ISIJ International, Vol. 34 (1994), No. 5, pp.

401-407

Dephosphorization Reaction of ChromiumContaining Molten lron

by CaO-basedFlux

Shin-ya Hiroyuki AOK11)and Kazuo OKOHIRA2)

KITAMURA.
Formerly Yawata R&D Laboratory. Technical Development Bureau. Nippon Steel Corporation. Nowat Process Technology
Research Laboratories. Technical Development Bureau, Nippon Steel Corporation. Shintomi. Futtsu. Chiba-ken. 299-12
Japan. )
1 Yawata Works. Nippon Steel Corporation. Tobihata. Tobata-ku, Kitakyushu 804.
2) Deceased. Formerly Yawata R&D Laboratory. Technical Development Bureau. Nippon Steel Corporation.

(Received on October 18. l993, accepted in final form on February 18. 1994)

Dephosphorization tests of chromiumcontaining molten iron by CaO-basedflux have been carried out
by an induction furnace. The following results were obtained:
(1) Whenthe carbon content in metal was high enough, chromium containing molten iron was able
to be dephosphorized by CaO-CaF2-FeO flux, In this case, oxygen partial pressure at the slag and metal
interface was calculated to be about I 0-12 Pa.
(2) The higher ratio of CaF21(CaO+ CaF.) and of Ca0/0 in flux were effective for the dephosphorization
without the oxidization of chromium.
Based on these results, 200ton scale industrial tests have been carried out by the hot metal pretreatment
station at Yawata Works. By the combination of the flux injection method and the oxygen top blowing
method, it was clarified that the dephosphorization without chromium oxidation and the suppression of
temperature drop were able to be achieved simultaneously.
KEYWORDS:
dephosphorization; stainless steel; activity coefficient; mass transfer rate; hot metal
pretreatment; oxygen gas.

for efficient dephosphorization without chromium oxi-

l. Introduction
For the dephorphorization of high chromiumand low
carbon molten steel by the oxidization reaction, a high-
ly basic fiux like BaOis necessary to prevent the oxidi-
zation of chromium. It has been reportedl,2,3) that
the dephosphorization by the CaO-basedflux becomes
possible for chromium containing high carbon molten
iron beacuse of the increase of phosphorous activity and
the decrease of chromium activity as carbon content
increased.
Recently, the smelting reduction of chromiumore has
been developed as an economical process to produce
ferro-chromium alloy.4) In this process, the phosphorus
content in the produced alloy increases as a large amount
of carbonecious material is used. It has been problematic
to makestainless steel by this process. Fortunately, the
product by smelting reduction is a carbon saturated high
chromiumalloy. The economical dephosphorization by
the CaO-basedflux can be applied and someresearch
has been achieved.2,3) Nakajima and Mukai reported2)
that the efficient dephosphorization without chromium
oxidation wasachieved by the injection of CaOand CaF2
with oxygen gas. The oxygen potential of this process
was reported to be IO ~ 8 Pa. This value is extremely high
comparedwith the optimumoxygen potential
proposed by Ohnuma
et al.5) The operation
dation has not yet been madeclear.
Onthe other hand, the melting point of high chromium
and carbon saturated iron is muchhigher than that of
carbon saturated iron without chromium. For industrial
application, the suppression of temperature decreases
during the treatment is also indispensable. Nevertheless,
the oxygen supplying rate for the control of temperature
drop is too large comparedto that for dephosphorization
without chromiumoxidation.
In this research, small scale fundamental tests were
carried out to clarify optimum dephosphorization con-
ditions. Also, the experimental results were analysed
by a coupled reaction model6,7) to comparethe oxygen
potential and slag basicity to the conventional hot metal
dephosphorization. According to the results, 200 ton
scale industrial tests were carried out. By the combination
of the flux injection methodand the oxygen top blowing
method, it was clarified that the dephosphorization
without chromium oxidation and the suppression of
temperature drop were able to be achieved simultane-
ously.
2. Experimental Method
(10 ~ I o pa) The small scale tests were carried out by the induction
conditions furnace with MgO
Iining. The chromiumcontaining iron

401 C 1994 ISIJ

 ISIJ International, Vol. 34 (1994). No. 5

Table 1. Composition range of dephosphorization flux.

Flux composition (wto/o) Oxygen(glkg)

Amountof 02
fiux (g/kg) (Nl/kg)
CaO CaF2 Si02 FeO O o'

Fundamental test 23~1 22~54 25 21~5 30 O 1.4-3.0 l.l~3 .O

Industrial test 44~62 l 6-22 o l 6~O 26,7-69.8 2,
I~LO l .O5.9 4.(~9.8

Dephcsphorizat ion rl ux + N2
(CaO~aF2- FeO) Exper i ment O
C~Icu tat i on
02 Gas o 6.0
o
-Si
a9
~ L 5 ,o

/l\ 19.0

o e
ola~ 18.0
e
o
Cr~onta ining Torpedo Car
Molten lron
Fig. 1. Hot metal pretreatment equipment at YawataWorks. 0.070
a
(about 70 kg) was melted at 1770 and the flux ofK 0.050 e
CaO-CaF2-Si02-FeOsystem was added at the bath
surface. As the fiux, the mixture of reagents
were used o 9.
o
and addedat 120 sintervals for 10 times. Thecomposition
o e o
is shownin Table 1. During the experiment,
8.
o
argon gas
(1.5 x l0~4Nm3/s) was injected from the porous brick 7,
o
located at the bottom of the furnace. The chromium O 5 1O 15 20
Time, mi n
content of the bath was about 18 o/o, while the content .

Fig. 2. Typical experimental result and the calculation

of carbon varied from to olo. 2 6 by coupled reaction model.
curves
The industrial tests were achieved by the hot metal
pretreatment plant at Yawata Works (Fig. 1). As a
reactor, a torpedocar was used. The mixture of CaO,
CaF2and iron-oxide was injected with carrier gas (N2) mass balance as the change of chromium content was
small comparing with the scatter of the analysis. In Fig.
into the hot metal (about 200 t), and also oxygen gas
was 3, the influence of carbon content
on Kp and ACr is
blown from the top blow lance. The chromiumcontent shown. As the carbon content decreased from the satu-
of hot metal waschangedover the range from to 17 "lo 5 ration value, an abrupt increase in ACr and a decrease in
chromiumand carbon content was about 6 o/.. The flux Kp Wasfound. Figure 4 shows the influence of Ca010
composition is shownin Table l. In Table indicates l,
O on KFand ACr. In this figure, Ca0/0 indicates the mixing
the weight of oxygen in iron-oxide and O' indicates the
ratio of CaO to the oxygen in iron-oxide.8) As the Ca010
total of oxygen weight in iron-oxide and that in
oxygen increased, ACr decreased and the increase of Kp Was
gas. found. It can be noted that the oxidization of chromium
wasalmost completely suppressed whenCa0/0 exceeded
3. Experimental Results 7, which
was very high comparing to the value of I or
29) for normal hot metal dephosphorization
3.1. Small Scale Tests process.
Figure 5showsthe influence of CaF2content in the flux
Typical experimental results are shownin Fig. 2. The
on Kp and ACr. The CaF2content was presented as the
curves in this figure indicate the calculation result of the mixing ratio of CaF2to the total content of CaOand
coupled reaction model explained in the following
CaF2' As the CaF2mixing ratio increased, a decrease in
section. As the carbon content
was high enough, de- ACr and an increase in Kp Were observed. The curves
phosphorization of the high chromium iron by CaO- in these figuresalso indicate the calculation results of
based fiux was able to be achieved. In the following
the coupled reaction model explained in the following
analysis, the experimental results
were evaluated by the section.
apparent dephosphorization rate (Kp I/min) which was
;
calculated under the assumption of the Ist order reaction 3.2. Industrial Tests
and the changeof chromiumcontent in metal before and The change of chromium and phosphorous content
after the test (ACr; olo). ACr
wasestimated by the (Cr203) before and after the treatmentis shown in Fig. 6. The

content in slag and the weight of slag calculated by the effective dephosphorization was achieved without the

C 1994 ISIJ 402

 ISIJ International, Vol. 34 (1994). No. 5

Influenee of CCr) on Inf luence of (Cl~03 )on Act ivi ty Coef f icient Nbss Transfer CDefficient Ef f i c i enev
Activity Coefficient Act ivi ty Coeff icient of (Cr203) of Slag Phase $ ks(cmls) af C~OF'/*
of p)in Metal Pt,ase of (FX)5) in Slag Fhase C~se A o.
2 0.0 o67 100
(:i~se
1 O: Cbnsidered O : Considered C~se B l.O O.O 134 100
(~se 2 X: NegIeoted O : Considered e~tse c o.
2 0.0 o67 75
(~se 3 o: Cbnsi dered x : Negleoted
Exper ~ment
(:i~
Icu lat i on
ease A Case B C~tse C
~! Exper,ment
Cia Icu Iat i Qtl ~:
c:ase I Case 2 C~se 3 e
0.0
6o o
0.0
6o Initial CCr] 10.0 -1 8.0'/.
Initial Cp J 0.05O~).1 OO
~
'/.
~l'
Init ial CCr) I 0.0-15.0'/.
\~
\ InitiaICP] O.O50~.100'/. 0.0 4o o
l
0,0 4o \
aL.
'~'_'J'
e) \\\ e
11)

~ ,,,,,,
,\ ~
1' 0.0
2o e
~ ll
0,0 2o ,,,,,
~?
II

,) e
~? e
o
o
e 0,3

ol~:\~
o.
4 e o
'.~ 6 0.2
e
~ o.
2
,,
'lb
'b
1~
,,
0.1
, o
'l~

o -t J
o 0.3 0.5 O.
7
o 2.0 4.0 6.0 CaF2/(Ca0+caF2 )
Initial CC) '/.
Fig. 5. Influence of CaF21(CaO CaF2) in fiux on Kp and
, +
Fig. 3. Influenceof carbon content on Kp and ACr for ACr for fundamental tests. (Ca0/0= 5, [C]: 5-60/0)
fundamental tests.
(Ca0/0 = 5, CaF2/(CaO+ CaF2)= O.d~0.5)

Exper i ment e
\E Cia Icu Iat ~on e
0.0 6o X
InitiaICCr) 1 5.0 -20.0 '/.

*,
\ Init ial (P J 0.0 50~. I OO'I.
t
~ 0,0 4o
o
~
~ 0.0
2o e
i e o o
~? 6 10 14 18
o CCr) ,
'/.

o
Fig. 6. Changeof phosphorousand chromiumcontent before
0.3
and after treatment for industrial tests.

o 0.2 , e
e mental conditions. Ca0/0' (based on the total oxygen
of iron-oxide and oxygen gas) was below 3.2. This con-
o.
1
o dition was considered that the oxidation of chromium

o would take place after the result of the small scale tests.
3.
o o 5, 7,
o Nevertheless, in the industrial test, visible chromium
CaOlO oxidation was not observed. In Fig. 8, the influence of
Fig. 4. Influence of Ca0/0 in fiux on Kp and ACr fot
fundamental tests.
Ca0/0 based on the injected oxygen in iron-oxide on
([C]: 5-6 o/o, CaF2/(CaO
ACr' and the apparent dephosphorization rate (K~) is
+ CaF2): 0.4~).5) shown. Where, K~ is calculated by the phosphorous
content before and after the treatment and the treat-
visible decrease of chromiumcontent. The influence of ment time. In a similar chromiumcontent range. ACr'
the oxygen gas ratio (010') on the decrease of tem- decreased and K~ increased as the increase of Ca010
perature and the change of chromium content before coincident with the small scale tests. Theseresults indicate
and after the treatment (ACr') is shownin Fig. 7. It can that the dephosphorization reaction is controlled by the
be found that the temperature decrease can be suppressed oxygen in iron-oxide injected with CaO,and that the top
as the increase of the oxygen gas ratio, without the blown oxygen has little effect. In industrial tests, good
increase of chromium oxidation. Under these experi- dephosphorization behavior was observed instead of a

403 C 1994 ISIJ

 ISIJ International. Vol. 34 (1 994), No. 5
The reaction rate of Eq. (2) is expressed by Eq. (3) on
0.6
Ca0/0' ~ 3.2 the assumption that the masstransfer of the slag and the
metal phase are the rate controling steps.
0.4

o [M]+n'[O]=(MO~) ....,....
.....,....(2)
0.2 o
o JM= {k~ p~/(100 NM)} {[oloM]b [o/oM]i}
-
' ' '

o OO NMo.)} {(o/oMO~)i (o/oMO~)b}

= {k. ' p./(1 ' '

- *(3)
o
For decarburization, a phenomenological rate param-
o eter for COevolution (Gco) is introduced and the re-
~50 action rate is expressed by Eq. (4).
~:

H o Jc = {k~ p~/(1 OO Nc)} {[o/oC]b [oloC]i}

_tOO
-too
ee
' ' '

-
= Gco'(P~o/PI ~ l) ........
...,......(4)

o The equations for the equilibrium partition ratio of

each reaction at interface can be expressed by Eq. (5)
20 40 60 80 1OO (effective equilibrium constant).
Rat io of Oxy9en Gas, '/.

Fig. 7. Infiuence of oxygen gas ratio on ACr' and ATduring EM= (o/oMO~)i/([o/oM]i '
a~")
treatment for industrial tests.
(lOO C NMOfM KM)/(p.'yMo~) .(5)
....
""
By these equations, the molar fiux density (J) of each
0.0
6o o element can be expressed as a function of a~ and by the
, o
relation of electrical
determined numerically.
neutrality (Eq. (6)), ab can be

'~
0.0 4o o 2JSi+ 1.5Jc.+JF.+2.5Jp+Jc~J0=0 .........(6)

\
CCr] '/.
The change of concentration for each element in
0.0 2o ,
the metal and the slag phase can be calculated by
a
~: e 5~ 8.5
4.

using a~.
o 1 6~1 7
In the present analysis, for the masstransfer coefficient
o in the metal and the slag phase and for the decar-
0.8 burization rate, the values obtained by the dephos-
~~~..

phorization and desiliconization tests under the same

experimental conditionslo) were used.
O 0.4
To determine the effective equilibrium constant, the
o equilibrium constant and the activity coefficient in the
e o metal phase can be calculated based on the various data
o already reported.11) On the other hand, the activity
O 5.0 10.0 coefficient of each component in the slag containing
Ca0/0 of IniectionFlux Cr203 and CaF2Wasnot clear. In this study, the activity
Fig. 8. Infiuence of Ca0/0 of fiux on K~ and ACr during was determined by the following considera-
coefiicient
treatment for industrial tests. tion.
(1) The activity coefficient of FeO: Based on the
mixing ratio of CaF21(CaO +CaF2) below 0.3, as the experimental result for CaO-CaF2-FeO system by lchise
fine powder of CaO used. No significant damage and lwase,12) the activity coefficient of FeO
was was defined
was observed for the refractory of the torpedo car. as I .5. Theinfluence of slag composition and temperature
were neglected.
4. Discussion (2) The activity coefficient of Cr203: The activity of
Cr203 in a CaO-CaO Cr203 Saturated slag system can
OxygenPotential
'
4.1. Evaluation of
be calculated by Eq. (9). This relation was introduced
In the coupled reaction model, the reactions shownin by the standard formation free energy of Eqs. (7) and (8)
Eq. (1) are considered for this research,
reported by Takiguchi and Sano,13)
[Si]
+ 2[O] = (Si02) 2[Cr] + 3[O] + CaO= CaOCr203 '

[Cr] + I .5[O] = (CrOl .5) AG~= - I 036 830+372.9T [J/mol] ..,..(7)

Fe + [O]=(FeO) ......
.,........(1)
2[Cr] + 3[O] Cr203
=
[p] +2.5[O]=(P02.5) AG'2 = 797 070+275.5T [J/mol] ........(8)
-
[C] + [O] = CO
In ac,,o,
= ~28870/T+ II .7 ...............(9)
C 1994 ISIJ 404
 ISIJ International. Vol. 34 (1994), No. 5
Table 2. Parameters used by the present calculation.

k,~ k* Gco
(mol/cm2/s)
Esi Ep* Ec Ep Ecr
(cm/s) (cm/s)

0.067 0,0067 66x 10~e 7x 1012 540 l 6520 l 5x 1013 4x 106

The activity coefficient of Cr203 was defined as 0.2
based on the activity calculated by Eq. (9) ~nd the average
concentration of Cr203. The influence of slag composi-
tion and temperature were neglected.
(3) The activity coefficient of P205: For CaF2con-
taining CaO base flux, Suito and Inouel4) have proposed
an empirical equation to calculate the influence of slag
composition and tempierature on the activity coefficient
of P20s' By using this equation, no good correspondence
between the calculated and experimental values were
obtained. Then, the constant value of Suito's equation
was changed to agree with the calculated concentration
changeof phosphrouswith the typical experimental result
and Eq. (lO) was obtained. Where the coefficient on
Cr203 in the equation was deduced by the relationship
between the values of the coefficient for each oxide in
the equation and the attractive force between each ion
and oxygen.Is)
10g(yp,0=)
= ~ I .02 x (23Xc.o + 20Xc.F, + 17XMgo
+ 8XF*o~ 26Xp,0=+ 5Xc*,o,)
-22 900/T+ 16.341 .......................(lO)
Case 3: The influence of
slag composition on the
activity of (P205) was neglected.
coefficient
The calculated line for case 3does not agree well with
the experimental result, in which only the activity change
of phosphorous in metal phase is considered. On the
I
cOntrary, the calculated lines for case and 2 agree well
with the experimental values. By the calculation for case
3, it
can be found that unless the increase of the activity
of (P205) by the increase of (Cr203) content was
considered, Kp increased with the decrease of the carbon
content. As the decrease of carbon in metal, FeOand
Cr203 content in s]ag increases. So, the reason of the
increase in Kp Would be the increase of the oxygen
potential at the interface with the decrease of carbon
content. Also, it should be mentioned that the influence
of carbon content on the dephosphorization can not be
explained by only the activity change of phosphorous in
metal phase. According to these results, it can be con-
sidered that not only the factor
(~) and (~), the following
O
but also the factors
factors, have to be
among
considered to explain the infl'uence of carbQn content on
the dephosphorization:

In Table 2, the typical values of each parameter used O

content.
Theactivity changeof phosphorousby the carbon
in this study are summarized.
~) The activity change of (P205) by the (Cr203)
The calculation results are shownin Fig. 2. By the use
content.
of Eq. (lO) for the estimation of the activity coefficient
of P205, a sound correspondence betweenthe calculation
C) The change of oxygen potential at reaction
interface by the carbon content.
curves and the experimental values was obtained. In this
calculation, the activity of oxygen at the slag and metal Ontheother hand, regarding the effect of Ca0/0, the
interface was about 2.6 x 10~5 (4 x lO~i2 pa, expressed
comparison of the calculation curves with the experi-
mental resu]ts was shown in Fig. 4. Sound agreement
in the partial pressure of oxygen) and the effective
equilibrium constant of phosphorous (Ep) was about was found for ACr and Kp. Based on the calculation,
whenCa0/0 increased from 3 to 7, partial pressure at
1013. Onthe contrary, Ohguchi et al.7) have reported for decreased by about 1/3 and the activity of
the interface
the hot metal dephosphorization tests by the CaObased
flux (CaO/Si02 3) that the activity of oxygen is about
P205 decreased about l/10. These decreases can be
~ considered as the reason why the dephosphorization
2x l0~3 (10~9Pa) and Ep is about 109. According to without the chromium loss can be achieved with the
these calculations, it
can be found that the oxygen increase in Ca010.
potential of this experiment is lowered by the decrease
of FeOcontent and the activity of P205 is also lowered
Figure 5showsthe influence of the CaF2mixing ratio.
In this figure, the calculation curves for the following
by the increase of CaO/Si02 compared to the normal
three cases are drawn.
hot metal pretreatment condition. Under these condi-
Case A: A11 factors shown in Sec. 4.1 were taken
tions, effective dephosphorization without the visible
into accouht.
oxidiation of chromiumcan be achieved.
Case B: The activity coefficient of Cr203 was as-
sumedto be I .O and the masstransfer coefficient in the
4.2. Influence of CarbonContent and Flux Compositions slag phase was increased to be twice that of Case A.
Figure 3showsthe comparison betweenthe calculation Ca,se C: 75"/o ofthe added CaOwas assumedto be
and the experimental values for the influence of carbon effective.
content on ACr and Kp. In this figure, the calculation The influence of CaF2 cannot be explained by the
results by the following three conditions were drawn. calculation result of CaseA. In the condition of Case B,
Case I : A11 factors discussed in the previous section the activity coefficient of Cr203 was determined by the
were taken into consideration. fact that the activity of Cr203 increased from 0.01 to
Case 2: The influence of chromium content on the 0,1 when CaF2/(Ca0+CaF2)exceeded 0.616) and the
activity coefficient of phosphorous was neglected. change of the masstransfer coefficient in the slag phase

405 C 1994 ISIJ

 ISIJ International, Vol. 34 (1 994), No. 5
was determined as the viscosity of slag decreased by the As the amount of slag exists at the hot spot, re-
little
addition of CaF2' The 6alculation results in case B
cor- phosphorization hardly take place and the infiuence
on
responded to the experimental results of the large mixing the dephosphorization behavior would be small.
ratio of CaF2' The condition of Case Cwas considered
based on the results that the unslagging CaOwasobserv- 5. Conclusrons
ed whenCaF21(CaO +CaF2)decreased below 0.2. The
calculation results in Case C corresponded to the ex- Dephosphorization tests of chromium containing
perimental results of the small mixing ratio of CaF2' molten iron by CaO-basedflux have been carried out by
By these calculations, the influence of CaF2 can be an induction furnace. The following results were ob-
considered as the results of the activity change of Cr203, tained :
the viscosity change of slag and the enhancementof (1) Whenthe carbon content in metal was high
slagging of CaO. In this report, the change of the liquid enough, chromiumcontaining molten iron was able to
fraction of slag was not taken into consideration be dephosphorized by CaO-CaF2-FeO flux. In this case,
the oxygenpartial pressure at the slag and metal interface
quantitatively, and investigations should be continued in
order to clarify the effect of slagging by the addition of was calculated to be about lO~ 12 Pa.
CaF2' (2) The higher ratio of CaF2/(CaO+ CaF2) and
Ca0/0 in flux were effective for the dephosphorization
4.3. Influence of Top Blown Oxygenon ChromiumOx- without the oxidization of chromium.
idation Based on these results, 200 ton scale industrial tests
Deducing from the experimental results of industrial have been carried out by the hot metal pretreatment
tests, it
can be considered that the dephosphorization station at YawataWorks. By the combination of the fiux
reaction would be controlled by the injected oxide but injection method and the oxygen top blowing method,
the top blown oxygen gas had little influence on the it
was clarified that dephosphorization without chromi-
reaction. The calculation results of the temperature umoxidation and the suppression of temperature drop
dependencefor fundamental test conditions is shownin were able to be achieved simultaneously,
Fig. 9. In this calculation, the temperature dependence
Nomenclature
of each factor was defined based on the previous
a~ : Henrian of oxygen at slag/metal in-
activity
report. I o)
The temperature of hot spot formed by the top blown terfaceon a weight percentage basis
Ct : total molar concentration in slag (mol •cm~3)
oxygen was reported to be about 2000K. Around this
temperature, as the decarburization reaction mainly
f: Henrian activity coefficient

occurred, the oxidation of chromium and dephos- Gco: Phenomenological rate parameter for COevo-
cm~ 2 s~
lution (mol l)
phorization do not take place visually. • .

By this reason, the top blown oxygen has the role to J: molar flux density (mol•cm~2.s~1)
proceed the decarburization without the oxidation of K: equilibrium constant
chromium, and the exothermic reaction by the de- k : masstransfer coefficient (cm s~1) '

carburization N: molecular weight (g 'mol~1)

suppresses the temperature decrease.
p~o : super saturated pressure of COgas at reaction
interface (Pa)

~ 6o 0,0
Exper i ment
C~Icu Iat i on
o
T:
t:
temperature (K)
time (s)
X: mol fraction
.~
\~~ y: activity coefficient with respect to the pure
:}~"
0.0
4o substance as standard state

~ , p: density (g'cm~3)
1 0,020 Subscri pt
ll
.o
2o
metal
~?
o
e m: slag
s:
o M: nameof metal composition
MO~: nameof slag composition

i~'~.'
0.3 Superscript
b: bulk
T
O 0.2 i: interface
e ee
0.1

o REFERENCES
l) T. Usui, S. Inoue. K. Yamada,Y. Kikuchi. Y. Kawai and H.
o 1700
OO 1900 21 OO
Mori: Tetsu-to-Hagan~, 72 (1986), A25.
2) Y, Nakajima and M. Mukai: ISIJ Int., 33 (1993), 109.
Temperature , K 3) M. Matsuo, K. Mayaand K. Kamegawa:Tetsu-to-Haganb, 78
Frg. 9. Influence of temperature on ACr and Kp for the (1992), 231 .

condition of fundamental tests calculated by coupled 4) M. Fujita, H. Katayama, M. Kuwabara, C. Saito, H. Ishikawa
reaction model. and H. Kajioka: Tetsu-to-Hagan~, 74 (1988), 680.

C 1994 ISIJ 406

 ISIJ International, Vol. 34 (1 994), No. 5
5) Y. Ohnuma,N. Sumida, T, Sakuraya, A. Nanba, T. Nozaki, A. 1l) Steelmaking Data Sourcebook Revised Edition, ed. by The 19th
Tamida, H. Nomuraand H. Bada: Tetsu-to-Hagan~, 72 (1986), Committee(Steelmaking), The Japan Soc. for the Promotion of
~Sci., ISIJ, Tokyo, (1984).
A21 .

6) S.Kitamura, Y. Mizukami. T. Kaneko, T. Yamamoto, R. l 2) H. Ichise and M. Iwase: Physical Chemistry of Steel Refining
Sakomura,E. Aida and S. Onoyama:Tetsu-to-Hagan~, 76 (1990), Reaction and Process Engineering, ISIJ, Tokyo, (1985), 31 .

1801. 13) S. Takiguchi and N. Sano: Tetsu-to-Hagan~, 74 (1988), 809.

7) S. Ohguchi. D. G. C. Robertson, B. Deo, P. Grieveson and J. 14) H. Suito and R. Inoue: Physical Chemistry of Steel Refining
H. E. Jeffes: Ironmaking Steelmaking, 11 (1984), 202. Reaction and Process Engineering, ISIJ, Tokyo, (1985), 82.
8) T. Kaneko, S. Mizoguchi and T. Yakushiji: Tetsu-to-Hagan~, 69 l 5) S. Ban-ya: The 122nd and 123rd Nishiyama Memorial Seminar,
(1982), S18. ISIJ. Tokyo, (1988), 1.
9) S, Kitamura, T. Kitamura, E. Aida, R. Sakomura, T. Kaneko 16) M. Iwase: The 122nd and 123rd Nishiyama Memorial Seminar,
and T, Nuibe: ISIJ Int., 31 (1991), 1329. ISIJ. Tokyo, (1988), 87.
lO) S. Kitamura. T. Kitamura. K. Shibata, Y. Mizukami, S.
Mukawaand J, Nakagawa:ISIJ Int., 31 (1991), 1322. (Originally published in Tetsu-to-Hagan~, 79 (1993), 1242, in Japanese)

407 C 1994 Is]J