(SANITIZED)UNCLASSIFIED SOVIET PAPERS ON INVESTIGATION OF A TOROIDAL DISCHARGE IN A STRONG MAGNETIC FIELD(SANITIZED)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80T00246A007500620002-9
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 63
• Document Creation Date: December 21, 2016
• Document Release Date: April 16, 2009
• Sequence Number: 2
• Content Type: REPORT

Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
Next 1 Page(s) In Document Denied
Iq
Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246A007500620002-9
INVESTIGATION OF A TOROIDAL DISCHARGE IN A STROI
MAGNETIC FIELD
BY:. G. G. DolgovSavelievs V. S. Mukhovatov, V, S, Strelkovn
Ms, N, Shepeleva N. A, Yavlineki
Abstract - Results of investigation of a circular plasma produced-in a
toroidal chamber with a strong magnetic field are presented. No macroscopic
oscillations have been observed under conditions when the Shsfranov-Hruskal
stability condition is satisfied..
The plasma radiation was investigated. in the visible and ultraviolet regions.
It to shown that in a. metallic chamber with a limiting pressure of 1 4. 2 x 1,66 mm
H~ .Most.of.the'emitted energy.is due to impurity ions.
It is well known that a current carrying plasma column is unstable. For
stabilization a magnetic field may be employed. One possibility in this direction
is to impose a longitudinal field uniformly distributed over the cross section of
the chamber. 1) The strength of the longitudinal fields, Ho over the chamber section
must then excee the field strength of the discharge current, H1 a 21/Ca
2)
where a is the column radius. As demonstrated in ref. the following condition
must be'met in order to obtain stable conditions
(I)
where L is the column length. For a circular plasma in a toroidal chamber of
radius R this length is 2 r R. If condition (1) is fulfilled Alistortions with a
not ensure absolute stability. Sole distortions ofthe cylindrical shape of the
wavelength of 9LrA or less are supprefted in a toroid. This method does
Approved For Release 2009/04/16: CIA-RDP80T00246A007500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246A007500620002-9
_2
plasma column remain unstabilized but theoretical considerations show that they
should not lead to displacement of the ring relative to its equilibrium position.
For confinement ofh ring which tends to end under the action of repulsive
electrodynamic forces a conducting screen can be employed in toroidal systems.
This method of stabilization possesses a number of good points. In particular
stability of the discharge column does not depend-on the distribution of the.
current over the cross section and it is not required that the current flow in a
thin-skin layer as the theory developed in ref. I) does. A plasma ring in a strong
longitudinal magnetic field (which does not interact with the chamber walls) can
be heated at a given current I to a much higher temperature than when it is
(2)
(3)
One of the problems of an experimental study of stabilization is the investigation
of stability of a plasma ring as a Auction of K.
in a toroidal Oystem stabilized by a frozen "paramagnetic* field.
Expression (I) can be. rewritten as follows
z~/cam
where K is the stability factor.
Then K Z
2. Experimental arrangement. Method of measurement.
The experiments described below were performed on the "TokCmak" assembly which
is a thick wall copper torus with a large diameter of 225 cm and small diameter of
50 cm. A closed stainless steel chamber with 0.,, Imm walls is located Inside. the
torus. The space between the copper chamber and inner chamber was evacuated to
a pressure of I A 16 The pressure of residual gases. in the working
?a
chamber was I.d A/0 n m 14 . The experimental apparatus was designed ? by V. S.
Vaail#vaky and his colleagues.
Approved For Release 2009/04/16: CIA-RDP80T00246A007500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
-3_
The discharge was produced by the voltage in the torus induced by discharging
a condenser bank through the primary coil. (19 turns) of the air transformer. A
quarter period of-the voltage in various experiments was 1.5 or 3m sec. The
magnetic field in the operating volume was produced by discharging. a o?j
condenser bank charged to a peak voltage of 5 KV through a coil wound on the torus.
The half-period of the magnetic field was 80m sec.
The energizing system has been described in ref. 3)
Measurements were carried out at pressures between 2.1r0'h and 5.10-3 mm Hg.
Working gases employed were deuterium, argon and helium.
The longitudinal magnetic field strength, NO, in our experiments was varied
between 600 and 12000 Oe. The 'initial electric field strength along the chamber
axis could be- changed from 0.2 to 006 V/cm.
In order to produce a circular plasma detached from the walls and in order
to diminish its interaction with the chamber walls two 2mm thick stainless steel
diaphragms were mounted inside the chamber. In our first experiments the apertures
in the diaphragms were 26mm in dUngter and their centers coincided with the axis
magnetic asymetry. It is possible, in particular, that it is due to stray fields
penetrating through the copper tube joints.
The main measurements were carried out frith diaphragms in which the apertures
were displaced relative to the axis in accordance with the.experimentally
of the torus tube.
As Is Well known the equilibir-mt position of a'plasma ring in a torus is
displaced, relative to the aide. A photograph of the diapbrogn removed from the
chamber after several hundred,pulses is shown in fig. I. The plasma ring was
displaced toward the chamber Vail and olightly upward. The shift in the equatorial
plane correapgnds to the calculated value. The upward shift may be due to a slight
Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
-k -
established position of the' plasma ring. The diaphragms were moreover supplied'
with coils. By sending a current through the coils the magnitude and shape of
the longitudinal field in the region adjacent to the diaphragm could be varied.
Two Rogovsky coils were moun"d on each of the two diaphragms. The coils
were enclosed in a thin walls stainless steel tube. The wall thickness of the
latter was chosen in such way as not'to shield oscillations with frequencies up
to 200 Re/sea. One of the coils measured the current Xg flowing through the
diaphragm apertures and the other measured the current v D flowing through.
the whole cross section of the chamber. One other toil measured the total current
flowing through the chambers including that in the wall of the internal chamber.
electron density in the plasma was inferred from microwave (73.000 and 130.000 =eeec)
absorption measurements. The technique applied in'ouch measurements is well known
and is described, for example* in ref..3).
By connecting the coils so that their e.m.f. were in opposite direction we could
measure the current flowing to the diaphragm j?~ Direct measuremment
of the e.m.f. of the coil yielded the current derivative di dt . As usual
the voltage per turn of the dischargechamber,was measured with a loop plated in
the equatorial plane of the.torus.
Besides electric characteristics same other properties of the plasma turn
were also investigated. Streak photographs of the total radiation in the visible
region of the spectrum were made. A feet C model streak-camera was'employed-
for this put-pose. The spectrum of the eradiation !the discharge was studied
with a A C C - 6 vacua spectro,raph which, according to its certificate., can
be used in the wave length range from 2200 to 60 A.'. The time variation of
various spectral lines from deuterium and impurities was studied, with a 3 MP-3
monocbramator with an electron photomultiplier placed behind the exit slit. The
Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246A007500620002-9
3. Results of measurements
The most graphic data were thoee'obtained by streak photography of the
discharge column. In fig..2 are shown streak photographs of the discharge in
argon made at values of No equal to 0.67, 2.35 and 6.e where
z r
The most photograph (fig. 2a) shows that the plasma column'le unstable. The
luminous region reaches the diaphragm about 100 ,A sec after initiation of the
discharge and after 300 /sec the plasma ring leave the boundaries defined by
the diaphragm apertures.
The photograph in fig. 28 is-distinguished by the uniform glow and the absence
.of any significant curvatures in the column. Finally, a feature of fig. 2C is that
not only is the glow uniform but that in less than 600 /sec it begins to diminish
over the whole cross section of the chamber. Intense emission of light in the
upper part of the chamber can be noted in all photographs. This is probably due
to the fact that the holes in the diaphragm do not coincide with the equilibrium
position of the plasma ring.
Photographs of the discharge in dueterium are similar to those presented
above. However the photographs are not as distinct due to the'low intensity of
the emitted.light.
It should be of interest to compare the photographs shown above with other
experimental date.
Oscillograme ?of' the current voltage per turn V1 derivative of
current in gas dx6 A and current in gas ];G measured for argon are presented in
fig. 3.. Oseillograms 3a, 3c and 3e correspond to the conditions under which the
photographs of tlAe plasma ring were made.
Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
- 6'-
There are characteristic points on the current and current derivative
oscillograms which refer to the same instant of time. An inflection in the current
curve can be observed which signifies that the rate of-increase of the
cur-rent is beginning to decrease. At the same time the derivative of the current
d I/d f rapidly decreases and simultaneously large oscillations appear on the
derivative oscillogram. With increase of NO the amplitude of oscillations of the
current derivative becomes smaller. The characteristic points on the current and
current derivative oscillograms concur with the appearance of the current in the
diaphragm. This was established by comparing the oscillograms of currents and
I t~ and also by a more detailed study of the current oscillograms, _Z
The oacillograms obtained for discharges in dueterium were found to be similar.
Some examples are shorn in fig, 4.
The effect of Ko which characterizes,the stability margin, can be illustrated
by oacillogr of..the time variation "Of the deuterium
t~. C+. d-JbJl t i7lS41C315 4w 83 r~7 , 3 r.. 25
ti E ? 9,`8 6 ya. Iral
life :ia~;tnetl-,~,~ ~08cillograme of this type are presented in fig. 5. In order to
facilitate comparison the upper cscillogram which refers to the total current
is also presented. The small current at the beginning of the process is
that flowing in the closed inner chamber.
For go = 2.8 and 1.7, after breakdown of the gas the intensity of line
first increases rapidly and in about 50 J''sec reaches a maxim.t and then begins
to decrease. After about 300-400 the line intensity stops decreasing and
then slightly increases. The state of affairs significantly changes with decrease
of'the stability margin. For go < 1 Dirge oscillations of the intensity are
observed as. long as the current in the gas exists.
Under working conditions involving a strong magnetic field a sufficiently intense
44 line is observed 1 m sec after the gas current ceases. With decrease of Ho
Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
-7-
Oscillograms of the currents .Z-*a & J current derivative
voltage deuterium 2,)9 line intensity and diaphragm current 2 for
two values of Ko are shown in fig. 6a and 6b. The variation of the electrical
conductivity d? during the discharge averaged over the cross section of the
plasma column is shown below the oscillogram on the same time scale. The'electrical
conductivity was calculated from the formula
Z R $ r o dT 9 IU L ' t SE
where the inductance of the plasma ring is assumed to be constant and equal
to L. 47rf? .h where .6 is the radius of the screening wall.
Comparison of the oacillogram of the diaphragm current with other oscillograms
shows that cessation of-decrease of the PA line intensity concurs 'with the'
appearance of the diaphragm. current. Simultaneously the current derivative increases
during a"time of the. order of 30-50/sec. At about the seine time the electrical
conductivity of the:plasms begins to decrease.
t~?+~ ~/ti r, 1r''f ~. .G~1 ..a?, .., .. vs ...:~: .. .r. W .. ... ._ _ ?c5.;: ? c"'r , '~ ;i ,;0 _?`,~..
Oscillograins of the CIII apectr$l. line, (doubly ionized carbon~A o~650
synchronized with the oscillogram of the diaphragm current are given in fig. 7.
The intensity of this line increases with increase of the longitudinal magnetic
field strength, Ho.
The effect of impurities in the operating gas on energy losses in the plasma
was studied by spectral methods. A spectrogram made with &A OCd 6 vas
%
spectrograph is shown in fig. S. The spectrogram was obtained after 150 ..,..pulses
under stable conditions when K. ;;P1 . The photographic film was sensitized
by sodium salicylate, the quantum yield of which is constant throughout the-
wavelength range from 3000 to 581A?'. This type of exposure was made in order to
obtain normal blackening of the most intense lines, L ?C (A =111-5,344")
r
and 1J 76 A J . Only the deuterium and carbon lines were recorded on the
spectrogram. The black lines of various height yield the relative energy'emitted
Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
by a given spectral line. It is assumed that the total energy of all lines
recorded on the spectrogram is equal to one., It can then be seen by comparison
that the deuterium radiation comprises only about 1/15 - 1/20 of the total energy
carried off by light in this spectral region.
The absolute magnitude of radiative energy losses could be roughly estimated
in the following experiments. A-photographic film which was preliminarily
calibrated with aid of a thermopile was mounted In :& tube connected to the discharge
chamber. The film was sensitized with sodium salicylate. A third of the film
was covered with a glass filter, another by a quartz filter and the rest of the
film was not covered at all. By measuring the blackening of the film one could
determine the absolute and relative amount of radiant energy in various parts of
the spectrum. Preliminary measurements showed that about 5% of the energy released
in the plasma is carried off by visible and near ultraviolet light. Most of the
energy is carried off by more penetrating radiaations. An estimate shows that
radiation carries off about 50% of the energy.
Investigation of a stable plasma column produced in a straight porcelain tube
with electrodes showed 5) that most of the energy imparted to the plasma is
radiated by impurity ions. The electron density of the plasma was determined by
probing the. plasma with high frequency radio waves. At a frequency of 130..000 sic/bee
'transmission of waves was observed b00-5Q0/ ' sec after the current started in
the gas. Afterwards transmission'of the''radlo waves ceased Completely. Thus the.
electron density was not smaller than 2.5 x 1014 cm-3. Our measurements moreover
show that the electron density was still large 1 m see after cessation of the
discharge. Zhis is consistent with the behavior of the 40A line during the saw
period of timed
Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
-9-
So. For /(,>I the ratio X continues to decrease but at a much smaller rate.x)
grams. The value of and,r are those which correspond to the mammal value
of the current in the gas. For, KIC I this ratio rapidly drops with growth of
4. Discussion of results and conclusions
A. Stability., The ratio of the diaphragn current to gas current re )
as a function of Ko is plotted in fig. 9. The curve was derived from the oscillo-
x) This dependence was found to be of a slightly different type when the centers
of the diaphragm apertures coincided with the axis of the toroid tube (see the
report of I!. A. Leontovich at Harvell, in-April 1959). A large diaphragm current
is observed for Ko m) and it in.due to the fact that the diaphragm apertures do
not coincide with the stable position of the plasma ring in the torus.
The break lathe curve for a value of No close to unity indicates that there
is a critical current which characterises the boundary between the stable and
unstable states of the plasma ring. The magnitude of the critical current can be
determined as follows
4 C0. /
Icy, R o
For working conditions such that > .,2p i.e. for conditions with K1
the current gas at the beginning of the process is less than the critical current
and hence the gas column should be stable. Subsequently, the gas current begins
to increase and for K e 1 it attains the critical value after which the plasma
column becomes unstable and an appreciable diaphragm current appears'.
The points in fig. 10 represent the gas current at the instant correspond
ing to the appearance of. the diaphragm current for various values of the longitudinal
Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9
magnetic, field, H0.The dotted line represents the calculated critical current.
Besides discharges in deuterium similar functions have been derived for discharges
in helium and argon. Unfortunately, we could not carry out the investigations at
still higher values of Ko since further increase of the longitudinal magnetic
field strength was not feasible. Increase of go by decreasing the gas current
was not advisable as in this case the energy involved per pulse Would be small
and under certain conditions even smaller than that required for dissociation and
ionization of the gas in the chamber. For K. the diaphragm current in our
experiments was smaller than the threshold sensitivity of the measuring device.
Intensity measurements of the p line are also illustrative of the effect
of the longitudinal magnetic field on the stability of?the plasma. For K.
when the discharge current exceeds the critical current, the intensity of line.
is higher than when Ko > t and moreover varies in a disorderly manner. This-D,
may possibly be due to bending of the plasma wing which leads to strong: interaction
between it and the diaphragm and perhaps the chamber Wall. Naturally di-ionization
of the plasma occurs in this case as well as repeated excitation of neutral atoms
entering the discharge. With increase of the, field strength, when Pto a 8 the
plasma Ving?is stable and in h00-500 sec the intensity of. the 46 line
drops. This may be regarded as meaning that ionization has stopped in the'region
in which. the discharge current is flowing.
B. Ionization - the data available at present are not sufficient to permit
one to determine the density of charged particles in a plasma ring.
Microwave probing of plasmas indicate that the electron density in a plasma
ping is at least 1.5 time larger than the initial density of deuterium atoms. It
is possible that this higher electron density is due to ionization of the impurity
or deuterium atoms which enter the plasma viry from the peripheral regions.
Approved For Release 2009/04/16: CIA-RDP80T00246AO07500620002-9 ;
 Approved For Release 2009/04/16: CIA-RDP80T00246A007500620002-9
? - 11 =
The average of the electrical conductivity over the cross section of the
plasma wing continues to increase after the density of the Z)o lines begins to
decrease. A possible explanation of this is that ionization of most of the neutral
gas has stopped. Microwave absorption measurements lead to a similar conclusion.
Growth of the electrical conductivity after cessation of ionization can apparently
be explained by heating of the plasma.
Another indication of a high degree ionization is the existence of intense
spectral lines in the discharge from doubly charged ( C Ilz} and triply charged
( m carbon ions. The ionization potentials are 22.28 and 47.55 v. Thus the
plasma contains electrons with energies which exceed the 'ionization potential of
deuterium by several times.
Finally, one might mention the results of microphotoimetric measurement of
streak photographs of the plasma ring. Analysis of these date with account of
the geometric factor shows that the luminosity in the center of the plasma is
less than that of the outer sheath of the column. In all probability the outer
parts of the plasma which are in.contact with the sheath of?neutral atoms radiate
most of their energy in the visible part of the spectrum. In fact, rough probe
measurements show that the density of charged particles sharply drops with the
distance from the axis of the plasma ring. It follows from these measurements
that the density of charged particles near the chamber walls does not exceed 1Os'
particles/cm3.
C. Electron temperature. As is well known, the electron temperature Te in
a totally ionized plasma can be derived from the magnitude of the'electrical conduct-
ittity d By applying some familiar formulas it can be shown that the mean
electron temperature Te , 7ev if the average of the electron conductivity over the
cross section of the plasma ring is 2-3 x 1014 COSE.
Approved For Release 2009/04/16: CIA-RDP80T00246A007500620002-9
 Approved For Release 2009/04/16: CIA-RDP80T00246A007500620002-9
An estimation of Te on basis of the relative intensity of the impurity lines
(in our case lines CI, CII etc.) yields an electron temperature of about 15 evo
It was mentioned above that coils were mounted on the diaphragms in the
chamber. A current pulse from the condenser bank was sent through these coils.
As a result the diaphragm current decreased and the total gas current IG increased.
Under certain conditions the current increased 1.1 - 1.5 times. The voltage per
turn in this case did not change. It is very probable that the Increase of the
current was related to growth of the cross section of the plasma column throughout
the chamber asa result of deformation of the longitudinal field lines. However,
in the region near-the diaphragms the cross section of,the plasma did not increase
and possibly decreased somewhat,
On basis of the foregoing the following conclusions can be drawn.
1. A plasma ring produced in a toroidal chamber in the presence of a longitud-
tonal magnetic field is stable and unstable for
UK ~b _ rllatri}U\~n`:>it `til:~_.~~ Lido_.u'V{ d~..CS{ '~i0'.' (' 1~..1 (2 .? Y' Ti^o
Ah _o~el ~.._. G~, .,~ , ?IFu:+ 'a'Ca _ - tst?4. ~.'r it i~.
f4: ~a~~s ~cY