Arch. Metall. Mater.
62 (2017), 2, 911-918
DOI: 10.1515/amm-2017-0134
P. DROŻDŻ*#
INFLUENCE OF COOLING CONDITIONS ON A SLAB’S CHILL ZONE FORMATION
DURING CONTINUOUS CASTING OF STEEL
The cooling conditions of a slab during continuous casting of steel have an impact on the crystalline structure formation.
Numerical methods allow real processes to be modelled. Professional computer programs are available on the market, so the
results of their simulations allow us to understand the processes that occur during the casting and solidification of the slab. The
study attempts to evaluate the impact of the intensity of the secondary cooling on the chilled zone size. The calculations show the
differences in the structure of a slab cast with various speeds while maintaining industrial cooling parameters during operation of
a continuous casting machine.
Keywords: crystallisation, chilled grains, continuous casting of steel, ProCAST
1. Introduction or can have a globular structure. When the macrostructure
consists of columnar and equiaxed grains, then a columnar-to-
The cast strand macrostructure is largely related to the equiaxed transition (CET) occurs. The CET transition in a slab
casting temperature (superheat), strand geometry, cooling and often involves a negative effect of segregation, the so-called
steel composition. Usually, cast strands have a fine equiaxed white bands, which may appear when electromagnetic stirring
zone near the surface (the chilled grain zone). This zone forms is applied [1-2]. Figure 1 shows a typical steel slab structure
during a fast crystallisation beginning. Depending on the casting diagram, divided into zones of chilled, columnar and equiaxed
parameters, the central area usually solidifies as a combination grains.
of columnar dendrites from the chilled grain zone towards the On the basis of numerical calculations in the ProCAST
centre of the strand and equiaxed dendrites in the central part software, the paper determines the influence of secondary cooling
of the slab. When the superheat temperature is very high, the intensity on the size of the forming chill zone in a continuously
equiaxed zone may not exist, so the macrostructure is comprised cast slab intended for flat products. The computations allowed us
of columnar dendrites entirely. When the superheat is very low, to indicate differences in the structure of a slab cast with various
the slab may solidify with the equiaxed dendrite zone only. speeds, while maintaining industrial cooling conditions during
The equiaxed zone can consist of randomly oriented dendrites operation of the continuous caster.
a) b) c)
Fig. 1. Steel slab structure diagram. a), b) 1 – chill zone; 2 – columnar zone; 3 – equiaxed zone, c) cross-section perpendicular to the slab axis [3]
* AGH UNIVERSITY OF SCIENCE AND TECHNOLOGY, FACULTY OF METALS ENGINEERING AND INDUSTRIAL COMPUTER SCIENCE, AL. A. MICKIEWICZA 30, 30-059 KRAKÓW, POLSKA
# Corresponding author: [email protected]
912
2. Cooling zones in the continuous steel casting process speed. In addition, the secondary cooling intensity, by selecting
an appropriate cooling schedule, depends on the steel grade that
Steel solidification in the process of continuous casting is on the carbon and alloy contents. Table 1 presents the area
starts within the mould, which constitutes the primary cool- of all slab cooling zones, whereas Figure 2 shows the change
ing zone, and subsequently it is continued in the water spray in the secondary cooling intensity versus the casting speed for
or air-mist spray zone located within the secondary cooling a selected cooling schedule.
chamber. After leaving the chamber the strand is cooled with
air and partially with guide rolls. The mould is built of copper TABLE 1
slabs, which are intensively water-cooled. The slabs form the Strand cooling zone in the continuous strand casting process broken
internal shape of the cross-section of the working part of the down into all zones
mould, which is adjustable to give the required dimensions to
the cast strand. During cooling in the mould, a solidified strand Cooling zone Surface, m2
Mould 2.91
layer – the shell – forms. Its strength and thickness are the basic
1st zone 0.94
parameters of the continuous casting process. The main task of
2nd zone 0.52
the secondary cooling zone is to continue the solidification of the
3rd zone 5.87
strand leaving the mould. The secondary cooling system com- 4th zone 11.80
prises a few separate zones in which the strand cooling intensity 5th zone 13.28
can be controlled. The slab quality, in particular the quality of 6th zone 13.31
their surface and macrostructure, depends on the operation of 7th zone 10.70
the secondary cooling zone. Air cooling zone 21.44
The distance that is covered by the steel from the metal me-
niscus in the mould to the exit from the last segment of strand car-
rying rolls is called the metallurgical length. In this study, a slab 3. Heat transfer coefficients α
with a cross-section 220 mm × 1600 mm, cast on an arc-shaped
continuous caster with a radius of 10500 mm, characterised by Secondary cooling is performed with nozzles supplied either
a metallurgical length of 23296 mm, was analysed. The strand with water or with air-mist. The density of heat flux exchanged
length to the strand straightening point is 18103 mm, whereas to between the strand and the coolant in the area of water stream
the point of exit from the secondary cooling chamber it is 18438 impact is heterogeneous, and therefore in many cases empirical
mm. The mould is cooled with a constant cooling intensity, inde- dependences are assumed for computing. They take into account
pendent of the casting speed. The secondary cooling is divided the most important parameters decisive to the cooling intensity,
into 7 zones, and their cooling intensity depends on the casting arising from the nozzle design, cooling water stream density, its
24 1.60
3rd zone
22
4th zone 1.40
20
Secondary cooling flow rate, dm3 /s
18 1.20
Casting speed, m/min
16
1.00
14
12 0.80
5th zone
10
6th zone 0.60
8
6 1st zone 0.40
4 7th zone
0.20
2 2nd zone
0 0.00
16:19 16:48 17:16 17:45 18:14 18:43 19:12 19:40 20:09 20:38
Time, hh:mm
1st zone 2nd zone 3rd zone 4th zone
5th zone 6th zone 7th zone Casting speed
Fig. 2. The change in secondary cooling intensity versus casting speed for a selected cooling schedule
913
pressure and speed. The heat flux densities, and thus the heat into account the actual water flow rate in each secondary cooling
transfer coefficients depend on the temperature of the strand zone. However, Table 3 presents the calculated average values
surface and the cooling water temperature. The distance of the of heat transfer coefficients α for the narrow and wide side of
cooling nozzle outlet to the strand surface is an important factor the slab after taking into account all cooling zones. The heat
[4-5]. In the actual continuous steel casting process, depending on transfer coefficient values α for the mould and for air cooling
the cooling intensity, individual cooling zones may be character- are taken from literature. The calculated values will be used as
ised by a diversified heat transfer average value αspray expressed in boundary conditions for modelling the temperature distribu-
W/m2/K calculated on the basis of the simplified formula (1) [6]: tion on the strand using numerical methods in the ProCAST
software.
D spray 10Q 107 0.688Q Vspray (1)
where: TABLE 3
ν – water drop speed, m/s, The average values of heat transfer coefficients α for the narrow
V·spray – water stream density, m3/s/m2. and wide side of the strand cast
This formula allows the heat transfer coefficient values to be
Casting speed, m/min 0.8 1.0 1.2
estimated, when the surface temperature is unknown. The value
Heat transfer coefficients α, W/m2/K
of the water stream density V·spray, and the speed of water drops
Cooling zone
Mould 1600 1600 1600
ν falling onto the strand surface cooled need to be estimated. 1st zone 550 694 819
The value of the water stream density was calculated knowing 3rd zone 501 616 733
the secondary cooling zone area and the water flow rate on the Wide side 4th zone 213 260 307
basis of the formula (2): 5th zone 170 210 247
6th zone 136 166 196
V
V (2) 7th zone 145 208 236
S zone Air cooling zone 85 85 85
where: Mould 1600 1600 1600
V – water flow rate, m3/s, 2nd zone 490 615 724
Szone – zone area, m2. 3rd zone 501 616 733
Narrow side
The strand withdrawal speed during the process of steel 4th zone 213 260 307
5th zone 170 210 247
casting varies, which forces a change in the cooling parameters
6th zone 136 166 196
resulting in a change in the spray flow rate. The spray water flows
7th zone 145 208 236
through the same nozzles all the time. Therefore the speed of Air cooling zone 85 85 85
the water drops outflow from a nozzle can be calculated on the
basis of the formula (3):
V 4. ProCAST – thermal model
Q (3)
Snozzle
Nowadays, advanced computer programs are used for
where: Snozzle – nozzle cross-section area, m2. numerical modelling of the continuous steel casting process
Table 2 presents the speed of the water drop outflow from [7]. The ProCAST software package is a solution, which allows
a nozzle for the selected strand withdrawal speeds, after taking computing of casting processes to be performed with the finite
element method. Results of simulations in the ProCast software
TABLE 2
include temperature distribution, strand metallurgical length and
The speed of the water drop outflow from a nozzle for the selected shell thickness. The advantages of numerical modelling allow
strand withdrawal speeds
these methods to be applied to effectively design new casting
Casting speed, m/min 0.8 1.0 1.2 technologies and to optimise existing ones.
Cooling zone The speed of the water drop ν, m/s The thermal model of the steel continuous casting process
1st zone 21.3 26.4 30.6 for slabs was formulated on the basis of the actual process engi-
2nd zone 20.3 25.1 29.2 neering parameters. To this end, the slab geometry was prepared,
3rd zone 23.4 28.4 33.3 taking into account the division into the existing cooling zones.
4th zone 10.3 12.5 14.6 The same steel grade with the chemical composition presented
5th zone 10.8 13.2 15.5
in Table 4 was assigned to all strand elements.
6th zone 9.3 11.3 13.3
7th zone 11.6 16.5 18.6
TABLE 4
The chemical composition of the steel taken into account in computing
Al, % C, % Cr, % Cu, % Mn, % N, % Ni, % P, % S, % Si, %
0.0282 0.0966 0.0209 0.0222 0.736 0.0053 0.0731 0.0126 0.0105 0.0184
914
After entering the steel chemical composition into the Pro- the FEM mesh applied, with declared boundary and initial con-
CAST program, the material data of the steel grade concerned ditions, constitutes a complete data set allowing the simulation
was computed and then verified and adjusted. The material data to be performed. The computations resulted in obtaining the
includes thermophysical properties, such as: thermal conductiv- temperature distribution on the whole strand length versus the
ity, density, enthalpy, solid fraction share and the liquidus and casting speed. Table 5 presents a summary of the simulation
solidus temperatures. The specific heat, along with the melting results in the thermal model for various casting speeds. Figure 3
heat of the material used in computations, are amongst more shows the selected points on the 1/4 of the cross-section of the
essential properties. After assigning the appropriate material to strand cast; these points were used for plotting the temperature
each part of the model, along with the verified material data, distribution. Figures 4-6 present the temperature distribution
the boundary conditions and initial conditions for the specific at the selected points along the whole strand length versus the
process were determined. The prepared geometric model with casting speed.
TABLE 5
The summary of simulation results in the thermal model for various casting speeds
Total solidification Length of the liquid Time of staying Shell thickness under
Casting speed, m/min Casting time, s
time, s core, mm within the mould, s the mould, mm
0.8 1843 1036 13810 59 24
1.0 1467 1027 17120 47 20
1.2 1255 996 19920 40 18
Fig. 3. Selected points at 1/4 of the strand cross-section. Point 1 – slab corner, Point 2 – centre of the slab narrow side, Point 3 – centre of the
wide side, Point 4 – slab centre
1700
1600
1500
1400
1300
1200
1100
Temperature, qC
1000
900 Point 4
800 Point 3
700
Point 2
600
Point 1
500
400
300
200
100
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Strand length, m
Fig. 4. The temperature distribution at the selected points along the whole strand length for the casting speed V = 0.8 m/min
915
1700
1600
1500
1400
1300
1200
1100
Temperature, qC
1000
900 Point 4
800 Point 3
700
Point 2
600
Point 1
500
400
300
200
100
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Strand length, m
Fig. 6. The temperature distribution at the selected points along the whole strand length for the casting speed V = 1.0 m/min
1700
1600
1500
1400
1300
1200
1100
Temperature, qC
1000
900 Point 4
800 Point 3
700
Point 2
600
Point 1
500
400
300
200
100
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Strand length, m
Fig. 6. The temperature distribution at the selected points along the whole strand length for the casting speed V = 1.2 m/min
Steel casting and cooling in the continuous casting process distribution along the whole strand length. These properties are
during one sequence should enable a slab with a consistent qual- related to the design parameters of the continuous caster, and
ity to be obtained regardless of the casting speed. For various maintaining them at an appropriate level ensures the casting
logistic reasons in a steel production cycle, the casting speed process is failure free. Too thin a shell under the mould can re-
is altered, in order to maintain the casting continuity, which sult in a breakout and spillage of the liquid steel onto rolls. The
involves a change in the strand cooling intensity. Despite these strand should fully solidify before shearing. On the other hand,
changes, a slab should feature similar values as regards the the length of the liquid core should not be shorter than the strand
length of the liquid core, the shell thickness and the temperature length upstream straightening; when the soft reduction system
916
is applied it should be closely related to this system. Numeri- surface and within its volume to be simulated and the growth
cal computing in the thermal model, illustrated by the selected of individual primary phase grains to be analysed. The model-
steel grade, taking into account actual process parameters, show ling results in the history of changes of the interface position,
certain differences in results as the casting speed changes. An and the determination of the grain boundaries after comple-
increase in the shell thickness under the mould is observed as tion of strand solidification. The CA method computations are
the casting speed decreases. Assuming a constant mould cooling coupled with the simulation of the strand temperature field.
intensity, this relationship is as expected. It was observed that The nucleation at the strand surface and within its volume is
when the cooling intensity in the secondary cooling zone and the heterogeneous with random distribution of nuclei. It is assumed
casting speed decreased, the liquid core length changed, which that the dependence of the amount of active nuclei on superheat
was directly related to the different temperatures of the strand is described by the Gauss normal distribution function with the
centre along the strand for various casting speeds. The strand mean ΔTμ and the standard deviation ΔTs. The values of surface
surface features a similar temperature regardless of the casting and volumetric nuclei density assumed in the modelling, and
speed, which may show that the casting parameters in the actual the Gauss function parameters used in the model, are presented
process are selected correctly. in Table 6.
TABLE 6
5. ProCAST – the CAFE model Nucleation parameters assumed in the modelling
Volume nucleation
Attempts to model processes of formation of the slab
Average undercooling, ΔTμ 22
structure, taking nucleation and grain growth into account, are
Standard deviation, ΔTs 6
available in literature [8-11]. Numerical models consider the Maximum number of nuclei, Nmax, m–2 200 000
heat flow issue in the scale of the whole strand, whereas nu- Surface nucleation
cleation and grain growth in the micro scale. Unfortunately, in Average undercooling, ΔTμ 15
some models it is not possible to determine the columnar zone Standard deviation, ΔTs 9
boundaries and the relationships between the length and the Maximum number of nuclei, Gmax, m–3 650 000
cross-section of grains. The CAFE model, which is now avail-
able as a ProCAST software module, was used in this study to The slab structure in the cross-section obtained by the
describe the nucleation and grain growth of the primary phase. simulation with a cellular automaton is presented in Figure 7.
In this model cellular automata (CA) were used. The cellular A narrow equiaxed chill zone is located near the slab surface.
automaton method enables the grain nucleation on the strand In the case analysed, the equiaxed zone in the central part is
a)
b)
c)
Fig. 7. Results of modelling of the primary grain structure in the cross-section of a steel strand with dimensions of 220 x 1600 mm for various
casting speeds: a) V = 0.8 m/min, b) V = 1.0 m/min, c) V = 1.2 m/min
917
0.5
0.45
Mean surface of the analyzed grains, cm 2
0.4
0.35
0.3 Casting speed
0.25 0.8 m/min
1.0 m/min
0.2
1.2 m/min
0.15
0.1
0.05
0
0 10 20 30 40 50 60 70 80 90 100 110
Distance from the slab surface, mm
Fig. 8. The change in the mean surface of analyzed grains in the slab cross-section versus casting speed
practically non-existent. The columnar zone dominates the Acknowledgements
whole cross-section. Figure 8 presents the change in the mean
surface of analyzed grains in the slab cross-section versus the The study was funded from the means of the Ministry of Science and Higher
casting speed. An increase in the casting speed slightly influences Education as a statutory work of AGH-UST No 11.11.110 293.
obtaining a finer strand structure. A similar structure may exist
in slabs.
The results of the modelling of the slab structure obtained REFERENCES
with the CAFE method may reproduce well the actual structure
of the slabs obtained by continuous casting. However, they do [1] T. Pikkarainen, V. Vuorenmaa, I. Rentola, M. Leinonen, D. Porter,
not allow some defects, such as e.g. axial porosity, to be forecast. Effect of superheat on macrostructure and macrosegregation in
continuous cast low-alloy steel slabs. 4th International Conference
on Advances in Solidification Processes (ICASP-4), Materials
6. Conclusion Science and Engineering 11 (2016).
[2] Liang Bai, Bo Wang, Honggang Zhong, Jie Ni, Qijie Zhai, Jieyu
The CAFE module of the ProCAST software is a useful tool Zhang, Experimental and Numerical Simulations of the Solidifi-
for modelling the formation of the structure of slabs obtained cation Process in Continuous Casting of Slab. Metals 6, 53 (2016).
by continuous casting. In accordance with the results of mod- [3] Z. Kudliński, Technologie odlewania stali. Wydawnictwo Poli-
elling the solidification process of a slab with a cross-section techniki Śląskiej, Gliwice, 2006.
of 220×1600 mm, a structure was obtained, which consists of [4] K. Milkowska-Piszczek, J. Falkus, Calculation of the boundary
a narrow equiaxed chill zone and a wide columnar zone. The conditions in the continuous casting of steel process. Metalurgija
equiaxed zone in the central part is practically non-existent. An 53, 4, 571-573 (2014).
increase in the casting speed with a simultaneous increase in the [5] Z. Malinowski, M. Rywotycki, Modelling of the strand and mold
secondary cooling intensity involves a change in strand solidifi- temperature in the continuous steel caster, Archives of Civil and
cation conditions. A change in the cooling intensity taking into Mechanical Engineering 9 (2), 59-73 (2009).
account the local industrial conditions results in obtaining a liquid [6] A. Buczek, A. Burbelko, P. Drożdż, M. Dziarmagowski, J. Fal-
core in the strand with various lengths, whereas the temperature kus, M. Karbowniczek, Tomasz Kargul, K. Miłkowska-Piszczek,
distribution at the strand surface shows slight differences. An M. Rywotycki, K. Sołek, W. Ślęzak, T. Telejko, L. Trębacz,
increase in the casting speed causes a slight grain size reduction E. Wielgosz, Modelowanie procesu ciągłego odlewania stali –
at the strand cross section. monografia, Radom 2012.
918
[7] B. Thomas, Modeling of the Continuous Casting of Steel- Past, [9] M. Rappaz, Ch.-A. Gandin, Probabilistic modeling of micro-
Present and Future. Electric Furnace Conference Proceedings, structure formation in solidification process. Acta Metallurgica
ISS 59, 3-30. (2001). et Materialia 41, 2, 345-360 (1993).
[8] A. Burbelko, J. Falkus, W. Kapturkiewicz, K. Sołek, P. Drożdż, [10] Ch.-A. Gandin, M Rappez, A coupled finite element-cellular
M. Wróbel, Modeling of the grain structure formation in the steel automation model for the prediction of dendritic grain structures
continuous ingot by CAFE method. Archives of Metallurgy and in solidification processes. Acta Metallurgica et Materialia, 42,
Materials, Polish Academy of Sciences. Committee of Metallur- 7, 2233-2246 (1994).
gy. Institute of Metallurgy and Materials Science 57, 1, 379-384 [11] Ch.-A. Gandin, J.-L. Desbiolles, M. Rappaz, Ph. Thevoz: A three-
(2012). -dimensional cellular automation – finite element model for the
prediction of solidification grain structures. Metallurgical and
Materials Transactions A 30, 3153-3165, December (1999).
