ISIJ International, Vol. 36 (1 996), No. 4, pp.

388-394

Ef tect of Stirring on Oxidation Rate of Molten Steel

Katsuhiro SASAIand YoshimasaMIZU KAM

11
)
NagoyaR&D Laboratories, Nippon Steel Corporation, Tokai-machi, Tokai. Aichi-ken, 476 Japan.
1) Formerly Nagoya R&D Laboratories, Nippon Steel Corporation. Nowat Japan R&D Cente~ for Metals. Toranomon,
Minato-ku, Tokyo, I 05 Japan.

(Received on August 75. l995.• accepted in final form on November29. 1995)

As basic research to quantify the oxidation rate of molten steel in tundish, oxidation experiments were
performed on nonkilled molten steeland aluminum-killed molten steel on laboratory scale, and the effect
of stirring on the oxidation rate was studied. The fol[owing conclusions were derived. The oxidation rate
of the nonkilled molten steel by air is not affected by stirring and is controlled by the diffusion of O, gas
in the gas phase. Stirring changes the oxidation rate because of breaking the oxide film in the oxidation
process of the aluminum-killed molten steel by air, The oxidation rate of the aiuminum-killed molten steel
in the still state is controlled by the diffusion of oxygen in the oxide film, while the oxidation rate of the
aiuminum-killed molten steel in the stirred state is controlled by the diffusion of O, gas in the gas phase,
KEYWORDS:
secondary oxidation; stirring; tundish; oxidation rate; diffusion; oxygen.

1. Introduction 2. Experimental Methods

In continuous casting process of steel, a tundish is 2.1. Oxidation Experiments of Molten Steel in Still
positioned between a ladle and a mold and is used for State
the purposes of distributing molten steel into the mold In the oxidation experiments of molten steel in the still
and removing nonmetallic inclusions from the molten state, an atmospherecontrollable Tammann furnace was
steel. The progress of secondary refining technology has used. 500 of electrolytic iron was melted in an alumina
significantly improved the cleanliness of molten
and steel
g
crucible, having inside diameter of and height of 40mm
has conversely revealed the contamination of the molten
steel in the tundish. Therefore, how to prevent the
150
mmin Arironatmosphere.
electrolytic
The chemical composition of
shown in Table
Is After the molten 1.
contamination of the molten steel in the tundish has steel temperature wasadjusted to 1600'C, aluminumwas
becomean extremely important issue. Entrapment of added to a target concentration of O.Imasso/o for
ladle slag, reaction with tundish cover powder and wear aluminum-kllled molten steel (Fe-AI melt) in order to
of refractorles are cited as important factors responsible facilitate the estimation of the oxidation rate. For
for the contamination of molten steel in the tundish. nonkilled molten steel (Fe melt), the electrolytic iron was
Secondary oxidation of the molten steel by air has the used with no alloy additions. The oxidation experiment
significant effect and is considered the principal cause of of nonkilled molten steel was carried out to compare
contamination of molten steel in the tundish.1) The with that of aluminum-killed molten steel. In someex-
stirring condition of molten steel in the tundish greatly periments, sulfur was added in an amount of O. I masso/o
varies from part to part, and this has the possibility of to both the nonkilled molten steel and the aluminum-
changing the oxidation rate of the molten steel by the killed molten steel. Ar02 gas mixture
was introduced
air. Production of
a highly clean steel needs quantitative through the bottom of the Tammann furnace at a flow
determination of the oxidation behavior of the molten rate of 5000Ncm3'min1for 2min to adjust the gas
steel by location in the tundish and implementation of phase to the desired partial pressure of 02 gas. The flow
appropriate measures for preventing the oxidation. rate of Ar-02 mixed gas was reduced to 2OOO Ncm3'
In this study, the effect of stirring on the oxidation min~ 1, and the molten steel oxidation experiment was
rate of molten steel by air has been investigated as started. The gas flow condition to control the partial
basic research to quantify the oxidation rate of molten
steel by location in the tundish of a continuous caster. Table l. Chemical composition of dectrolytic iron used in
the experiments. (ppm)

C Si
Mn P S O
25 14 5 31 32 135

C 1996 ISIJ 388

 ISIJ International, Vol, 36 (1996), No. 4

pressure of 02 gas was determined by measuring the the above-mentioned oxidation experiments, in order to
partial pressure of 02 gas In Tammann furnace in facilitate the observation of oxides. The partial pressure
advance. As an experimental condition, the partial of 02 gas in the gas phase was adjusted to 20 kPa by the
pressure of 02 gas in the gas phase was changed over samemethodas in the oxidation experiments in the still
5
the range of to 50 kPa. Molten steel samp]es were taken state, and the oxidation experiment of the molten steel
wlth a quartz tube of 6mm
inside diameter at specified wascarried out. The experimental time waschangedover
time Intervals and analyzed as to O
concentration for O
the range of to 20 min, and the molten steel was rapidly
the nonkilled molten steel and as to Al concentration for cooled in the water together with the crucible after the
the aluminum-killed molten steel. To investigate the effect oxidation experiment. The solidified sample was longi-
of the oxide film on the oxidation rate, the molten steel tudinally halved and examined by optical microscopy
was electromagnetically stirred (20 kW, 500 Hz) In some and electron probe microanalyzer (EPMA)for the con-
experlments. dition of oxides formed on the molten stee] surface.

2.2. Oxidation Experiments of Molten Steel in Stirred

3. Experimental Results
State
In the oxidation experiments of molten stee] in the In the oxidation experiments, the molten steel surface
was visually observed. In the oxidation experiments of
stirred high-frequency
state,
a vacuuminduction melt-
ing furnace (50kW, 2.7kHz) was used. 70 to 80kg of the nonkilled mo]ten steel in both the still state and the
electrolytic iron was melted in a magnesia crucible, stirred state, small globular liquid oxldes repeatedly
having inside diameter of 220 and height of 400 formed and dissolved, never covering the entire molten
mm
in Ar atmosphere at a total pressure of 40kPa. The
mm steel surface. In the oxidation experiments of the
total pressure was determined a little lower than atmo- aluminum-killed molten steel, stirring changed the
spheric pressure in order to prevent the air leak of the condition of the oxide film formed on the molten steel
surface. In the still state, the oxide film uniformly
vacuuminduction melting furnance. The temperature
and composition of molten steel were the same as in covered the entire molten steel surface. In the stirred
the oxidation experlments in the Tammann furnace. Ar state, numerous cracks formed in the oxide film,
gas and 02 gas were introduced into the vacuuminduc- composedof solid particles, and exposed a very large
tion melting furnace. The partial pressure of 02 gas area of the molten steel surface.
in the gas phase was adjusted to the desired value, 4 to The optical micrographs of the gas-metal interface
20 kPa, at a total pressure of 80kPa, and the oxida- obtained in the rapid quenching experiments are shown
tion experiment of molten steel was started. Molten steel in Fig. l, and the EPMA
Iine analysis results of surface

samples were taken at specified time intervals and oxides are shown in Fig. 2. A
thin FeO(Fe/O molar
analyzed as to O
concentration for the nonkilled molten I I
ratio of about .O to .1 as calculated from the EPMA
steel and as to Al concentration for the aluminum-killed line analysis results)
was partly formed on the nonkilled
moten steel. molten steel surface in the oxidation process. The oxide
accounted for a very small proportion of the entire molten
2.3. Rapid Oxides Quenching Experiments steel surface, and this situation did not changewith time.

To investigate the oxides formed on the molten steel 1

The melting point of FeOis 37'O to 390'C.2) This 1
surface in the oxidation process, the Tammann furnace indicates that FeO was in the liquid state at the
was used. 25g of electrolytic iron was melted in an experimental temperature. On the other hand, a thick
alumina crucible, having inside diameter of II mm
and oxide film with a scatter of granular FeAl204 (Fe/Al/O
height of 100 in Ar atmosphere. Nonkilled molten molar ratio of abut 1.0 to l.6 to 3.8 as calculated from
mm
steel had the chemical composition of the electrolytic the EPMA
Iine analysis results) in FeO(Fe/O molar ratio
iron, and aluminum-killed molten steel had aluminum of about 1.0 to 1,1 as calculated from the
EPMA
Iine
concentration increased to 0.2 masso/o as comparedwith analysis results) formed the entire aluminum-killed
on

/
Oxide
Fig. 1.

Microstructures of gasmetal interface obtained

~.~,*~, by rapid quenching experiments.
~}
(a) Metal (b)
Metal ~ ~ (a) Fe melt, Po,=20 kPa, after min5
(b) Fe-AI melt, Po,=20kPa, after 5min
L-l
40 m
p
L--J
1OO m /~e

389 @1996 ISIJ

 ISIJ International, Vol. 36 (1996), No. 4

l OO 50
60 1oo 70
L'~-,
Fe melt .
L-=] Fe-AI melt lo/*m
P02=20kPa ! 10!' m p02=20kPa
60
50 atter 5mln
80 40 FeO l after 5min 80 Feo FeAI204
50
Fe _J,'

60 30
~ Fe 40
~~.'
60
'

~~.' ~~. ~~.'

40
i lo
~~.'

(,)

cT'
'r'
~E 30 ~ ao

E E
cT'
E E
~ 40 ~~
30
~ 40 o 20 -o :~

20
f
f v .i
f'-
'-'~

t
'

1lit1l
' '
20

l!
, .

'- AI
20 10 20 .
j~

j
'
10
10 If

~/
l~!

o o o o o
Fig. 2. EPMA
Iine analysis results of oxides on molten steel surface.

O08
.

[] P02=5kPa In still

Fe melt
state
Op02= 5kPa In still

Fe-AI melt
state

o 20
Cl p02=10kPa WF.=500g Op02=10kPa WF*=500g
P02=20kPa 1600~)
Op02=20kPa
.
l] 16001C
l P02=50kPa o 06
.

~ROO ~RO
,,)
cl'
0.15 'l)
c')

cO
,e

E E
12
o 04
,

~ 0,10
O O
O02
o 05 ,
.

o o
O 5 1O 15 20 25 30 O 5 1O 15 20 25 30
Time (min) (min)
Time
Fig. 3. Change with time in oxygen absorption [O]Ab* of Fig. 5. Change with time in
oxygen absorption [O]Ab, of
nonkilled molten steel in still state. aluminum-killed molten steel in still state.

0.12
C] P02=8kPa In stirred
Fe melt
state
Op02= 4kPa In stirred
Fe-AI melt
state

o 20
DI P02=15kPa WF.=70kg 0.lO Op02=10kPa WFe=80kg
.

l P02=20kPa 1500t e p02=20kPa 1600t)

~~ ~R
o 08
~
E
0.15 U,
U,
Ce
E
,

o 06 .

!:
0,10 ~
O O o 04 ,

o 05 .

o 02 .

o o
O 5 1O 15 20 25 30 O 5 1O 15 20 25 30

Time (min) Time (min)

Fig. 4. Change with time in
oxygen absorption [O]Ab* of Fig. 6. Change with time in
oxygen absorption EO]Ab* of
nonkilled molten steel in stirred state. aluminum-killed molten steel in stirred state.

molten surface. At the experimental temperature,

steel point is
21
35'C.2) When measuredby an image analyzer,
FeOwas considered to be in the liquid state, and FeAl204 the average thicknesses of the oxide fiim were 220, 225
was considered to be in the solid state because its melting 5
and 262 ktm at O, and 20 min respectively after the start

@1996 ISIJ 390

 ISIJ International. Vol. 36 (1996), No. 4
of the experiments. From these results, it may be gas partial pressure in the gas phase, Po,'
considered as follows. The oxidation reaction of the The
relationship between the oxidation rate of the
aluminum-killed molten steel rapidly proceeds initially. molten steel in the still state, d[O]Ab*/dt, and the partial
Oncea thick oxide film forms, the oxidation rate of the pressure of 02 gas in the gas phase is shownin Fig. 7.
alumlnum-killed molten steel decreases, and so does the Therelationship betweenthe oxidation rate of the molten
growth rate of the oxide film. steel in the stirred state and the partia] pressure of 02

The changes with time in the amounts of absorb- O gas in the gas phase is shown in Fig. 8. The oxidation
ed, [O]Ab~, by the nonkilled molten steel and the rate of the nonkilled molten steel increases in direct
aluminum-killed molten steel in their oxidation proc- proportion to the first power of the partial pressure of
ess, obtained by analysis of molten steel, are shown in 02 gas, irrespective of whether it is stirred or not. The
Figs. 3 to 6. WF. is the weight of each molten steel. oxidation rate of the aluminum-killed molten steel
O absorption of the aluminum-killed molten steel is increases in direct proportion to the l/2 power of the
calculated by assumingthat all of absorbed reacts with O partial pressure of 02 gas in the still state and to the first
Al to form Al203. Under all experimental conditions, O powerof the partial pressure of 02 gas in the stirred state.
absorption linearly increases with time, and the rate of
increase in O
absorption increases with increasing 02 4. Discussion

The oxidation reaction of the nonkilled molten steel

p02XI0~3 (Pa)
and the aluminum-killed molten steel can be considered
5.0 10.0 20.0 50.0 to consist of elementary steps, a) to e) and a) to g).
co
respectively.
\ state
In still
e)O diffuslon control
wFe=500g, 1600t - a) Diffusion of 02 gas in gas phase
~~ROO
~.
,,)
,,) - 3.5 .
b) Chemical reaction at gas-oxide interface
ct;
C]fe melt c) Diffusion of
oxygen in oxide film
E g)~I dlftuslon control
!] Fe-S melt d) Chemical reaction at oxide-metal interface
JD

~4.0
e) Diffusion of
Oin molten steel
O~ OFe-AI melt
D Chemical reaction between Oand A1 in molten steel
10 OFe-Al-S melt ~l 1 g) Diffusion of Al in molten steei
~ "'
~
OFe-AI melt,
EMS
1 Of the above elementary steps, the slowest one con-
trols the oxidation rate of the molten steel. Generally,
ol:)
-4 . 5
chemical reactions proceed fast at high temperatures.
l:, ~(] 1 The oxidation rate of the molten steel is measured at
2 relatively low 02 gas partlal pressures in this study and
o)
o _5 . O is lower than the oxidation rate of molten steel by
pure
3.6 3.8 4.0 4.2 4.4 46 48 02 gas, which is reported to be controlled by the chemical
Iog P02 Cp02 (Pa)] reaction at the gas-oxide interface. 3) Taking these factors
Fig. 7. Relationship between oxidation rate ci[O]Ab.lcll of into account, there is the small possibility of either
molten steel in still state and partial pressure ofoxygen interfacial chemical reaction b) or d) becoming the
Po, in gas phase. rate-determining step. The chemical reaction between

P02XI0~3 (Pa)
deoxidizing elements and
Q
is reported to be extremely
fast,4) and
O
absorption linearly increases with time in
5.0 10.0 20.0 this study. For these reasons, chemical reaction f) can
also be excluded as the rate-determining step. Diffusions
co In stirred state e)~1 diffusion control
\~~ -3.0 WF5=80kg, 1600~
L_
a), c) and e) in the nonkilled molten steel and diffusions

,,,
_ _ . _ a), c), e) and g) in the aluminum-killed molten steel
c/) [] Fe melt g)A! diffusion control
are studied to see if
one them is the rate-determining
of
~
ct;

E -3 . 5 Fe-S melt _L._ step.

__ _
J2
O Fe-AI melt
4.1. Reaction Rate Model by Rate-determining Step
O1) C) Fe-Ai-S melt
1:: -4 . O 4.1.1. Model Regarding Diffusion of 02 Gas in Gas
c"
Phase as Rate-determining Step
JD
If the oxidation rate of the molten steel is assumedto
O~ 1:~
-4 5 1 be controlled by the diffusion of 02 gas in the gas phase,
1:)
1 Eq. (1) holds.
o'
o _5 ' O
dno/dt=2A •kG• (Co,~ Co,(G-F)
"- ) * ..
""'(1)

3.4 3.6 3.8 4.0 4.2 4.4 4.6 where no represents numberof moles of Oabsorbed by
log P02 Cp02 (Pa)] the molten steel (mol); t, time (s); A, surface area of the
Fig. 8. Relationship between oxidation rate d[O]Ab*/dt of
molten steel (cm2); kG, mass transfer coefficient of 02
1); Co,, molar concentration
molten steel in stirred state and partial pressure of gas in the gas phase (cm s~ '

oxygen P02 in gas phase. of 02 gas in the gas phase (mol cm~3); and subscript
•

391 C 1996 ISIJ

 ISIJ International, Vol. 36 (1996), No. 4
G-F, gas-oxide interface. Whena state equation of gas I .3.
4. Model Regarding Diffusion of Oin Molten Steel
is applied to Eq. (1), the oxidation rate of the molten as Rate-determining Step
steel expressed by the change in the concentration of
O If the oxidation rate of the molten steel is controlled
absorbed by the molten steel, d[O]Ab~/dt, is given by by the diffusion of Q
in the molten steel, the oxidation
rate of the molten steel is given by
d [O]Ab~/dt
= 200MOA kG (Po. ~ P0=(G- F))/(p
' ' ' ' V' R ' T)
.(2) clno/clt=A 'kM,o'(Co(F-M)~Co)
"" "'
where [O]Ab* represents concentration of Oabsorbed by where kM,o represents mass transfer coefficient of O in
the molten steel (massolo); Mo, molecular weight of O the molten steel (cm s~ l) and Co, molar concentration
'

(g mol - 1); Po,, Partial pressure of 02 gas in the gas phase

' of O in the molten steei (mol•cm3). Whenthe molar
(Pa); p, density of molten steel (g'cm~3); V, volume of concentrations in Eq. (7) are converted into mass con-
molten steel (cm3); R, gas constant (Pa'cm3'mol1 ' centrations and arranged, the oxidation rate of the
K~1); and T, absolute temperature (K). The partial molten steel is given by
pressure of 02 gas at the gas-oxide interface is reduced
in the nonkilled molten steel and the aluminum-killed A kM,o ([O](F_M) - [O])/V .......(8)
d[O]Ab,/dt= • '

molten steel, and

presumedto be lower than the 02
is where [O] represents Oconcentration in the molten steel
gas partial pressure of 8.1 x l0~4 Pa5) In molten FeOin (massolo). If O concentration at the oxide-metal inter-
equilibrium with the molten steel at 1600'C. Since P0= face is in equilibrlum with molten FeO [O] '

, (F_M) rs
4
in this experimental study is greater than kPa, Po,(G-
F)
0.23masso/o at 1600'C.5) Since concentration is O
is negllgible. extremely low for the aluminum-killed molten steel and
If the oxidation rate of molten steei is controlled by is low for the nonkilled molten steei at the beginning
of
the diffusion of 02 gas in the gas phase according to the the experiment, [O] is lower than [O](F_M). If the
above results, a linear relationship holds between [O]Ab~ oxidation rate of the molten steel is controlled by the
and t, and d[O]Ab*/dt increases in proportion to Po,' diffusion of O
in the molten steel, therefore, [O]Ab,
4.1 .2. Model Regarding Diffusion Oxide of Oxygenin
linearly changeswith t, and [O]Ab*/dt becornes constant,
d
irrespective of P02'
Film as Rate-determining Step
If the species of
oxygen diffusing in the oxide film is 4.
I .4. Model Regardlng Diffusion of Al in Metal Phase
simp]y assumedto be oxygen (O) in the oxide film when as Rate-determining Step
the oxidation reaction of molten steel is proceeding, If the oxidation rate of the aluminum-kiiled molten
the dissolution reaction of 02 gas in the oxide film is steel is controlled by the diffusion of Al in the molten
given by Eq. (3), and the apparent equilibrium constant steel, Eq. (9) holds.
K' is given by Eq. (4).6)

11202(g) (O) ..........(3)

dnAl/dt
= A kM.AI • ' (CAI
- CAl(F-M)
) .* "' .*
""'(9)
= ........ where nAI represents numberof moles of Al that reacts
K' C(o)/ Pg~5 .(4) with Oabsorbed by the molten steel; kM,Al, masstransfer
= . . . .. . . ..

coefficient of A1 in the molten steel (cm s~ 1); CAl' molar

where C(o) represents molar concentration of (O) in the
'

concentration of Al in the molten steel (mol•cm~3).

oxide film (mol cm~ 3). If the oxidation rate of the molten
'

steel is controlled by the diffusion of (O) in the oxide Whenthe molar concentrations in Eq. (9) are converted
into mass concentrations, the oxidation rate of the
film whenthe growth of the oxide film can be ignored,
aluminum-killed molten steel is given by
Eq. (5) holds.

d[O]Ab,/dt = 3MoA kM,AI ([Al]

- [Al](F_ M))/(2 MAI V)
= A k~
' • '
dno/dt (C(oxG-F)~ C(oxF-M)) -""""(5)
•
' '

.(lO)
where k~ represents mass transfer coefficient of (O) in
the oxide film (cm s~ 1) and subscript F-M, oxide-metal
where [Al] represents mass concentration of Al in the
'

molten steel (masso/o). Since molten FeOis formed at

interface. WhenEq. (4) is applied to Eq. (5), the oxidation
the oxide-metal interface when the aluminum-killed
rate of the molten steel isgiven by
molten steel is oxidized, Al concentration at the
d[O]Ab,/dt= 100Mo'A'kF'(Pg•jG-F)- P0.5 )/(p V) oxide-metal interface, [Al](F_M), can be regarded as
o.(FM)
'

'
(6) being extremely low. If the oxidation reaction of the
.

aluminum-killed molten steel is controlled by the

where kF represents rate constant or k~'K' (mol- diffusion of A1 in the molten steel, therefore, d[O]Ab~/dt
Pa~0.5 .cm ~ 2 s~ 1). The partial pressure of 02 gas at the
.
varies with [Al] and does not depend
gas-oxide interface, P0=(GF), rs considered to be equal on Po,'
to the partial pressure of 02 gas in the gas phase, Po,' 4.2. Effect of Son Oxidation
Rate of Molten Steel
Furthermore, the partial pressure of 02 gas at the Whenthe gas absorption reaction of the molten steel
oxidemetal interface, Po.(F-M), is lower than the partial is controlled by chemical reaction step at the gas-metal
pressure of 02 gas in molten FeOin equilibrlum with interface, such as gas adsorption and desorption, Sin
the molten steel and is an extremely small value. If [O]Ab~ the molten steel is reported to reduce the reaction rate.7)
linearly increases with t and d[O]Ab~/dt varies with Pg•5, In this study O. I masso/o sulfur wasaddedto the nonkilled
=is
accordingly, the oxidation rate of the molten steel molten steel and the aluminum-killed molten steel, and
control]ed by the diffusion of oxygen in the oxide film. the effect of Son the oxidation rate of the molten steel

C 1996 ISIJ 392

 ISIJ International. Vol.
was investigated. The results are shown in Figs. 7 and
8. The nonkilled molten steel and the aluminum-killed
molten steel exhibited no drop in the oxidation rate with
the sulfur addition,regardless of whether or not they
were stirred. These results support that the oxidation rate
of the nonkilled molten steel and the aluminum-killed
molten steel is not controlled by an interfacial chemical
reaction process but by a diffusion process. It is thus
considered reasonable to exclude the above-mentioned
elementary steps b) and d) as steps that control the
oxidation rate of the molten steel.
4.3. Effect of Stirring on Oxidation Rate of Nonkilled
Molten Steel
Diffusion steps a), c) and e) may be considered as
steps that control the oxidation rate of the nonkilled
molten steel. Using the experimental results and reac-
tion rate models, which step controls the oxidation rate
of the nonkilled molten steel is studied as discussed
below.
In the oxidation process of the nonkilled molten steel,
the surface of the molten steel is not covered with oxides
that are visually observable, irrespective of whether or
not the molten steel is stirred. Whenthe oxidation rate
of the molten steel is controlied by step c), it varies with
the 1/2 power of the partial pressure of 02 gas. This
dependenceof the oxidation rate on the partial pressure
of 02 gas is not observed in the oxidation process of the
nonkilled molten steel in both the still state and the stirred
state. Whenthe oxidation rate of the molten steel is
controlled by step e), it does not depend on the partial
pressure of 02 gas and is constant. The oxidation rate
of the nonkilled molten steel is affected by the partial
pressure of 02 gas in both the stlll state and the stirred
state and does not meet the condition for rate-deter-
mining step e). As shown in Figs. 7
and 8, the oxida-
tion rate of the nonkilled molten steel is smaller than
that calculated by Eq. (8), or is controlled by step
e). Therefore, it is judged that steps c) and e)
can be
excluded as rate-determining steps for the oxidation
reaction of the nonkilled molten steel. The masstransfer
coefficient of Oin the molten steel in Eq. (8)
was the
value obtained from the results of N2 gas absorption
experiment in a 100 o/* N2 atmosphere by the application
of a model that assumesthat N2 gas absorption rate of
the molten steel is controlled by the combination of
chemical reaction at the gas-metal interface (the rate
constant of chemical reaction as reported by Harashima
et al.8)) and diffusion of N in the molten steel.
Whenthe oxidation rate of the nonkilled molten steel
is controlled by step a), the O
absorption linearly
increases with time, and the oxidation rate increases in
proportion to the first power of the partial pressure of
02 gas. These relations completely agree with the
oxidation behavior of the nonkilled molten steel in the
still state and the stirred state. Taking into account the
fact that neither of the above-mentioned steps c) and e)
becomethe rate-determining step, the oxidation rate of
the nonkilled molten steel is not influenced by stirring
and is controlled by step a), or diffusion of 02 gas in the
gas phase.
393
36 (1996). No. 4
4.4. Effect of Stirring on Oxidation Rate of Aluminum-
kil]ed Molten Steel
The oxidation rate of the aluminum-killed molten steel
is controlled by one of the steps a), c), e) and g). Which
is the rate-determining step for the aluminum-killed
molten steel is discussed below by using the experimental
results and reaction rate models.
If the oxidation rate of the molten steel is controlled
by step e), it does not depend on the partial pressure of
02 gas and is constant. The oxidation rate of the
aluminum-killed molten steel changes with the partial
pressure of 02 gas in both the still state and the stirred
state. As evident from Figs. 7 and 8, the oxldation rate
of the aluminum-killed molten steel is smaller than that
calculated by Eq. (8) or is controlled by step e). If the
oxidation rate of the molten steel is controlled by step
g), it should be infiuenced by Al concentration of the
molten steel, O absorption of the aluminum-killed molten
steel, be it stirred or not, Iinearly increases with time,
and Al concentration dependenceof the oxidation rate
is not recognized.
Whenthe oxidation rate is controlled
by step g), its values are calculated by Eq. (lO) and given
in Figs. 7 and 8. The measured oxidation rate of the
aluminum-killed molten steel is lower than that cal-
culated. Al concentration in the molten steel is put at
O.1 mass~/*. It is thus thought that the oxidation rate of
the aluminum-killed molten steel is controlled by a step
other than either step e) or g).
The surface of the aluminum-killed molten steel in the
still state is completely covered with an oxide film
composedof FeOand FeAl204. O is absorbed through
the oxide film. Furthermore, the diffusion of O
in the
oxide film is believed to occur not in the FeA1204phase
but mainly in the FeOphase9) where the interdiffusivity
of iron and oxygen is reported to be large. The oxidation
rate of the aluminum-killed molten steel in the still state
is proportional to the l/2 power of the partial pressure
of 02 gas and meets the condition applicable whenit is
controlled by step c). To makeclearer the effect of the
oxide filrn on the oxidation rate, an oxidation experiment
was conducted by causing waves in the molten steel
surface using the electromagnetic stirrer of the Tammann
furnace and breaking the oxide film. The results are
shownin Fig. 7. The breakage of the oxide film increased
the oxidation rate of the aluminum-killed molten steel,
and it becamethe same as the oxidation rate of the
nonkilled molten steel that is controlled by the diffusion
of 02 gas in the gas phase.
Manycracks were formed in the oxide film of the
aluminum-killed molten steel and expose a very large
area of the molten steel surface in the stirred state. O
is
mainly absorbed through the exposed portions created
by these cracks. The oxidation rate of the aluminum-
killed molten steel in the stirred state is proportional to
the first power of the partial pressure of 02 gas and
meets the condition applicable whenit is controlled by
step a). As shown in Fig. 8, the oxidation rate of the
aluminum-killed molten steel in the stirred state agrees
with the oxidation rate of the nonkilled molten steel that
is controlled by the diffusion of 02 gas in the gas phase.
C 1996 ISIJ
 IS]J International, Vol. 36 (1 996). No. 4
The above results show that stirring changes the ef- oxidation process of the aluminum-killed molten steel by
fect of the oxide film on the oxidation rate of the air. The oxidation rate of the alurninum-killed molten
aluminum-killed molten steel. The oxidation rate of the steel in the still state is controlled by the diffusion of
aluminum-killed molten steel in the stil] state is controlled oxygen in the oxide film, while the oxidation rate of the
by the diffusion of oxygen in the oxide film, while the aluminum-killed molten steel in the stirred state is
oxidation rate of the aluminum-killed molten steel in the controlled by the diffusion of 02 gas in the gas phase.
stirred state is controlled by the diffusion of 02 gas in
the gas phase. REFERENCES
l) H. Tanaka, R, Nishihara. I. Kitagawa and R. Tsujino:
5. Conclusions Telsu-lo-Hagan~, 79 (1993), 1254.
2) lron and Steel Handbook. Vol, l. Fundamentals, lOth ed.,
As
basic research to quantify the oxidation rate of Maruzen, Tokyo, (198 l), 446.
molten steel in tundish, oxidation experiments were 3) T. Emi, W. M. Boorstein and R. D. Pehlke: Meta!!. T,'a,Is..
5
performed on nonkilled molten steel and aluminum- (1974), 1959.
killed molten steel on laboratory scale, and the effect of 4) Y. Miyasita: Telsu-to-Hagan~, 52 (1966), 1049,
stirring on the oxidation rate was studied. The following
5) T. Cho and M. Inoue: Telsti-to-Haganb, 65 (1979), 1712.
6) S.Banya and D. Jae: Te!su-to-Haga,1~, 66 (1980), 1631.
conclusions were derived: 7) T. Choand M. Inoue: Tetsu-lo-Hagan~, 54 (1968), 19.
(1) The oxidation rate of the nonkilled molten steel 8) K. Harashima. S. Mizoguchi. M. Matsuo and A. Kiyose: ISJJ
by air is not affected by stirring and is controlled by the 32 (1992), II l.
Inl.,

diffusion of 02 gas in the gas phase. 9) K. Mori and K. Suzuki: Tc'!sti-to-Hagcln~, 54 (1968), 1199.
(2) Stirring changes the effect ofthe oxide film in the
(Originally published in Te!su-to-Hagan~, 81 (1995), I138, in Japanese)

C 1996 ISIJ 394