See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/261216791

Effect of casting speed on continuous casting of steel slab

Article · January 2014

CITATIONS READS

0 874

2 authors:

Ambrish Maurya Pradeep Jha

Amity University Banka Eyecarere and childcare hospital and research centre
16 PUBLICATIONS   17 CITATIONS    47 PUBLICATIONS   314 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

In-Mold Electromagnetic Stirring View project

All content following this page was uploaded by Ambrish Maurya on 28 April 2015.

The user has requested enhancement of the downloaded file.

Int. J. Mech. Eng. & Rob. Res. 2014 Ambrish Maurya and Pradeep Kumar Jha, 2014

ISSN 2278 – 0149 www.ijmerr.com

Special Issue, Vol. 1, No. 1, January 2014
National Conference on “Recent Advances in Mechanical
Engineering” RAME – 2013
Research Paper © 2014 IJMERR. All Rights Reserved

EFFECT OF CASTING SPEED ON CONTINUOUS

CASTING OF STEEL SLAB
Ambrish Maurya1 and Pradeep Kumar Jha1

A three dimensional mathematical model has been developed to study the effect of degree of
superheat and casting speed in continuous casting of steel slab. Study is based on investigation
of heat transfer and solidification behaviour of steel within mold and Secondary Cooling Zone
(SCZ). The Computational Fluid Dynamics (CFD) software Ansys Fluent 13.0 has been used to
solve the discretized equations with realizable k- turbulence model. For solidification, Enthalpy-
Porosity technique was used which treats the mushy region (partially solidified region) as a
“pseudo” porous medium. Liquid fraction of steel in the mushy region has been traced to find the
solidified shell thickness. However, calculations were made to find the temperature distribution
and metallurgical length at different degree of superheat and casting speed. Result shows that
superheat has a little effect on temperature distribution and metallurgical length of strand while
casting speed has a significant effect. High casting speed may cause inadequate thickness of
the solidified shell at the mold exit to withstand the Ferro-static pressure of the molten metal
below the mold. It may also leads to breakout due to sticking of solidified shell and mold because
of lack of slag film for lubrication between the two.

Keywords: Solidification, Continuous casting, CFD, Heat transfer, Mathematical modelling

INTRODUCTION solidified to sufficient depth (Choudhary et al.,

Continuous casting is the primary method in 1993; Meng and Thomas, 2003; and Amimul
production of steel billets, blooms or slabs.The Ahsan, 2011) around the inner surface of the
process starts by transferof molten steel from mold, forming a solid shell with molten metal
ladle through tundish to the water cooled in centre of strand.Moreover the solid shell
copper mold by submerged entry nozzle, formed should have an adequate amount of
whereby molten metal get solidified into thickness to withstand the ferro-static pressure
‘‘semis’’ and subsequently pulled/rolled out into (Mazumdar and Ray, 2001) of the molten
final product. The cooling intensity of the mold metal.Cooling rate has also a great influence
should be high enough so that steel get on the formation of various defects within the
1
Indian Institute of Technology, Roorkee, Roorkee, Uttarakhand, India.

This article can be downloaded from http://ijmerr.com/special-Issue.php

13
Int. J. Mech. Eng. & Rob. Res. 2014 Ambrish Maurya and Pradeep Kumar Jha, 2014

cast product (Lee et al., 2000; Meng and fraction lies between 0 and 1. This mushy zone
Thomas, 2003; Li and Thomas, 2004; Thomas, is modelled as a “pseudo” porous medium
2006; Hibbeler et al., 2009; and Hadala et al., where porosity decreases from 1 to 0 as the
2011). So an appropriate control of the strand material solidifies. When the material has fully
cooling and shell growth is to be made tohave solidified in a cell, the porosity becomes zero
defect free products. and hence the velocities also drop to zero.
Hence it can be said that the pull velocity is
Atthe exit of the mould, strand enters to the
included for the movement of the solidified
secondary cooling zone. Secondary cooling
material as it is continuously withdrawn at
zone have rollers to support the strand and
casting speed.
assist in bending and straightening. Water
sprays are present to extract the heat from the In the present work degree of superheat and
strand. These sprays are grouped as spray casting speed has been considered as the two
zones according to cooling rate required which process parameters which are varied to see
is controlled independently by valves. Later, their effect on the solidification behaviour of
strand moves to radiation zone (Versteeg and the cast product. Variation of degree of
Malalasekera, 2012) where it cools naturally superheat is inevitable as when the next ladle
and cut off after particular length for further starts pouring into the tundish and then further
processing. into the mold. Casting speed is important as
high speed casting is required to improve
Many authors have investigated cooling and
productivity and production yield. High speed
solidification models (Lee et al., 2000;
Mazumdar and Ray, 2001; Meng and Thomas, Figure 1: Schematic of Steel Continuous
2003; Gonzalez et al., 2003; and Amimul Casting Process
Ahsan, 2011) for the continuous casting in
mold. It is very difficult to measure the data and
process experimentally, so mathematical
modelling remains a reliable tool for obtaining
required information and preventing costs of
experimental investigations in order to
increase the productivity and quality.In the
present work, a precise heat transfer model
simulating the solidification process within
mold and secondary cooling zone of
continuous castinghas been developed. An
enthalpy-porosity technique (ANSYS
FLUENT) has been used using CFD
softwareAnsys Fluent 13.0. In this technique,
liquid fraction is calculated,which indicates the
fraction of cell volume that is in liquid form. A
mushy zone region is formed in which the liquid
This article can be downloaded from http://ijmerr.com/special-Issue.php
14
Int. J. Mech. Eng. & Rob. Res. 2014 Ambrish Maurya and Pradeep Kumar Jha, 2014

casting is also reported tocause breakout The source term QL, has two terms in it;
(Meng and Thomas, 2003; and Li and explicit latent heat term and convective term.
Thomas, 2004) and other surface defects In a single phase solidification model QL can
(Hakaru et al., 1984; Harada et al., 1990; Yeo be expressed as
et al., 1996; and Thomas, 2006) in the cast
due to active flow of liquid steel and low = + .∇
solidified shell thickness in the mold. Hence,
The latent heat has been released from the
in the present investigation, temperature
mushy zone. In the continuous casting the
distribution, metallurgical length and solid shell
solidifying shell is pulled out at a constant
thickness at narrow and wide face has been
casting velocity u pull . That means the zone
calculated by varying the process parameters
values. having solid fraction “fs” equals to one, will
move downward with the casting speed. It is
MATHEMATICAL to be noted that sum of liquid fraction “” and
FORMULATION solid fraction “fs” is always 1. The liquid fraction
can be calculated by determining the
Heat Transfer
temperature as
Energy conservation equation for solidification
can be defined as

+ ∇. ( )=∇ ∇ + ⎧ 0 <
⎪ −
= < <
where, H is the enthalpy of the material and ⎨ −
⎪ 1 >
can be computed as the sum of sensible heat ⎩
(h) and latent heat content (H),
Governing Equations
H = h + H
The continuity equation can be expressed as
Sensible enthalpy and latent heat content
are defined as, ( )=0

ℎ=ℎ + While, transient Navier-Stokes equation for

momentum conservation can be expressed as
H = L
( ) + ∇( )
where, L = latent heat of material,  = liquid ( )
fraction, =
= −∇ + ∇ (∇. ) + +
href = reference enthalpy, Tref = reference
temperature, where, eff = l + t is the effective viscosity
due to turbulence and k- model is used for
keff = effective conductivity, cp = specific this, l is the dynamic viscosity and t is the
heat,  = density, turbulent viscosity. In the enthalpy-porosity
u = velocity,,QL = source term. technique mushy zone is treated as porous

This article can be downloaded from http://ijmerr.com/special-Issue.php

15
Int. J. Mech. Eng. & Rob. Res. 2014 Ambrish Maurya and Pradeep Kumar Jha, 2014

medium and the porosity of every cell is set Sinks are added to all of the turbulence
equal to the liquid fraction in that cell. The equations in the mushy and solidified zones
porosity equals to zero if the zone is fully to account for the presence of solid matter. The
solidified, which extinguishes the velocity in sink term (11) is very similar to themomentum
that zone. Thus the momentum sink term “S” sink term.
was added to the right hand side of the Navier- (1 − 2 )
Stokes equation. The presence of this term = ℎ
( 3− )
allows the newly solidified material to move where  represents the turbulence quantity
downward at constant pull velocity. The being solved (k, , , etc.), and the mushy zone
momentum sink can be expressed as constant, Amush, is the same as the one used
(1 − )2 in equation above k and  are the inverse
= ℎ −
( 3− ) effective Prandtl numbers, C1 and C2* are the
model parameters, Gk is the generation of the
where,  is a small number (0.001) just to
turbulence kinetic energy due to mean velocity
prevent zero in denominator, Amush is mushy
gradients.
zone constant. The mushy zone constant
measures the amplitude of the damping; the C1 = 1.44, C2 = 1.92, C = 0.09, k = 1.0, 
higher this value, the steeper the transition of = 1.3(9)
the velocity of the material to zero as it
solidifies. Very large value may cause the BOUNDARY CONDITIONS
solution to oscillate. In the momentum sink term, AND ASSUMPTIONS
the relative velocity between the molten liquid The following assumptions were made during
and the solid is used rather than the absolute formulation of the solidification model to
velocity of the liquid. simplify the governing equations:
For simulating turbulence, the realizable k- • Liquid steel as Newtonian incompressible
 turbulence model was used, which is found fluid.
to be very much suitable. The turbulence • Only two dimensional heat transfers (lateral
viscosity is given by direction) is considered.
2
= • Convective boundary condition has been
taken into account for extraction of heat from
The two partial differential equations for
mold and strand.
turbulent kinetic energy (k) and dissipation rate
()(9)are given by • Density and specific heat of steel are
invariant.
+ (∇ )=∇ ∇ + + +
• Mold oscillation, bending of strand, effect
of segregation, etc., have been ignored.
+ (∇ )=∇ ∇ + 1
• Perfect contact between the shell and mold
2
∗
is considered as shrinkage due to
− 2 +
solidification is ignored.
This article can be downloaded from http://ijmerr.com/special-Issue.php
16
Int. J. Mech. Eng. & Rob. Res. 2014 Ambrish Maurya and Pradeep Kumar Jha, 2014

• No slip boundary condition prevails at the solid shell were set to move with casting
walls. velocity along the casting direction. In
Based on the above assumptions, the secondary cooling zone, different heat transfer
material properties and standard boundary coefficients were quantified for four sections
conditions used for analysis in present work modelled with consideration of different cooling
are listed in Table 1. rateswhile moving down the strand.
Outlet
Table 1: Material Properties and Boundary
Conditions Since, the solidified shell is pulled out with a
specified constant velocity, so a velocity inlet
Material Property, Boundary Conditions Value
-3
boundary condition at the exit. The velocity at
Steel Density, kg.m 7200
-1 -1
the outlet is equal to the casting speed toward
Viscosity of Liquid Steel, kg.m .s 0.0067
casting direction.
-1 -1
Thermal Conductivity, W.m .k 41
Specific Heat, J.kg .K -1 -1
750 COMPUTATIONAL
Latent Heat, J.kg -1
272000 PROCEDURE
Liquidus Temperature, K 1800 The computational domainand the grid were
Solidus Temperature, K 1770 created using Gambit. The model has total
Casting Speed, m/s 1.0, 1.2, 1.4 strand length of 12 m including mold with a
Liquid Steel Superheat, K 15, 20, 25 circular inlet of radius 79 mm at the topand its
Mushy Zone Constant 100000 other dimensions are shown in Figure 2. As
Inlet shown in the figure, geometry has been broadly
divided in two parts: mold and Secondary
The mold is fed by a simple rectangular inlet
Cooling Zone (SCZ). SCZ is further subdivided
port with velocity inlet of the molten steel. The
into four sections, as cooling rate varies while
velocity component at the inlet is only in z-
moving down the strand. Second order implicit
direction (casting direction) while, inlet velocity
scheme and realizable k- turbulence model
was obtained by balancing the inlet flow rate
were used to solve the fluid flow equations by
with the casting speed. However, inlet
finite volume method. For solidification, instead
temperature (Tinlet) of the molten steel was fixed
of tracking liquid-solid front explicitly, enthalpy-
according to the superheat (T) provided to
porosity technique has been used as
the steel above the liquidus temperature (TL).
explained above. The solution was executed
The inlet temperature can be expressed as:
in transient state with time step of 0.01 second.
Tinlet = TL + T To avoid divergence during calculation,
Wall reduced under relaxation factor has been
To avoid the computational difficulties used. The solution convergence has been
associated with the heat extraction from steel achieved with momentum residuals < 10-4 and
through cooling water flowing within mold, heat energy residuals < 10-7.
transfer by convection (2, 12) has been In order to check the accuracy of present
considered for the mold walls.The walls with work, model is validated with the calculation
This article can be downloaded from http://ijmerr.com/special-Issue.php
17
Int. J. Mech. Eng. & Rob. Res. 2014 Ambrish Maurya and Pradeep Kumar Jha, 2014

Figure 2: Model Geometry (a) Thinner Side

temperature distributions were calculated at
(b) Wide Side (c) Isometric View the centreline of the faces of the strand and
solid shell thickness at the symmetric plane of
two axes along the casting direction. The liquid
fraction having value less than 0.5 is
considered to be solidified.
Effect of Superheat
In the first part of investigation, effect of
superheat has been studied by the temperature
distribution and solid shell thickness. Figures
4a and 4b shows the temperature distribution
at the broad face and narrow face respectively
of mold and SCZ. As heat loss from the mold
is very high, a sharp drop of temperature can
be seen in the mold region, while along SCZ

Figure 4: Temperature Distribution

of solid shell thickness at the narrow wall of at Different Superheat (a) Broad Face
the model of Nakato et al. as shown in Figure (b) Narrow Face
3. Form the figure, it can be seen that there is 1800

Superheat 15K
a good match between them for solid shell 1700 Superheat 20K
Superheat 25K
thickness. 1600

1500
Temperature (K)

Figure 3: Validation of Model

1400

1300
10
1200

8 1100
Shell Thickness (mm)

1000
0 2 4 6 8 10 12
6
Distance below meniscus (m)

1500
4
Superheat 15K
1400 Superheat 20K
Simulated Result Superheat 25K
2 Experimental Result of Nakato et al.
1300
Temperature (K)

0
0 100 200 300 400 500 600 700 1200

Distance from meniscus (mm)

1100

RESULTS AND DISCUSSION 1000

Model calculations of temperature and solid

900
shell thickness are performed with 0 2 4 6 8 10 12
Distance below meniscus (m)
solidification and melting model. The
This article can be downloaded from http://ijmerr.com/special-Issue.php
18
Int. J. Mech. Eng. & Rob. Res. 2014 Ambrish Maurya and Pradeep Kumar Jha, 2014

heat loss reduces down the strand Effect of Casting Speed

consequently the increase and decrease in To study the effect of casting speed in
strand temperature results from the change in continuous casting of steel, investigation has
cooling conditions. On the other hand, been performed with three different casting
temperature drop in broad (near) face is speeds. Figures 6a and 6b shows the
observed less as compared to narrow (far) face temperature distribution at broad and narrow
because of difference in heat flux at the two
face of the strand for different casting speed.
surfaces. It can also be seen that the change
It can be noticed that change in casting speed
in superheat has a very little effect on slab
affects the temperature distribution
temperature and solid shell thickness, as
significantly which in turns affect the solid shell
shown in Figures 5a and 5b. Thus small
thickness in mold and SCZ, as shown in
change in degree of superheat may not affect
Figures 7a and 7b. It was observed that the
the process significantly. But large changes
may reduce solid shell thickness at the mold metallurgical length of the strand increases with
region and hence breakout and other defects increase in casting speed and at the speed of
can occur. 1.4 m/s metallurgical length of the strand

Figure 5: Solid Shell Thickness at Different Figure 6: Temperature Distribution

Superheat (a) Thin Section (b) Wide at Different Casting Speed (a) Broad Face
Section (b) Narrow Face
0.10 1800

Superheat 15K 1.0 m/s

Superheat 20K 1700 1.2 m/s
0.08 Superheat 25K 1.4 m/s
1600
Solid shell thickness (m)

Tempetature (K)

0.06 1500

1400
0.04
1300

0.02 1200

1100
0.00
0 2 4 6 8 10 12
1000
Distance from meniscus (m) 0 2 4 6 8 10 12
Distance below meniscus (m)
1400
Superheat 15K
0.8 Superheat 20K 1.0 m/s
Superheat 25K 1.2 m/s
1300 1.4 m/s
Solid shell thickness (m)

0.6
Temperature (K)

1200

0.4
1100

0.2
1000

0.0 900
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Distance below meniscus (m) Distance below meniscus (m)

This article can be downloaded from http://ijmerr.com/special-Issue.php

19
Int. J. Mech. Eng. & Rob. Res. 2014 Ambrish Maurya and Pradeep Kumar Jha, 2014

Figure 7: Solid Shell Thickness at Different

thickness has been calculated to study the
Casting Speed (a) Thin Section (b) Wide effect of superheat and casting speed. It was
Section observed that small change in degree of
0.10 superheat will not affect the process and has
1.0 m/s
1.2 m/s very little effect on the metallurgical length. The
0.08 1.4 m/s
change in casting speed has more
Solid shell thickness (m)

pronounced effect on the chance of formation

0.06
of defects such as bulging, cracks, etc. Casting
0.04
speed can be increased or decreased by
properly maintaining the cooling in the mold
0.02 and secondary cooling zone.

0.00
0 2 4 6 8 10 12
REFERENCES
Distance below meniscus (m)
1. Amimul Ahsan (2011), “Modeling
1.0 m/s Solidification Phenomena in the
0.8 1.2 m/s
1.4 m/s Continuous Casting of Carbon Steels”, in:
Two Phase Flow, Phase Change and
Solid shell thickness (m)

0.6
Numerical Modeling, pp. 121-148.

0.4
2. ANSYS FLUENT® Academic Research,
Release 13.0, Help System, Solidification
0.2
and Melting Guide, ANSYS, Inc.
3. Choudhary S K, Mazumdar D and Ghosh
0.0
0 2 4 6 8 10 12
A (1993), “Mathematical Modeling of Heat
Distance below meniscus (m) Transfer Phenomena in Continuous
Casting of Steel”, ISIJ International,
became more than the computational domain
Vol. 33, pp. 764-774.
and could not get solidified completely at the
caster exit. Significant decrease in solid shell 4. Gonzalez M, Goldschmit M B, Assanelli A
thickness at the mold exit was observed which P, Berdaguer E F and Dvorkin E N (2003),
may cause breakout at mold exit. Due to high “Modeling of the Solidification Process in
casting speed bulging, inner and surface a Continuous Casting Installation for Steel
cracks and other defects may be found in final Slabs”, Metallurgical and Materials
cast. So the casting speed is to be limited to Transactions B, Vol. 34, pp. 455-473.
prevent the formation of any defect. 5. Hadala B, Cebo-Rudnicka A, Malinowski
Z and Goldasz A (2011), “The Influence of
CONCLUSION Thermal Stresses and Strand Bending on
A three dimensional numerical model has Surface Defects Formation in
been developed to study the solidification of Continuously Cast Strands”, Archives of
continuous casting of steel slab. The Metallurgy and Materials, Vol. 56,
temperature distribution and solid shell pp. 367-377.
This article can be downloaded from http://ijmerr.com/special-Issue.php
20
Int. J. Mech. Eng. & Rob. Res. 2014 Ambrish Maurya and Pradeep Kumar Jha, 2014

6. Hakaru N, Ozawa M, Kinoshita K, Habu Y 11. Mazumdar S and Ray S K (2001),

and Emi T (1984), “Factors Affecting the “Solidification Control in Continuous
Formation of Shell and Longitudinal Casting of Steel”, Sadhana, Vol. 26,
Cracks in Mold During High Speed
pp. 179-198.
Continuous Casting of Slabs”,
Transactions ISIJ, Vol. 24, pp. 957- 965. 12. Meng Y A and Thomas B G (2003a),
“Heat-Transfer and Solidification Model of
7. Harada S, Tanaka S, Misumi H, Mizoguch
S and Horiguchi H (1990), “A Formation Continuous Slab Casting: CON1D”,
Mechanism of Transverse Cracks on CC Metallurgical and Materials
Slab Surface”, ISIJ International, Vol. 30, Transactions B, Vol. 34, pp. 685-705.
pp. 310-316. 13. Meng Y A and Thomas B G (2003b),
8. Hibbeler L C, Thomas B G, Santillana B, “Modelling Transient Slag-Layer
Hamoen A and Kamperman A (2009), Phenomena in the Shell/Mold Gap in
“Longitudinal Face Crack Prediction with Continuous Casting of Steel”,
Thermo-Mechanical Models of Thin Slabs Metallurgical and Materials
in Funnel Moulds”, in European Transactions B, Vol. 34, pp. 707-725.
Conference Continuous Casting of Steel,
pp. 1-10. 14. Thomas B G (2006), “Modeling of
Continuous Casting Defects Related to
9. Lee J-E, Yeo T-J, Oh K H, Yoon J-K and
Mold Fluid Flow”, Iron and Steel
Yoon U-S (2000), “Prediction of Cracks
Technology, Vol. 3, pp. 1-6.
in Continuously Cast Steel Beam Blank
Through Fully Coupled Analysis of Fluid 15. Versteeg H K and Malalasekera W
Flow, Heat Transfer and Deformation (2012), “An Introduction to Computational
Behaviour of a Solidifying Shell”, Fluid Dynamics: The Finite Volume
Metallurgical and Materials Method”, Longman Scientific &
Transactions A, Vol. 31, pp. 225-237. Technical.
10. Li C and Thomas B G (2004), 16. Yeo K K I M T, Hwan K and Lee D (1996),
“Thermomechanical Finite-Element
“Effect of Carbon and Sulfur in
Model of Shell Behavior in Continuous
Continuously Cast Strand on Longitudinal
Casting of Steel”, Metallurgical and
Surface Cracks”, ISIJ International,
Materials Transactions B, Vol. 35,
pp. 1151-1172. Vol. 36, pp. 284-289.

This article can be downloaded from http://ijmerr.com/special-Issue.php

21
View publication stats