SOVIET ATOMIC ENERGY VOL. 57, NO. 3

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP10-02196R000300050003-4
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 101
• Document Creation Date: January 4, 2017
• Document Release Date: February 22, 2013
• Sequence Number: 3
• Publication Date: September 1, 1984
• Content Type: REPORT

Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
,~J IJJN UUSti-5S I7.
Russian Original Vol. 57, No. 3, September, 1984
~ /~ March, 1985
~~
SATEAZ 57(3) 57772 (1984)
SOVIET
ATOMIC
ENERGY
ATOMHAfI 3HEPt'Nfl
(ATOMNAYA ENERGIYA)
U
TRANSLATED FROM RUSSIAN
CONSULTANTS BUREAU, NEW YORK
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
O ~' ~ ~ I Soviet Atomic Energy is a translation of Atomnaya Energiya, a
publication of the Academy of Sciences of the USSR.
ATOMIC
ENERGY
Soviet Atomic Energy is abstracted or in-
dexed in Chemica/ Abstracts, Chemical
Titles, Pollution Abstracts, Science Re-
search Abstracts, Parts A and B, Safety
Science Abstracts Journal, Current Con-
tents, Energy Research Abstracts, and
Engineering Index.
An agreement with the Copyright Agency of the USSR (VAAP)
makes available both advance copies of the Russian journal and
original glossy photographs and artwork. This serves to decrease
the necessary time lag between publication of the original and
publication of the translation and helps to improve the quality
of the latter.. The translation began with the first issue of the
Russian journal.
Editorial Board of Atomnaya t`nergiya:
Associate Editors: A. I. Artemov, N. N. Ponomarev-Stepnoi,
and N. A. Vlasov
I.A. Arkhangel'skii
I. V. Chuvilo
I. Ya. Emel'yanov
I. N. Golovin
V. I.II'ichev
P. L. Kirillov
Yu. I. Koryakin
E. V. KuloV
B. N. Laskorin
V. V. Matveev
Mailed in the USA ~ by Publications
Expediting, Inc., 200 Meacham Ave-
. nue, Elmont, NY.1 1003.
iPOSTMASTER: Send address changes to
Soviet Atomic Energy, Plenum Publish-
ing Corporation, 233 Spring Street, New
York, NY 10013.
A. M. Petras'yants
E. P'. Ryazantsev
A. S. Shtan
B. A. Sidorenko
Yu. V. Sivintsev
M. F. Troyano
V. A. Tsykanov
E. I. Vorob'ev
V. F. Zelenskii
Copyright ?1985, Plenum Publishing Corporation. Soviet Atomic Energy partici-
pates in the. Copyright Clearance Center ICCC) Transactional Reporting Service.
The appearance of a code line at the bottom of the first page of an article in this
journal indicates the copyright owner's consent that copies of the article may be
made for personal or internal use. However, this consent is given on the condition
that the copier pay the flat fee of $ 8.50 per article Ino additional per-page fees)
directly to the Copyright Clearance Center, Inc., 27 Congress Street, Salem,
Massachusetts 01970, for all copying not explicitly permitted by Sections 107 or
108 of the U.S. Copyright Law. The CCC is a nonprofit clearinghouse for the pay-
ment of photocopying fees by libraries and other users registered with the CCC.
Therefore, this consent does not extend to other kinds of copying, such as copying
for general distribution, for advertising or promotional purposes, for creating new
collective works, or for resale, nor to the reprinting of figures, tables, and text ex-
cerpts. 0038-531 X/ 84 $ 8..b0
Consultants Bureau journals appear about six months after the publication of the
original Russian issue. For bibliographic accuracy, the English issue published'by
Consultants Bureau carries the same number and date as the original Russian from
which it was translated. For example, a Russian issue published in December w""ill
appear in a Consultants Bureau English translation about the following~June, but the
translation issue will carry the December date. When ordering any volume or particu-
lar issue of a Consultants Bureau journal, please specify the date and, where appli-
cable, the volume and issue numbers of the original Russian. The material you will
receive will be a translation of that Russian volume or issue.
Subscription (2 volumes per year)
Vols. 56 & 57: $560 (domfstic), $621 (foreign) Single Issue: S 100
Vols. 58 & 59: $645 (domestic), $715 (foreign) Single Article: $8.50
CONSULTANTS BUREAU, NEW YORK AND LONOON
233 Spring Street
New York, New York 10013 '
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
SOVIET ATOMIC ENERGY
A translation of Atomnaya Energiya
March, 1985
,Volume 57, Number 3 September, 1984
CONTENT8
ARTICLES
Cost of Information in Nuclear Power - Ya. V. Shevelev. .
Using In-Reactor Measurements to Reduce the Indeterminacy of the
Physical Calculation of Fields of Energy Release - V. K. Goryunuv
and Ya. V. Shevelev.
Incorporating Reliability in Optimizing Units in Nuclear Power Stations
Containing Water-Cooled-Water-Moderated Reactors - N. E. Buinuv,
S. M. Kaplun and L. S. Popyrin.
Effects of Irradiation on the El.astoplastic Deformation of Zr + 1% Nb
Fuel Pin Cladding - V. A. Matushkin, A. A. Medvedev, Yu. V. Mi].oserdin,
B. D. Semenov, Yu. K. Bibilashvili, and I. S. Gol.ovnin .
Damage Summation in Annealing and Repeated Irradiation of Pressure-Vessel.
Steel - V. A.' Nikolaev, V. I. Badanin, and A. M. Morozov. .
Effect of Chemical Composition and Annealing Conditions on the Radiation
Embrittlement of the Metal of Low-Alloy Welded Seams
- V. A. Nikolaev, A. M. Morozov, V. I. Badanin, A. S. Teshchenko,
and R. P. Vinugraduv
Erosion of Fe-Cr-Ni Alloys and Vanadium Alloys during Bombardment with
Helium Ions - B. A. Kalin, I. I. Chernov, V. L. Yakushin, V. 1. Badanin,
I. P. Kursevich, V. A. Nikulaev, and V. N. Kulagin .
Resonance Effects in the Interaction of 0.2-0.8-MeV Neutrons with
g6Fe Nuclei - A. A. Sarkisov, I. N. Martem'yanov, A. M. Boguslavskii,
V. N. Ivanud, and C. N. Ivanov
Application of Gamma Spectrometry in Integrated Experiments on Reactor
Physics - A. V. Bushuev and V. N. Ozerkov. .
Penetration of Radioactive Industrial Waters from the North Sea into Centr 0, the economic advisability of doing the experiments is proven.
The function p(yllyo, oo, 6~) which enters into the expression (6) should satisfy the
equation
A (y I yo, oo) = J P (yt I yo, 60> oi) P (y I yt+ o~) dyi ? ( ~ )
Actually, one can treat the averaging of y as a two-stage process. In the first stage an
inaccurate experiment gives the parameter yl with the probability p(yilyo, Qo, a~)., after.
which an accurate measurement gives the value y with the probability p(y~y3, 6,.). The final
distribution of y cannot differ from that existing prior to the experiment p(ylyo, oo), as
long as the possible results of its execution are being discussed. In particular, we find
from Eq. (7), using the expressions (1), (3), and (4),
yo= ` P (y I yo, oo) y dy = ~ P (y~ l yo, 60> 6i) dyt J P (y I yi+ o~) y dy , _ ~ P ~yt I yo, 60, ot) yi dy,
J P(ytlyo+ 60> Qi)y~dyi=yo; (8)
60 = ~ A (y I yo+ oo) (y - yo)2 dy = ~ P (yi I yo, Qo, ot) dye , A (H I y1, o~) (yi - yo ~-
+y-yi)2dy= ~ P (yf~yo, oo, at) dyi I(yt-yo)a-}-oi);
(9)
P (y~ I yo, 60, 6i) (y~ - yo)2 dyi = ~a2.
If there are grounds for assuming p(y~yo, Qo) and p(ylyl, 61) to be normal distributions,
p(y~lyo, co, Q~) is also a normal distribution with a mathematical expectation yo and a dis-
persion Doe.
Simplest Quadratic Model. Let the choice of x be unconstrained in any way, and let f
he approximated with good accuracy by a quadratic function of x and y. One can, without
restricting generality, figure x from the optimum point at y = yo and f from its value at
this optimum. Then
~ = fb ' (y - yo) ~- fey ' (y- yo)2/2 -~ fxxx2/2 ~-- fxy~' ~y - yo), (10)
fzx~0. (11)
Integrating the expression (10) with the weight p(ylyo, Qo), we obtain according to the ex-
pression (1)
(f)o = f by6ol2 -i- fa:x2/2.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
The optimal x is equal to. zero, so that
min (f)o= fuyou/2. (12)
x
Integrating the expression (10) with the weight p(ylyl, ol), we obtain, according to Eq. (3),
(f )1= fu' (/f - yo) ~-- fyu' [(yt - yo)a -}- ail /2 ~-- fxxx2/2 + fxux ' (yt - yo)
The optimal x is equal to :- fxg?(yl - yo)/fxx, :so that:
min(f)~=fy?(y~-yo)-f-fy??1(y~-yo)Z-{-oil/2- (fx,~)2(yi-yo)2/(2fxx)? (13)
x
Averaging this quantity with the weight p(yllyo, oo, 61), we obtain, according to the expres-
sions (7) and (8),
(min (f)i)o~ =1by - (~aa -I- oi) l2 - (fxu)2 ~o2/(2f xx)
~ x
Substituting Eq.,(2) and the expression (14) into formula (5) and using the quantity (4), we
find
Ef = {fubool2} - {0 f -I- f boo/2 - Doe (fxy)2/(2fxx)},
Ef=C wa-0f,
C - (fxy)a/(2f xx)
The quantity C~62 is the economic cost of the information, i.e., the usefulness of the
experiment. If its value is greater than the cost of the experiment ~f, the experiment is
profitable. The economic value of the information turns out in this case to be proportional
to the information content of the experiment X02. One can call the proportionality coefficient
C the information cost. According to the expression (11), this is a positive quantity. If
the experiment is accurate, X02 = cso. The usefulness of an accurate experiment CQo coincides
with the disadvantage from a lack of information prior to the experiment:
Possible Generalizations of the Quadratic. Model. A rather simple generalization of the
model consists of replacing the scalars x and y with vectors x and y. The expansion (10)
takes the form (to shorten the writing we set yo = 0)
f=ytly-{- 2 ytfyuy+ 2 ~fxxx-~xtfxutJ, (18)
where "t" denotes transposition, fy is a vector whose components are the derivatives of f
with respect to the components of y, and fyy, fxx, and fxy are matrices whose elements are
the corresponding derivatives. In order that x = 0 be a minimum point of f at y = yo= 0,_the
matrix f~ should be positive definite. Then
min (f)o= 2 (ytf~yJ)o, (19)
x
where ()o is, as before, the averaging symbol for the distribution of probabilities for y
prior to performance of the experiment. This distribution is .now characterized. (besides the
mathematical expectation yo = 0) not by the number oa but by the covariant matrix (yyt)o with
the elements (yiy~)o. We shall find the loss from inaccurate knowledge of the value of y.
Assuming that the planned experiment gives an accurate value of y, we obtain in place of the
expression (5)
Ef- min (f)o-df -(min f)o. (20)
x x 1{
The optimal value of x for an accurately measured value of y is equal to -fzzfxby? Substi-
tuting it into the expansion (18) and averaging, we obtain '
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
(min f)o = 2 (y~f,,~y)o - 2 (ytfiyfwsfx~y)o,
x
Ef = 2 (yt/sUfxxtxgl/)o - Of.
The economic loss from a lack of information is equal to
i 7
where Cij are the elements of the cost matrix
C = 2 fxUfzxlx~? (23)
The matrix C, just as fxx; is positive definite. Therefore the loss calculated from formula
(22) is always positive.
Constraints can be imposed on x. If at y = 0 the .optimal x does not lie on the boundary,
formulas (22) and (23) remain in force for covariant matrices (yyt~o which are sufficiently
small in the norm, since the probability of the optimal value of x lying on the boundary is
small for a small indeterminacy in the values of y.
Let us consider the case in which at y = 0 the optimal value of x lies on the boundary
of the authorized region. If the boundary is smooth in the neighborhood of the optimal value
of x, one can reduce the problem to the preceding one by selecting new generalized coordi-
pates on the boundary. The new vector x will have a smaller dimensionality, and its choice
will not be constrained in any way. If the optimal (at y = 0) value of x is located at a sin-
gular point of the boundary (at a "corner"), so that the optimum point is not shifted upon
small changes in y, the cost of the information for a matrix (yyt)o which is sufficiently small
in the norm is equal to zero (ignorance of the value of y does not impose a loss).
Finally, if upon a change in y in the region in which the probability is not small the
optimal value of x can switch from the boundary into the volume or from one boundary to anoth-
er, the analytic calculation of the usefulness of the experiment is complicated. One can
do it by using the Monte Carlo method.
A Model of a Different Type. In a number of cases the transition of the parameter x
through some boundary whose position is not known completely accurately will result in large
economic losses. Moreover, if there were not these losses, the optimum would lie on the
boundary. The simplest mathematical model of such a situation is expressed by the following
dependence of the expenditures on the parameter being controlled x and the poorly known value
of the parameter y:
f(x, y)==cx-}-IIIE(x-y), c>0, III>0. (24).
Here the function E is equal to zero for nonpositive and to unity for positive values of the
argument. The statistical properties of y are determined by the expression (1); yo = 0, and
the distribution for y is Gaussian:
P (y I yo, Qo) = exp [ - ya/(2Qo)] /(Y2i~ aa).
In addition to a local minimum at x = y, the function f formally has a global minimum as
x-}~. However, we shall assume that larger values for:x are forbidden.
Averaging f over the distribution (25), we obtain
s
(~o = - cx -~ III , P (y I yo, Qo) dy
.III > c V 2~ ao,
(25)
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
a minimum of ~f~o is observed at
dP ~xont Iyo, Qo)ldxo~t > 0,
whence according to the distribution (25),
~8
xoPt= -60 ~ln 27t60C8
Substituting the expression (29) into formula (26), we find
ff IIIa t/2
min (f)o = hoc ? l ~ In Znaoc$ ~ 1--
-~ ~ erfc [ (1n iIIa ~ t/ZJ }
aac 2naoc$ ~
erfc (z) . ~ J e- t' dt ~ ~ ~ e-=a.
v-
An approximate formula which is valid for large values of z gives
( a
min (f )o%?ooc?L(ln ~as~t/Z-~(ln m28)_t~z~+ ~ ?1.
x 2~taoc 2a[voc 2s[ aoc
(28)
If y were to become known accurately as a result of the performance of an experiment,
we would have
xopt = y, min f = - cy,
x
(min f)o=0
x
and the economic loss from a lack of information is
(33)
L~ min (f)o?
(34)
x
In agreement with the expression (32), this loss vanishes when ao =
0.
However, in contrast
to the preceding models, the loss is not proportional to the dispersion oo. In the first
approximation it is proportional to.6o, i.e., to the root of the dispersion (the proportion-
ality coefficient depends on cso logarithmically, i.e., very weakly), and with good accuracy
dL z c ? ~ In Ina ` 1~2
dap 2naocE ) '
The loss increases slowly as III increases and is almost proportional to c.
An Example of the Application of the Simplest Quadratic P4ode1. Let it be known that the
annual volumes of plutonium extracted from fast reactors which are subject to refabrication
vary linearly with time:
4 = 40 -I- 4t (36)
The rate of growth. of the refabrication volumes q/qo is equal to the rate of growth of
the total installed capacity of breeder reactors. ide shall assume that it is limited by the
arrival of excess plutonium from these reactors. Actually, thermal reactors also supply plu-
tonium. But ignoring this source of plutonium, we overestimate the cost of the information
which we are talking about. Consequently, we shall assume that q/qo is proportional to the
excess of the value of the breeding ratio of the breeder reactors above unity.
We shall denote the relative error in this excess, and this also means in q, as oo.
Very large expenditures have been consumed throughout the world in the last decade for a
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
decrease of oo, and right now the absolute error of the breeding ratio is 3.5% [5, 6], so that
Qo,? 0.10.* Was there an economic necessity to refine the breeding ratio? Should one refine
it further?
Automated assembly lines should be constructed for the refabrication of plutonium. We
shall assume that the design capacity of the assembly line Q is not expressed in that part
of the instantaneous flow rates which is proportional to the load of the undertaking. One
can ignore this part of .the flow rates by selecting the design capacity of successively
installed assembly lines Qj, j = 0, 1,... In other words, we shall represent the expendi-
tures for construction and operation of the assembly line reduced to the time of its startup
which are not associated with the load of the undertaking as a linear function of Q [7, 8]:
At time to = 0 a line should be introduced with a capacity Qo subject to determination.
The times of introduction of the lines tl, ti, ... are uniquely determined by the design
capacities Qo, Q~, ..., since it is advisable to introduce a new line prior to exhausting
the production reserve of the lines already operating:
t~+1-ti=Qtl4, 7>1; tt-to=(Qo-4o)l4. (38)
The true value of q will be determined up to the time of exhaustion of the production capa-
city Qo.
The part of the reduced costs subject to minimization is equal to
/ = W (Qo) -I-- ~ ~P (Qi) exP (- Pty), ( 39 )
~=i
where P is the discount norm, i.e., the norm of the reduction of nonsimultaneous costs to a
single time [2]. When q becomes known, the equal intervals between the introduction of the
lines and accordingly the equal values of the design capacity
tz-t,=ta-tz=... =0t,
(40)
Q~=Qz=...=Q
(41)
,,ill be optimal by virtue of the time-independent nature of the problem, so that we obtain
from the expressions (37)-(41)
f=/t'-I-'kQo~--(~L'+kq~t)exp[-(Qo-go)I'~4] [1-exp(-POt))'i.
~42)
Let us introduce the dimensionless parameters
z- Pot; a = (Qo- 40) k/:%L'; ? = P:%!,'/(kq).
We obtain
(43)
- 1~-zl? ex
(--
a)
1 k /
J
l
'
'
(44)
P
?
-{-a-{
:
~
L
-
L
instead of the expression (42). After q, and this means u also, are determined, one should
for minimization of the costs find a z (and this means ~t-also) such that the expression
is a minimum. Equating the derivative with respect to z to zero, we obtain an equation for
z [8]:
*If b.r. = 1.5(1 +e), where e is the relative error of the breeding ratio, the relative
error (b.r. - 1) will amount to e b.r./(b.r. - 1) = 3e.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
exp (z)-1-z=?.
With this equation taken into account the expression (44) is transformed:
fl ~ 40/ h,' ~- a ~-- ? exp (- ?a).
Here the parameter being optimized is a, and u is undetermined. In order to reduce the prob-
lem to the simplest quadratic model discussed earlier, we shall assume
- Ego = P~C/'(k (4)0); (49
exp(zo)-1-zo=?o (51)
It follows from the expressions (43) and (47) that
4 = (4)0 (1-i- y)~ (52)
so that the relationships (1) with yo = 0 are valid for y. It is easy to see that for y = 0
the optimal value of a is equal to ao, and accordingly the optimal value of x is equal to
zero. Next
fY/~ = ao- ao ? (1 -}- ? -{- z) exp l - ?ao' (1 -F- x));
fxx/~' = ?oao' (1 -~ ?o + zo) exp (- ?oao);
fxbl~ = Iao (1 -{- ?o ~- zo) ?u - .
- ao ? (?b + Zv)) exP (- ?oao),
where
?u = - ?o; Zu = ?v~IeXp (ZO) -1) ,
We obtain from formula (16) after simple transformations using the ,expressions (50) and (51)
C /~L'- 1?o r 1 - Z0 z (53)
2 \ zo + ?o. ?o
The dependence of C/~C on the optimal dimensionless intervals between introductions of assem-
bly lines zo is given in Table 1 (the corresponding values of uoare also given there).
The intervals between the introductions of lines should scarcely exceed 10 yr, so that
with P X10-1 yr'1 we obtain zo fwt ,
for fbt Y Z and NZx > N~ , then we have a partial failtue;
if (X Z -I- YZ) ?x2 >YZ and NZx ~ N~ or x.Z = (Xz ~' ~'z)+ then we have a complete failure. (8)
Here NzX is the power of a unit on the failure of xz elements in item z. The reliabili-
ty parameters for subsystem Zm are found from the scheme for the series connection of zZ items
separately for the complete and partial failures. The equipment power and the reliability
parameters for system m on failure of ~m out of Lm technologically parallel-subsystems may
be defined as follows if the items are identical in reliability:
Nam ='YImN (Lm - lm)~I'm> rlm (t) = til,n (t):
!m i lm-1 Lm !m-1
~(t).2 wi,n(t)ze,,,(t.) CLmICLm-1~
where N is the installed power of a unit; YZm is the safety factor.on output for subsystem
Zm~ WZm(t)~ TZm(t) are the fault flux parameter and the mean recovery time for subsystem Zm;
Cym and Cy ~1 are the numbers of combinations from Lm taken ~m at a time and from Lm 1 taken
m m
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
TABLE 1. Initial Data on Equipment Re-
liability
Parameter estimates
Element or
part of station
Power
Mly ~
ii mean
pessimrstic
ptimistic
I
mxto3
h.-i
l
z. h
~?xto~.I
h-i
T. h
~,x~n~
h-i '
T.h
I
Reactor-
9000
U,2
135
0,25
150
0,1
100
2000
0,3
150
0,4
200
0,15
125
MCP
j 250
0,025
100
0,04
120
0,02
fi0
l 500
0,04
120
0,06
150.
0,025
900
Steam genera-
r 250
0,09
200
0,12
225
0,06
150
for
l 50U
0,13
225
0,15
250
0,09
200
Main pressor .
250
0,025
60
0,04
80
0,015
40
ized equip-
` 500
0,035
100
0,05
-150
0,025
60
ment
500
0,4
100
0,5
130
0,35
70
Turbine
~ 10E>D
U,6
130
0,75
150
0,45
100
2i>n0
0,7
140
0,85
165
0,55
110
Feed. pump ..
1 500
n,15
25
0,2
35
0,1
20
t 1000
0,21
3n
0,3
40
0,15
25
ondensate
250
0,12
10
0,18
15
0,09
5
pump
~ 500
0,18
15
0,25
20
0,12
10
100(1
0,2
20
0,3
25
0,15
15
Low-pressure
500
0,045
10
0,06
15
0,03
5
heater
{ 1000
0,06
15
0,08
20
0,04
10
High-pressure
500
0,04
1u
0,055
95
0,03
5
heater
{ 1000
0,055
15
0,07
20
0,04
l0
TABLE 2. Mean Estimates of the Differ-
ences in Capital Investment in Units
for Various Forms of Structure
Number of
loops and tur-
bines
4---2'
4~-1
2-x-2
2~-i
Capital investment differ-
ence, mtllton rubles
with MPE wiFhout
MPE
y
-5,9
-5,8
-19,7
*Taken as the initial form for
containing a WAR-1000.
'Taken as the initial form for
containing a VVER-2000.
Zm-1 at a time correspondingly. The form of failure in
be determined from the relation between the parameters N
-1,4
-7,:3
-7,5
-93,4
Ot
-14,6
the unit when Zm subsystems fail may
and Ntm by analogy with the above
expressions. The reliability parameters in the unit mayV~e found from the scheme for the
series-connected systems. The calculations give the reliability parameters for .all the hie-
rarchic levels in the unit for each year of operation t.
At the end of 1982, there were over 150 water-cooled power reactors working in the world,
whose total service life exceeded 2000 reactor-years. The conditions of use vary widely as
do the standard dimensions and the designs, which means that the information. on the reliability
and cost is inhomogeneous and ambiguous [3-6]. Therefore, the initial data for theoretical
researches on reliability may be specified as the average and the limits to the ranges corre-
sponding to the average pessimistic and optimistic evaluations (Table 1). The initial cost
595
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
parameters were varied from + 40% to -10% of the mean estimates. Various conditions for emer-
gency repair of the first circulation loops were also considered: shutting down the unit for
repair immediately after fault is observed (mode I), shutting down the unit for repair in ac-
cordance with the management service during the period of falling load in the EPS (mode II),
and repairing the individual components of the switched-out loop. with its main pressurizing
equipment (MPE)if possible without shutting down the reactor (mode III).
In the calculated example, forms of scheme were drawn up for units containing water-
cooled-water-moderated reactors of power 1000-2000 MW, which differed in the number of loops
in the first circuit, the presence or absence of MPE, in the number of turbines, and corre-
spondingly in the capital investment (Table 2). The transition from a double unit to a single
one containing a VVER-1000 gives an economy in the reduced costs on account of a reduction in
capital investment of an average 1.5 million rubles per year. In the case of the VVER-2000,
Co is reduced on average by 3.7 million rubles per year on going from a double unit to a
single one. On the other hand, the reliability parameters of a single unit are lower than
those of a double one because of the larger number and longer duration of the complete fai-
lures, which increases the cost for maintaining a reliable energy supply in the EPS, namely
Cu (Table 3). From the value of C for the overall reduced costs, one can say that forms of
system containing two turbines are economically more efficient than single units because
~Et is reduced by 10-25% (Table 3), with the corresponding reduction in the necessary emer-
gency backup in the EPS. This result agrees with the conclusion for units containing water-
graphite reactors [6].
The optimum number of circulation loops is dependent to a considerable extent on the
mode of transfer to repair conditions and on the presence or otherwise of MPE. If the repairs
are conducted in modes I and II, no matter whether there is MPE, the forms of units contain-
ing VVER-1000 reactors with two loops are more efficient than forms containing four loops be-
cause of the reduction in costs not only due to equipment failures but also to capital in-
vestment. The improved reliability in a unit containing two loops 'is explained by the re-
duced number of equipment items liable to complete failure. If, on the other hand, the MPE
provides for sealing off a disconnected loop and ensures conditions for safe repair of the
failed circulation loops in course (mode III), then the reduction in number of complete fai-
lures in the unit in a scheme with four loops reduces the power deficiency by 8-10% by com-
parison with the two-loop scheme. Correspondingly, the savings in overall reduced costs for
a four-loop scheme containing MPE are from 0.4-1.1 million rubles a year.
Eliminating the MPE from a four-loop first circuit in a VVER-1000 gives a saving in
capital investment of up to 1.5 million rubles, together with a relative reduction in the
deficiency in energy production in mode I by 5-8% or by 4-6.5% in mode II, which reduces the
reduced costs by 1.0-2.5 and 0.5-1.7 million rubles a year, respectively. The lower economy
in mode II occurs because the power loss from the unit on shutdown for repair is less than
that when a loop in the MPE is shut down.
We not only determined the costs with various estimates for the initial data but also
found the optimum levels of equipment reliability improvement (Fig. 1). The following rela-
tionship applies between the reliability-change coefficient f3i and the costs of the elements:
where Qoi and Qi are the costs of equipment type before and after raising the reliability,
while woi~ wi+ Toi, and Ti are the fault-flux parameters and the mean recovery times, and
ai is a coefficient given values from 0-0.6 in our calculations. Equation (10) was used to
determine the limiting permissible capital investment BkTe going to reliability enhancement.
According to Fig. 1, the optimum reliability level is attained for a single unit containing
a WAR-2000 on raising Ri by more than afactor of 2.5,while the permissible increase in the
capital investment is 20-40% of the original equipment cost.
Table 3 shows that the forms of units with two turbines are more efficient than single
units with the various reliability estimates; on the other hand, there are uncertainties in
the data, so we examined the necessary increase in the reliability in a single unit required
to make it more efficient than a double one. The calculations showed that this requires a
reduction in the fault-flux parameter-for the turbine equipment in a single unit containing
a VVER-1000 of 20-50% in accordance with the element reliability estimates and the value
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
TABLE 3. Comparison of Reliability and
Reduced Costs for Forms of a Unit Scheme
Containing VVER-1000 and VVER-2000
without MPE (Mode I)
3
~
N
0~
o~
Reliability
I
o ~~
~,
~
o T
~
`
estimates
~
'~
30
o
p C
~ ~
*w
~ J1
~ .C
a
z~
K
~ ~
ti
4-1-2
7.3,5
86,5
23,2
127 8
4-1-1
Average
16,2
83,8
26,7
12:1,,;
2-}-2
72,6
87,4
21,8
124,3
2-1-7
15,3
84,7
25,ii
12fi,5
4+2
18,9
81,1
32,2
136,3
1000
4-~1
Pessimistic
21,6
78,4
35,2
137,8
2-~2
17,6
82,4
30,4
132,9
2-1-1
2(1,5
79,5
33,7
134,6
4-1-2
7,3.
92,7
12,8
116,9
4-x-1
Optimistic
9,2
90,8
15,5
118,1
2-~2
7,4
92,6
13,0
115,5
2-1-1
9,3
911,7
15,fi
176,5
4-~2l
Average 1
22,01
78,111
73,61
269,3
4-}-1
24,8
75,2
78,7
27(1,7
200(1
4-f-21
Pessimistic 1
29,81
711,2.1
97,61
293,3
4-}-1
33,7
67,3
104,5
29(1,5
4~-2 Optimistic 13,6
1
86,4 45,9 241,6
1
1
4-}-1 14,5
85,5
50,8
242,8
*In the fourth year of operation with a S-
yr repair cycle the repair durations for
a unit containing a VVER-1000 were 1080,
1440 and 1080 h a year and those for
VVER-2000 were 1180, 1460, and 1180 h a
year.
Fig. 1. Dependence of the relative reduced costs dC
and capital investment dkTe in a single unit con-
taining a VVER-2000 on the reliability-change coef-
ficient Ri with the following estimates for the re-
liability: optimistic (a), average (b), and pessi-
mistic (c) with the values of a = 0 (1); 0.4 (2);
0.5 (3); 0.6 (4); dC is given relative to the value
of Co before optimization.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
/3-p~~ 1
/ /~
S
- / //
~ G~
/ a
-2 -1 0
9 2 d91,
b
/
/ /
~
~~l
~/ 3
------- ~ ~/?
-----
Fig. 2. Equal-performance limits for a single unit (a)
and a double unit (b) containing VVER-2000 with the op-
timistic reliability estimate (1), the average estimate
(2), and the pessimistic one (3) in relation to the dif-
ferences between them in providing the planned output
0~ _ ~d - ~m and as regards capital investment Ak =
kd - km, (the subscript m corresponds to a single unit,
and subscript d to a double one): ) mode I; - - - -)
mode II; the hatched rectangle shows the relation be-
tween Ak and ~~ for the average estimates for the re-
liability and equipment cost.
of a. The additional specific capital investment required to improve the reliability should
not exceed 7.7-1.7 rubles/kW, which appears feasible.
At present, there are, discussions on the numbers of cylinders for turbines of power 500
or 1000 MW and above, as well as on simplifying regeneration systems while simultaneously in-
creasing the heat output to users and the desirability of other changes in turbines that af-
fect the cost, reliability, and efficiency. As regards the economy in direct use of metal
and labor, turbines of higher power have potential advantages, and the same applies as re-
gards the cost of the control and monitoring equipment. Theoretical studies enable one to
estimate the effects from these improvements, as well as from changes in the parameters of the
nuclear. power station as a whole and of the EPS, which influence the numbers of turbines and
other items in a unit. A boundary zone of equal efficiency has been identified-for single
and double units for various estimates of the equipment reliability (Fig. 2). Figure 2 shows
the relation between reliability and capital investment for double and single units, which
enables one to determine the effectiveness of various designs intended to raise the reliabil-
ity in VVER-2000 systems. Similar relationships have been derived for units containing WER-
1000.
The following conclusions are drawn:
1) It is important to consider equipment reliability in defining the schemes for nuclear
power. station units;
2). zones of equal ef.f.iciency have been defined for structural forms of units containing
VVER reactors, as well as the zones of higher efficiency subject to alterations in reliability,
equipment cost, and other factors;
3) it has been shown to be economically desirable to increase the reliability of a single
unit containing VVER reactors by making additional capital investments that enable one to re-
duce the costs for providing EPS backup and equipment repair;
4) an economic comparison of different forms shows that it may be desirable to use two
circulation loops in a unit containing a VVER-1000 subject to the condition that the required
reactor safety is provided.
1. L. S. Popyrin, Mathematical Simulation and Optimization for Thermal Power Systems (in
Russian], Energiya, Moscow (1978).
2. B. A. Kozlov and I. A. Ushakov, Handbook on Calculating Reliability for Automatic and
Electronic Equipment [in Russian], Sov. Radio, Moscow (1975).
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
3. A. I. Klemin, Statistical Engineering Calculations in the Design of Nuclear Reactors
[in Russian], Atomizdat, Moscow (1973).
4. S. M. Kaplun, Optimizing Reliability in Power Systems [in Russian], Nauka, Novosibirsk
(1982).
5. Yu. V. Smirnov, D. D. Sokolov, and I. D. Sokolova, The Nuclear Industry in Foreign Co-
untries [in Russian], Atomizdat, Moscow (1980).
6. I. Ya. Emel'yanov, A. I. Klemin, Yu. I. Koryakin, et al.,'''A comparative analysis of
reliability and economy in nuclear power station units containing one and two turbine
systems," At. Energ., 53, No. 2, 67-70 (1982).
.EFFECTS OF IRRADIATION ON THE ELASTOPLASTIC
DEFORMATION OF Zr + 1% Nb FUEL PIN CLADDING
V.
A.
Matushkin, A. A. Medvedev,
Yu.
V.
Miloserdin, B. D. Semenov,.
Yu.
K.
Bibilashvili, and I. S. Golovnin
It is extremely important to have information on the mechanical properties of structural
materials under actual working conditions in order to improve the performance of nuclear re-
actorsa In particular, in order to choose the optimum conditions for running the reactor up
to power one needs experimental results obtained at very low neutron fluences (up to 1024
m-2). Also, such data are of some value in researching .radiation damage-accumulation. In
this connection, we have examined the effects of reactor. irradiation on the resistance to
elastoplastic strain in standard fuel-pin sheaths of diameter 9.15 X 0.65 mm made of Zr + 1%
Nb alloy.
In accordance with the standard method [1-3], the tubular specimens underwent cyclic
sign-varying torsion with a constant strain amplitude o.f ? 0.7%. The small tube thickness
enabled us to analyze. the results by means of standard. relationships for the shear deforma-
tion Y and the tangential stress T:
R M
~ - ~ l ~ - ~ ~Ra+T2) ~R-T) . ~
where ~P is the angle of torsion in the working part of the specimen of length Z, Rand r are
the outside and inside radii of the cylindrical tube, and M is the torsional moment.
The plastic strain yc was. determined as the difference between the total y and the elas-
tic strain:
t
Y~=1'- ~ ,
where G is the shear modulus.
The tests were performed with TsIRK apparatus [4] as follows. Initially the elastoplas-
tic strain in the fuel-pin sheath. was examined under laboratory conditions over a wide range
in strain rate: 8 x 10-4-8 x 10-9 sec`'. at 295-780?K. At each temperature, we used 20-30
strain cycles with a. rate a = 8 x 10-4 sec`1. The cyclic-strain diagram after temperature
change stabilized within 3-5 cycles. In some cycles,- the material was kept at constant load
(in creep testing) for short times (up to 2 x 104 sec) or under monotonically decreasing load.
Such tests enable one to estimate virtually all the parameters required to forecast the ac-
cumulation of .plastic strain in a given state of stress [1]: the temperature dependence of
the elastic moduli, the yield point, the sensitivity to the strain rate, and the change in
deformation resistance during steady conditions under load [5, 6].
After this, the apparatus together with the specimen was placed in the core of an IRT-
2000 reactor. The flux densities for the thermal neutrons and fast ones (E > 0.1 McV)were
1.2.1017 and 1.8~1016tq 2?sec-1,respectiyely. The.tests.under irradiationwere.performed at 540-
Translated from Atomnaya Energiya, Vol. 57, No. 3, pp. 162-165, September, 1984. Orig-
inal article submitted August 1, 1983; revision submitted March 21, 1984.
0038-531X/84/5703- 0599$08.50 8 1985 Plenum Publishing Corporation 599
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
~ 2s
., .
,o
>~oa
60 ~zv
Fig. 1. Temperature dependence of the
shear modulus G(~, Oj and of the yield
point To.z(~, v~ ?, ~) before irradia-
tion; O, ~) ~ = 1.5 X 1023 m 2; A, o )
after irradiation.
Fig. 2 Fig. 3
Fig. 2. Dependence of the yield point for Zr + 1% Nb alloy on.ther-
mal-neutron fluence at 540?K.
Fig. 3. Cyclic-deformation diagrams for Zr + 1% Nb.
7.70?K. During the reactor shutdown, the temperaturefell to 320?K. During 95% of the time, the irra-
diation was under conditions of self-heating at 540?K.
The graph of the temperature dependence of the cyclic shear modulus (Fig. 1) shows that
irradiation to a thermal-neutron fluence ~ = 1.5 x 1023 to z increases .the shear modulus by
10-13% at all temperatures, with the change in modulus occurring at ~ = 1021 m-2 and per-
sisting in the range 1021-1023 m 2. Also, the results obtained with the reactor working and
after it had been shut down are described by a single temperature dependence, which shows
that this low flux density has no effect on the elastic modulus.
The cyclic yield point "CO,z was determined from the steady-state diagrams for sign-vary-
ing deformation in the range 300-780?K. Figure 1 also shows the temperature dependence of
the yield point obtained before irradiation, during irradiation, and during reactor shutdown.
The increase in the yield point was determined only by the neutron fluence and was ndepen-
dent of the flux density.. Figure 2 shows the effect of the fluence on the yield point at
540?K,
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
LR ~ca
~"~ .
p~ a "-1, 2S
20 75 70 S 0 5 70 >S
zo- z+/i, MPa zd-z, MPa
Fig. 4. Dependence of In (Yca/Yc) on the
difference in the effective stresses. during
irradiation at 540?K: o) creep at T = 81 MPa;
D, ~) relaxation at To 60 and 41 .MPa, respec-
tively.
~ 70
a
so ~ 0,5
4 .
QI I ~ i i i i i 0 1 ~ i I i ~ i
300 500 T, K J00 400 500. 600 700 T, 'K'
Fig. 5 Fig. 6
Fig. 5. Temperature dependence of the parameters r(o), p(~), and A(o)
prom (4).
Fig. 6.' Dependence of a on temperature with allowance for dynamic
strain aging ( ) and without such allowance (- - -): ~) before
irradiation; O) during irradiation; ~) after irradiation; - - -) a* _
dln Yc6?/d('[a - s); - ) a = dln '~~c /d(ia - ti ~- li) .
The increases in shear modulus. and plastic-deformation .resistance on.irradiation sub-
stantially affect the diagram for cyclic sign-varying strain (Fig. 3). To estimate the,ef-
fects of irradiation on the accumulation of plastic strain, over 1000 creep and stress-relaxa-
tion tests were performed. Orovan's relation was used in processing the results:
where b.is Burgers vector; pm is the mobile-dislocation density; V(T - Ti) is the mean speed
of the mobile dislocations, which is a function of the effective stresses Te = T - Ti; T is
the applied stress; and Ti is the inverse internal stress, which is dependent on the accumu-
lated plastic strain and on the loading history.
If we assume that the mobile-dislocation density and the internal stresses are dependent
only on the accumulated plastic strain at the given fluence, flux density,,and temperature,
then one can derive a relationship between the plastic strain rate Yc at the stress T(Yc) and
the rate Ica under conditions of strain rate a:
Yca a V ~tia NYC)-~i (YC)1
V ~ti (Yc) -tii (YC)1 '
Vc
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
Fig. 7. Strain diagram for Zr + 1% Nb on changing the strain
rate a by a factor 103 at T = 593?K: -) experiment; - - -)
~alrulatinn frnm (31 and (41_
where Ta(Yc) is the resistance to strain with rate a.
For an exponential relationship between the mean speed of the mobile dislocations and the
effective stress we get
The plastic-strain rate at a constant rate of displacement in the loading tie is defined
Yert=al (1+av~ l G + c ~J~
where G is the equipment rigidity and dTa/dYc is the current plasticity modulus.
Then the coefficient a defining the sensitivity of the material to strain rate (a =
dlny~ )can be found from the slope of the experimental. curve constructed from the creep-
dt I T, Vc
test data or from stress-relaxation results in ln(Yca/Yc)-(Ta - T) coordinates (Fig. 4). The
agreement between these results indicates that the. initial assumptions are correct. Then
knowing a and the strain diagram, one can"calculate the rate of plastic strain accumulation
from
Y?=[a/1+dYc l~+c~~exp{a(i-
However, the dependance of ln(Yca/Yc)_?n Ta - T is substantially nonlinear in the range
500-600?K (Fig. 4), which is due to dynamir_ strain aging, which has a considerable effect on
the behavior of Zr + 1% Nb at the working temperature.
The increase in strain resistance after a temporary reduction in the strain rate or par-
tial unloading due to dynamic strain aging may be related to the increase in the internal
stresses by the value f3 [6]. Then to retain the given .plastic strain rate it is necessary to
increase the applied stresses by DT = R. Therefore, one can judge the changes in the internal
stresses from the changes in the applied ones under conditions of variable-rate strain.
The results show that the degree of strain aging is dependent on-time and on the plastic
strain. With an infinitely. small increment in the internal stresses, one can write the fol-
lowing approximation for this material [6]:
d~ = r(A - S) dt - p~ d yC, (4 )
where r, A, and p are parameters to be determined from experiment which are dependent on tem-
perature (Fig. 5). A notable result from these studies is that irradiation does not affect
the dynamic strain aging parameters.
The change in the inverse stresses on aging during creep or stress relaxation must be
incorporated in determining the sensitivity to strain rate:- the coefficient a is calculated
from the slope of the curve in ln(Yca/Yc)-(Ta - T ~'S) coordinates (Fig. 4),. In accordance
with Fig. 6, the data obtained before irradiation, during it, and afterwards lie on a common
curve, which shows that the sensitivity of Zr + 1% Nb to strain rate remains unchanged within
the limits of the fluence and neutron flux density used.
Joint consideration of (3) and (4) enables one to describe the fairly complicated pheno-
mena observed in tests at variable strain rates and under the corresponding working condi-
tions in ,fuel-pin sheaths when the reactor power varies. Figure 7 compares the calculations
with experimental data obtained under conditions of nonstationary loading (with the strain
rate reduced by a factor of 1000).
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
These results indicate that one can use (3) to forecast the accumulation of plastic
strain under conditions of brief nonstationary loading at high temperatures and low fluences.
In practical calculations, allowance must be made for the changes in Ti"and a due to the dy-
namic strain aging and for the changes in G and Ta caused by the irradiation conditions. Ad-
ditional experiments are required to establish whether these formulas can be used for more
prolonged irradiation and higher flux densities.
1. B. D. Semenov, Methods of Examining Creep and Relaxation in Refractory Materials under
Sign-Varying Torsion over a Wide Temperature Range: Ph. D. Thesis [in Russian], MIFI,
Moscow (1975).
2. Yu. V. Miloserdin, V. N. Chechko, and B. D. Semenov: Proceedings of the All-Union Sym-
posium on Few-Cycle Fatigue at Elevated Temperatures [in Russian], Issue 1, ChPI,
Chelyabinsk (1974), p. 123.
3. Yu. V. Miloserdin, V, N. Chechko, and B. D. Semenov, Inventor's Certificate No. 532032,
Byull. Izobret. No. 38 (197"6).
4. A. S. Glinskii et al., Reactor Tests on Materials [in Russian], Energoatomizdat, Moscow
(1983).
5. Yu. V. Miloserdin et al., Techniques in Radiation Experiments [in Russian], Issue 9
(1981)., p. 116. .
6. V. A. Matushkin et al., Prob1. Prochn., No. 6, 80 (1981).
V. A. Nikolaev, V. I. Badanin, UDC 621.039.531
and A. M. Morozov
Reducing heat treatment is a means of increasing the working life of pressure vessels
of"water-cooled-water-moderated reactors (WER), but its use requires a knowledge of the laws
of radiation embrittlement in materials under conditions of repeated irradiation in order to
provide a reliable forecast of viability on further use. Since it is far from always desir-
able or possible. to bring about complete recovery of the properties of the material by re-
ducing heat treatment, it is important to establish how the residual damage sums with the
radiation damage produced by repeated irradiation. The available information is very re-
stricted because the corresponding studies are very complicated [l, 2].
In the early experiments [1], it was found that the residual shift in the embrittlement
temperature OTres (50?C), persists .in A350-LFl steel after irradiation to a neutron fluence
of 3 x 1023 m-2 and annealing at 307?C, and this should be borne in mind in estimating the
total shift ~Tt after repeated irradiation. It is possible to do this on the assumption that
the material has already accumulated some damage at the start of the repeated irradiation,
this damage corresponding to an equivalent neutron fluence Fe,-which causes a radiation em-
brittlement effect of ~Tres? This approach is-less conservative than the ine.thod of indepen-
dent (arithmetic) summation of Tres with the shift due to the neutron fluence on repeated
irradiation (ATF2), and so if it is to be used in calculations one requires a reasonably reli-
able experimental basis for it, the more so since ambiguous results were obtained in a series
of experiments on the irradiation of welded joints in A533-B steel: a difference-from the
results from repeated irradiation after annealing at 399?C, which confirmed the results of
[1], while the changes in Tt resulting from annealing at-343?C and subsequent irradiation
agreed best with the view that the damage sums"independently [3].
In this connection, experiments were performed on repeating .the irradiation of Soviet
reactor steel following variable intermediate-annealing temperature and determination of
Translated from Atomnaya Energiya, Vol. 57, No. 3, pp. 165-167, September, 1984. Orig-
final article submitted October 13, 1983; revision submitted January 31, 1984.
0038-531X/84/5703- 0603$08.50 ?1985 Plenum Publishing Corporation
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
~Tres? We examined 15Kh2MFA steel made commercially with phosphorus and copper contents of
0.013 and 0.15%, respectively. An abbreviated program was also used in testing 15Kh2NMFA
steel (0.012% phosphorus, 0.12% copper). As it was necessary to irradiate a large-number of
specimens simultaneously under identical conditions in order to determine Tt, the embrittle-
ment temperature, for 15Kh2MFA steel, we used impactor specimens of size 5x5 x27.5 mm. The
15Kh2NMFA steel was tested in the form of standard specimens.
The cassettes containing the specimens were placed in ampules of identical internal di-
ameter and were irradiated in the core of the VVR-M reactor. The cassettes were inserted
in the ampules as sliding fits and the irradiation temperature was not more than 250?C. The
temperatures of the specimens during the first irradiation were monitored with 8 Chromel-
Alumel thermocouples. On repeated irradiation, the specimens were sealed in by means of a
conical fitting ground into the head of the ampule and the edges were rolled over by remote
control. In that case, it was impossible to insert thermocouples, and diamond indicators
were used to monitor the temperature.
The initial irradiation of the 15Kh2MFA specimens was performed to a neutron fluence
of F, = 7.5.1023 m 2 (E >0.5 MeV). The irradiation temperature varied over the height of the
ampule over the range 180-200?C at the start of the exposure and up to 215 ? 15?C in the closing
stage. The rise in the embrittlement temperature ~TF1 under these conditions was 100?C, which
corresponds to an irradiation embrittlement coefficient of AF1 = ~TF1/(F1.10-22)1/3 = 24. An-
nealing for 100 h at 240, 275, or 325?C produced reductions in the shift in Tt correspondingly
of 80, 50, and 20?C, i.e., the degrees of recovery were 20, 50, and 80%.
The repeated exposure was given to previously irradiated specimens either with or without
annealing as above, as well as to control specimens, which had not previously been irradiated.
In the latter, the irradiation to a fluence of F2 = 8.7.1023 m 2 raised Tt by 130?C, which
corresponds to AF2 = 29. The value of AF2 is higher than AF1 evidently because of the lower
temperature in the repeated irradiation (100-120?C).
Specimens annealed at 240, 275, and 325?C showed additional increases in Tt of 70, 70,
and 100?C, respectively. The smallest shift in Tt on repeated irradiation (50?C) occurred
in the material that had not been annealed. Therefore, the embrittlement of 15Kh2MFA steel
on repeated irradiation, which is characterized by the additional shift in Tt, is the less
the more the material has been embrittled after the preliminary treatment: irradiation or
annealing after irradiation (Fig. 1).
The first exposure of the 15Kh2NMFA steel specimens was at TF = 160-180?C with a neutron
fluence F1 of 8.3 x 1023 m 2, which gave OTF1 = 160?C; annealing the specimens at 350?C for
100 h resulted in Tres = 20?C. After repeated exposure (F2 = 9.1023 m 2) ~TF for the un-
0 SO 100 d T~~, ?C
Fig. 1
0, 2 0, 4 0, 6 O,B ~
Fig. 2
Fig. 1. Dependence of the rise in viscous-brittle
transition temperature ~TF2 on repeated irradiation
on the preceding shift ~Tt due to preliminary irradia-
tion and incomplete annealing. The curves have been.
constructed from (4): p) steel 15Kh2MFA, fluence
8.7 x 1023 m 2; ~) steel 15Kh2NMFA,`neutron fluence
9 x 1023 m-2.
Fig. 2. Dependence of the increase in permissible
neutron fluence on the degree of recovery in Tt on
annealing steels 15Kh2MFA (1) and 15Kh2NbFA (2).
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
annealed specimens increased to 220?C, i.e., by 60?C. In specimens that had been annealed,
the additional shift in Tt was 160?C, as against 180?C in the control specimens. Therefore.,
the experiments indicate unambiguously that there is nonadditive summation of the residual
damage (after annealing) and the damage caused by repeated irradiation.
We now transfer to considering the damage summation principle and first consider the
behavior of 15Kh2MFA steel under conditions of stepwise change in irradiation temperature
from T1 to T2, where T1 > TZ, where consequently there is no annealing of the defects formed
at Ti during the second stage of irradiation. We assume formally that the value of ~TF1 cor-
responds to Fe at T2, i.e., Fe = (~TF1/AF2)9. Then the total shift is
0T? = AF2 (Fe ~- Fz) i~3 - (AF~F'i -~ AF2Fz)1/s (1)
The justification for summing the damage in accordance with (1) can be checked by means
of the above experiments with 15Kh2MFA steel: ~Tc = (243.75 +293.87)1/9 ='147?C .with the
experimental value ~Tc = 150?C.
The summation for T1 > Tz was checked in independent experiments, whose detailed descrip-
tion falls outside the scope of this paper. We mention only one result, which was obtained
for the metal in a welded joint, in 15Kh2Pg'A steel. The specimens of this material were orig-
inally irradiated at 290 ? 10?C to a fluence of 4.4 x 1023 m 2 and then additionally to a flu-
ence of 2.7 x 1023 m-2 at 225 ? 10?C. The increment ~Tt due to that irradiation was 140?C,
whereas (1) gives ~Tc = 152?C.
A similar-result was obtained for 15Kh2NMFA steel, for which, however, the change in Tt
satisfies a dose dependence of the form
where CF is a coefficient dependent on the irradiation temperature that characterizes the
radiation stability of the steel. Statistical processing was applied to data from 17 experi-
ments in which the fluence varied from 2 x.1023 to 3.9 x 1024 is 2, which .showed that with ir-
radiation temperatures of 200 ? 20?C and 240 ? 10?C the .power-law dependence of the embrittle-
ment for this steel applies with powers of m = 0.446 and 0.532, respectively. In this con- .
nection, the summation formula for 15Kh2NMFA steel was similar to {1), but AF should be re-
placed by CF, while the powers 1/3 and 3 should be replaced by 1/2 and 2. In the. above ex-
periment, Cgi = 17.5, CF2 = 19, and the calculated value of ~Tc was 240?C, while the actual
value was 220?C.
We now consider the damage summation on partial recovery during annealing (Tres > 0) and
repeated irradiation. By analogy with (1), we get for 15Kh2MFA steel that
~T?=(47reS -~AFZI'z)ila
The values of ~Tc calculated from this formula withTres of 20, 50, and 80?C as given
by experiment were correspondingly 128, 130, and 137?C, which agree with the experimental
values (120, 120, and 150?C) within the error with which OTt is determined.
If the dose dependence of the radiation embrittlement is described in general by a power
law with power m, then the summation formula becomes
Calculation from (4) for 15Kh2NMFA steel (m = 1/2, QTres = 20?C, F2 = 9.1029 m 2, CF2 =
19) gives ~Tc = 182?C while the experimental value is 180?C. '
These relationships imply important conclusions on the calculation of the limiting per-
missible neutron fluence under conditions of irradiation after annealing. Consider a pres-
sure vessel made of 15Kh2MFA steel for which the established limiting permissible fluence is
[F1] _ ([~Tt]AF)3, where [~TtJ is the limiting permissible shift in Tt allowed by the working
conditions. If [F1] has been attained and the vessel is annealed with a degree of recovery
in the properties of the material n = {[~Tt] - QTres} [~Tt], then in accordance with .(3) the
permissible fluence for the subsequent operation is
605
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
foTt13-oT? - (5)
- ----- ---- -F- _ _ _. -
Figure 2 shows the dependence of [Fz]/[F1] on n as expressed by (5). We see that to in-
crease [F,,] by 80% it is sufficient to have p = 0.4, i.e., 40% recovery in the properties
of the material. As to increase n requires a corresponding increase in the annealing tempera-
ture, one assumes that recovery of the properties by 40-60% is an optimal specification for
the annealing. For structures made of 15Kh2NMFA steel, the power in (5) varies by a factor_of
2, as previously, and the corresponding curve is flatter (Fig. 2).
1. U. Potapovs, J. Hawthorne, and C. Serpan, Nucl. Appl., 5, No. 6, 389 (1968).
2. L. Steele, At. Energy Rev., 7, No. 2, 3 (1969).
3. J. Hawthorne, in: Irradiation Embrittlement, Thermal Annealing and Surveillance of Re-
actor Pressure Vessels, IAEA, Vienna (1979), p. 181.
EFFECT OF CHEMICAL COMPOSITION AND ANNEALING CONDITIONS
ON THE RADIATION EMBRITTLEMENT OF THE METAL OF LOW-
ALLOY WELDED SEAMS
V. A. Nikolaev, A. M. Morozov, V. I. Badanin, UDC 621.039.531:621.791.053
A. S. Teshchenko, and R. P. Vinogradov
The metallurgical factors determining radiation embrittlement of low-alloy reactor steels
have been studied in sufficient detail. However, for the metal of welded joint, mainly de-
termining the operating efficiency of the structures on the whole, data about the mechanisms
of the effect of chemical composition and heat treatment on radiation embrittlement are quite
few [1-3]. This has served as the basis for the study of radiation embrittlement of the metal
of welded seams as a function of their annealing conditions and the contents of certain
alloying and impurity elements.
In order to explain the role of alloying elements in the radiation embrittlement of the
metal of seams, the effect of the contents of nickel and manganese,-which improve the strue-,
ture of the metal weld, its strength, and resistance to brittle fracture [3], was studied,
but, judging by the data for the base metal [4-6], they can have an unfavorable influence on
the radiation stability. As the content of impurity elements (phosphorus and copper) is an
important factor controlling the radiation embrittlement of steel and the metal of welded
seams [1-3, 7], it was advantageous to study in detail the effect of these elements on the
embrittlement of the metal of specific seams.
Material and Investigation Procedure. The effect of chemical composition was studied
on the metal of laboratory welded seams, obtained by the use of experimental batches of wire
of Cr-Mn~Ii-Mo composition, which is the basis. of Sv-08KhGNMTA wire [7] and which is intended
for the automatic welding of the pressure vessels of water-cooled/water-moderated power re-
actors. In the metal of the seams investigated, the phosphorus content .varied within the
limits of 0.008-0.022%, copper 0.03-0.35%, nickel 0.76-2.16%, and manganese 0.47-0.95%. In
each individual series of compositions the content of one element was varied as a rule, the
contents of the others being almost unchanged.
Grade E-12 (0.18% Cu, 0.01% P) and 03ZhR (0.01% Cu, 0.004% P) irons were used in smelting
the metal of the experimental compositions. The fusions, with a mass of 160 kg, were carried
out in an open induction furnace. The elements being varied were added in the furnace (nickel,
manganese) or in the ladles (copper, phosphorus) in proportion with the casting of the metal
in ingots of mass 16 kg. Wire with a diameter of 5 mm was obtained from these ingots by
forging and subsequent drawing. Welded samples were prepared under grade 48NF-18M flux
(Technical Specification 5.965-4011-72); plates of 15Kh2NMFA.steel with a content of 0.017%
Translated from Atomnaya Energiya, Vol. 57, No. 3, pp. 167-172, September, 1984. Orig-
final article submitted October 13, 1983.
606 0038-531X/84/5703- 0606$08.50 ?1985 Plenum Publishing Corporation
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
TABLE 1. Chemical Composition, Cooling Conditions after Annealing and Irradiation,
and Increase under Irradiation of the Critical Temperature of Brittleness Tc of the
Metai of the Welded Seams Investigated
Series
Content of varied elements, maSs%
~
Irradiation
and com-
I
I
I
Cooling
neutron flue- I
temperature,
Tc, ~
position No.
Mn
xt
r
cu
ence, cm- 11
~
1-i
1,0
1,5
0,008
0,04
In the furnace to
4,3.1019
270-340
20
i-2
1,0
1,5
0,012
0,04
300~,then in air
4,3.1018
270-340
20
1-3
i,0
i,5
0,015
0,04
-
4,3,1018
270-340
30
i-4
1,0
i,5
0,020
0,04
4,3.1018
270-340
40
i-5
1,0
1;5
0,015
0,32
4,3.1018
270-340
80
i-6
i,0
1,5
0,020
0,35
4,3.1018
270-340
90
2-1
0,63
4,28
0,011
0,07
In water
2,5?.1010
320-340
70
2-2
0,63
1,28
0,015
0,07
2,5?!019
320-340
80
2-3
U;63
1,28
0,018
-0,07
2,5.1019
320-340
90
2-4
0,63
.1,28
0,022
0,07
2,5.1019
320-340
100
3-i
0,63
1,28
0,010
0,07
in water
2,5.1018
320-340
60
3-2
0,63
1,28
0,010
0;10
2,5.1018
320-340
80
3-3
0,63
1,28
0,010
0,135
2,5?.1010
320-340
120
4-i
0,75
1,30
.0,011
0,03
In the furnace to
2,6.1019
290-310
40
4-2
0,75
1,30
0,011
0,10
300`C, then in air
2,6.1018
290-310
50
4-3
0,75
1,20
0,011
0,18
~
2,6.1019
290-310
70
4--4
0,75
1,20
0,011
0,27
2,6.1018
290-310
90
5-i
0,47
U,91
0,012
0,19
In the furnace to
2,6.1018
290-310
6(1
5-2
0,47
1,70
0,012
0,19
300.-then in air
~
2,6-1019
290-310
80
5-3
0,50
?0,76
0,012
0,17
2,4.1019
270-300
50
5-4
0,50
1,42
0,012
0,17
2;4.1018
270-300
70
5-5
0,50
2,16
0,012
0,17
2,4.1018
270-300
90
6-i
0,47
l,20
0,012
.0,19
In the furnace to
2,6.1019
290-310
60
6-2
0,64
1,20
0,011
0,19
300`C, then in air
2,6?!019
290-310
60
6-3
0,75
1,20
6,011
0,18
2,6.1019
290-310
70
6-4
0,82
1,20
0,012
0,19
2,5.1019
3(10-310
8b
6-5
0,83
1,20
0,011
0,19
2,5.1018
30^-310
80
6-6
0,85
1,20
0,012
0,19
2,5.1018
300-310
90
6-7
0,95
1,20
0,011
0,19
2,5.1019
300-310
90
7-i I
0,60 I
1,i
0,007 I
0,26 I
In the furnace to ~
2,8.1019 I
290-310 (
40
7-2
0,63
1,0
0,004
0,26
300, then in air
2,8.1019,
290-310
40
phosphorus and 0.14% copper and with a thickness of. 50 mm were used. Welding was carried
out in the following conditions: Iweld = 600-650 A, Uarc - 38-40 V, Vweld = 24 m/h, reverse
polarity, and temperature of heating up of the plate before welding was 150_200?C. The chem-
ical composition of the metal of the welds investigated is given.in Table 1.
The heat treatment of the samples was high tempering with the following conditions,:
heated to 650?C and held for 10 h. Most of the welded samples (see Table 1, No, 1, 1-6; No.
4, 1-4; No. 5, 1-5; No. 6, 1-7; No. 7, 1-2), in accordance with. a realistic regime of heat
treatment of the articles, were cooled in the furnace to 350?C~ (rate of cooling 20?C/h), and
then in air. The materials .of two series of compositions, with a variable phosphorus and
copper content, after annealing were cooled in water in order to study the effect of impurity
elements with a high cooling rate of the welds.
The tendency of the materials to radiation embrittlement was estimated by the values of
~Tc - the shift of the ductile~rittle transition temperature Tc caused. by irradiation. The
values of Tc were determined by the impact viscosity serial curves, using notched samples with
a size of 5 x 5 X 27.5 mm (radius of the bottom of the notch was 0.25 mm and the .depth 1 mm).
The testing temperature, at which the impact viscosity amounted to one-half of the maximum
value, was used for Tc.
The samples were irradiated in the core of the WR-M reactor by the usual procedure [4].
The temperature was measured with a Chrome!-Alumel thermocouple, spot-welded to the samples.
*As in the Russian original; the temperature given in Table 1 is 300?C - Publisher.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
dT~;c
7S
SO
0005 0015 P,% - 0 0,1 0,2 0,3 CU,% ! 0
Fig. 1. Effect of the content of phosphorus (a) and copper (b, c) on
the increase in Tc of the metal of welded seams of Cr-~In-~1i-~1o composi-
tion, by the action of neutron irradiation. a: l) compositions 1-1 to
1-4 (~), neutron fluence 4.3.1019 cm z at 270-340?C; 2) compositions
2-1 to 2-/+ (V), 3-1 (C~, 2.5.1019 cm-Z at 320-340?C; 3) compositions 4-4
(O), 7-1,
7-2 (~) (2.6 to 2.8)?1019 cm-Z at 290-310?C; b: 4) composi-
tions 3-1
to 3r-3 (O), 2-1 (0), neutron fluence 2.5.1019 cm-Z at 320-
340?C; 5)
compositions 4-l to 4-4 (O),. 6-3 (~), 2.6.10i9 cm 2 at 290-
310?C; 6)
compositions 1-3,1-5 (~), 4.3.1019 cm 2 at 270-340?C; c: ~)
1.95-2.01% Ni; ^) 1.45-1.64% Ni, neutron fluence with energy E > 1 MeV
2.8.1019 cm Z at 288?C (l, 2]; ~) 1.65-1.75% Ni, neutron fluence with
energy E > 1 MeV 3.5.1019 cm-2 at 288?C [1, 2].
There was no possibility of regulating the temperature of the samples during irradiation be-
cause, during the reactor run, a monotonic increase in temperature was observed within the
limits of 20-40?C. In view of the fact that in this situation the temperature conditions of
the final stage of exposure [8] set the determining value for radiation damage to the mate-
rial, as the principal characteristic of the irradiation regime we assumed the maximum values
of the temperature of the samples.
The fluence of neutrons with energy E ? 0.5 MeV was determined on the basis of the power
output of the reactor during the time of irradiation, preliminary measurements of the flux
density at the site of location of the samples, and additional monitoring_ by irradiated acti-
vation detectors together with the samples. The 5`'Fe(n, p)54Mn reaction was used, with a
cross section of ~ 92 mb (1 b = 10-ze m2) for neutrons with energy in excess of 0.5 MeV of
the energy spectrum of the VVR-M reactor core.
Results of the Investigation. The experimental data about the effect of irradiation on
the increase in Tc fora different phosphorus concentration in the metal of the welded seams,
with a copper content of 0.04-0.07% and different cooling rates after annealing, are given
in Table 1 and also in Fig. la. The data presented were obtained for irradiation at a maxi-
mum temperature of 340?C. It follows from Fig. la that with increase in the phosphorus con-
tent, the values of ~Tc increase almost linearly. In this case, for the material cooled in
water (curve 2), despite the somewhat lower neutron fluence, markedly greater variations of.
Tc were obtained than for materials cooled in the furnace (curve 1). An even sharper inten-
sification of embrittlement with increasing phosphorus content was found for the metal of
welds containing 0.26-0.27% copper and subjected to slow cooling after annealing (curve 3).
If the high values of ~Tc in this case can be partially explained by the lower irradiation
temperature (ti 310?C),then the steeper slope of curve 3 confirms the intensification of the
effect of phosphorus in the presence of copper.
The effect of the copper content on radiation embrittlement of the metal of the welds
investigated is described in Fig. lb. On the curves of the concentration relation, con-
structed on the results of tests of individual series of compositions, the points are plotted
corresponding to the experimental data for materials with identical cooling conditions. and
with a similar content of other elements and with similar irradiation conditions. It can be.
seen that for all materials the shift of OTc increases linearly withr.increase of the copper
concentration from 0.03 to 0.3%. For materials having been slowly cooled after annealing,
the curves of the concentration relationship have a lesser slope and, in addition, they are
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
disposed lower, even in the case of irradiation at a lower temperature. With copper concen-
trations < 0.05%, a tendency to merging of the curves is observed. For metal with slow cooling
after annealing, a neutron fluence of ti 2.5.1O19cm 2, and an irradiation temperature of
300-310?C, an increase in the copper content by 0.2% causes an increase in ~Tc by approxima-
tely 50?C, while in the same conditions an increase in the phosphorus content by 0.005% led
to an increase in ~Tc by approximately 40-45?C. Thus, the effect of copper on embrittlement
of the welds is more than a factor of 10 weaker than the effect of an equal amount of phos-
phorus.
For comparison with the experimental data obtained, Fig. lc shows the results of a study
of radiation :embrittlement of the metal of experimental welded seams of similar composition
(2.SCr-i~In-iJi-~Io) with a copper content varying .within the limits of 0.01-0.27% and with a
phosphorus content of 0.-007-0.012% [1, 2]. After annealing, the welded samples were cooled
at the rate of 56?C/h. Samples_ of these welds were irradiated at 288?C. It follows from a
' comparison of Figs. 1c and la that with a concentration of < 0.05% copper in the seams, the
values of ~Tc almost coincide with those obtained. experimentally. With a copper concentra-
tion of > 0.1%, the published experimental points [l, 2] in conjunction with the intermediate
values of the rate of cooling after annealing are disposed between the curves obtained for
the materials having been cooled after annealing either in water or in the furnace (one should
also bear in mind. the slight difference in-the irradiation temperature of the welded mate-
rials).
In comparing the effect of copper on the radiation embrittlement for the seam metal and
steel 15Kh2MFA [4, 5] at an irradiation temperature of 340=350?C, a definite similarity can
be noted in the behavior of these materials, shown in the linear dependence of ~Tc on the
copper content within limits of up to 0.3% and with a phosphorus concentration of >_ 0.01%.
When the irradiation temperature is reduced, the effect of the influence of copper increases.
Judging by the experimental data obtained for welded seams, an, increase of the cooling rate
also shows the same effect.
The effect of the nickel concentration on the radiation embrittlement of the metal of
the welded seams, with a phosphorus content of 0.010-0.012% and 0.17-0.19% of copper, sub-
jected to slow cooling after annealing, .is shown in Fig. 2a. The.data given relate to an
irradiation temperature of 300-310?C. It can be seen that with a change in the weld metal
of the nickel. concentration from .0.76 to 2.16%, the value of ~T~ is almost doubled ,(from 50
to.90?C). Comparison of the results obtained with the-data of [4-6] concerning the effect
of nickel on the radiation embrittlement of the base metal of 15Kh2MFA steel showed that for
the metal of the welded seams and for deformed steel, the nature of the radiation embrittle-
ment versus the nickel concentration is almost identical. The linear dependence with almost
d ~;C
40' ~ ~ ~
gs 9,0 7,5 N~,%
dT~?C
Fig. 2. Effect~of the content of nickel (a) and
manganese (b) on the increase in Tc of the metal
of welded seams of Cr-~in~Ii-~Io composition by
the action of neutron irradiation: a) composi-
tions 5-1 to 5-5, 6-1 (see Table 1); ~, 0) neutron
fluence 2.61019 cm z at 290-310?C; O) 2.5.1019
cm z at 270-300?C; ~) 3.8.1019 cm z at 50-80?C;
b) compositions 6-1 to 6-7, 5-4: O neutron flu-
ence 2.6.1019 cm ? at 300-310?C; O) 2.5.1019
cm z at 300-310?C; ?, ~) 3.8.1019 cm 2 at
50-80?C.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
the same angle of slope is maintained even in the case of a reduction of the irradiation tem-
perature to 50-80?C. Similar results were obtained when studying the effect of nickel in the
case of the metal of welded seams of Cr-~In--~1i--~Io composition [2].
Manganese, similarly to nickel, intensifies the tendency of the seam metal to radiation
embrittlement, which is demonstrated in Fig. 2b. In this figure, the set of points corre-
sponding to the experimental values of ~Tc for the material of almost one series, obtained
with an irradiation temperature of ti 310?C, shows that with increase in the manganese concen-
tration approximately from 0.5 to ti 1%, an increase in ~Tc is observed from 60 to 90?C. Irra-
diation at 50-80?C leads to stronger radiation embrittlement of the materials investigated,
but the nature of the dependence of ~Tc on the manganese concentration is maintained. Simi-
lar observations were obtained also for Cr-3Vi-Mo-V steel [5, 6]. As confirmation of the ne-
gative effect of manganese, the data on the embrittlement of welded seam metal of Cr-Mn.~Ii-Mo
composition [2] can be considered also.
From a comparison of the OTc curves of metal of the seams investigated versus the man-
ganese and nickel concentrations, it follows that manganese exerts a stronger influence on
embrittlement than the same amount of nickel, although it cannot be excluded that the result
is characteristic only for the metal of the melt being considered.
Discussion. Consideration of the experimental data obtained indicates that for the
metal of welded seams of studied composition, in the main the same regularities of the effect
of alloying and impurity elements on radiation embrittlement are qualitatively observed as
for the principal metal - chrome-molybdenum steel. Consequently, there are grounds for as-
suming the mechanisms of this effect also to be identical.
Relative to the role of phosphorus in the intensification of radiation embrittlement of
steel and alloys of iron, the opinion is expressed in [9] that atoms of this element, owing
to its capability of forming loose complexes with vacancies or mixed dumbbell-type pairs with
interstitial atoms, segregate on imperfections of the crystal structure of ferrite, in parti-
cular on accumulations of point defects or on the grain boundaries. In consequence of the
low surface energy of phosphorus and the deformation of the lattice in these zones.of segre-
gation, the interatomic bonds are weakened and embryo microcracks can originate most easily
here. If the irradiation temperature is sufficiently high for annealing the defects (above
300-350?C), the embrittlement caused by the presence of phosphorus in its principal appear-
ances is similar to thermal brittleness. The difference, in essence, reduces to the accelera-
tion of the diffusion redistribution of phosphorus under the action of superequilibrium va-
cancies. The similarities of these two phenomena concern also the effect of alloying ele-
ments. Nickel and manganese intensify the segregation processes, first and foremost in con-
sequence of the reduction of solubility of phosphorus in ferrite [10], but in radiation em-
brittlement their role is not limited to this. Judging by the fact that with increase in
concentration of these elements in steel and alloys, a marked increase in radiation hardening
is observed in the case of irradiation at a temperature in excess of 300?C, and the atoms of
nickel and manganese in some way participate in the formation of a defect structure. Definite
hypotheses have been expressed concerning the mechanism of their interaction with defects
[11], but reliable arguments still have not been found.
The effect of copper on embrittlement likewise is explained by the increase in radia-
tion hardening [5, 12]. The cause of this effect is associated in [12] with the hetero-
geneous origin of vacancy accumulations at complexes of vacancy-copper atoms, owing to which
the volume concentration of defects is increased, resolvable in the electron microscope.
Certain data, characterizing the change of physical properties of iron-copper alloys in con-
ditions of irradiation (11], coincide with this concept. According to this concept, the in-
crease in sensitivity of steel and the weld metal to the presence of copper, with increase in
the phosphorus content, can be traced as a consequence of the increase in the volume concen-
tration of intercrystallite defect zones, enriched with phosphorus.
It cannotbe excluded that welded seams, in preserving the characteristic of the structure
of cast metal, such as crystallization nonuniformity of distribution of the elements, have
certain specific properties of behavior by comparison with forged steel. Both phosphorus
and copper show a tendency to liquidate in the interdendritic layers of cast steel [13], in
connection with which it is possible to suppose that these elements, entering into competi-
tion, mutually change the volume concentration of individual structural components of the
weld metal.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
Fig.. 3. Dependence of Tc of the metal of welded seams of
Cr~1n-Ni-~io composition on the temperature of final annealing:
~ initial state; ~) after irradiation up to a neutron flu-
ence of 5.1019 cm Z.
-BOi i ~ ~
SOO 550 600 650
Temperature. ~
Fig. 4. Dependence of Qtr on the tempera-
ture of testing: 1) metal of a welded
seam of Cr-i~lm~li~io composition (wire
Sv-08KhGNMTA); 2) steel 15Kh2MFA; 3) steel
lOKhN1M; 4) steel A-302-B.
The marked dependence of the radiation embrittlement of the weld metal on the rate of '
cooling of the samples after annealing is interpreted most naturally as the result of a change
in the ferrite of the concentration of limitedly soluble elements, affecting the radiation
tendency to damage. Thus, it was established that .previous prolonged irradiation (500-1000 h)
maintained at 450-500?C, causing an increase in Tc for steel, is inclined to thermal brittle-
ness, and at the same time it reduces its sensitivity to embrittlement with subsequent irra-
diation [14]. This fact was described by the partial transfer of phosphorus atoms from the
solid solution at the boundaries of the former austenitic grains during aging, owing to which
the corresponding fraction of phosphorus atoms is excluded from the processes of interaction
with radiation defects. It is slightly probable, however, that in the weld metal containing
less than 0.020% of phosphorus, the grain-boundary segregation processes derived an appreciable
development during slow cooling of the samples (ti 15 h) after annealing. The experiment,
moreover, did not reveal any significant difference in the angle of slope of the curves of
'the dependence of embrittlement on the concentration of phosphorus in the case of a change of
the cooling rate, whereas in the case of a deficit of phosphorus in the solid solution, such
a difference should be. expected. '
The circumstance that for rapidly cooled welds the maximum angle of slope is observed
for the curves of the dependence of embrittlement on the content of copper gives the basis
for supposing a possible responsibility of this element for the observed effect. The limit
of solubility of copper in iron does not exceed 0.15% by mass at 450?C [15], and in alloyed
ferrite it can only be less. The separation of excess copper from the solid solution during
slow cooling of the welds from 500-350?C, in principle, is capable of weakening the tendency
of the metal to radiation embrittlement. This assumption, admissible in relation to welds
with a copper .content of > 0.1%, does not explain, however, the reason for the sensitivity
to the rate of cooling of the welds containing 0.04-0.07% copper (see Fig. la). Meanwhile,
the nonrandomness of this result coincides with the data of a special experiment carried out
on the metal of a welded sample, performed with commercial wire grade Sv-08KhGNMTA and heat.
treated in industrial conditions in the following regime: annealing 650?C; 16 h; cooling in
the furnace to 350?C, then in air. The contents of phosphorus and copper in the weld metal
amounted to 0.013 and 0.05%, respectively. This sample was divided into four parts, three
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
of which were subjected to additional annealing according to different conditions: 600?C,
10 h; 600?C, 10 h + 500?C, 10 h; 550?C, 10 h. The samples in all cases were cooled in water.
As a result of the additional heat treatment, the initial value of Tc (-40?C) varied within
limits of ? 20?C, and in proportion with the reduction of temperature of the final annealing,
an increase in Tc was observed. After irradiation up to a neutron fluence of 5.1019 cm z at
300-330?C, the values of Tc were found to be identical, independently of the conditions of
the previous annealing. The minimum shift of OTc (20?C) is appreciable for metal subjected
to stepwise annealing with a concluding temperature of 500?C, and the maximum shift of ~Tc
(60?C) -for metal after annealing at 600?C (Fig. 3). Intermediate values of OTc were ob-
tained for metal after regular heat treatment, i.e., after annealing at the highest tempera-
ture (650?C), but with cooling in the furnace.
Thus, even with a low content of copper, a change of cooling conditions after annealing
changes somewhat the sensitivity of the welds to radiation embrittlement.
A change of the cooling rate after annealing can also change the concentration in the
solid solution of implanted atoms (carbon, nitrogen). The effective influence of these ele-
ments on the increase in radiation hardening and embrittlement of alloys of iron, and also
low-alloy steels, is well known. Although until recently there is no single universally
accepted hypothesis for explaining the mechanism of the effect of implanted atoms on the
formation of defects in alloys based on iron [16, 17, 11], the role of these elements in ra-
diation embrittlement is quite important. It is understood, owing to the alloying of the
metal of welding wire with carbide- and nitride-forming elements, in the first place chromium,
that the concentration of carbon and nitrogen in the ferrite of the welds after high annealing
cannot be so high as in technical iron or carbon steel. Nevertheless, concerning the presence
of dissolved implanted elements in the weld metal, the nonmonotonicity can be confirmed of
the change with temperature of the transient tensile strength a'tr (Fig. 4). The extreme de-
pendence of the strength and plasticity of iron and steel in the range 200-400?C reflects the
phenomenon of dynamic deformation aging (blue brittleness), to the cause of which is attri-
buted blocking of the loose dislocations by atoms of implanted impurity, mainly atoms of ni-
trogen [18]. A comparison of the metal of the weld with the studied composition and certain
structural steels shows (Fig. 4) that steel of grade 15Kh2MFA, alloyed in addition to chro-
mium with such a strong carbide- and nitride-forming element as vanadium, manifests the mini-
mum tendency towards deformation aging, and for foreign steel A302-B (C~In-Mo composition),
according to the data of [19], this effect is expressed in the greatest degree. The metal of
welds made with wire of grade Sv-08KhGNMTA, occupies an intermediate position and is similar
to steel lOKhN1M, which has a similar system of alloying.
1. U. Potapovs and J. Hawthorne, Nucl. Appl., 6, 27 (1969).
2. J. Hawthorne, E. Fortner, and S. Grant; Welding J., 49, No. 10, 453 (1970).
3. V. V. Ardentov et al., Welding Materials for the Mechanized Welding of Reactor Pressure
Vessels of Nuclear Power Stations of Increased Power [in Russian], Automatic Welding, No.
6,
51
(1981).
V.
A.
Nikolaev and V.
I. Badanin, At. Energ., 37, No. 6, 491 (1974).
V.
A.
Nikolaev and V.
I. Badanin, in: Radiation Effects of the Change of Mechanical Pro-
perties of Structural Materials and Methods of Their Investigation [in Russian], Naukova
Dumka, Kiev (1977), p. 75.
6. V. I. Badanin and V. A. Nikolaev, Metalloved. Term. Obrab. Met., No. 9, 21 (1979).
7. V. I. Badanin et al., Inventor's Certificate No. 528161, Byull. Izobret., No. 34, 30
(1976).
8. L. Steele and J. Hawthorne, New Information on Neutron Embrittlement and Embrittlement
Relief of Reactor Pressure Vessel Steels, ASTM STP 380, 283 (1965).
9. V. A. Nikolaev, V. V. Rybin, and V. I. Badanin, in: Effect of Impurities on the Radia-
tion Embrittlement of Low-Alloyed Steel. Physics of Brittle Failure. Collection of Re-
ports of the Third All-Union Conference, Kiev [in Russian], Institute of Problems of
Material Behavior of the Academy of Sciences, Ukrainian SSR, Pt. 2 (1976), p. 89.
10. Phase Diagrams of Metal Systems, Issue 15 [in Russian], All-Union Institute of Scientific
and Technical Information, Moscow (1970), p. 226. -
11. V. A. Nikolaev, Problems of Nuclear Science and Technology. Series: Physics of Radia-
tion Damage and Radiation Material Behavior [in Russian], No. 2 (13) (1980), p. 47.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
12. F. Smidt and H. Watson, Metal. Trans., 3, 2065 (1972).
13. Yu. A. Nekhendzi, Cast Steel [in Russian], Metallurgizdat (1948).
14: V. I. Badanin and V. A. Nikolaev, At. Energ., 41, No. 3, 209 (1976).
15. A. E. Vol, Structure and Properties of Binary Metallurgical-Systems [in Russian], Vol.
2, Moscow (1962), p. 749.
16. E. Little and D. Harries, Metal. Sci. J., 27, 701 (1970).
17. G. Seidel, Phys. Status Solidi, 25, 175 (1968).
18. A. Cottrell, Philos. Mag., 44, 829 (1953).
19. L. Steele, Neutron Irradiation Embrittlement of Reactor Pressure Vessel Steels, IAEA,
Vienna (1965).
B. A. Kalin, I. I. Chernov, V. L. Yakushin,
V. I. Badanin, I. P. Kursevich,
V. A. Nikolaev, and V. N. Kulagin
Austeni.tic heat-resistant alloys and alloys of refractory metals, in .particular vanadium,
are considered to be prospective structural materials for thermonuclear reactors. It has
been established that these materials are subjected to considerable erosion in consequence
of helium blistering, and it is well known that the nature and degree of damage to the sur-
face by blistering depends on the alloying and purity of the starting material. Because of
this, a comparative investigation was carried out of the .erosion of a number of trinary
Fe~r~Ti alloys with fcc lattices and alloys of vanadium with bcc lattices, by bombardment
with helium ions with energies of up to 40 and 100 keV at a temperature of 290-970?K and an
ion dose of (8-50)?1021 m-2. The chemical composition and heat treatment of the alloys are
given in Tables 1 and 2. All the alloys are homogeneous solid solutions with dispersed se-
TABLE 1. Chemical Composition,-Treatment, and Certain Mechanical Properties (at
290 ?K) of Vanadium Alloys
Item
T
t-
Mass fraction of elements""%
No.
Material
rea
ment
Melt
~ I
Ti I
C I
H I
others
an
MPa
b' ?~0
H?, MPa
1
V~L-2
CD
I
Base
-
0,02
1.10-?
0,01A1
335
-
1550f120
2
VTsU
TO-1
I
Same
-
0,4
~
5?!0
0,03Fe
2,5Zr
410
20
10801100
3
V-7Nb
TO-2
I
? ?
-
0,017
.3.10-8
0,01Y
7,ONb
540
18
20401180
4.
V-14Nb
TO-2
I
? ?
-
0,017
3.10-8
14,0Nb
635
17
25901230
5
V-8Cr
TO-2
I
s ?
-
0,016
3.10-8
8Cr
515
18
1870f150
6
V-5Ti
TO-2
I
? ?
5,0
0,017
4.10-8
-
445
15
17801180
7
8
V-15Ti
V (technical)
TO-2
TO-2
I
L
? ?
15,0
0,017
4,5.10-8
-?
-
770
9
2020f200
? ?
-
0,035
2.10
-
-
-
18901180
9
V-10Ti
TO-2
L"
? ?
10
0,046
1,2.10-4
- .
-
-
1990f190
10
V-20Ti
TO-2
L
? ?
20
0,034
2,9.10-~
-
-
-
2020200
11
V-30Ti
TO-2
L.
? ?
30
0,033
1,7.10-4
-
-
-
2170220
!2
V-4UTi
TO-2
L
n ?
40
0,037
4.10-4
-
-
-
2320f230
*CD - cold deformation; TO-1 and TO-2 - annealing at 1470 and 1370?K during 1 h, re-
spectively; I, L -industrial and laboratory melts.
Translated from Atomnaya Energiya, Vol. 57, No. 3, pp. 1'73-178, September, 1984. Original
article submitted December 8, 1983.
0038-531X/84/5703- 0613$08.50 ? 1985 Plenum Publishing Corporation 613
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050003-4
TABLE 2. Chemical Composition, Treat-
ment, and Microhardness (at 290?K) of
Fe-C~Ni Alloys
TABLE 3. Coefficients of Erosion of Vana-
dium Alloys, Irradiated with He+ Ions,
10-2 atom/ion
E=4okeV, n=z
?iozz m_
`;
..
~
x
o
~~
xo
~
x
x
x
x
?.~
~
~ ~~
~
o
0
o
a
o
i
ro
.n