SOVIET ATOMIC ENERGY VOLUME 18, NUMBER 4

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP10-02196R000700010004-3
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 174
• Document Creation Date: January 4, 2017
• Document Release Date: September 24, 2013
• Sequence Number: 4
• Publication Date: April 1, 1965
• Content Type: REPORT

Declassified and Approved For Release 2013/09/24: CIA-RDP10-.02196R000700010004-3
,
Volume 18, Number 4
, April, 1965 ?
SOVIET
ATOMIC
ENERGY
ATOMHAFI 3HEP114F1
(ATOMNAYA iNER9IYA)
TRANSLATED FROM RUSSIAN
CONSULTANTS BUREAU
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2913/09/24: CIA-RDP10-02196R000700010004-3
NEW JOURNALS 1965
? in cover-to-cover
translation from Russian
SOVIET JOURNAL OF ORGANIC CHEMISTRY (Zhurnal organischeskoi
A ne? w journal devoted chiefly to synthetic organic into the chemistry of ,natural compounds. The
chemistry. Reports describe new methods of syn- , adjacent problems of biochemistry, and the chem-
thesis, new Classes of organic compounds, results istry and mechanisms of action of physiologically
of inves* tigations of Organic reactions by Methods active substances are-Covered.
of synthetic organic chemistry, and researches Annual subscription (12 issues): $160.00
SOVIET ELECTROCHEMISTRY
."(..lektrokhimiya)
A hew Academy of Sciences journal, coordinating
work on theoretical and applied electrochemistry.
Reports research on problems of electrochemical ;'
kinetics and thermodynamics, 'properties of liquid
and solid systems ,with ionic, conductivity, and
eleCtrochemistry of the rare elements; work" on
fuel cells; electrochemical controlling devices,
electrochemical protection, and the electrochem-
istry of semiconductors.
Annual subscription (12 issues): $125.00
?
CHEMICAL AND PETROLEUM ENGINEERING (Khimicheskoe i neftyanoe mashinostroenie)
A journal devoted to problems of the design and
operation of_equipment for the basic technological
processes of the chemical, petrochemical, petro-
leum refining, paper, and oxygen industries?cal-
culation, design, and construction of apparatus
and machines for precipitating, filtering,' centri-
,
fuging; crystallization, sorption, extraction, distil-
, .lation, contact-catalytic, thermal, pOlymerization
and other processes, and equipment for air sepa-
ration, transferring liquids, and compression and, '
expansion of gas.
Annual subscription (12 issues): $125.00
PROTECTION OF METALS (Zashchita metallov)
A new journal designed to correlate progress in
the various approaches to prevention and control
of corrosion: electrochemical; porcelainizing, ano-
dizing, etc.; electrocrystallization; application of
galvanic coatings; inhibitors and lubricants; re-
ELECTRONIC PROCESSING OF MATERIA
A new journal reporting the latest research and
experience in the practical applications of electric
fields and eletric discharges to the processing of
materials: theoretical bases of electronic process-
ports dealing with the thermodynamics and kinet-
ics of cOrrosion processes, intercrystalline corrosion,
control of high temperature corrosion, and cor-
rosion of semiconductor Materials.
Annual subscriptior (6 issues): $80.00
(Elektronnoya obrabotka'materialov)
ing; dimensional processing with electrical pulses
(electrospark and electroerosion); electrical,meth-
dds' of producing a surface with set physicochem-
ical properties.
Annual subscription (6, issues): $80.00
ENTER YOUR SUBSCRIPTION NOW
@CONSULTANTS BUREAU 221W. 17th St., New York, N.Y. 10011
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700610004-3
Declassified and Approved For Release 2013/09/24 : CIA-RDP10-02196R000700010004-3
ATOMNAYA kNERGIYA
EDITORIAL BOARD
A. I. Alikhanov M. G. Meshcheryakov
A. A. Bochvar M. D. Mill ionshchikov
N. A..Dollezhal' (Editor-in-Chief)
V. S. Fursov P. N. Palei
I. N. Golovin V. B. Shevchenko
V. F. Kalinin D. L. Simonenko
N. A. Kolokol'tsov V. I. Smirnov
(Assistant Editor) A. P. Vinogradov
A. K. Krasin N. A. Vlasov
A. I. Leipunskii (Assistant Editor)
V. V. Matveev
SOVIET ATOMIC
ENERGY
A translation of ATOMNAYA ENERGIYA,
a publication of the Academy of Sciences of the USSR
@ 1966 CONSULTANTS BUREAU, A DIVISION OF PLENUM PUBLISHING
CORPORATION, 227 west 17th Street, New York, N. Y. 10011
Vol. 18, No. 4 April, 1965
CONTENTS
RUSS.
PAGE PAGE
Interaction of Modulated Heavy-Current Electron Pulse Beams with Plasma in a Longitudinal
Magnetic Field?A. K. Berezin, G. P. Berezina, L. I. Bolotin, Yu. M. Lyapkalo,
and Ya. B. Fainberg
407
315
Interaction of a Straight Plasma Pinch with a Varying Magnetic Field of Quadrupole
Configuration?D. V. Orlinskii
- 415
323
Experimental Study of Plasma Injection into a Programmed Magnetic Field?O. I. Fedyanin.
422
329
A Pulsed Neutron Generator?G. E. Murguliya and A. A. Plyutto
428
336
Feasibility of Using Thorium in Fast Power Reactors?A. I. Leipunskii, 0. D. Kazachkovskii,
S. B. Shikhov, and V. M. Murogov
434
342
Transmission of Neutron Radiation from a Reactor through Hydrogen-Free Media
?S. G. Tsypin, B. I. Sinitsyn, and V. K. Daruga
442
350
Interrelation between the Grain Orientation and the Radiation Growth of Uranium Rods
?V. E. Ivanov, V. F. Zelenskii, V. V. Kunchenko, N. M. Roenko, A. I. Stukalov,
M. A. Vorob'ev, and A. V. Azarenko
451
357
Center
In Memoriam: Andrei Vladimirovich Lebedinskii
456
insert
Radiation Stability of Vitrified Radioactive Preparations?F. S. Dukhovich and V. V. Kulichenko .
459
361
Concentration of Water Samples for Determining the Tritium Content?Ya. D. Zervenskii,
D. A. Nikolaev, V. S. Tatarinskii, and V. A. Shalygin
466
367
Uranium and Arsenic in the Hydrothermal Process?V. E. Boitsov and T. M. Kaikova
v Method for Calculating the Radioactive Impurity Concentration in the Water and the Bottorri
473
373
Layer of Stagnant Reservoirs?F. Ya. Rovinskii
480
379
Rules for Depositing (Storing) Articles
486
383
ABSTRACTS OF DEPOSITED ARTICLES
Intensive Muon Beams in the OIYaI Synchrocyclotron?Yu. M. Grashin, B. A. Dolgoshein,
V. G. Kirillov-Ugryumov, A. A. Kropin, V. S. Roganov, A. V. Samoilov, andS.V.Somov .
487
384
Conversion of the 1.5-m Cyclotron for the Acceleration of Multicharge Ions?V. V. Batyunya,
Pai Fu-wei, G. N. Vyalov, B. A. Zager, and A. F. Linev
488
384
Decreasing the Energy of Beams of Multi-Charge Ions on the 1.5-Meter Cyclotron
?R. Ts. Oganesyan, G. Indreash, and B. A. Zager
-489
385
Investigation of How the Aging of Co6? Radioactive Impurities on St. 3 Affects
the Efficiency of Chemical and Ultrasonic Deactivation Methods?S. M. Kochergin
and S. K. Moiseenko
490
385
Annual Subscription: $95
Single Issue: $30
Single Article: $15
All rights reserved. No article contained herein may be reproduced for any purpose what-
soever without permission of the publisher. Permission may be obtained from Consultants
Bureau Enterprises, Inc., 227 West 17th Street, New York, N.Y. 10011, U.S.A.
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
CONTENTS (continued)
RUSS.
PAGE PAGE
Space-Energy Distribution of Neutrons in an Infinite Absorbing Medium?I. A.Kozachok . . . .
492
386
REVIEWS OF GENEVA PAPERS
Thermoelectric and Thermoemissive Converters?N. N. Ponomarev-Stepnoi
494
387
Fast Reactors?O. D. Kazachkovskii
499
390
Constructional Metals and Alloys for Nuclear Reactors?A. S. Zaimovskii ...... . . . .
505
395
Reserves of Nuclear Materials, Prospecting, Extraction, and Treatment of Uranium Ore
?A. A. Zadikyan
512
400
LETTERS TO THE EDITOR
Contribution to the Theory of Longitudinal Focusing of Radiation-Accelerated
Charged-Particle Bunches?V. V. Yankov
515
402
Calculation of the Mean Square of the Recoil Nucleus Momentum in Evaporation
?F. P. Denisov and V. P. Milovanov
517
403
Spectra of Fast Neutrons in Heavy Media and Water?D. L. Broder, A. S. Zhilkin,
and A. A. Kutuzov
519
404
The Weight-Effectiveness Index of Two-Component Materials Used for Shielding Against
Neutrons and Gamma Rays?G. A. Lisochkin and F. A. Predovskii
524
408
Angular Distribution of Fast Neutrons Scattered by Medium and Heavy Nuclei
?A. G. Guseinov, M. N. Nikolaev, A. G. Dovbenko, V. E. Kolesov, and V. N. Morozov.
526
409
Numerical Calculations on the Penetration of y -Quanta through Matter?V. S. Galishev ? . . .
533
415
Spatial Energy Distribution and Dose Rate of y-Radiation from Unidirectional
and Isotropic Co6? Sources at the Ground?Air Interface?S. M. Ermakov,
B. A. Efimenko, V. G. Zolotukhin, Yu. A. Kolevatov, and V. I. Kukhtevich
534
416
A
Induced y-Activity in Polyethylene as a Result of Neutron Irradiation
?N. A. Dubinskaya, A. Yu. Lyul', and L. L. Pelekis
538
418
Asymptotic Solution of the Kinetic Equation and the Diffusion Characteristics
?Ya. I. Granovskii and A. A. Kostritsa
540
419
Rod-Liquid Interaction in Control and Protection Systems?R. R. Ionaitis
546
422
Improving the Accuracy of the Radiometric Analysis of Multicomponent Specimens
?S. I. Babichenko, L. N. Krylov, V. S. Raikov, and A. P. Utekhin
552
426
Heat Generation in Highly Radioactive Solid Preparations in Connection with the Problem
of their Burial and Utilization?P. V. Zimakov, B. S. Kolychev, V. V. Kulichenko,
and Yu. P. Martynov
556
428
SCIENCE AND ENGINEERING NEWS
Conference on Experimental Research Reactor Techniques?A. M. Demidov
561
432
All-Union Seminar on Industrial y-Ray Flaw Detection
564
432
New Public Health Regulations Governing the Design and Operation of High-Level Isotope
Facilities?V. I. Sinitsyn
566
435
REVIEWS
New Books
569
437
The Russian press date (podpisano k pechati) of this issue was 4/14/1965.
Publication therefore did not occur prior to this date, but must be assumed
to have taken place reasonably soon thereafter.
Publisher
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
INTERACTION OF MODULATED HEAVY-CURRENT ELECTRON-PULSE
BEAMS WITH PLASMA IN A LONGITUDINAL MAGNETIC FIELD
(UDC 533.9)
A. K. Berezin, G. P. Berezina, L. I. Bolotin,
Yu. M. Lyapkalo, and Ya. B. Fainberg
Translated from Atomnaya Energiya, Vol. 18, No. 4,
pp. 315-322, April, 1965
Original article submitted July 1, 1964
Results of experiments on the interaction of modulated heavy-current electron-pulse beams with plasma
in a longitudinal magnetic field are presented. The plasma is formed by the beam itself. It is shown
that under certain conditions the modulated electron beam interacts much more strongly with the plasma
than an unmodulated beam. Longitudinal waves with a considerably greater electric field strength (some
seven times) than in the absence of initial modulation are excited in the beam and the plasma. An ex-
planation of the results is offered.
As was shown theoretically in [1, 21! and experimentally in [3-5], one of the causes of the effective interaction
of an initially unmodulatedelectron beam with plasma is automodulation, leading to coherent interaction of the
beam with the plasma. The depth of modulation is determined by the field strength of the excited oscillations, and
the field strength depends substantially on the initial perturbation amplitude, which for originally unmodulated beams
is determined by comparatively small fluctuation fields. Hence automodulation becomes appreciable only at the
end of the interaction region, where the field strengths become considerable owing to exponential growth.
Thus we may expect that, if the beams pass through a region of modulating field at the entrance into the in-
teraction region, the effectiveness of the interaction between electron beam and plasma will rise considerably [2,
6-8]. This arises from several causes. Firstly, the amplitude of the initial field intensities rises considerably as com-
pared with the fluctuation fields, and, secondly, under certain conditions, when the beam passes through an external
region of hf field, particle bunching develops, reinforcing the effects of coherence.
It must be noted that, for a very large depth of modulation, in the long run, oscillations increasing neither in
time nor space are established in the beam- plasma?system.
The interaction of modulated charged-particle beams with plasma may be used for heating the plasma, cap-
turing particles in traps, developing new methods of accelerating charged particles, and amplifying and regenerating
hf oscillations [2].
The exponential growth of instability is caused, as we know, by the fact that the fields arising on instability
increase the degree of bunching of the particles, and this in turn amplifies the field. Deep modulation at a given
frequency leads to a situation where the degree of bunching no longer changes with further increase of field strength,
and the growth of the field ceases. Preliminary modulation disrupts the bunching of particles at frequencies differing
from the modulation frequency, and leads to the collapse of the whole instability spectrum. Here we must watch that
the coherency conditions a< Xpi, where a = length of particle bunch, Xpi= 27rvf /coo (vf is the phase velocity of the
wave, wo is the electronic Langmuir plasma frequency) for the excitation of characteristic plasma frequencies are
not satisfied. For this it is sufficient that the wavelength of the modulation Xm= a/2. It should be noted that, in re-
moving the ordinary instabilities by modulating the beam, there may arise other instabilities engendered by para-
metric resonances, but, since the breadth of parametric resonances is small, the considerable inhomogeneities and
collisions in an actual plasma lead to the collapse of these instabilities.
407
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
be
0
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
2
IJ [>7] m. El vJ [7>H(p) for 10=1400 A: for po =0.06 mm Hg: 0, 0)for U0=0 and 12 kV respectively,
ti= 54sec; A, A)for U0=0 and 12 kV respectively, t2= lOptsec; 0, IN)for U0=0 and 12 kV respectively,
t3=50 psec; x) p0=2?10-6 mm Hg.
the hf field is larger (on account of the field of induction currents) than when the pinch is absent, and hence the
signal from the probe increases. When, however, as a result of the spreading of the plasma, the probe falls inside
the pinch, the signal sharply diminishes.
Typical distribution curves for the hf field along the tube radius are shown in Fig. 4 (Hr between the rods).
For convenience of comparison, all the curves are referred to the same value of current through the rod (7 =830 A).
Soon after the appearance of the plasma pinch (t1=51isec), the field in the central part of the tube becomes con-
siderably larger than in the case when 10=0 (broken curve). At the following instant (t2= lOusec), however, owing
to the redistribution of the "collapsing" plasma, the difference between the distribution curves for Uo= 0 and 12 kV
diminishes; subsequently (e.g., at t3=50?sec) the magnetic field for U0=12 kV passes out of the plasma to a greater
extent than for U0= 0. This is apparently because the longitudinal current through the plasma causes an increase in
both the degree of ionization and the temperature. A similar type of variation in field distribution occurs for other
discharge conditions as well. The plasma conductivity a estimated from the thickness of the skin layer (as seen in
Fig. 4, the skin layer is about 1 cm thick) is approximately 2 ? 1013 cgs. If we consider that collisions of electrons
with neutral atoms predominate, then for a =2 ? 1013 the electron concentration tie 2. 1013 cm-3 and the degree of
ionization is < 1%.
The distribution of the magnetic field Ficp0 of the longitudinal current 10 along the tube radius was measured by
a magnetic probe differing from that used in measuring the hf field only by the larger number of turns in the meas-
uring coil. The probe sensitivity was ?2.5 k0e/V.
From the measured distribution of Hcoo(r), the distribution of current through the gas over the tube cross section,
jo(r), was calculated. Here it was assumed that the axial symmetry of the discharge-current distribution was pre-
served while the Hcoo(t) oscillograms remained satisfactorily reproducible. Figure 5 shows curves of the relative dis-
tribution of j0(r) for various instants of time from the onset of the current pulse through the gas. As seen from the
418
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
?
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
rel. units
7
6 t=2 ?sec 2
5
4
3
1= 800,A
t'-6 //sec
4
3
2
4
tO Msec
?
Fig. 5. Distribution curves of direct-current density j0(r) through the gas at various instants of
time from the onset of the current; a) hf field absent, J =0, for t> 6 iisec the. symmetry of the
discharge is disrupted; b) discharge between electrodes begins after ignition of hf discharge, cur-
rent flows the whole time around the walls, the pinch is absent; c) hf field applied,to formed
pinch 0 to 6 Msec after onset of direct discharge; current undergoes redistribution out of the cen-
tral part of the discharge tube into the peripheral region.
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Fig. 6. Oscillograms of longitudinal flux (I., for Po =
0.05 mm Hg, U0=15 kV (Iomax= 1.6 kA), time markers
every 201isec. Scale 1000 Mx/cm: a) J =0; b) J 0.
j0(r) curves of Fig. 5a, for J =0 the breakdown takes place
on the chamber axis, after which the pinch gradually
broadens. Some 64sec after the beginning of the dis-
charge, the symmetry of i0(r)is broken. If the break-
down occurs after the hf discharge is ignited, the current
appears at once at the walls of the chamber (see Fig. 5b)
and, growing, continues there the whole time. Finally,
when the hf field is switched on after the breakdown (see
Fig. Sc), the symmetry of the j0(r) distribution relative
to the axis is preserved, but the current is gradually re-
distributed. The proportion of current flowing along the
axis diminishes, and that near the tube walls increases.
Thus the curves show, first, that the rate of current in-
crease (dI0/dt.:J108 A/sec) is insufficient to displace the
discharge at the initial stage to the periphery of the tube
with subsequent constriction and skimming of the neutral
gas to the axis (thickness of the skin layer, determined
by the derivative dIo/dt, exceeds tube radius) [3]; and,
secondly, that the value of current 10 bounded by the
stability condition (1"), i.e., by the value of J, is insuf-
ficient for constriction in the case when the straight-
forward discharge begins after the high-frequency one
and the current Io flows around the tube walls. Unfortun-
ately, an increase in tube diameter leads to a fall in
aH/ar, and an increase in 10 demands a simultaneous
increase in J, i.e., in the hf-generator power.2
In the bending of the compressed plasma pinch, the magnetic field of the current flowing through it is dis-
torted, and a longitudinal component appears. From the change in the size of this component we may also judge
the stabilizing effect of the hf field.
In order to record the flux of the longitudinal magnetic field, a 20-turn coil was wound on the discharge tube.
The signal from the coil was integrated in its self-inductance and shunting resistance (time constant ?0.1 msec). In
the oscillograms of Fig. 6 we see that, on applying the hf field, the signal from the coil sharply diminishes. The
maximum flux recorded by the coil without the hf field was ?250 Mx, which for a 4-cm diameter of the plasma-
pinch spiral corresponds to H=20 Oe. However, despite the small value of the magnetic flux, the oscillograms
indicate the suitability of this method of recording the instability of the plasma pinch in such experiments.
The results of our preliminary study of the interaction between a straight, self-constricted discharge and a hf
magnetic field of quadrupole configuration indicate the stabilizing effect of the hf field (see Fig. 1, 2, and 6), al-
though there remain some doubts connected with the fact that the hf field itself (independently of its configuration)
may affect the course of the straight discharge. It is more expedient, however, to check this for large currents Io
and correspondingly large.
We see from the photographs of the discharge that, when the hf field is absent, the plasma pinch coils into a
spiral gradually. In order to stabilize the perturbation with the minimum observable wavelength (X ;17.5 cm), the
current through the gas must not exceed 200 A. In our experiments, a stable, though' not long-lasting, torn-off plas-
ma pinch existed for 3000 A, while, according to the stability condition, perturbations could only be suppressed
for X 60 cm. This gives rise to the idea that, on stabilization of the long-wave perturbations, the development of
2For Po= 0.02 mm Hg and electron temperature T =15 eV, the current 10 necessary for constriction of the plasma must
exceed ?)4c2NkT 25 kA, which corresponds (according to the stability criterion for X= 60 cm) to current J 5 kA.
To obtain this current, we must raise the generator power some 40 times. Special measures taken to match the gen-
erator with the low-Q "plasma" load raised the power absorbed by the plasma to 2 MW and the current through the
rods to J =2.5 kA. It was not possible, however, to achieve the desired change in the processes taking place. For a
proportionate increase in Io, the qualitative picture of the discharge remained as before.
420
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
short-wave perturbations is also hindered. In the case of the stabilization of a plasma pinch by a steady magnetic
field, this effect cannot be detected, since, on satisfying the stabilization condition for large wavelengths, that for
the stabilization of short-wave perturbations is automatically satisfied.
Both the photographs of the discharge and the current density distribution curves WO along the radius of the
discharge tube indicate the brief duration of the stable pinch. For a more prolonged containment of the plasma, it
is (in our view) necessary to raise the current through the gas, and hence also that through the rods, substantially.
The author considers it his duty to express thanks to S. M. Osovets and Yu. F. Nasedkin for advice and discus-
sion of the results, and also V. M. Atamanov, G. M. Balkov, V. P. Mokshantsev, V. G. Nikolaevskii, and A. Ya. Cher-
moshentsev for assistance in the work.
LITERATURE CITED
1. S. M. Osovets, ZhtTF, 39, 311 (1960).
2. M. L. Levin and M. S. Rabinovich, ZhTF, 33, 164 (1963).
3. N. A. Borzunov, D. V. Orlinskii, and S. M. Osovets, "Atomnaya energiya," 4, 149 (1958).
All abbreviations of periodicals in the above bibliography are letter-by-letter transliter-
ations of the abbreviations as given in the original Russian journal. Some or all of this peri-
odical literature may well be available in English translation. A complete list of the cover-to.
cover English translations appears at the back of this issue.
421
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
EXPERIMENTAL STUDY OF PLASMA INJECTION
INTO A PROGRAMMED MAGNETIC FIELD
(UDC 533.9)
0. I. Fedyanin
Translated from Atomnaya Energiya, Vol. 18, No. 4,
pp. 329-335, April, 1965
Original article submitted April 22, 1964; revision submitted December 21, 1964
The authors give experimental results on the capture of an injected plasma by a rising magnetic field.
It is shown that a long-lived plasma pinch is formed, localized at a distance from the wall of the vacu-
um chamber: the plasma density is ?102 cm-3.
In the study of high-temperature plasmas, one of the main problems is how to fill magnetic traps with plasma.
Most work on plasma injection into traps is connected with plasma?plasma and plasma?magnetic field interaction
and with interaction between beam and plasma in the magnetic field, etc. [1-4]. In this paper.we discuss the pos-
sibility of using irreversible processes in a plasma compressed by a rapidly rising magnetic field to secure injection
into the magnetic trap.
Experimental Method
The principal purpose of this experiment was to study the possibility of lateral injection into a programmed
magnetic field.
Figure 1 shows the topography of the magnetic field at various times. The field is uniform and quasistationary
and has "wells" in time and space. The plasma is injected when the field on the axis is zero (see Fig. la). The
field is then rapidly increased so that the lines of force become parallel and the plasma is compressed (see Fig. lb
and c).
In Fig. 2 the magnetic field is plotted against time in the center of the system. The maximum intensity is
?4000 0e, the duration of zero field is 20 ji sec,and the rate of increase of the field (dH/dt)max is1.7 ? 109 0e/sec;
the field is increased according to the relation H =H0(1?et/T), where T .-:.:(2.5-3)? 10-6 sec.
422
Fig. 1. Topography of magnetic field.
The magnetic system contains two parts: a solenoid giving
a quasistationary field (half-period 4 msec) and two compensating
coils which modulate the field in time and space in the injection
region. The plasma is injected from two diametrically arranged
injectors at time t,3 (see Fig. 2). The plasma sources were elec-
trode injectors with the discharge on the surface of a plastic die-
lectric. The vacuum chamber (positioned within the solenoid) is
a glass tube of diameter 12 cm, length 220 cm. The central part
has two ground joints for the injector assembly. The system is
evacuated to 5 ? 10-6-10-5 torr.
In studying the injection, we used the following measuring
devices: a local electrically shielded probe, a luminescent screen
[5], hf probes (frequencies 3000 and 10000 Mc)[6], and annular elec-
tric and magnetic probes for measuring the current and plasma
discharge. In some experiments, where detailed localization of
the measurements was not needed, the local probe (giving the
plasma current density at a point) was replaced by a flat probe
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
a
composed of seven concentric collector rings (ring probes).
The other methods used were the standard ones employed
in laboratory studies of plasma physics.
The plasma-current configuration escaping from the
magnetic trap created by the programmed field was in-
vestigated mainly by means of the luminescent probe.
Figure 3a gives screen figures showing the electron image
of the stable plasma-pinch configuration. It is seen that
the plasma pinch, which is isolated from the walls of the
vacuum chamber, lasts for a long time (t 1()-4 sec); its
width is 4-6 cm.
ti
0
Fig. 2. Time dependence
field; b) compensating field;
in injection region.
of field.
c)
4
a) Quasistationary
total programmed field
These results were obtained by a long series of measurements, intended to determine the effect of physical
conditions on the formation of a plasma pinch. The criterion of optimum conditions was the presence of a localized
plasma current on the axis. The following conditions were found to be necessary for a stable pinch:
1. The amplitude of the compensating field (see Fig. 2) must be equal, at least on the axis, to the axial field;
the allowable deviation is 5-10%.
2. The working conditions may be either periodic or aperiodic; for the first current half-wave ? must be par-
allel to H (where H is the direction of the axial uniform magnetic field, ? the magnetic moment of the
injector current).
3. The .two sources must work identically (injecting the same number of particles).
4. The interval between the moment of injection and the beginning of increase in the magnetic field (t4?t3)
must be 3-6 ?sec.
If these conditions are satisfied, the configuration obtained resembles that shown in Fig. 3a. Although a lo-
calized pinch is observed when the conditions are broken, the screen figures are then nonrepeating (see Fig. 3b),
showing that there is no stable plasma configuration.
Local measurements of the plasma densities in two cross sections were made by means of screened electric
probes. Figure 4 plots the current density distribution at distances z= 25 and 60 cm from the injector axis. The
curves show that there is a plasma pinch which lasts for 80-100 ?sec.
From probe measurements it was also established that, if the plasma injectors work aperiodically (At 10-.6 sec),
the plasma stream obtained by compressing the increasing field has a double structure: a rapid low-density leader
is first formed (v r--,107 cm/sec, n p.-.1011 cm-3), followed by movement of the main plasma body (v = 5 ? 106 cm/sec,
n 1012 cm-3). When the sources are changed to periodic working [At ,---..1(5-6)? 10-6 sec], the plasma has a different
structure.
To measure the total number of trapped particles and obtain additional information on the radial plasma dis-
tribution, we investigated the pinch parameters by means of ring-shaped electric probes. The data given here refer
to stable imaging of the localized pinch on the luminescent screen. The probe was 44 cm from the axis of the in-
jectors. Figure 5 shows oscillograms for a ring-shaped probe. From similar oscillograms, we can plot the plasma_
currentdistribution on the ring at various times (Fig. 6a) and, assuming axial symmetry of the plasma pinch, the
density distribution, by converting nv to the plasma density (see Fig. 6b).
From, the measurements with ring-shaped probes we can conclude that:
1. It is possible to make a differential measurement of the plasma current with separate rings, as the arith-
metical sum of the signals from the separate rings agrees with the electrical sum, i.e., with the total cur-
rent on all the rings connected together.
2. The total number of particles trapped by injection into the programmed magnetic field is N 1015.
3. For n=1011 cm-3 the rate of density expansion is v =8 ? 106 cm/sec; for n=1012 cm-3, v 2.5? 106 cm/sec.
4. A satisfactorily localized plasma pinch is obtained: when the radius is altered to 2.5 cm, the density de-
creases about tenfold.
423
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
-28:6 sec
_ 37 p-iec
Fig. 3. Screen figures of plasma pinch (H=2000 Oe, Hcomp= 1400 Oe): a and b are plasma
configurations, respectively; the times shown are measured from the beginning of the rise of the
compressing magnetic field.
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
108
120
a
------s- 37
50
87
100
125
Fig. 4. Relative distributions of plasma-current den-
sities at 25 (a) and 60 cm (b) from the injector axis,
at various times.
5. The plasma caught in the magnetic field is dis-
tributed practically without loss along the lines
of force. Measurements of the total current, ?
made at various distances from the injection
site, give a constant value within the region of
existence of the magnetic field.
A high-frequency probe was used to afford an ad-
ditional method, not associated with measurement of the
plasma current, for measuring the resonance density of
the plasma.
Fig. 5. Oscillograms of probe current to ring-shaped
probe; a) compensating field; b) total signal for all
rings; c) signal to first ring (diameter 1.1 cm, area
1 cm2); d) signal to third ring (mean diameter 2.7 cm,
area 3 cm2, amplification factor 2.8); e) signal to
seventh ring (mean diameter 9 cm, area 40 cm2, am-
plification factor 7.6); f) calibration signal, f =100 kc.
The hf probe was placed 60 cm from the injector axis. Figure 7 gives oscillograms with the probe working at
wavelengths X=10 cm (n0=1011 cm-3) and X=3 cm (no 1012 cm-3). In the region R4 cm, the density is n= 10u
cm-3, and in the region R 2.5 cm, n 1012 cm-3; the plasma with density 1011 cm-3 lasts much longer (-10-4 sec).
These results satisfactorily confirm the idea of localization of the plasma pinch.
To study the influence of electrodynamic forces on the mechanism of capture of the plasma by a rapidly in-
creasing magnetic field, we made measurements of the induction current (magnitude < 0.1 kA). The compressed
plasma was found to have little disturbing effect on the outer magnetic field.
425
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
s .? t''2$11.see ? ?
1 2 3R, crn 0 ;
Fig. 67 Plasma-current distribution over chamber radius
at various times.. a) Current distribution; b) mean plas-
ma-current density distribution (the characteristic den-
sities are given below. the curves).
Discussion of Results
It has been shown that lateral plasma injection is
possible into a programmed magnetic field. So far we
have obtained a plasma pinch of diameter 5-6 cm, iso-
lated from the walls of the vacuum chamber for 80 ?sec.
With quasistationary field strength H kOe the
maximum plasma density is 3.1012 cm-3. Preliminary
measurements of the capture coefficient showed that this
injection method gives adequate efficiency. The sources
inject ..(0.7-1)? 1016 particles, while the percentage of
these captured is ?20-50%.
The plasma current travels 100 cin along the mag-
netic field with practically no losses.
The mechanism of plasma capture can be quali-
latively represented as follows. The injected plasma fills
the magnetic trap, the configuration of which is given in
Fig. 1, and partly begins to flow along the lines of force.
For slow compression, with dH/dt? 8- 108 0e/sec, the
capture efficiency (i.e., the number of captured particles)
is 15-20 times less. After the trap is filled, compression
begins. The compressed plasma forms a pinch, which
spreads symmetrically along the axis away from the in-
jection region.
The initial distribution of the plasma density within
the trap is of great importance in the mechanism of cap-
ture. The presence of a residual non-compensated magnetic field in the region of distribution of the plasma sources
leads to interaction between the injected plasma and these fields, and this may disturb the uniformity of filling. If
the zero-field region is not symmetrically filled, the plasma pinch becomes unstable.
The compression of the plasma by the increasing magnetic field can be explained either by means of an MHD
model or by considering the motion of individual particles in the drift approximation.
Using an idealized conception of the motion of a cylindrical layer of plasma under the action of a linearly
increasing outer field [71, we can estimate the order of magnitude of the characteristic compression time. The plas-
ma layer reaches the axis after a time
2n T2
t = 147483-t-r'6
2i/67H
where ro is the initial radius of the plasma, po the initial density, H the maximum field; dH/dt=H/r; H =Hze?Hzi
=47rj is the drop in the magnetic field at the plasma-vacuum boundary. For our apparatus, the compression time,
estimated from the above formula, was t2 10-6 sec. Thus, even at relatively low plasma currents (j< 100 A), this
very approximate model gives quite acceptable values for the duration of the motion.
Owing to unequal filling of the sources, lack of synchronization at the moment of action, or residual fields in
the injection region, the trap may be nonuniformly filled with plasma: the pinch is then formed at a distance from
the axis and later moves to the wall of the vacuum chamber. As the field in the injection region increases with dis-
tance from the trap axis, the force acting on the plasma is F = grad (pH), where g is the magnetic moment of the
plasma-current vortex; the presence of this force is a possible reason for the ejection of the plasma to the wall.
The compression of the plasma by the increasing magnetic field can also be explained by considering the mo-
tion of individual particles. The plasma, finding itself within the force tube (which in our case has the shape of a
concentric funnel), becomes deformed as the dimensions of this tube alter. The displacement of the plasma pinch
is due to drift of the plasma in the curvilinear field, which is completely removed when the force tube is symmet- ?
rically filled..
426
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
100 kc
Fig. 7. Oscillograms of potential
length X=10 cm, n0=1011
The author wishes to thank M.
the results, V. F. Pis'menko and E. N
collective under the leadership of V.
= 4 cm
of high-frequency probe. a) Wavelength X=3 cm, no= 1012 cm3; b) wave-
cm-3; Ar is the distance between the probe and the chamber center.
S. Rabinovich and I. S: Shpigel' for valuable advice and useful discussion of
. Sobolev for help in the investigations, and N. V. Perov and the mechanics'
P. Solov'ev for help in building the apparatus.
LITERATURE CITED
1. J. Tuck, Phys. Rev. Lett., 3, 313 (1959).
2. S. Yu. Luk'yanov, I. M. Podgornyi, and V. N. Sumarokov, ZhETF, 40, 448 (1961).
3. Eubank and Wilkerson, Phys. Fluids, 4, 1407 (1961).
4. M. A. Ivanovskii and G. M. Batanov, ZhTF, 35, 66 (1965).
5. M. A. Ivanovskii et al., In collection: "Plasma Diagnostics" [in Russian], Gosatomizdat, Moscow (1963), p. 263.
6. B. P. Kononov, A. A. Rukhadze, and G. V. Solodukhov, ZhTF, 31, 565 (1961).
7. L. A. Artsimovich, Controlled Thermonuclear Reactions [in Russian], Fizmatgiz, Moscow (1961).
427
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
A PULSED NEUTRON GENERATOR
(UDC 621.373.2:539.172.84)
G. E. Murguliya, and A. A. Plyutto
Translated from Atomnaya Energiya, Vol. 18, NO. 4,
pp. 336-.342, April, 1965
Original article submitted February 8, 1964
The authors give the results of physical investigations on a'neutron generator with spark-ion source,
. yielding pulsed-neutron.fluxes from the reactions D+ D and D+ T, with mean yields ?7-106 and
per pulse respectively. The total pulse length is ?100-25011sec and the potential across the
accelerator gap is ?110_1(V.
A number of models of pulsed-neutron generators with spark ion sources have recently been developed [1, 21.
These are used in nuclear physics.and its applications. However, they have not so far been subjected to detailed
physical investigation. The working characteristics of a neutron generator are, of course, determined by the param-
eters of the ion beam (composition, current strength and time characteristics), the magnitude. and stability of the '
accelerating voltage, and the target properties. We have studied the influence of these factors on the?wOrking of
one type of neutron generator.
Description of Apparatus and Experimental Method
We first developed a suitable method and apparatus for the problem in hand.
The Construction principles of the neutron generator (Fig. la) are similar to those of the model described in
[3]. It differs from the latter in that a grid 3 is positioned in the selection and accelerator gap,and is connected to
a capacitative potential divider (C1 and C2) and screen 7. This grid forms an intermediate electrode; when the gen-
erator is working it acquires a floating potential and prevents electrical breakdown. Screen 7 retains secondary elec-
trons emitted from the target and protects the sides of porcelain chamber 4 from the deposition of a conducting layer,
thus avoiding breakdown along the chamber sides. These modifications made it possible to increase the diameter of
hole 2 in the limiting electrode to 16 mm, raising the ion current three or four times and increasing the accelerating
potential to VG= 100-110 kV.
By introducing a large inductance, L = 300-6004H, into the spark circuit, it was possible to stabilize the evap-
oration of the working substance and increase the neutron yield. The working substance was evaporated from com-
bined-electrode 9'and burnt out at a depth of ?5 mm, preserving the metal walls of the electrode channel. The
secondary electron beam burns the working substance in the spark gap, which is situated on the axis of the spark
source; for this reason, we used only the auxiliary gaps of the standard source.
To study the ion-beam composition and energy scatter, we used a mass spectrograph and the method of parab-
olae [4] (see Fig. lb). The narrow ion beam was passed through a 1-mm-diameter hole in the target 6 and two col-
limating diaphragms 13, and then traversed a region of electrical and magnetic fields 15, where it was analyzed; it
finally fell on a photographic plate 14, fluorescent screen, or Type II scintillation. counter. By photography with MP
plates, we studied the ion-beam composition averaged over ?100 pulses. The photographs were subjected to photo-
metry (making a correction for the blackening produced by ions of different masses), and this enabled us to form
some idea of the quantitative beam composition. When necessary, visual observations were made on the screen for
qualitative control of the beam composition in each individual pulse. The time characteristics of the analyzed
beam were studied by means of a Type H scintillation counter [CsJ(Ti.), FEU-29].
At the same time we investigated the neutron yield. The mean yield for neutrons from the D+ D reaction was
measured by the silver-activation method (see Fig. lb). The apparatus was calibrated with standard Ra+ Be and Po+ Be
428
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
IRs? 10 - 300megohmC, 0.06 p F
a
To galvanometer (oscillograph) .
To
oscillograph
Fig. 1. Schematic cross section of appardtuS, a) Neutron generator b) mass spectrograph and other diagnostic devices. 1) Spark source; 2) selection
grid; 3) intermediate electrode (grid) 4) porcelain ChatiTher; 5) diaphragm; 0 targets; 7) Screen; 8) LD insert; 9) combined electrode; 10) cassette
for targets; 11) rod for adjusting targets 12) handle foradjusting ion source; 13) diaphragms 14) photographic plate; 15) magnet pole; 16) porcelain
tube; 17) insulators for leads; K, A) spark gaps; L, inductance and capacitance .iri spark circuit ; C1, C2) capacitances in selector cirCuit;.R1-, R2') .
discharge resistances; Rs) current monitor in-selector circuit; 18) apparatus for,activatiOn.registration of neutrons ;'I, II) scintillators for recording
neutrons and ions, respectively.:,
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
neutron sources. The neutron yield from the D+ T re-
action was measured by the copper-activation method.
The time characteristics of the pulse were studied with
a Type I counter, consisting of a plastic scintillator and
an FEU-24 [photocell]. The absolute number of neu-
trons in each separate pulse was determined by analyz-
ing the tracks on the oscillograms [5]. A Type I counter
was used to study the stability of the neutron yield from
one pulse to another.
The total ion-induced current in the selector cir-
cuit was registered by measuring the potential across
122 = 10 O. Spark gap A protects the measuring circuit
from high-tension leakages occasioned by flashovers.
The mean ion current to the target was determined by
the thermal method, corrections being made for cooling.
The D+_ ion current to the target was determined from
the neutron yield.
Experimental Results
Ion-Beam Composition and Its Influ-
ence on the Neutron Yield. Mass-spectrosco-
pic investigations showed that the neutron yield depends
markedly on the ion-beam composition. The ion beam
suffers definite changes owing to the working of the ion
source and the heating up of the working substance. This
is illustrated by the mass spectrograms shown in Figs. 2a,
b and c. The El and D+ lines are superimposed, and
their separate values are calculated from the lines of
the molecular ions (HD)+ and D.
.
During the first thousand or so pulses, the working
substance is usually evaporated inefficiently. H+,
C+ and 0+ ions predominate in the beam and there are
few Li + ions (Fig. 2a). After aging?about 103-104 pul-
ses?the source enters its optimum working state and the
working substance is evaporated intensively. D+ ions
then predominate, sometimes reaching 80% of the beam (see
Fig. 2b). Impurity ions (Lit, C+, 0+) and molecular ions EDI ,
(HD)+, Hill are found only in small quantities. Li + ions
adhere chemically to the source, as they react with ele-
ments in the porcelain. After about 5 ? 105 pulses, this
leads to destruction of the porcelain tube. The D+ con-
tent of the ion beam remains stable at above 50% for
Fig. 2. Mass spectrograms of beam composition (E = ion about 105 pulses. After this, the number of impurity
energy, kV). a) Initial working stage of spark gap; b) ions (Hf, H2+, Li, C+ and 0+) then rises, and finally,
optimum stage; c) beam composition after 3 ? 105 pulses. after about 2-3.105 pulses, these ions become predomi-
nant (see Fig. 2c).
As the beam composition changes, so does the neutron yield. This is clearly seen from Fig. 3, which plots the
relative intensity of D+ and Lf+ ions, ASD+/ASLi+ (I), and the relative neutron yield, N (II), versus the number of
pulses, n (i.e., the degree of burn-out of the working substance). The values during the optimum period are taken
as unity. Curve I was obtained by photometry of a number of photographs like those in Fig. 2. In measuring curveII,
it was necessary to eliminate the effect of target fatigue. For this purpose, the beam was directed on to the target
for the short time necessary for the measurements, and the target was then covered with the diaphragm.
430
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
A
if
0 50 100 150 200 n , x 10 3
Fig. 3. ASDi./ASLi+, the relative D+ content, and N, the
neutron yield, versus n, the number of pulses. During the
optimum period, ASD+/ASLi+ and N are normalized to
unity.
a
Fig. 4. Oscillograms of ion-electron current in accelerat-
ing gap (a) and neutron pulse (b) for reactions D+D (I) and
D+ T (II). t= duration of ion-current pulse, ?sec.
small Vs is due to increase of the D+ current at target 6 (see
fact that the ion beam begins to fall outside the edges of the
growth of the spark-plasma density, and consequently of the i
At VG =110 kV and Vs =70 kV, the maximum neutron
length ?2501jsec were ?7 ? 105 and ?109 respectively.
By means of similar experiments, it was found
that the number of pulses over which the neutron yield
decreases by a factor of two (by comparison to the
optimum yield) is about 1.5 ? 105, whatever the spark
gap. The over-all lifetime for the production model
[6] is about 105 pulses.
The working substance can be a volatile organic
compound (e.g., hydrocarbon) based on deuterium.
This should preferably have a high deuterium content
and not contain chemically active substances. By this
means the source lifetime can be considerably increased.
Total Ion Current and Its Influence
on the Neutron Yield. The mean ion current
per pulse to the target, measured by a thermal method,
increases proportionally to the spark power and reaches
0.5-1 A when the potential across the spark condenser
is Vs =70 kV and the mean pulse duration is ,-100?sec.
The time characteristics of the ion current were
determined from oscillograms of the total ion current
and the secondary-electron current in the selector cir-
cuit. The ion current was corrected for the coefficient
of secondary emission, which in these experiments was
3.5-4 and did not vary appreciably in the accelerating
voltage range VG =60-100 kV.
Figure 4a shows oscillograms of the total ion
current: this oscillates with the natural frequency of
the spark circuit and decreases proportionally to the
oscillating current in the spark. Ion-current pulses
are observed when the cathode spots fall on the com-
bined electrode 9 (see Fig. 1).
In general, the neutron yields from the D+D and
D+T reactions (see Fig. 4b) are strongly correlated in
time with the ion current. The total ion current, de-
termined from the neutron yield (taking account of the
relative D+ content [7]) at VG = 80-100 kV, does not
differ appreciably from the value determined by the
thermal method.
At constant VG, the neutron yield per pulse from
the D+D reaction increases with the power of the spark
discharge (Fig. 5). The spark power was increased by
increasing the initial potential Vs across the spark ca-
pacitance C3; the gap K was also widened. The pulse
repetition frequency was kept constant by varying R2
(see Fig. la). The rapid rise of the neutron yield for
Fig. la). The bend in the curve of Fig. 5 is due to the
target. The absence of saturation is here caused by
on current, with increasing Vs.
yields from the D+D and D+ T reactions per pulse of
431
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
7
6
4
3
2
20 30 40 50
6? Vs , k V
Fig. 5. Neutron yield per pulse versus
spark power at VG= 106 kV (Vs= initial
potential across spark condenser C3, N =
number of neutrons per pulse).
Fig. 6. Oscillograms of ion-electron current in ac-
celerator gap (a) and D+ current in analyzer (b).
Energy Scatter of Ions in the Beam and Its
Influence on the Neutron Yield. The mass spectro-
grams (Fig. 2a, b, c) reveal that the ions have an energy scat-
ter of up to ?15 kV, due to decrease in the accelerating voltage at large pulse currents. By increasing C1 and C2,
this falling-off in voltage can be reduced, but there is then an increase in gas evolution during flashover, and this
hinders the operation of the generator. These adverse effects are especially noticeable at low accelerating voltages.
At voltages 100 kV, voltage fluctuations reduce the neutron yield only by 10-15/0.
In most cases, the sections of the mass-spectrogram parabolas (Fig. 2a, b, c) have a layer structure. This is
due to current oscillations in the analyzer. Figure 6 illustrates this effect with oscillograms of the total ion-electron
current in the selector circuit (a) and the D4-- ion current (b) in the analyzer. The beam usually enters the analyze
only during the decreasing stage of each oscillation of the total ion current. Between successive entries of the beam
into the analyzer, the accelerating voltage falls off and the beam departs from the parabola, imparting a layer struc-
ture to the blackening. In some cases the analyzer current vanishes altogether, or does not at all correspond in am-
plitude and duration to the total ion current. These deviations in the analyzer current are apparently due to fluctua-
tions in the emitting surface of the plasma; these cause changes in the conditions for transmission of the relevant
beam component through the narrow opening in the target. The lack of correspondence between the analyzer current
and the total ion current might cause errors in determining the beam composition: to ensure reliable analysis, we
therefore averaged over a large number of pulses.
482
30
20
10
a
0 I_ A
70 80 90 100 110
30
20
10
.1.
30
20
10
70 80 90 100 110 120
.4.
?
60 70 80 90 In
Fig. 7. Histograms showing stability of neutron yield, constructed from separate consecu-
tive series of measurements taken over ,-104 pulses. In= amplitude of neutron pulse, arbi-
trary units; W =number of pulses. a) VG =96 kV; b) VG =90 kV; c) VG = 87 kV.
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
?
a
??
?
? '
??
? ??
?
?
Stability of Neutron Yield
T a r get Fatigue. The stability of the neutron yield depends
on the working of the ion source and on target fatigue. These two
factors were studied separately.
0.5The neutron yield was measured, with the generator working in
8 16 24 32 On.xIO3 the optimum state, for each of 100-150 pulses. The target was then
Fig. 8. Fatigue of TT target versus num- covered with the diaphragm and a further series of measurements made
ber of pulses, n. N =relative neutron yield, over ?104 pulses, and so on. Similar measurements were made for all
normalized to unity at n=5 ? 103 pulses. six spark gaps in the production-model ion source. On passing from
one gap to another, the total neutron yield can change by a factor of
two or three. For an individual gap, the neutron yield varied by about ?5-10%. In the conditions of our experiments,
the vacuum spark thus developed with adequate stability. Figure 7 gives typical histograms of the neutron yield.
Fatigue of the target might be caused either by the formation of an organic film on its surface (oil from the
diffusion pump) or by sputtering from the beam.
The fatigue of TiT targets was investigated experimentally by the following method. The generator was
brought into its optimum working stage, and the relative neutron yield from the D+T reaction was measured, keep-
ing the accelerating voltage constant at VG= 90 kV. Measurements were made over ,?100 pulses out of each ? 103
pulses. The yield was measured by a monitor based on the short-lived induced activity of lead (T112=0.8 sec) [8].
A special electronic circuit ensured registration of the activity belonging to Pb207n, formed mainly by the reaction
pb2os 2711
2n)Pb0 . Fatigue of the TT target scarcely reduces the neutron yield during ?4.104 pulses (Fig. 8).
Our investigations of this neutron generator with spark-ion source have thus enabled us to make considerable
improvements in the working parameters of the first model [3]. The simple construction of the generator and associ-
ated electrical circuit make this a convenient device for laboratory investigations in nuclear physics and its applica-
tions. In particular, it has been used to study short-lived isotopes and isomers (T112.- 1 msec) formed in (n, 2n) re-
actions with various nuclei.
The authors would like to thank I. P. Selinov and I. M. Rozman for their interest and valuable advice.
LITERATURE CITED
1. J. Gow and H. Pollock, Rev. Scient. Instrum., 31, 3, 235 (1960).
2. B. Carr, Nucleonics, 18, 75 (1960).
3. G. E. Murguliya and A. A. Plyutto, Pribory i tekhnika eksperimenta, 5, 28 (1961).
4. A. Dempster, Rev. Scient. Instrum., 7, 46 (1936).
5. G. E. Murguliya, A. A. Plyutto, and I. M. Rozman, Pribory i tekhnika eksperimenta, 1, 54 (1962).
6. A. A. Plyutto, K. N. Kervalidze, and I. F. Kvartskhava, Atomnaya Energiya, 8, 153 (1957).
7. E.-G. V. Aleksandrovich and V. A. Sokovishin, Pribory i tekhnika eksperimenta, 5, 7 (1961).
8. L. Ruby and J. Rechen, Nucl. Instrum. Methods, 15, 74 (1962).
433
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
FEASIBILITY OF USING THORIUM IN FAST POWER REACTORS
(UDC 621.039.526)
A. I. Leipunskii, 0. D. Kazachkovskii,
S. B. Shikhov, and V. M. Murogov
Translated from Atomnaya tnergiya, Vol. 18, No. 4,
pp. 342-350, April, 1965
Original article submitted July 9, 1964
This paper gives data on the feasibility of using U233 and thorium in a breeding-reactor system. From
the viewpoint of doubling time, the most promising method is the simultaneous use of U233-Th and
Pu239-U239 in a mixed-fuel cycle in fast reactors; the thorium being distributed in the screens and the
U233, Pu239 and U238 in the cores. Consumption and breeding of U233 and Pu239 is arranged so that their
amounts remain in constant ratio. The doubling time of a fast-reactor system working with a mixed-
fuel cycle is considerably less than that of reactors using only U233 or thorium. The main raw material
is thorium. A method is indicated for obtaining isotopically pure U233 containing ?10-4% u232.
The use of thorium for U233 breeding encounters substantial difficulties, which make this system uneconomical
as compared, for example, with fast breeder reactors using plutonium.
These difficulties, which impede the use of tl.23 for nuclear power production, are as follows:
1. The low breeding rate in the U233-Th thorium fuel cycle.
2. U233 has a higher activity than U235 or Pu239, owing to accumulation of U232 and its daughter isotopes.
The available data on the nuclear constants of U233 and Th232 [1] indicate that the U233-Th cycle can in prin-
ciple be used to breed U233 from thermal or fast neutrons.
In the optimum case, taking account of all losses, the breeding ratio of thermal thorium reactors is up to ?1.1,
while for fast reactors it is about 1.3. The minimum doubling time for such systems, allowing for the practical fac-
tors involved in nuclear power production, is 15-20 years [2]. For intermediate neutron energies (1 eV to 1 keV),
vat-, the number of fission neutrons per absorbed neutron, is less than two for U233: breeding is thus impossible under
these conditions.
Fast plutonium-breeder reactors, in which the breeding ratio is ?1.5-2.0 and the doubling time can be much
less than 15 years, and much more efficient than reactors employing U233 and Th.
However, it is found that the so-called mixed-fuel cycle, used with fast neutrons, can substantially improve
the characteristics of fast reactors using thorium; a system can be devised with a doubling period nearly equal to
that obtained with a pure plutonium fuel cycle. In addition, the activity of the U233 produced can be considerably
reduced.
Breeding in Fast Reactors With a Mixed-Fuel Cycle
A "mixed-fuel cycle" is a breeding-reactor system using two fissionable isotopes simultaneously as nuclear
fuels: these are Pu239 and U233, and the raw materials from which they are bred are U238 and Th232.
For a fast reactor working in such a cycle, the neutron balance is much better than that for the thermal tho-
rium cycle, owing to the additional contribution from U238 fission and the high 'Jeff of PU239.
A mixed-fuel cycle can be realized in either of two ways:
1. The nuclear fuel is placed in two reactors working in parallel. Thorium is placed in the breeding zone of
both reactors. In the core of one reactor (which we shall provisionally call"Type 9"), the main fissionable isotope
434
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
is Pu239, while the breeder raw material is U238. In the core of the other (which we shall call "Type 3"), the main
fissionable isotope is U233 and the raw material is again U238. U233 is formed in the thorium used in the screens of
both reactor types, and after chemical separation part of it is used to compensate for its consumption in the Type-3
reactor and the rest to construct new cores for reactors of the same type. Part of the secondary Pu239, formed in the
U238 of the Type-3 and-9 cores, is used to compensate for consumption in the Type-9 reactor, and the remainder is
extracted in the chemical plant and used to construct new cores for reactors of the same type.
2. The fissionable isotopes?Pu239 and U233 ?are regarded as a single-composite nuclear fuel, employed in the
core of a single reactor, with some fixed isotopic ratio of U233 to Pu239.
The plutonium and U233 can be either homogeneously mixed in the core, or placed separately in the inner and
outer sections, or in separate cassettes. The breeding blanket of such a reactor contains thorium. The U233 formed
in this is extracted in the chemical plant, and part of it 'is used to compensate for consumption of the mixed fuel.
Part of the secondary Pu239, formed in the U238 of the core, is used to compensate for depletion. As excess of
Pu239 and U233 is extracted from the thorium, new mixed fuel is available for the construction of new cores.
The dynamics of mixed cycles can conveniently be analyzed by means of Usachev's theory pf the evolution
of breeder-reactor systems [21.
Let us assume that all the excess fuel (U233 and Pu239) produced by the mixed-fuel cycle is used in the con-
struction of new reactors of Types 3 and 9. The appropriate non-stationary equations, allowing for the effect of
delay, are:
dN9
dt . ,
/9 (t)? t9kt? Ta9), (1)
dN3
citi3(t)? 13 (t ? T 33), (2)
A3 BRC 3i3 (t ? ?
19 (t)= i9 (t ? Tag ? p.c 3) [1 ? A9 ( 1 ? BRC 9)189 +11/-39M (3)
4p.c3) 89
13(t) =13 (t ? Ta3 Tp.c3) (1 ? A3) 82 + A3 8733 N3 (t ? Tssc Tps.sc) 83
e
BRS 9e 8 + Z93 T Ba S
p.SC Ng. (t ? n:9? n.sc) 83. (4)
13 (t ? T.3 ? Tpe.sc) A3 BRS 363 + 9 19 (t Ta9 Te ) A9
A9
In these equations the history of each fissionable isotope is traced through time. The indices 3 and 9 refer to
parameters associated with U233 and Pu239 respectively.
i3(t) and WO are the over-all rates of introduction of new cassettes in the cores of the working and newly-
constructed reactors of Type 3 and 9 at time t. A convenient unit for i(t) is the number of cassettes in a core. In
this case i(t) can be taken as the rate at which a given type of core passes through the reactor.
M3 and Mg are the initial charges of the corresponding fissionable isotopes in a core. Thus M3i3 and M9i9 are
the rates at which the corresponding isotopes enter the cores, expressed in kg/year.
N3(t) and N9(t) are the numbers of reactors of each type working at time t. Then (1) and (2) represent the
rate of change of the number of reactors of each type, and this is equal to the rate of introduction of cassettes (in
core units) from the processing plant to the reactor installation at time t, minus the rate at which cores leave the
reactor for the processing plant at time t, which is equal to their rate of introduction to the reactor at time t-Ta,
where Ta is the run period of a reactor, i.e., the mean time during which the cassettes are kept in the core.
Equations (3) and (4) indicate which elements contribute to the rates of entry into the cores of the fissionable
isotopes Pu239 and U233.
6,3 and 1l3 are the amounts of fissionable isotopes burnt up, relative to their initial loadings.
The breeding ratios are broken down into two components. These are BRC (the breeding ratio of a core, also
called the inner breeding ratio) and BRS (the breeding ratios of the screens): BRS=BRSs+ BRSe, where the indices
s and e refer to side and end screens, and each component of the breeding ratio can have the index 3 or 9 accord-
ing to the reactor type. Thus BRC3 and BRC9 are the rates of formation of Pu239, relative to the rates of burn-up of
435
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
TABLE 1, Main Initial dhdtaeterigties Of Sy-Ste-111S
, .
? ?
Filet type
Fuel cytles
Fuel composition
Plutonium
Pu02 ?UO2
Oxide fuel
Mixed
Pu02?
Thorium
U23302?
Metallic
Plutonium
Pu?U
fuel
Mixed
Pu? U233 ?
Thorium
U2"?Th
U23302 ?
T h02
U2"
U23802
clo Porosity of fuel in core
27.0
27.0
27.0
28.5
28,5
25.0
Maximum allowable temperature
at center of fuel element, ?C.
2400
2400
2400
700
700
950
Maximum allowable burn-up, kg
fragments per 1 t fuel
100
100
100
50
50
50
Heat-transfer agent
Sodium
Mean rise in temperature of heat-
transfer agent, ?C
230
Temperature of heat- transfer
agent on entering reactor, ?C. .
300
Maximum velocity of heat-
transfer agent, m/sec
10
Thickness of fuel-element
jacket, mm
0.4
Flattening of core (D/H)
1.4
Screen material
UO2
Th
Th
UO2
Th
Th
Fuel density in screen, g/cm3. .
10
11
11
10
11
11
Construction material
Stainless steel
Note. The composition and dimen
sions of the side screen were the same for all the systems: &fuel= 0.60;
ei\ja = 0.25; cft= 0.15. The screen t
hickness is 60 cit.
U233 and Pu239 respectively, in cassettes or core of a Type 3 or Type 9 reactor; BRS3 and BR.% are the rates of formation
of U233 in the thorium screen of a Type 3 or 9 reactor, relative to the rates of burn-up of U233 and Pu239, respectively.
The losses of fissionable isotopes U233 and 1311238 in the processing plant are represented respectively by 63 and
69; (1-63) and (1-69) are the amounts of irreversible loss during processing, relative to the total amount of a given
isotope entering the plant.
Tc is the lifetime of U233 in a thorium side screen.
s
The thorium end screen is discharged at the same time as the core; this is reflected in the third and fourth
terms of (4).
Tp.c and Tp.sc are the processing times of cores and screens respectively; Tp.c can be different in different
types of reactor (Tpx3 and Tp.c 9). Tp.sc for a side screen can be different for T for an end screen (Ts and
Tp.sc p.sc
T e
p.sc)? However, reactors of Types 3 and 9 have the same Tp.sc, as the same raw material is used in both cases
for the screens.
The asymptotic solution of (1)-(4) is:
11T9(t)=e)t, N3 (t)= ce" ,
i9 (t) = cte" , i3 (t) -=
A system of simultaneously operating reactors, in which the secondary fuel is expended only on the construc-
tion of new reactors, ultimately takes the form of (5), irrespective of the initial values of N3(t) and N3(t). The ratio
of these quantities must finally attain a constant value, c =N3(t)/N9(t), which is derived below. This means that,
given all the initial reactor characteristics, the presence of U238 in the cores and Th232 in the screens unambiguously
determines the rate of production of secondary fuel (Pu239 and U233), and also their corresponding rates of expendi-
ture in the construction of new reactors, which also ultimately satisfy the asymptotic relations (5).
(5)
436
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Mcr, kg
900
700
SOO
300
3
2
1
lb
la
100 300 500 Q, kw/liter
Fig. 1. Critical mass of reactors work-
ing with mixed plutonium and thorium
oxide fuel, core volume 2000 liter, plot-
ted versus mean power intensity Q. 1)
Reactor with mixed fuel; (la, b) quan-
tities of U233 and Pu239, respectively, in
critical mass of reactor with mixed fuel;
2) plutonium reactor; 3) thorium reactor.
Critical mass r
1.5
1.4
1.3
1.1
1.1
1.0
0
2
4
1 2 3 S
Core volume, m3
Fig. 2. Critical mass ratio, Mcr, and (7)-1
--Cle) f, for Type 3 and 9 reactors, plotted
versus core volumes for various mean power
intensities 0, kW/liter: 0) 300; A) 400;
0) 500; V) 600; 1, 2)Men/MU 9 for oxide
and metallic fuel, respectively; 3, 4)
(i- 1?Tc) of u233
in reactors with oxide and
(17-1-Z)-6f Pu239
metallic fuel, respectively.
Substituting from (5) in (1)-(4) and eliminating the three unknowns, a, b,and c from the four equations obtained,
we find the following characteristic equation (when e=e3=e s):
,
? e
(Ta3+Tp.c3 ) (1 ?,A3) 1? e
0.:TaT3 a3 -0) (7s. c3+TP?s
_ (Ta3H-Tp.ca)
A3 BRc.3e
= (Ta9+TP?c 9)11? A9(1- BRc9)1
[e (7'a 3-I.-Tp.c.3) A
A3 BRS 83 ? A3
e (T
BRS 3 e
p. sc
? ?T
a9
? e
BRS 9 a 9 -0 (Ts +T s )
A9 BRSs9 e ?se'3
CaTa9
from which we can calculate co. The quantity T2 = In 2/co clearly has the physical meaning of the doubling time
a system of reactors working on a mixed-fuel cycle.
We obtain the following expressions for the coefficients a, b, and c:
co)
a? b =--
-aaa9
1?e
1
--o)T --e
Mg 1? e a 6 e
c, = ?
-0)(T59 frp,, [1_ A3 (I BRC 9)]
?0) (Ta3+T p. c 3)
43 BRc3e
(6)
in
(7)
It must be emphasized that T2 is the doubling time for Pu239 and U233 simultaneously. From (5) and (7) it fol-
lows that, in a mixed reactor system, the ratio of the numbers of reactors of Types 3 and 9 remains constant:
N3 (t) b= i 3 (t)
=_-
N9 (t) a i9 (t)
(8)
If we now consider a system of reactors with cores containing mixed fuel and U238 and screens containing tho-
rium, (1)-(4) and (6) retain their form, provided that the core consists of a mixture of two types of cassette, i.e.,
Type 3, containing U238 enriched with U233, and Type 9, containing U238 enriched with Pu239. The definitions given
above then retain their meanings and (8) shows that, in reactors with mixed fuel, the number of Type 8 and 9 cas-
settes remains constant.
43'7
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
For a system of reactors working with a mixed-fuel cycle, we can introduce the idea of a combined breeding
ratio,
BR --=A79-E'n'a
. .
Qi+ Q3 (9)
where N9^ and Ng are the breeding rates of secondary plutonium in U238 and secondary U233 in thorium, and Qg^ and
are the rates of burn-up of Pu and U233 in the corresponding cassettes.
As the breeding ratio can always be split into an inner term BRC and an outer term BRS,
BR =BRC+ BRS, (10)
where
BRC = ./V9 .
Qs+ Qs
/:r3
BRS = ? ,
Q9+93
As Pu239 is produced in cassettes of Types 3 and 9,
&3 =Q3 BRC3 + 63BRC9,
whence
BRC ?BRC3+ 13 BRC3
1+13
(12)
(13)
where 6 =Q3/Q9. In the thorium screens U233 is formed by burning of U233 and Pu239 in Type 3 and Type 9 cassettes
respectively. Therefore
where
Consequently,
Numerical Results
Ng =Q3 BILS3 + Qg 13169,
BRS =BRSs+BRSe.
BRS ? BRS3+ 13 BRS3
1+ )3
(14)
As an exaniple, we give comparative characteristics of fast power reactors working in plutonium, thorium and
mixed-fuel cycles. The comparison is made for reactors with oxide and metallic fuel, with core volumesVc of 1000,
3000, and 5000 liter, working at optimum power intensity, i.e., that corresponding to minimum doubling time.
The main parameters of the systems considered are given in Table 1 and are typical for planned fast power
reactors [2]. The table also gives some characteristics of fuel materials us'ed in the systems studied.
Let us consider the second method of realizing a mixed cycle, i.e., a fast reactor with mixed fuel.
The calculations were based on a 26- group system of constants [1], and were carried out with the aid of com-
puters.
The results are given in Table 2 and Figs. 1-4.
The reactor characteristics were determined for mean isotopic composition of the core fuel, corresponding to
a given depth of burn-up and taking account of accumulation of fission products and transuranic and transplutonic
elements.
As seen from Table 2 and Fig. 1, the critical mass Mer in reactors with mixed fuel is about 10% less than that
of plutonium reactors, because (-77-1?a)?af for U233 is higher than for Pu239 (see Fig. 2). At the same time,
438
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
TABLE 2. Physical Characteristics of Systems with Oxide and Metallic Fuels
Characteristic
liter
Vcore ,
Qcore, kW/liter
&fuel
8Na
&steel
Mcr9, kg
Mcr3, kg
?f Pu239, barn
-
af U233, barn
of U238, barn
of Th232, barn
veff Pu239
veff U233
Contribu- U238
tion to BR Th232
BR
BRC
T2, years
liter
''core'
Q
core, kW/liter
8fuel
&Na
csteel
Mr, kg
M3cr, kg
-?Pu239, barn
-c3f U233, barn
-0 f u238, barn
-af Th232, barn
Veff PU238
Veil' U233
I
Contribu- U238
tion to BR Th232
BR
BRC
T2, years
Plutonium reactor
Pu02- UO2
Reactor with mixed fuel
Pu02-U23302-U23802
Thorium reactor
U23302-Th02
Oxide fuel
1000
3000
5000
1000
3000
5000
1000
3000
5000
500
400
300
500
400
300
500
400
300
0.439
0.389
0.425
0.439
0.389
0.425
0.429
0.380
0.419
0.360
0.425
0.397
0.360
0.425
0.397
0.368
0.430
0.403
0.201
0.186
0.178
0.201
0.186
0.178
0.203
0.190
0.178
515
1140
1790
215
645
1200
250
365
430
520
1100
1740
1.94
2.01
2.02
1.90
1.96
1.98
2.85
3.12
3.20
2.70
2.88
2.92
0.050
0.044
0.044
0.054
0.045
0.043
0.0030*
0.0028*
0.0027*
0.0116
0.0097
0.0092
2.42
2.36
2.40
2.46
2.39
2.38
2.26
2.21
2.20
2.28
2.26
2.23
0.24
0.23
0.19
0.18
0.19
0.18
0.010
0.006
0.005
0.034
0.030
0.027
1.57
1.50
1.55
1.33
1.35
1.40
1.22
1.17
1.19
0.51
0.69
0.87
0.44
0.65
0.81
0.40
0.56
0.69
7.1
6.9
7.4
12.4
10.6
10.8
19.9
19.7
21.0
Metallic fuel
1000
3000
5000
1000
3000
5000
1000
3000
5000
500
400
300
500
400
300
500
400
300
0.379
0.351
0.414
0.379
0.351
0.414
0.464
0.412
0.464
0.357
0.422
0.393
0.357
0.422
0.393
0.349
0.414
0.386
0.264
0.227
0.193
0.264
0.227
0.193
0.187
0.174
0.150
575
1320
2120
250
870
1640
-
255
375
400
640
1360
2150
1.84
1.85
1.84
1.81
1.83
1.82
-
2.50
2.61
2.58
2.36
2.46
2.45
0.045
0.040
0.039
0.048
0.040
0.040
-
0.0024*
0.0022*
0.0022
0.0124
0.0102
0.0096
2.54
2.50
2.51
2.57
2.52
2.53
-
2.31
2.29
2.30
2.35
2.23
2.26
0.34
0.33
0.29
0.27
0.35
0.30
-
0.008
0.005
0.004
0.044
0.041
0.046
1.83
1.85
2.05
1.58
1.67
1.87
1.33
1.30
1.34
0.78
1.05
1.38
0.69
0.98
1.28
0.45
0.63
0.78
6.0
5.0
5.0
8.1
7.3
7.0
17.5
14.7
14.3
"For neutron spectrum in breeding zone.
substitution of U238 for thorium in the core spectrum leads to some increase in reactivity, so that the critical mass
of a mixed-fuel reactor is somewhat less than the critical mass of U233 in a thorium reactor.
As well as the relative efficiencies of U233 and Pu239, Fig. 2 gives the critical mass ratios of Type 3 and 9
reactors with oxide and metallic fuel, other conditions being equal. It is seen that the critical mass of U233 reac-
tors with oxide fuel is 1.4 times less, and with metallic fuel 1.2 times less than the critical masses of the corre-
sponding plutonium reactors.
439
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
BR, BRC
t6
1.7
44
2
3
la ?
2a \
?.`
3a
5.
-5. .5...
?
100 200 300 400 SOO 600 Q, kW /liter
Fig. 3
T2,, years
32
24
16
2
3
100 200 300 400 SOO 600 Q , kW lifter
Fig. 4
Fig. 3. Total ( ) and inner (-----) breeding ratios of reactors with oxide
fuel, working in various fuel cycles with core volume 2000 liter, plotted versus mean
power intensity -0: 1, la) plutonium reactors; 2, 2a) mixed-fuel reactors; 3, 3a) tho-
rium reactors.
Fig. 4. Doubling times for reactors systems with oxide fuel, working in various fuel
cycles with core volume 2000 liter, plotted versus mean power intensity -0: 1) thorium
reactors; 2) mixed-fuel reactors; 3) plutonium reactors.
In a mixed-fuel reactor (see Table 2 and Fig. 3), where Pu239, U233 and U238 are located in the core, i.e., in
the region with the hardest neutron spectrum, the combined breeding ratio is greater than the breeding ratio for a
thorium reactor. This is explained, on the one hand, by the fact that the contribution to the BR from U238 fission
is greater than that with thorium, thus increasing the BR by 0.2-0.3 for metallic fuel and 0.1-0.15 for the dioxide;
on the other hand, the contribution of Pll238 is higher than that of U233 because of its greater veff. It must be noted
that, owing to the presence of thorium in the screens, there is little protactinium poisoning in mixed-fuel reactors.
As a result, the breeding ratio is higher than that for a pure thorium cycle, by 0.1-0.2 for oxide fuel and by 0.25-
0.5 for metallic fuel. At the same time, it is close to the value for a plutonium reactor.
In calculating the doubling times, the following characteristics of the fuel cycles were taken into account:
1. The period of operation was calculated for a given burn-up and reactor power (for a load coefficient of
0.85).
2. The time for processing in the chemical plant was taken as 0.5 years for all types of fuel and raw material,
including the time required to prepare and transport new fuel elements. The processing losses were taken as 1.5%
of the material processed.
3. Tsc, the delay time for production of secondary fuel from the screens, was determined on the assumption
that the amount of secondary fuel accumulated was, in the limiting case and on average, 10 kg U233 per 1 t raw
material.
The advantages of the mixed over the thorium fuel cycle are seen especially clearly when we consider the
doubling times.
Table 2 and Fig. 4 show that the doubling time of a reactor system using thorium is much greater than T2 for
reactors working with mixed fuel. This is due to the differences in breeding ratio and critical mass. However, there
is little difference in doubling time between a mixed-fuel cycle and pure plutonium: the difference is especially
small for the variants where the greatest contribution to the breeding ratio and power comes from U238 fission.
Let us consider some quantities associated with the rates of Th and U238 consumption in an evolving system of
fast breeder power reactors. In this case the breeding requirements are mainly determined by the rate of construction
440
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
of new reactors and are associated with the expenditure of thorium and uranium in the creation of new breeding
blankets and cores, respectively. The thorium charge in the screen is about ten times greater than the UM charge
in the core; this also determines their relative rates of consumption.
In addition, the raw materials are expended by being burnt up, and their requirements for this purpose are
small by comparison with those for constructing breeding blankets and cores.
In an isolated reactor, the requirements for breeding materials are determined only by the burnup rate and
processing losses. If the latter amount to 1.5% of the total processed, the ratio of the rates of expenditure of tho-
rium and uranium in the reactors under consideration varies from one to four.
In the general case, the relative total thorium requirements (by comparison with uranium) vary over a wide
range?from 1 to 15?and depend markedly on the type of fuel and the construction and sizes of the various parts
of the reactor.
As well as accelerating breeding, the distribution of thorium in the screen of a fast reactor somewhat reduces
the difficulties linked with the use of secondary UM . The U233 and thorium activity in the screen is much less than
when thorium is used in a thermal reactor or in the core of a fast reactor. In the first place., there is a marked re-
duction in the cross section of the reaction Th232(n, 2n)Th231, which leads to formation of U232, the ancestor of a
number of,high- activity elements [3]. Furthermore, in the screen of a fast reactor the absorption cross section
a(n,)')Pa231 (an intermediate reaction leading to the formation of U232) is about a hundred times less than in ther-
mal thorium reactors. Owing to these factors, the U233 activity obtained in the screen of a fast reactor may be about
an order of magnitude lower than that obtained in thermal reactors and in the cores of fast reactors. In the second
place, the separate disposition of U233 and thorium reduces the concentration of Th
228 and Th232, on account of sys-
tematic removal of U233 from the screen, and thus also of U232, a source of accumulated Th228.
Summary
The use of mixed-fuel cycles thus opens new opportunities for using thorium in the breeding of nuclear fuel,
in conjunction with U238. Mixed-fuel cycles have the following advantages:
1. A high combined breeding ratio is obtained, close to that for plutonium reactors, owing to the high con-
tribution of U238 fission to the BR and to the high veff of Pu233 in the neutron spectrum of a fast reactor.
2. The critical mass of a mixed-fuel reactor is somewhat less than that of a plutonium reactor.
3. As a result, the doubling time of a mixed system is close to that for a system of plutonium breeder reac-
tors and is much less than for a system of thorium reactors.
4. Thorium is taken up by a mixed-type reactor system faster than uranium. In an isolated mixed-fuel re-
actor, the rate of consumption of thorium is comparable with that of uranium.
5. The use of thorium in the screen of a mixed-fuel reactor has the consequence that protoactinium poisoning
is reduced to a negligible level.
6. The removal of thorium into the screen of the fast reactor makes it possible to reduce by an order of mag-
nitude the activities of Th232 and the resultant U233, in comparison with the usual methods of using thorium in ther-
mal reactors or in fast reactor cores.
In conclusion, the authors wish to thank M. F. Troyanov and L. N. Usachev for helpful discussion of the results,
and A. N. Shmelev for performing the computer work.
LITERATURE CITED
1. L. P. Abagyan et al., Group Constants for Reactor Calculations [in Russian], Gosatomizdat, Moscow (1964).
2. A. I. Leipunskii et al., Report No. 369, presented by the USSR to the Third International Conference on the
Peaceful Uses of Atomic Energy, Geneva (1964).
3. M. Benedict and T. Pickford, Chemical Technology of Nuclear Materials [Russian translation], Atomizdat,
Moscow (1960).
441
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
TRANSMISSION OF NEUTRON RADIATION FROM A REACTOR
THROUGH HYDROGEN-FREE MEDIA
(UDC 621.039.538:539.171.4)
S. G. Tsypin, B. I. Sinitsyn, and V. K. Daruga
Translated from Atomnaya tnergiya, Vol. 18, No. 4,
pp. 350-357, April, 1965
Original article submitted April 28, 1964
The results of experimental and theoretical investigations into the transmission of neutrons through
hydrogen-free, uniform materials and mixtures are discussed. The relaxation lengths obtained for
neutrons of different energy groups are discussed. It is shown that under certain assumptions the mag-
nitude of the flux attenuation for neutrons of the upper energy group (En 3 MeV) in a hydrogen-free
medium can be obtained from very simple empirical expressions by using well-known constants: the
removal cross section ?rem, the cross section obtained from the reciprocal of the relaxation length
(avx.) and the asymptotic cross section cras. Values are derived for these constants.
A vast quantity of experimental and theoretical data have been accumulated concerning shielding from the
neutron radiation from nuclear reactors and can be found in numerous papers by various authors (see [1-251). As a
consequence of this the necessity has arisen for generalizing and systematizing this information and for developing
general laws and relationships which will facilitate the evaluation and planning of specific types of reactor shield-
ing. These laws are obtained on the basis of experiments as well as on the basis of precise calculations. For this,
all the existing data can be presented in the form of quite simple algorithms and approximations which will enable
accurate solutions to be obtained for many problems without recourse to complex calculations.
In order to derive the simple relationships it is necessary, obviously, to reduce the number of parameters de-
scribing one or another of the processes in the shielding to a minimum. The significant successes in this direction
are shown by the example of investigations into hydrogenous shielding (water, hydrocarbons, concrete, etc.), where
it has been found possible on the basis of the numerous experimental data and various methods of calculation, to
solve almost any problem concerning the construction of a specified reactor shielding by using a specified number
of known parameters and simple empirical expressions [1-3]. However, hydrogenous materials, no matter how good
they are from the point of view of neutron attenuation and simplicity of calculation, are not so versatile if more
rigorous technological requirements are set: temperature stability, thermal conductivity, etc. In this respect, the
investigations which have been carried out recently into hydrogen-free shielding materials are of great value. There
are now adequate data of an experimental nature available concerning the transmission of reactor spectrum neutrons
through uniform media (graphite, sodium, iron, lead) as well as through mixtures (iron-graphite, iron ore, boron
carbide). For certain materials the calculations for the transmission of fission spectrum neutrons are well known.
Calculations by the method of moments have been carried out for beryllium, beryllium oxide, carbon, and iron.
It should be noted that the calculation of hydrogen-free shielding is much more complex and less accurate
than the calculation for hydrogenous shielding. The inaccuracy in the values of the elementary interaction cross
sections of neutrons with the nuclei of the shielding material are displayed to a great extent in hydrogen-free media
since in hydrogenous media the nature of the neutron attenuation function is determined mainly by hydrogen, whose
cross section is known with high accuracy and the presence of other components is taken into account as a perturba-
tion leading to some absorption. Nevertheless, in the case of hydrogen-free shielding, at least for neutron groups
with energies in excess of 3 MeV, all the data obtained can also be interpreted by quite simple empirical relationships.
For this, in addition to the well-known constants used in calculations of hydrogenous shielding (for example,
the removal cross section), the cross section ovx. or the asymptotic cross section are also used.
442
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
TABLE 1. Relaxation Lengths of Neutrons in Shielding
with Minimum Hydrogen Content
Medium
Composition (wt.
?/0) and density
(3/cm3)
Relaxation length
X, cm
E>0.5
MeV
E 3 MeV, Calculated by Formula (4) and Obtained
Experimentally,
Medium
Xcm, cm
experi-
meat
calcu-
lated by
formula
(4)
Iron?carbon ( =4 g/cm3) [9]
Iron?oxygen ( =2.6 g/cm3) [17]
9.6
14.6
10.4
16
A calculation by the method of moments is
given in [18] for the transmission of neutrons through
iron. The divergence from experimental data [15] in
determining the value of the relaxation length is
about 15010.
Iron-Carbon Mixture. A heterogeneous
mixture of iron with carbon (graphite) is for the time
being one of the best hydrogen-free shielding materials
and it has all the essential qualities for operating at
elevated temperatures.
According to the data in [9], the optimum shield-
ing concentration lies within the limits of 70-80 wt. %
iron. The relaxation length of the BR-5 reactor-spec-
trum neutrons varies weakly as a function of energy and distance (for a thickness 60-90 cm). Carbon obviously closes
the gap in iron in the region of En 1 MeV and completely defines the relaxation length for neutrons of intermediate
energy. According to measurements with an indium resonance indicator, a relaxation length X =9.1 cm is obtained.
In the fast neutron region X does not exceed 11 cm (for an iron-carbon mixture with a carbon content of 26 wt. %).
The calculated verification, on the basis of the conclusions in [16], was carried out according to the empirical
formula (the additivity principle was used):
1
cm XFe XC
(1)
where PFe and Pc are the volume concentrations of iron and carbon; XFe and XC are the relaxation lengths in cm of
neutrons with En> 3 MeV in iron and carbon respectively; X is the neutron relaxation length in cm in the mixture.
The value of Xem calculated by formula (1), agrees within the limits of experimental error with the experimental
value determined by measurements with a S(n, p) detector with an effective threshold Eeff= 3 MeV (see Table 5).
Iron-Oxygen Mixture. The investigation in [17] of neutron attenuation in an iron ore medium with an
iron content of ?60 wt. clio and oxygen ,-,30 wt. % is interesting. The content of other elements (Si, Mg, Ca etc.) is
10010 and can be considered as a certain perturbation in the medium which does not play a significant role. Thus,
the question actually is one of a homogeneous mixture of iron and oxygen (essentially Fe203).
Comparison of the neutron attenuation in an iron-ore medium, iron and carbon mixtures (all the investigations
were carried out with the BR-5 reactor neutron spectrum) indicates the significant role of oxygen in the over-all at-
tenuation of the shielding for fast as well as for intermediate neutrons in particular.
The neutron relaxation lengths, calculated according to the data in [9, 15, 17] are given in Table 3 for a
comparable nuclear density of iron of 1.54 ? 1022 cm-3. It can be seen that oxygen also effectively screens the gaps
in the iron cross section, just as in the case of carbon.
For the region En> 3 MeV, a calculated verification was carried out using the removal cross section according
to the formula
1 _ _
'
?V
:cm rem( O) (o)+ 0-rem (Fe) V(Fe) remV7
(2)
where ?rem (0, Fe) and v(0, Fe) are the removal cross sections in barn (Table 4) and the nuclear densities in cm-3
for oxygen and iron in the mixtures respectively; Zrem and 17 are the mean removal cross sections and nuclear den-
sity of the remaining elements. The value of Xem obtained from expression (2) was compared with the experimental
data ES (n, p) detector with and effective threshold Eeff= 3 MeV]. The results agree within the limits of measure-
ment error. For this, it was assumed in [17] that ore/n(0)=0.74 barn. The value of are/n(0)=0.79 barn, obtained
from data for B203 and boron, (see Table 4), has been used for calculating XeminTable 5.
The prospects for utilizing iron-ore shielding are linked with the possibility of obtaining a higher density of
the medium (-3-3.5 g/cm3) or the introduction of stable hydrogen-containing additives (Cal-I2, serpentine-3Mg0
?S i02 ? 2H20 [5]) into that portion of the iron-ore shielding which is external relative to the reactor core, where as
a result of the good thermal conductivity of the inner portion the temperatures are quite low. Such a "two-layered"
447
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Shield would represent a ornbihatibil of the qualities of an iitii-Carbon shield and ati iron-water shield (see Table 1
and 2) with relatively inexperitiVe starting Material and With a high therrnal stability:
Lead. The attenuation of neutrons from a water-cooled,, water-moderated reactor has been studied in [11]
by means of a neutron scintillation counter. The spectrum of the neutron flux incident on a lead assembly is iden-
tical with the fission spectrum for neutron energies in excess of 2 MeV. It is shown that the relaxation length of
fast neutrons is weakly dependent on the energy (X 10 cm for the energy region En from 3 to 11 MeV) and on the
shielding thickness (X varies by several percent as a result of doubling the thickness). The cross sections ?IA, cal-
culated according to these measurements for En> 3 MeV is close to the value of Grern = 3.3 barn [19] and Gas for
the group 2.5-.0 MeV (see Table 4).
General Laws of Neutron Attenuation in Hydrogen-Free Media.
Empirical Relationships and Parameters
The experimental and calculated data which have been discussed above, concerning the transmission of neu-
trons through hydrogen-free shielding enable the conclusion to be drawn that the nature of attenuation of neutrons
with En> 2 to 3 MeV in the shielding material can be approximated with adequate accuracy to a function of the
form
CD (x)cv exp (?qx).
(3)
Here q is some empirical parameter, which will be defined below. The range of applicability of expression (3) is
included within the limits of variation of x from 2 to 15 mean free paths [1, 2, 8, 9, 10, 13, 14, 15, 17, 18] inside
a thick shield.
It is essential to note that for very large shield thicknesses the neutron relaxation length, because of filtration
by the medium, will be determined by the minimum neutron-interaction cross section in the energy range of the
reactor spectrum.
The following quantities can be used as the parameter q [16]:
1) the macroscopic removal cross sections Erem, if the neutron group with energy in excess of 3 MeV is
considered;
2) the macroscopic cross sections Evx, obtained from the reciprocal relaxation lengths for neutron groups
with energies in excess Of 3 MeV;
3) the asymptotic macroscopic cross sections; obtained from the solution of the one- velotity kinetic equation
in transport approximation, using the system of group constants [24, 25] for the energy groups 1.4 to or 2.5 to MeV.
Data concerning the parameters mentioned, taken from various published papers, are presented in Table 4.
It can be seen from the table that there is satisfactory agreement between the three stated parameters. The maxi-
Mum deviation from the mean value for these quantities does not exceed 10%. This agreement can be explained
qualitatively as resulting on the one hand from the determination of the stated parameters, and on the other hand
from the physical significance of the procetses Which occur: It is well known that the minimum distance Rmin in
the hydrogenous material, in which the removal cross section is Measured, depends on the energy threshold of the
detector and decreases at the energy threshold is increased [16]. At the same time, the magnitude of the removal
crosS section depends weakly on the threshold Of the detector up to an energy of 7-8 MeV [19-21]: in principle,
there exists such a value for the detector threshold for which Rmin-)0 and, consequently the removal cross section
must be identical With the reciprocal neutron-relaxation length measured in this medium by a detector with the
same energy threshold. As mentioned above, this threshold is equal to Eeff= 3 MeV. The conformity of the asymp-
tOtic trots section and the cross section calculated from the reciprocal relaxation length, lot neutron energies in
excess of 3 MeV it a consequence Of the fact that the rieUtron spectrum in the medium dbet not vary strongly (asymp-
totic approximation is Valid) with distance (within the limits stated above) and the neutron absorption is not very
great (transport approxitnatiOn is valid).
The description of neutron attenuation in a shield by a funetiOri of the form of Eq. (3) is generally Valid for
homogeneous materials at Well as for homogeneous and heterogeneous mixtures.
If the shielding consists of a mixture of several elements, then the parameter q in Eq. (3) tan be determined
448
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
from the expression
71
yllIcml= olpi
Ai Ae CM
i = 1
(4)
where oi is a constant for the i-th element in the mixture, taken from Table 4; NA is Avogadro's number; Pi is the
concentration by weight of the i-th element in the mixture; Ai is the atomic weight of the i-th element.
By way of example of the application of formula (4), the determination of X for neutrons with energy En>
3 MeV for the iron-carbon and iron-oxygen media described in this paper can be cited (see Table 5). It can be
seen from Table 5 that within the limits of measurement error (Ax =10%) the calculated values agree with the ex-
perimental results and verify the validity of applicability of expressions (3) and (4) for calculating the transmission
of neutrons in the upper energy group through hydrogen-free mixtures.
Thus, it can be seen from a review of the data concerning the investigation of the transmission of reactor
spectrum neutrons through hydrogen-free media, that a considerable portion of the information is obtained from ex-
perimental projects. The experimental investigations in the majority of cases have been carried out for actual
thickness of shielding under so-called infinite geometry conditions with a unidirectional source most convenient
for comparison with the data from calculations.
Further experiments are desirable with the objective of elaborating the data obtained for narrower energy
groups of neutrons in the materials already studied (beryllium, B4C) and mixtures (iron-carbon, iron-oxygen). In
addition, it is important to investigate experimentally certain materials for supplementing the data and verifying
the general laws. These materials are lithium, nickel, molybdenum and tungsten. The almost total absence of
data (experimental as well as theoretical) concerning the transmission of reactor-spectrum neutrons in the interme-
diate energy region should be particularly noted. It is obvious that this problem will be resolved after the develop-
ment of special detectors (of the "intermediate" chamber type).
By comparison with experimental data there is considerably less calculated data concerning hydrogen-free
shielding. It is extremely essential to have accurate calculations (for example, by the method of moments) on the
transmission of neutrons in B4C, sodium, iron, and lead. Even one precise calculation should be carried out for a
mixture of the type iron-oxygen or iron-carbon especially as there are experimental data for practically infinite
media.
The parameters used in hydrogen-free shielding calculations, which are given in Table 4, can be assumed to
be completely satisfactory from the point of view of their accuracy (the deviation from the mean value is 10%).
The exception is oxygen. For oxygen, the known values of the removal cross section fluctuate from 0.74 to 0.99.
Obviously, "pure" experiment (for example, with liquid oxygen) will give the answer.
The authors render thanks to A. I. Leipunskii and I. I. Bondarenko for valuable comments and advice in the
preparation of the present paper.
LITERATURE CITED
1. G. G. Gol'dshtein, Primary Reactor Shields [in Russian], Gosatomizdat, Moscow (1961).
2. R. Aronson and C. Klahr, Reactor Handbook. III. Part B, ch. 9. Ed. E. Blizard, New York (1962).
3. D. L. Broder et al., Atomnaya tnergiya, 16, 26 (1964).
4. B. Price, K. Horton, and K. Spiney, Protection from Nuclear Radiations [Russian translationT, Izd-vo, IL,
Moscow (1959).
5. I. A. Arshinov, In the collection "Problems in the Physics of Reactor Shielding" [in Russian], Edited by
D. L. Broder et al., Gosatomizdat, Moscow (1963), p. 327.
6. V. P. Mashkovich et al., Atomnaya tnergiya, 17, 65 (1964).
7. D. Wood, Nucl. Sci. Engng., 5, 45 (1959).
8. J. Moteff, The Use of Threshold and Resonance Foils for Neutron-Spectrum Determinations, Caius College,
Cambridge, England (August 16-29, 1958).
9. V. K. Daruga et al., Atomnaya Energiya, 17, 60 (1964).
10. D. L. Broder et al., In the book "Proceedings of the Second International Conference on the Peaceful Uses
of Atomic Energy" [in Russian], Report of Soviet Scientists, 2, Atomizdat, Moscow (1959), p. 674.
11. A. P. Veselkin et al., Atomnaya tnergiya, 16, 32 (1964).
12. S. G. Tsypin, Atomnaya Energiya, 12, 300 (1962).
449
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
13, V. K. Daruga et al., Atomnaya ,Energiya, 17, 145 1964).
14. A. Kania et al., ,Long-D istanee Propagation of 'Neutron Fluxes in Fast Reactors lin French], Seminar on Reactor
Physics, 5, International Atomic Energy Agency, Vienna (1961).
15. V. P. Mashkovich and S. G. T,sypin, Atomnaya Energiya, 11, 251 (1961).
16. B. I. Sinitsyn and S. G. Tsypin, Atomnaya tnergiya, 12, 306 (1962).
17. V. K. Daruga, I. I. Lazutkin, et al., Atomnaya tnergiya, 17, 63 (1964).
18. A. P. Suvorov et al., In theCollection "Problems in the Physics of Reactor Shielding" [in Russian], Edited by
D. L. Broder et al., Gosatomizdat, Moscow (1963), P. 44.
19. J. Bourgeois et al., Report No. 1190 presented by France at the Second International Conference on the Peace-
ful Uses of Atomic Energy [Russian translation], Geneva (1958).
20. S. F. Degtyarev et al., In the Collection "Neutron Physics" [in Russian], Gosatomizdat, Moscow (1961).
21. V. I. Kukhtevich and B. I. Sinitsyn, Atomnaya tnergiya, 10, 511 (1961).
22. R. N. MacDonald and H. H. Baucom, Nucleonics, 20, 158 (1962).
23. G. Chapman and C. Storrs, Effective Neutron Removal Cross Sections for Shielding, Report AECO-3978 (1955).
24. I. V. Gordeev et al., Handbook of Nuclear Physics Constants for Reactor Calculations [in Russian], Atomizdat,
Moscow (1960).
25. L. P. Abagyan et al., Group Constants for Nuclear Reactor Calculations [in Russian], Atomizdat, Moscow (1964).
450
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
INTERRELATION BETWEEN THE GRAIN ORIENTATION
AND THE RADIATION GROWTH OF URANIUM RODS
(UDC 621.039.548)
V. E. Ivanov, V. F. Zelenskii, V. V. Kunchenko,
N. M. Roenko, A. I. Stukalov, M. A. Vorob'ev,
and A. V. Azarenko
Translated from Atomnaya tnergiya, Vol. 18, No. 4,
pp. 357-361, April, 1965
Original article submitted May 4, 1964
The results obtained in radiation tests of 3-heat-treated uranium rods at temperatures of 200-300;
450-470; 480?C are given. Dependences of the radiation-growth coefficients Gi on the growth index
GI characterizing the grain orientation of the tested specimens were plotted. The elongation com-
ponent resulting from the radiation growth due to grain orientation was determined. The dependence
of the radiation-growth coefficient on the test temperature for weakly-pronounced grain orientations
is given. It is shown that the mean values of the linear thermal-expansion coefficient a, measured in
one direction only, do not provide information on the character and degree of the grain orientation if
the latter is not uniaxial.
The grain orientation which appears in uranium as a result of its mechanical treatment is one of the main
causes of irreversible changes in the shape of parts exposed to radiation. In [1], a quantitative relationship between
the degree to which the grain orientation [010] is pronounced and the radiation growth coefficient was established,
and it was shown that the latter tends -to zero in a metal with a quasi-isotropic structure. It -is known that such a
structure can be obtained in uranium subjected to cooling in the 5- or y-phase. However, according to data sup-
plied by various authors, the radiation-growth coefficient of such a metal is not equal to zero; it is equal to 15-30
[2-4]. This may be caused by at least two factors: 1) the presence of weakly pronounced grain orientations in hard-
ened uranium, the existence of which was confirmed by many authors [5, 6]; 2) the conditions under which the fuel
elements are tested, where the observed irreversible changes may be caused by factors unconnected with the grain
orientation (swelling, creep, thermal oscillations, etc.).
The separate effect of each of these factors on the over-all change in the dimensions of fuel elements con-
taining unalloyed hardened uranium is of great importance, since it provides the possibility of controlling the varia-
tion of the shape of fuel-element cores by controlling the degree of grain orientation in uranium and the test conditions.
In the present article, we shall analyze the grain orientations arising in uranium during its heat treatment,
and we shall establish a quantitative relationship between the degree of grain orientation and the radiation-growth
coefficient of uranium. This investigation was performed in connection with the development of rod fuel elements[7].
Material and the Investigation Method
For the investigations, we used uranium of 99.78-99.80% purity, where the percentage of each of the basic
impurities (Fe, Al, Si, C) did not exceed 0.02%. The diameter of the specimens was 4 mm.
Grain orientations with different characteristics, marked in various degrees, were produced by regulating the
parameters of the 5-heat-treatment of uranium: the characteristics of heating to the temperatures at which the
3-phase is present, the exposure time at these temperatures, and the cooling conditions [8].
The grain orientation of hardened uranium was investigated by using the x-ray structural and dilatometric
analysis methods.
451
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
111
150 )013
, 112
131 //
113
33 114
010
43 0 0 13
100
Fig. 1. Reverse pole figure for a uranium rod whose axial thermal expansion
coefficient was a =12.5 ? 10-6?C-1. a) Longitudinal section; b) end-face section.
The "growth index" method, proposed by the authors of [9], was used for describing the grain orientation in
connection with anisotropic radiation growth. The quantitative parameter?the growth index GI?was derived from
x-ray measurements similar to those used earlier for plotting reverse pole figures [10]. The growth index is found
by determining the product between the volume of grains with a certain orientation and the weighting function of
the deformation tensor of anisotropic growth.
The choice of the tensor weighting function is based on the observed deformation of uranium (at constant vol-
ume), while it does not depend on the deformation mechanism. The expression for the growth index can be written
in the following manner:
GI =I{Pi? Pr} (eos2 pi ? cos2 a1),
gi/g2
where Pi= is the density of poles, expressed in terms of the intensities (3i), of the x-ray interferences
1
-- V i1311
n
observed on the specimen under investigation and corresponding to the interference (37) of a standard specimen
with disordered structure; the index i pertains to the (hid) planes; Pr= 1 is the pole density of a specimen with an
isotropic structure; 5 i and ai are the angles between the crystallographic planes (hk/) and the planes (010) and (100),
respectively.
The growth index yields a single quantitatively defined value for the given grain orientation.
The coefficients of linear thermal expansion (a) along the rod axis were measured in the temperature range
from 20 to 100?C by means of a dilatometer on specimens with a length of 100 mm. The measurement error was
equal to ?1'o[8].
The radiation tests were performed in experimental gas loops on enriched-uranium specimens at core temper-
atures of 200-480?C. The temperature was controlled by means of miniature thermocouples, which were placed in-
side the cores. On the average, the heat release density was 15 MW/t. In order to prevent the corrosion of uranium
specimens in the gas stream: they were covered with magnesium-based alloys.
452
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
No)
(0)
a
'040)
069 111(1:0000$01#F..".010)
or, 1.70
(240)c
ass
U4 ,
Q524, (z4r) (tit
0626 1 i242/ 1 O. 4
746e. ea76
0.94 ? (223) 8
19
0.54 0(131) 0,98. (no 4D tO6
*1,0
ft49 0.776(132) ?'94? 013)
? (152) 0.13 ?(03) 469 *088 ?0.8Z
*099 *074
/
1,45 1,03 496 0,85 0,63
ae 0:9V4 1.23
i 4.11.."
4, as ft .7.0. ? 7.1 4111.1,
1.1 t
? 1.1 1.0
? 1.0
f./1,
0.9 ? 0.9
tO1 ift62 0.85
(010.0.68 (041) i1/ 6Z) 21) 43)
0.7 1,2 ,Z 1,7
(no)
1.3
08
114)
(n3)40)
0.15
0,9 0.9
3
0.7
0.9
1.0
1.64
.0,
0,9
(310) (too)
0.8
? 1.2
? 1.0
00.9
08
(240)
(a to)
(iso)
Fig. 2. Reverse pole figures for the end-face sections of 13-heat-treated rods.
a) and b) Specimens 3 and 4 with a = 15.3 ? 10-6?C-1; c) and d) specimens 1
and 2 with a =14.8 ? 10-6?C-1.
Results, Their Discussion, and Conclusions
Comparing the results of x-ray studies of the grain orientation and of dilatometric measurements of the linear
expansion coefficients of the same specimens, we found that the grain orientation [010] was formed along the speci-
men's axis in uranium wire that had not been subjected to heat treatment. The more pronounced the [010] grain
orientation, the lower the values of the linear expansion coefficient.
In metal subjected to 13-treatment, the degree of the [010] grain orientation is relatively low, although under
certain treatment conditions (slow heating to a temperature of 740?C, exposure over a period of 20 min, and harden-
ing in water), it can be considerable (Fig. 1). In correspondence with this, the coefficient of linear thermal expan-
sion of the specimens was much smaller than the isotropic expansion coefficient; its value was in the (12-13)? 106?C'
range.
By varying the conditions of 13-heat-treatment, the type and degree of grain orientation in hardened uranium
can be varied in a wide range. In this case, there is no well-defined mutual correspondence between the degree of
grain orientation and the linear expansion coefficient, since, as a rule, a double or a more complex grain orientation
is observed in the direction of the rod axis (Fig. 2). Consequently, the mean linear expansion coefficient, measured
in one direction (for instance, along the rod axis), may have the same value for grain orientations which are different
with respect to their degree and character. Thus, this coefficient does not provide information on the character and
degree of the grain orientation if the latter is not uniaxial. This, in turn, does not make it possible to use such dil-
atometric data for predicting the anisotropic radiation growth of uranium rods.
For a certain given type of fuel elements and the chosen irradiation conditions, it is possible to use the growth
index method for predicting three-dimensional changes in the rods resulting from their anisotropic growth due to ir-
radiation. We used this method for analyzing the results of radiation tests of several batches of specimens with the
aim of determining the effect of the character and degree of grain orientation and of some other factors on the elon-
gation of rods. By using the radiation test results and taking into account the grain orientation of "indicator rods,"
we plotted the dependences of the radiation-growth coefficients Gi on the growth index GI.
Figure 3 shows the dependence of Gi on GI for specimen batches 1, 2, and 3 with positive as well as negative
growth index values, which were tested in the temperature range from 200 to 300?C. As a result of radiation tests,
453
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
.
20
Al
IS
x
10
i
/17
//
/
N?
V
/
/ ./..---?
- - -
I
61 -1,5 -70-05 0,?,(-1?::
- --
-- ...-----:--..:
7...4.
,/
05
1.0 1 5 GI
? ...!...7
r'
5
/
//
G,
Fig. 3. Dependence of Gi on GI for speci-
mens tested at diffeient temperatures. I)
Batch 6, temperature, 480?C; II) batch 4,
temperature, 470?C; III) batch 5, temper-
ature, 460-470?C; IV) batches 1-3, tem-
perature, 200-300?C.
8
7
6
5
4
3
2
1
1
?
1
1
\
1
n.-..........,0\
0
A
?
\,,,
?
\
x ?
k
250 900
350
400 450 i,?C
Fig. 4. Dependence of G1 on the
test temperature. A) GI=1.0;
0) GI= 0.5.
elongation or contraction of the corresponding specimens should
have been observed [9]. However, curve IV in Fig. 3 was shifted
with respect to the axis by approximately eight Gi units. Con-
sequently, the elongation of rods during the irradiation process
was caused not only by radiation growth as a result of the grain
orientation, but also by other factors. The elongation which
is due only to radiation growth can be determined by means of
the corresponding dashed curve IV, obtained by translation of
the curves to the coordinate origin.
A similar dependence of Gi on GI was plotted for the fourth batch of specimens, which were tested at a nomi-
nal temperature of 470?C under a load of 0.25 kg/mm2 to a depletion of 0.23%. In this case, during 10-15% of the
over-all testing time, the specimens were at a lower temperature (as low as 300?C) as a result of reactor shut-downs
and temperature fluctuations. The measured elongations were converted to the corresponding values of the radiation
growth coefficient, while the obtained curve II was also shifted along the axis by approximately 11 Gi units.
The contribution of radiation growth to the over-all elongation can also be estimated by means of the other
.dashed curve II passing through the coordinate origin. The solid curves in this figure show the corresponding depend-
ences for batch 5, which was tested at a temperature of 460-470?C to 0.18% depletion (curve III); and for batch 6,
which was tested at 480?C to 0.25% depletion (curve I).
The translation of the curves to the coordinate origin (dashed curves) makes it possible to estimate with respect
to their slope the "pure" radiation growth for specimens with a certain given grain orientation. A dependence of the
radiation growth on temperature, which agrees well with the general concepts, is observed. With an increase in the
test temperature, radiation growth caused by the grain orientation diminishes; it almost vanishes at ?500?C.
The dependence of the radiation-growth coefficient on the test temperature for a certain given grain orientation
(GI= const), which is shown in Fig. 4, indicates that the behavior of the dependence of Gi on the test temperature is
roughly the same for weakly as well as strongly pronounced grain orientations [11].
The following conclusions were drawn from an analysis of the results obtained:
1. Factors unconnected with the initial grain orientation may exert a considerable influence on changes in the
shape of uranium rods with a weakly pronounced grain orientation in radiation tests.
2. An increase in the testing temperature to 460-480?C leads to a comparatively small increase in elongation,
which is probably due to swelling (curve I in Fig. 3).
3. A load of 0.25 kg/mm2 applied axially to the specimens at a test temperature of 470?C causes a negligible
increase in elongation (curve II in Fig. 3).
454
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
LITERATURE CITED
1. Buckley, Report submitted at the Spring Symposium on Uranium and Graphite at the Institute of Metals
(London, March, 1962). Khar'kov, Institute of Physics and Technology, Academy of Sciences, USSR (1962).
2. Fut, In the book: Nuclear Power Engineering Metallurgy and the Effect of Radiation on Materials. Reports
by foreign scientists at the International Conference on the Peaceful Uses of Atomic Energy (Geneva, 1955)
[Russian translation], Metallurgizdat, Moscow (1956), p. 89.
. 3. Kittel and Paine, In the book: Transactions of the Second International Conference on the Peaceful Uses
of Atomic Energy. Selected reports by foreign scientists [Russian translation], 6, Atomizdat, Moscow (1959),
p. 310.
4. Metallography of Reactor Materials. Surveys of the Battle Institute [Russian translation],!. Gosatomizdat,
Moscow (1961).
5. J. Store, M. Englander, and M. Gotron, In the book: Transactions of the Second International Conference
on the Peaceful Uses of Atomic Energy. Selected reports by foreign scientists [Russian translation], 6, Atomiz-
dat, Moscow (1959), p. 515.
6. A. S. Zaimovskii, V. V. Kalashnikov, and I. S. Golovnin, Atomic Reactor Fuel Elements [in Russian], Gosato-
mizdat, Moscow (1962).
7. P. I. Khristenko.et al., In the book: Transactions of the Second International Conference on the Peaceful
Uses of Atomic Energy. Reports by Soviet scientists [in Russian], 3, Atom izdat, Moscow (1959), p. 655.
8. V. E. Ivanov et al., Atomnaya tnergiya, 16, 325 (1964).
9. E. Strurcken and W. McDonall, J. Nucl. Materials., 7, 85 (1962).
10. G. B. Harris, Phyilos, Mag., 43, 114 (1952).
11. G. Ya. Sergeev, V. V. Titova, and K. A. Borisov, Metallography of Uranium and Some Other Reactor Materials
[in Russian], Atomizdat, Moscow (1960), p. 78.
455
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
IN MEMORIAM: ANDREI VLADIMIROVICH LEBEDINSKII
Translated from Atomnaya Energiya, Vol. 18, No. 4,
2 page insert following p. 360. April, 1965
We are grieved to report the untimely death, on January 3, 1965, of Professor Andrei Vladimirovich Lebedinskii,
Member of the Communist Party of the Soviet Union, Acting Member of the Academy of Medical Sciences of the
USSR, a highly-esteemed active scientist of the Russian Soviet Federated Socialist Republic, a Major General in the
medical services, and Doctor of Medical Sciences, in the 63rd year of his life.
Soviet and world science has suffered the loss of an outstanding physiologist, scientist, and true continuator of
the work of I. P. Pavlov and L. A. Orbel'.
During the four decades of his scientific and teaching activities, Andrei Vladimirovich had a lengthy and active
career, his scope of activities ranging from medical practitioner to one of the most prominent scientists in Soviet
physiological science. A. V. Lebedinskii belonged to that generation of physiologists whose scientific viewpoints were
formed after.the Great October Socialist Revolution, during the period of the creative fluourishing of ideas of the
remarkable galaxy of Russian physiologists, such as: I. M. Sechenov, I. P. Pavlov, N. E. Vvedenskii, L. A. Orber,
A. A. Ukhtomskii.
Even while enrolled in course IV of the Military Medical Academy, A. V. Lebedinskii embarked on his career
of scientific research activities on joining the staff of the laboratory supervised by L. A. Orber. After completing
his studies at the Academy, he worked as neuropathologist from 1924 to 1928, combining his practical activities with
scientific research at the psychophysiological laboratory of the Military Aviation Fleet. Thereafter his activities for
the next 25 years were inseparably connected with the Military Medical Academy. During that period, A. V. Lebe-
dinskii performed his duties in the physiology department of that institution, first as instructor, and later as its di-
rector. At the same time, A. V. Lebedinskii took upon his shoulders the management of the physiological sector of
the psychophysiological laboratory of the GVF Leningrad Institute of Engineers, the physiology laboratory of the Len-
ingrad Ophthalmological Institute, the physiological sector of the V. M. Bekhterev Brain Institute, and the theoretical
sector of the Leningrad A. L. Polenov Neurosurgical Research Institute, occupying these several posts for a number
of years.
456
From his very first steps in scientific activities, A. V. Lebe-
dinskii began to concern himself with the problems of nerve trophic
activity, and particular its biophysical fundamentals. This prob-
lem was a determining factor in the entire range of his scientific
activities.
Andrei Vladimirovich was one of the pioneers in the use of
electronics as a research tool in studying trophic effects of the
sympathetic nervous system and afferent pathways. The richest
experimental materials and theoretical conclusions on that topic
were generalized by A. V. Lebedinskii in 1945 in the form of the
monograph "On mechanisms underlying neurogenic dystrophy."
A. V. Lebedinskii was the country's leading specialist in the
field of physiological optics. The research findings he published
in this field have received high recognition from ophthalmologists
and have become incorporated into clinical practice. The scope
of Andrei Vladimirovich's interests was amazingly broad. He did
not limit himself to research solely in the field of the sight ana-
lyzer. The work done by A. V. Lebedinskii and his co-workers on
the interaction of various afferent systems were awarded the 1936
I. P. Pavlov Medal by the Leningrad Society of Physiologists.
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Over a span of many years, Andrei Vladimirovich Lebedinskii supervised specialized research in the field of
radiation medicine. In 1953 he was appointed Director of the Institute of Biophysics under the Ministry of Public
Health of the USSR, in which capacity he continued to expand and deepen investigations in the field of radiation
physiology.
A. V. Lebedinskii was one of the founders of the Soviet school of radiobiology. Beginning in the Fifties, he
proceeded to concentrate the brunt of his attention on the study of the response patterns of the organism to exposure
to ionizing radiation. He was one of the first to advance the notion of the high sensitivity of the central nervous
system to ionizing radiation. The research of Andrei Vladimirovich and his disciples in the field of studies of the
response of the nervous system to radiation effects acquired particular importance. The major conclusion arrived
at in these investigations was the postulation of a change in the functional state and the breakdown of the adapta-
bility of all levels of nerve regulation from the cortex of the cerebral hemispheres to the post-ganglion neuron system.
In 1958, A. V. Lebedinskii expressed the hypothesis that the principal point of application of radiation effects
on the nervous system is that of synaptic formations, where depolarization processes take place. This thought found
confirmation in the research done by co-workers at A. V. Lebedinskii's laboratory.
A. V. Lebedinskii and his associates carried out some impressive work on the study of changes in endocrine
regulation in an organism exposed to radiation, resulting in the inference that post-irradiation breakdown of the
normal "hormone ensemble" ensues in the organism.
Focusing on the most crucial problems in radiobiology, Andrei Vladimirovich proceeded to solve some of them
from an overall physiological vantage point. His laboratory unearthed data providing evidence of not only harmful
action, but also stimulatory action by ionizing radiation. A. V. Lebedinskii was one of the first to turn particular at-
tention to the problem of low-dosage effects of ionizing radiation on the organism.
One salient feature of the entire lifelong scientific activity of Andrei Vladimirovich was the striving to study
the biophysical fundamentals of physiological processes. He is considered by right the leading scientist of the nation
in various topics in biophysics, and physiological and pathological manifestations of metabolism.
The study of the fundamentals of shielding of the human organism and of animals from radiation effects oc-
cupied a special place in the creative efforts of A. V. Lebedinskii. Under his guidance, the staff of the laboratory
participated in programs to develop a radiation safety system serving the world's first nuclear-powered icebreaker,
the LENIN.
A. V. Lebedinskii devoted much attention to radiation safety problems in outer-space travel. But it is not only
for that the name of A. V. Lebedinskii is indissolubly linked to the founding and development of this new branch of
biological science: aviation and space biology and medicine. For he began his research in this direction as far back
as the Twenties, returning to it in his later years. In collaboration with the leading physiologists of the nation,
A. V. Lebedinskii developed the fundamentals of space physiology. Under this guidance, a large staff of assembled
specialists pursued research on the mechanisms of extreme responses occurring under the conditions encountered in
space flight. A. V. Lebedinskii is one of the founders of the Soviet school of space biology.
A. V. Lebedinskii was not only a great scientists but also a brilliant teacher. He devoted much time and effort
to the training of the young generation of doctors and research scientists. Under his leadership, 7 doctoral and over
30 candidate theses were accepted. The textbook co-authored by A. V. Lebedinskii and A. D. Ginetsinskii is a
splendid tool for students and a reference for a large number of doctors and research scientists.
Andrei Vladimirovich carried out intensive activities in the service of the scientific community. He was a
member of the board of directors of the All-Union Society of Physiologists, a Vice-Chairman of the board of directors
of the Moscow Society of Physiologists, chairman of the scientific councils of the USSR Academy of Sciences on the
problems "Radiobiology" and "Specialized Physiology." A. V. Lebedinskii was assistant editor of the periodicals
Radiobiologiya, Byulleten' eksperimentarnoi biologii i meditsiny [Bulletin of Experimental Biology and Medicine],
a member of the editorial staff of the "Radiobiology" section of the Great Medical Encyclopedia, a member of the
editorial council of the Fiziologicheskii zhurnal SSSR, and a member of the editorial staff of Excerpta medica.
Andrei Vladimirovich also acted over a protracted period as part of the editorial staff of our Atomnaya tnergiya.
From 1955 to 1958, A. V. Lebedinskii was the representative of Soviet radiology in the Scientific Committee
on Atomic Radiation of the U. N. In his work in the U. N. he actively defended humanistic policies, on behalf of our
government and expended many efforts and energies in the cause of the struggle for cessation of nuclear-weapons
testings and for the abolition of nuclear weaponry.
457
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
The- socio-political arid scientific activities of A. V. Lebedinskii were highly esteemed by the Soviet govern-
ment. He was awarded two Orders Lenin, the Order of the Red Flag, the Order of the Red Star, two Orders of the
Labor Red Flag, and miscellaneous medals.
A. V. Lebedinskii was not only an outstanding thinker and scientist, and but also a highly principled splendid
and charming individual, possessed of inexhaustible optimism and fervid energy.
A. V. Lebedinskii died in the flowering of his creative forces. The glorious memory of this outstanding Soviet
scientist and Communist will always retain a warm place in the hearts of the Soviet people.
458
A Group of Comrades
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
RADIATION STABILITY OF VITRIFIED RADIOACTIVE PREPARATIONS
(UDC 539.12.04)
F. S. Dukhovich and V. V. Kulichenko
Translated from Atomnaya Energiya, Vol. 18, No. 4,
pp. 361-367, April, 1965
Original article submitted March 12, 1964
Data are cited on the effect of ionizing radiations on the chemical stability of vitreous materials in-
tended for radioactive waste disposal. It is shown that exposure in air leads to a change in the chemi-
cal composition of the surface of specimens as a result of heterogeneous radiation-chemical reactions
involving the solid preparations and components of the air. Suggested reaction products are carbon-
ates, hydroxides, and nitrates
A good deal of research of singular importance for the solution of problems encountered.in the safe disposal
of radioactive wastes has been devoted to the production of fused vitrified preparations containing uranium fission
products. Vitrified- preparations are also in wide use as 8-radiation sources. The excellent chemical and thermal
stability of these materials provides grounds for considering them safe for use in waste disposal operations [1-6].
But the radiation stability of fused concentrates and the effect of ionizing radiation on the behavior of radio-
active isotopes present in them have not received adequate study. It has been mentioned in some papers [7, 81 that
irradiation of vitreous materials of this type with Co60 gammas does not result in any substantial modifications in
the structure of the preparation irradiated, nor in its mechanical strength of chemical stability. No effect of 8-radi-
ation or y -radiation on the chemical stability of conventional glasses has been discovered either [9]. Only in one
paper [10] has any increased solubility of glass after irradiation been reported, this being a case of glass exposed to
radon alphas.
With the object of studying the effect of irradiation on the properties of vitreous materials recommended for
immobilizing radioactive wastes, preparations were made by fusing dessicated and calcined residues of solutions
simulating radioactive wastes in composition with various fusing additives. Selected fusing additives.for use in this
context were silicon dioxide, soda, boric acid, fluorspar, etc. Some of the preparations contained uranium fission
products and featured specific activities as high as 2.6 Ci/g. Preparations containing no radioactive isotopes were
irradiated on a Co60 source at an absorbed dose rate of 10, 50, 100, and 200 rad/sec. Either fused materials previ-
ously pulverized to a specified fraction, or preparations coated in melt form on a substrate with a known surface,
were subjected to irradiation. In contrast to other investigations [9], the specimens were not subjected to any ad-
ditional post-irradiation machining or other mechanical treatment, and were not rinsed, prior to the solubility de-
termination. The chemical stability with respect to distilled water was determined at 30?C by the method of meas-
uring the electrical conductivity of the solution, and was expressed in terms of the damage depth 6 (A) of the speci-
men by water in a specified time. The value of 6 was found from a familiar formula [11].
The fused materials studied are vitrified bodies whose natural surfaces were homogeneous as a rule. Surfaces
of the fused preparation obtained in reflected unpolarized light and magnified 550 times appear in Fig. 1. Begin-
ning with an absorbed dose ?5.107 rad, there appear on the external surface of ireparations Irradiated in air by
Co6? gammas some small dark spots visible under the microscope, with size and concentration increasing with the
absorbed dose. The internal layers of the preparation undergo no marked changes in the process. The corrosion
products are fairly stable and do not change once the irradiation is terminated. However, they are weakly associ-
ated with the preparation and may be removed to an appreciable extent by simply using a soft cloth, or partially
by playing water on the preparation (cf. Fig. 1). After rinsing the exposed preparations with water, we discovered
the presence of nitrate ion in the solution, as well as a content of alkali and alkali earth metals greater than that
found in an unexposed preparation. Preparations containing Sr90?>Y9? with specific activity of 20 mCi/g suffer
similar changes when exposed to internal 8-radiation (Fig. 2).
459
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Fig. 1. Surface of fused preparation of com-
position B (3310 Si02, leo 13203, leo Na20,
and 47% mixture of oxides CaO, Fe203, Cr203,
MgO): a) after being left in air for six months
without exposure; b) after exposure to Co" y -
radiation, D=5 ? 107 rad; c) same, D =5 ? 108 rad;
d) same, D =1.3- 109 tad; e) same as d, but after
wiping with dry cloth; f) same as d, but after
treating with water for 15 min at 30?C; g) unir-
radiated preparation after exposure to moist
atmosphere for six months.
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
8, A
50
80
40
30
20
10
50 120
180 240 300 c min
Fig. 3. Variation in chemical stability of preparation of
composition B in response to exposure to Co89 gammas;
1) unirradiated; 2) D= 4- 108 rad; 3)D =7 ? 108 rad; 4)D=
1.3 ? 109 rad.
Fig. 2. Surface of fused preparation of com-
position B, specific activity 20 mCi/g, tagged
with Sr99-*Y99; magnified 550 times. a) 3 h
after preparation; b) one month after prepara-
tion, D = 1.2-107 rad; c) six months after pre-
paration, D = 7.2 ? 107 rad.
The following patterns are typical of the dissolu-
tion of an irradiated preparation: _a sharp increase in
the rate of dissolution corresponding to increased ab-
sorbed dose for the first 5 to 10 min of contact with
water, and a decrease in the rate of dissolution com-
pared to that of unirrathated specimens in the subse-
quent period (Fig. 3).1 The difference between the
damage depth in irradiated and unirradiated specimens
5 min after dissolution commences is taken to ,be the
magnitude of the radiation effect. This magnitude de-
pends on the absorbed dose and is independent of dose
rate in the 10 to 200 rad/sec range (Fig. 4).
The greatest degradation is that suffered by prep-
arations containing a heightened amount (over 10-20%)
of sodium oxide or boron oxide. Even at doses of 5 ?
108 rad, specimens enriched with boron oxide are readily
covered with a flaky crystalline scale (Fig. 5) whose
thickness increases with increased absorbed dose. X-ray
structural analysis has revealed the layer of radiation-
induced degradation in a fused preparation with an en-
hanced content of boron oxide and sodium oxide to con-
sist principally of H3B03 and NaNO3. The greatest stability is that exhibited by preparations containing no boron
oxide ot sodium components (or else containing these only in slight quantities); even at doses of 108 to 109 rad the
change in the surfaces of such specimens is comparatively modest (Fig. 6).
No changes in the chemical stability of the preparations are induced by irradiation in vacuum, nor do any
visible structural-damage effects appear in the surfaces of the preparations.
1In all the diagrams T indicates the time required for the specimen to become dissolved.
461
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
61,A
So
10
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
0 1.108 2 3 I S G 7 8 P, rad
Fig. 4. Magnitude of radiation effect plotted vs
absorbed dose at dose rate indicated in rad/sec:
A) 10; 0)100; x) 200.
Fig. 6. Surface of fused preparation consisting of
35% Si02, 9% A 1203, 4% Ti02, 1% CaF2, and 51%
of a mixture of the oxides CaO, MgO, FeO, Fe203,
Cr203 (550 times magnification): a) unirradiated;
b) after exposure to Co" gammas, D =5 ? 109 rad.
Fig. 5. External form (unmagnified) of fused prep-
aration consisting of 20% CaO, 10% Na20, 10%
Si02, 60% B203: a) prior to irradiation; b) after ex-
posure to Cot? gammas; D =1.3 ? 109 rad.
The external appearance of the corrosion products
depends on the chemical composition of the specimen.
Figure 7 shows a 550-times magnification of the surface
of preparations enriched with strontium, borate, or sodi-
um components, after irradiation at an absorbed energy
dose of 6 ? 108 rad. Various crystal formations stand out
on the surface of the preparations, clearly as a result of
the irradiation.
The investigations we conducted revealed that de-
gradation of the surface of all the silicate materials in-
vestigated will be observed in irradiation in air. The
absence of damage in the inner-lying layers of the speci-
mens, as well as the absence of corrosion in irradiation
in vacuum, militate in favor of the view that surface de-
gradation is due to heterogeneous chemical reactions of
the fused materials with air constituents. The process of
atmospheric corrosion of glass takes place as a rule even
in the absence of irradiation; but at a negligible reaction
rate. The radiation-chemical reaction products decom-
pose on the surface in the form of a layer only weakly
bound to the basic structure. The thickness of the layer
of degradation products depends on the chemical compo-
sition of the preparation. At identical values of absorbed
dose, the degree of corrosion will increase with any in-
crease in the content of alkali or alkali earth metals, as
well as boron oxide. The special role of the oxides of the alkali metals and of the rare earths is confirmed by the
increased content of these elements in water when the irradiated materials go into solution. The radiation instability
of boron-containing compounds is confirmed by the presence of boric acid in the degradated layer.
Radiation effects may escape notice in an investigation of chemical stability if the preparations undergo gran-
ulation after the irradiation exposure. The fraction of the surface subjected to radiation damage may then become
negligibly small (Fig. 8). Radiation effects may again escape notice if mineral acids in which the solubility of cer-
tain glasses rises steeply are used as solvents. These circumstances appear to account for the discrepancy in the re-
sults we obtained and the data reported elsewhere [8, 9].
462
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
tS,A
ZO
10
9
0
9
0 10 ZO 30 GO 50 'r, min
Fig. 8. Radiation-damage depth for water in a fused prep-
aration of composition B: X) prior to irradiation; 0) gran-
ulated after exposure to a Co89 source, D =1.3 ? 109 rad.
Fig. 7. Surface of fused preparations of several
compositions after exposure to Co" gamrnas,
D =6 ? 108 rad: a) 20% CaO, 10% Na20, 60%
B203, 1010 Si02; b) 35% Chasov-Yarsk clay,
10% Na20, 15% B203 and 40% mixture of oxides
CaO, Fe203, Cr203, MgO; c) 40% Sr0, 10%Na20,
307o Si02, 20/0 B203.
On the basis of experimental findings, we may
suggest the following mechanism to account for radia-
tion-chemical destruction of vitreous preparations.
Oxides of alkali and alkaline-earth metals are corn-
pletely or partially bound to the silicate structures.
Scission of old bonds and the formation of new bonds
may occur in response to ionizing radiation through
the excitation of ,the silicate molecule. The chemical
effect of the irradiation is due to the formation of
active centers in the solid and in the surrounding gas-
eous medium. The process may be schematized as
v H20, CO2, NO2
3MeSiO3 ---> Me (OH)2+MeCO3
+Me (NO3)2-1-3Si02 '.2
As a result of this reaction, the metal loses its linkage to the silicon-oxygen framework of the glass and forms a film
in the form of a hydroxide, carbonate, or nitrate on the surface of the preparation. Corrosion products pass into the
solution, thereby bringing about a pronounced increase in the rate of dissolution of the irradiated preparation (Fig. 9).
The surface of the preparation becomes enriched with silica in the process, and that in turn results in a subsequent
decline in the rate of dissolution of the irradiated specimen causing it to lag behind the rate of dissolution of the
unirradiated counterpart (cf. Fig. 3). From the suggested mechanism we infer that: the higher the rate of dissolution
of the irradiated preparation in the first period (5 to 10 min), the lower the value the rate will have to have later on,
and this is confirmed experimentally. Oxides of metals present in the preparation in forms other than silicate com-
pounds are also capable of participating in the radiation-chemical reactions, it is quite evident.
Radiation damage to the surface results in a greater rate of passage of radioactive isotopes into the solution.
Figure 10 shows the variation in damage depth and leachability of fission products when a fused radioactive prepara-
tion is contacted with water and is left in air for different periods of time after being prepared. The preparation
was made from process solutions; the specific activity of the preparation was 2.6 Ci/g.
2The nitrogen oxides form in air in response to the ionizing radiation [12].
463
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
ctA
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
21
5
4
3
2
0
,r___
1
IA
$
\
_ _
?
0
A
A
0
30
40
50
'C, IT
in
8,A
200
f 50
100
50
3
-20 10 20 30 40 SO m
Fig. 9 Fig. 10
Fig. 9. Ratio of rates of dissolution in water of irradiated (v) and unirradiated (v1) fused preparation of
composition B at absorbed doses: 0) 4.108 rad; A) 7 ? 108 rad; X) 1.3 ? 108 rad.
Fig. 10. Effect of internal irradiation on chemical stability of fused preparation with specific radio-
activity of 2.6 Ci/g, containing the sum of the uranium fission fragments (age of fragments: 300 days):
1) 30 min after preparation; 2) determination repeated 16 days after preparation, D =7 ? 108 rad; la)elut-
ability of radioactive isotopes A, expressed in mCi/cm2 per 1 mCi of specific activity recorded in the
first experiment; 2a) elutability of radioactive isotopes in repeated solubility determination 16 days later.
As a result of the surface character of the radiation,
chemical reactions, only an insignificant fraction of the
40 3 compounds included in the composition of the preparation
and present on its surface will participate in those reactions.
30 In the case of vitreous fragment concentrates, disposed of
in waste burial in the form of metal-clad blocks, for which
the surface to volume ratio is small and the contacting
toV area with air is slight, radiation effects may be inconse-
quential. To minimize radiation effects, it is advisable
to devise preparations containing no boron oxide and no
oxides of the alkali metals, or at most containing them
in only very slight quantities.
In the case where 5-radiation sources are prepared
in the form of an emamel, where the volume to surface
ratio is low, and further an appreciable amount of boron
in order to reduce viscosity, we may expect severe radiation
damage to the preparation. A protective.silicate glass film would hardly act to spare the source from radiation
damage over a protracted span of time, since even chemically stable glasses fall victim to surface corrosion when
bombarded by ionizing radiations (Fig. 11). The correct choice of radiation-stable compositions, as well as an ef-
fective solution to the problem of reliable hermetic sealing, is of paramount importance, then, in the design of safe
sources,of ionizing radiation.
2
20
1
0 10 20 30 40 SO r, mil
Fig. 11. Effect of Cou y-radiation on the chemical
stability of "Druzhnaya gorka" glass: 1) prior to ir-
radiation; 2) D= 4. 108 rad; 3) D= 7 ? 108 rad.
and sodium components are introduced into the enamel
LITERATURE CITED
1. Watson et al., "1958 Geneva Conference of the Peaceful Uses of Atomic Energy," Selected reports
of foreign scientists, Vol.9, Atomizdat, Moscow (1959), p. 187.
464
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
2. P. V. Zimakov et al., "1958 Geneva Conference of the Peaceful Uses of Atomic Energy," Selected reports of
Soviet scientists, 4, Atomizdat, Moscow (1959), p. 247.
3. P. V. Zimakov and V. V. Kulichenko, Disposal of radioactive wastes, IAEA, Vienna (1960), p. 431.
4. P. V. Zimakov and V. V. Kulichenko, Atomnaya energiya, 10, 58 (1961).
5. C. Amphlett, Progr. in Nucl. Energy, III, Process Chemistry, 2, Pergamon Press (1958).
6. M. Goldman et al., Report No. 2004 presented by the U.S.A. at the Second International Conference
on the Peaceful Use of Atomic Energy, Geneva (1958).
7. L. Watson, Glass Ind., 41, 264 (1960).
8. M. Elliot et al.,'Industr, Chemist, 37, 368 (1961).
9. T. Mike, B. Steierm'an, and E. Degering, J. Amer. Ceram. Soc., 43, 405 (1960).
10. I. Vista'', Steklo i keramika, No. 2, 30 (1937).
11. V. S. Molchanov, Zhur. prikladnoi khim., 13, 934 (1940).
12. S. Ya. Pshezhetskii, Mechanism of radiation-chemical reactions [in Russian], State chem. press, Moscow
(1962), p. 148.
465
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
CONCENTRATION OF WATER SAMPLES FOR DETERMINING
THE TRITIUM CONTENT
(UDC 621.039.332)
Ya. D. Zel'venskii, D. A. Nikolaev,
V. S. Tatarinskii, and V. A. Shalygin
Translated from Atomnaya tnergiya, Vol. 18, No. 4,
pp. 367-372, April, 1965
Original article submitted March 23, 1964
A method of concentrating the tritium in water samples by means of distillation under vacuum
(-100 mm Hg) is suggested and experimentally verified. The degree of tritium enrichment is
determined by the change in concentration of stable isotopes (oxygen or deuterium) on distilla-
tion of the sample. Two variations of the calculation are proposed; for the stationary state with
the attainment of the maximum possible degree of fractionation and cessation of enrichment at
the nonstationary state.
In order to determine the tritium content in natural waters, it is necessary to carry out a preliminary concen-
tration of the sample by a factor of 100-1000 or more. It is obvious that the method of preliminary concentration
of the sample should ensure not only sufficient enrichment of the sample in tritium but also the possibility of deter-
mining accurately the degree of enrichment.
For this purpose two methods can be used for concentrating the tritium; an electrolytic and a distillation
method. Electrolytic enrichment takes place as a result of the electrolytic reduction of 250 ml of sample to a vol-
ume of less than 2.5 ml [1].
In [2], the degree of enrichment of the sample in tritium after distillation of the water in a packed column
was determined relative to the volumes of liquid in the feed reservoir, in the vat and in the column packing on the
basis of the tritium material balance equation. However, this method of determining the degree of enrichment is
inaccurate and can be used only for low enrichment; moreover, the distillation process proceeds for a long time.
In this paper, the principal results of an investigation into the distillation of water, undertaken with the pur-
pose of developing a more efficient method of concentrating samples in tritium, are discussed. The proposed method
is based on the possibility of determining the degree of enrichment of a sample containing tritium according to ana-
lytical measurement taking place simultaneously with the change of concentration of the stable isotopes oxygen and
deuterium. The experimental verification of this method of determining the degree of enrichment of water in trit-
ium was taken into account in the investigation and also the explanation of the optimum conditions for distillation
of the water.
Principle of the Method
Suppose we have a distillation column with an upper feed reservoir of sufficiently large volume, filled initially
with water. As a result of the distillation process, the water in the lower section of the column will be enriched in
tritium and other isotopes (deuterium and 018), forming a less.volatile species of water. The partition coefficients
of the isotopes differ but little from unity, and the isotopic effects in the kinetics of mass exchange as a result of
distillation are probably so insignificant that they can be neglected. Then, for given conditions of distillation
izT no = nu,
(1)
where nT, no, and nD are the number of theoretical separation stages (or the number of transference units) with re-
spect to tritium, oxygen and deuterium respectively. Hence it follows that enrichment of the water in the various
466
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Water
Fig. 1. Diagram of distillation appa-
ratus; 1) condenser; 2) reflux gauge;
3) feed reservoir; 4) surrounding jacket;
5) heater coil; 6) distillation tube with
packing; 7) cold trap; 8)coolant; 9)vat
with packing; 10) differential manometer.
g1110
15
10
5 f.
V
3"
50 100 150 t, h
Fig. 2. Enrichment of water in tritium at
atmospheric pressure and loading, kg/(m2 ? h);
1) 320; 2) 640; 3) 945.
isotopes (tritium, deuterium and 018) should take place in a defi-
nite relationship with one another, determined by the value of
the corresponding partition coefficients. For the stationary state,
as a result of distillation with total reflux (without takeoff) on
the basis of Eq. (1) we have by the well-known Fensk equation
log KT -= log Ko 1?gaT ?lo Knl?gaT ,
log ac, g al)
(2)
where KT, Ko and KD are the degrees of fractionation (the ratio
of the concentration in the upper and lower sections of the col-
umn); aT, ao, and aD are the partition coefficients with respect
to tritium, heavy oxygen and deuterium respectively. The val-
ues of aT, ao, and aD are determined with high accuracy. Thus,
if by means of the. appropriate analytical procedures the degree
of fractionation of the water is determined with respect to the
heavy-oxygen isotope Ko or deuterium KD, then 'Eq. (2) enables
the degree of fractionation in the same sample of water to be
found with respect to tritium (KT) for stationary conditions.
If a high-efficiency column is used, then a specified sam-
ple enrichment can be obtained without reaching the stationary
state, which permits a several-fold shortening of the duration of the distillation. According to the data from iso-
topic analyses on the content of deuterium or 018 in several samples taken off over known intervals of time, the
number of theoretical degrees of fractionation n of the column can be determined by the equation for nonstationary
distillation. Using the value found for n and the value for aT corresponding to the distillation conditions, then by
the same equation the required degree of fractionation of water with respect to tritium KT can be calculated.
Experimental Procedure
Concentration of the water was carried out in a packed distillation column with a diameter of 25 mm; the
height of the column packing layer was 1920 mm. Pieces of a spiral 2 x 1.5 mm noncorroding steel wire with a
0.2-mm diameter were used as the packing (a diagram of the apparatus is shoWn in Fig. 1). A weak solution of
tritiated water, to which was added 1% heavy-oxygenated water H2018, was distilled in order to increase the accu-
racy of the analytical control after varying the concentration of 018.
The change of H2018 content at the ends of the column was controlled by a mass-spectrometric method. The
samples of the solution for analysis were subjected to isotopic exchange with a known quantity of potassium carbon-
ate K2CO3, which were decomposed by phosphoric acid; the carbon dioxide gas liberated was admitted to the inlet
system of a mass-spectrometer. A Soviet-produced instrument of the type MI-1305 was used. The tritium concen-
trations at the ends of the column were measured by a scintillation method.
467
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
TABLE 1. Distillation of Water for Various Values of Pressure and Load
-
Pressure
in upper
section of
column;
mm Hg
Load
(reflux
density),
kg/(m2. h)
Time
elapsed
from
start of
experi-
ment, h
Activity due to
content of HTO,
imp/ min
'
Concentration
of H2018, 010
base of
column
top of
column
base of
column
top of
column
50
400
45
7501
181.0
1.5930
0.987
59
8530
178.0
1.5941
0.983
- 17
5381
220.0
1.4515
1.021
30
6562
203.0
1.5380
1.011
48
7402
141.0
1.5989
0.999 '
50
212
-? 65
8023
133.0
1.6530
0.991
80
8760
-
1.7259
98
9840
127.0
1.7289
0.986
20
4440
181.3
1.2531
1.010
50
176
33
6530
180.7
1.5002
0.996
16
2227
-
-
_
17
2270.
-
1.3574
-
100
575
28
2685
120.9
-
-
41
3063
120.0
1.4989
0.980
29
3700
128.0
1.6415
1.1089
100
320
45
4539
109.0
1.8301
1.0763
59
4977
98.0
1.6500
0.9953
72
5947
90.6
1.8185
0.9942
6
660
-
1.2289
-
100
176
19
2177
-
1.4699
-
32
2795
-
1.8339
-
50
4295
2.0293
-
72
5423
91.0
2.0890
1.0193
300
960
13
2800
180.0
1.4500
1.016
25
3611
179.0
1.4604
1.002
300
480
13
2500
171.0
1.4403
1.014
25
3830
170.0
1.4802
1.011
300
192
' 39
4536
172.0
1.5070
1.011
53
5980
168.0
1.5350
1.003
14
1030
180.8
1.2430
1.014
750
945
23
1370
178.3
1.2461
1.014
750
640
33
1790
178.0
.1.2630
1.012
44
2005
176.8
1.3111
1.011
750
320
52
2318
176.0
1.349
1.008
60 ?
2480
175.0
1.372
1.006
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
TABLE 2. Efficiency of a Packed Column for Enriching Water in Tritium and the Heavy-Oxygen Isotope
Distillation conditions
Number of theo-
retical fractiona-
tion stages
EHTS,
cm
Maximum degree
of fractionation*
Degree
' of frac-
tionation
with re-
spect to
tritium
calcu-
lated by
Eq. (2)
pressure in
uppersection
of column,
mm Hg
average
temper-
aturein
column,
?C
loading,
kgAtn2 ? h.)tritium
with
respect
to 018
with re-
spect to
tritium
with re-
spect to
oxygen,
isotope
(K0)
with re-
spect to
(KT)
50
50
50
100
100
100
300
300
300
750
750
750
41.0
42.0
44.0
52.5
53.5
55.0
75.0
75.5
76.5 '
100.0
100.0
, 100.5
Fensk's equation
176
212
400
176
320
575
192
480
960
320
640
945
for the
?84
79
74
105
96
82
115
104
93
112
100
94
mean value
83
80
75
103
95
84
114
104
92
112
101
93
of a.
2.29
2.40
2.58
1.84
2.02
2.31
1.68
1.84
2.09
1.73
1.92
2.03
1.84
1.77
1.69
1.96
1.84
1.69
1.75
1.64
1.55
1.38
1.32
1.29
.
,
308
216
148
400
240
110
112
67.5
40.7
15.2
11.5
9.5
316
219
151
392
280
120
117
59.3
35.6
12.6
9.1
7.1
*Calculated by
TABLE 3. Results of Verifying the Calculation of the Degree
of Fractionation of Water Containing Tritium According to Data
from the Nonstationary Enrichment of Water in 018
Distillation conditions
Distillation,
time, h
Degree of frac-
t ionat ion with re-
spect to tritium
pressure in
upper sec-
tion of col-
loading,
kg/(m2 ? h)
experi-
ment
calcu-
lation
umn, mm Hg
50
212
48
52.5
50
100
176
72
59.5
63
100
575
41
25.5 .
26
300
192
53
35.6
37
750
320
52
13.2
13.3
Experimental Data and Discussion
For the purposes of determining the optimum regime for enriching water in tritium and for experimental veri-
fication of the stated method under different conditions, experiments were carried out in the apparatus described for
distilling water under a pressure of 50-750 mm Hg over the range of loads 176-960 kg/(m2 ? h). The experimental
data obtained are given in Table 1.
It was assumed for the calculations that distillation of two binary mixtures takes place: H20"-H20'6 and
HTO-H20, since the presence in the water of small quantities of other isotopic species of water has practically no
effect on its properties.
469
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
K018
1.75
1.5
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
1.25
0 50
\I
2
100 150 200 t, h
Fig. 3. Enrichment of water in the heavy-oxygen
isotope 018 at a pressure of 100 mm Hg and a load-
ing, kg/(m2 :h): 1) 176; 2) 320; 3) 575.
Ica?
300
200
100
2 ,
0 200 400 600 760
p, mm Hg
Fig. 4. Dependence of the maximum (stationary)
degree of fractionation of water with respect to
tritium on the pressure for a loading, kg/(m2 ? h):
1)176; 2)320; 3) 575.
Because the stationary state was not attained during
the time of the experiment, the calculation was carried out
according to the nonstationary distillation equation proposed
by Berg and James [3]; a calculation according to the equa-
tions of S. I. Babkov and A. M. Rozen [4, 5] gives similar
results.
The values for the partition coefficient ao for the sys-
tem H2018?H201 were taken from [6]. According to data
from our laboratory [7], the partition coefficient of the iso-
topes hydrogen-tritium for the distillation of water (system
H20?HTO) over the temperature range 40-100?C is equal to
38,80
log ar ----- 0.0935 .
(3)
In order to calculate the column efficiency the mean value
of a at the temperature of the upper and lower sections of
the column was used.
The results of some of the experiments and calculations
which were carried out are shown in Figs. 2 and 3; the ex-
perimental data are shown by points and the curves are drawn
according to Berg and James' equation [3] for the values
found for the number of theoretical degrees of fractionation
of the column. It can be seen from the figures that the ex-
perimental data agree well with the calculated data. Simi-
lar graphs were obtained for other distillation conditions.
The equivalent height of a theoretical stage, EHTS,
and the stationary degree of fractionation for operation with-
out takeoff (K =all) were calculated from the value found
for n by the method described above. The results of the
calculation are presented in Table 2. It can be seen from
the table that the values for the number of theoretical de-
H018
grees of fractionation, calculated by the change in concen-
tration of tritium and 018, agree amongst themselves, i.e.,
the experimental data confirm the validity of Eq. (1) and,
consequently also, Eq. (2) which results from it.
/.7 A comparison of the degree of fractionation of water
1.6 with respect to tritium, calculated by the proposed method,
with the experimental values is presented in Table 3. It
can be seen from the table that there is satisfactory agree-
ment between the calculated and experimental data for
1.3 stationary as well as for nonstationary distillation. It is oh-
0 vious that similar results could be obtained if, for determin-
ing the degree of fractionation of water with respect to trit-
ium, in place of the data concerning the change of concen-
tration of the oxygen isotope the data from the isotopic anal-
ysis of the water with respect to deuterium are used.
If it is necessary to increase the concentration of 018
or deuterium in order to increase the accuracy of the iso-
topic analysis, then a known quantity of heavy-oxygenated water or deuterated water respectively can be added to
the sample of water to be analyzed, prior to its distillation.
The results of the experiments which were carried out enable conclusions to be drawn also concerning the
optimum distillation conditions. The relationship between the degree of fractionation of water with respect to the
1.9
1,8
2 1
1,5
1,4
3
200 400 600 760
p, mm Hg
Fig. 5. Dependence of the stationary degree of
fractionation of water in the oxygen isotope on
the pressure for a loading, kg/(m2 h): 1) 176;
2) 320; 3) 575.
470
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
100
50
TABLE 4. Characteristics for Enriching Water in Tritium by Distillation
in a Packed Column at a Pressure of 100 mm Hg
? Process parameters
Specified degree of fraction-
ation of water with respect
to tritium
100
500
1000
For stationary state of column:
Number of theoretical de-
grees of fractionation
80
105
115 '
Height of packed portion
of column, cm
185
242
265
Degree of fractionation
of water with respect
to 018
1.64
1.93
2.04
Degree of fractionation
of water with respect
to deuterium
58
200
340
Operating time of appa-
ratus, h*
500
600
650
For nonstationary distillation:
Number of theoretical de-
grees of fractionation
100
130
150
Height of packed portion
of column, cmt
190
247
285
Time during which a speci-
fied enrichment is
achieved, ht
120
140
160
Degree of fractionation
of water with respect
to 018
1.9
2.0
2.36
Degree of fractionation
of water with respect
to deuterium
30
.,
?
200
650
*For column operation with a loading
of 575 kg/(m2 .h).
t For column operation with a loading of 180 kg/(m2 Oh).
200 400 600 760
p, mm Hg
Fig. 6. Dependence of the degree of fraction-
ation of water in tritium on the pressure, for
nonstationary distillation and a loading of
176 kg/(m2 h).
heavy isotopes and the pressure for the stationary state of the
column, operating with total reflux (when the maximum frac-
tionation K is achieved) is given in Figs. 4 and 5. It can be seen
from the figures that the maximum degree of fractionation of
water for tritium as well as for the heavy-oxygen isotope at dif-
ferent loadings is observed at a pressure of 100-120 mm Hg, (up-
per section of column). The presence of an optimum pressure,
corresponding to the maximum degree of fractionation, results
from the nature of the dependence of a and the EHTS on the
pressure (temperature) and it is typical for distillation in columns
with efficient packing [8, 9]. The same interval of pressure can
be assumed optimum in the case of nonstationary distillation
also (Fig. 6). For identical duration of distillation, the maxi-
mum degree of fractionation of water with respect to tritium
(and also with respect to the other isotopes) is achieved under a
pressure of 100-120 mm Hg.
471
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
On the basis of the data obtained for enriching water in tritium, distillation under a pressure of 100-200mm Hg
(in the upper section of the column) can be recommended. The pressure used by Smith and Rawson [2], equal to
40 mm Hg, is not the optimum pressure.
It can be seen from the experimental data obtained that a significant relationship is observed between the
degree of fractionation of water with respect to the different isotopes and the loading (reflux density or vapor rate)
of the column. By decreasing the loading, as a consequence of the reduction of the EHTS (other conditions being
equal) and increase in the enrichment is achieved. However, the stationary state is considerably more rapidly es-
tablished at high loadings. Therefore, if enrichment is carried out under nonstationary conditions, it is advantageous
to operate at small loadings, whereas for achieving a stationary state it is more advantageous to carry out the dis-
tillation at higher loadings. In the present project the reflux density amounted to 170-200 kg/(-1-12 ? h).
The parameters of the column necessary for achieving different degrees of fractionation of a sample with re-
spect to tritium are shown in Table 4, and also the enrichment in the stable isotope 018 and deuterium obtained with
it. Data are presented for two possible variants of the project: the stationary state with the attainment of the maxi-
mum possible degree of fractionation K and the cessation of enrichment at the nonstationary state. Table 4 is com-
piled for the following distillation conditions: the pressure in the upper section of the column is 100 Mm Hg, the
packing is of the same type as in this project, the loading of the column for operation up to the stationary state is
575 kg/(m2?h), and for.nonstationary distillation it is 180 kg/(m2 ? h).
It follows from the data presented in Table 4 that if the attainment of the maximum possible enrichment for
a given column (stationary state) is abandoned, then it is possible to considerably shOrten the dutation of the distill-
ation by a small increase of the column height. However, as a result of this the calculation of the degree of frac-
tionation is made more complex, since in place of the simple Eq. (2), the more complex relationships of nonstation-,
ary distillation must be used.
Conclusions
A method for enriching samples of water in tritium is proposed and proved experimentally by means of dis-
tillation under vacuum with the determination of the degree of fractionation on the basis of isotopic analysis of the
distillate with respect to the heavy isotope of oxygen or deuterium.
It is shown that the maximum degree of fractionation is achieved by distillation under a pressure of 100min Hg.
LITERATURE CITED
1. H. Ostlund and E.. Werner, Tritium in the Physical and Biological Sciences, 1, IAEA, Vienna (1962), p. 95.
2. D. Smith and D. Rawson, Tritium in the Physical and Biological Sciences, 1, IAEA, Vienna (1962), p. 105,
3. P. Berg and G. James, Chem. Engng. Progr., 44,307 (1948).
4. S. I. Babkov, N. M. Zhavoronkov et al., Kernenergie, 5, 219 (1962).
5. A. M. Rozen, Theory of isotope separation in columns [in Russian], Atomizdat, Moscow (1960).
6. 0. V. Uvarov, N. M. Sokolov, and N. M. Zhavoronkov, Kernenergie, 5, 323 (1962).
7. Ya, Zel'venskii et al., Atomnaya tnergiya, 18, 46 (1965).
8. 'Ya. D. Zel'venskii, A. A. Titov, and V. A. Shalygini Khim. proni-st', No. 2, 116 (1963):
9. A. A. EfternoV and Ya. D. Zei'venskii, Khim. se, No. 3, 201 (1964).
472
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
URANIUM AND ARSENIC IN THE HYDROTHERMAL PROCESS
(UDC 549:533.495+553.497)
V. E. Boitsov and T. M. Kaikova
Translated from Atomnaya tnergiya Vol. 18, No. 4,
pp. 373-378, April, 1965
Original article submitted May 30, 1964
The authors give a brief description of the structural characteristics of hydrothermal uranium deposits,
together with data on the mineral composition of their ores. The deposits are of the uranium sulfide
type, mineral formation being a multistage process. The distinctive feature of the ore composition is
extensive occurrence of native arsenic, which forms a paragenetic association with pitchblende. An
analysis of the mineral composition of the ore and of the trends of paragenesis formation gives reason
to suppose that there was a high arsenic concentration in the hydrothermal solutions, from which pitch-
blende was deposited at low temperatures and at shallow depths.
Industrial uranium deposits have a wide range of mineral associations. The distinctive feature of some deposits
is extensive occurrence in quartz and carbonate, veins of arsenic-containing minerals associated with the main ura-
nium minerals, i.e., pitchblende and coffinite.
The geochemical characteristics of uranium deposition from arsenic-containing hydrothermal solutions have
mostly been based on studies of arsenide or five-mental deposits, in which pitchblende is present with nickel and
cobalt diarsenides, or (far less frequently) sulfoarsenides.
Arsenic is a particularly-useful element for a geochemical analysis of mineral associations because of its wide
range of valence: As2--As1--As??As2+?As3+?As5+. Arsenic may be present as anions (in reducing conditions, in
which nickel and cobalt diarsenides, arsenopyrite, cobaltite and gersdorffite are deposited from the solutions), or as
cations (in more oxidizing conditions, in which realgar, orpiment, arsenic-containing fahlerz and enargite are
formed). Replacement of a (pitchblende-cobalt and nickel diarsenides) paragenetic association by a (pitchblende-
sulfoarsenides) paragenetic association therefore indicates an increase in the solutions redox potential. This increase
is within the range corresponding to transition of As2,- to Asl- and does not impair uranium deposition from these
solutions.
Present concepts of uranium and arsenic behavior in the hydrothermal process can be expanded by the new
data obtained by the authors in a study of uranium sulfide deposits containing pitchblende and (presumably) coffinite
in association with arsenic and arsenic-containing minerals. The characteristics of the geological structure of two
deposits, and their mineralization, are summarized below.
Structural Characteristics of the Deposits
The region is composed mainly of Cambrian rocks contorted into a steep Caledonian anticline. Major tectonic
zones controlling granite massifs are traced in the axial part of the anticline. The deposits are associated with rocks
of a Cambrian effusive-sedimentary Cambrian complex and lie in the east and west exocontact zones of one of the
massifs (composed primarily of granodiorite). This massif has breached and changed the Cambrian rocks and is over-
lain transgressively by Devonian red sandstone. The granite has been breached by alaskite granite stocks. In the re-
gion of the deposits the dike rocks are represented by aplite: pegmatite, alaskite granite, plagiaplite, dioritic and
diabasic porphydrites. The latter occur very extensively around the western exocontact of the granodiorite massif,
whereas aplite dikes are predominant around the eastern exocontact.
The region contains major tectonic zones oriented in a submeridional direction conformably with the eastern
and western contacts of the granodiorite massif.' The zones are up to 200-m thick and consist of several geosuturec,
in echelon or virtually parallel.
473
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
.
Mineral 'IP.
Mineral formation stage
. , .
first
(quartz-sulfide)
-....
second
(carbonate-sulfide)
third '
(arsenic-pitchblende)
Quartz
Potash feldspar
Chlorite
Magnetite,
Rutile
Pyrite
'Arsenopyrite
Pyrthotine
Molybdenite
'Chalcopyrite.
Marcasite
Sphalerite
Tennantite
Calcite ..
Serpentine
Galena
Pitchblende '
Native arsenic
Cobaltite
Gendorffite
Niccolite
Smaltite-Chloanthite .
Safflorite7rammelsbergite .
Realger
Orpiment
ted
:
black ?
white
1111....m?Omb
ins
d.....
?
.....--
?M?11
?
tomows...?
?..
?
=I
?????
11111111.1111111111
.
?
?.
?
?..,
.......
Tectonic crushing .'ha,
....
?
ore structures and2.)
textures. confirming mineral
Mediation in stages, .
eTypical
T xtures:
c hypidiomor-
.62 phic -granu-
t; tar. corrosion
7.. poi kilitic.
r.
so
4 ;Textures:
T. r,
.4..?
.
-241
,..
..0.oc't.'
c..-
.6 .
a 2. er 2
2 0 c 0
?
'Intersection of brec-.
ciation zone by pitch -
bilende veins ?
Textures: cbllomorphic (pitch -
blende forms nodules in quartz
and native arsenic), g,ranular.
substitution and cataciastic.
Typical and most widely oc-
cursing Mineral associations.
Quartz-chalco
p rite4nolyb-
dent te -amend-
p rite.
Calcite-arseno-
pyrite: calcite- .
sphalerite -galena-,
tennantite.
Qua rtz-pitehblende , pitcliblende-
native arsenic: pitchblende-sinaltite-
chloanthite: pitchblende - safflorite -
rammelsbergite: pitchblende -calcite..
Fig. 1. Mineral formation sequence in the deposit in the western exocontact
of the granodiorite massif. (Arsenic-containing minerals underlined.)
The tectonic zones are older and control the Caledonian granodiorite massifs. Subsequent tectonic movements
overthrust the effusive-sedimentary rocks -onto the Devonian sandstone.
Uranium mineralization is usually absent in the major faults of virtually meridional strike, being mainly con-
centrated in the feather faults.
Uranium mineralization in the deposit associated with the western exocontact of the granodiorite massifsoccurs
in Cambrian limestone penetrated by diabasic porphyrite dikes, whereas in the eastern exocontact deposit it is pre-
sent in the amphibolites at the edge of this massif.
Mineralization Stages and Mineral Segregation Sequence
Three hydrothermal veinlet types of different composition can be distinguished in the western contact deposit.
1. Sulfide-quartz veinlets. Their thickness is generally less than 1-1.5 cm; they contain quartz, potash feld-
spar, chlorite, calcite, magnetite, rutile, pyrite, arsenopyrite, pyyrhotine, molybdenite, chalcopyrite, marcasite,
sphalerite and fahlerz. Pyrite, chalcopyrite and rnolybdenite occur Most frequently in the ore minerals, and quartz
in the vein minerals. The mineral contents vary widely both in the ore and vein minerals. One can therefore find
the following facies varieties of quartz-sulfide veinlets: quartz, potash feldspathic-quartz and chlorite-quartz vein-
lets, containing ore rninerals?mainly sulfides (the amount varying from 3 to 20%;
2. Sulfide-carbonate veinlets of thickness 2-3 mm to 20 cm*. The Calcite and (to a lesser extent) the quartz,
which compose these veinlets and contain an impregnation of ore Minerals, also serve as cement in the brecciation
474
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Fig. 2. Thread-like segregations of coffinite ( ?) (gray) in native arsenic (white) in quartz (dark) (x 1200).
zones (thickness 2.5 m). Ore minerals in these formations are represented by pyrite, arsenopyrite, sphalerite, galena,
fahlerz and (less frequently) chalcopyrite and marcasite. The content of sulfides is usually less than 3-5% and that
of sphalerite and galena reaches 10-15% only in a few veinlets.
3. Arsenic-quartz uranium-containing hydrothermal formations. These are usually fine veinlets of thickness
up to 5-8 mm, single or contiguous at distances of a few meters. One sometimes observes zones (thickness up to
15-30 cm) of branching veinlets, passing over into brecciation zones at some points.
Individual veinlets of thickness up to 8-10 cm, and brecciation zones 1-1.5-m thick, are also found in the
deposit.
The composition of the uranium-containing veinlets includes quartz, calcite and serpentine, pitchblende,
coffinite (?), native arsenic and several arsenides and sulfides (Fig. 1). In most cases quartz and calcite are the
only really important vein minerals in the ore veinlets. Serpentine is found near the contacts with the diabasic por-
phyrite dikes, from which magnesium was apparently acquired. The ore textures and age relationships of the differ-
ent veinlet types indicate three mineralization stages in the deposit's formation. Formation of the sulfide-quartz
veinlets to the second stage, and that of arsenic-quartz uranium-containing veinlets to the third one. In this deposit,
as in the great majority of hydrothermal deposits of sulfide-uranium formation, the ore stage is also the final one.
The diagram of mineral formation sequence (Fig. 1) highlights the extensive occurrence of arsenic -containing min-
erals in the deposit. Arsenopyrite and fahlerz (Chemical analysis confirmed this as tennantite) were formed in the
first and second stages. In addition to these, native arsenic, cobaltite, gersdorffite, nicolite, realgar and orpiment
were formed in the ore stage. Native arsenic is the most widely-occurring mineral; very small amounts of nickel and
cobalt diarsenides are found (mostly near the diabasic porphyrite dikes).
A very characteristic feature of the ore veinlets is the presence of a constant spatial and paragenetic associa-
tion of uranium minerals?pitchblende and coffinite ( ?)?and native arsenic which form co-segregations in fine-
grained chalcedony-like quartz and (less frequently) calcite.
Pitchblende occurs most frequently in ore minerals of the third stage. X-ray analysis of three monomineralic
specimens showed lattice parameters ao of 5.41, 5.42 and 5.43 A respectively. Pitchblende forms two segregation
varieties, probably corresponding to different generations.1
Pitchblende of the first generation consists of segregations of various shape, dispersed in the quartz, the size
varying from thousandths to tenths of a millimeter. In veinlets with the highest uranium contents, fine segregations
lln accordance with Betekhtin's definition [1], by "generations" we mean noncontemporary formations of a mineral
in the same mineralization stage.
475
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
TABLE 1. Interplanar Spacings and Line Strengths
of Native Arsenic
From the east-
contact deposit
From the west
contact deposit
Reference
sample No. 32
.[3]
4
6.34 .
6
6.20
1
6.50
? --
7
6.17
3
(3.50)
4
3.42-
7
3.45
.9
3.17
3.15
8
3.14
3
(3.10)
2
2.98
3
2.81
10
2.73
9,
2.72
10
2.74
. 4
2.50. ?
5
2.50
7
2.51
2
2.25
? 3
2.20
3
2.25
2
2.41
2
2.15
3
2.12
3'
2.04
5
2.10
9
2.04
4
1.948
4
4.89
,5
.1.95
4
1.847 ?
7
1.88,
10
1.867
1
1.840
2
1.85
3
1.837
1 ?
1 .760
5
1.79
8
1.76
4
'1.661
5 '
1:67
8
1.65
1
1.596
2 .
1.62
.3
1.59
7
1.547
8
1.57
10
1.53
4
1.440
4. ?
1.43
5
1.433
2
1.386
1
1.37
7
1.380
3
1.350
2 .
4.35
5
1.363
4
1.196
5
1.21
8
1.195
Note. Conditions as follows: BSV tube; unfiltered
iron radiation; camera diameter 57.3 mm; diameter
of cylindrical sample, 0.6 mm. The line strengths
are based on a scale. of 10.
TABLE 2. Ultimate Analyses of Organic Matter
in Pitchblende
CO2
H20
S
C
H
S
Ash content
Wt. of
sample
mg
mg
%
mg
%
10.350
0,335
0.941
-
O.830.87
-
9.57(192,46
10,895
0.302
0.978
-
0.76 0,80
-
10,20593.66
11,230
0.290
0.882
-
0.71
0.88
-
10.48093.32
72.88
-
-
0.36
-
-
0.49
-
-
79.01
-
-
0.41
-
-
0.52
-
-
75.94
-
-
0.38
-.
-
050
-
-
,
of pitchblende, saturating the quartz, form-in conjuction with native arsenic (or sometimes with coffinite)-a dense
impregnation, to which is attributable the black color of the ore quartz veinlets.
Second-generation pitchblende forms thread-like segregations or veinlets of thickness up to 2-2.5 mm, devel-
oped in or near the contacts in ore veins containing an impregnation of first generation pitchblende. Second genera-
tion pitchblende is also frequently associated with native arsenic. The ratios of these minerals vary over a wide
range. Intimate concretions of their finest segregations, as well as individual nodules, are observed in native arsenic
incrustations.
Microscopic investigation of specimens in native arsenic surrounded by quartz revealed branching thread-like
segregations of a radioactive mineral differing from pitchblende by its lower reflectivity and.the.shape of the segre-
gations (Fig. 2). The diagnostics of this mineral could not be established satisfactorily because these segregations
476
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Mineralization stage
Mineral
(quartz
first
-sulfide)
1
sccond
carbonaic
.
third
( arsenic -pitchblende)
Oxides of quartz
OPal
Pyrite
Carbonate
Chalcopyrite
Sphalerite
Galena
Molybdenite
Marcasite
Arsenopyrite
Pitchblende
Native arsenic
Sraimerite 1
Pyrrhotine
Mineral
Realgar
Orpiment
Quartz
Chalcedony and quartz
Siderite
dolomite,
Dolomite
....,'
calcite
Dolomite
Calcite
Tectonic crushing
1 Typical paragenetic associations.
Quartz-pyrite
Quartz -dolo-
mite -chalco-
PYtite -pyrite
Dolomite -sphal-
erite -galena
Dolomite-
molybdenite
Carbo-
nate-
quartz
Calcite -
marcasite
Dolomite -pyrite
Dolomite -pitchblende -arsenic .
Quartz-dolomite-arsenic -sphalerit
Chalcedony -dolomite -pitchblende
arsenic
Quartz -calc ite -arsenic: calcite -
realgar
Quartz-calcite -pyrite
Fig. 3. Mineral formation sequence in the east -exocontact deposit.
were very small. However, from their optical and physical properties they may be attributed to coffinite. These
thread-like segregations of coffinite ( ?) are very similar to those noted by Yu. M. Dymkov [2] in ore deposits of the
Erzgebirge.
Like first-generation pitchblende, native arsenic forms very small segregations (frequently barely discernible
under the microscope) in quartz, occupies the space between the nodular pitchblende segregations and coffinite ( ?)
threads, and is itself nodular in some places; it forms vein-like segregations in the quartz and incrustation-fringes
at the contact of the ore veinlets. Some of these fringes have complex structures: arsenic forms dendrites oriented
perpendicular to the contact. In some cases the thickness of the lens-like segregations and veinlets reaches 3-4 mm.
Finally, arsenic is also found in pitchblende itself, forming accumulations of nodules a few thousandths of a milli-
meter in size. Submicroscopic arsenic segregations very probably frequently "saturate" the pitchblende and quartz
segregations.
The hardness and reflectivity of the large native arsenic segregations vary, most probably as a result of the
different degrees of saturation of arsenic by the smallest pitchblende and coffinite segregations. The arsenic is fre-
quently markedly anisotropic, which enables one to establish the fine-grained structure of its aggregates. Having a
cryptocrystalline structure, it is sometimes isotropic.
The powder patterns of native arsenic show all the principal lines of this mineral (Table 1).
Formation of the deposit in amphibolites of the east exocontact zone also took place in three stages (Fig. 3).
Sulfide-quartz veins distinguished by the presence of white or grayish-white mixed quartz and pyrite, were formed in
?
:477
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
the first stage. The first-stage veins are intersected by carbonate and carbonate-quartz veinlets containing uranium
minerals. Veinlets composed of carbonates, the composition of which varies from siderite to calcite, were formed in
the second stage. Marcasite fringes are sometimes observed at the selvages of the calcite veinlets.
Veinlets formed in the second stage are associated with the same fractures, but those composed mainly of
siderite are earlier formations and are intersected by calcite veinlets (these do not go beyond the siderite veinlets).
Spectral analysis showed that siderite and calcite contain the same characteristic elements (admixtures)?arsenic
and barium.
Ore veinlets consisting mainly of carbonate and quartz were formed in the third stage. Pitchblende and native
arsenic are characteristic and widely occurring ore minerals in the third-stage veinlets, but carbonate-quartz vein-
lets with symmetrical-banded structure and without ore minerals also belong to this stage.
Third-stage veinlets have a more complex structure than the others, brought about by repeated intrastage
movements in which minerals of the early generations were crushed and their fragments cemented by minerals of
later generations. At the same time, some minerals were dissolved and redeposited.
Pitchblende is the deposit's principal uranium mineral of practical importance. X-ray analysis revealed a
uraninite structure. Chemical analysis showed that its organic matter content is negligible (Table 2). It dissolves
in hydrochloric acid, forming a silica gel.
Pitchblende sometimes forms very small segregations in the quartz veinlets. Quartz containing a dense im-
pregnation of pitchblende becomes dark.
In addition to quartz, brown spar, native arsenic and occasionally galena, sphalerite and molybdenite are found
in paragenetic association with pitchblende. In brown spar, as in quartz, pitchblende forms a fine dissemination; the
ratios between them vary in a wide range. The uranium content in the carbonate veinlets reaches 2%.
Carbonate forms several generations. Late-generation carbonate is segregated together with native arsenic and
corrodes the pitchblende. In some places, only relicts of pitchblende (0.01-0.03 mm) are retained in the carbonate
veinlets.
Microscopic examination of native arsenic shows a fine-grained structure (sometimes fine polysynthetic twin-
ning). Its diagnostic properties are the same as the reference specimens. Anisotropy is very weak in some segrega-
tions. X-ray analysis data of native arsenic are given in Table 1.
Native arsenic has an intimate spatial relationship with pitchblende. Early-generation arsenic is characterized
by relatively, large segregations; they are sometimes divided by veinlet-like segregations of pitchblende. Late-gener-
ation arsenic is segregated together with pitchblende, forming a paragenetic association, both of them being observed
as very small modules or plates (some of which require more detailed examination because they resemble coffinite
in their main diagnostic properties).
Together with quartz, native arsenic often corrodes fragments of early pyrite grains; reaction fringes of arseno-
pyrite or realgar, which one might expect to be present as a result of the action of arsenic-containing solutions on
pyrite [1], are not observed at the contact of pyrite and native arsenic. It may therefore be supposed that the pyrite-
native arsenic association is chemically in equilibrium.
Cleiophane and small amounts of pyrrhotine, chalcopyrite, arsenopyrite and molybdenite are also found in
paragenetic assocation with native arsenic.
Our study of the ore composition of deposits, in whose formation uranium- and arsenic-enriched solutions
participated, thus reveals the following characteristics.
1. Mineral formation took place in three stages. A study of the mineral complexes and their segregation se-
quence reveals common features between these deposits and uranium deposits belonging to the uranium sulfide
formation.
2. The most important feature of these deposits is the extensive occurrence of a quartz-pitchblende-native
arsenic association in the ore stage veinlets. This association may be accompanied by coffinite as well, particularly
in ores of deposits located at the west exocontact.
478
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
1
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
3. The solutions from which pitchblende was deposited had initially a high sulfur concentration; somewhat
later and before the end of pitchblende deposition they had a high concentration of sequentially ,oxidizable arsenic.
This is confirmed by the fact that the Asl- (in arsenopyrite) arsenical minerals formed were subsequently replaced
by As? (in native arsenic) minerals, As3+ appearing (in realgar) after the end of pitchblende deposition.
The presence of arsenic-containing minerals like native arsenic and realgar is an indirect indication of the,
shallow depths at which the deposits were formed. It may therefore be concluded that these minerals and pitchblende
were deposited from hydrothermal solutions at low temperatures.
LITERATURE CITED
1. A. G. Betekhtin, In the symp. "Fundamental Problems in the Study of Magmatogenic Ore Deposits"
[in Russian], Moscow Izd-vo AN SSSR (1953).
2. Yu. M. Dymkov, Uranium Mineralization of the Erzgebirge [in Russian], Moscow, Atomizdat (1960).
3. V. I. Mikheev, X-Ray Identification of Minerals [in Russian], Moscow, Gosgeoltekhizdat (1957).
479
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
METHOD FOR CALCULATING THE RADIOACTIVE IMPURITY CONCENTRATION
IN THE WATER AND THE BOTTOM LAYER OF STAGNANT RESERVOIRS
(UDC 621.039.7:628.515)
F. Ya. Rovinskii
Translated from Atomnaya Energiya, Vol. 18, No. 4,
pp. 379-383, April, 1965
Original article submitted May 17, 1964
The present article is concerned with certain trends in the migration and' redistribution of tadioactive
impurities in Stagnant reservoirs after they have been contaminated only once. The capture of the
dissolved impurities by the bottom layer occurs as a result of ion-exchange and molecular adsorption
processes. On this basis, we derived equations describing the variation in the concentration of radio-
active isotopes in dependence on the time of their presence in the water and the bottom layer. The
derived equations make it possible to calculate the impurity percentages in the components of stag-
nant reservoirs.
It was shown in [1] that radioactive isotopes are distributed among the basic reservoir components (the water,
the bottom deposits, and the biological Mass) in such a manner that the amount of radioactive isotopes in the bio-
logical mass can be neglected. Consequently, a stagnant reservoir can be considered as a two-component system.
In order to predict the contamination levels of the water and the bottom deposits, we shall consider a reservoir
where the water volume is V m3, the surface area of the bottom layer is S m2, while the average depth is small, not
exceeding 4-5 m. Such a reservoir, which has the shape of a shallow basin, is characterized by intensive turbulent
and convective mixing of the water mass, which leads to the interaction of the entire water mass with the bottom
layer [2].
We shall assume that the radioactive impurity was introduced only once in an amount equivalent to A Ci,
which initially entered only the water mass in the reservoir, so that this amount was instantaneously distributed
throughout the entire volume V. We shall denote by Q(t) the amount (supply) of the radioactive impurity in the
water and by P(t) the amount (supply) of the radioactive impurity absorbed by the bottom layer, which vary during
the time t. Then, we can use the following initial conditions:
t=0, Q0= A, P0=0,
where Q0 and Po are the initial impurity amounts in the water and the bottom layer, respectively. Since the impurity
introduced in the reservoir will be subsequently redistributed only between water and the bottom layer, then, without
considering for the moment radioactive detay, we obtain
A=Q(t)+P(t).
(1)
The capture of the dissolved impurity by bottom deposits occurs as a result of ion-exchange and molecular ad-
sorption processes, and, therefore, in the general case, the change in the amount of impurity in the bottom layer
(Fig. 1) can be described by the following equation:
dP (t)
= p,iQ (t)? 1i2P (t),
di
where ?1 and 112 are constants determining the sorption and desorption rates, respectively.
480
(2)
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
r
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
pes
With a consideration of the initial conditions and Eq. (1), the
solution of Eq. (2) is given by
P (t) _= A (1e?g2(1+ ?t17,2-)-t);
1+---
1-11
Q(t)=A e---112(1-1121)t)
LT
P4i
(3)
(4)
pand q(t) the mean values of the radioactive impurity's surface density in
ration in water:
P (t) A 1 1t
S = S (1? e u'20+ 11.2 I ),
1+ IL2-
(5)
) (t) A 1 112 -1,2(1-1--)t
== ? e (6)
V V 4 112 P.2
[Li
- these quantities for t --> 00, we reach the conclusion that, in the course of
ourity is established between the water and the bottom layer in the reservoir:
where zr and 17) are the equilibrium concentrations in the water and the bottom layer, respectively.
By using Eqs. (1) and (7) and introducing a correction for the decay, we finally obtain:
A ,
p (t)=_- p (1? e-112A-ps' et,
q (t)
A
-
ql A- e
qV t
v
(7)
(8)
(9)
Consequently, if the values of A, V, and S are known, the prediction of the impurity concentrations in the
water and the bottom layer consists in determining Cr, and I12 (or Ai).
We shall assume that we know q1 and q2-the results of measurements of the volume concentration at the in-
stants of time t1 and t2. Then, after eliminating p2, we obtain
(42?F) V (qt cr) V s't2/ti
A --qV A? q-V )
TABLE 1. Hydrochemical Composition of the Lake Waters
(10)
1"--
Reservoir
Composition, mg/liter
HCOil
Nei
K*1
Mg42
Ca*2
CI-1
SO2
Si02
Total
First
105.3
9.3
56.4
25
57.1
354
125
4.7
759
Second
943
45.6
110.5
8.5
866
1282
127
3
3493
481
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
log q(t)
-2.5-
Se
-15 - reservoir
coand rese.rv: ? ., ?
-(0 -
?
-(5 ? ?
-2.0
-2.5
-3.0
-3.5
-4.0
-4.S
First reservoir
0 5 10 15 20 25 30 35 40 45 SO 55 60 65
Time, months
Fig. 2. Comparison between the actual (*) variation (1) of the
concentration of isotopes in the Water and the calculated (0)
curve (2) of Sr" concentratiOn in the waters of the experimental
reserVoirs.
log q(t) log q(t)
- LS
-1.0
-2.5
-3.0
43
log 1(t1
-.10
pe ? Rti]
TIZu= 18 days
t? 11
0 10 20
-4.0
-45
Tce= 24 days TR,. 105 days
,.-50 , , , , , .
30 40 50 60 70 80 9O 1.00 0 60 120 180 240 300 36 0 420
Tirne,-days Time, days
a b
Fig. 3. Determination of the half-life of Rul" and Cei44 in the first (a) and the second (b) reservoirs.
It is convenient to express q in explicit form if t2 2. In this case, .
ti
Aq2?V
(11)
? v (14-q2--") In A-91/
' - Ati (A + V .72-2Vq1) (q q2) V
( 12)
In praCtice, it is, of course, advisable to perform More than two measurements of q(t), while the times of mea-
stretilent should be chosen in the form of terms of a geometric progression with a denominator equal to 2,
482
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Second reservoir
-3.0
First reservoir
?
0 S IS ZO 25 30 Is 40 45 SO 55 60 GS 70
Time, months
a
log pa)
-2,0
-2.5
-3,0
-4,0
-5,0
-1,5'
-2.0
-1.5
-3.0
-3,5
-4,0
-4.5
-5,0
Second reservoir
First reservoir
0 5 f0 f5 ZO 25 JO 35 40 45 50 55 60 65 70
Time, months
Fig. 4. Calculated concentration curves for radioactive isotopes in water (a) and in the bottom layer (b) of the
experimental reservoirs. 1) Sr"; 2) Rul"; 3) Ce; 4) sum of three isotopes.
Thus, the above scheme of redistribution of the radioactive impurity in a stagnant reservoir makes it possible
to calculate the percentages of the radioactive impurity in the reservoir components at any time after a single in-
stance of contamination:
We shall apply the derived equations to the case of artificial contamination of two stagnant reservoirs by a
mixture of Sr", Rti106 and Ce 144 isotopes.
Eutrophic lakes, one with a surface area of 11:3 km2 (the first reservoir) and the other with a surface area of
4.5 km2 (the second reservoir), were Used for the experiments: The lakes have shallow, saucershaped bottoms. They
have large silt deposits, which have completely smoothed out the initial bottom relief. The shores are partially
overgrown with reed; there is an abundance of submerged plants: milfoil (Myriophyllum), aquatic plant (Cerato-
phyllurn demersum), and pond weed (Potamogeton). Good conditions for the development of the biological Mass
prevail in the lakes: the summer temperatures are high, there is a sufficient amount of oxygen and organic Matter
in the water, the water is well illUrninated throughout its depth, etc. The hydrochernical composition of the lake
waters is given in Table 1.
ThereStilts of the measurements of the concentration of iSotopes in the Water, which were performed over a
period of five years after the isotopes Were introduced in the lakes, were made available to us. The total activity of
water Sainples was determined during the first three years, and radiochemical determinations of Sr" were performed
during the next two years: The results of these experimental observations are given iP Fig. 2.
483
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
TABLE 2. Half-Life T of Some Isotopes in Stagnant
Water Reservoirs
Reservoir
Average
depth
m
T, days
Y90
Cel 44
Ru106
SOO
First
Second
1.0
1.9
1.1
10.9
1.0
24
18
105
116
180
Moreover, the constants A, V, S. q, andIT, (besides 112) were found experimentally for calculation by means of
the derived equations. The 112 constant for Sr" was determined by means of Eq. (9), where the q(t) and t values found
from curves 1 of Fig. 2 were substituted. It was assumed here that, beginning with the third year after the intro-
duction of radioactive isotopes in the reservoirs, the contribution of Rul" and Ce144 to the total activity of water was
small in comparison with the contribution of Sr" because of their radioactive decay and adsorption by the bottcim
layer. This assumption was confirmed by special analyses of the percentages of the above isotopes in the water.
The values of the 112 constant for Sr" for the first and the second reservoir were equal to 1.9 ? 10-4 and 3.9 ? 10-4
day-1, respectively.
Then, on the basis of the known constants, the Sr" concentration for the entire period of time was calculated.
A comparison between the actual variation of the isotope concentrations in the water and the calculated q(t) curves
for Sr" is given in Fig. 2.
The constant 112 for a certain isotope is connected with its half-life in water (T)1 by the following simple
relationship:
112= 0.693 q'
AT
The value of T for Rul". and Ce144 for the first and the second reservoirs can be found by the well-known method of
graphical analysis of a complex curve. The points on the curves for instants of times sufficiently close to to repre-
sent the concentrations of three isotopes, i.e., [Ce+Ru+ Sri. If we subtract the calculated curves 2 from curves 1
(see Fig. 2), the thus obtained difference curves will correspond to the variation in the concentration of two compo-
nents in the water [Ce + Ru] (Fig. 3a and b). The rectilinear section of the curves [Ce+Ru] corresponds to the varia-
tion of [Ru] in time due to migration (since a correction for the decay of Rul" has been introduced here). The slope
of this section of the curve can be used for determining the T value for Rul".
Furthermore, if we subtract the [Ru] straight line from [Ce+ Ru] , we can separate the straight line corresponding
to changes in the Ce144 concentration in water due to migration; the r value for Ce144 can readily be found with re-
spect to the slope of this straight line (see Fig. 3, a and b).
The half-life constitutes the quantitative characteristic of the migration of radioactive isotopes from the water
to the bottom layer in stagnant reservoirs.
The Eq. (9) given above consists of two parts: a certain constant Ff and the variable
A
,
=
A
V
A-0.6931/T.
A-01 -71T
9
e_1-12
It is obvious that r characterizes the rate at which q(t) tends to ;T. The larger the r value, the slower the rate
at which equilibrium between water and the bottom layer is established in the reservoir, and, conversely, the smaller
the r value, the higher the rate at which the equilibrium state is established. Consequently, T characterizes the
equilibrium establishment time, but does not determine the equilibrium concentrations of isotopes in the reservoir.
1The term half-life of an isotope in the water of a stagnant reservoir denotes the time interval during which the
isotope concentration in the water is reduced by one half solely as a result of isotope migration in the reservoir.
484
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Table 2 provides the T values for four isotopes, where the T values for Ce, Ru106 and Sr9? were determined
with respect to the experimental data given in Fig. 2, while the r value for Y9? was found independently with re-
spect to the shift of radioactive equilibrium between Sr9? and Y90 [3].
It is obvious from the table that the T values for Y" and Ce 144 are close to each other for both reservoirs.
This was to be expected, since the above isotopes are chemically similar, and theform of their state in a solution
with pH = 7-9 is the same. Therefore, the processes of their absorption by the bottom layer must also be identical.
Moreover, it should be mentioned that the T values for all the radioactive isotopes were larger in the second reser-
voir than)in the smaller first reservoir.
The found T values make it possible to place the isotopes in order with respect to the rate at which equilibrium
is establiShed in the reservoir: rare earths, yttrium > ruthenium > strontium.
Thus, on the basis of the experimental data, we obtained the constants necessary for calculating q(t) and p(t)
for Sr90, Ru106 and Ce144 in two reservoirs by means of the equations derived by approximating a stagnant reservoir by
a two-component system. The calculation results are shown in Fig. 4, a and b.
The variation of the total concentration of Sr90, Ru106 and Ce144 in water (curves 4 in Fig. 4a) is in fairly good
agreement with the actual behavior of the concentration of radioactive isotopes obtained by measuring the over-all
3-activity of samples (see Fig. 2).
The p(t) curves (see Fig. 4b) have characteristic maximums, the existence of which can also be demOnstrated
analytically.
Thus, the results obtained in calculating the concentration of radioactive impurities in the water and the
bottom layers of stagnant reservoirs are in fairly good agreement with the factual data available to us. The use of
the equations derived also made it possible to determine some other characteristics of the behavior of Sr90, Ru 10 6
and Ce114 in stagnant reservoirs.
LITERATURE CITED
1. A. L. Agre and V. I. Karogodin, Med. Radiologiya, No. 1, 67 (1960).
2. B. B. Bogoslovskii, Limnology [in Russian], Moscow, lad. MGU (1960), p. 84.
3. G. A. Sereda and F. Ya. Rovinskii, Atomnaya Energiya, 14, 326 (1963).
485
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
RULES FOR DEPOSITING (STORING) ARTICLES
Translated from Atomnaya Energiya, Vol. 18, No. 4,
p. 383, April, 1965
Articles of interest to a limited number of specialists are deposited either on the authors' request or in accor-
dance with the decision of the Editorial Board of the journal.
Detailed abstracts of the deposited articles will be published in the journal, while the complete articles will be
kept for five years at the editor's office and sent COD to readers on request. The volume of the abstracts to be pub-
lished should not exceed 1/10 author's sheet (about two pages of typewritten text), while the maximum volume of
the deposited text should not exceed 1 sheet. On the authors' request, diagrams, tables, the basic equations, etc,
can be included in the abstract within the limits of its over-all volume.
- An abstract will be published not later than three to four months after the article has been received at the
editor's office (if the article is deposited on the authors' request) or after the authors have consented to deposit it (if
this decision has been brought by the Editorial Board).
Deposited articles are regarded as scientific publications and are taken into consideration in the defense of
dissertations.
The articles submitted for deposition must be suitable for photographic reproduction, and the text, equations,
tables, etc., must be clear; the drawings must be made on tracing paper, included in the text and pasted on the
paper, and provided with captions.
The price of a single copy of a deposited article is 40 k. In ordering copies of deposited articles, reference
must be made to the registration number of the article given at the end of the abstract.
Orders should be mailed to the editorial office of the journal at the following address: 18 Kirov Street,
Center, Moscow.
486
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
ABSTRACTS OF DEPOSITED ARTICLES
INTENSIVE MUON BEAMS IN THE OlYal SYNCHROCYCLOTRON (No. 8/3155)
(UDC 621.384.611.1)
Yu. M. Grashin, B. A. Dolgoshein, V. G. Kirillov-Ugryumov,
A. A. Kropin, V. S. Roganov, A. V. Samoilov, and S. V. Somov
Translated from Atomnaya Energiya, Vol. 18, No. 4,
p. 384, April, 1965
Original article submitted December 7, 1964; abstract submitted February 6, 1965
A strong-focusing channel for intensive muon beams with a low percentage of the pion admixture was put into
operation at OlYal by the end of 1963. The channel, whose aperture has a diameter of 20 cm, consists of 28 quadru-
pole magnetic lenses and a three-section analyzing magnet.
A negative muon flux of 3 ? 104 sec-1 over an area of 80 cm2 with a momentum of 130 MeV ? c-1 polarization
(70 ? 20%), and a pion admixture of less than 0.4% at the maximum of muon stoppings was obtained from the syn-
chrocyclotron's internal target. Moreover, a negative muon flux of 3.5 '104 sec-1 with a momentum of 280 MeV ? c-1
was produced.
487
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
CONVERSION OF THE 1.5 -m CYCLOTRON FOR THE ACCELERATION
OF MULTICHARGE IONS (No. 9/3127)
(UDC 621.384.611)
V. V. Ba,tyunya, Pai Fu-Wei, G. N. Vyalov,
B. A. Zager, and A., F. Linev
Translated from Atomnaya Energiya, Vol. 18, No. 4,
p. 384, April, 1965
Original article submitted November 9, 1964; abstract submitted February 8, 1965
The 1.5-m cyclotron U-150 is designed for accelerating deuterons and a-particles to an energy of 12 MeV per
nucleon for a magnetic field strength H= 14.1 k0e.
Investigation's of nuclear reactions between complex nuclei constitute the main task of the Nuclear Reactions
Laboratory at OIYaI. For the production of the multicharge ions required in experiments, the U-150 cyclotron was
converted for operation under conditions where such ions are accelerated for an A/Z ratio equal to 2.6-3.2. Accel-
erated NV, CV, 0165, etc., ions with a sufficiently high intensity at an energy of 6-7 MeV per nucleon were obtained
at the terminal radius R= 66 cm as a result of this conversion.
In converting the cyclotron for use under the new operating conditions, it was necessary to increase the maxi-
mum wavelength of the hf oscillator from 34 to 40 m. New chamber lids, whereby the gap was reduced from 210 to
180 mm, were made for this purpose. The hf oscillator was modified accordingly.
The magnetic field was adjusted by shimming at a strength of 16.7 k0e. Radial coils were mounted in the ex-
ternal part of the magnet gap for ensuring the optimum field decay. The main effort was directed toward producing
an internal beam of CV ions with a sufficiently high intensity and bringing it to the terminal radius with a relatively
small intensity drop and a satisfactory orbit shape. An internal C41-' beam with a maximum intensity of up to 30 AA
was obtained for an orbit center shift and a deviation from the median plane of ? 1 cm. The radial spread of ions
on the target at the terminal radius was equal to 5-6 mm. The voltage between the dees was 200-220 kV.
After its formation at the terminal radius, the CV beam was extracted from the cyclotron chamber by means
of an electrostatic deflector with a nonuniform field. Initially, the deflecting voltage was equal to 70-75 kV. It
was then reduced to the minimum value of 35-40 kV by reducing the deflector's radial aperture, increasing the
terminal radius by 10-15 mm, and using the optimum angle of ion entrance to the deflector. After subsequent ad-
justment, the operating deflecting voltage was in the 50-60 kV range. The intensity of the external beam of CV
ions, focused to an area of 1.5 cm2, attains 10 AA. The extraction coefficient is equal to 30-40%.
488
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
DECREASING THE ENERGY OF BEAMS OF MULTI-CHARGE IONS
ON THE 1.5-m CYCLOTRON (No. 7/ 3156)
(UDC 621.384.611)
R. Ts. Oganesyan, G. Indreash, and B. A. Zager
Translated from Atomnaya tnergiya, Vol. 18, No. 4,
p. 385, April, 1965
Original article submitted December 7, 1964; abstract submitted February 3, 1965
The industrially manufactured V-150-1, 1.5-m cyclotron is designed for accelerating light ions (protons,
deuterons, and a-particles). Readjustment of the cyclotron for acceleration of multi-charge ions made it possible
to accelerate CV and 445 ions to energies of 6 and 7 MeV per nucleon, respectively. The magnetic field was
shaped by annular iron shims with Ho = 16.7 k0e; the drop 61-If at the final radius amounted to 2.2%. The require-
ment of increased energy necessitated an increase of the magnetic field to Ho = 17.5 k0e; the drop 6Hi at the final
radius amounted to 3.4%. It was proposed that the additional drop (61-Ff-6Hf = 1.2%) be compensated by means of
an external annular shim.
Magnetic measurements showed that it was possible by this method to adjust the magnetic field to 17.5 k0e,
as required. No decrease in intensity was observed; the energy of the accelerated ions was increased by 10-14% and
amounted to ?7 MeV per nucleon for C.
489
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
INVESTIGATION OF HOW THE AGING OF Co6? RADIOACTIVE IMPURITIES
ON St. 3 AFFECTS THE EFFICIENCY OF CHEMICAL AND ULTRASONIC
DEACTIVATION METHODS (No. 11/2936)
(UDC 621.039.75)
S. M. Kochergin and S. K. Moiseenko
Translated from Atomnaya tnergiya, Vol. 18, No. 4,
pp. 385-386, April, 1965
Original article submitted May 14, 1964; abstract submitted February 20, 1965
In order to obtain repeatableresults, we studied the effect of a number of factors on the degree of Co6? con-
tamination of carefully cleaned and degreased specimens. As a measure of the degree of contamination of the
materials, we used the coefficient of susceptibility (the ratio of the specific activity of the specimen to the specific
activity of the solution, expressed as a percentage). After being contaminated under fixed conditions in solutions
with specific activities of 0.2, 0,4, and 0.8 pCi/cm3, the specimens were washed with distilled water and then dried,
after which the activity of each side was measured at least twice with a B-2 instrument. The results used were aver-
ages of at least four independent measurements.
It was found that the variation of the coefficient of susceptibility of St. 3 and duralumin as a function of the
different factors was similar in nature. There is very little contamination of these metals in strongly acid media.
The maximum deposit of Co6? is observed in almost neutral media, in which the coefficient of susceptibility of St. 3,
after 3 h of exposure at 20?C, was 2-5 times as large as the coefficient of susceptibility of duralumin, depending on
the specific activity of the solution used. In addition to the production of Co6? by an ion exchange mechanism,
there was also deposition of Co6? in the radiocolloidal form. When the contact time was increased to three days, the
repeatability became considerably poorer, owing to "aging" of the solutions and to corrosion processes. Raising the
solution temperature to 60?C increased the coefficient of susceptibility by a factor of 1.5, but if repeatable results
are desired, the contact time should not exceed 10 min.
Specimens of St. 3 contaminated at 20?C, with exposure times ranging up to 2 h inclusive, were subjected to
deactivation at room temperature by an ultrasonic method in a cavitation field (power 1-1.5 W/cm3, frequency
21-23 kc) and a chemical method (on a shaking apparatus). In individual cases the degree of contamination of the
specimens was ten or more times as high as the maximum allowable contamination level of equipment before clean-
ing [1]. For our deactivation media we used the formulas found most effective [21 for the ultrasonic deactivation of
duralumin: 10% H2SO4-m- 15 g/liter KMn04; 10% HNO3; 10% H2SO4-4- 15 g/liter K2Cr307. In these solutions a degree
of deactivation (percentage of the activity removed) of more than 99.5% was reached after 2.5 min, 3 min, and
5 min, respectively, and the average weight loss in the specimens was 0.5%, 2.5%, and 1%, respectively. It was
also shown [2] that when the duralumin contamination time was extended to four days, the ultrasonic treatment time
required to attain the same degree of deactivation was 1.5-2 times as long; the weight losses of the specimens be-
came About twice as large.
It was found that when St. 3 was deactivated in the same solutions by the ultrasonic method immediately after
contamination, more than 96% of the Co6? was transferred to the solution in less than 15 sec. When the specimen
was treated in the shaking apparatus, 90% of the Co6? was transferred into the solution in 1-1.5 min. In order to
reach a degree of deactivation of the order of 99.5%, the total time consumed was 1.5-2 min in the ultrasonic
method and 13-20 min in the chemical method, depending on the solutions used.
Aging the impurities on the specimens for four days led to an increase in the proportion of radioactive sub-
stances more strongly bound to the surface of the metal, which reduced the efficiency of both methods. In the case
of ultrasonic deactivation, the time required was increased by a factor of 2-2.5, and the percentage weight loss of
the specimens by a factor of 2.5; in the c2e of chemical deactivation, the time and the weight loss were increased
490
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
by a factor .of 1.5-2; Under similar conditions, the time required for treatment on the shaking apparatus was 10 or
more times as long as the time required for ultrasonic treatment. The weight losses in nitric acid solutions were
2-3 times as large as in sulfuric acid solutions containing potassium permanganate and potassium dichromate, but
the nitric acid solutions compared favorably with the others with regard to their rate of deactivation in chemical
treatment. On the basis of the experiments we conducted, we recommend the following as optimum compositions
for the deactivation of St. 3 and duralumin: 10% H2SO4+ 15 g/liter KMn04 and 10% H2SO4+ 15 g/liter K2Cr207 for
the ultrasonic method; 10% HNO3 for the chemical method.
LITERATURE CITED
1. Sanitary Rules for Work with Radioactive Substances and Sources of Ionizing Radiation [in Russian], Moscow,
Gosatomizdat (1960), p. 68.
2. S. M. Kochergin and S. K. Moiseenko, Determination of effective compositions for. the deactivation in an
ultrasonic field of duralumin and St. 3 steel contaminated by Co60. Zh. prikl. khim., 38, No. 5 (1965).
491
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
SPACE-ENERGY DISTRIBUTION OF NEUTRONS
IN AN INFINITE ABSORBING MEDIUM (No. 12/2792)
(UDC 539.121.73:539.125.5)
I. A. Kozachok
Translated from Atomnaya tnergiya, Vol. 18, No. 4,
p. 386, April, 1965
Original article submitted January 4,1964; abstract submitted March 5, 1965
The article discusses the stationary problem of decelerating the neutrons from a monochromatic point source
in an infinite medium with an arbitrary variation of the effective absorption cross-sections as a function of energy.
It is assumed that in the center-of-mass system the scattering of the neutrons is isotropic and that its cross section is
independent of energy. In order to overcome the well-known difficulties involved in solving such a problem, the
problem is forMulated in the form of an integral equation whose kernel is the Green's function of the transport equa-
tion for neutrons in a medium where absorption is neglected.
The solution of the integral equation for the neutron collision density was found by the method of successive
integration. The final results for an isotropic source are of the form
To (r, u)= Goo (r, (r, u),
f3uB
p (r, u)=--(1-1 1.5 R2 R )exp [aB (u?I3 ? I (0, u
13u\1 r puD
1 ? 0-02 R
2.5 Pup ) al (?' u) x exp aD (1 )1 X L1 ( 1-- 1 ,
31iR R?)2
R2
I?(0, du' ila((u2 , r242+02u2,
0
where Goo is the space-energy distribution of the neutrons in the medium without absorption; a, B, and D are
parameters (the analytical expressions for these are given in the article); the other symbols are those of the generally
accepted notation. By analogy with the well-known solution of the problem of the distribution of neutrons with re-
spect to energy alone, the function p(r, u) may be regarded as the probability that a neutron will avoid capture when
it has been decelerated to a lethargy of u at a distance r from the source.
Thus, in the general case, the absorption requires a change not only in the energy distribution of the neutrons,
but also in their spatial distribution. The greater the anisotropy of the angular distribution of neutrons in an analo-
gous medium without absorption, the more pronounced will be this redistribution in space.
From an analysis of the resulting formulas and of the numerical calculations performed by means of the formu-
las for dry silica and for a mixture of silica and water (20% by weight), it was found that p(r, , u), the probability of
avoiding resonance capture, decreased monotonically as r increased. Consequently, the effect of the decrease in
neutron density daused by neutron absorption is more pronounced at large distances from the source than near the
source. The dependence on r of the probability of avoiding resonance capture increases with increasing absorption
intensity and also with increasing concentration of hydrogen nuclei in the medium (owing to the increased anisotropy
of the angular distribution of neutrons). In a "dry" medium this dependence is smoother than in a hydrogenous
medium (in the last case it was almost linear). If there is not much absorption, then the depence of p or r in such a
medium is weak and may be neglected in a first approximation.
492
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Thus, the solution found in this study differs in principle from the results yielded by the age theory. The main
difference is manifested in the marked change of the spatial distribution of the neutrons, owing to neutron capture,
which is not taken into account in the age-theory approximation. The age theory is found to be valid in a first
approximation only for media with low nitrogen content and relatively weak absorption. The probability of avoiding
resonance capture agrees with the known expressions obtained from the solution of problems dealing with the energy
spectrum of neutrons integrated over the entire space.
Most chemical elements have the property of intensively absorbing slow neutrons, and this property is utilized
in prospecting for certain minerals (for example, boron). The formulas found in this study are suitable for numerical
calculations and may be useful in examining questions concerned with the theory of neutron prospecting methods.
493
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
REVIEWS OF GENEVA PAPERS
THERMOELECTRIC AND THERMOEMISSIVE CONVERTERS
N. N. Ponornarev-Stepnoi
Translated from Atomnaya Energiya, Vol. 18, No. 4,
pp. 387-390, April, 1965
The physical principles of methods for the direct conversion of heat energy to electricity have been known for
a long time. However, interest in this branch of physics and technology has been intensified only recently, in con-
nection with the development of nuclear energy. This increased interest is due to the great similarity between
nuclear heat sources and direct-conversion elements.
Thanks to the absence of moving parts in direct-conversion systems, we may anticipate the possibility of set-
ting up reliable installations with long operating lifetimes; because the converters are compact, it will be possible
to build energy -producing units with high power per unit weight and per unit volume, especially in the case where
the heat sources are integrated with the conversion elements.
The properties of nuclear heat sources (high energy capacity and high specific power) can be advantageously
combined with the properties of direct-conversion systems. The combination of these properties offers good prospects
for the use of nuclear power units for the direct conversion of heat energy to electrical energy where there is a need
for self-contained, reliable, and compact sources of electrical energy with long operating lifetimes.
Direct-conversion systems, especially those of the thermoemissive and magnetohydrodynamic types, are high-
temperature systems, and this offers a possibility of increasing the efficiency of energy-producing units, especially
when the direct-conversion elements used have the form of a high-temperature stage in combination with a mecha.L. -
cal converter. Nuclear sources can supply the high temperatures required for direct-conversion systems, offering
good prospects for the future use of stationary nuclear installations with direct conversion.
At the Third Geneva Conference the questions of direct energy conversion were discussed at a special sectional
meeting, at which the participants listened with great interest to papers1 devoted to work done in the field of thermo-
electric, thermoemissive, and magnetohydrodynamic converters.
THERMOELECTRIC SYSTEMS
(217, 218, 318, 873)2
The thermoelectric effect was regarded for a long time as unsuitable for energy production because of its low
a2
efficiency of converting heat energy to electrical energy, which was due to the low quality factor Z = of
metallic thermocouples (where a is the coefficient of thermoelectromotive force, S is the electrical conductivity,
and X is the thermal conductivity).
The development of semiconductor components, such as Bi2Te3, PbTe, SiGe, and others, made it possible to
raise the quality factor of the equipment by a factor of several dozen, thus yielding acceptable values of conversion
efficiency. Figures 1 and 2 (217) show how the dimensionless parameter ZT varies as a function of temperature for
several semiconductor materials with n-type and p-type conductivity. Thermoelectric materials may be divided
into three groups with regard to operating temperature: low -temperature materials based on Bi, Pb -based materials
for the intermediate temperature range, and high-temperature materials such as the alloy SiGe. Material based on
SiGe is promising as a thermoelectric material for high-temperature units, including outer-space systems (218), since
it has an operating temperature of over 1000?C, a low vapor pressure at this temperature, and good mechanical pro-
perties and makes possible the electrical switching of thermoelements by metallurgical means.
1A list of the papers delivered by Soviet scientists was published in Atomnaya Energiya, Vol. 17, No. 3, p. 235 (1964),
and a list of the papers delivered by non -Soviet scientists appeared in Atomnaya Tekhnika Za Rubezhom , No. 9 ,p. 27(1964).
2The numbers in parentheses are the numbers of the papers.
494
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Zr
1.0
0.8
0.6
0.4
0.2
Biz Te
0 100 200 300 400 SOO .600 700 800 900 1000
Temperature, ?C
Fig. 1. Characteristics of fl-type thermoelectric
materials.
36
32
0 28
8
4
1600 2000 2400
Cathode temperature, ?K
Fig. 3. Converter output power as a
function of cathode temperature
(for optimum values of cesium n pres-
sure and anode temperature): 1) tting -
sten cathode, nickel anode; 0.13 trim;
2) UC+ ZrC cathode, 1.0 mm gap.
Zr
1.0
as
0.6
a4
a2
Ti!
Bi Te
SilTe
6eSi
0 100 200 300 400 500 600 700 800 SOC 1000
Temperature, ?C
Fig. 2. Characteristic's of p -type thermoelectric
materials.
15,
i0
5
0
Inter -electrode distance, Mtn
Fig. 4. COnverter output power as a func-
tion of inter-electrode distance at a cath-
ode temperature of 2100?k and optimum
cesium-vapor pressure: 1) tungsten cath-
ode; 2) uranium nionacarbide cathode.
Isotopic Sources Of Current. A number of
electrical generators using thetnioelectric converters with
radioactive isotopes as their heat source have been devel-
oped and are operating at the present tune. Soviet scien-
tists reported (318) oil two sources of this type; One of
these sources uses Po210; the heat source is a flat container
(60 x 60 x 13 mm) with five Pon' capsules having activity ovahies of 1000-2000 Ci and is surrounded by a battery of
SiGe heat converters. The generator Was successfully tested for several thousand hours. it had a power of ? 6 W and
an efficiency of ? 2.4%. The temperature on the hot surface of the converter was 760?C, and the radiator tempera-
ture was 230?C. The second type of generator used Ce144, which supplied electrical energy for an atitbrnatic radio
weather station. It had an electrical power of ? 5 W. the use of a special storage battery made it possible to in-
crease the powet to 150 W. Variations in thermal power were cbinpensated by a Special ternperaniteLcantrol system:
The generator used solid solutions of6i2Te3+ BiSe3 (ii-type) and Bi2Te3+ Sb2Tei (p-type) lot its thetinal elements.
Work on the development of isotopic energy sources with thermoelectric converters has been done in the
United States (211). Units With electrical power values of 2 14/ to 60 W have been built and are in operation; the
isotopes used are: Sr9? Pui3; and Po16. These units are designed to supply electrical energy to Weather stations,
navigational buoyS, and satellites.
The testing of isotopic generatOrS under operational conditions both in the Sdidet Linibri and in the United States
has shown that, thanks to their compactness; reliability, self-contained construction, atid.long operating lifetime,
isotopic generators give great promise for use as low -power sources of electrical energy for inaccessible regions of
the earth and for outer space.
495
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
(1)
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
3 4 5 6
Work function in vacuum, eV
Fig. 5. Comparison of the effective work function
in cesium vapor with the work function in a vac-
uum, at a cathode temperature of 1700?K:
1) cesium -vapor pressure 8 ? 102 mmHg; 2) cesium -
vapor pressure 0.8 mm Hg.
2- Z000
H 1500
Min
11111=11
10-2 10-1 1
Pressure, mm lig
Fig. 6. Experimentally determined boundaries of
operating modes of a converter with molybdenum
cathode (gap 0.5 mm): A) quasi-vacuum mode;
B) diffusion mode; C, D, E) arc modes.
React or Schemes. It is possible to obtain high elec-
trical power values in units with thermoelectric generators if
a nuclear reactor is used as the heat source. The first opera-
ting unit'of this type is the Romashka high-temperature reac-
tor-converter (8733), which began producing electric current
on August 14, 1964. This unit is characterized by a complete
absence of moving parts or mechanisms in the energy-conver-
sion process. The heat generated by uranium dicarbide fuel
elements is distributed by thermal conductivity and, passing
through a beryllium reflector, is transmitted to a thermoelec-
tric converter. A converter with silicon-germanium thermal
elements is in contaet with the external surface of a radial
reflector. The heat in the converter elements is partly con-
verted into electricity. The unutilized portion of the heat is
removed by radiation into the environment.
A somewhat different principle of operation is used in
the SNAP-10A thermoelectric-converter nuclear power unit,
developed in the United States (218). This unit is now in the
start-up and adjustment stage. The SNAP-10A unit uses a
homogeneous thermal-neutron reactor. The moderator used is
zirconium hydride, containing about 10% U235 by weight. The
heat from the reactor is transmitted by means of a liquid -
metal coolant (a eutectic NaK alloy) to the thermoelectric
generator and is removed from the cold junctions of the thermo-
elements by radiation into the environment. Of the two possi-
ble materials, PbTe and SiGe, preference was given to SiGe,
since it is capable of operating under high-vacuum conditions,
10 while the use of PbTe requires special jacketing of the thermo-
elements in order to prevent sublimation.
The converter of this system,designed for 500 W, is di-
vided into 120 modules. The heat from the reactor is trans-
mitted by the coolant to the tubular part of the module.
Twenty-four cylindrical thermoelements are arranged along
each tube; these elements are electrically insulated from the tube by thin disks of aluminum oxide. On the hot side,
the elements are connected to copper bus bars, and on the cold side, to aluminum radiation plates. All the inter-
faces between the tubes and the aluminum plates are metallurgically connected. The elements are designed with
low thermal and electrical resistance, which makes it 'possible to prevent any parasitic heat flows,. Each set of three
successive modules in the converter is connected into a section. Forty sections are mounted on the surface of a cone
frustum, forming parallel coolant loops between the input and output collectors. The 120 modules weigh 60 kg and
have a radiating surface of 5.8 m2. The sections of the converter are electrically connected in a series-parallel cir-
cuit, which improves the reliability of converter operation.
In addition to the main converter, the SNAP-10A unit has an auxiliary thermoelectric converter used for feed-
ing a d-c electromagnetic pump. Two parallel PbTe thermocouples short-circuited across the operating part of the
pump, supply 700 A of direct current. The heat reaches the thermocouples through the coolant tubing walls, and
the unutilized heat is removed by radiation from aluminum radiation ribs. A 2400-gauss magnetic field is provided
by a permanent magnet. The overall efficiency of the pump for the head so set up is 1%.
It is expected that when the development work has been completed, the SNAP-10A unit will be launched into
outer space, where it will supply 500 W of power for about one year.
3 This paper was published in Atomnaya Energiya, Vol. 17, p. 329 (1964).
496
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
7 6 5
Fig. 7. Schematic diagram of the electric-
ity-generating element of the reactor-con-
verter: 1) heat-generating core with cathode;
2) cesium gap; 3) anode; 4) insulation; 5)nio-
bium tubes; 6) coolant; 7) moderator.
6
0. 2
0.
1
0
0
7,--
/,-
r
/
/
/
to 10-`
to-3 to-2 to-' i to
Pressure, mm Hg
Fig. 8. Variation of the output power of a
converter with a cathode made of a solid solu -
tion of uranium carbide and zirconium car-
bide as a function of cesium-vapor pressure,
at the following temperatures: 1) 1600?C;
2) 1550?C; 3) 1500?C; 4) 1250?C.
THERMOEMISSIVE CONVERTERS
(44, 132, 219, 317)
The physical foundations of the thermionic emission process
are determined by the high operating temperatures of thermoemis-
sive converters. The output power of the converter increases sharply
with temperature (Fig. 3) (219). Above 1500?K the specific power
of the thermoemissive converter becomes more than 1 W/cm2,
which makes it very promising for use in conjunction with nuclear
energy sources.
Earlier studies on thermionic emission were devoted to the
investigation of vacuum operation of the converter. The use of
this type of operation led to serious design and engineering difficul-
ties, since in order to eliminate the space charge of electrons, the
distance between the electrodes had to be reduced to 10 p. Through
the use of cesium vapor to compensate the space charge, it is
possible to develop elements with a reasonable inter-electrode dis-
tance. Figure 4(219) shows the variation of the converter output
power as a function of the inter-electrode distance at a cathode
temperature of 2100?K and optimum cesium vapor pressure.
The presence of cesium in the inter-electrode gap also mani-
fests itself in a reduction of the work function at the cathode and
the anode, thereby improving the characteristics of the converter.
The effect of cesium on the work functions of various materials,
determined theoretically and experimentally, is shown in Fig. 5,
which graphically compares the effective work function in cesium
vapor with the work function in a vacuum for a number of materials
(219). The curves have been plotted for two values of cesium
vapor pressure and a constant radiator temperature. A higher value
of the work function in a vacuum means that the effective work
function in cesium vapor will be lower. The quantitative data for
specific materials may differ considerably because of the way these
parameters are affected by surface structure, crystal orientation,
presence of impurities in the cesium vapor, etc. Thus, for example,
the work function of molybdenum in cesium vapor with an admix-
ture of fluorine is 1.36 eV, whereas the value for pure cesium is 1.68 eV.
An analysis of the converter operating modes as a function of cesium vapor pressure yields a theoretical de-
scription of the diffusion mode (317). Figure 6 shows the boundaries of operating modes of a converter with a molyb-
denum cathode, as determined from experimental data.
The development of thermionic elements requires not only physical research but also the elimination of a num-
ber of technological difficulties. In particular, rigid requirements must be imposed on.the insulating compounds,
which must maintain a vacuum seal and a high dielectric strength (up to 1000 V/cm) at temperatures of up to
1000?K. Satisfactory properties were provided up to 900?K by aluminum oxide combined (by means of copper solder)
with niobium. Higher temperatures of up to 1400?K require the use of vanadium-based solder or diffusion com-
pounds (219). Figure 7 shows a schematic diagram of a thermoemissive converter (TEC) combined with a fuel ele-
ment (the electricity-generating element of the reactor-converter). The reactor-converter consists of a cathode en-
closing a heat-generating core (in some cases it is possible to use an unjacketed core, for example, a solid solution
of uranium carbide and zirconium carbide). The cathode is surrounded by a cylindrical anode, and the gap between
the electrodes is filled with cesium vapor. The individual components are connected in series, which increases the
output voltage of the converter. The unconverted heat is removed by the coolant flowing past the metallic jacket
which contains the series-connected components, separated from the jacket by a layer of electrical insulation.
497
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
The operating conditions of the TEC in the reactor aggravate the difficulties resulting from the high tempera-
ture of the TEC and create additional difficulties due to the presence of reactor radiation, the effect of fission frag-
rnents, and the distribution characteristics of the heat generation.
The accumulation of fission products at high burnout values for the heat-generating core material causes the
core to expand. The diffusion of gaseous fission products in the cesium space affects the operating characteristics of
the converter. Thus, the fission products whose condensation temperature is higher than the anode temperature will
be deposited on the anode and will affect both its work function and its emissivity.
The strong temperature dependence of the electrical characteristics of the TEC makes it necessary to equalize
the heat generation in the reactor-converter by redistributing the concentration of fuel or moderator.
Important results for the development of electricity-generating components of a reactor were obtained by in-
vestigating TEC specimens in a reactor (317). Tests were conducted on converters with cathodes constructed in
various forms: as a rod made of a UC+ ZrC solid solution with no jacket and as UO2 rods in a molybdenum jacket
with or without an outer coating of emissive material. Because of the uneven temperature distribution along the
cathode, the experimental results were averaged over a range of temperatures. Figure 8 contains typical curves
showing how the output power of a converter with a uranium-zirconium carbide solid-solution cathode varies as a
function of cesium vapor pressure for different cathode temperatures. Tests conducted in a reactor showed that the
TEC characteristics were fairly stable when the inter-electrode space was periodically pumped out. TEC operating
lifetimes of seVeral hundred hours were attained in the course of loop tests.
498
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
FAST REACTORS
0. D. Kazachkovskii
Translated from Atomnaya Energiya, Vol. 18, No. 3,
pp. 390-395, April, 1965
One of the Conference sessions was devoted to the physics and technology of faSt reactors and 17 reports were
heard.I In addition, certain questions relating to the problem of fast reactors were discussed by other sections (nu-
clear fuel and its chemical processing, reactor kinetics, economics. Interesting data concerning the operation of
fast reactors was presented by several countries at the scientific-technological exhibition (especially Great Britain,
USSR and USA).
Physics of Fast Reactors
At the Conference, in contrast to the two previous ones, problems of measurement procedure (and also of
theory) of nuclear-physical parameters used for calculating fast reactors were hardly considered: In certain papers
(259 et al.)2 the present-day level of knowledge of the nuclear-physical constants was assessed from the indications
of the principal results obtained in recent years. It was found, in particular, that on the basis of precise experiments
carried out the value of the fission cross section of U235 (4) in the energy range from 50 keV to 1 MeV is given
somewhat lower (by 5/0in comparison with the values assumed earlier). This implies that other quantities also, for
the measurement of which the value of c1.1 was used as the reference cross section and above all the fission cross
section of other isotopes; should also be reduced correspondingly. The excellent agreement Of the measurement re-
sults for the quantity v(E), obtained in different laboratories and by different methods, testifies to the reliability and
the high accuracy of the valtie of this parameter.
Great progress hal been achieved in tneaShrernents of one of the most important quantities Oc, for determining
the breeding ratio (the ratio of the radiative capture cross section to the fission cross section) for the principal fissile
isotopes. The development of nanosecond techniques and the use of effective scintillation detectors has enabled the
knowledge of the parameters of inelastic scattering and the radiative capture cross section oc for fast neutrons to be
considerably improved. Progress in the range of measurement of oo for inactive isotopes is particularly significant,
for which the experimental data were previously extremely meager. It should be noted, however, that in certain
cases quite a large discrepancy is observed between the measurement data for oc in different laboratories; these data
fall outside the limits of the prescribed accuracy of the results. The maximum discrepancy (approximately a factor
of two) is observed in the region of neutron energies of order tens and hundreds of keV for Au, Ta, In.
Summary tables and graphs of the interaction parameters of neutrons with nuclei, including the most recent
refinements, have been presented in several Special publications [1-31, which were distributed at the Conference:
At the same time, it was pointed out in public addresses that such a form of information storage is becoming cumber-
some and obsolete. Increasingly more data are being accumulated everywhere (thousands of points for each isotope)
and the use of all the data necessary for calculating reactors is in each case an extremely tedious and time-con-
suming manual operation. Iii certain laboratoties methods have been developed for storing the appropriate data
which are suitable for feeding the data directly into a machine (mainly on punched cards). The necessity was em-
phasized for creating on an international scale a Single unified system for data storage, which would permit quite
simple processing of the bulk of data; with adequate operability, and to introduce the incoming changes and refine-
ments. The programs whieh have been developed enable a system of multigrotip constants to be produced on com-
puters, containing among them a very large nun-Met (several hundred) of groups. The latter are used for detetmining
in simple geometry the detailed sttuctute of the spectrum of fast reactors.
IThe list of reports by Soviet scientists is published in Atomnaya Energiya, 17, No. 3, 235 (1964); and a list of re
ports by foreign scientists in Atomnoi tekhnike za rubezhom, No. 9, 27 (1964).
2
The numbers of the reports are shown in parentheses.
400
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
TABLE 1. Comparison of Theoretical
and Experimental Values of Critical Masses
Index of
assembly
Critical mass kg U235
experiment
theory
2 (Great Britain). . . .
416
414
3 (Great Britain). . . .
79.6
61.2
1A (USSR)
76.5
78.2
12 (USSR)
709.4
682
25 (USA)
661.0
581.6
36 (USA)
237.0
242.7
TABLE 2. Comparison of Theoretical
and Experimental Values of the Breeding
Ratio
Index of
Breeding ratio
assembly
experiment
theory
1A
1,21 ? 0.15
1.27
6A
1.57
1.40
8
1.07? 0.08
1.00
12
1.13 ? 0.07
1.15
In some papers (18, 166, 259, 510) data are reported
concerning the reference systems of multigroup constants
which are used and which have been prepared by various authors. These systems, in comparison with those used pre-
viously, are characterized by an increase in the number of groups (up to 26-33) which occurred principally because
of the introduction of additional lower groups. The reference systems provide for calculating the dependence of the
group cross sections On the concentration of the corresponding isotope. This effect, which is associated with a homo-
geneous resonance blocking, appears for heavy nuclei at neutron energies below 25 keV.
Fast reactor calculations were carried out mainly in diffusion and Se-approximations (166,259,368). Some
results of critical mass calculations for large diluted systems and comparison with experiments carried out in critical
assemblies are presented in Table 1.
In the majority of cases the calculation permits the critical mass to be predicted with an accuracy of ?5-6%.
The neutron distribution density in the active zone and the relationships between the reaction intensities are calcu-
lated with approximately the same accuracy. However, the calculated lifetime of neutrons and correspondingly the
effect of boron on the reactivity is found to be systematically too low in comparison with the experimental data
(according to data from Report 265, by 30%). This indicates nonagreement between calculation and experiment in
the very soft region of the neutron spectrum and shows the necessity for further improving the multigroup constants
in the last groups. The value of the breeding ratio (more precisely, the conversion factor in the U235?U238-13039
cycle) has been measured only in experiments carried out by Soviet authors (Table 2).
Relatively more consideration has been given to the calculation of quantities which are essential for studying
the kinetics of fast reactors. The fast components of the reactivity power factor have the most important signifi-
cance, among which the Doppler effect makes the primary contribution because of the presence of soft neutrons in
the spectrum. The most progress has been achieved in calculating the Doppler coefficient; in almost all cases which
are of interest this coefficient is negative and has a value of the order of ? 10-5AkrC. It is shown in (41) that by
taking account of the partial screening of the U235 and Pu239 resonances by U238 (there is no inverse effect) this leads
to intensification of the over-all negative Doppler effect.
Data are presented (41,259,539) concerning the calculation of the sodium reactivity coefficient. For this,
the appearance of voids in the active zone is borne in mind, for example, because of effervescence of the sodium,
etc. It should be noted that the contribution from thermal expansion of the sodium to the reactivity coefficient is
almost negligible and amounts to ? 10-76,1
) Efor (I.
s< 0,
? lit 171814e ___(e_irsh2k ;field
515
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
where the distance z is measured from the forward boundary of the layer ,which faces the incident wave Eoe ?i((iJt?kz).
The above equation reflects the general regularity characteristic for any bounded system of charges, which consists
in the fact that the dissipative wave radiation pressure produced by the interaction between particles does not act on
the bunch elements adjacent to its back boundary (z = d) and that it continuously increases with an increase in the
distance from this boundary toward the bunch interior to a value s :s X/8 for e > 0 or right to the forward boundary
s = d for e 0 (X is the wavelength inside the bunch), i.e., it focuses the medium by pinching it. For instance, in
the coherent case for a very thin layer (d ? X/47r), the radiation forces monotonically decreases in the direction of
propagation of the incident wave, ,f(z) = 4R(d?z)/27rd2, where R is the over-all coefficient of reflection from the
layer. Moreover, an analysis of the specific structural characteristics of the interference field immediately in front
of the medium layer (z < 0) reflecting the radiation shows that a rather thin layer of radiation-accelerated plasma
generates returning forces, which retain individual particles that have left the bunch for any chance reason:
r 3E8 (e ? 1) lie sin 2k lie d
F (s')= X cos.2ks' >0 (0 is the mean standard deviation of the L-th harmonic's contribution to the scattering cross section;
is the mean correlation moment befween harthonics L and L'.
Since the errors of the expansion coefficients depend only weakly on the atomic weight it follows that the
matrix of the correlation moments cannot be very different for the other elements.
The thin continuous lines in Fig. 1 represent the angular distributions for the above elements, calculated on
the optical model of the nucleus (v0=1.45 ? 10-13 cm). The calculations were made on the assumption that the cross
sections of the reactions taking place via a compound nucleus are isotropic. Thus the anisotropy of the calculated
angular distributions was caused by anisotropy of the elastic scattering proper, which was calculated by the formula
a s, 1 (E, I1) r=
da
dt.c
a
=-- 2_1 (21-j- 1) (1-- 111)!El
2
barn
for all energies falling within the upper group, with varying steps of 0.25, 0.5 and 1 MeV. The values of 74 were
calculated for the Woods-Saxon potential,
Vo (1+
V (r) ? ?
1+exp aR
where Vo = 45 MeV, a = 0.5 ? 10-13 cm and g = 0.1. The method of calculating as, 1(E, was given in [3].
as, /(E, bi) was then averaged over the upper group, taking account of the detector efficiency:
CO
S cp (E) h (E) dE
El
(3)
(4)
(5)
where .cp(E)qh(E) is the weighting function, p(E) being the fission-neutron spectrum above E1 in MeV and crTh(E) the
fission cross section of Th232. The final theoretical cross section for a given element was determined by the formula
d? Laic.=
da I
(11)+K (ac (E))] barn,
where is the cross section for inelastic processes, averaged over the upper group:
S.w (E)a (E) dE
(a =_-_, (E)) El
5 q (E) dE
(6)
(7)
here K is a dimensionless coefficient which takes account of the fraction of inelastic scattering which leaves the neu-
tron in the group. K was calculated from the system of multi-group constants [4]. A comparison of the experimental
and theoretical data (see Fig. 1) shows that there is satisfactory agreement for most of the elements. In the cases
when there was a clear discrepancy, we calculated a further curve (dotted lines in Fig. 1) for which ro was taken as
-1.30 ? 10-13 cm.
The angular distributions OW are shown in Fig. 2 in order of atomic weight.
527
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
528
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
1
A (W) 1? R21471 (1L0W20 ?P2)
A (P) 11,0
1? = P2 (1 0 W20 P2)
Lo
(17)
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Considering that W20 = W1 =(3n) z, we obtain
so that
, 1 3L2'
20?P2 (z/ ( ? ) '
3/.2 ( 3L2)
L L-1-1 411, 11,
In
L?/ , 3L2) (1 3L2) ?
I m )
(18)
Rayleigh scattering, for which 6 = 1/2, is an important particular case of the example considered above. By
solving (18) with respect to L2 (for L >> /), we find the expression for the diffusion length in the case of Rayleigh
scattering:
L2= _Ha ( Sdr,
3
Fick Law in the Theory of Neutron Transport
(19)
The basic law in the elementary diffusion theory is the proportionality of the flux of neutrons to the gradient
of their density (Fick law). Since, in the case of asymptotic behavior, all Nk depend exponentially on the distance
with the attenuation distance L, the validity of the Fick law is obvious. In the case of plane geometry, the diffu-
sion coefficient is given by
D -=vL N1
N 0 ?
(20)
In the case of asymptotic behavior, it can be considered that all Nk values satisfy the homogeneous system (7)
with the determinant A = 0. Due to this relationship, the Nk/No ratios are fixed in a well-defined manner. In or-
der to find Ni/No, we shall subtract the first equation of system (7), multiplied by P1, from the second equation:
Ept., (QB ?Pic;0)N; = 110 (Qi ?P1Q0) Ar0=--
Consequently, Ni/No = P1-14, and
(21)
D =-_vL (Pi ?110)?
(22)
In the case of steady-state conditions, taking into account Eq. (15), we arrive at the well-known equation
vL2
D = (23)
/a '
i. e., the factor of proportionality between the flux and the gradient of neutron density in the case of asymptotic
behavior is related to the attenuation distance in the same manner as in the diffusion approximation. The effect
of the scattering anisotropy on the diffusion coefficient is limited to the change in the diffusion length L, which is
determined from (14). The anisotropy of sources does not affect the L and D values.
In the case of unsteady-state condition this important property of the asymptotic behavior [see (23)] vanishes.
For instance, for a uniformly moving source,
vL2
D= (1
1,
where u is the velocity of the source.
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
(24)
543
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Since all, the.Ni/No values.areideterinined.in a well-defined manner by system (7), one readily arrives at the
other characteristic of the distribution of neutrons that have experienced a large number of collisions: the angular
distribution does not depend on the anisotropy of the source. The dependence of the asymptotic neutron density
Nas on the intensity and the angular distribution of the source consists in the constant factor No, which, or course,
cannot be determined on the basis of the homogeneous Eq. (2).
In the presence of sources, the solution of the nonhomogeneous equation leads to the following system of
equations instead of (7):
for Fourier transforms, where
+Fri
S (q, co) (' (0) Ph d52
Ph= =--S(q, (pa,
qv
where f(e) is the angular distribution of the source neutrons.
The Fourier transform of the function to be determined is given by
iS (q, (a)
N = ?. A .13 ? is (q, f (0)
(cos 0)
4.itqv (z?cos 0)
43tqv (z ?cos 0) A 3
(25)
(26)
(27)
where A. differs from A, determined by Eq. (11), by the substitution of the (,90, (p2, values for the j-th column,
Reverting from the Fourier transform (27) to the neutron density, we obtain the solution of the problem of
neutron distribution about the source. Let us consider the asymptotic part of the solution, which is determined by
the contribution of the pole for A = 0 in the particular case of a source defined by the expression
Here,
S (q, co, 0)r_6 (w?uqx) (qy) ? (gz) /(0). (28)
From (27), we obtain the asymptotic density of neutrons in front of a moving source:
pjAiPj (cos 0)
Nas? Ae-(x-nt)IL
Z ? cos 0
so L 216
A? 810 v
where 6 = A'(i/L), = L was used in
From the equation
544
Z =
_x-ut
Nh=42-tAe L.Ah,
(29)
(30)
(31)
(32)
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
Declassified and Approved For Release 2013/09/24: CIA-RDP10-02196R000700010004-3
we find
and, consequently,
A dN
1 o
j =vN ?vL -Ao dx
(33)
D =vL ?
Ao (34)
In order to calculate the ratio of determinants in Eq. (34), we note the following expressions derived from (29):
No (x, t) = 4J1,4e-(x-u1)1L
N i(x, t) 4nAe-(x-u1)IL Eik,A1(2; 1.00) ,
3 1
N2(x, t) 41tAe-( x--11t) (p 2 E 1,,,A;Q; ? PIRA ?
The relationships between the determinants can readily be obtained from Eqs. (32) and (35):
3 1
6.2= [ (P1? RO) ( -2-Pi ? -y ) ? ?
(35)
(36)
(37)
From (34) and (36), we naturally obtain Eq. (23), while, from (36), (37), and similar equations, we find, after
substitution in (29), that the angular distribution Nas does not depend on the source.
Finally, it follows from (36)-(37) that, for u = 0,
v/
3
(1+2 N2 ) dNo
j--- ? -----.
No dx
(38)
(38)
Expression (38) is a consequence of the generalized Fick law [3], which was derived for isotropic sources.
This relationship holds for asymptotic behavior and for anisotropic sources. In using the method of spherical harmon-
ic and cutting off the infinite system at the P1-approximation, it is assumed that
2 N.2 =2?A2 = 1 ? 1