JPRS ID: 9115 EAST EUROPE REPORT SCIENTIFIC AFFAIRS

• Document Type: CREST
• Collection: JPRS
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82-00850R000200090001-9
• Release Decision: RIF
• Original Classification: U
• Document Page Count: 21
• Document Creation Date: November 1, 2016
• Sequence Number: 1
• Content Type: REPORTS

APPROVE~ FOR RELEASE: 2007/02/08: CIA-R~P82-00850R00020009000'1 -9
i ~ J ~
~ ,~~l~E ~~~0 f F~l~~ ~

APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFF[CIAL USE ONLY
JPRS L/9115
2 June 1980
- .
- E a st E u ro e R e o rt -
p p
- SCIENTIFIC AFFAIRS -
CFOUO 5/80)
,
FB~$ FOREIGN BROADCAST INFORMATION SERVICE
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
NOTE _
JPRS publications contain information primarily from foreign
newspapers, periodicals and books, but also from news agency
transmissions and broadcasts. Materials from foreign-language
sources are translated; those from English-language sources
are transcribed or reprinted, with the original phrasing and
- other characteristics rPtained.
- Headlines, editorial reports, and material enclosed in brackets
are supplied by JPRS. Processing indicators such as [Text]
or [Excerpt] in the first line of each item, or following the
last line of a brief, indicate how the original information was
processed. Where no processing indicator is given, the infor-
~ mation was summarized or extracted.
Unfamiliar names rendered phonetically or transliterated are
enclosed in parentheses. Words or names preceded by a gues-
tion mark and enclosed in parentheses were not clear in the
original but have been supplied as appropriate in context.
Other unattributed parenthetical notes within the body of an
item originate with the source. Times within items are as
given by source.
Tl,.e contents of this publication in no way represent the poli-
cies, views or attitudes of the U.S. ~overnment.
For further information on report content
call (7031 351-3060. _
COPYRIGHT LAW~ AND REGULATIONS GOVERNING OGI~TERSHIP OF
MATERI.P.LS REPRODUCED HEREIN REQUIRE THAT DISSEi~IINATION -
OF THIS PUBLICATION BE RESTRICTED FOR OFFICIAI, USE ONI,Y.
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY -
. JPRS L/9115
2 June 1.9 80 -
EAST EUROPE REPORT
$CIENTIFIC AFFAIRS
' (FOUO 5/so)
= CONTENTS
CZECHOSLOVAKIA
Use of Decision Tables for Protection of Data in Computers _
- tt3:nnd Date, Banka
(Cestmir Sra~hans; AUTOMATIZACE, Feb 80) .1
_ High-Quality Steel From 0].a Jet Engines
(Bohdan Irmler; HUTNiK, Feb 80) 10
- a - [IZI - EE ~ 65 FOUO]
- FOR 0~'FICIAL USE ONLY
b
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY
CZECHOSLOVAKIA
UDC 381.32:517.1
USE OF DECISION TABLES FOR PROTECTION OF DATA IN CONIPUTERS AND DATA BANKS '
Prague AUTOMATIZACE in Czech No 2, Feb SO pp 40-43
[Article by Engr Cestmir Srajhans, CSc, College of the National Security
' Corps, Prague]
- [Text] The article discusses the problem of protecting
_ data in computers and data banks. The decision algorithm _
in automatic protection systems is fairly complex. The =
~ decision table is a suitable tool for ths solution of
this demanding problem. If the protection system is suit-
ably designed, then the computer automatically puts into
effect measures co~nensurate with the degree of threat to
the stored or processed data. The article presents one
of the feasible variants for designing data protection.
Topic tags: data protection, decision tables.
Introduction
Advances in computer technology and the constantly growing volume of data
- stored and processed in computers and data banks create various problems.
One such problem is effective and preferably automatic protection of the
data from destruction, tampering, theft or misuse by an unauthorized person.
- If we disregard natural hazards (fires, earthquakes, floods, etc.), then
_ the main threat to data security is man--the penetrator. By this concept
we mean the organization or individual who intentionally or accidentally
enters the system and by his action causes the data to be destroyed, dis-
torted or stolen, to the user's detriment.
Indentifier's Role
The pretection of data from penetration thus consists essentially of iden-
tifying the persons who have authorized access to the data, and of accurate-
ly defining the rights and privileges of these authorized persons. The vari- :
ous types of protection must therefore permit the differentiation of the in- -
dividual users as authorized or unauthorized. An authorized user may Y.ave -
the following rights:
1 -
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIr1L USE ONLY
To alter the recorded data; '
To read the data;
To exercise his privileges that may be full or li.mited;
To enjoy a certain priority in relation to other users.
If a user wants to access the system (computer or data bank), he first must
prove that he is an authorized person, and that he performs activities and
requires services only for which he has authorization. As soon as he ex-
ceeds his authority, he becomes an unauthorized person. The protection sys-
tem then must immediately react, adopt the necessary measures and uncover
the unauthorized activity. The agreed-upon and unambiguously defined method
that is able to identify the user and to define his privileges is called the
identifier. If the protection system is to be effective, it is not enough -
to identify merely the user. It is also necessary to unambiguously identify
further elements of the system. Such elements are particularly: _
Hardware elements: terninals,
processors,
memories,
teleco~nunications channels;
Software elements: programs,
sets,
elements of the operating system;
- Privileges: priorities,
authorization of activity in the system,
secrecy classification.
Decision Tables
From the preceding it is evedient that it is by no means easy to decide
whether a user is an authorized person, and wnether he conducts only activ-
ities for which he is authorized. A decision matrix that could describe
this situation would have to be multidimensional and very complex. There-
fore it seems expedient to describe the decision algorithm with the help of -
decision tables. Decision tables are a suitable and by now well-developed
tool, understable to both the computer and to man. Depending on the number
of conditions, decision tables may contain scores and even hundreds of pos-
sible rules. The total number of all rules p is given by the equation:
P = 2q
where q is the number of conditions. In the case of three conditions, then,
we get eight different rules. If there are five conditions, the number of
dif~erent rules is 32; and in the case of 10 conditions, their number is
102a~. In such complex decision tables, many different rules lead to the -
same decisions. We then say that the decision table is redundant. It then
- becomes necessary to reduce the decision table and to exclude the redundant
2
- FOR OFFICI~,L USE ONLY _
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY
rules. When~complex decision tables are being reduced, it may easily happen
that some of the basic rules are omitted. If a decision table does not con-
tain all the rules that lead to different decisions, we say that the deci-
sion table is incomplete. Combinations of the conditions from which the
rules are formed may lead in some instances to contradictory situations in
_ which it is not possible to make any decision. If the decision table con-
tains such rules, we say that the table is contradictory (conflicting).
If the computer is to make correct decisions automatically on the basis of
the adopted decision table, this table must be complete and n~nredundant,
and it must not contain contradictory rules. When constructing a decision
table, therefore, it is essential to always check the table, correcting it _
when necessary.
During the logical analysis of the individual rules we often find ti~at some
of ttie conditions do not particip~te in the resultant decision; whether or
not the conditions is met does not influence the decision. In such cases
it is expedient to introduce in the quadrant of rules the so-called incon- -
sistent value N. It indicates that the condition does not participate in
the decision; thus for N we may substitute a value of 1 as well as a value
of 0. In certain types of decision tables it is convenient to introduce the
ELSE; rule. These are tables that lead only to two different decisions and
employ the logic function AND or OR. In decision tables of tnis tvpe only
one rule results in a certain d~acision, and all the other rules result in
the opposite decision. Such a table is thus highly redundant. We remove -
the redundancy by retaining only the one rule and combining all the other
rules into a single rule, the so-called ELSE rule. A decision table that
contains an ELSE rule is complete and r~onredundant.
Design of an Automatic Protection System
Figure i is a flowchart in which the decision algorithm of one of the pos-
sible systems for protecting data in computers or data banks is written
with the help of decision tables. It is of course possible to write a sin- _
gle decision table for this purpose, but such a table would be too complex.
,
The~refore three decision tables have been used: RT-1 (permission), RT-2
(execution), and RT-3 (verification). Let us now describe the mentioned
flowchart, check the individual decision tables and analyze their logic.
Wh~n accessing the system the user must identify himself with valid identi-
fiers. The first decision must be made, illustrated in the flowchart by the
diamond "Access procedures fulfilled." If the user does not know the valid
identifiers or makes a mistake during identification, he is rejected and
warned that he did not proceed properly. The concept "access procedures"
may include:
Tdentification of the user,
Identification of the terminal,
Identification of the processor, memory or channel.
- 3
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY
~'O O ~ O ~c
' ~4' O O K _
[i� ~ O ~c
_ ,p n ~ 'C .
v
~ - p
~N ~ ~ O ~ j ~
, 'v .F
~ ~ ~ Z e~ K i.a .
~4 ~
- O
~ ~3 ~
~ O ,N ~S O
~ v -~i` C ~p t~
a ~ C > > a O N U
,j C b c~i ~ N Q v U1 -
V~ ,l p ~ 1 ~ a y
~ v ~ ~ ~ ti t~, N
L ~Q ~ t ~v ~ ~ L U
Q U
l ~u O '1 C ~G O ~ C C ~ ^ ~
c�, ~ ~C ~ '4i ~ai�J ~ ~j'~ ~O ~ N 1~1 ~
~ ,b L
, Q~ ~ C ~ y
j E o ~l L~< 'C j W
'~n Q 44 V 4 u ~
� -
~ 'a' ~ N Ifl
.M.. Ol O r-I N N N ~-i r-~ N �~i
~ ~N.. N v ~I
r e v ~
~
. ro
L(1 ~ � ^
~ Q '
,G v y~' O ~ ~C ~m b~
a, o ?
C y o ~ ~N 2 p k a
IV i j w n k v~ e~ a~
~ � ~ .
~
~ ~ G
t J C
~ ~
~ ~ ~ ~N O N
o l ~ N ~.1 .
o ~ ~ 4 ~d Q)
o , ~ ~x
~ i 'v 'm -
v ~ y C ~C~ ~ ~ ~ ~ 3 3~1
7 Q ~ �p O iV I ~ ~ O ~
-'o~. Q 1~ ~ o y� ~ ~ w u
. o y ~ ~ y ~ ~ l~ C ,
� ~ Q e C C . 4~e ~ C ~ s~ �
~ C ~ i C ,O r-1
oi ~
~ � � 4 n a ~ ~ ~ . ~
4 4~ 4 Q, , ~ o o O 4 Q ~
.
W
~ '-,I rN-1 r-1 t7' tn tD f~ M Op
~ v v v ~ r1 r-I
~ v v ~ .
_ ~
FOR OFFICIAL USE ONLY �
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
i
FOR OF'FICIAL USE ONLY
,
Key to Figure 1: ~
1. Start 16. Authorized to read or alter data
2. Access procedures fulfilled 17. Switch to RT-1 -
3. No 18. Execute instruction
4. Warning 19. RT-3 (verification)
5. Yes 20. Additional identifiers do not
- 6. Wait agree
7. RT-1 (permission) 21. Memory blocked
8. ELSE 22. Sys~~em not functioning propErly
9. Authorized access to program 23. Simulate operation, issus alarm
10. Authorized access to set 24. Interrupt, institu�"te checks -
11. Priority user 25. Eliminate fault
12. Switch to RT-2 26. Simulation
13. Switch to RT-3 27. Alarm
?4. RT-2 (execution) 28. Stop
_ 15. Restriction on set or program
- If the user fulfills +the accessing procedures, he is allowed access to the
system and is checked further. Pennission table RT-1 investigates whether
- he hz.s authorized access to the desired program. The user nnist thus know
the identifier of this program. The identifier is valid for one access only
= and is changed after each access. It is actually a relative address record-
ed in the system's central memory as a two-digit number that is automatically
printed out for the user if he has been permitted access to the program. The
program the.~ selects the appropriate set. The user is aqain checiced whether
he has authorized access to the set. The se~'s identifier c~~ntains the fol-
lowing: the set's name, the descriptor of the form of recorciing, the user's
identifie.r, and the identifier of the medium. If the user passes also this
_ check, the ~:ermission table inv~stigates further whether he has any prior-
ity, whether he may claim preferential service. If the system has only one
_ user, this test may be omitted. The priority identifier is related to the
user's function and is expressed numerically (the lower the number, the
higher the function and the priority). If all condi~ions in Table RT-1 have
been met, the user may be served. In this case he is transferred to execu-
tion table RT-2. If any of the conditions in RT-1 is not met, the user be-
comes suspect. He is then transferred to verification table RT-3, which in- ~
r�estigates whether there is an att~mpt at penetration.
Table RT-1 contains the ELSF. rul~; hEnce it is nonr~dundant and complete,
- and therefore it does not have to be analyzed further.
Lei: us assume that the user has successfully passed all the tests, has -
reached table RT-2 and demands service. Before a decision to execute the
instruction, it is necessary to determine whether the user is within his
privileges for the system. Therefore a check is made whether. he has any re- -
strictions regarding the set and the program, and whether he is requesting
execution of an instruction for which he is authorized. His rights are ex-
pressed by the access iclentifiers. RT-2 thus checks whether these rights
5
FOR OFFICIAL USE ONLY �
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-00850R040240090001-9
FOR OFFICIAL USE ONLY
have been exceeded. If the user has not exceeded his rights, the desired
instruction is executed and the user is served.
F'secution table RT-2 contains two conditions. Hence from formula p= 2q
it follow~ that the table must contain four rules. These are:
P1 P2 P3 P4
1 1 0 0
1 0 1 0
Let us subject these rules to logical analysis. According to rule P1, the
user is authori2ed to read or alter the data, but this authorization is not
unlimited. He is restricted i.n relation to either the program or the set.
~ If the requested instruction is in conflict with his rights, the user is -
returned to RT-1 and is checked again. If it is established that the user
- accessed the system but does not have any rights (see rule P2 and P4), then
he is under suspicion of having stol~n some of the identifiers, or that he
removed or circumvented protection. He is then regarded as a p~netrator
and is transferred to verificatio:~ table RT-3. We see that in this case
the first rule has no significant validity. Therefore we may introduce the
inconsistent value N and combine rules P2 and P4. Then solely rule P3
leads to the execution of the instruction. According to this rule, the user
is under no r~striction and is authorized ~o request the instruction. With ,
the execution of the instruction, the algorithm ends.
Table 22T-2 is likewise nonredundant and complete. I~ contains tr.ree rules,
each of which results in a different decision. Rule P2, which contains the
inconsistent value N~, actually represents two rules: P~ and P4. `
Verification table RT-3 contains three conditions. Theoretically it thus
- shenld have eight rules. In our case, however, inconsistent values N are
introduced into two rules, whereby the number of rules is reduced oy two.
Let us subject these rules to logical analysis. Rules P1 and P3 result in
~he same decision. A basic condition for both these rules is a third condi-
- tiona If this condition is fulfilled, then the system has lost control over `
- itself and security of the data is compromised. If t_o this condition there
is added a further fault, then it becomes necessary to issue an alarm (acous- `
tical or optical), to disconnect all users, to interrupt operation and insti-
tute checks. Rule PS likewise leads to a similar decision. But here the
system's improper functioning is not linked to any furicher fault. In this
case there is evidently a tecr.nical breakdown that must be correct~d without
delay.
Interesting is the decision pursuant to rule P2. In this case the protec- -
tion system iinds, with a high d~gree of prob~bility, that there is an at-
, tempt to penetrate the system. The user either did not know tY,e additional
identifiers or acted so clumsily that the protection system classified him
as a penetrator. If communication with him w~re broken off on the basis of
6
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
~
FOR OFFICIAL USE ONLY
- this finding, an actual penetrator would imanediately recognize that he has
been discovered, would strive to conceal his clues and to escape. Therefore
~ in this case a secret alarm is issued. Normal operation is simulated, but -
the penetrator receives false data. Thereby time is gained to catch the
p~ane�,trator in t:he act.
Rule P4 solves the case when one aser must allow another user with higher
priority to precede him. According to rule P6, the result of the check on
- the user is positive, and the protection system has no objection to provid-
- ing service for him.
We still have to check whether RT-3 is nonredundant and complete. In this
case, however, the situation is not so clear as in the case of decision ta-
bles RT-1 and RT-2.
Checking the Decisi~n Table -
In the case of complex decision tables we check their nonredundancy and com-
pleteness by successively reducing them into elementary decision tables whose
completeness and nonredundancy we are able tc recognize directly. During
the redt~ction into elementary de~ision tables is suificient to investi-
gate only the quadrant of c~nditions and the quadrant of rules. We distin-
guish in all six types of elementaxy decision tables with the following
properties: ~
> P Decision table incomplete.
Qr C
2 P -
= Q~ N Decision table nonredundant and com-
~ plete.
3. Pr Pi
Decision table incomplete and redundant.
Q, C C ,
~
4. Pr Pi
Q, C C Decision table nonredundant and com-
~ plete. `
_ 5 pi Pz Decision table redunandant, possibility
Q~ ~ ~ of contradiction not excluded.
- 7
- FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE OhZY
_ 6 P' P1
N N Decision table redundar~t, possibility
Q~ of contradiction not excluded. "
~ The symbols in these tables are as follows: ~
Q1~ Q2~ Q3 are the conditions,
, pl, P2, are the rules.
C,., is a consistent value of 1 or 0,.
_ C... is a consistent value of the opposite type, 0 or 1,
N... is an inconsistent value of 1 or 0.
The checked decision table is nonredundant and complete if all the Elemen-
_ tary decision tables it contains are either of t~pe 2 or type 4. If any
other type of elementary decision table occurs, then the checked decision
table is either redundant, incomplete or contains rules that lead to con-
tradicticn.
- Let us now reduce our decision table RT-3. We prepare the reduction An the
basis of a selected key condition. We strive to choose a condition that does
not contain the value N. In our casp we select the first condition, Q1. On
~he basis of this key candition we reduce the decision table into two partial -
_ decision tables, A and B. Table A will comprise the 1 values of key condi-
tion Q1, and table B will comprise the 0 values of key condition Q1. Thus
A B � ' _
_ p~ p1 p~ p pf p6
Q~ N N Oi 1 f 0 0
Q~ 1 0 0~ 1 0 1 0
Partial tables A and B are not elementar~ tables. Therefore we continue our
reduction. For the key condition in both cases we select Q3. In the same
- manner we obtain _
aa ae ea ea
p~ p p~ ps p` p6
- ~ 0~ N C= N 0= 1 0 Q2 1 0 -
~
All tliese par�tial tables are elementary tables and cannot be reduced further.
Elemen~ary decision tables AA and AB are of type 2, and elementary decision
tables BA and BB are of type 4. Table RT-3 does not contain elementary de- ~
cision tables of any other type. Hence it is nonredundant and complete.
_ The advantage of the described algorithm for checking decision tables is its ~
simplicity. Reduction and checking can be run on a computer. In the case
- of complex decision tables containing several hundred rules, the ~dvantage
' of this method of checking is obvious.
8 -
FOR OFFICIAL USE ONLY -
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY
Conclusion
It is the duty of every top official to pratec� data and information in gen-
eral. Moderr.i computers of the later generations, the extensive dat~ banks,
and particularly the automatic control systems demand that the problem of
data protec~ion be solved seriously and with full respansibility at every
level of management. This does not apply merely to systems ~hat process
data which are state or official secrets. Al1 data, even nonsecret data,
are of considerable value to the user. For this reason ~t is necessary to _
guard such data with effective and reliably functioning ~rotec~ion systems.
Since the storage, processing and retrieval of data in .~~3ern systems is
highly automa`ted, it is essential that also the protection system function
- automatically. Hence suitable methods and tools must be sought for this _
purpose. The article points out that decision tables aXe very suitable
tools for this purpose.
BIBLIOGRAPHY
1. Gerasimov, V., and Vladislavskiy, V.: "Methods of Protecting Iriforma-
tion in Automatic Control Systems," a series of articles in ZARUBEZHNAYA
RADIOTEKHNIKA, Nos 2, 6, 8 and 10, 1975.
_ 2. Hoffman, L.: "Security and Privacy in Comguter Systems," Melaille
~ Publishing Co, Los Angeles, California, 1973.
- 3. Chvalovsky, V.: "Rozhodovaci Tabulky" (Decision Tables), SNTL [State -
Publishing House for Technical Literature], Prague, 1974.
4. Work paper for the Seminar on Decision Tables, Research Institute for
Control and Automation, Prague, 1970.
- 5. Srajhans, C.: "Protection of Inf~rmation in Automatic Control Systems," _
science-candidate dissertation, College of Special Sciences and Re-
search, Prague, 1979. ~
COPYRIGHT: ShTL, Nakladatelstvi Technicke Literatury, Prague, 1980.
1014
CSO: 2402
= 9 -
- FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED F~R RELEASE: 2007102/08: CIA-RDP82-00850R000200090001-9
~
- FOR OFFICIAL USE ONLY
CZECHOSLOVAKIA
_ HIGH-QUALITY STEEL FROM OLD JET ENGINES
Prague HUTNIK in Czech No 2, Feb 80 pp 57-60
[Article by Eng Bohdan Irmler, Trinec Iron Works VRSR [Trinec Iron Works
of the Great October Socialist Revolution~]
[Text] Discarded jet engines of aircraft made of highest quality construc-- -
tion materials containing Ni,Cr,Al constitute a very valuable raw material
for. metallurgical plants. But disassembly of discarded ~et engines into
individual elements is very time consuming and Kovosrot collection centers
perform it only to a limitPd degree. -
_ Originall~ the TZVRSR used to purchase discarded jet engines together with
- assorted highly allayed scrap for remelting in an open-hearth furnace to
recover steel of known chemical composition. Following the modification
of the remelting technology in an open-hearth furnace to remove phosphorus,
the chromium contained in the aircraft eng~ne was burned away which led to
a 4earch for a new method of procesa~ng aircraft engines. Following re-
melting in an electric arc f�urnace (the engine was added to the basic melt)
" the resul.cing average chemi~~al composition in percentages was as follows:
= C=0.61; Mn=1.04; Si=5.4~; P=0.049; Cu=1.16; Cr=7.7; Ni=10.15; Mo=0.59;
W=0.73; Ti=0.26; A1=11.0.
The high carbon content stems from the rubber components of the aircraft =
- engine. The true Si content will be considerably lower because following ~
the add3tion of the aircraft engine t.he furnace slag was completely de-
prived of oxygen and an unknown quantity of Si was extracted by reduction ~
from the ladle lining. In reality, the phosphorus content is lower; chem- _
:ical analyses of individual parts of the a','.rcraft engine disclosed that
the steel and the alloys used in the construction contained less than
0.03~ percent P. The average sulfur content was not determined because
- of complete desulfurization of the steel bath following the addition of
the aircraft engine (0.003 percent S). Also the aluminum eliminated with
the slag was considered.
_ Cheraical analysis reveals that it is most advantageous to process dis-
carded aircraft engines by oxidation remelting in the production of CrNi
10
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
- FOR OFFICIAL USE ONLY
_ (Mo) class 17 steel. The direct remelting technology was developed at
- TZVRSR between 1976 and 1977. Aircraft engines are remelted in an alkaline
12-ton electric-arc furnace equipped with electrohydraulic control. The
transformer input is 2.5 MVA.
The firat melts were characterized by excessive meltoff of the furnace
bottom and walls as a result of the high temperature of the steel bath -
following oxidation. The bath temperature could not be determined because
TZVRSR did not possess a device for measuring bath temperatures around
2,100�C. Fundamental improvement was not ach~tved until the following
technological changes were made:
a. Nickel plate was replaced with nickel oxide in the melt.
b. The steel-bath temperature before ovidation was increased from 1,580- -
1,600�C to 1,630-1,670�C. ~
c. Oxidation with oxygen was terminated when the carbon content in the
bath dropped to 0.12-0.15 percent.
The first two technological measures were designed to postpone the oxida-
tion of A1 and Si dissolved in the steel bath partly into the last melting
stage, mainly into the post-tieating stage. This resulted in a sharp in-
crease in the temperature of the steel bath at about 1,600�C at which
point the temperature of the melt rose by SO�C within 10 minutes. -
The third technological measure was based on the experience that, in dif- _
ference to melts conducted by the standard technology where intensive carbon
- combustion terminates at a 0.05 to Ov08 percent carbon content in the bath,
intensive carbon burnoff is blocked at a content of 0.12 to 0.15 percent
carbon in steel. Prolonged oxidation of inelts containing aircraft engines
- results in increased melting loss of Cr, excessive temperature rise of the
steel bath and slag and the simultaneous undesirable meltoff of the furnace -
bottom, walls and lid.
Tests designed to postpone oxidation till the final melting phase in order
to make use of the chemical heat of A1 and Si dissolved in the bath were
- unsuccessful. They resulted in high Cr losses (about 3 percent) and high
meltoff of the furnace hearth lining.
~ In view of the high Ni0 and low alloy and low phosphorus scrap content in
the burden which yields melts with a P content lower than 0.035 percent and
after oxidation a carbon content of 0.10 to 0.16 percent, aircraft engines -
are now being remelted exclusively in the production of steel according to
- CSN (Czechoslovak Norm) 17,251 of chemical composition of at most 0.2Q per-
cent C, 1.5 Mn, 2.00 Si, 0.035 P, 0.035 S, 18 to 21 Cr and 8 to 11 percent
- Ni.
Charging and Melting Technology -
Charging is accomplished in two batches which consist of (in kg):
11 -
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
_ ' FOR OFFICTAL USE ONLY
First batch: low P and Ni scrap ~,000
heavy high alloy Cr,Ni scrap 3,000
carbon Fe-Cr 500
_ Second batch: discarded aircraft engine 800
nickel oxide 500
high-alloy Cr-Ni chips 3,800
Total 11,600
Table I. Change in the Content of Elements During Oxidation
T(vb) IZo~b)eg o> (d~ Obenh .
a C~ I C~ I C~~' Cr~ I d(;r I ~[n, I Alrt~ I d\in I F.li I 91~ I de1 I Tli ( AI~'� -
- s leG.~c. priimcr 47 0,?A'? (1,130 13,79 13,28 0,508 0,793 0,502 0.289 0,8RG U,209 0,0,7 0,14�l 1,204
moto� mici. 0,10 0,()0 11,84 11,1? -0,(30 0,69 0,33 0,1�2 U,3�l 0,08 0,04 0,04 0,6:i
rom rozputi
(e) g) max. 0,72 0,19 1b,44 15,47 3,19 2,05 1,03 1,02 1,22 0,47 0,03 0,23 1.0~
bez (f) prum~Sr 33 0,348 0,0~~ (f,,4U* 14,61 1,79 0,691 0,370 0.3~1 Q,3(38 0,0~31 0,277 O,Q4'? -
letcc. min. 0,18 O,Of, 13,34 11,95 0,82 0,45 0,~7 0,20 0,21 0,05 0,11 ~,01 -
motorii rozputi
max. 0,62 0,14 19,85 17,69 3,05 1,32 0,47 0,70 0,08 0,19 0,67 0,14 -
Key:
- a. Melt e. With aircraft engine
b. Value f. Without aircraft engine
c. Number of Melts g. Average, minimum, range,
d. Content (percent) maximum
Comments:
Index 1--content of element in the bath following melting
Index 2--content of element in the bath following oxidation
Delta--difference in the contents of elements in the bath before and after
oxidation
~ *--the high average meltdown chromium content is due to the fact that 22
compared melts out of 33 were o~iginally conducted as reduction remelts, _
but because the required meltdown content of carbon was exceeded, melt- ;
ing had to be conducted by oxidative remelting
**--in reality the aluminum contents will be lower; due to lack of chemical _
homogeneity the upper layers of the bath are richer in aluminum. Alu- ~
- minum determination is not quantitative and is rather difficult and
expensive in this type of steel. For these reasons, aluminum content -
_ was determined only in the case of 14 melts.
12
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY
After the first batch has melted by 60 to 70 percent, the aircraft engine -
is added and heaped with nickel oxide and alloy chips. The first batch -
must not be completely melted. Following immersion of the aircraft engine =
into a cumpletely melted batFi the aluminum alloys of the aircraft engine
flare up and burn intensely. The meltdown slag is not collected; by chem- _
ical composition and appearance it exhibits reducing properties. It is
- oxidized by oxygen when it reaches the steel bath temperature of 1,630 to .
1,670�C with the oxidation completed at a 0.12 to 0.15 percent carbon con-
tent of the bath. Chromium is not recovered from the slag by reduction
(median Cr203 content in 17 oxidation slag samples is 11.56 percent), the
oxidation slag is skimmed off and oxygen removal precipitation is accom-
plished by 3 kg/t aluminum forged to a pole. New slag forms consisting
of lime and fluor spar. The bath is left to cool and is then alloyed by
low carbon ferrochrome, nickel plate or recycling scrap. Following dis-
solution of the added alloys and homogenization of the bath, the slag is ~
skimmed off and alloying of the melt to the desired composition is com-
pleted on the basis of chemical analysis.
This technology is characterized by decarbonization of an actual C-Mn-Si-
Cr-Ni-A1-Ti melt. The change in the content of individual elements due to
oxidation is indicated in Table I. The same table contains, for comparison,
values found in comparison melts without the addition of aircraft engines.
The comparatively high A1 content in the bath before oxidation influences -
the increase of the meltdown content of Mn, Si and Ti. Heat generation due
to A1, Si and Ti combustion at the start of oxidation raises the tempera-
ture of the steel bath rapidly to 1,850�C and higher. At these tempera-
tures, carbon burns off before chromium which is supported conclusively by -
the melting loss of chromium by 0.508 percent.
The Relationships Established
The basic statistical characteristics were calculated by correlation and
regression analysis on an IBM 370/148 computer programmed for mathematical-
statistical analysis and relationships between selected technological
parameters and chromium loss were established. The most important factors
affecting chromium loss were found to be the carbon content in the steel -
bath following oxidation, the chromium content in the bath following melt-
down., the te~?perature of the melt before oxidation and the melting loss of
Si during oxidation.
The equation of the regression hypersurface from which the Cr melting loss
can be theoretically derived has for a given number of 47 melts (the crit-
ical correlation coefficient value is 0.287) the following form:
Cr = 4.973 - 12.68 C2 + 0.25 Crl - 0.00404 T+ 0.443 0 Si (1)
wher e
- ~ Cr...chromium melting loss during oxidation (percent)
13
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY
C2.....carbon content in the bath following oxidation (percent)
Crl.,..chromium content in the bath following melting (percent)
_ T......bath temperature before oxidation (�C)
~ Si...silicon melting loss during oxidation (percent)
The calculated coefficient of multiple correlation for the given equation -
is 0.737; therefore, it can be said that the chromium melting loss c~n be
explained by the given equation by almost 55 percent as indicated by the
- graph in Figure 1. Very surprising is the finding that during oxidation
the chromium melting loss increases with increasing silicon melting loss.
Other important findings established include in addition the relationship
between the carbon content and the silicon content in the bath following
oxidation. The regression equation found has the following form:
C2 = 0.106 + 0.118 Si2;r~, = 0.442 (2)
~ Cr~
1,00 . . . � ~
_ 4 ~ � t~. ~ ' -,y{ry:
a'� r~ ?..T: . . .�r;�:,:
0,80
. dsi
0,60
~ 0 40 � rXy = 0, ~3~
~ d = 54 3
~ N = 47 ~aveb
� 0, 20 T .
melts
~ D o0 '
~ .
-q 20 Cs
0,09 0,11 0,93 0,15 0,1~ CI J _
11,0 13,0 15,0 Crf C'.)
150D . 9600 1~00 T(�C J
0,30 0,50 0,'~D 0,90 a Si
Figure 1. Graphic Representation of Chromium Melting Loss According to
Equation (1)
1~+
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY
where
C2....carbon content in the bath following oxidation (percent)
Si2...silicon content in the bath following oxidation (~,~rcent)
The graphic representation is in Figure 2.
g 1s -
_ ^
~ q~4
_ ~N
_ ~ ~7Z ~ � _ . .
/
~ ~
Oj10 . .
~ g~ R3 gs 4~ -
siZ !/.1~ � .
Figure 2. Graphic Representation of the Dependence of the Carbon Content
in the Bath Following Meltdown According to Equation (2)
The effect of the aluminum content in the bath before oxidation could not
be determined for lack of reliable data, even though it is of overriding -
importance in the technology under study.
The chemical composition of the slag after the melt and oxidation differs
considerably from that of inelts conducted according to standard technology _
as is evident from Table II. -
The comparison of analyses of ineltdown slags resulting from individual
technologies reveals that the content of oxides of elements with lower
affinity to oxygen than a?uminum is several times lower in melts contain-
ing aircraft engines and is replaced by increased content of A1203 which
averages 52.3 percent. The average content of Cr203 is 2.94 percent com-
pared to 12.1 percent in comparative melts. Oxidation slags in melts with
aircraft engines have an Fe0,Mn0 and Cr203 content two to three times
lower than the oxidation slags of the comparison melts and are again re-
placed by A1203 and partly by MgO. In both technologies, the Ca0 contents
are on approximately the same level both after meltdown and oxidation.
Since the addition of lime to the charge is the same in both technologies,
- the conclusion can be reached that the weight of the meltdown and oxida- -
tion slags is the same in both technologies.
15
FOR OFFICIAL USE ONLY
s ~
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY
A marked difference exists in the alkalinity of the slags which can be
calculated by the simple~equation
Ca0
- V =
Si02
and by the expanded equation
' Ca0 + Mg0
V =
Si02 + A1203
In view of the high A1203 content in slags from melts containing aircraft
_ engines, the more correct calculation of alkalinity is derived from the
expanded equation. Median alkalinity of slags calculated in this way is
0.551 in melts containing air~raft engines and 0.537 following oxidation
which is 2.32 times or 2.33 times lower than in melts not containing air-
craft engines. _
The Economic Gain
The introduction of direct remelting of aircraft engines in the production
- of steel in accordance with CSN 17,251 has resulted in considerable saving
in ferroalloys, power and in increased output. Compared with the original
technology the following savings were achieved:
Reduction in the consumption of FeSiCr 4.41 kg/y
FeSi 10.Y9 kg/t
MnSi 0.56 kg/t
Mn aff 1.06 kg/t
FeCr IV 17.27 kg/t -
FeCr II 25.78 kg/t
power 110.7 kWh/t
- Increased EOP output 0.220 t/h
_ Conclusion
The electric steel department of the TZVRSR has developed a technology of
direct processing of discarded ~et aircraft engines by oxidative remelting -
of high-alloy scrap in st~el production according to CSN 17,251 resulting
in considerable savings of alloying additives, power conswnption and in
increased output. The decarbonization of the C-Mn-Si-Cr-Ni-A1-Ti melt
with an average content of 1.204 percent A1 was tested and introduced in
practice. The technology of producing steel in an alkaline fu.rnace using
highly acidific slag of inedian post-oxidation alkalinity of 0.537 at post-
oxidation steel temperatures of around 2,100�C was mastered.
1.6
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200090001-9
FOR OFFICIAL USE ONLY
Ta~le II. The Chemical Composition of Slag Following Melting and Oxiaation
~b) ~d~ _ (e) Obenh [�a~
A Rtrueka iloilnotn o~ Ou0 Cu0 ~i4s0 !
E ~y Fu0 ~ DirtO I SIO, i A1,0, I Cu0 I\[RO I Cr,O, I T10, I NIO ( tii0, I~i0, f.\I,U,
I
~ po rozta- pri~mur 1,3G 0,9~ 4,11 5'l,3 �_-~,4 Fi,Gl 'L,94 2,U0 ~0,137 5,~J~ 0,651
o veni min. 0,35 0,3~ 1,85 ?9,5 15 1 3 50 U,70 0:i~ 0 012 7:3 0 33
~ rozpgti ' ~ ' ' ' '
o~ ~g~ i) max. 17 2,85 1,48 7,60 64,0 38,9 10.40 4.97 3,50 1,020 lfi,21 1,61
~ po oxidaci priim~r I 2,91 2,28 13,?4 3fi,1 10,18 lfi.32 11.5f, 2,5fi 0,30 0.7Ei8 0,537
m min. 2,19 1,81 9,8 �8,5 7,7 1U,7 3,01 1,10 0,010 O,~Fi 0,?1
(h~ rozpgtf
�D mns. 20 3,69 3,2D 1fi,2 4?,1 ll,b 2~,? 18,9 4,70 0,059 0,9; 0,77
M po roztn- pr~Ymi;r 3,70 4,Ofl '13.3 ~,7 24,0 l~,:i 1?,l l,p3l ~,~~7
o ~�cni miti. 0,8'l 0,4fi 13,4 3,5 13,0 !!,1) 1,0 p~~ p~~l
o rozputt
~ . mnx. �0 6,17 7,b8 34,8 11,~ 42,8 32,~ 25,"_ 2,3 2,31
y -
` po oxiclnci priimi~r i U,7? 4,8'2 ]4,05 3.fi~ 1f1,:i~J i I�.R1 37.:S~1 - 0,7O'l 1,'L�iH
N min. G,28 3,03 8,30 2,12 5,7 7,U ~1,8 0,4 0,70 ~ -
~ rozp8ti ~
max. 20 ~13,01 7,�8 15,80 5,5;i 1~.1 �8,3 45.f, 1,0 1,7~
Key:
a. Melts f. With aircraft engine, without
b. Slag aircraft
c. Value g. Following melting
d. Number of Melts h. Following oxidation
e. Content (percent) i. Average, minimum, range, maximum
COPYRIGHT: SNTL, Prague, Nakladatelstvi technicke literatury n.p., 1980
8664 END
CSO: 2402
17
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200090001-9