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

net/publication/328058000

A FERRITE POTENTIAL INFLUENCE ON HEAT TRANSFER CONDITIONS IN

INDUSTRIAL MOLD DURING THE CONTINUOUS CASTING OF STEELS

Article in Revista de Engenharia Térmica · September 2018

DOI: 10.5380/reterm.v8i1.61876

CITATIONS READS

3 351

6 authors, including:

Carlos Raimundo Frick Ferreira

Federal University of Rio Grande do Sul
14 PUBLICATIONS 153 CITATIONS

SEE PROFILE

All content following this page was uploaded by Carlos Raimundo Frick Ferreira on 30 November 2020.

The user has requested enhancement of the downloaded file.

Tecnologia/Technology

A FERRITE POTENTIAL INFLUENCE ON HEAT TRANSFER

CONDITIONS IN INDUSTRIAL MOLD DURING THE
CONTINUOUS CASTING OF STEELS

V. Karlinskia, ABSTRACT
W. Fogazzib, This paper aims to apply a solidification mathematical model to the process
B. V. de Souzac, of the continuous casting of steel. Heat transfer coefficients in the mold
were determined by the inverse method and they are related with both
C. R. F. Ferreirad, macrostructure conditions and carbon equivalents of carbon steels from
peritectic reactions. Both structure characterization and ferrite potential
C. A. dos Santose, were established by solidification parameters and chemical composition
and J. A. Spimf after casting. Samples were cut at different positions of the metal/mold
interface, whereas the selected sections were polished and etched with a
reagent for the metallographic examination. The image processing system
f
Universidade Federal do Rio Grande do Sul was used to analyze as-cast structure for every selected position. It was
Centro de Tecnologia/LAFUN observed that during casting, numerical predictions about metal/mold heat
transfer coefficients along the mold are in accordance with type-A and B
Bairro Agronomia
steels ferrite potential, due to both its sticking and depression tendency.
CP. 15.021, Porto Alegre, RS, Brasil
[email protected] Keywords: ferrite potential, solidification, continuous casting,
mathematical model.

NOMENCLATURE especially during the initial stage of the solid shell

formation inside the mold.
Cp equivalent carbon In order to understand the mold behavior, it is
Cr’ equivalent chromium important to know how the heat transfer from the
dist distance metal to the cooling fluid occurs. For a heat transfer
FP ferrite potential global analysis, the relative influence of every active
h heat transfer coefficient, W/m²K heat transfer way must be initially analyzed. In this
L liquid stage, both liquid and solid metal, lubricant, physical
MDF finite difference method space or air gap formed in the metal/mold interface,
Ni’ equivalent nickel as well as the mold wall and the cooling fluid are
T temperature, K considered as sources of thermal change. The solid
Tp thermocouple metal layer formed along the mold works as an
increasing barrier, which stems the heat flow, that is,
Greek symbols it constitutes a thermal resistance that relatively
increases the influence of the solid layer thermal
δ delta ferrite conduction on the heat global fluid. The mold walls,
γ austenite however, do not represent an important thermal
resistance to the heat conduction in the system,
Subscripts because, besides the order of magnitude of the walls
thickness, mold materials, usually copper and its
i initial value alloys, have a great capacity of transferring heat due
m/m metal/molde to their high thermal diffusibility.
n limit value For the thermal contact between both metal and
sim simulated mold, three regions of contact can be observed: one
exp experimental from the meniscus up to the initial point of
solidification, where the liquid metal is separated
from the mold by a layer of mold powder; the second
INTRODUCTION one has a good thermal contact between the solid
metal and the mold wall, due to the ferrostatic
The occurrence of heat transfer during primary pressure on the solidified layer by the liquid metal.
(metal/mold), secondary (sprays zones) and tertiary Furthermore, its behavior has an important influence
(free radiation) cooling stages is the main on superficial quality of the ingot. In the third one,
phenomenon concerned with the continuous casting there is either a physical separation or a well-defined
of steels. Either the thermal behavior prediction or air gap formation between both metal and mold,
determination of such heat transfer systems becomes making difficult the heat flow. This separation can
important as it directly interferes in solidification, vary according to the steel chemical composition and

18 Engenharia Térmica (Thermal Engineering), Vol. 8 • No 01 • June 2009 • p. 18-23

i.exe
Tecnologia/Technology Karlinski at al. A Ferrite Potential Influence on Heat …

its ferrite potential, which represents a sticking or The routine acts interactively. Initially, the
depression tendency, as shown by (Brimacombe, mesh determines a temperature profile of both metal
1973; Brimacombe et al., 1984; Mahapatra et al., and mold in the meniscus region, and it also
1991; Mills et al., 1991; Chandra et al., 1993). compares the simulated temperature with the
The aim of this paper is to apply a numerical experimental one in the monitored point from a
simulation model of solidification to a continuous convenient initial value of hm/m.
casting of steels. This model is capable of predicting Compared temperatures within the range of
the heat transfer coefficient (h) evolution along the ±1ºC are accepted. The test is carried out when the
metal/mold interface, in different faces of a square numerical mesh moves from the meniscus region to
150 x 150 mm section mold according to the sticking the next monitored point.
or depression tendency presented by different types At this time, the hm/m representativeness is
of steels. The results are compared with the verified by analyzing whether or not the simulated
macrostructure obtained from ingots by monitored temperature is within the tolerable range. If
heat, during the process of solidification. necessary, the system can either increase or decrease
the hi value and also repeat calculations from the
THEORY previous monitored point with a new hi value. In
each stage, the processing follows up to the distance
Heat Transfer in the Mold of the next thermocouple, and successively, along the
mold.
The problem concerning about the metal/mold
interfacial heat transfer has been of great interest of
studying by researchers, due to its important
influence on either the rate or speed of solidification.
Several works have been developed aiming at
characterizing the physical mechanism of heat
transfer between both metal and mold contact
interfaces, as it is shown by (Hills, 1969; Grill and
Brimacombe, 1976; Lait et al., 1974; Lait and
Brimacombe, 1984; Krishnan and Sharma, 1996;
Santos et al., 2001).
In this work, the model by Barcellos et al.
(2006) is employed to determine heat transfer
coefficients (h) along the metal/mold interface. This
method involves the measure of experimental
temperatures obtained by thermocouples placed in
the metal and/or mold in known positions. Those
temperatures are compared to profiles of simulated
ones by a numerical model of both solidification and
heat transfer, previously examined. The procedure
used to determine h is performed by adopting an
initial value of h (hi), by which, temperatures of each
position in a numerical mesh, are calculated by a
mathematical model for both mold and metal at
different time intervals. Every interaction, the hi
correction is carried out with either increasing or
decreasing of a Δhi value and therefore, new Figure 1. Flow-chart of the optimization algorithm
temperatures are estimated. for determining heat transfer coefficients in the
metal/mold interface.
Optimization Strategy for Searching the Heat
Transfer Coefficient along the Mold Peritectic Reaction

Search routines for determining heat transfer Solidification of steels starts with either δ-
coefficients in the metal/mold interface (hm/m), along ferrite or γ-austenite precipitation. In the case of
the mold length, were developed by optimization carbon steels, the peritectic reaction L + δ → γ
strategies for increasing both the processing speed occurs with a C concentration rate between 0.09 and
and the results accuracy obtained from the 0.53 in weight %. Its effectiveness can be obtained
mathematical model of solidification. Figure 1 shows by adding elements such as C, N, Ni, Mn, Cu,
the flow-chart of the optimization algorithm for responsible for austenite formation. Due to the
determining heat transfer coefficients in the difference of thermo-physical properties between γ-
metal/mold interface. austenite and δ-ferrite, this reaction originates

Engenharia Térmica (Thermal Engineering), Vol. 8 • No 01 • June 2009 • p. 18-23 19

Tecnologia/Technology Karlinski at al. A Ferrite Potential Influence on Heat …

undesirable phenomena such as stress, volumetric FP = 5,26.(0,74 - [%Ni' /%Cr' ])

variation, segregation and inclusions precipitation, (5)
according to Matsuura et al. (1995). Figure 2
schematically shows the peritectic reaction, where δ- Some studies have been carried out to
ferrite, firstly formed, is encapsulated by γ-austenite understand changes in the heat flux by the peritectic
from the L + δ reaction. reaction. Singh and Blazek (1974) analyzed the
medium heat flux in the mold region by the carbon
percentage (up to 1.7 %C). For concentrations up to
0.1 %, the heat flux decreases, presenting a minimum
C concentration value. Over this concentration rate,
the heat flux increases and it is kept constant
(concentrations over 0.25%). Grill and Brimacombe
(1976) proposed a mechanism based on phases δ → γ
reaction to explain the heat flux range. They
Figure 2. Schematic drawing of the peritectic compared the steel 0.10%C to others with larger C
reaction, showing δ-ferrite encapsulation by γ- concentrations and observed that the former suffers a
austenite. more intense reaction in the solid state than the latter,
with a volume contraction of 0.38%, which reduces
In general, the effect of alloying elements is the heat flux through the mold wall. Similar results
described by means of equivalent concentrations. For were found by Suzuki et al. (1996).
either carbon or low-alloy steels, the effect of
alloying elements can be expressed by equivalent EXPERIMENTS
carbon, as follows:
In order to investigate the FP influence on the
Cp = [%C] + 0,02.[%Mn]+ 0,04.[%Ni]- 0,1.[%Si] heat transfer coefficient, two types of steels were
selected with different chemical compositions: one
- 0,04.[%Cr]- 0,1.[%Mo]
with low carbon content and the other with medium
(1) carbon content, 0.16 %C and 0.53 %C respectively.
The chemical composition was analyzed and
The equation was applied to 12 steels and classified into two categories according to the ferrite
compared with the ferrite fraction – or Ferrite potential, such as: type A and type B. Heat transfer
Potential (FP) – obtained from solidification in coefficients both along the mold length and in three
experiments of differential thermo-analysis (Wolf different mold faces were determined from
and Kurz, 1981). This concept outlines a steel temperatures data obtained by thermocouples
tendency to either shrink or expand during strategically positioned in the mold walls. In both
solidification inside the mold. For either carbon or cases, a 150 mm x 150 mm section cooling copper
low-alloy steels, the ferrite potential is defined as: mold was monitored with type K thermocouples
inserted along the mold height. Heat transfer
FP = 2,5(0,5 − %C P ) (2) coefficients were correlated with the steels ferrite
potential, which tend to stick or depress from the
For steels with a tendency to form depressions wall. Experimental data were obtained from an
(type-A steels), FP values range between 0.85 and industrial plant by temperatures measured in a mold
1.05, whereas for those with a tendency to stick to of the continuous casting machine. Both relevant
the mold wall (type B steels), these values are specifications to the machine and experimental
FP>1.05 or FP<0.85. working conditions are shown in Tables 1, 2 and 3.
To calculate the ferrite potential for stainless The mold was monitored with 30 thermocouples
steels, the presence of both equivalent chromium and inserted in the three faces central position of the
nickel is taken into account, as follows: mold, arranged in the following way: 10
thermocouples in the center of the outer face, 10
Cr' = [%Cr] + 1,5[%Cr] + 1,38 [%Mo] + 2[%Nb] thermocouples in the center of the inner face, and 10
(3) thermocouples in the center of the side face.
+ [%Ti]
The thermocouples arrangement along the mold
is shown in Figure 3, from the meniscus region (the
Ni' = [%Ni] + 0,31[%Mn] + 22[%C] + 14,2[%N]
(4) liquid metal level at the mold top) to 170 mm from
+ 1[%Cu] the mold top, up to 31 mm above the mold exit,
which results a total of 10 (ten) strategic points of
The ferrite potential is given by the following faces monitoring. Thermocouples were placed at 4.4
equation: mm from the mold hot face (metal/mold interface).
The Tp2 (Thermocouple 2) means the reference of
meniscus temperature.

20 Engenharia Térmica (Thermal Engineering), Vol. 8 • No 01 • June 2009 • p. 18-23

Tecnologia/Technology Karlinski at al. A Ferrite Potential Influence on Heat …

For the macrographic analyses, two blocks of

Table 1. Machine specifications billets were taken from steels employed in this
Billet section 150x150 mm experiment. They were submitted to both a cutting
Mold type Parabolic process and a grinding machine with band saws.
Mold material Copper Figure 4 shows both an image of one of those blocks
Mold dimensions Length: 801 mm and a schematic drawing of a cut “slice” for the
Thickness: 11 mm metallographic procedure. After surface adequacy,
Machine length 24 m the samples were ground on a SiC paper (#1200),
polished on cloth with diamond abrasive and finally
Table 2. Monitored heat data in the continuous etched with 10 % Nital reagent.
casting machine for a low-carbon steel.
Chemical composition (wt %)
C Si Mn Mo Ni Cr
0.16 0.22 0.65 0.21 0.09 1.1
Operation data
Liquidus temperature 1513 ºC
Solidus temperature 1487 ºC
Pouring temperature 1577 ºC
Casting Speed 2.1 m/min
Section 150x150 mm
Ferrite Potential 0.98 (type A) Figure 4. A sample block obtained from the steel
billet and a schematic drawing of the sample used for
Table 3. Monitored heat data in the continuous the procedure.
casting machine for a medium-carbon steel.
Chemical composition (wt %) RESULTS AND DISCUSSION
C Si Mn Mo Ni Cr
0.53 1.3 0.67 0.04 0.11 0.66 Figures 5 and 6 show a set of both monitored
Operation data data of temperatures and results of heat transfer
Liquidus temperature 1474 ºC coefficients obtained during solidification for type-A
Solidus temperature 1404 ºC and B steels. Regarding mold walls temperatures, a
little difference among monitored temperatures in the
Pouring temperature 1514 ºC
mold faces was observed. The inner ray face presents
Casting Speed 2.0 m/min
the lowest temperatures in the mold wall due to the
Section 150x150 mm
mold curvature ray, which reduces the billet contact
Ferrite Potential 0.35 (type B) by the ferrostactic pressure. The side face, however,
shows a different behavior for both steels. For type-
A steel, between its inner and outer faces,
temperatures values are intermediate, whereas for B-
steel, they increase in relation to the outer face. By
analyzing temperatures from the meniscus region, a
great similarity among maximum temperatures
values (approximately 140ºC) was observed, in
despite of the difference among pouring
temperatures (approximately 60ºC). For type-B steel,
however, the region of maximum temperature
extends from the meniscus up to 300 mm under it.
It is also observed that coefficients achieve
their maximum values in the meniscus region due to
the best thermal contact between the liquid metal and
the mold wall. As these coefficients moves along the
mold, they drastically decrease, keeping both a
constant feature for type-A steel, and a decreasing
variable one for type-B steel. As previously
described in Perictectic Reaction, type-A steel has a
depression or shrinkage tendency, whereas type-B
steel, a sticking or expansion one. This situation has
an influence on the behavior of metal/mold heat
Figure 3. Thermocouples arrangement along the transfer coefficient along the mold. Therefore,
copper mold of the continuous casting machine. average values around 1500 W/m2.K and 2000

Engenharia Térmica (Thermal Engineering), Vol. 8 • No 01 • June 2009 • p. 18-23 21

Tecnologia/Technology Karlinski at al. A Ferrite Potential Influence on Heat …

W/m2.K are obtained for type-A steel and B steel

respectively.

Figure 6. Monitored data of temperatures and results

from heat transfer coefficients for B-steel.

Metallographies in Figure 7 show a good

correlation with h for each mold face. Furthermore,
by means of macrographic images, it can be
observed that type-A steel presents a grain structure
with both proportional size and orientation among its
faces. In other words, the heat extraction rate through
the metal/mold interface is homogeneous. For type-B
steel however, it was observed that the outer ray face
presents the greatest columnar grains, whereas the
inner ray face presents the smallest ones. Such
observations agree with h results for each mold face
shown in Figures 5 and 6. In type-A steel, a well-
defined columnar equiaxed transition can be noted
with columnar grains that are perpendicular to the
metal/mold interface.

Type A Type B

Figure 5. Monitored data of temperatures and results

from heat transfer coefficients for A-steel.

20 mm 20 mm

Figure 7. Macroestructures of type-A and B steels.

On the other hand, type-B steel presents an

irregular columnar-equiaxed transition, fine
columnar grains with a non-perpendicular growth to
the metal/mold interface, as well as an equiaxed
nucleus dislocated from the center of the billet.

CONCLUSIONS

22 Engenharia Térmica (Thermal Engineering), Vol. 8 • No 01 • June 2009 • p. 18-23

 Tecnologia/Technology Karlinski at al. A Ferrite Potential Influence on Heat …

The mathematical model was efficient to Krishnan, M., and Sharma, D. G. R., 1996,
determine both the ferrite potential and its influence Determination of the Interfacial Heat Transfer
on the heat transfer regime during solidification of Coefficients h in Unidirectional Heat Flow by Beck’s
steels the along parabolic mold with a square section Nonlinear Estimation Procedure, International
of 150 x 150 mm and 800 mm of height. Calculated Communications on Heat and Mass Transfer, Vol.
heat transfer coefficients were in accordance with the 23, No. 12, pp. 203-214.
heat transfer tendency, presenting both high values Lait, J. E., Brimacombe, J. K., and Weinberg,
for the region under the meniscus, where the thermal F., 1974, Mathematical Modelling of Heat Flow in
contact is more effective, and low ones as steel the Continuous Casting of Steel, Ironmaking and
solidification develops. Steelmaking (Quartely), Vol. 2, pp. 90-97.
These coefficients were also important to Lait, J. E., and Brimacombe, J. K., 1984,
indicate the difference existent between type-A and Solidification during Continuous Casting of Steel,
B steels heat transfer. The macrographic analysis Continuous Casting - Heat Flow, Solidification and
revealed that internal structures of both types of Crack formation, Vol. 2, pp. 171-183.
steels are influenced by the metal/mold heat transfer Mahapatra, R. B., Brimacombe, J. K.,
regime and the liquid-solid interface advance. Samarasekera, I. V., Walker, N., Paterson, E. A., and
Young, J. D., 1991, Mold Behavior and its Influence
ACKNOWLEDGMENTS on Quality in the Continuous Casting of Steel Slabs:
Part I, Industrial Trials, Mold Temperature
The authors would like to thank CNPq Measurements, and Mathematical Modeling,
(National Council of Technological and Scientific Metallurgical Transaction B, Vol. 22, pp. 861-874.
Development) for the financial support and Mahapatra, R. B., Brimacombe, J. K.,
GERDAU Aços Especiais Piratini for the partnership Samarasekera, I. V., Walker, N., Paterson, E. A., and
for studies on the process of solidification by Young, J. D., 1991, Mold Behavior and its Influence
continuous casting. on Quality in the Continuous Casting of Steel Slabs:
Part II, Mold Heat Transfer, Mold Flux Behavior,
REFERENCES Formation of Oscillation Marks, Longitudinal Off-
Corner Depressions, and Subsurface Cracks,
Barcellos, V. K., Fogazzi, W., Ferreira, C. R. Metallurgical Transaction B, Vol. 22, pp. 875-888.
F., Santos, C. A., and Spim, J. A., 2006, Matsuura, K., Kudoh, M., and Ohmi, T., 1995,
Desenvolvimento de Rotinas para Determinação e Simulation of Peritectic Reaction during Cooling of
Avaliação dos Coeficientes de Transferência de Iron-Carbon Alloy, ISIJ International, Vol. 35, No.
Calor ao Longo de um Molde durante o 6, pp. 624-628.
Lingotamento Contínuo de Aços, Anais do 61º Mills, K. C., Billany, T. J. H., Normanton, A.
Congresso Anual da Associação Brasileira de S., Walker, B., and Grieveson, P., 1991, Causes of
Metalurgia e Materiais, pp. 995-1005. Sticker Breakout During Continuous Casting,
Brimacombe, J. K., 1973, Design of Ironmaking and Steelmaking, Vol. 18, No. 4, pp.
Continuous Casting Machines Based on a Heat Flow 253-265.
Analysis, State of the Art Review, Canadian Santos, C. A., Quaresma, J. M. V., and Garcia,
Metallurgical Quartely, Vol. 15, pp. 1-13. A., 2001, Determination of Transient Interfacial Heat
Brimacombe, J. K., Samarasekera, I. V., and Transfer Coefficients in Chill Mold Castings, Journal
Lait, J. E., 1984, Continuous Casting – Heat Flow, of Alloys and Compounds, Vol. 319, No. 1-2, pp.
Solidification and Crack Formation, Iron & Steel 174-186.
Society of AIME. Singh, S. N., and Blazek, K. E., 1974, Heat
Chandra, S., Brimacombe, J. K., and Transfer and Skin Formation in a Continuous
Samarasekera, I. V., 1993, Mould –Strand Casting Mold as a Function of Steel Carbon Content,
Interaction in Continuous Casting of Steel Billets – Journal of Metals, pp. 17-27.
Part 3, Mould Heat Transfer and Taper, Ironmaking Suzuki, M., Yu, C. H., Sato, H., Tsui, Y.,
& Steelmaking, Vol. 20, No. 2, pp. 104-112. Shibata, H., and Emi, T., 1996, Origin of Heat
Grill, A., and Brimacombe, J. K., 1976, Transfer Anomaly and Solidifying Shell
Influence of Carbon Content on Rate of Heat Deformation of Peritectic Steels in Continuous
Extraction in the Mould of a Continuous Casting Casting, ISIJ International, Vol. 36, pp. 171-174.
Machine, Ironmaking and Steelmaking, Vol. 3, No.2, Wolf, M., and Kurz, W., 1981, The Effect of
pp. 76-79. Carbon Content on Solidification of Steel in the
Hills, A. W. D., 1969, A Generalized Integral- Continuous Casting Mold, Metallurgical
Profile Method for the Analysis of Unidirectional Transactions B, Vol. 12, pp. 85-93.
Heat Flow During Solidification, Transactions of the Received: April 13, 2008
Minerals, Metals and Materials Society – AIME, Revised: May 13, 2008
Accepted: June 13, 2008
Vol. 2445, pp. 1471-1479.

Engenharia Térmica (Thermal Engineering), Vol. 8 • No 01 • June 2009 • p. 18-23 23

View publication stats

i.exe