Process Metallurgy
Mould Powder Requirements for High-speed Casting
J.A. Kromhout1), 3), S. Melzer1), E.W. Zinngrebe1), A.A. Kamperman2) and R. Boom1), 3)
1)
Corus Research, Development and Technology, PO Box 10000, 1970 CA IJmuiden, The Netherlands
Corus Strip Products IJmuiden, Direct Sheet Plant, PO Box 10000, 1970 CA IJmuiden, The Netherlands
3)
Netherlands Institute for Metals Research, Delft University of Technology, PO Box 5008, 2628 AL Delft, The Netherlands
2)
Mould powders play an important role in the stability of the continuous casting process of steel. The main functions of mould slag (i.e.
molten powder) are to provide sufficient lubrication and to control the heat transfer between the developing steel shell and the mould.
Sufficient lubrication requires an undisturbed melting of mould powders and uniform infiltration of mould slag. Based on the casting practice
in IJmuiden, it is found that these demands become even more important for the applied high casting speeds in thin slab casting at 5 to 6
m/min. At Corus RD&T, mould powders were characterised by X-ray diffraction and subsequent fully quantitative Rietveld analysis.
Additionally, the melting of mould powders has been studied in-situ using high-temperature X-ray diffraction, to gain crucial knowledge
about melting relations. Slag rims obtained from the thin slab caster mould were characterised using extended microscopic techniques in
order to describe the mechanisms of rim formation and growth. Finally, slag films obtained after casting were characterised. As a result, not
only the melting process of mould powder, but also the mechanism of formation and growth of slag rims is much better understood. This
knowledge will be applied to define the demands on the composition and properties of mould powder for even higher casting speeds.
Keywords: thin slab casting; mould powder
DOI: 10.2374/SRI07SP069-79-2008-143-148; submitted on 7 June 2007, accepted on 27 July 2007.
Introduction
Mould powders play an important role in the stability of
the continuous casting process of steel at all casting speeds.
The main functions of mould slag (i.e. molten powder) are
to provide sufficient lubrication and to control the heat
transfer in horizontal direction between the developing
steel shell and the copper mould.
Several authors have reviewed the composition,
properties and operational experience of mould powders
for conventional slab casting [1-2]. However, only a few
publications have been issued on the design and use of
mould powders for thin slab casting [3-4]. Compared with
conventional slab casting, process control in thin slab
casting is more critical and it can be assumed that mould
powder demands for thin slab casting become more critical
as well [5-6]. Thus a more extensive understanding of
mould powder properties and functions is needed to meet
the increasing demands of high-speed thin slab casting.
Nevertheless, a lot of improvements (mould powder
developments) are based on trial and error and are
focussing on casting speeds up to 2 m/min (conventional
slab casting). A general approach is to relate the chemical
composition of a mould powder to the operational
behaviour (in-mould behaviour) during casting. In doing
this, some physical properties of the mould slag like the
viscosity and the melting point are addressed as well.
However, this approach will not automatically result in a
suitable mould powder or an in-depth knowledge on
powder design and required slag properties.
At Corus IJmuiden, a project started in the framework of
the Netherlands Institute for Metals Research with the aim
to develop mould powders suitable for high-speed thin
slab casting with a maximum casting speed up to 8 m/min.
For this work, a fundamental understanding and
quantification of the melting and solidification behaviour
steel research int. 79 (2008) No. 2
of mould slag as well as the functions of mould powder is
needed.
Thin Slab Casting at Corus IJmuiden
Corus IJmuiden introduced the thin slab casting – direct
rolling technology in 2000, in parallel with the existing
conventional slab casting – hot strip mill route. The role of
the Direct Sheet Plant (DSP) is to turn part of the liquid
steel produced in BOS No. 2 into an annual 1.3 Mton of
high valued, thin-gauged products. The DSP has been in
operation for nearly 6 years and has reached a stable
production level. The product range is still developing and
is well accepted in the market.
The casting process is one of the many important links
in the production chain of the DSP. The liquid steel,
produced in the adjacent BOS No. 2 in 320 ton batches
and treated in the ladle furnace, is cast into thin slabs. The
caster has one strand and is equipped with several features
like a funnel shaped mould and an adjustable multiple pole
ElectroMagnetic Brake (EMBr) to enhance quality level
and casting speeds up to 6 m/min. Liquid core reduction
decreases the slab thickness from 90 mm to 70 mm. A
summary of the main specifications of the DSP is listed in
Table 1; an illustration of the mould and SEN is given in
Figure 1.
Over the last years, several technological developments
were implemented in order to support the casting
operations at the DSP [7]. The operational performance of
the DSP is approaching the design level of 1.3 Mton/year
(coils). The HSLA part of production is now over 30%.
Today, the maximum operational casting speed is 5.7
m/min; the average value is over 5.0 m/min. For the HSLA
steel grades, the same data apply and the same mould
powder is used. The maximal sequence length is ten ladles
143
Process Metallurgy
fluid flow, mould level control, mould heat transfer and
secondary cooling. Special attention has been given to the
design of mould powders for high-speed casting.
Mould Powder Developments for High-speed Casting
Figure 1. DSP mould and SEN.
Table 1. Main specifications of Direct Sheet Plant (DSP) at Corus.
Steel grades
Casting speed (max.)
Mould/slab thickness
Strip thickness
Strip width
Capacity
low carbon, HSLA
6.0 m/min
90/70 mm
0.7-2.5 mm
1000-1560 mm
1.3 Mton/year (coils)
Table 2. Chemical composition (wt%) of mould powders for thin
slab casting (DSP), as measured at Corus IJmuiden.
Mould powder
constituents
I
Powder type
II
III
CaO/SiO2
SiO2
CaO
MgO
Al2O3
Fe2O3
Na2O
K2O
Li2O
F
C free
CO2
LOI (1000°C)
1.0
33.1
33.4
0.5
2.8
0.4
11.8
0.4
0.0
7.9
4.0
6.8
12.8
1.0
32.4
32.5
2.3
3.9
0.3
10.7
0.1
0.4
7.7
4.8
5.4
13.2
1.0
33.9
35.5
3.9
0.4
0.8
8.9
0.1
0.0
9.1
3.9
4.9
10.8
(12 hours casting). Trials are performed, applying a
maximum casting speed of 6.0 m/min.
In 2004, it was decided to increase the production of the
DSP from the current level of 1.3 Mton/year to 1.8
Mton/year (coils) in 2010 with one caster strand. To meet
this demand, the steel-in-mould time has to be increased
from the current 72% to approx. 85% and the maximum
casting speed will be increased from 6.0 m/min to 8.0 m/min.
A project was started to develop and implement the
essential technology for this strategic target. The main
technological developments have been focussing on mould
144
A project started with the aim to develop mould powders
suitable for high-speed thin slab casting. For this work, a
fundamental understanding and quantification of the
melting and solidification behaviour of mould slag and the
functions of mould powder has to be a major requirement.
Starting point is the definition of windows of physical
properties for powders, acting under this condition. The
physical properties of mould powders are directly related
to the chemical and mineralogical composition i.e. the raw
material choice of a mould powder. Thus, the work will
concentrate on the translation of process data from the
caster to physical properties and to the chemical and
mineralogical requirements.
Previous work on mould powders resulted in the
definition of windows of their physical properties for
conventional slab and thin slab casting. Based on an
inventory on mould powders reported in 2000 and
compared with conventional slab casting, the values for
viscosity for thin slab casting mould powders are lower,
indicating the importance of slag infiltration, i.e. an adequate
and constant lubrication and mould heat transfer [8].
Recently, the mould slag formation during thin slab
casting has been investigated and related to the organic
components of the mould powders, i.e. the choice,
distribution and amount of free carbon within the granules
[9]. This work revealed that the free carbon source and an
even distribution within the granules are essential to
guarantee a stable slag formation during casting. This can
be obtained by using one grade of fine carbon black,
possibly supplemented with one grade of fine coke.
Another important conclusion of this work is that the
physical properties of mould powders are not as critical as
expected, when the additional mineralogical requirements
are taken into account.
The work builds on earlier investigations on free carbon
in mould powder [10-11]. However, the approach at Corus
IJmuiden is to relate the slag formation and additional
caster data to the physical properties of mould slag, results
of chemical analyses and mineralogical investigations [9].
It is clear that this approach will be used again in the
project regarding the mould powder developments for
high-speed thin slab casting. The mineralogical
investigations will concentrate on all raw materials and not
only on the free carbon sources.
Operational Experiences
During the first year of DSP operation, several mould
powders were used with both high and low basicity
(CaO/SiO2). After approximately a year, a low-basicity
mould powder (CaO/SiO2 = 1.0) was selected as the
standard and with this powder, the casting speed and
sequence length has successfully been increased over the
years to the current level. In Table 2 an overview is given
of the chemical composition of mould powders.
steel research int. 79 (2008) No. 2
Process Metallurgy
Mould powder I has been the standard mould powder of
the DSP for approx. 5 years. Mould powder II is an
alternative for this powder. Mould powder III is a
development of mould powder I with the aim to improve
(i.e. decrease) the friction.
The performance of mould powder is evaluated using
several operational criteria like the slag formation (liquid
pool depth and rim formation), uniformity and ratio of
mould heat transfer, friction of the strand, meniscus
stability and scale formation in the tunnel furnace.
The liquid pool depth is measured by immersing a steel
strip (carbon steel) or a combination of a stainless steel
strip and a copper strip into the mould. The combination of
copper and stainless steel strips is used when the EMBr is
operating, i.e. in the presence of a magnetic field. At the
meniscus, the steel melts and the mould slag solidifies and
adheres on the steel sheet. Compared with carbon steel, the
adherence on the stainless steel strip is less. The
combination with the copper strip – which melts at the slag
surface – indicates the liquid pool depth when operating
the EMBr. The operators observe rim formation n. Mould
heat transfer is calculated using mould cooling water
temperatures and water flow.
Strand friction data are obtained by the hydraulic
oscillating system of the DSP. The meniscus stability is
measured using a radiometric system. Scale formation is
observed by visual inspection of the rolls in the tunnel
furnace.
Mould powder I is the standard mould powder. This
powder shows a very constant behaviour, which can be
related to the constant chemical (and raw material)
composition. A disadvantage of powder I is the formation
of some rims during casting. These rims are strong and do
not melt easily. However, excessive rim formation during
normal casting practice has not been reported. Mould
powder II showed less rim formation and a comparable
friction as powder I. This powder resulted in excessive
scale formation (wustite; FeOx) in the first meters of the
tunnel furnace of the DSP. Mould powder III showed
similar in-mould behaviour as mould powder II (less rim
formation) but no scale formation. The excessive scale
formation with mould powder II can be related to the
interaction between the specific mould slag and the steel
surface. This is being investigated.
During normal casting practice, the liquid pool depth is
below 6 mm (all mould powders). The current work
concentrates on the melting of mould powder and the
formation of slag rims and slag films (solidification).
Mould Powder Characterisation
The chemical composition as given in Table 2 does not
give any information on the mineralogical composition
(raw material choice) of the mould powders. The
mineralogy is a very important theme in the present work.
Mineralogical analyses were done using X-ray diffraction
(XRD) techniques. Additionally, optical microscopy and
SEM/EDS were applied.
A starting point was the determination of some physical
properties; the viscosity and the melting trajectory. Results
are given in Table 3.
steel research int. 79 (2008) No. 2
Table 3. Physical properties of mould powders, as measured at
Corus IJmuiden.
Mould powder properties
Viscosity at 1300°C (Pa s)
Crystallisation point
Viscosity at 1300°C (Pa s)*
Softening point (°C)
Melting point (°C)
Fluidity point (°C)
* measured by the supplier
I
0.14
1140
0.13*
Powder type
II
0.19
1110
-
III
0.07*
1119
1120
1031
1055
1077
1106
1118
1131
Table 4. Mineralogical composition (wt%), as measured at Corus
IJmuiden (Rietveld-XRD, errors range between 20% relative for
minor phases (<20 wt%) and 10% relative for major phases (>20
wt%)).
Mould powder composition
I
Powder type
II
III
50.4
1.9
0.0
2.7
13.8
0.0
54.8
5.1
0.0
2.3
0.0
5.6
48.9
2.0
5.0
2.5
0.0
0.0
17.0
0.0
6.9
4.1
20.0
0.0
12.7
1.5
14.9
4.8
19.0
1.8
0.0
1.4
0.0
Silicates:
wollastonite (CaSiO3)
quartz (SiO2)
forsterite (Mg2SiO4)
diopside (CaMgSi2O6)
feldspar ((Ca,Na)(Al,Si)4O8)
spodumene (LiAl(SiO3)2)
Fluorites:
fluorite (CaF2)
cryolite (Na3AlF6)
Carbonates:
natrite (Na2CO3)
calcite (CaCO3)
Others:
periclase (MgO)
The viscosity (rotating cylinder method) of powder I and
II shows similar behaviour. The melting temperatures (hot
stage microscope) of powder II are lower than for I and III.
XRD with subsequent quantitative Rietveld analysis
revealed the quantitative composition of the crystalline
mould powder constituents i.e. the raw material choice of
the three mould powders. Amorphous components cannot
be recognized with XRD and do not figure in the analyses.
Results are given in Table 4.
For each mould powder, significant differences in
composition can be observed. In general, the mineralogical
composition can be divided into three groups: the silicates,
the fluorites and the carbonates. It can be assumed that
there will be no chemical equilibrium between the raw
materials. Consequently, the composition will change
continuously during temperature increase until a mould
powder melts.
The change in mould powder composition during
heating was studied in-situ using high-temperature XRD
(HT-XRD). The analyses showed a sequence of disappearance of primary minerals (raw materials) and
appearance of secondary phases, before melting takes
place. A few minerals dominate this process (cuspidine,
combeite). An example of raw data is shown in Figure 2.
A summary of the results of the analyses is given in
Table 5.
145
Process Metallurgy
7
6
4
6
5
4
3
2
1
Figure 2. High-temperature XRD analysis (mould powder I)
showing the raw materials natrite (1), feldspar (2), quartz (3),
wollastonite (4), fluorite (5) and the secondary phases combeite (6)
and cuspidine (7).
Table 5. Melting sequence (ºC), as measured at Corus IJmuiden
(HT-XRD)
Primary minerals
phases
Calcite-out
Natrite-out
Feldspar-out
Cryolite-out
Spodumene-out
Forsterite-out
Quartz-out
Fluorite-out
Wollastonite-out
Melting
I
n.d.
750
850
absent
absent
absent
900
650
800
1050
1025
1100
1200
Powder type
II
650
n.d.
absent
800
850
absent
950
675
850
1025
1075
1100
1200
Secondary
III
n.d.
750
absent
absent
absent
850
875
725
925
1125
1050
1125
1225
Combeite-in
Cuspidine-in
Combeite-out
Cuspidine-out
n.d.: not detected
First, the carbonates disappear; next some silicates
including feldspar (mould powder I), spodumene (mould
powder II) and forsterite (mould powder III). Then the
intermediate (secondary) phases are formed. Finally the
frequently used constituents fluorite and wollastonite
disappear and the mould powder melts. Based on the
composition, each mould powder shows a specific
sequence of phase transformations and reactions before it
melts.
In previous work, the phase composition of mould
powders has been studied as a function of temperature [12].
The presence of secondary phases like cuspidine and
combeite has been demonstrated. Their occurrence is
related to annealing temperature and the powder
composition. However, this work is based on (batch)
experiments at fixed temperatures; powder samples were
annealed (sometimes for hours) and quenched, before
XRD-analyses were performed.
It is generally known that cuspidine (3CaO·2SiO2·CaF2)
is predominantly formed as a crystal phase during
solidification of mould slag. The presence of cuspidine
plays an important role in the control of heat transfer
during casting [1]. Recently, the formation and stability of
cuspidine was thoroughly investigated and a phase
diagram was defined [13-14]. High-temperature reactions
of mould powder are used in order to describe the
146
formation of cuspidine, i.e. to control mould heat transfer
during casting. This work includes the effect of Na2O on
the crystallisation behaviour of cuspidine [15-16].
The information obtained (XRD-methods) will be used
in investigations on rim formation as well as in the design
of mould powders.
Additional mineralogical analyses showed the presence
of fine particles of carbon black as free carbon source in
mould powder I and III. This carbon source is well
distributed within the granules, comparable with previous
work [9]. Powder II has two free carbon sources; large
amounts of fine particles of carbon black and coarse
particles of coke. The cover of the carbon black particles
is much more pronounced in powder II than in powder I
and III.
Characterisation of Slag Rims
During continuous casting, slag rims are formed. Rims
are composed of mould powder, mould slag and sintered
products (secondary phases) and adhere to the mould walls.
Rims can melt, providing some mould slag. Usually, slag
rims are small (nearly visible during casting) and play a
role in the infiltration of mould slag. However, rims can
grow and disturb the casting process [1, 17-18].
Based on our experience, excessive rim formation is
undesired because the slag infiltration can be hindered,
parts of the rim can break and the rim can form a bridge,
especially at the narrow faces or between the SEN and the
copper mould plate. Serious disturbances and even
breakouts can occur. Due to the mould dimensions (thin
slab caster) and the high casting speed, it is clear that the
control of rim formation is very important.
The formation of rims can be related to the meniscus
behaviour during casting, the mould heat transfer near the
meniscus area and the mould powder composition and
properties [19]. Previous work at Corus revealed the role
of free carbon (melting behaviour) on slag formation,
including the formation of rims [9]. The present work aims
at the influence of the inorganic mould powder
constituents on the formation and growth of slag rims.
Rims obtained at the DSP during normal casting practice
were prepared for characterisation for both mould powder
I and II. Characterisation was done using optical
microscopy and SEM/EDS.
In both cases, the rims are initially formed by a
“painting” mechanism. Thin layers of mould slag and
mould powder adhere to the (cold) copper walls. However,
the mechanism of growth is not the same.
Rims corresponding to mould powder I consist of a
layered structure of mould slag, mould powder and several
intermediate products (secondary phases) together with
some small slag and Fe droplets and particles of carbon
black. Thin crystals of a condensed Na-compound coat
and cement these layers. As a result, the rims are dense
and strong. The growth of the rims can also be related to a
“painting” mechanism. The rims are formed during a
longer period and will not melt easily.
Rims from powder II show a loose structure, containing
a lot of small slag droplets, particles of coke and a
condensed tar binder. The slag droplets also contain
steel research int. 79 (2008) No. 2
Process Metallurgy
secondary phases, comparable with the melting sequence
(Table 5). The droplets are “sprayed” and quenched and
this “spraying” mechanism causes the growth of the slag
rim. Comparing with rims from mould powder I, these
rims are not only loose, but are formed in a shorter period
and can be melted more easily.
The operational experience (rim formation) with
powders I and II corresponds to a large extent with the
results of the mineralogical analyses.
An excess of sodium, especially at lower temperatures,
possibly enhances the formation of dense rims in powder I.
Mould powder II and III do not show significant rim
formation during casting. Mould powder II uses
spodumene and periclase; in mould powder III forsterite is
used whereas mould powder I contains feldspar and larger
amounts of natrite. The carbon and tar may originate from
the abundant free carbon in powder II. The mechanism of
spraying (slag droplets) is not fully understood. This can
possibly be related to the amount of carbonates, i.e. the gas
formation (CO2) during casting.
Rims from mould powder I show small droplets of Fe.
Up to now, the presence of these droplets is not fully
understood. Simulation experiments (laboratory scale) on
mould powder melting revealed small metallic droplets in
the slag, originating from the steel [20]. The work suggests
that reactions at the steel-slag interface cause these small
droplets. Other work on steel slag interfacial phenomena
revealed the presence of small droplets of metal,
distributed in the mould slag. The droplets are formed via
a reduction of metal oxides by carbon. The metal oxides
originate from interfacial reactions between slag and liquid
steel and from mould powder raw materials. The reduction
occurs in the mould slag and at the interface between the
slag and the sinter layer [21].
Characterisation of Slag Films
During casting, mould slag infiltrates into the channel
between the mould and the strand forming a thin slag
film. The slag film is moving downwards and solidifies
into glassy and crystalline structures. This process is very
complex. It is important to consider that the slag film
realizes the main mould powder functions.
Characterisation of slag films will help in understanding of
the mould powder functions and will guide mould powder
design.
Pieces of slag films were collected by means of a steel
plate, which is constructed between the mould and the first
segment of the caster. Due to the local strong watercooling, the slag film pieces are quenched after leaving the
mould, possibly crumbled and distributed under the mould
area. However, a lot of small pieces were collected and
several samples were prepared for characterisation.
Microscopic analysis was done using optical microscopy
and SEM/EDS-techniques. All the pieces are related to
mould powder I.
Most slag films show a face which has been in contact
with either Fe-metal (strand) or Cu-metal (mould). The
side opposing the flat face consists of curved, wavy
surfaces, apparently derived from breaking along a line of
bubbles (vesicles). Thus, the observed film pieces are
mainly fragments of the slag film; the thickness of the
steel research int. 79 (2008) No. 2
fragments varies between 50 µm and 500 µm. Oscillation
marks were not visible. Films in contact with the strand
(Fe) show large and fine cuspidine crystals
(3CaO·2SiO2·CaF2) in an Al2O3 and SiO2 enriched glass
matrix. FeOx-scales are sometimes attached to these films.
Films in contact with the mould (Cu) show predominantly
fine quenched crystals (cuspidine) in a glass matrix,
enriched in Na2O and F.
2
3
1
_________________
25 µm
Figure 3. Slag film fragment showing glass (1) and cuspidine
crystals (2) in contact with FeOx (3 - strand side).
Furthermore, the films show small droplets of Fe
(comparable with those observed in the slag rims),
fragments of Cu, some mould powder granules and even
particles of carbon black, probably originating from the
mould powder. An illustration of a slag film fragment is
given in Figure 3.
As mentioned above, a gradient in Fluorine (F) and
Sodium (Na) components between the strand- and mould
side is observed; the mould side showing the highest Fand Na-concentrations. Based on the local composition, it
is proposed that the bubbles are probably formed through
evaporation of NaF components in the melt. It is assumed
that the properties and functions of the slag film will be
influenced by the changes in slag film composition and
structure, including the presence of bubbles. Most
fragments are split off and the slag film fragments are
quite thin.
The observations partly confirm results obtained at
AvestaPolarit where a change in slag film structure and
composition as a function of casting time and location was
found as well as a presence of sub-layers, pores and gas
bubbles [22]. Slag films obtained at conventional slab
casting showed the presence of pores, glassy and
crystalline phases and also metal droplets. The thickness
of the slag films was generally between 1.5 and 2.5 mm
[23]. This comes up to expectations [24].
Discussion and Progress
Future work on HT-XRD will concentrate on the
solidification of mould slag in order to increase knowledge
147
Process Metallurgy
on the mould powder functions and the role of the slag
composition (phase relations).
Based on the slag consumption, the average liquid film
thickness can be calculated as:
dl = Qs / ρ ≈ Qs / 2600
(1)
where dl is the average thickness of the liquid film, Qs is
the slag consumption (kg/m2) and ρ the mould slag density
[25]. For high-speed casting, a mould powder consumption of less than 0.1 kg/m2 at casting speeds up to 8
m/min has been reported [26]. Assuming average slag
consumption at the thin slab caster of 0.1 kg/m2 at most
and a slag density of 2600 kg/m3, the liquid film thickness
is approx. 40 µm. This thickness represents a fraction of
the total film thickness. As a rule of thumb, the average
liquid film thickness can be considered to be at least a
tenth of the total film thickness. This indicates an average
total film thickness of approx. 400 µm or more. Although
this is a rough estimate, it gives an indication (and some
confirmation) of the film thickness.
As a next step, slag films obtained from inside the
mould will be investigated in order to get more knowledge
on the film thickness, the chemical and mineralogical
composition including the crystallisation, the presence of
metallic droplets and bubbles etc.
Recent trials with mould powders developed from
powder III (without feldspar) did not show significant rim
formation. This seems to confirm the suggestion on the
role of sodium during rim formation and rim growth.
Conclusions
During heating, a mould powder shows a specific
sequence of phase transformations before melting takes
place.
Rim formation can be related to the mineralogical
composition of the mould powder. At least two
mechanisms of rim growth can be defined where sodium
probably enhances rim formation.
Slag films are mainly composed of cuspidine crystals,
glass phases and a line of bubbles. Most films break along
this line of bubbles. Furthermore, the films show a
gradient in fluorine and sodium.
In addition to conventional characterisation methods,
mineralogical characterisation is essential for a further
understanding of mould powder behaviour and mould
powder functions.
Acknowledgements
The authors would like to thank F.H.W. Grimberg for
assistance during the casting trials and F.J.L. van der Does
and J.J. Koster for support during the characterisation of
mould powder and mould slag. This paper is based on a
presentation at the 2006 International Symposium on Thin
Slab Casting and Rolling, TSCR2006, held in Guangzhou,
China on 11-13 April 2006.
148
References
[1] K.C. Mills, A.B. Fox, R.P. Thackray and Z. Li: The performance and
properties of mould fluxes. Proc. 7th Int. Conf. on Molten Slags,
Fluxes and Salts, Cape Town, 2004, The South African Institute of
Mining and Metallurgy, p. 713-721.
[2] K. Koyama, K. Nagano, Y. Nagano, T. Nakano: Design for chemical
and physical properties of continuous casting powders. Nippon Steel
Tech. Rep., 34 (1987), 41-47.
[3] F. Neumann, J. Neal, M.A. Pedroza, A.H. Castillejos, F.A. Acosta:
Mold fluxes in high speed thin slab casting. Proc. 79th Steelmaking
Conf., Pittsburgh, 1996, Iron and Steel Society, p. 249-257.
[4] D. Blevins, M. Ingold, A, Schaefer, J. Neal, F, Neumann, C. Sowa:
Ironmaking Steelmaking, 27 (2000), No. 3, 85-88.
[5] M. Kawamoto, K. Nakajima, T. Kanazawa, K. Nakai: ISIJ Int., 34
(1994), No. 7, 593-598.
[6] P. Tang, G. Wen, Z. Wang, S. Fan, Z. Zhang: Investigation on mold
powder of low carbon steel for CSP thin slab continuous casting.
Proc. ISSTech 2003, Indianapolis, Iron & Steel Society, p. 567-573.
[7] M.C.M Cornelissen, J.A. Kromhout, A.A. Kamperman, M. Kick and
F. Mensonides: High productivity and technological developments at
the Corus DSP thin slab caster. Proc. 5th Eur. Continuous Casting
Conf., Vol. 1, Nice, 2005, p. 322-328.
[8] J.A. Kromhout, V. Ludlow, S. McKay, A.S. Normanton, M. Thalhammer, F. Ors, T. Cimarelli: Ironmaking and Steelmaking, 29
(2002), No. 3, 191-193.
[9] J.A. Kromhout, A.A. Kamperman, M. Kick, J. Trouw: Ironmaking
and Steelmaking, 32 (2005), No. 2, 127-132.
[10] B. Xie, J. Wu and Y. Gan: Study on amount and scheme of carbon
mixed in CC mold fluxes. Proc. 74th Steelmaking Conf.,
Washington, 1991, Iron and Steel Society, p. 647-651.
[11] J-W. Kim, S-K. Kim, Y-D Lee: Effects of carbon particle size and
content on the melting rate of mould powder. Proc. 4th Eur.
Continuous Casting Conf., Vol.1, Birmingham, 2002, IOM
Communications Ltd, p. 371-377.
[12] P. Grieveson, S. Bagha, N. Machingawuta, K. Lidell and K.C. Mills:
Ironmaking and Steelmaking, 15 (1988), No. 4, 181-186.
[13] T. Watanabe, H. Fukuyama, K. Nagata: ISIJ Int., 42 (2002), No. 5,
489-497.
[14] H. Fukuyama, H. Tabata, T. Oshima, K. Nagata: ISIJ Int., 44 (2004),
No. 9, 1488-1493.
[15] M. Hanao, M. Kawamoto, T. Watanabe: ISIJ Int., 44 (2004), No. 5,
827-835.
[16] M. Hayashi, T. Watanabe, H. Nakada, K. Nagata: ISIJ Int., 46
(2006), No. 12, 1805-1809.
[17] M. Emi: The mechanisms for sticking type break-outs and new
developments in continuous casting mold fluxes. Proc. 74th
Steelmaking Conf., Washington, 1991, Iron and Steel Society, p.
623-630.
[18] C. Perrot, J.N. Pontoire, C. Marchionni, M.R. Ridolfi, L.F. Sancho:
Several slag rims and lubrication behaviours in slab casting. Proc.
5th Eur. Continuous Casting Conf., Vol. 1, Nice, 2005, p. 36-46.
[19] R.B. Mahapatra, J.K. Brimacombe, I.V. Samarasekera:. Metall.
Mater. Trans. B, 22B (1991), 875-888.
[20] M. Görnerup, M. Hayashi, C.-Å Dåcker, S. Seetharaman: Mold
fluxes in continuous casting of steel – characterization and
performance tuning. Proc. 7th Int. Conf. on Molten Slags, Fluxes and
Salts, Cape Town, 2004, The South African Institute of Mining and
Metallurgy; p. 745-752.
[21] P.R. Scheller: Redox reactions in slags during continuous casting of
stainless steel. Proc. Mills Symposium, Vol. 2, London, 2002, NPL
London, p. 487-493.
[22] P.O. Hooli: Study of the mould flux film between the mould and
steel shell. Proc. 4th Eur. Continuous Casting Conf., Volume 1,
Birmingham, 2002, IOM Communications Ltd, p. 379-386.
[23] V. Ludlow, B. Harris, S. Riaz and A.S. Normanton: Ironmaking and
Steelmaking, 32 (2005), No. 2, 120-126.
[24] K.C. Mills: A review of ECSC-funded research on mould powders.
Synthesis report EUR 13177 EN, 1991.
[25] R. Saraswat, A.B. Fox, K.C. Mills, P.D. Lee and B. Deo: Scand. J.
Metallurgy, 33 (2004), 85-91.
[26] T. Kanazawa and M. Kawamoto: Latest technology for QSP process.
Proc. Int. Conf. on Continuous Casting – Past, Present & Future,
Jamshedpur, 2005, The Indian Institute of Metals and TATA Steel
Ltd., p. 227-230.
steel research int. 79 (2008) No. 2