee or contractor of the Commission, or employee of such contractor, to the extent that
such employee or contractor of the Commission, or employee of such contractor prepares,
disseminates, or provides access to any information pursuant to his employment or contract
with the Commission, or his employment with such contractor.
LAN
END



n
'
S