KEMPROCAST TUNDISH POWDERS

PAN: Insulating covering powders

BASIC: High reactivity towards steel inclusion tundish basic powders

Powder functions and properties

A tundish covering powder is placed on top of molten steel in the tundish.

In general, the tundish covering powder should have the following functions:

1. act as a thermal insulator

2. act as a barrier between air and liquid steel to prevent reoxidation
3. assimilate the inclusions that separate from
the steel during casting

Tundish powders are mainly divided in two

categories: acid and basic. The main component of
acid powders is SiO2 while lime (CaO) is the main
components of basic powders. When using an acid
covering powder the steel cleanness can be
weakened by oxygen pick-up reoxidation from the
covering powder because the oxygen potential of an acid slag is by far higher than that of the
steel.

For high grade steels basic tundish covering powder is preferred. It can get mixed without
great problem with the rest of the basic ladle slag. Basic covering powders in the tundish are
also beneficial for the lining because they do not readily react with basic refractories.
Tundish covering powders with high basicity still have problems when casting long sequences
without changing tundish because the covering layer becomes hard in a few hours. This creates
big problems when the steel level in tundish varies, for instance when changing ladle, or when
blocking the stopper rod, as well as causing problems in sampling and measuring temperature
in the tundish.

Usually, basic powders are used together with acid powders (the so-called double covering
layer) in order to fulfill the three functions listed above as follows:

At first, the basic (low melting) powder is put on the steel forming a liquid layer with
high inclusion adsorption capability. It forms an efficient barrier against reoxidation but
does not act as an efficient thermal insulator
Then an acid powder is added (forming the top powdery layer), making an excellent
thermal insulation but neither capable against reoxidation nor participating in the
inclusion absorption.

Both in case of one powder (acid) or two powders (basic + acid), the total covering layer
become a complex multi-layer system.
 Along the layer thickness there is a variation of physical status, chemical composition and
structure of the oxide system.

During the service a continuous variation of this system occurs, with consequent
degradation of performance: loss of thermal insulating ability and other desirable abilities like
inclusion adsorption and low oxidising tendency.

The degradation is caused by dissolution of refractory materials (stopper rod, ladle shroud
and tundish lining), inclusion capture and, over all, ladle slag carry over. This is in many cases a
limit for the casting sequence for several steel grades.

Thus, in many cases the performance of the covering powder is a bottle neck for
steelmaking productivity and a problem for steel quality. Nevertheless, a complete physical and
chemical knowledge of the mechanisms acting and controlling the properties of the covering
system has not yet been achieved.

State of the art

Considering that the primary function of the tundish cover powders is threefold:

1. thermal insulation to minimize heat loss from the steel surface

2. prevention of reoxidation of the steel by the atmosphere, and
3. assimilation of non-metallic inclusions

The tundish powder performance during use depends on melting and solidification
temperature, viscosity, thermal insulation, refractory compatibility and oxide absorption
capability.

Thermal Insulation and re-oxidation prevention

In tundish, provision of thermal insulation and prevention of re-oxidation is obtained by:

1. retaining a reasonably thick and well dispersed powder layer covering the molten slag
layer for thermal insulation;
2. developing and retaining a reasonably thick slag barrier between molten steel and the
external oxidising atmosphere.
Thermal insulation

It has been reported that the cover powder

could lose its thermal insulation ability and allow the
slag to crust because of temperature and composition
gradient through the covering layer. Different theories
of crust formation have been proposed. For example,
according to one theory, firstly the crust forms at the
top surface of the liquid slag layer and then it grows
upwards to the powdery layer because of the thermal
gradient.

The fairly fluid slag can wet the grains of the tundish cover powder, causing them to
agglomerate. Along the thickness of the tundish powder layer there will be a point where the
powdery layer, on its point of fusion, is in contact with the slag on its point of solidification,
hence creating a rigid layer. This rigid raft on the steel surface causes problems for taking
tundish samples as well as to the movement of the ladle shroud. On the chemical side, it is
observed practice that increasing C level in Alumino-Silicate slags changes the heavy rigid
nature of the crust to the highly viscous but plastically deformable layer but the exact reason
for this behaviour is not known.

At present, the exact mechanism of crust formation has not been quantified as well as
the effect of external oxidation, cast sequence length, ladle slag carryover and internal
chemical reaction.

In the past, acid powders, i.e. rice husks rich in SiO2 (85-90%), were used as an excellent
thermal insulator but it was found that they are not well suited for inclusion absorption.
Moreover, in order to prevent steel re-oxidation, some acid tundish fluxes also contain Carbon,
this could cause a problem of C-pick up by steel. Similarly H pick up by steel is also reported
when using acid fluxes.

Basic powders enriched in CaO and/or MgO for improving thermal insulating ability
have been used, but these powders are less efficient to control steel re-oxidation.

It has also been reported that the insulation properties of basic powders can be
improved by changing their chemistry. For example, it is shown that the addition of TiO 2 to
CaO-Al2O3-SiO2 powders reduces the radiative component of heat transfer, which provides a
means of improving insulation in the tundish slag and decreasing the possibility of crusting.
However, using TiO2-including tundish powders could be problematic when casting some
special steel grades more sensitive to carbide formation.
Reoxidation prevention

One of the main functions of tundish powders is to prevent the reoxidation of the steel melt.
This reoxidation is due to:

1. the exposure of the liquid steel to the atmosphere particularly in the turbulent area
around the ladle shroud (extrinsic oxidation)
2. chemical reactions inside the melt (chemical or intrinsic oxidation)

The prevention of extrinsic reoxidation of the steel requires a continuous impermeable

barrier between the steel and the atmosphere. Normally, powders with fusion temperature
lower than the temperature of the liquid steel are fed when the tundish is only partially filled in
order to limit the exposure time of the molten steel surface to the oxidising atmosphere.

The steel reoxidation in tundish is mainly at the boil area around the ladle shroud; this re-
oxidation is responsible for powder chemistry changes.

An inappropriate slag chemistry, for instance with high levels of reducible oxides (such as
FeO, MnO and SiO2) becomes a source of oxygen through redox reactions and oxygen
transport. Hence tundish powder chemistry must be adjusted to maintain low levels of
reducible oxides. For example, in non-desulphurised Al-killed steel, reoxidation products (FeO,
MnO, SiO2 etc.) are present in addition to Al2O3. In this case the slag becomes a source of
reoxidation through its reaction with the dissolved aluminium in the steel and also by acting in
a transport mechanism for moving oxygen from air through highly oxidised slag. There is a co-
relation between FeO+MnO contents of the tundish slag and the steel total oxygen contents.

It is also possible that redox reactions (as shown below) occur at the steel surface which
may transfer silicon and manganese to the steel melt hence increases the Si and Mn to an
undesired level:

2Mn + 3SiO2 2MnO.SiO2 + Si

4Al + (3+x) SiO2 2Al2O3.xSiO2 + 3Si

It is also recognized that the change in composition of tundish slag depends on the steel
chemistry in the ladle before tapping to tundish.
Inclusion absorption

An important role of avoiding contamination of

the steel and facilitating further separation of
inclusions is given to the tundish. The idea of
increasing retention time for inclusion floatation in
the tundish has been pursued by gradually increasing
the tundish capacity.

In general, non-metallic inclusions will be

removed from the liquid steel if they come in contact
with the free surface (the interface between liquid
steel and tundish cover powder). The physical,
thermophysical properties of the interface are very important in assimilating the floating
inclusions. For example, the viscosity of tundish flux at the interface plays an important part in
absorbing inclusions. Too high a viscosity (mainly acid powders) limits inclusion absorption
capability while too low a viscosity (mainly basic powders) can lead to entrainment of the flux
into the steel during unstable conditions, particularly if the metal level drops too low causing
vortexing. It is also noted that Al2O3 addition increases the viscosity of the flux so initially high
Al2O3fluxes should not be used. Since the prime requirement for absorbing most of the
deoxidisation and reoxidation products is the presence of liquid slag, it is recommended the
use of low melting fluid basic fluxes (sometime termed Active powders) from the start.
This is best achieved by insuring that the initial melting and solidification temperatures are well
below steelmaking temperatures.

The other side of the picture is that unfortunately any flux, which has a high capacity for
absorbing inclusions, will also tend to be aggressive towards tundish refractories (particularly
Al2O3 based). For the same reason, mould powders are not used as tundish powders since they
contain high amount of fluidisers such as calcium fluoride (CaF2). In order to minimise
refractory corrosion, it has been suggested to use either:

1. low viscosity with low Al2O3 slags, although these are aggressive towards refractories at
the start but soon become high in viscosity or
2. use two fluxes: a simple lime-alumina flux in the pouring area, so that refractory erosion
is minimised, and a more complex flux (low in alumina) covering all other areas.

Low interfacial tension between the tundish slag and steel is suggested to enhance
inclusion assimilation. However, a very low interfacial tension between steel and molten
tundish powder (mainly due to high S and O in steel) results in a Marangoni effect. In this case,
non-metallic inclusions may transfer back to the body of the liquid steel.

Some work on the operational parameters and processing has also been carried out in the
past to improve the inclusion assimilation. For example, increasing the residence time in the
tundish by incorporating weirs and dams in the tundish geometry has been shown to increase
the chance of inclusion floatation out to the interface.
Present situation and open problems

The simple solution of achieving good thermal insulation and inclusion assimilation capability
of tundish powders is to keep their chemistry constant during the operation, but it is very
difficult under practical dynamic conditions.

The slag chemistry can be controlled to some extent. Possible actions, currently pursued,
are:

1. elimination of the reoxidation source within the tundish; no significant amount of MnO
and FeO are desirable since these lead to reoxidation of steel;
2. use of basic powders; a fairly high lime content is required to provide a reasonable
basicity while some MgO would help to reduce the attack on MgO-based refractories;
3. use of multi-layer with two mixtures; it is noted that tundish flux chemistry suitable for
inclusion absorption is rather aggressive towards Al2O3-based refractories. So two
layers or multi-layer powders particularly with high Al2O3 in the top layer could be used
to avoid severe refractory attack by the flux;
4. fresh powder addition to compensate local formation of crust.

All of these actions are based on thermodynamic considerations and practical experience.
 KEMPROCAST MOULD POWDERS

LING: Lubricating powders for slabs, billets, rounds, thin slab

Introduction

The primary functions of the mould flux are:

1. to form a liquid slag pool which protects the surface of the steel from oxidation
2. to provide thermal insulation which prevents the surface of the steel from freezing
3. to lubricate the newly-formed shell
4. to control the horizontal heat transfer between shell and mould
5. to absorb inclusions from the steel

The most important of the five functions of the mould flux are
providing lubrication and the right level of heat transfer. However
the continuous casting process is a very complicated process and
everything counts.

The casting powder must be tailored to perform satisfactorily

for the casting conditions.

The casting powder cannot act on its own and satisfactory operation requires that other
casting variables should be carefully adjusted, for example:

good steelmaking practice (minimise Al2O3 formation, inclusion engineering)

good mould level control
careful design and positioning of the SEN to ensure good temperature and steel flow
pattern at meniscus
mould oscillation characteristics.

Modern synthetic powders usually contain:

1. raw materials with similar melting points (eg lithium feldspar, sodium feldspar,
wollastonite) to provide more uniform melting
2. a minimum number of raw materials to achieve the desired composition so as to keep
the recipe simple and simplify quality assurance.

Spray dried powders were introduced to provide better quality control and minimise health
hazard.
Pulverised powders do have the following disadvantages:

fine dusts present health problems to plant operations

some inhomogeneity of supplies can arise because fine particles tend to gravitate
towards the bottom of bag.
 For the production of Spray dried spherical powders, the various powdered raw materials
are weighed and mixed with water for > 5 hours to form a slurry or suspension. It is then
sieved and pumped into a stirrer (to prevent segregation) and is then fed to an atomiser in a
spray tower. Fine spherical droplets are formed. Hot air is introduced and the evaporation of
water in droplets is very rapid since the drops have a high ratio surface area/ volume.
The result is dry spherical hollow particles, which are then sieved, cooled, sieved and packaged.

Chemical composition of mould powders

Continuous casting powders are made up of several components as shown in the following:

SiO2 17 - 56% CaO 22 - 45%

Al2O3 0 - 13% MgO 0 - 10%

Glass formers Basic Oxides
B2O3 0 - 19% BaO 0 - 10%

Fe2O3 0 - 6% SrO 0 - 5%

Na2O 0 - 25% F 2 - 15%

Fluidiser
Alkali Li2O 0 - 5% MnO 0 - 5%

K2O 0 - 2% Melting rate control C 2 - 20%

The effects of these various components on the viscosity, melting point and
solidification temperature are summarized in the following table

Increase Viscosity Solidification Point Melting Point

CaO Decrease Increase Increase

SiO2 Increase Decrease Decrease

CaO / SiO2 Decrease Increase Increase

Al2O3 Increase Decrease Increase

Na2O Decrease Decrease Decrease

F Decrease Increase Decrease

Fe2O3 Decrease Decrease Decrease

MnO Decrease Decrease Decrease

 MgO Decrease Decrease Decrease

B2O3 Decrease Decrease Decrease

BaO Decrease Decrease Decrease

Li2O Decrease Decrease Decrease

TiO2 No Change Increase Increase

K 2O Decrease Decrease Decrease

Functions of mould powders

Powders are added continuously to the mould and form 4 layers over the steel surface:

1. raw powder layer

2. sintered layer
3. mushy layer (mixture of slag globules and carbon particles)
4. a pool of liquid slag

Chemical protection
The presence of a liquid slag pool above the meniscus and carbon particles in the sintered and
powder layers prevent oxidation of the steel surface

Thermal insulation
The principal factors affecting thermal insulation is the bulk density.

Lubrication of the steel

Liquid slag infiltrates into the gap between the solid slag film formed against the water-cooled
mould and provides lubrication for the solidified steel shell. It is desirable that this layer of slag
should remain molten through the whole length of the mould.

Uniform heat transfer between steel and mould

Longitudinal cracking is known to be related to both the excessive magnitude and variability of
the heat flux. The heat flux is controlled by the slag film formed between the steel and mould.
The principal factors affecting the heat transfer between steel and mould are

thickness of the slag layer, this increases with increasing solidification or break
temperature and viscosity of the flux
the crystallisation tendency of the slag film that acts as a good shield to the radiant heat
interfacial resistance caused by the air gap formed resulting from the steel shrinkage.
Absorption of inclusions
The principal non-metallic inclusions formed in the continuous casting process are Al2O3 and
TiO2, in Ti-containing grades. Alumina is formed principally by oxidation of the steel particularly
in Al-killed grades; consequently the powder has to maintain its properties after alumina
absorption.

Physical properties of fluxes

The physical properties of the casting flux are important to the success of the casting process

Melting rate
Carbon, in the form of either carbon black, graphite or coke dust is added to the mould powder
to control its melting rate and to provide insulation to both the steel and the liquid slag.

Carbon particles are non-wetting to molten slag and thus separate the mineral particles
and slow down the agglomeration of molten slag globules. Thus the more C, the more time
required to agglomerate, the slower the melting rate.
The powder bed above the steel meniscus consists of four layers, powder, sinter, mushy
(mixture of slag globules and carbon particles) and molten slag pool.

The melting rate is controlled by several factors

1. the vertical heat flux density which is affected by various casting parameters such as
casting speed, superheat, turbulence, etc.
2. the free carbon content of the powder
3. the type of carbon and particle size of carbon
4. the carbonate content of the mould flux (as the CO2 developed from carbonates
decomposition causes turbulence in the mould powder layers so favouring its melting)
5. the presence of exothermic agents which can raise the melting rate.

Viscosity

The viscosity of the casting powder is a key physical property since it

influences:

the lubrication of the steel shell

slag entrapment (lower with higher viscosity)
depth of oscillation marks (lower with higher viscosity)
SEN erosion (higher with lower viscosity)

Several investigators have determined the effects of different additions on the viscosities of
mould powders:

Al2O3 additions cause increase in viscosity since they favour in a prone polymerised slag
Li2O, Na2O and CaF2 cause a decrease in viscosity since they result in a more
depolymerised melt
 Additions of TiO2 result in a slight decreases in viscosity up to the solubility limit of
about 10% beyond which the viscosity increases because of the solids present.
Addition of ZrO2 probably have a similar effect.

Measurements of viscosity as a function of temperature are made only when a new powder
is designed. Usually values of viscosity are calculated by means of mathematical models.

Break (Tbr) Solidification (Tsol) and Melting (Tmelt) Temperatures

The break temperature is the temperature below which there is a marked change in
viscosity and represents the point where liquid lubrication starts to break down.
Break temperatures are obtained from viscosity measurements during the cooling
cycle. Break temperatures are dependent upon the cooling rate and are usually quoted
for a cooling rate of 10C/min but cooling rate in the mould could be > 50 times that.
Consequently, Tbr values in the mould could be substantially lower than those quoted
for 10C/min.
It should be noted that high-viscosity fluxes such as those used in the billet-casting have
no break temperature since they form super-cooled liquids.
The solidification temperature is the temperature where the flux starts solidifying
(formation of crystalline phase) and is sometimes called Crystallisation temperature. It
is usually measured during cooling cycle in Differential Thermal Analysis (DTA) or
Differential Scanning Calorimetry (DSC) measurements.
The melting temperature is the temperature where flux becomes totally fluid and is
usually measured on the heating cycle in DTA/DSC experiments.

Differential Thermal Analysis are made only when a new powder is designed. Usually, the
characteristic temperatures (softening, melting and fluidity) are determined by means of
heating Microscope.

Mould powders characteristics

As the mould powders performance influences both continuous casting operations and
slabs surface quality, mould powders need to have the following fundamental characteristics:

Melting temperature suitable to guarantee, at a fixed casting temperature, a sufficient

slag fluidity for the infiltration into the gap between steel and mould. The casting
powders are essentially mixtures of oxides, present as different mineralogical phases,
which do not melt simultaneously; for this reason the heating of these powders give
rise to a temperature melting range instead of a melting temperature
a viscosity suitable for both the infiltration into a very narrow gap and minimization of
the friction between steel and mould
a melting rate suitable to guarantee a proper molten slag reserve
Convenient values of the break and solidification temperatures that are very important
for their influence on the slag lubricating ability and heat flux regulation.
 When a slag changes from the molten phase to doughy and then glassy phase, it
maintains a certain lubricating ability due to the fact that a glass is an under cooled
liquid. When a slag changes from the molten phase to a crystalline one, its lubricating
ability is completely lost

Concernings the heat flux regulation it has to be remembered that a glassy slag layer
between steel and mould does not shield the radiant heat and it is suitable to cast steels not
subjected to longitudinal cracks. A slag with high tendency to the crystallisation forms a good
shield to the radiant heat and it is recommended for casting of steels generally subjected to
longitudinal cracks.

How to read a mould powder technical data sheet

Chemical Analysis

The basicity index (IB = CaO/SiO2) gives an idea of the slag tendency to crystallisation:
- IB > 1.0 generally means that the powder has a certain tendency to crystallisation and
so it is best suited to cast steels subjected to longitudinal cracks;
- IB < 1.0 generally means that the powder is prone to form a glassy film between
mould and steel, so the powder has good lubricant ability but does not shield the
radiant heat.
The Free Carbon content indicates the melting rate of the powder: low free C
corresponds to high melting rate and consequently high molten pool depth (keeping
constant operative parameters and powder viscosity).

Physical Analysis

The characteristic temperatures (softening, melting and fluidity), determined by means

of Heating Microscope are important, especially the fluidity temperature which
determines the thickness of the slag rim. When this temperature is too high it can cause
a big slag rim that can inhibit the slag inflow into the channel between mould and steel.
Viscosity calculated as a function of temperature. Generally, the viscosity has to be
lower for higher casting speed to guarantee a good lubrication.

Free Carbon content and viscosity are the most important parameters in determining the
molten slag pool depth.

Free Carbon content controls how much liquid slag is produced; viscosity controls how
much liquid slag infiltrates; in practice that means that if you need to reduce the viscosity the
free Carbon content has to be reduced too, in order to keep constant the liquid slag pool depth.