Surface Effects and Contact Mechanics X 85

A study of breakout of a continuously cast steel

slab due to surface cracks
F. Kavička1, J. Dobrovská2, K. Stránský1, B. Sekanina1, J. Stetina1,
M. Masarik3, T. Mauder1 & Z. Franek4
1
Brno University of Technology, Czech Republic
2
VSB-TU Ostrava, Czech Republic
3
EVRAZ VITKOVICE STEEL, a.s, Czech Republic
4
Faculty of Business Administration, Karvina, Czech Republic

Abstract
The solidification and cooling of a continuously cast slab and the simultaneous
heating of the mold is a very complicated problem of three-dimensional transient
heat and mass transfer. The solving of such a problem is impossible without
numerical models of the temperature field. Experimental research and
measurements must be carried out simultaneously with numerical computation.
An important area of the caster is the so-called secondary cooling zone, which is
subdivided into thirteen sections. In the secondary cooling zone, where the slab
begins to straighten, a breakout of the steel can occur at the points of increased
local chemical and temperature heterogeneity of the steel, due to increased
tension resulting from the bending of the slab, and also due to high local
concentration of non-metallic, slag inclusions. The changes in the chemical
composition of the steel, during the actual continuous casting, are especially
dangerous. In the case of two immediate consequent melts this could lead to
immediate interruption of the continuous casting and to a breakout. The
temperature field of a slab was calculated by means of original numerical model
before, as well as after the breakout and the calculated parameters were
collected. If the dimensionless analysis is applied for assessing and reducing the
number of these parameters, then it is possible to express the level of a risk of
breakout as the function of five dimensionless criteria.
Keywords: continuously cast slabs, oscillation marks, hooks, chemical
composition, breakout.

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
doi:10.2495/SECM110081
86 Surface Effects and Contact Mechanics X

1 Introduction
Oscillation marks are transverse grooves forming on the surface of the
solidifying shell of a continuously cast slab. The course of individual marks is
rough and perpendicular to the direction of the slab movement. The formation of
the marks is sometimes the result of the bending of the solidifying shell during
the oscillation of the mould, which depends on the frequency and the amplitude
of the oscillation and on the casting (movement) speed. The hooks are solidified
microscopically thin surface layers of steel – as described by various authors
(Badri et al. [1], Thomas et al. [2], Ojeda et al. [3]). They are covered with
oxides and slag. Their microstructure is different to that of the base material of
the solidifying shell. The formations of the oscillation marks and hooks are
related. The depth of the oscillation marks and also their shapes, size and the
microstructure of the hooks vary irregularly. An increasing extent of these
changes leads to a defect in the shape of a crack, which reduces the thickness of
the solidified shell of the slab upon its exit from the mould, and it causes a
dangerous notch.
In the secondary cooling zone, where the slab begins to straighten, a breakout
of the steel can occur at the points of increased local chemical and temperature
heterogeneity of the steel, due to increased tension resulting from the bending of
the slab and also due to high local concentration of non-metallic, slag inclusions.
The changes in the chemical composition of the steel, during the actual
continuous casting, are especially dangerous. The consequences of this
operational immediate change in the chemical composition of the steel, which
are not prevented by a breakout system directly inside the mould (Pyszko et al.
[4]), could lead to immediate interruption in the continuous casting and to a
breakout at a greater distance from the mould than usually, thus leading to
significant material losses and downtime.

2 Interruption of continuous casting

The following case was recorded during the process of continuous casting of
250×1530 mm steel slabs of the grade A with the 0.41 wt.% carbon content and
9.95 wt.% chromium content (melts 1 to 3), and grade B steel with the 0.17 wt.%
carbon content and 0.70 wt.% chromium content (melt 4). The casting of the first
two melts of the grade A took place without any significant problems, after the
casting of the third melt of the grade A, the fourth melt of the grade B followed.
The change in the chemical compositions of the steels of both grades was carried
out very quickly by changing the tundish. Inside the mould, the steel B mixed
with the steel A of the previous melt. The pouring continued for another 20
minutes, but then, a breakout occurred between the 7th and 8th segments and the
caster stopped at the unbending point of the slab, at a distance of 14.15 m away
from the level of the melt inside the mould. The height difference between the
level inside the mould and the breakout point was 8.605 m. This tear in the shell
occurred on the small radius of the caster (Fig. 1a). A 250 mm thick sample was
taken from the breakout area using a longitudinal axial cut. The structure of this

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
 Surface Effects and Contact Mechanics X 87

sample and the distribution of sulphur was analyzed on the basis of the Bauman
print (Fig. 2) as well. The numbers 1 to 11 indicate the positions of the samples
in the places around the breakout intended for analysis (Fig. 1b). Simultaneously,
significant 25 mm sulphide segregations were discovered – very heterogeneous
areas created by the original base material of the slab (melt 3, steel of the grade
A), the new material of the slab (melt 4, steel of the grade B) and between them,
and also in the areas of mixed composition. Beneath the surface of the slab, at a

Figure 1: a) Macro-structure of the breakout; b) Sampling scheme.

Figure 2: Baumann print.

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
88 Surface Effects and Contact Mechanics X

depth of 75 to 85 mm, there were cracks and a zone of columnar crystals oriented
towards the surface of the slab on the small radius. This was identical to the
orientation of the groove, which gradually turned into a crack (Fig.1b – direction
4 – 6) and, on the opposite surface of the slab, the hook, which was covered by
melt (position 8). During the first phase of the analyses, the objective was to
determine the material, physical, chemical and technological parameters, which
differed in both melts 3 and 4 (besides the already introduced chemical
composition). Table 1 presents the individual parameters of both melts.

3 Numerical model of the slab temperature field

The original numerical model takes into account the temperature field of the
entire slab (from the meniscus of the level of the melt in the mould to the cutting
torch) using a 3D mesh (Stetina et al. [5]). The solidification and cooling of a
continuously cast slab is a global problem of 3D transient heat and mass transfer.
If heat conduction within the heat transfer in this system is decisive, the process
is described by the Fourier-Kirchhoff equation. It describes the temperature field
of the solidifying slab in its all three states: at the temperatures above the
liquidus (i.e. the melt), within the interval between the liquidus and solidus (i.e.
in the mushy zone) and at the temperatures below the solidus (i.e. the solid state).
In order to solve these, it is convenient to use the explicit numerical method of
finite differences. Numerical simulation of the release of latent heats of phase or
structural changes is carried out by introducing the enthalpy function dependent
on temperature. The numerical model fully considers the non-linearity of the
task, i.e. the dependence of the thermo-physical properties, especially of the
continuously cast material and the mould on the temperature, as it does the
dependence of the boundary conditions on surface temperature and other
influences (shift rate, cooling intensity, etc.). The 3D model had first been
designed as an off-line version and later as an on-line version, so that it could
work in real time. After correction and testing, it will be possible to implement it
on any caster, thanks to the universal nature of the code. The off-line model
analyses the temperature field of the actual continuous casting, while it passes
through the primary-, secondary- and tertiary cooling zones, i.e. through the
entire caster. The model is fully functional and accurately conducts general
analyses of the influences of the various technological measures on the formation
of the temperature field of the entire continuous casting. In a similar way, it
helps the staff to analyse the failure state of the caster.
The off-line version of the temperature model was used now to simulate the
temperature field of the steel slab from the melt No. 3 (steel grade A), steel slab
from the melt No. 4 (steel grade B) and a mixture of A and B grades. We
consider that each thermo-physical property (thermal conductivity, heat capacity,
density and enthalpy) of the mixture A + B is the arithmetic average of the
properties of steels A and B. The dependence of these parameters on temperature
was observed (Dobrovska et al. [6]). The model solves the temperature history of
every point of the cross-section during its movement through the whole caster.
The model also provides the curves of isoliquidus (red) and isosolidus (blue).

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
 Surface Effects and Contact Mechanics X 89

Fig. 3 shows the comparison of isoliquidus and isosolidus in the first axial
longitudinal section of the slab, the Fig. 4 in the second axial longitudinal
section. The areas between isoliquidus and isosolidus, i.e. in the solidification
interval, are the so-called mushy zones.

Figure 3: Isosolidus and isoliquidus in the first axial longitudinal section of

the slab.

The calculated mushy zone (the area between isoliquidus and isosolidus
curve) in the cross-section of the slab, in which the break out has occurred (at the
distance of 14.15 m from the level of melt in the mould), is shown for steel A in
Fig. 5 for the grade A, in Fig. 6 the grade B, and in the Fig. 7 for the A + B. It is
possible to determine the area of mushy zone Fmushy. Furthermore, surface
temperatures of the slab in the small and large radius of a slab were calculated.
The course of the surface temperatures of all three variants is shown in Fig. 8.

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
90 Surface Effects and Contact Mechanics X

Figure 4: Isosolidus and isoliquidus in the second axial longitudinal section

of the slab.

Figure 5: The computed mushy zone in the breakout cross-section (Steel A).

Figure 6: A computed mushy zone in the breakout cross-section (Steel B).

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
 Surface Effects and Contact Mechanics X 91

Figure 7: A computed mushy zone in the breakout cross-section (Steel A+B).

Figure 8: The surface temperature of the slab in the large (upper curve-blue)
and small radius (lower curve-red).

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
92 Surface Effects and Contact Mechanics X

Table 1 presents the determined material, physical, chemical and

technological parameters of steels, which differed in both melts 3 and 4 (row 1-
10). These individual parameters of both melts enter into the off-line version of
the numerical model. The calculation results are given in the rows 11 to 16.

Table 1: The parameters characterizing the continuous casting of the melt 3

(steel grade A) and melt 4 (Steel grade B).

Item Parameter Symbol Units A – melt 3 B – melt 4

#
1 Pouring speed w [m.s-1] 0.0130 0.0126
2 Dynamic viscosity  [m-1.kg.s-1] 0.00570 TL 0.00562 TL
 = . [m-1.kg.s-1] 0.00772 TS 0.00615 TS
3 Density  [kg.m-3] 7560.7 7600.9
4 Latent heat of the phase L [m2.kg.s-2] 246×103 259×103
change
5 Specific heat capacity cp [m2.s-2.K-1] 632.6 611.0
6 Mould oscillation S [m] 0.0060.003 0.0060.003
amplitude
7 Oscillation frequency f [s-1] 1.533 1.533
8 Solidus temperature TS [°C] 1427.0 1480.6
9 Liquidus temperature TL [°C] 1493.9 1512.3
10 Difference between the TL – TS [°C] 66.9 31.7
liquidus and solidus
temperatures
11 Max. length of the hSmax m 21.07 19.72
isosolidus curve from
the level*
12 Min. length of the hSmin m 19.92 18.69
isosolidus curve from
the level**
13 Max. length of the hLmax m 14.50 16.20
isoliquidus curve from
the level*
14 Min. length of the hLmin m 13.70 15.20
isoliquidus curve from
the level**
15 The area of the mushy Fmushy m2 0.05366 0.04100
zone on half of the
cross-section of the
breakout +
16 The surface temperature Tsurf °C 934 1097
of the slab++

Notes (referring to Table 1):

*) of the steel inside the mould to a position 0.650 m from the edges of the
1.53 m wide slab;
**) of the steel inside the mould to the centre of the slab;
+) the overall area of half of the cross-section is Fslab = 0.19125 m2;
++) in the material 15 mm around the groove (Fig. 1).

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
 Surface Effects and Contact Mechanics X 93

4 Dimensionless criteria
If the method of a dimensionless analysis is applied for assessing and reducing
the number of parameters in Table 1, in the first approximation, then it is
possible to express the breakout risk level as a function of the five dimensionless
criteria contained in Table 2 (units [m], [kg], [s], [K]).

Table 2: Dimensionless criteria characterizing the breakout.

L .f S. f  .S 2 . f Fslab TL  TS

Criterion Fslab  Fmushy
c p . .TL .S w  TL
steel A 5124.78 1.179 172.77 1.3900 0.044782
steel B 6237.96 1.217 197.87 1.2729 0.056404*

Notes (referring to the Table 2):

*) The maximum temperature difference inside the mixture zone
( TL  B  TS  A ) / TL  B .

5 Susceptibility to breakout and its risk

The risk of a breakout grows in accordance with the first criterion directly
proportionally to the latent heat L released from the mushy zone and inversely
proportionally to its dynamic viscosity η. The second criterion (the Strouhal
number) includes transient, oscillation movement, including the amplitude of the
mould and also, implicitly, a susceptibility to marks and hooks, which precede a
breakout. The third criterion has similar significance but, in addition, includes
also dynamic viscosity. The first three criteria increase the breakout risk during
the melt 4 (steel B) more than during the melt 3 (steel A). The fourth criterion
characterizes the reduction of the load-bearing cross-section of the slab (by
28.1% in the melt 3 and by 21.4% in the melt 4) by creating a mushy zone,
which indicates a greater risk of a breakout in the melt 3. The last criterion
considers the effect of the mixed zone of the melt 3 and a common effect of the
mixed zone of the melts 3 and 4. The first three criteria are of a dynamic nature
and their product in the melt 3 is 1.044×106, while in the fourth melt it is
1.502×106, i.e. the mixture melt has a 50% greater risk of a breakout. The
product of all five criteria of the melts 3 and 4, considering their partial
homogenization, is 1.078×105 in the melt 4 and 6.498×104 in the melt 3. The
quotient of the product for the melts 3 and 4 is 0.603, which predicts a reduced
risk of a breakout in the melt 3. If you consider the influence of temperature on
the surface of the slab during the melt 3, and in the place of the groove during
the melt 4, it is clear that the effect of the groove during the straightening of the
slab is connected with tensile stress, then in the place of the groove (Fig. 1) the
effect must have been compensated for at a temperature of 1097°C, i.e. at a
temperature of 163°C higher than the temperature of a completely straight
surface of the slab of the melt 3. The data was obtained from the investigation

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
94 Surface Effects and Contact Mechanics X

into the causes behind a transversal crack that occurred in a different steel slab
(Dobrovska et al. [6]). In order to clarify this, it was necessary to conduct a
series of ductility tests at the temperatures ranging from 20°C to the solidus
temperature. Table 3 contains the test results from temperatures that are close to
the temperatures in the row 16 of Table 1. A comparison of the mechanical
values indicates that the tensile strength at 914.5°C and the pulling force are 1.5
times greater than at 1093.0°C. In addition to this, there was an 8.605 m column
of melt working on the mushy zone at the point of the breakout, where the
mushy zone reached hSmax =21.07 m from the level in the mould, i.e. at least
6.92 m beyond the breakout point. It is therefore possible to assume that the
main factor that significantly increased the risk of a breakout was the
superposition of the causing effects of the parameters occurring in the first four
criteria of Table 2.

Table 3: Ductility testing at 1093.0°C and 914.5°C [5].

Sample Testing Tensile Strength Diameter Contraction Deformation Breaking

temperature strength before work
breaking
[°C] [N] [MPa] [mm] [%] [mm] [ J]
1 1093 817 28.9 3.90 58.0 12.0 7
2 914.5 1247 44.1 5.35 21.5 5.5 6

6 Discussion and conclusions

Following a fast change of the tundish, there was a period of 20 minutes when
there was a mixture of the grade A and grade B steels. The liquidus temperature
1493.9°C of grade A increased to 1512.3°C and, simultaneously, the latent heat
of the phase change increased from 246 kJ/kg (grade A) to 259 kJ/kg (grade B).
This led to an increase in the temperature of the melt and to the re-melting of the
solidified shell of the original grade A steel. Furthermore, there was an increase
 3.melt  h L  3.melt = 21.07  13.70 = 7.37 m)
in the length of the mushy zone (up to hSmax min

and also in its temperature heterogeneity. The temperature of the mushy zone –
following the mixing of both qualities – could find itself anywhere between the
maximum temperature of the liquidus of grade A and the minimum temperature
of the solidus of the grade B (i.e. within the interval TL B  TS  A = 1512.3 –
1427.0 = 85.3°C. During the 20 minutes of pouring of the grade B steel (the 4th
melt), which began immediately after pouring of the grade A steel (the 3rd melt),
marks and hooks formed as a result of the oscillation of the mould and continued
to form during the unbending of the slab (Fig. 1 – where the groove is 50 mm
wide and 15-16 mm deep with an opening angle of 115°). The tensile forces in
the vicinity of this groove and the re-melting of the solidified shell caused a
breakout in the wall of the small radius of the slab at the unbending point.
One way of reducing the breakout risk and the successive shutdown of the
caster is to modify the values of the parameters in the first criterion in Table 2,
i.e. to select two consecutive melts of such chemical compositions and the

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)
 Surface Effects and Contact Mechanics X 95

corresponding physical and chemical parameters (from which the dimensionless

criteria are determined) that the criteria predict the zero-breakout.

Acknowledgements
This analysis was conducted using a program devised within the framework of
the GACR projects Nos. 106/08/0606, 106/09/0940, 106/09/0969 and
P107/11/1566.

References
[1] Badri, A. et al., Metallurgical and Materials Transactions B, Vol. 36 B,
p.373-383, 2005.
[2] Thomas, B.G., Sengupta, J. & Ojeda, C., Mechanism of Hook and
Oscillation Mark Formation in Ultra-Low Carbon Steel. Proc. of the Second
Baosteel Biennial Conference, Shanghai, PRC, Vol.1, pp. 112-117, 2006.
[3] Ojeda, C. et al., Mathematical Modelling of Thermal-Fluid Flow in the
Meniscus Region During an Oscillation Cycle. AISTech 2006, Vol.1,
pp. 1017-1028, 2006.
[4] Pyszko, R. et al., On-line Numerical Model of Solidification for Continuous
Casting Integrated with Monitoring and Breakout Prediction System. Proc.
of the MicroCAD 2008 International Scientific Conference. University of
Miskolc, Miskolc, pp. 95-97, 2008.
[5] Stetina, J. et al., Optimization of technology and control of a slab caster.
I. Off-line numerical model of temperature field of a slab a parametric
studies. Hutnicke listy LXIII, No. 1, pp. 43-52, 2010.
[6] Dobrovska, J. et al., Materials Science Forum Vols. 567-568, pp. 105-108,
2008.

WIT Transactions on Engineering Sciences, Vol 71, © 2011 WIT Press

www.witpress.com, ISSN 1743-3533 (on-line)