University of Wollongong

Research Online
University of Wollongong Thesis Collection
University of Wollongong Thesis Collections
1954-2016

1995

Improvement of surface quality of continuously

cast steel control of cast structure and straightening
temperature
Koesnohadi Soeyanto
University of Wollongong

Unless otherwise indicated, the views expressed in this thesis are those of the author and do
not necessarily represent the views of the University of Wollongong.

Recommended Citation
Soeyanto, Koesnohadi, Improvement of surface quality of continuously cast steel control of cast structure and straightening
temperature, Master of Engineering (Hons.) thesis, Department of Materials Engineering, University of Wollongong, 1995.
https://ro.uow.edu.au/theses/2488

Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library:
[email protected]
IMPROVEMENT OF SURFACE QUALITY OF CONTINUOUSLY
CAST STEEL BY CONTROL OF CAST STRUCTURE AND
STRAIGHTENING TEMPERATURE

A thesis submitted in fulfilment of the requirement for

the award of the degree of

HONOURS MASTER OF ENGINEERING

from

THE UNIVERSITY OF WOLLONGONG

by

KOESNOHADI SOEYANTO, Ir (ITB)

Department of Materials Engineering

1995
 ACKNOWLEDGEMENTS

The author wishes to thank:

1. The supervisor, Professor N. F. Kennon, for his invaluable guidance and

constant encouragement during this research work, for his great patience
and precious comment in reading the drafts and for his understanding of
his student.

2. Ir. Djoko Subagyo and Ir. Fazwar Bujang MM., for the chance given to
join this research program.

3. Top management of Slab Steel Plant PT. Krakatau Steel in year 1992­
1994 for the facility given for this research work.

4. Mr. Uus Kusmana, Mr. Arif N., Mr. Desmardi, Mr. Gandhi, and Mr. Chairul
for providing the observation of experimental data especially for
measuring the straightening temperature.

5. Ir. Koswara MSc. and Ir. Wawan Setiawan for providing the metallo­
graphy facility for this research.
 SYNOPSIS

In recent years, most world wide mass produced steels have been m ade
by continuous casting. Despite significant advances in continuous casting
technology, including advances made at PT. Krakatau Steel in Indonesia,
transverse cracking remains one of the most common defects
encountered during normal production of steel slabs containing 0.12­
0.18% carbon.

The presence of transverse cracks has undesirable influences on the

product because they can degrade quality as well as increasing the risk of
failure during applications of the material in finished products. The
problem of transverse cracking can be eliminated by removing the defect
by scarfing the corner of the slab, but this process involves a degree
material loss and associated expenditure. Therefore, decrease in the
incidence of transverse cracks is dem anded to satisfy the customers as
well as to obtain a significant yield increment and to improve process
efficiency. Consequently, examination of these defects and de­
termination of means of controlling the incidence of them is a matter of
considerable importance.

This thesis presents the results of a program of work designed to identify

means by which the incidence of transverse cracking of continuously cast
steel slabs can be reduced in a production machine such as that
operating at PT. Krakatau Steel.

In Chapters 2 and 3, the process of continuous casting of steel is examined

and the way in which solidification of the continuously cast strand takes
place is described.

The effects of operational parameters such as steel composition,

secondary cooling water, tundish temperature and straightening
temperature, which are the principal effects in causing cracking in the cast
material, are set out in Chapter 4. Particular attention is directed to
transverse cracking and the operational variables causing this defect are
identified.
In Chapters 6 and 7 the experimental procedure for determ ining the
influence on cracking of secondary w ater rate and titanium in the cast
m aterial are described and the results of the experimental work are set out.
A m athem atical m odel for estimating the strand surface tem perature is
presented in C hapter 5 and used in C hapter 7 to assist in analysing the
experim ental results.

It is co n clu d e d that to reduce the incid en ce of transverse cracking,

austenitic grain growth must be suppressed by titanium add ition up to
0.01 7% and the precipitation of allotriomorphs of ferrite must be a vo id e d
by lowering the w ater rate from 0.756 l/kg to 0.736 l/kg.
 CONTENTS

page

1. INTRODUCTION 1

2. CONTINUOUS CASTING OF STEEL 5

2.1. Continuous Casting Machine Components and Functions 6
2.2. Strand Deformation during Bending dnd Straightening 9

3. SOLIDIFICATION AND COOLING PROCESS IN CONTINUOUS CASTING 13

3.1. Cooling Process in the Mould 15
3.2. Secondary Cooling 20
3.3. Processes at the Solidification Front 24

4. DEFECTS IN CONTINUOUSLY CAST STEEL SLAB 26

4.1. Longitudinal Surface Cracks 27
4.2. Star Cracks 28
4.3. Transverse Cracks 29
4.3.1. Appearence of Transverse Cracks 30
4.3.2. Mechanism of Transverse Cracking 30
4.3.2.1. Effect of Coarse Austenitic Grain Structure 31
4.3.2.2. Effect of y/a Transformation 34
4.3.3. Prevention of Surface Transverse Cracking 37
4.3.3.1. Control of y Structure 37
4.3.3.2. Control of Precipitation 39

5. MATHEMATICAL HEAT TRANSFER MODEL FOR SOLIDIFICATON OF

CONTINUOUSLY CAST CAST STEEL SLABS 42
5.1. M athem atical Model 44
5.2. Numerical Solution 47
5.2.1. Determination of the Temperatures at Nodes 2 through (n-1) 49
5.2.2. Determination of Strand Surface Temperature 50
5.3. A pplication of the Model to Continuous Casting of Slabs 52

6. EXPERIMENTAL 54
6.1. Methods 58
6.1.1. Effect of Steel Composition 59
6.1.2. Effect of Straightening Temperature 62
6.1.3. Com bined Effects of Titanium and Straightening Temperature 64
6.2. M etallography 55
6.3. A pplication of the M athem atical Model 67
7. RESULTS AND DISCUSSION 70
7.1. Statistical Survey 70
7.2. Water Rate 76
7.3. M athem atical Model 83
7.4. Effect of Titanium 88
7.5. Com bined Effects Water Rate and Titanium 90
7.6. M etallography 93
7.6.1. Subsurface Structures in the Vicinity of Transverse Cracks 95
7.6.2. Grain Size of the Subsurface Cast Structure 99

8. CONCLUSIONS 103

REFFERENCES 107

Appendix A 111
COMPUTER PROGRAM FOR CALCULATING THE TEMPERATURE OF
STRAND DURING CONTINUOUS CASTING

Appendix B 119
INDEX OF TRANSVERSE CRACKS AND GRAIN SIZE OF THE CAST
STRUCTURES OF STEELS A, B, C AND D HAVING VARIUOUS TITANIUM
CONTENTS, CAST USING WATER RATES 0.756 l/kg AND 0.736 l/kg

Appendix C 120
MAIN SPECIFICATION OF CONTINUOUS CASTER

Appendix D 121
SPECIFICATION OF THE MOULD POWDER
 1

C h apter 1

INTRODUCTION

At the present time, steel is mainly produced by a sequence ot several

processing steps comprising iron and steelmaking, rolling, drawing and
other primary forming operations followed, if necessary, by secondary
forming.

Iron is the principal raw material for the steelmaking process, and is
!
obtained by reducing iron ore to molten iron in the blast furnace or by
direct reduction in the Hyl reactors and/or Midrex reactors. In these
processes, the iron dissolves from 2.0 to 4.0% carbon, but as commercial
steels contain considerably less than 1.0% carbon, the excess carbon
must be removed during the conversion into steel. The excess carbon is
removed by controlled oxidation of mixtures of molten iron and steel
scrap in steelmaking furnaces including the basic oxygen furnace and
the electric arc furnace to produce carbon steels with the required
carbon content. Various elements such as manganese, silicon, titanium
and vanadium may be added singly or in combination to molten steel in
the treatment ladle to produce alloyed steel. When the molten steel has
attained the required chemical composition, it is teemed Into the moulds
where it solidifies to form ingots. The ingots, offer being removed from
the moulds, are reheated and rolled into shapes known as blooms, slabs,
or billets and referred to as semifinished steel.
 2

However, since the early 1960s, the production of semifinished steels by

ingot casting routes has decreased with increasing development of
continuous casting in most steel industry worldwide. Continuous casting
offers excellent cost efficiency due to a substantial improvement in yield
of about 15% com pared with ingot casting. The improvement is a
consequence of elimination of processing stages such as blooming and
slabbing which result in a certain degree of material loss. In continuous
casting, feed stocks are produced directly in the form blooms, slabs, or
billets and, almost all steel plants, including PT. Krakatau Steel in Cilegon,
Indonesia, in the world today use this process for commercial production
of most steels. !

PT. Krakatau Steel, a steel com pany owned by the Indonesian

Government, is one of the largest steel producers in South East Asia with a
crude steel production capacity of 2,500,000 tonne per year. At present,
the steelmaking plant consists of the Billet Steel Plant (BSP) commissioned
in 1978, Slab Steel Plant No.l (SSP 1) commisioned in 1983 and the new
and modern Slab Steel Plant No.2 (SSP2) commisioned at the end of
1994. The BSP is equipped with four 65 tonne arc furnaces and two
continuous casters producing 600,000 tonne of mainly commercial steel
grades for wire rods. The SSP 1, has four 125 tonne arc furnaces and two
bow-type continuous casters with one straightening point, producing
1,200,000 tonne of mainly low and medium carbon steel grades for
drawing sheets, high grade pipes and structural purposes. The SSP2, has
two 125 tonne arc furnaces and one continuous caster with several
straightening points, producing 700,000 tonne of mainly ultra-low and low
carbon steel grades for deeo and normal drawing sheets for automotive
parts, cans and other special purposes.

Although most steel products made by PT. Krakatau Steel generally

satisfy many applications, improvements in quality as well as
diversification of products have been m ade during the last few years.
The medium carbon steel currently produced in SSP1 is one typical
example of the diversified steel products from PT Krakatau Steel. It is
produced to anticipate the increasing dem and for high grade pipes for
oil and gas transmission and also for structural proposes.

A problem encountered in the production of the medium carbon steel in

PT. Krakatau Steel is the formation of transverse (and other) cracks on the
surface of continuously cast slabs, particularly for steels containing 0.12
to 0.18% C. In practice, the cracks are currently removed by substantial
scarfing to avoid the formation of defects on the hot rolled steel as well
as on end products originating from it. Although the surface transverse
cracks can be removed by this means, the practice involves a degree of
material loss and other expenditure such as labor and consumable costs.

Recent developments of rolling without scarfing and of hot charge rolling

of continuously cast slab are being practiced at various steel makers to
save energy and improve yield. To operate this latter process therefore,
it is important to avoid surface cracking and the consequential defects
which would occur in the absence of scarfing.
 4

In the present study, factors affecting the surface quality of continuously

cast slab and the preventive measures which can be taken to avoid

surface defects are examined. Investigations into the nature of, and
mechanism of formation of, transverse cracks are described, as are the
methods used to reduce the incidence of this cracking. Steel
composition, secondary cooling water distribution, and slab surface
temperature were found to be major factors influencing the cracking.

To appreciate the origins of the several kinds of surface defects, it is

necessary to examine the process of continuous casting and to consider
the ways in which solidification and subsequent cooling occurs in a
caster.
 5

C h ap ter 2

C O N T I N U O U S C A S T I N G OF S T E E L

Continuous casting is a relatively recent developm ent in the steel industry

and has had a dram atic im pact on steel production throughout the world.
C om pared with conventional ingot casting, continuous casting has had
greatest effect on improving the efficiency of production. Additionally, it
has the advantage that the products of continuous casting approximate

more closely to the final product dimensions than the products of ingot
casting with corresponding reduction in the number of forming steps
required. The conventional ingot casting route requires soaking pits for
ingot heating, and primary mills for rolling the ingots into blooms, slabs, or
billets. In continuous casting, all these facilities are eliminated and the feed

stock is produced directly in the form of blooms, slabs, or billets. Such a

process not only leads to a considerable improvement in yield, but also

reduction in energy consumption. The process yield is more than 95%

com pared with approximately 80% for the production of semi finished

products by rolling from a cast ingot[l ,2].

Today, continuous steel casting is performed with a w ater-cooled open

ended copper mould which is separated from the unit supplying the
molten steel. At the com m encem ent of the casting operation, the mould

is closed at the bottom by a so-called dum my bar, then filled with the

molten steel after which the dum my bar is continuously withdrawn in the
casting direction. At the same time, the mould is oscillated in sinusoidal
movement to accommodate movement of the strand and to prevent
sticking of the strand to the mould wail. The steel shell which forms in the
mould contains a core of liquid steel which gradually solidifies as the
strand moves through the machine. The main components[l ,2] of the
machine are shown schematically in Fig.2.1.

Ladle Cover

Fig. 2.1 Schematic diagram of a continuous casting machine[2]

2.1. Continuous Casting Machine Components and Functions

There are three essential requirements of the continuous casting machine:
one is to guide the solidifying metal, the second is to remove heat from
the steel at the required rate, and the other is to deliver a solidified strand
at the strand device. A continuous casting machine (Fig.2.1) consists of a
liquid metal reservoir and distribution system (a tundish), liquid metal
 7

shrouding, a mould with water cooling system, secondary cooling zones in

association with a strand containm ent section, bending rolls, a
straightener and cutting e quipm ent^,3],

2.11 The fundish is essentially a rectangular box, but some tundishes are T

shaped or L shaped m ade from steel plate and lined with refractory
bricks, and a nozzle is located in the bottom of the tundish. The
tundish is designed to provide a number of important functions
including:
(a) control of metal flow patterns to enhance the stability of the
metal stream entering the casting mould,
(b) provision of a metal reservoir to facilitate casting a sequence of
heats, and
(c) a means to facilitate the separation of inclusions and slag
entering the tundish from the ladle.

2.12 Liquid metai shrouding is employed to avoid the metal stream

absorbing oxygen from the air, and thereby forming deleterious
inclusions in the liquid steel. Shrouds are normally placed between
the ladle and the tundish as well as between the tundish and the
mould. A shroud is normally tubular in form and manufactured from
alum una/graphite refractory. In addition, in most modern continu­
ous casting plant, argon is introduced into the refractory tube to
avoid aspiration of air through pores and joints due to the venturi
effect of the moving metal stream.
 8

2.13 The mould is constructed from a copper-based alloy as an open

ended box structure with an inner lining which serves as the interface

with the steel being cast and which determines the shape of the cast
section. The primary function of the mould system is to contain and
start solidification of the liquid steel. Control of heat transfer in the
mould is accomplished by a forced convection cooling water
system which is designed to accom m odate the high heat transfer
rates that result from the solidification process. In addition, mould
powder is introduced into the mould as the lubricant to minimize the
frictional force between the mould surface and the solidified cast
steel shell , and it is an absorber of oxide inclusions and acts as a
protective layer preventing reoxidation of the liquid m e ta l.

2.14 Secondary cooling is designed to produce a final cast section which

has the required shape, and surface quality. To accomplish these
results, the solidification section leaving the mould is cooled in a
series of spray zones and is contained and withdrawn by a series of
roll assemblies until the solidified cast section reaches the cut-off
machine. The secondary cooling system is normally divided into a
series of zones to control the cooling rate as the strand progresses
through the machine.

2.15 Strand containment is provided by a series of retaining rolls which

extend across the two opposite faces of the cast section in a
horizontal direction. The basic functions of the mechanical strand
containment and withdrawal equipment, which form an integral part

of the secondary cooling system, are:

 9

(a) to support and guide the strand from the mould exit to the cut-off
operation, and

(b) to drive the strand through the caster at a controlled speed.

In these two functions, the final objective is to minimize the

m echanical stress and consequential strain imposed upon the strand.

2.16 The sfraighiener is installed following the completion of unbending

which, as the name implies, straightens the strand and completes the
transition from the vertical to horizontal orientation. During straigh­
tening, the strand is "unbent" to reverse the tension and compression
forces in the horizontal faces of the strand. The series of rolls which
guide the strand through a prescribed arc from the vertical to the
horizontal plane must be strong enough to withstand the bending or
straightening reaction forces. During movement from the mould to
the torch cutter, the strand is subjected to a number of loads which
lead to defects in the product if the resultant stresses and the
associated strains becom e too high to be sustained by the cast cross
section and steel grade. The loading phenomena which occur are:
(a) tensile stresses resulting from the withdrawal forces,
(b) bending of the strand shell from the vertical orientation into a
curve, and
(c) straightening of the strand shell or the solidified strand into the
horizontal orientation.
The stresses associated with these mechanical phenomena may be
exacerbated by additional thermal stresses [3].
 10

2.2. Strand Deformation during Bending and Straightening

Strand support involves restraint of the solidifying steel shape which
consists of a solid steel shell and a liquid steel core. During bending, the
inner radius of the solid shell is placed in compression and the outer radius
in tension. On the other hand, during straightening the inner radius is
placed in tension and the outer radius is placed in compression^].
Excessive strain may result in metal failure and strand defects (cracks). The
strain e , which arises during straightening from the curvature radius to
straight in the horizontal direction is a function of the caster radius J?c, and
strand thickness f, and can be derived from consideration of Fig.2.2.

Fig. 2.2 Diagram showing strain due to straightening

in continuous casting process[4]
 11

From Fig. 2 , the strain €t, in the interior of the slab, can be formulated:

Ic-li [

« = -------- (2.1)

where, lc # Rc .tan fias

h & (Rc - dj) fan pas p -+ 0

when p~*0,

Rc Jan p - (Rc - d>) fan p

G = ------------------------------------------ (2 .2 )
c.
R fan p

d>
i.e. S] - ------ (2.3)
Rc

0
The strain due to straightening will be minimum, for d> & , in the center of

slab and will be maximum for d; * f/2, on the surface. Thus, the strain on

the outer fibre € St can be described as[4]:

f
Ss = ----- (2.4)
2 Rc

3 0009 03143400 9
 12

In SSP1-PT. Krakatau Steel, tor which the caster radius (Rc) is 9700 mm and
the strand thickness is 200 mm (f), the strain on the outer fibre during

straightening is:

200
6 = ------------ = 1.03 % (2.5)
2x9700

Generally, strain values up to 1.5% on the outer fibre are tolerated,

because the solidified steel in the outer fibre has sufficient strength to
accom m odate the strain. On the other hand, for the internal fibre, the
value must not exceed 0.5 %, because the solidification front is very
sensitive to cracking due to the lack of ductility when tensile strain arises
at the solid/liquid interface [4].

To minimize the development of cast strand defects, particularly surface

defects due to excessive strain, and which may be exacerbated by
additional thermal stresses, it is necessary to control both the temperature
levels and thermal gradients in the strand over the time of the casting
process. In practice, this aim can be achieved by controlling the cooling
of the solidifying strand. Solidification and the cooling process that occur
in the strand are described in Chapter 3.
i
 13

Chapter3

S O L ID IF IC A T IO N AND COOLING P R O C E S S
IN C O N T I N U O U S C A S T I N G

The continuous casting process is essentially a continuous

solidification process, with the molten metal travelling through a water-
cooled mould and a series of rollers [and cooling sprays. Steel is poured
into the water cooled copper mould having a predetermined shape and
solidifies under forced cooling conditions in the mould and progressively in
the spray cooling zone. During solidification, 15% of the heat is extracted in
the mould, 30 to 40% in the water spraying zone, and 10% in other parts of
the machine. This leaves 35 to 45% of the total heat still in the product at
exit and then sensible heat may be lost to the atmosphere[5].

The phase transformations involved during the solidification process can be

determined from consideration of the Fe-C constitutional diagram (see
Fig.3.1). For alloys containing 0.12 to 0.18% carbon, cooling from an initial
temperature higher than the liquidus to a temperature slightly below it (A),
results in nucleation of primary 8-dendrites in the liquid until recalescence

occurs due to the heat released from the growing nuclei. During
recalescence, nucleation ceases and existing nuclei grow rapidly into
dendritic grains which impinge on each other at the end of recalescence.
Growth of 8-dendritic grains is then succeeded by coarsening of the

dendritic arms until solidification is com plete (B). When the temperature
decreases further, austenite precipitates from the delta phase until all 8-
 14

phase is transformed. The austenite phase cools through the austenite

zone until the temperature reaches Ar3 at which ferrite begins to precipitate
(C) . Finally, at or below Àr3, the remaining austenite transforms to pearlite
(D) by nucleation at interfaces and coupled growth of ferrite and
cementite phasesfó]. ■

Fig. 3.1 Iron-carbon constitutional diagram[6].

During solidification, transformation of the liquid involves the extraction of

two kinds of the heat i.e. sensible-heat and latent-heat. Sensible-heat, is
the heat component which results from the difference in the actual liquid
metal temperature and the transformation temperatures. Latent-heat, is
the heat which is evolved during solidification and phase changes in the
solid state during cooling as the cast strand leaves the caster. Dissipation
of these heat components takes place by radiation, conduction and
 15

convection in the three cooling zone (Fig. 3.2), which are the primary
cooling zone (in the mould ), the secondary cooling zone (spray cooling
including the roll cooling system) and the zone in which heat is exclusively
transferred by radiation to the environment[7].

Meniscus

Fig. 3.2 Schematic representation of the three heat

transfer zones in continuous casting[7],

3.1. Cooling Process in the Mould

Solidification of liquid steel in the mould begins by the formation of the
strand shell in contact with the walis of the mould just below the liquid
meniscus (Fig. 3.3). This phase of solidification continues progressively and
is characterized by the presence of oscillation marks that form periodically
at the meniscus due to mould reciprocation. The movement of the mould
prevents the newly formed strand shell from sticking to the mould wall[8,9].
 1 2 3 4 5 6 7 1 2 3 4 5 6 7

Negative s tr ip time Negative s tr ip time

Fîg. 3.3 Diagram showing formation of strand shell and

oscillation marks [10].

The formation of oscillation marks has been studied [10] in relation to the
solidification in the meniscus region. The study suggested that a partially
solidified meniscus could play a significant role in the formation of
oscillation marks during slab casting. There is close relationship between
the distance between the oscillation marks I, the oscillation frequency f,
and the casting speed Vc. This relationship can be expres$ed[10]:

Vc
/ = - (3.1)
f

The formation of oscillation marks is largely the result of the interaction

between the mould flux and the mould movement, and has been related
previously to the stroke S, frequency of mould oscillation, and more
generally to the negative-strip time which is the time interval in each
 17

cycle of mould oscillation in which the downward velocity of the mould

(Vm) exceeds the withdrawal rate of the strand. The negative-strip time can

be expressed as [10]:

1 Vc
tN = ------arc cos ( ---------- ) (3.2)
iJ.f 171 $
There are two types of oscillation marks based on the presence or absence

of a small hook in the sub surface structure adjacent to each oscillation

mark. Figure 3.3 shows a schematic representation of the formation of the

two types of oscillation marks i.e., with and without adjacent subsurface

hooks.

As shown in Fig. 3.3, during the negative-strip time (stages 1-3), the

meniscus is pushed by the positive pressure, generated in the mould flux,

away from the mould wall. Then in the ensuing positive period (stages 4-7),

the meniscus is drawn back toward the mould wall. It is most likely that the

meniscus is drawn back uniformly, hovj/ever, because the upper part of the

meniscus skin is farthest from the cooling influence of the mould wall, it

therefore should be the hottest and weakest. As a result, the upper part of

the skin is expected to be drawn back more by the the negative flux

pressure and inertial force of the surging liquid steel. The difference
 18

between the two types of oscillation marks then arises because of

differences in the mechanical strength of the meniscus skin. In the case of

oscillation marks with subsurface hooks, the skin is relatively strong, owing to

the greater thickness) low superheat or minimal bath movement). Thus the

top of the skin resists being bent back fully toward the mould wall, and

liquid steel overflows it (stage 4, Fig. 3.3a) to form a subsurface hook. On

the other hand, with oscillation marks having no sub subsurface hooks, the

skin is weak and behaves more like a liquid. Thus, at beginning of positive

strip, the top of the skin is easily pulled back with the liquid toward the

mould so that overflow does not occur (stage 4, Fig. 3.3b). In addition,

formation of pronounced hooks may cause weakening of the subsurface

cast structure, due to the presence of non-metallic inclusions and/or

segregated structures in the vicinity of the hook region.

During solidification in the mould, heat transfer from the strand shell surface

to the mould is probably the least understood and most complex of the

heat transfer steps[11 ]. The salient feature of this heat transfer is the liquid­

solid shrinkage and the resulting tendency for an air gap to form between

the strand shell and mould surface. As the air gap is formed, heat transfer

changes from mainly conduction to radiation across the gap with a

resulting decrease in heat flux. This effect can be reduced by designing

 19

the mould space with a taper of the narrow faces of the mould to follow

the shrinkage in the cross section, while the broad faces are set parallel to

each other. In addition, the complexity of heat transfer at the mould inner

surface may be exacerbated by introducing mould powder into the mould

and also by the formation of oscillation marks at the strand surface.

Although the mechanism of heat transfer in the mould is complex, for

practical purposes, the average heat transfer from the mould can be
calculated from a heat balance applied to the cooling water which
accom m odates the high heat transfer rates resulting from the solidification
process. The average heat transfer qm (heat flux), due to strand shell

formation and cooling process in the mould, can be described as[l 1]:

M wCw
qm = ----------- ( To - Tf) (3.3)
L.F

thus,

jCV.QyyCTw
qm ( T o- T j ) (3.4)
L.F
 20

where, qm heat flux of mould cooling (W/m2)

Mw mass-rate of water (kg/sec.)
pw density of water (kg/m3)
Qw flow rate of water (m3/sec.)
CvV specific heat of water (J/kg/K)
L mould length (m)
F mould width (m)
T J o water temperature in,out (K)

After the strand shell is formed and primary cooling occurs in the mould,
additional heat is removed in the secondary cooling zones below the
mould where solidification is completed.

3.2. Secondary Cooling

Secondary cooling of the strand has a decisive influence on the degree to
which the strand surface and center remain crack free[l 1]. The purpose of
secondary cooling is to produce and maintain a strand shell having
sufficient strength to withstand bulging, bending and straightening below
the mould, and is provided in cooling zones consisting of a series of water
sprays and water cooled support roll units (see Fig. 3.4). The main heat
transfer functions of the water system in the secondary cooling zone are to
provide:
(a) the required amount of water to complete the solidification process,
(b) the capability to regulate the thermal conditions of the strand from
below the mould to the cut-off operation, and
(c) cooling of the containment rolls.
 21

Fig. 3.4 Diagram showing heat extraction during

secondary cooling

It is necessary to control both the temperature levels and thermal gradients

in the strand to avoid the occurrence of surface and internai defects such
as improper shape and cracksfl 1]. At high temperature, the strength
properties of the steel shell are critical to the ability of the shell to withstand
the external and internal forces that are imposed by the casting operation.
In particular, the ductility of steel close to the solidus temperature and to
the temperature at which the y/a transformation occurs, is low and at these
temperatures, the shell is therefore susceptible to crack formation. It is also
important to control temperature gradients because consequential
thermal strains can exceed the strength of the steel resulting in cracks.
 22

Excessive thermal strains may result from changes in the heat extraction
rate by over cooling. Thus, in the design of a secondary cooling system,
the thermal conditions along the strand must be established to satisfy the

required product integrity and quality. For these purposes, the surface
temperatures along the strand are generally specified to be above 900 °C,
particularly for avoiding the formation of surface defects such as transverse
cracks which can occur due to lack of ductility in relation to the y/a
transformation. Based on this inform a tio na l], the cooling rate along the
strand may be determined from heat transfer equations. Important
parameters in these calculations include the convection heat transfer
coefficient of the water spray and the water flux which is the amount of
water provided per unit area of surface contact per unit time. In addition,
changes to the water flux can be m ade to compensate for changes in
casting conditions, such as casting speed, strand surface temperature and
steel grade.

Many investigations!! 2,13] have been concerned with determination of

the heat transfer coefficients under virtual production conditions. The
results of these studies show that the quantity of heat extracted is largely
dependent on the water flux Vs (l/m2/sec), of spray water. The values of
heat extracted qs (W/m2 ), by spray water cooling can be described by
the empirical relationship between qs and Vs:

qs = 160.000 (Vs10J5 (3.5)

 23

Recently, the spray intensity of secondary cooling water has been

considerably reduced in most continuous casting operations to obtain the
required quality of cast steel slab, particularly surface quality[14]. As a
result, there is an increase in the radiation com ponent of the total heat
extracted due to effective replacement of a portion of the spray cooling

with the radiant cooling.

The quantity of heat extracted due to radiant cooling qr (W/m2), is largely

dependent on the surface temperature Ts (K), of the strand, and may be
expressed [15]:

qr = £<y (Ts4 -Te4) (3.6)

where, s : emissivity
a : Stefan-Bolttzman constant (5.67x10 ~8, W/m 2 K4)

Te : surrounding temperature (K)

In addition, it is neccessary to consider the heat extracted by a series of

water cooled support rolls to determine the strand surface temperature.

Practically, the heat transfer coeficient of the contact between the strand

surface and the rolls in this case can be expected to be in the region 300 -

1200 W/m2.K[l 6].

 24

3.3, Processes at the Solidification Front

During continuous casting, solidification commences with nucléation of

dendrites of 6-phase. Crystals becom e attached to the gradually forming

solidification front and have compositions dictated by the solidus line of

the phase diagram (see Fig. 3.1). Owing to the increase in concentration

of alloying elements such as carbon, manganese and silicon in the vicinity

of the solidification front, it is here that the liquid temperature is lowest.

However, the temperature in the liquid phase is higher than the

temperature of the solid phase so that heat will flow from the liquid phase

to the solidification front and if th!e temperature of the molten steel

decreases to below the liquidus temperature, constitutional supercooling

occurs and solidification commences within the liquid pool. Once the

superheat has been removed, usually in the region 1 to 3 m below the

meniscus, a two phase layer of liquid plus 6-phase crystals, which may be

up to 5 mm thick, forms on the solidification front[l 7].

In the region where superheat is still present, formation of the fine-grained

edge zone is followed preferentially by dendritic crystallization. After

removal of the superheat from the liquid phase, equiaxed crystals form in

the vicinity of the solidification front, and as they have higher density than

the remaining liquid metal, they sink and sedimentate. Convection

 25

currents in the liquid core of the strand exert some influence on this process.

The rate of solidification and the temperature gradient in and at the

solidification front are the main factors affecting dendritic crystallization

but the factors which cause change in the solidification morphology from

dendritic to equiaxed crystallization have yet to be accurately defined.

Nevertheless, low temperatures in the tundish (low level of superheat) can

be expected to result in larger regions of equiaxed crystallization [1 7,18].

The thermal behavior and microstructural evolution in continuously cast

steel have been investigated in numerous studies. At the high temperature

at which continuous casting is carried out, defects such as cracks

associated with phase transformation and microstructural changes tend to

form during the process. Normally, these defects are not acceptable in the

final products and therefore, to prevent such defect formation, it is

absolutely necessary to understand the mechanism of genesis of these

defects. The defects that commonly occur in continuous casting and

relevant preventive measures are described in Chapter 4.

 26

Chapter 4
DEFECTS IN CONTINUOUSLY CAST STEEL SLAB

Continuous casting-direct rolling techniques have been adopted

increasingly in recent years to improve both efficiency and yield
productivity in modern integrated steel production plant. However, to
operate these processes effectively, it is necessary to prevent the cracking
that often occurs on the slab surfaces.

Surface cracking on continuously cast slab can occur in various forms of

which longitudinal cracks, star cracks and transverse cracks are
significantly important[19]. All these surface defects 'have undesirable
influences on the product because they can degrade quality as well as
increase the risk of failure during application of the material in finished
products. Figure 4.1 shows the appearance of surface defects in cast slab.

Fig.4.1 Schematic representation of surface cracks in cast slab

 27

Although almost all surface defects can be removed by scarfing the slab,

this practice may in fact increase defect probability in finished products,

particularly if the scarfing shells are not removed completely and still

remain in the slab surface. Additionally, scarfing involves a degree of

material loss and other expenditure including labor costs. Therefore,

preventing surface defects is most important to assure the consistency of

steel product quality as well as to achieve efficiency of the process.

4.1. Longitudinal Surface Cracks

A longitudinal surface crack is a crack which occurs, on the surface of slab,

parallel to the casting direction and usually originates in the m ould[l 9,20].

This defect occurs as a result of weakening of the grain boundaries during

formation of the strand shell. It has been suggested that fluctuation of

temperature due to non-uniformity of heat transmission below the mould

meniscus probably contributes to formation of the defect. This

phenomenon might occur if any defect such as scratch or a non-uniformity

inflow of mould pow der occurs in the mould thereby resulting in diminished

co nta ct between the strand shell and the mould. As the heat transmission

resistance between the strand surface and the inner face of the mould

decreases in the small area just below the meniscus, heat conducted from

the interior liquid metal will reheat the solidified shell. At the same time,
 28

thermal shrinkage and ferrostatic pressure applied to this area results in

crossectional stresses in the casting direction leading to weakening of the

grain boundaries.

To prevent longitudinal surface cracking, it is necessary to control the

uniformity of the heat transmission[2 0 ,2 1 ] by maintaining the mould surface

regularity as well as controlling the uniformity of mould powder inflow.

Control of these factors limits the incidence of longitudinal cracking to a

few percent only of slabs.

4.2. Star Cracks

Star cracks are probably caused by the phenomenon of hot shortness in

which, for example, a low melting phase penetrates the casting through
I
the grain boundaries[23]. In view of the high strand shell temperature, the

penetrating phase remains liquid and causes weakening of the inter­

granular bond and in the continuous casting context, hot shortness is

usually attributed to copper released through abrasion of the mould plates.

In most cases, segregated copper has been found in and near the star

crack in the cast slab confirming the proposal that embrittlement of the

cast slab at a high temperature results from action of copper derived from

the mould copper plate .

 29

In early practice, the inner face of the mould was chromium plated to

prevent release of copper by mould abrasion. The plating, however,

detached from the lower part of mould surface after long-time operation,

resulting in occurrence of star cracking[23,24,25]. Therefore, regular control

of coating thickness over the mould surface could be effective in

preventing star cracks. Generally, as with longitudinal cracks, these defects

occur in only a few percent of continuously cast steel slabs.

Although these and other surface defects do occur in continuously cast

slab, transverse cracks are probably the most com m on and therefore have

attracted numerous studies and still continue to be studied .

4.3. Transverse Cracks

At PT. Krakatau Steel, transverse cracks occur in approximately 68% of the

total production of continuous cast steel slabs, mainly in 0.12-0.18 %C

steels. Insignificant cracking occurs in steel containing less than 0.09 %C.

The incidence of these defects is most easily observed after the oxidized

surface layer on the narrow side of the slabs is removed by scarfing. To

obtain the better understanding of effective preventive measures for

transverse crack prevention it is necessary to understand the mechanism of

genesis of the defect.

4 .3 .1 . A p p e a r a n c e of T r a n s v e r s e Crac ks

In continuous slab casting, transverse surface cracks usually occur on the

corners of the slabs (see Fig.4.l), perpendicular to the casting direction,

and usually along the valley of the oscillation marks. In general, these

cracks penetrate from a few mm to about 10 mm into the surface, and

occur only on the top face of cast slab, on which a tensile stress is

generated during straightening [26,27]. In addition, microscopical study

indicates that the cracks are mostly intergranular. These observations

suggest that the occurrence of a transverse crack is a consequence of

interaction between the straightening strain and the proof stress of the slab

material. Based on this consideration, models have been developed to

explain the mechanism of transverse cracking in relation to the effects of

grain boundary embrittlement.

4.3.2. Mechanism of Transverse Cracking

In this section, the mechanism of genesis and means of prevention of

transverse cracking is described using models which emphasise the factors

that significantly affect the formation of cracks and the relationship with

limitations of practical processes.

 31

By com bining plant experience and fundamental knowledge, as discussed

by Brimacombe and Sorimachi[28], the models of cracking are based on

i
relationships between cracking, the formation of the coarse austenite

grains during solidification, and subsequent precipitation of carbides

and/or nitrides at the austenitic grain boundaries.

4.3.2.1. Effect of Coarse Austenitic Grain Sructure

The surface cracking susceptibility of low alloy steel slabs during continuous

casting processes depends largely on the carbon content, and is a

maximum in the range 0.1-0.15% (referred to as a medium range). This

phenomenon has been studied[29] in relation to the peritectic reaction

during solidification in the mould. In this study it was confirmed that the

primary solidified shell for 0.1-0.15% C becomes uneven in the mould due

to shrinkage associated with transformation of ^-ferrite to austenite. The

shrinkage results in a marked reduction of heat extraction owing to lack of

close co nta ct between the metal and the mould. Consequently, the

temperature gradient between the liquid steel and the thinner parts of the

unevenly solidified shell is quite small, the thin metal is close to the solidus

temperature, and it has low strength and/or ducTility. Therefore, it will be

susceptible to cracking. In addition, a local delay of cooling in the thinner

parts of the unevenly solidified shell result in larger grains of austenite in the
 32

surface region of the cast strand. This phenomenon is also to be noted as

the second factor which causes high transverse cracking susceptibility.

The combined effects of uneven solidification and metallographic

changes with carbon content are shown in Fig.4.2

L+6 or
y L+K+&
K. l+K.
Mould 1 | Mould j J

Liq

(b)

Fig. 4.2 Schematic representation of the difference of y grain size

between (a) low carbon steels and (b) the medium carbon
steels, owing to uneven solidification in the mould [29].

Figure 4.2 shows that if the contact between the solidified shell and the

mould is poor due to thermal shrinkage, as occurs in medium carbon

steels, a coarse grain structure will be produced in the thin solidified shell.
 33

On the other hand, if the contact is good, as is known to occur for low

carbon steels, the grain structure will ¿e fine.

Although the carbon content corresponding with the maximum austenite

grain size agrees well with the carbon content for maximum surface

cracking susceptibility (Fig. 4.3), it is considerably less than the 0.18 %C

peritectic point in the Fe-C constitutional system. This can be explained in

terms of the effects of alloying elements such as Mn, Si, Ni and etc. which

reduce the carbon content at the peritectic point. The effect of these

elements can be represented as the carbon equivalent (CE), and

described using the empirical relationship [29]:

CE (96) = %C + %Mn/6 + %SU24 + 96NU40 + 96Cr + 96Mo/4 + 96V/14 (4.1)

The carbon dependence of austenitic grain size also corresponds to that

for ductility loss and is in good agreement with the variation of austenite

formation temperature in the Fe-C constitutional diagram, as shown in

Fig. 4.3.
 34

C (%)

C (%)

C (%)

Fig.4.3 Diagram showing the dependence on carbon content of (a)

surface cracking frequency (b) ductility (RA) and y grain size
( D r ) of as cast steel and (c) their relation to the peritectic
transformation [28].

4.S.2.2. Effect of y/a Transformation

Practically, transverse surface cracking of continuously cast slabs occurs

mainly at low strain rates in the straightening operation[30]. The total strain

at the slab surface is estimated to be fairly small, probably less than a few

percent (see Sec. 2.2). The strain to which the strand is subjected becomes

particularly critical as it passes through the Ar3 temperature (850-750°C), at

 35

which the y/a transformation occurs. It has been suggested[31,32] that

transverse cracking occurs at the \Ar3 temperature and is related to

intergranular embittleme.nt caused by allotriomorphs of ferrite precipitating

along the austenitic grain boundaries. The embrittlement phenomenon

occurring at Ar$ has been $tudied[33,34] by hot tensile testing (Fig. 4.4.).

Fig.4.4 Diagram showing the influence of test temperature on hot

ductility [34].

Figure 4.4 shows hot ductility as a function of test temperature for low alloy

steels, and indicates that ductility decreases to a minimum value at about

800 °C. This temperature approximates to the beginning of y/a

transformation that occurs at the austenitic boundaries of the coarse­

grained cast material. Additionally, there is a sharp decrease in the values

for reduction of area below 600 °C.

 36

The tensile stresses generated during straightening will concentrate strains

preferentially within the soft ferrite seams, and therefore, plastic

deformation will occur predominantly in this ferrite[34,35]. Plastic

deformation will then be followed by (austenite) grain boundary sliding

causing wedge type cracks at the grain boundary triple points. Final

fracture will occur as the result of micro-void coalescence. From these

consideration, a mechanism of intergranular ductile fracture of austenite

with dynamic precipitation of carbide and/or nitride particles has been

postulated as shown in Fig.4.5.

<r

Fig.4.5 Schematic diagrams showing (a) the dynamic precipitation of

carbides and/or nitrides on austenitic grain boundaries, (b)
nucléation of allotriomorphs of ferrite and strain concentration
within soft ferrite seams along austenitic grain boundaries in the
initial stage of deformation, (c) microvoid formation by de­
cohesion of precipitate/matrix interfaces and (d) coalescence of
microvoids resulting in ductile intergranular fracture of austenite
[35].
 37

4.3.3. Prevention of Surface Transverse Cracking

It should be emphasised again that the transverse surface cracking which

occurs in continuously cast steel slabs depends largely on the austenitic

structure a nd /o r the presence of ferrite at the austenitic grain boundaries.

Therefore, control of structure and suppression of precipitation should be

the first priority in preventing this defect.

4.3.3.1. Control of y Structure

As discussed in Section 4.3.2.1, the grain size of the austenitic structure

significantly affects transverse cracking susceptibility. The relationship

between the final grain size, Dy, and cooling rate, dT/df, has been studied

[36] using an empirical relationship developed from the nucléation law of

solidification of cast iron, and it was found that the final grain size can be

described by the equation:

Dy =0.336 (d lld t) ]i2 (4.2)

Equation (4.2) shows that a fine-grain structure can be achieved by

increasing the cooling rate. Therefore, there is theoretical justification for

reduction of transverse cracking susceptibility of continuously cast steel

 38

slab (associated with the coarsening ot the structure), by increasing the

rate ot cooling ot the slab.

Although increasing the cooling rate can suppress austenitic grain growth,

in practice it is not feasible to achieve sufficiently rapid cooling before the

end of grain growth at temperatures above 1300 °C in the limited mould

cooling. Consequently, under the limited cooling condition in practical

continuous casting, the austenitic grain size after solidification depends

largely on the concentrations of alloying elements, including carbon.

Therefore, it is possible that effective prevention of surface cracking might

be possible through addition of small amounts of micro-alloying elements

such as titanium, boron and zirconium which can refine the austenitic grain

structure. Addition of titanium can result in precipitation of TiN during

solidification[37], thereby increasing the number of nuclei resulting in grain

refinement. Moreover, at lower temperatures, the precipitation of TiC on

austenitic grain boundaries will suppress austenitic grain growth and result

in a grain refinement effect. It is unlikely that the precipitates of TiC will

prom ote form ation of allotriomorphs of proetectoid ferrite as they may

retard the y/a transformation (see Sec. 4.3.2.2J. Both boron and zirconium

additions have similar effects in the grain refinement of austenite, but these

elements are seldom used for econom ical reasons.

 39

The Influence of titanium, boron and/or zirconium in reducing transverse

cracking susceptibility associated with the grain refinement, is shown in Fig.

4.6.

600 700 800 900 1000

Def ormat i on t e mp e r a t u r e . °C

Fig. 4.6 Hot ductility at low strain rate for low alloy steels containing
microalloying additions [37]

Figure 4.6 shows hot ductility determined at low strain rate such as occurs

in continuous casting of low alloy steels containing micro-alloying

additions. Clearly, the addition of a small amounts of titanium, boron or

zirconium can significantly improve ductility of steels in the critical

temperature range of 850-750 °C.

4.3.3.2. Control of Precipitation

The second factor controlling transverse cracking Is dynamic precipitation

of carbides and/or nitrides at the austenitic grain boundaries. Suppression

 40

of precipitation should be effective in preventing transverse surface

cracking, since precipitates such as niobium and vanadium carbides

and /o r aluminium and vanadium nitrides which are formed at

temperatures close to Ar 3, will prom ote the formation of aliotriomcrphs of

proetectoid ferrite and cause embrittlement at the grain boundaries[38].

If the secondary cooling pattern below the mould can be controlled to

avoid dynam ic precipitation, the formation of proeutectoid ferrite might

be avoided thereby preventing embrittlement at the boundaries.

However, as it is difficult to control the secondary cooling due to the

limitations of the production process, this means of control is not

practicable. Consequently, in practice, the precipitation of carbides

and/or nitrides does frequently occur in the grain boundary regions.

However, addition of titanium has a strong effect in the refinement of

austenitic structures during solidification, and may also retard the y/a

transformation resulting in reduction of transverse cracking susceptibility by

suppressing the formation of allotriomorphic ferrite on the grain boundaries

[38].

An alternative m ethod for prevention of transverse cracking caused by

intergranular embrittlement associated with y/a transformation, is control of

 41

during straightening to above Ar 3, precipitation of allotriomorphs of ferrite

at the austenitic grain boundaries can be avoided altogether thereby

removing the embrittlement effect of strain in the straightening operation.

In practical processing, therefore, the surface edge temperature during

straightening must be controlled, if possible to higher than 900 °C [39].

Control of aluminium and nitrogen content and their solubility product AlxN

is also feasible in a production environment, as are taper, m oud powder,

spray design, spray arrangement, spray condition and m achine alignment.

The opportunity for extensive experimentation in the continuous casting

plant is very limited. Additions of small amounts of titanium and control of

surface edge temperature of the strand to above 900 °C are practicably

the only sensible preventive measures that can be exercised.

Measurement of the surface temperature as the strand progresses from the

tundish to the straightener is essential to control of cracking, but very

difficult to achieve directly. On the other hand, if is possible to calculate

these temperatures using a m athem atical m odel of the cooling process as

described in Chapter 5.
 42

Chapter 5

MATHEMATICAL HEAT TRANSFER MODEL FOR

SOLIDIFICATION OF CONTINUOUSLY
CAST STEEL SLABS

As has been discussed in Chapter 4, the quality ot continuously cast steel

slab depends largely on the operational conditions, particularly the

cooling processes from com m encem ent in the mould to completion of

solidification in the secondary cooling zones. Operational experience has

shown that appropriate control of the cooling processes, particularly in

spray cooling, is essential for achieving the proper quality of the cast

products [39]. Overcooling can lead to intergranular embrittlement and

associated transverse crack defects due to transformation which occurs

below Ar3. To minimize that cracking, the strand surface must be

m aintained in the austenitic range or, in practice, above 900 °C, until after

the straightening operation (refer to Sec. 4.3.2.2). With this limitation in

mind, and due to lack of appropriate operating plant and experimental

facilities, direct measurements of temperature distributions and

solidification rates on moving castings are very' difficult to make.

Consequently, it has becom e necessary to a do pt an alternative approach

to obtain thermal and kinetic data. The most appropriate alternate

 43

approach is to m athem atically simulate the heat transfer in a continuously

cast section, and then calculate the temperature distributions as a

function of the controllable variables of the process.

Simulation of heat transfer during solidification requires that a nonlinear

m athem atical problem be solved. As pointed out by Ruddle [40], there

are two m athem atical approaches to the problem, the analytical

approach and the numerical approach. While the analytical approach is

certainly the more elegant of the two, it does require a number of inexact

assumptions because of the complexity of the problem. Simplifying one

or more of these assumptions, such as invoking invariant thermophysical

properties, constant heat-transfer coefficients, and linear temperature

profiles in the shell, can introduce considerable uncertainly in the validity

of the results. Thus, this approach is not favoured. On the other hand,

numerical solutions, which are considerably more versatile, appear to be

better suited for solving solidification problems. Complex variations in the

boundary conditions and variable thermophysical properties can be

handled readily with this technique. Fortunately, numerical computations

to the problem can be obtained quite readily with use of the digital

com puter.
 44

In this present study, a m odel ot unidimensional heat transfer in

continuously cast slab is presented. The m ethod of solution using a digital

com puter is also included. Calculated temperature distributions during

secondary cooling along with attempts to verify the model are discussed.

5.1. Mathematical Model

As has been discussed in Chapter 3, the solidification of a slab during

continuous casting occurs as the slab passes through three distinct zones

of cooling. Accordingly, the m athem atical model developed for

simulating the process consists of three parts: solidification in the mould

zone, solidification in the spray cooling zone and solidification in the

radiant cooling zone.

A m athem atical m odel of the cooling process can be developed from a

heat balance on a horizontal slice of cast slab over the time period

required for the slice to proceed from the liquid meniscus in the mould to

the cut off station[41]. As the slice moves downward, heat is conducted

from the center line to the surface of the slab at a rate governed by the

surface boundary conditions and thermophysical properties of the metal

in the slice. As a consequence of heat loss from the surface, a

tem perature profile exists in the slab and as the temperature decreases
 45

progressively, so that profile will change. The m odel enables the profile

and the change in profile to be estimated in terms of the imposed

boundary conditions. The m odel is based on a heat balance derived from

the partial differential equation describing the unsteady state conduction

of heat in a m edium m oving at velocity u in direction z (casting direction)

that can be described[41 ]:

37 37 3 37 d dT 3 37
p .c — + p .c .u — - [ - (k — ) + - I k - ) + ~ ( k - ) ] = o (5.1)
dt 3z dx dx dy 3/ 3z 3z

where, p : density of steel (kg/m 3)

c : specific heat of steel (J/kg.K)
k : thermal conductivity of steel (W/m.K)
7 : tem perature (K)
x,y: transverse directions.

As p ointed out by Hills[41], conduction of heat in the transverse (x)

direction is very much greater than conduction of heat in the withdrawal

(z) direction. Also, conduction in the other transverse ( y) direction can be

justifiably ignored. As the horizontal slice moves dow nward at the velocity

of the casting, the relative velocity of the slice is zero and the bulk heat

transfer term therefore vanishes. Consequently:

 46

ÔT
( k — ) = Ot (5.2)
dz

■dT
( k — ) = 0, and ? (5 .3 )
dy

dT
p.c.u — = 0 (5 .4 )
az

Substituting Eqs. (5.2), (5.3) and, (5.4) into Eq. (5.1 ) leads to:

dT d dT
p c . — - — (k — ) = 0 (5.5)
df dx dx

Variation of thermal conductivity with temperature can be taken into

account by expansion of the second term of Eq. (5.5) giving:

»
dT m dT
CM
-Û

p.c. ----- k
o

(5.6)
h
1

dt dx2 dx

where b is the rate of change of thermal conductivity with temperature.

To enable digital programming of the differential heat transfer expression,

it is first necessary to reduce it to algebraic equations. The finite difference

 47

technique, w hich is one of many numerical schemes available, is suitable

for this purpose and is described in Section 5.2.

5.2. Numerical Solution

To a d a p t the m odel described in Eq. (5.6) for numerical analysis solution,

the horizontal slice is subdivided into a number of small elements each of

'w hich is represented by a node located at the center of the element. The

nodes are respectively identified as 7 at the center, 2, 3...... . n-7 and n

the surface element.

Figure 5.1 shows that the half slab slice passes through three distinct zones

of cooling with respective extracted heats. These extracted heats are qm

due to cooling in the mould, q s spray cooling, and qr due to radiant

cooling. In addition, the containm ent rolls are involved in the extraction

of heat qrc, by conduction during c o n ta ct with the strand surface. At time

f = o, the tem perature of each node is fixed by the tim e boundary

condition that the tem perature profile of the slice at the meniscus is equal

to the incom ing metal temperature. At the next tim e step At, the

temperatures of nodes 2 through (n-7) are calculated using the finite

difference form Qf the conduction equation and the known temperatures

for nodes at tim e f = 0. The temperatures at nodes n are then calculated

 48

using the surface boundary conditions and the temperatures of the

remaining nodes as calculated at the next time step At

Fig. 5.1 Node arrangement and boundary conditions of half slab slice

Under limitations of the assumptions of the mathematical modelling, the

calculation is repeated with the appropriate thermophysical properties of

the elements until the completion time is reached[41 ].

 49

5.2.1 Determination of the Temperatures at Nodes 2 through (n-1)

Developm ent of the numerical analysis solution of the m odel for

calculating the temperatures of nodes 2 through (n-7) at time t from

known node temperatures calculated at tim e (t-At), is derived from the Eq.

(5.6), expressed in the form:

dT l d2T dT
— [ k ( ------ ) + b ( - ) 2 ] (5.7)
dt p. c dx2 dx

Consider a slice w hich is to be divid ed into equal elements in the x

direction (refer to Fig. 5.1). The nodal points are designated as shown by

the J locations indicating the x increment. Then over the next time step,

At, the tem perature is calculated using the finite difference approxim ation

of Eq. (5.7), where[4l ]:

dT 77 - Tj
— = --------- , (5-8)
dt At

m (Tj+i - Tj )/Ax - [Tj- Tj., )/Ax

= Tj+, - T,., - 2 T j, and (5.9)
dx2 Ax.2

dT Tj+|- Tj.;
(------ )2 = (------------- )2 = Tj+,2- 2Tj+i.Tj+, + Tj.,2 (5.10)
dx 2 Ax
Subtituting Eqs. 5.8, 5.9 and 5 .10 into Eq. 5.7 leads to:
 50

77 -Tj 1 !-
-------- ---------- [K(Tj+r Tj.r 2Tj) + b/4(Tj+J2- 2Tj+].Tj+i + Tj-i?)] (5.11)
At Ax2,p c

where, Ij : temperature of J node calculated at time f, and.

T/ : temperature of J node calculated at time t+At

Ihus, the explicit form of the finite difference approximation tq. (5.11) for

calculating the interior node temperatures is

At
Tj = T j + -------- [K(TJ+r Tj-r 2Tj) + b/4(Tj+12- 2Tj+1J j +} + Tj-i2)] (5.12)
Ax2,p c

5.2.2. Determination of Strand Surface Temperature

The temperature of node n, at the strand surface, can be calculated using

an equation which is derived from the heat balance for the heat-in, qfn,

heat-out, q out, and accumulated heat on a half element n over the time At,

as shown in Fig. 5.2.

H eat-out

Fig. 5.2 Fleat balance on a half element.

 51

As shown in Fig. 5.2, the heat balance on a half element can be described
as

[A ccum ulated heat] = [Heat-in] - [Heat-out] (5.13)

For the heat which flows in direction, x, the following differential form of Eq.

(5.13) is appropriate:

UT Cjin | fx=x)- A y . A Z - C fo u t I fx = x + A x l'A y .A Z Qin j (x=Oj Qouf I (x-O+Ax)

p.c — = --------------------------------------------- = ----------------------------------- (5.14)

at Ax.Ay.Az Ax Ax

Equation (5.14) may be expressed in differential equation form:

dT dq
p .c — = (5.15)
dt dx

and forx = Ax/2 :

dl dq
P -C ----- = --------- | 2
(a x / } (5.16)
dt dx

Since dq can be expressed as k(dT/dx) - qout, Eq. (5.16) becomes

dT k[dT/dx) - qout
p .c — = ------------------- (5.17)
dt Ax/2
 52

thus,

dï 2
_ = ------------[k[dTId/) - q0ut] (5.18)
of Ax.p.c

The strand surface temperature carj now be calculated using Eq. (5.19)

derived from Eq. (5.18) using respective boundary conditions of the heats

extracted in each cooling zone as a function of time. For surface nodes n,

this equation, expressed as a finite difference, is:

2AI
Tn = T n + ------------ [k {Tn-i - Tn) -Ax.q0ut] (5.19)
zlx2.p.c

where, Tn, IV: strand surface temperature calculated at times t and t+At

5.3. Application of the Model to Continuous Casting of Slabs

As explained in Chapter 4, it is necessary to control the strand surface

temperature at above 900 °C to avoid transverse surface cracking

associated with intergranular embrittlement associated with the y/a

transformation below the Ar3 temperature. Therefore, control of spray

cooling during continuous casting is very important to minimize the

cracking.
 53

With the m athem atical model that has been developed in this present

study, control of the strand surface temperature during continuous casting

can be simulated for a given set of casting parameters. The model may

be used to assist in designing the cooling process, particularly if the spray

cooling is to be changed to improve the cooling process.

Development of the computer solution of the model is described in

Appendix A. The m athem atical model was coded to the Q-Basic

language for solution using Compal 486 computers. To run the

m athem atical model, the following input data were required:

(a) section size,

(b) casting temperature, speed and time,

(c) liquidus temperature of the steel, .

(d) water flux of spray cooling in each cooling zone, and

(e) thermophysical constants and operational conditions.

This study, was limited to a casting 200 mm thick, 1200 mm wide, of 0.16­

0.18% C steel for which the casting machine is used at PT. Krakatau Steel

and for which the transverse cracking problem is most prevalent. The

parameters used in the study are described in Chapter 6.

 54

Chapter 6

EXPERIMENTAL

Despite recent significant advances in continuous casting technology,

transverse surface cracking remains one of the most common defects

encountered during the production of continuously cast steel slab in most

world w ide plant. At PT. Krakatau Steel, the incidence of these defects is

mainly restricted to steel containing 0.12-0.18% of carbon, and is

insignificant for steel containing less than 0.09% C.

As discussed in Chapter 4, the presence of transverse cracks has

undesirable influences on the product because they can degrade quality

as well as increasing the risk of failure during applications of the material in

finished product. At present, the problem of transverse cracking is

attacked by removing the defect by scarfing the corners of the slab,

which involves a degree of material loss and associated expenditure.

Therefore, decrease in the incidence of transverse cracks is dem anded to

obtain a significant yield increment and to improve process efficiency.

Consequently, examination of these defects and determination of means

 55

of controlling the incidence of them is a matter of considerable

importance.

Theoretically, the opportunity to meet requirements of a stable quality of

slab greatly depends on the capability of the continuous casting machine

as well as the operational conditions of the process. These conditions

comprise the radius of the cooling strand, strand size, steel grade, casting

speed, distribution of cooling water, tundish temperature and strand

surface temperature. However, as the opportunity for extensive

experimentation in the continuous casting plant is very limited, the only

variables that could be examined in this study of transverse cracking were

operational process parameters which can be controlled in practice, and

that can be studied without interfering with production schedules.

Therefore, experimentation was carried out under this limitation. The two

parameters which were investigated were believed to strongly influence

the formation of transverse cracking. These parameters were:

(a) composition of steel, and

(b) water rate, with limited variation possible only for control of the strand

surface temperature.
 56

As is believed, addition of small amounts of titanium to the cast steel might

be effective in preventing the incidence of transverse cracking associated

with austenitic grain growth. Accordingly, to examine the effect of

titanium additions, the steels chosen for metallurgical study of transverse

cracking had four different titanium contents with similar carbon,

manganese, and aluminium contents, and other undesirable elements

such as phosphorus, sulphur and nitrogen which remain in steel in limited

concentration. The specified titanium contents of steel for this purpose

were:

(A) without titanium

(B) 0.005- 0.010%Ti,

(C) 0.015-0.020% Ti, and

(D) 0.025-0.030% Ti

Compositions of the steels are set out in Tables 6.», 6.2, 6.5 and 6.6.

For the purpose of controlling the strand surface temperature during

continuous casting, the rate of spray cooling water for strand cooling is

controlled through a flow-meter panel, located in the control room, and

designed to distribute water in the six secondary cooling zones comprising

the spray-ring, the zones-1 A and 1B, the zone-2, the zone-3 and the zone-4,

as shown in Fig. 6.1.

 57

Fig. 6.1 Schematic diagram showing distribution of water cooling

in the secondary cooling[43]

The transient heat transfer model developed in Chapter 5 to simulate the

strand cooling process was used to determine how the strand surface

temperature responds to variation of the secondary cooling water rate

during the continuous casting process.

 58

6.1. Methods

The hypothesis that both steel composition and straightening temperature

are the most important factors influencing the susceptibility to transverse

cracking d ictate d the experimental methodology.

Initially, a statistical survey of accum ulated production data was m ade to

identify susceptible alloys, and to determine the critical straightening

temperatures which are associated with poor ductility of the strand. For

this, the data were classified into three steel grades with compositions:

(a) C < 0.06%

(b) 0.07 < C <0.09%

(c) 0.12 < C <0.18% •

The incidence of transverse cracking in these steels had been routinely

determined by removal of the oxidized surface layer on the corner region

of the slabs followed by visual assessment. Results of the statistical surveys

are shown in Figs. 7.1 and 7.2 in Chapter 7. The survey clearly established

that the susceptibility of transverse cracking of the. cast steel slabs (c)

containing 0.12-0.18% C is very high, whereas the susceptibility of the steels

(a) and (b) is (negligibly) low. Therefore, the relationship between

straightening temperature and transverse cracking was examined only for

steel grade (c).

 59

The steel used for this study was a grade frequently produced at PT.

Krakatau Steel cast with the specified casting conditions:

(i) casting (tundish) temperatures of 1530-1550 °C

(ii) casting speed of 0.9 m/min., and

(iii) strand size of 200 mm thick, 1200 mm width

and having composition shown in Table 6.1.

Table 6.1 Chemical composition of steel for metallurgical

study of transverse cracks

Elements wt. %

C 0.16-0.18
Mn 0.80 - 1.00
A! 0.030 - 0.060
P 0.025 max.
S 0.020 max
N 0.0070 max.

6.1.1. Effect of Steel Composition

As discussed in Section 4.3.3.1, suppression of austenitic grain growth

during solidification, by addition of small amounts of titanium to refine the

grain structure during solidification, could be effective in preventing

transverse cracking associated with coarsening of the cast structure.

 60

To examine possible beneficial effects of titanium, four steels A, B, C and D

with compositions shown in Table 6.2, were examined with a secondary

cooling water rate of 0.756 l/kg, as specified for normal production

operation as shown in Table 6.3.

Table 6.2 Composition of the steels cast with water rate 0.756 l/kg

Chemical composition (wt. %) Nr. of

Steels Ti C Mn Al P s N slabs

A1 - 0.17 0.78 0.034 0.006 0.009 0.0046

A2 - 0.18 0.85 0.048 0.006 0.007 0.0054 17

A3 - 0.16 0.85 0.047 0.007 0.006 0.0036

B1 0.007 0.17 0.89 0.043 0.008 0.006 0.0048

B2 0.009 0.17 0.87 0.052 0.008 0.007 0.0041 17

B3 0.006 0.16 0.87 0.052 0.007 0.009 0.0039

Cl 0.017 0.17 0.86 0.043 0.013 0.008 0.0053

C2 0.01 7 0.17 0.85 0.050 0.011 0.009 0.0037 17

C3 0.018 0.18 0.82 0.049 0.006 0.007 0.0054

D1 0.026 0.17 0.86 0.044 0.008 0.007 0.0056

D2 0.028 0.17 0.90 0.048 0.008 0.008 0.0040 16

D3 0.025 0.18 0.84 0.041 0.007 0.008 0.0049

 61

It should be noted that water rate Wn is indicated by the litres (I) of water

used per kilogram (Kg) of steel, determined from the relationship:

Vs
Wr = ----------=
— (6.1 )
Vc. t . F . p

Where, V5 : water spray (l/min.)

Vc : casting speed (m/min.)
f : slab thickness (m)
F : slab width (m)
p : density of steel (kg/m3)

Table 6.3 The normal rate of spray cooling wafer

Utilized water rate

Cooling zones (for cast, speed = 0.9 m/min.)

Vs (l/min.) Wr (l/kg)

Spray-ring 89 0.052

Zone-] A 0.214
o
o

Zone-1 B 206 0.122

Zone-2 284 0.169

Zone-3 206 0.122

Zone-4 129 0.077

To t a 1 0.756
 62

The incidence of transverse cracks on the upper surface edge of slabs of

the four steels was then determined as a function of the titanium content

of the steel, observed by removing the oxidized layer by a slight scarfing.

The grain size of the subsurface cast structure was measured.

6.1.2. Effect of Straightening Temperature

As discussed in Section 4.3.3.2, suppression of precipitation of the

allotriomorphs of ferrite during straightening by maintaining the strand

temperature at above 900 °C might be affective in preventing the

incidence of the transverse cracking. Accordingly, the possibility of

controlling the straightening temperature using an empirical relationship

(see Fig. 7.3) between the water rate and straightening temperature was

examined. It was possible to vary the water rate only slightly by an overall

reduction of the water rate in zones-3 and 4 to achieve reducing

the water rate from 0.756 l/kg as specified for normal production

operation, to 0.736 l/kg, without significant interference with production

practice, and with an expectation of obtaining a straightening

temperature higher than 900 °C. The distribution of the reduced water rate

is shown in Table 6.4.

 63

Table 6.4. The reduced rate of spray cooling water

Utilized water rate

Cooling zones (for cast, speed = 0.9 m/min.)

Vs (l/min.) Wr (l/kg)

Spray-ring 89 0.052

Zone-1 A 360 0.214

Zone-1 B 206 0.122

Zone-2 284 0.169

Zone-3 183 • 0.108

Zone-4 119 0.071

To t a 1 0.736

To examine possible beneficial effects of straightening temperature, steel

A was examined with the two different secondary cooling water rates of

0.756 l/kg and 0.736 l/kg. For the water rate of 0.756 l/kg, the experimental

d a ta were obtained using steel with compositions A l, A2 and A3, shown in

Table 6.2. For the water rate of j 0.736° l/kg, three slightly different

compositions of steel A were used as shown in Table 6.5.

 64

Table 6.5 Composition of the steels cast with water rate 0.736 l/kg

Chemical composition (wt. %) Nr. of

Steels Ti C Mn Al P S N Slabs

A4 - 0.18 0.89 0.044 0.008 0.009 0.0048

A5 - 0.18 0.84 0.048 0.010 0.008 0.0049 16

A6 - 0.17 0.90 0.050 0.007 0.006 0.0040

During straightening, the surface edge temperature of the slabs was

measured using an optical pyrometer. Subsequently, the incidence of

transverse cracks on the upper surface edge of slabs cooled with two

water rates was determined as a function straightening temperature, and

the grain size of the subsurface cast structure was determined.

6.1.3. Combined Effects of Titanium and Straightening

Temperature

As discussed in Sections 4.3.3.1 and 4.3.3.2, suppression of austenitic grain

growth and/or precipitation of allotriomorphs of ferrite can reduce the

susceptibility of transverse cracking. Accordingly, com bined control of

both effects by addition of small amounts of titanium to the steel and by

maintaining the strand surface temperature to above Ar3 might be

especially effective in preventing cracking.

 65

To identify the optimum com bination of the limited variables, the steels A,

B, C and D shown in Table 6.6 were cast using the water rate 0.736 l/kg (see

Table 6.4).

Table 6.6. Composition of the steels cast with water rate 0.736 l/kg

Chemical composition (wt. %) Nr. of

Steels Ti C Mn Al P s N slabs

A4 - 0.18 0.89 0.044 0.008 0.009 0.0048

A5 - 0.18 0.84 0.048 0.010 0.008 0.0049 16

A6 - 0.17 0.90 0.050 0.007 0.006 0.0040

B4 0.006 0.16 0.95 0.043 0.006 0.006 0.0050

65 0.006 0.17 0.92 0.052 0.008 0.006 0.0033 17

BÓ 0.009 0.16 0.92 0.033 0.011 0.009 0.0034

C4 0.018 0.18 0.87 0.040 0.012 0.010 0.0038

C5 0.016 0.18 0.81 0.049 0.011 0.010 0.0045 17

CÓ 0.017 0.16 0.89 0.049 0.009 0.007 0.0050

D4 0.026 0.16 0.86 0.045 0.008 0.008 0.0054

D5 0.028 0.17 0.85 0.042 0.008 0.007 0.0054 16

D6 0.029 0.17 0.86 0.047 0.007 0.006 0.0039

______
 66

During straightening, the surtace edge temperatures ot the slabs were

measured using an optical pyrometer and the incidence of transverse

cracks on the upper surface edge of the slabs was used to identify the

optimum com bination of titanium content and straightening temperature.

6.2. Metallography

To obtain a better understanding of the mechanism of transverse crack

formation, especially in relation to grain boundary embrittlement, the

subsurface structure of the cast steel was examined metallographically to

determine:

(a) the appearance of transverse cracks, and

(b) the grain size of the cast structure.

Test pieces 200x200x50 mm (see Fig. 6.2) were oxy-acefylene cut from the

slabs of steels A, B, C and D, which have different titanium contents, as

shown in Tables 6.2 and 6.6 for water rates of 0.756 and 0.736 l/kg. Four

specimens measuring 20x20x50 mm were cut from each piece using a

machine saw equipped with a cooling system. The specimens were then

prepared by polishing and etching for metallographic investigation using

a Leitz Wetzlar-Metalux II microscope.

 67

Fig. 6.2 Schematic diagram showing location of

metallographic specimens.

6.3. Application of the Mathematical Model

Careful control of spray cooling conditions is essential for successful

operation of the continuous casting process. Under-cooling can result in

retention of liquid metal in the core of the strand with consequential

sensitivity to cracking due to the lack of ductility when tensile strain arises

at the solid/liquid interface. On the other hand, over cooling of the strand

surface to below Ar$ may lead to surface transverse cracking. Therefore,

control of spray cooling is essential and normally, the strand surface should

be m aintained at above 900 °C.

 68

The temperature ot liquid metal in the tundish can be measured using a

thermocouple, and in the straightening region with an optical pyrometer.

However, it is very difficult to determine temperature between these two

stations. With this limitation in mind, secondary cooling was investigated

using the model described in Chapter 5 to simulate the strand surface

!
temperature during continuous casting under a given set of casting

parameters. The model may be used to assist in designing the cooling

process, particularly if the spray cooling is to be changed to improve the

cooling process. Accordingly, the m athematical model was used to

calculate the surface temperature for spray cooling conditions of 0.756

l/kg and 0.736 l/kg. The surface temperature of the strand was calculated

as a function of time under limitations of the assumptions made during

developm ent of the model and which are discussed in Chapter 7, and

with the input data shown in Table 6.7.

 69

Table 6.7 Input data for running the m athematical model

Casting parameters Normal water rate Red. water rate

(0.756 l/kg) (0.736 l/kg)

(a) Slab w idth (mm) 1200 ' 1200

(b) Tundish temperature (°C) 1544 1544

(c) Liquidus temperature (°C) 1516 1516

(d) Casting speed (m/min..) 0.9 0.9

(e) Casting time (min.) 16.9 16.9

(f) Water rate (l/min.):

spray-ring 89 89

zone-1 A 360 360

zone-1 B 206 206

zone-2 284 284

zone-3 206 183

zone-4 129 119

 70

Chapter 7

RESULTS AND DISCUSSION

The work carried out in this study of surface quality of continuously cast

steel was concerned with preventing, or at least reducing, transverse

cracking by measures that are practicably sensible in routine production.

Experimental work was carried out as described in Chapter 6 to:

(a) identify susceptible alloys and) the important factors influencing

susceptibility to transverse cracking, and

(b) investigate the possible beneficial effects of controlling steel

composition and/or straightening temperature to minimise the

occurrence of transverse cracking.

For purpose (a), a stastistica! survey of accum ulated production data was

m ade and for purpose (b) a program of experimental work on transverse

cracking prevention was carried out. The recorded data are stored on the

com puter and available on request from PT. Krakatau Steel.

 71

7.1. Statistical Survey

PT. Krakatau Steel records relating to production from the No. I casting

machines for the period July 1992 to July 1993 were searched to

obtain information about:

(a) susceptibility to transverse cracking of various slab grades, Fig. 7.1,

(b) the relationship between straightening temperature and tranverse

cracking probability, Fig. 7.2, and

(c) the relationship between water rate and straightening temperature,

Fig. 7.3, including : 1

(i) the statistical variation of actual water rate for nominal standard

practice of 0.756 l/kg Fig. 7.4, and

(ii) the consequential variation of straightening temperatures for the

nominal standard practice of 0.756 l/kg, Fig. 7.5.

Figure. 7.1 is a histogram which shows the percentage of slabs which

contained transverse cracks for steels classified into three grades with

compositions:

(a) C < 0.06%,

(b) 0.07< C <0.09%, and

(c) .0.12 < C <0.18%

 72
70 N=508
g 60
0>
§ 50
I«> 40
12 30 M m wm M
£ 20 mkîMMWMM
2
N=1201
I 10 11
N=787
0 -----miummiUHiüiim.iwni----- 1----- H -----1----- M É g M M M l
<0.06 0.07-0.09 0.12-0.18
ste e l (a) ste el (b) ste e l (c)

Steel grades,(%C)

Fig. 7.1 Histogram showing the effect of carbon content on

transverse cracking susceptibility of cast steel.

For grade (a) 0.5% of 787 slabs, for which information was available, were

cracked; for grade (b) 7% of 1201 slabis were cracked and for grade (c)

62.8% of 508 slabs were cracked. These data clearly establish that the

susceptibility of cast steel slabs (c) containing 0.12 < C < 0.18% is

excessively high with 62.8% cracking incidence, whereas the susceptibility

of the steels (a) is negligable and of the steels (b) is acceptably low.

Consequently, attention was directed in this study to steels with carbon

content in the range 0.12-0.18%.

As discussed in Section 4.3.2.1, the high incidence of cracking for steels

containing 0.12-0.18% C might be explained in terms of microstructural

 73

changes by peritectic reaction during solidification in the mould. These

changes result in a coarse austenitic grain structure with consequential

increased sensitivity to cracking (see Sec. 7.6, Metallography). The

dependence of austenitic grain size on carbon content corresponds to

the range in which ductility loss occurs (Fig. 4.3) and is in good agreement

with the variation of austenite formation temperature by the peritectic

transformation as shown by the Fe-C constitutional diagram (see Fig. 3.1).

With respect to the relationship between straightening temperature and

transverse cracking, the survey established that for steels containing 0.12 <

C < 0.18%, the sensitivity was maximum at about 800 + 25 °C, Fig. 7.2. This

result indicates that the incidence of transverse cracking conforms with

the hot tensile testing results carried out using a Gleeble machine (see Fig.

4.4). Those test data showed that the ductility, as indicated by

percentage reduction in area, had a minimum value at about 800 °C

corresponding to:

(a) the temperature for which cracking was maximum (83.7%, Fig. 7.2), and

(b) the temperature at which significant precipitation of ferrite from

austenite is expected.

Furthermore, Fig. 7.2 shows that there is a significant decrease of cracking

to 31.2% for straightening temperatures of about 900 °C. Since the Ar3
 74

temperature at which precipitation ot ferrite from austenite commences

during cooling is about 900 °C for very low carbon steels, if would appear

that the decrease in cracking relates to microstructure at the straightening

station. The decrease to 56.2 % for straightening temperatures of 750 + 25

°C was a consequence of increase in the amount of ferrite in the cast

structures, resulting in improvement of ductility as shown in Fig. 4.4.

Fig. 7.2 Diagram showing the effect of straightening temperature on

transverse cracking susceptibility of continuously cast 0.12­
0.18% C steels. The number N refer to the statistical data of
slabs examined during normal production operation, with a
secondary cooling water rate of 0.756 l/kg, and ± 25 is 3
standard deviation.

As discussed in Section 4.3.2.2, the occurence of transverse cracking in the

temperature range 775-824 °C is apparently related to intergranular

 75

embrittlement (see Sec, 7.6, Metallography) caused by allotriomorphs ot

ferrite precipitated at the austenitic grain boundaries in com bination with

the tensile stresses generated on the upper slab surface during

straightening. As postulated in Fig. 4.5 (page 31), inter-granular

embrittlement arises because the tensile stresses concentrate strains

preferentially within the soft ferrite seams with the consequence that

plastic deform ation occurs predominantly in the ferrite and is then

followed by (austenitic) grain boundary sliding causing wedge type

cracks at the grain boundary triple points. Final fracture occurs as a result

of m icrovoid coalescence, resulting in transverse cracks.

The advantageous possibility of controlling secondary cooling to raise the

straightening temperature to about 900 °C is evident in the empirical

relationship between the water rate and straightening temperature shown

in Fig. 7.3. The d ata shown in this Figure clearly establishes that, as the

secondary cooling water is reduced from 0.771-0.790 l/kg (of steel) to

0.731-0.750 l/kg, the temperature at the straightening station increases

from about 755 °C to about 880 °C. The nominal water rate designated for

normal operaton of the caster is 0.756 l/kg, for which the straightening

temperature is 800-850 °C. Figure 7.3. shows that reducing the water rate

by 0.01 i/kg increases the straightening temperature by aproximately

 76

30 °C/ so that a temperature ot about 900 °C should result from reduction

of the water rate from 0.756 l/kg to about 0.736 l/kg. The effects of water

rate on straightening temperature and cracking are described in Section

7.2.

W a te r ra te (l/k g )

Fig. 7.3 Diagram showing an empirical relationship between

water rate and straightening temperature.

7.2. Water Rate.

The objective of controlling the water rate is to obtain a straightening

temperature in excess 900 °C. In normal operation of the continuous

caster, the nominal water rate is 0.756 l/kg which was established from

empirical work with particular concern only for the requirement that
 77

solidification of the strand must be com pleted before straigthening. This

requirement was imposed to avoid the incidence of internal cracking

which would be caused by liquid metal remaining in the core of the strand

under the tensile strain arising at the solid/liquid interface during

straightening (see Sec. 2.2). However, the opportunity exists to lower the

water rate slightly to 0.736 l/kg, as this practice had been examined

previously and was known not to impair the internal quality of stabs, or

interfere with production schedules.

There is no doubt that variations in water pressure, increases or decreases

in casting speed, changes in the tundish temperature and inconsistences

in other operational parameters result in a considerable range of

straightening temperatures for standard operating conditions, Fig. 7.4 and

7.5. Clearly, these statistical data show that for the nominal rate of 0.756

l/kg under normal production conditions of 159 slabs, the straightening

temperature varied from about 700 °Cp to over 900 °C with ~ 46% of the

slabs having a straightening temperature of 850 + 25 °C. Under this

condition, it is expected that significant precipitation of ferrite at the

austenitic grain boundaries will occur before straighfening with the

consequence that many slabs will contain transverse cracks.

 78

Variation in straightening temperature for normal production conditions,

shown in Fig. 7.5, might be attained by varying the actual water rate used

for the secondary cooling process. Figure 7.4 shows that for a nominal rate

of 0.756 l/kg, the actual utilized water rate ranges from about 0.730 l/kg to

about 0.790 l/kg.

50
40
30
20 '♦ >>f+:4prT^
' *V*.*"*vl'+ 'tV *+tv 't v**+++vf>*

10 L+J4+4+4

0 .¿¿¿i .t.r.tti.ritliiiiiitrfiiiMiUt.u
0.73 0.75 0.77 0.79
Vteter rate (l/kg)

Fig. 7.4 Histogram showing range of the actual water rate for normal
production operation with nominal water rate of 0.756
l/kg. The number N refers to the statistical data of slabs, as
examined in Fig. 7.2.
 79

Straightening temperature (°C)

Fig. 7.5 Histogram showing range of straightening temperature for cooling

conditions operation with nominal water rate of 0.756 l/kg.

To minimize the consequences-of normal production variations in water

rate, the rate was very carefully controlled at the nominal value of 0.756

l/kg during the production of 67 slabs, as described in Section 6.1.1, Table

6.2. Even with this careful control the straightening temperatures for the

slabs, shown in Fig 7.6, were in the range 775-925 °C with about 60% in the

850 + 25 °C range.

It is obvious from this experimentation that careful control of the cooling

water rate at 0.756 l/kg, would reduce the spread of the range of

straightening temperatures. Equally obvious, the majority of slabs have a

straightening temperature of about 850 °C and would be expected to be

 80

cracked, despite the caretul control. Thus tor production proposes, it is

evident that caretul control of the water rate at 0.756 l/kg will not

significantly reduce the incidence of cracking and it is further evident that

a slightly lower water rate, could be used to raise the straightening

temperature. Consequently, 66 slabs as described in Section 6.1.3, Table

6.6 were cast using a water rate of 0.736 l/kg to determine whether this

measure could reduce the incidence of cracking. The results of the

experimentation, shown in Fig. 7.7, indicate that the reduced water rate

did in fact result in a straightening temperature of ~ 900 °C as expected

from the relationship shown in Fig. 7.3.

Straightening temperature (°C)

Fig. 7.6 Flistogram showing the range of straightening temperatures for trial
operation with the water rate controlled to exactly 0.756 l/kg.
 81

80
70
60
50
g 40
30
20
10
0
725 775 825 875 925 975
Straightening temperature (°C)

Fig, 7.7 Histogram showing the range cf straightening temperatures for

trial operation with the water rate controlled to 0.736 l/kg.

The incidence of cracking in the slabs, which were cast using the two

different water rates, Figs. 7.6 and 7.7, without and with titanium additions

are shown respectively in Figs. 7.8 and 7.11 [see Sec. 7.5).

For steel A, Fig. 7.8 shows the incidence of cracking of 17 slabs and 16

slabs which were cast respectively using 0.756 l/kg and 0.736 l/kg of

secondary cooling wafer. Clearly, the effect of reduced water rate was

significant in decreasing the incidence of transverse cracking from about

60% of slabs to about 30% of slabs. The beneficial effect of control of

straightening temperature probably resulted from prevention of

 82

precipitation of allotriomorphs of ferrite by raising the temperature of the

austenite at straightening as indicated by comparison of Figs. 7.6 and 7.7.

The tensile stresses generated during straightening, as discussed in Sec

4.3.2.2, were therefore accomodated without failure in the grain boundary

regions of the austenite.

N=17
60
*

60 Steel A
m
o0
1 ü ®
40 *V' ^vT« :fV

2 N=16
t>
> 30 }î?î^ Â w w îw î*'i
CO W Ê ffls tè m
c
(0 M m v CC+IV
Critit»tlvi í
»>;£+r+>TtT« »?+.-♦:
20
o
X ♦Kw ä'

Ta
i
J
c 10
4XvVCCC+^Vl++viCv
« M Ä :
0 --------------------------- 1— — ♦— — I---------------------------

0.756 0.736
Wfeiter rate (l/kg)

Fig. 7.8 Histogram showing the effect of the two rates of secondary
cooling water on transverse cracking susceptibility of continuously
cast titanium free 0.16-0.18% C steels.

Attainment of a straightening temperature of ~ 900 °C is a consequence

of the complete cooling process that occurs between the tundish and the

straightening station. During secondary cooling the surface temperature

of the strand is progressively reduced from the known tundish temperature

 83

to the tem perature which can be measured at the straightening station. It

is not possible to obtain measures of the surface temperature of the strand

during this cooling despite the im portance of knowledge of the way the

tem perature changes during progress of the strand through the caster.

Consequently the m athem atical model, established in Chapter 5, was

used to calculate these temperatures and thereby to assist in

understanding the effect of secondary water rate on straightening

temperature.

7.3. Mathematical Model

For a particular tundish temperature, the straightening temperature will be

determ ined principally by:

(i) the dimensions of the strand,

(ii) the casting speed, and

(iii) the secondary cooling, including the strand containments and the rolls

cooling system.

A strand surface edge of above 900 °C is dem anded at straightening to

prevent precipitation of ferrite and consequential transverse cracking. To

attain a straightening temperature of (say) 900 °C it is essential to have

inform ation about the progressive decrease in temperature from the

tundish to a straightening station. However, It is very difficult to measure

 84

the temperature ot the strand directly in the continuous casiing machine,

and so the most profitable approach to determination of the temperature

profile of the strand surface is by calculation using the m athematical

model. For this purpose, the strand is represented as a thin slice which

progresses from the tundish to the straightening station, and which

undergoes cooling that can be calculated using the model set up in

Chapter 5.

The model is a finite element analysis by which the temperature of the

node of the surface element is ultimately calculated. This node is located

in the center of the wide face if the strand. However, the transverse

cracking occurs, not at the face center but, at the face edge.

Consequently, the critical condition which should govern whether or not

cracking occurs is the temperature of the strand edge at straightening.

The m athem atical processes for determining the edge temperature from

the calculated face center temperature are complex and were not

pursued in this work. Alternatively, a simple empirical relationship between

the two temperatures was obtained from direct measurement, near the

straightening station, for the specified casting conditions as described in

Sec. 6.1. This relationship was found to be :

 85

Tedge = (T fc c e - 90 ) ° C (7.1)

and was applied to the calculated results to obtain the required edge

temperatures.

It should be noted that the correction temperature of 90 °C is valid for the

conditions and the m achine involved and should not be applied to the

other situations indiscriminantly.

The input d a ta required to run the m athem atical m odel are shown in Table

6.6, and the results of the calculations are shown in Fig. 7.9 and in

Appendix A. From Fig. 7,9 it is generally evident that the temperature of

edge of the slice decreases sharply from the tundish/mould region to

about the end of zone-1 B. The high rate of decrease in this region is a

consequence of high cooling intensity due to the high ratio of water rate

to sprayed surface area (see Fig. 6.1 and Table 6.3 or 6.4). In these zones,

the total surface area of the top of the strand is approximately 1.34 m x 1.2

m and was cooled, as designated in Table 6.3 or 6.4, with about 650

l/m 2/m in of water. This rate may be com pared with the rate of about 52

l/m 2/m in in zone-2, for which the surface area is approximately

4.39 m x 1.2 m.
 86

*■ tiMe (see)

Fig. 7.9 Diagram showing the calculated edge temperature as the strand
moves through the continuous caster for water rates 0.756 l/kg and
0.736 l/kg.
 87

As the slice enters zone-2, the surface is reheated by conduction from

within the slab until the temperature gradient between surface and interior

of the slice becomes relatively small due to reduced cooling from spray

w ater (as discussed above). The temperature of the slice then decreases

again through zones-3 and 4, mainly due to spray cooling. When the slice

leaves zone 4, the cooling conditions change from forced convection

cooling to radiant cooling and consequentially, the surface temperature

increases due to conduction of heat from interior of the slice.

If due to severity of cyclic heating/cooling, the temperature gradients

during secondary cooling are too high, it will result in the strand surface

temperature to fall below 800 °C, as shown in Fig. 7.9, This can increase the

susceptibility to transverse cracking related to the intergranular

embrittlem ent caused by allotriomorphs of ferrite precipitated at the

austenitic grain boundaries in com bination with the tensile stresses

generated due to the withdrawal forces, strand bulging, roll eccentricity,

and poor alignm ent between the individual strand guide elements and

segments. Consequently, transverse cracks would be expected to occur

in continuously cast steel slabs, despite the reducing of water rate to

0.0736 l/kg, which was expected to give the straightening temperature of

above 900 °C to avoid the formatioH. of atlotriomorphs of ferrite during

straightening.

Figure 7.9 shows the edge temperature variations for secondary cooling

water rates of 0.756 l/kg and 0.736 l/kg. The effect of reduced water rate is

evident and results in a calculated increase in edge surface temperature

of about 10 °C in zone-3 and-4, and in the straightening region. It should

be noted that the reduced water rate is effective only in zones 3 and 4

(see Table 6.3, 6.4) so that there is no change to the temperature variation

from the mould to zone 2. Calculated straightening temperatures of 848

°C and 856°C are to be com pared with the respective measured

temperatures at the straightening station of 850°C and 900°C. It is obvious

from this comparison that the mathematical model under-estimates the

straightening temperatures, particularly for the lower water rate.

Nevertheless, the model provides much useful information about the

temperature variation during cooling and will be pursued further to

determine how it can be modified to provide more reliable predictions of

the exit temperature.

 89

7.4. Effect of Titanium

The work described in Section 6.1.3 establishes that addition of titanium to

cast steel has beneficial effects in reducing the incidence of cracking. The

effect is probably due to suppression of austenitic grain growth by

precipitates of titanium compounds. Consequently, the effects of titanium

additions on cracking were examined. (

As discussed in the experimental work (see Chapter. 6), titanium additions

were m ade to a normal 0.16-0.18 %C steel to provide four base alloys with

titanium contents in the ranges (Table 6.2):

(A) 0 % Ti,

(B) 0.005-0.010%Ti,

(C ) 0. 015-0.020 %Ti, and

(D) 0.025 - 0.030 %Ti

The additions were m ade so that TiN and /o r TiC would precipitate at the

austenitic grain boundaries and thereby pin those boundaries and refine

the grain structure. The effect of the additions on the surface quality of

slabs of the four alloys which were cast using a water rate of 0.756 l/kg,

with careful control of the process, are shown in Fig. 7.10 (see Appendix B).

Clearly the effect of titanium is significant up to approximately 0.01 7 % and

 90

results in a decrease in the incidence of transverse cracking from

approximately 60% of slabs to approximately 40%. Increasing the titanium

content to ~ 0.028 % seems to have had no additional influence on the

surface quality. The effects of titanium additions are probably due to

limited precipitation of TiC and/or TiN at the austenitic grain boundaries

to refine the grains during cooling in the mould in practical continuous

casting as discussed in Section. 4.3.3.1. Furthermore, the negligible effect

of increasing the titanium content from 0.01 7% to 0.028% is most likely

associated with the grain sizes of the cast structure for these two

compositions. Figure 7.17, which is discussed in Section 7.6.2, shows that

the distribution of the ASTM grain sizes for the steel containing 0.028 %Ti

was very similar to the distribution in the steel containing 0.017 %Ti.

Consequently, it is apparent that the increase in titanium content did not

alter the grain size distribution and so the cracking propensity of the steels

would be expected to be the same, resulting in similar surface quality

levels.
 91

Fig. 7.10 Diagram showing the effect of titanium on transverse cracking

susceptibility of continuously cast steel slabs for water
rate 0.756 l/kg. The number N refer to the experimentation
data of slabs of the steels A, B, C and D with 95% confidence
limit, composition shown in Table 6.2.

7.5. Combined Effect Water Rate and Titanium

The experiment work described in Sections 6.1.1 and 6.1.2 established that

reduction of water rate and addition of titanium to the cast steel each

have beneficial effects in reducing the incidence of cracking. The effects

are probably due to suppression of austenitetic grain growth and

supression of precipitation of allotriomorphs of ferrite by titanium additions,

 92

and raising of the straightening temperature by towering the water rate.

The results indicate that the beneficial effect of titanium addition and

reduced water rate might be com bined advantageously. Consequently,

the effects of reduced water rate on the cracking propencity of steels

containing several levels of titanium were examined as described in

section 6.1.3. For this purpose, the titanium contents of steels were

designed (Table. 6.5):

(A) 0 %Ti,

(B) 0.005-0.010%Ti,

(C) 0.015-0.020 %Ti, and

(D) 0.025-0.030 %T\.

Slabs of the four steels were cast using both the normal water rate of 0.756

l/kg and the reduced rate of 0.736 l/kg and the incidece of cracking in

these slabs is shown in Fig. 7.11 (see Appendix B).

Fig. 7.11 Diagram showing the effect of titanium on transverse cracking
susceptibility of continuously cast steel slabs for water rate
0.736 l/kg and 0.756 l/kg. The number N refer to the
experimentation data of slabs of the steels A, B, C and D with 95%
confidence limit, composition shown in Tables 6.2 and 6.6.

Clearly, the effect of titanium and reduced water rate combined was

significant in decreasing the incidence of transverse cracking up to

approximately 0.01 7 %Ti. As noted in Section 7.4, the relationships shown in

Fig. 7.11 indicate that addition of 0.017 %Ti reduced the incidence of

cracking from about 60% of slabs to about 40% of slabs for the standard

water rate. Additionally, reduced secondary cooling reduced the number

of cracked slabs by about 50% at each level of titanium. Further addition

 94

of titanium to 0.028 % had negligible influence on the surface quality.

Clearly, the beneficial effect of com bined control of titanium content and

straightening temperature can result in suppression of austenitic grain

growth and suppression of precipitation of allotriomorphs of ferrite by

retardation of the y/a transformation. T!pe tensile stresses generated during

straightening, as discussed in Sections 4.3.2.1 and 4.3.2.2, can therefore be

accom m o d ate d without failure in the grain boundary regions of the

austenite. Therefore, it can be concluded that, by control of the titanium

content to approximately 0.01 7% and straightening temperature to above

900 °C by lowering the water rate to 0.736 l/kg, it is possible to obtain

optimum measures to minimize transverse cracking. Both of these

measures can be applied to the production casting machine without

interference with the production schedule.

7.6. Metallography

As has been discussed in Section 3, the surface of continuously cast steel

slab contains oscillation marks formed perpendicular to the casting

direction, as shown in Fig. 7.12. This observation is considered to be

important with respect to the location of transverse cracks which

frequently occurred at the bottom of oscillation marks, on the upper

surface edge of the slabs, and perpendicular to the casting direction as

 9S

shown in Fig. 7.13. Normally, transverse cracks.can be clearly observed by

removing the oxidized layer on the corner of slabs, by a slight scarfing.

Defects revealed in this way are shown in Fig. 7.14.

casting direction

Fig. 7.12 Typical appearance of oscillation marks on the surface of

continuously cast steel slabs shown in cm scale.

bottom of oscillation marks

Fig. 7.13 Photomacrograph showing occurance of transverse cracks at

the bottom of oscillation marks.
Fig. 7.1 4 Photomacrograph showing the appearance ot transverse cracks
after removing the oxidized surface layer by scarfing.

To obtain the better understanding of the mechanism of transverse

cracking, and for controlling the incidence of them, it was considered

important to investigate the subsurface cast structure. For this purpose,

specimens for metatiographic investigations were prepared from slabs as

set out in Sec. 6.2 (see Fig. 6.2) and were examined using a Leitz Wetziar-

Metalux II microscope.

7.6.1. Subsurface Structures in the Vicinity of Transverse Cracks

Figures 7.15 and 7.16 show two different, but typical, transverse cracks

found at the bottom of oscillation marks in two specimens cut from

different slabs containing 0% Ti, and cast using a secondary cooling water
 97

of 0.756 l/kg. Both subsurface structures consist of pearlite and ferrite as

idiomorphs, allotriomorphs and some Widmanstatten plates. These

subsurface characteristics provide g ood sites for transverse crack

form ation at the bottom of oscillation marks on the upper surface of slabs,

at the straightening point.

Figure 7.15 shows the high volume fraction of allotriomorphs of ferrites in

the subsurface structure beneath the oscillation marks. The

photom icrograph also shows that the cracks initiated from that surface,

then propagated, mostly intergranularly.

 98

Fig. 7.15 Photomicrograph showing typical appearance of a transverse

crack, with no solidified hook formation, in the subsurface
structure below an oscillation mark, magnification of 100 x,
etched in nital.
 99

The crack shown in Fig. 7.15 might be associated with the allotriomorph of

ferrites formed by phase transformation at and below the Ar$ as a

consequence of casting with the water rate of 0.756 l/kg. The presence of

the allotromorphs resulted in intergranular embrittlement when tensile strain

arose during straightening. Consequently, it appears evident that the

incidence of transverse cracks in the slabs containing 0% Ti and cast

usingthe water rate of 0.756 l/kg would be too high. As shown in Fig. 7.8,

the incidence was, in fact, about 60 %.

On the other hand, the crack shown in Fig. 7.16 apparently initiated from

the segregation line between the fine and coarse grains consequent upon

the formation of a solidified hook (refer to section 3.1 and Fig. 3.3), then

propagated mostly intergranularly in the coarse grained region below the

hook. Additionally, it can be seen that the cracks propagated throught

regions of fine ferrite which are likely to be allotriomorphs formed at the

grain boundaries of a coarse grained austenite.

 100

Fig. 7.16 Photomicrograph showing typical appearance of a transverse

crack, with solidified hook formation, in the subsurface structure
below an oscillation mark, magnification of 75 x, etched in nital.
 101

Clearly, transverse cracks, such as shown in Fig. 7.16, are likely to

propagate in the most brittle zone, owing to the eftects of segregation

a nd/or discontinuities in the solidified shell due to the formation of a

solidified hook. The effect results in.coarse austenite grains which in turn

can result in increasing susceptibility to transverse cracking. Consequently,

transverse cracks would be expected to occur in continuously cast steel

slabs in which solidification hooks occur, despite addition of titanium to

refine the grain structure and/or controlling the straightening temperature

to avoid the formation of allotriomorphs of ferrite. As these two measures

cannot control the formation of solidified hooks this factor has an

independent influence on the susceptibility to transverse cracking and

clearly needs to be investigated,

7.6.2. Grain size of the Subsurface Cast Structure

Grain sizes of the cast structures of steels A, B, C and D, which have

different titanium contents, shown in Tables 6.2 and 6.6 for water rates of

0.756 and 0.736 l/kg, were measured using standard ASTM grain size charts

and were classified according to ASTM grain size number.

It was, in fact, quite difficult to obtain a measure of the grain sizes of the

cast structures, as they normally consisted of mixtures of small, medium

 102

and large grains. However, the grains which were measured, were re­

presentative ot the grain size distributions.

The results of the measurements are shown in Figs. 7.17 and 7.18 (See

Appendix B). Clearly, for both water rates (0.756 l/kg, 0.736 l/kg), the grain

size was sensitive to titanium content up to about 0.01 7%. Further increase

in the titanium content to 0.028% had little additional effect on the grain

size distribution. As the titanium concentration was increased from 0 to

0.01 7% the fraction of large grains (ASTM 6) decreased, the fraction of

small grains (ASTM 9) remained about the same, and the fractions of the

intermediate sized grains (ASTM 7,8) changed in opposing ways. As a

consequence of these changes, the grain structures becom e finer as the

titanium content increased, as expected.

The effect of water rate on the grain structure is indicated by comparison

of Figs. 7.1 7 and 7.18. For both rates, the average grain size for the steel

containing 0 %Ti was ASTM 7.65 and for 0.017 %Ti was ASTM 8.02. The

reduction of average grain size consequent upon addition of titanium is

almost certainly responsible for the reduced cracking incidence shown in

Fig. 7.10 and 7.11 because by reducing the grain size with addition of

small amounts of titanium significant improvement in the ductility of the

 103

steel in the critical temperature range of 850 - 750°C occurred, as

discussed in Section 4.3.3d (see Fig. 4.6)

0 0.007 0.017 0.027

T Itanium conte rrts (%)

Fig. /.I 7 Diagram showing the effect of titanium on grain size of

subsurface structure of the cast slabs for water rate
of 0.756 l/kg.

Fig. 7.18 Diagram showing the etîecT of titanium on grain size of subsurface
structure of the cast slabs for water rate of 0.736 l/kg.
 104

Chapter 8

CONCLUSIONS

Practical measures for reducing the incidence of transverse cracking

associated with austenitic grain growth and precipitation of allotriomorphs

of ferrite in continuously cast steel slabs have been studied. The

conclusions reached from a stastistical survey of PT. Krakatau Steel

production d a ta and experimental work were as follows.

1. The defect propensity in continuously cast steel slabs containing 0.12 <

C < 0.18 % is very high, for 0.07 < C < 0.09% it is acceptably low and for C

< 0.06% if is negligible.

2. The critical straightening temperature for which transverse cracking

susceptibility maximum was 800 + 25 °C.

3. Under normal production conditions, slabs were cast with nominal

water rate of 0.756 l/kg, resulting in straightening temperatures from about

700 °C to over 900°C with ~ 46% of the slabs having a straightening

temperature of 850 + 25°C.

 105

4. The range of straightening temperatures for a water rate of 0.756 !/kg

couid be reduced by careful control of the process, to the range 775 °C to

925 °C with ~ 60 % of the slabs having a -straightening temperature of 850 +

25 °C.

5. For a straightening temperature below about 900 °C, allotricmorphs of

ferrite, precipitated at the austenitic grain boundaries, cracked under

tensile stress generated on the top surface during straightening of the slab.

6. The incidence of transverse cracking associated with precipitation of

allotriomorphs of ferrite was reduced significantly from about 60 % of slabs

to about 30% of slabs, by raising the straightening temperature to above

900 °C by lowering the water rate to about 0.736 l/kg.

7. Straightening temperatures calculated as a function of water rate m ade

with a finite element m athem atical mode! were under-estimated

com pared with direct measurement. Nevertheless, the model provides

much useful information about the temperature changes in the strand

surface during cooling; these d a ta are very difficult to measure directly.

 106

8. The incidence of transverse cracking associated with coarsening of the

austenitic grain structure can be reduced significantly from about 60 % of

slabs to about 40% of slabs, by addition of titanium up to 0.01 7%.

9. Under limited cooling conditions, the grain size of the cast slab was not

sensitive to water rate, but was sensitive to addition of titanium up to

about 0.01 7%.

10. The negligible effect of increasing the titanium content from 0.01 7% to

0.028% was most likely associated with the similar grain size distributions in

the structures for these two compositions.

11. The fraction of large grains (ASTM 6) decreased, the fraction of small

grains (ASTM 9) remained about the same and the fraction' of the

intermediate size grains (ASTM 7,8) changed in opposing ways with

increase in the concentration of titanium from 0 % to 0.017%, and

remained essentially the same for further addition of titanium to 0.028%.

12. The average grain size for the steel containing 0 %Ti was ASTM 7.65 and

for the steel containing 0.01 7 %Ti was ASTM 8.02.

 107

13. The highest incidence of transverse cracking occurred at the bottom of

oscillation marks and occurred intergranularly.

14. The effect of com bined control of titanium content to approximately

0.01 7% and straightening temperature to above 900 °C by lowering the

w ater rate to 0.736 l/kg was most beneficial in minimizing transverse

cracking and can be applied to the production casting machine at FT.

Krakatau Steel without interference to the production schedule.

 108

REFERENCES

1. H.F. Schrewe: H isto rical D e v e lo p m e n ts o f C o n tinu ou s S te e l C a s tin g ,

Continuous Casting of Steel, Stahl Eisen, May 1987, p.1-11.

2. WJ. Lankford, N.L. Samways, R.F. Craven, H.E. McGannon: The M a k in g ,

S h a p in g a n d T re a tin g o f S te e l Association of Iron and Steel Engineers,
Copyright 1985, p.741-755.

3. H. F. Schrewe: C o n tin u o u s S te e l C a s tin g P la n t C o m p o n e n ts , Continuous

Casting of Steel, Stahl Eisen, May 1987, p.20-50>

4. H.F. Schrewe: S tra n d M e c h a n ic s , B e n d in g , S tra ig h te n in g a n d B ulg ing ,

Continuous Casting of Steel, Stahl Eisen, May 1987, p.l 38-144.

5. J. Pietryka, A. Chielens and P. Benoit: A N ew W a y To S p ra y , AISE

Conference, Pennsylvania, 1983.

6. J. Zou and A. A. Tseng: M a c r o s c o p ic M o d e llin g o f F u n d a m e n ta l P h as e

Transform ations in C o n tin u o u s C astin g s o f S te e l Metallurgical
Transactions A, Vol. 23A, February 1992, p.457-467.

7. G.J. Davies, R.S. Laki, I.G. Saucedo and Y.K. Shin: H e a t Transfer M o d e ls
and Their A p p lic a tio n to C o n tin u o u s C a s tin g , Department of
Metallurgy, University of Sheffield, 1985, p.l 23-125.

8. A.W. Cramb and FJ. Mannion: The M e a s u r e m e n t o f M e n is c u s M a rk s a t

B e th le h e m S teels Burn H a rb o r S la b C a s te r , Steelmaking Proc., Vol. 68,
Detroit 1985.

9. E. Takeuchi and J.K. Brimacombe: The F o rm a tio n o f O s c illa tio n M a rk s in

th e C o n tin u o u s C a s tin g o f S te e l S labs , Metallurgical Transactions B,
Vol. 15B, September 1984, p.493-496.

10. K. Kawakami, T. Kitagawa, M. Komatsu, FI. Mizukami, A. Masui and T.

Ishida: S tu d y on th e S o lid ific a tio n P h e n o m e n a in C o n tin u o u s C a s tin g
M o u ld , Nippon Kokan Technical Report, No. 36, 1982, p.17-24.
 109

11. G. Beirer: Solidific a fio n a n d H e a t Transfer M o d e l , Thyssen Stah! AG,

Technical Recomendation, January 1992.

12. A. Etienne and B. Mairy: H e a t Transfer in C o n tin u o u s ly C a s t S trand s,

CRM Report, No. 55, November 1979.

13. G. Kaestle and Associates: H e a t Flow a n d S o lid ific a tio n R a te in S tra n d

C a s tin g o f S lab , Steelmaking Proc. Vol. 65, Pittsburgh 1982.

14. A. Diener and A. Drastik: H e a t E x c h a n g e B e tw e e n Strands a n d G u id e

Roll in th e S e c o n d a r y C o o lin g Z o n e o f a S la b C o n tin u o u s C a s tin g
M a c h in e , Arch. Eisen Hutten Wesen, No. 1, 1981.

15. F.P. Incropera. In tro d u c tio n to H e a t Transfer , 2nd Edition, John Wiley &
Sons, 1990.

16. F. Schwerdtferger: M e ta llu rg ie d e s S tra n g g u ß e n s , Verlag Stahl Eisen

1992.

17. H.F. Schrewe: The S o lid ific a tio n Process in C o n tin u o u s S te e l C a s tin g ,
Continuous Casting of Steel, Stahl Eisen, May 1987, p.104-120.

18. K. Kawakami, T. Kitagawa, K. Murak-ami, Y. Moyashita, Y. Tsuchida and

T. Kawawa: F u n d a m e n ta l R e s e a rc h on S o lid ific a tio n In v o lv e d in
C o n tin u o u s C a s tin g o f Steel, Nippon Kokan Technical Report, No. 36,
1982, p.26-33.

19. Y. Miyashita, M. Suzuki, K. Taguchi, S. Uchida, H. Sato and M.

Yamamura: Im r o v e m e n t o f S u rfa c e Q u a lify o f C o n tin u o u s ly C a s t S lab ,
N ip p o n K o k a n T e c h n ic a l R eport, No. 36, 1982, p.55-64.

20. S. Komori, H. Miyano, Y. Horiuchi, M. Yoshino, T. Yam ada and M.

Uesugi: Q u ic k D e fe c tio n o f L o n g itu d in a l F a c ia l C ra c k s on
C o n tin u o u s ly
H o t C a s t Slabs, 103rd ISIJ Meeting, April 1982.

21. R. Bommaraju, R. Glennon and M. Frazee: A nalysis o f th e C a u s e s a n d

P re v e n tio n of L o n g itu d in a l M id f a c e C ra c k s and D epressions on
C o n tin u o u s ly Cast 1st European
F r e e -M a c h in in g S te e l Blooms,
Conference on Continuous Casting, Florence, Italy, September 1991,
p.1599-1601.
 110

22. D.E. Humpreys: A p p lic a tio n o f M o u ld T h e rm a l M o n ito r in g in The S tu d y o f

S la b S u rfa c e Q u a lify fo r H e a v y P lpfe G ra d e s a t S c u n th o rp e W orks,
British Steel, 1st European Conference on Continuous Casting,
Florence, Italy, September 1991, p.l 529-1540.

23. F. Schrewe: P ro p erties o f The C a s t S tra n d B efore a n d A fte r S u b s e q u e n t

Processing, Continuous Casting of Steel, Stahl Eisen, M ay 1987, p.l 54­
158.

24. N i-C r A llo y T h e rm a l S p ra y in g o f th e N a rro w F a c e o f C on tin u o u s C a s tin g

M o u ld , Nippon Steel Corporation, Mishima Kosan Co.,Ltd., May 1989.

25. S u rfa c e C o a tin g o f C o n tin u o u s C a s te r M o u ld s , N ip p o n S te e l C o r p o r a t­

io n , Mishima Kosan Co.,Ltd., May 1989, p.l -14

26. E. Takeuchi and J.K. Brimacombe: E ffect o f O s c illa tio n M a r k F o rm a tio n

o n th e S u rfa c e Q u a lity o f C o n tin u o u s ly C a s t S te e l Slabs, Metallurgical
Transactions B,Vol. 16B, September 1985, p.605-6! 7.

27. N.E. Hannerz: C ritic a l H o t P lasticity a n d Transverse C r a c k in g in

C o n tin u o u s S la b C a s tin g w ith P a rtic u la r R e fe re n c e to C o m p o s itio n ,
Technical Report, No. 149, Laboratory of Thermal Processing of
Materials, Technical University of Denmark, August 1984.

28. J.K. Brimacombe and K. Sorimachi: C r a c k F o rm atio n in C o n tin u o u s

C a s tin g o f Steel, Metallurgical Transactions B, Vol. 8B, September 1977,
p.489-503.

29. Y. Maehara, K. Yasumoto, H. Tomono, T. Nagamichi and Y. Ohmori:

S u rfa c e C r a c k in g M e c h a n is m o f C o n tin u o u s ly C a s t Low C a r b o n Low
A llo y S te e l S la b s , Materials Science and Technology, Vol. 6, 1990,
p.793-804.

30. Y. Maehara and Y. Ohmori: The P re c ip ita tio n o f A IN a n d N b C a n d The

H o t D u c tility o f Low C a r b o n S te e ls , Material Science and Engineering,
Vol.62, 1984, p.l 09-119.

31. K. Yasumoto, Y. Maehara, T. Nagamichi and H. Tomono: E ffe ct o f

T h e r m o -m e c h a n ic a l History on S u rfa c e C r a c k in g o f A s -c a s t Low
Iron and Steel Research Laboratories,
C a r b o n Low A llo y S te e l S labs,
Sumitomo Metal Industries, Ltd., May 1989.
 Ill

32. B. Mintz and J. M. Arrowsmith: H o t D u c tility B e h a v io r o f C - M n - N b - A I

Steels and Its R elatio n sh ip to C ra c k P ro p a g a tio n D u rin g th e
S tra ig h te n in g o f C o n tin u o u s ly C a s t S tra n d , British Steel Corporation,
May 1978.

33. G. A. Wilber, R. Batra, W.F. Savage and W.J. Childs: The Effects o f
T h e rm a l History a n d C o m p o s itio n on th e H o t D u c tility o f Low C a r b o n
S te e ls , Metallurgical Transactions A, Vol, 6A, April 1973,p.l 727-1 734

34. J. Hertel, H. Litterscheidt, U. Lotter and H. Pircher: L a b o ra to ry S im u la tio n

o f S tra n d S hell Stresses a n d Strains d u rin g C on tin u o u s C a s tin g , Thyssen
Technische Berichte, January 1991, p.31 -41.

35. H. Suzuki, S. Nishimura, J. Imamura and Y. Nakamura: E m b rittle m e n t o f

Steels O c c u r in g in th e T e m p e ra tu re R ange from 1000 to 6 0 0 °C,
Research Article (1 69), Nippon Steel Corporation, April 1983.

36. M. Rappaz and Ph. Thevoz: A cta Metal, Vol.35, 1987,p.l 487-1497.

37. L. Meyer, C. Straßburger and C. Schneider: E ffect a n d P re sen t

,
A p p lic a tio n o f th e M ic ro a llo y in g E lem ents N b, V Ti, Zr, a n d B in HSLA
Steels, Thyssen Stahl AG, Duisburg, Germany, 1991, p.29-44.

38. H. Suzuki, S. Nishimura, J. Imamura and Y. Nakamura: Im p r o v e m e n t o f

H o t D u c tility o f C o n tin u o u s ly C a s t C a r b o n Steels , Research Article ¡54),
Nippon Steel Corporation, May 1983.

39. J. K. Brimacombe: D e s ig n o f C o n tin u o u s C a s tin g M a c h in e s B a s e d on a

H e a t-flo w A nalysis, Canadian Metallurgical Quarterly, Vol.15, 1976,
p.l 63-1 67.

40. E. A. Muzikar: M a t h e m a t ic a l H e a t Tran sver M o d e l fo r S o lid ific a tio n o f

C o n tin u o u s ly C a s t S te e l S labs, Transactions of The Metallurgical Society
of AIME, Vol. 239, November 1967, p.l 747-1 753.

41. J.P. Holman: H e a t Transfer, SI Metric Edition, Mc.Graw Hill Book

Company, Copy Right 1984.

42. S la b C a s te r M a n u a l, PT. Krakatau Steel, 1983.

 112

Appendix A

COMPUTER PROGRAM FOR CALCULATING THE TEMPERATURE

OF STRAND DURING CONTINUOUS CASTING

( PID )
Appendix A
'*:H****************************************************************
'COMPUTER PROGRAM FOR CALCULATING THE SURFACE TEMPERATURE OF STRAND
' DURING CONTINUOUS CASTING

'DEFINING THE ELEMENTS

DIM B AS DOUBLE, E AS DOUBLE, G AS DOUBLE, c AS DOUBLE, Cw AS DOUBLE

DIM k AS DOUBLE, p AS DOUBLE, Rw AS DOUBLE, Vw AS DOUBLE, Rc AS DOUBLE
DIM Trs AS DOUBLE, Tin AS DOUBLE, Tout AS DOUBLE, Ta AS DOUBLE
DIM I AS INTEGER, Leu AS DOUBLE, LI AS DOUBLE, L2 AS DOUBLE, L3 AS DOUBLE
DIM L4 AS DOUBLE, L5 AS DOUBLE, L6 AS DOUBLE, L7 AS DOUBLE, L8 AS DOUBLE
DIM NrzlB AS INTEGER, Nrz2 AS INTEGER, Nrz3 AS INTEGER, Nrz4 AS INTEGER
DIM Nrwd ASINTEGER, N AS INTEGER, M AS INTEGER, i AS DOUBLE, ts AS DOUBLE
DIM A AS INTEGER, Tb AS INTEGER, RzlB AS DOUBLE, Rz2 AS DOUBLE, Rz3 AS DOUBLE
DIM Rz4 AS DOUBLE, RpzlB AS DOUBLE, Rpz2 AS DOUBLE, Rpz3 AS DOUBLE
DIM Rpz4 ASDOUBLE, Rpwd AS DOUBLE, hczlB AS DOUBLE, hcz2 AS DOUBLE
DIM hcz3 ASDOUBLE, hcz4 AS DOUBLE, hewd AS DOUBLE

'THERMOPHYSICAL CONSTANTS

B = -.03 '(the rate of change thermal conductivity with temperature, in W/m.KA2) )

E = .85 '(emissivity)
G = 5.669E-08 ' { Stefan-Boltzman constant, in W/mA2.KA4)
c = 747 '(heat capacity of steel, in J/kg.K)
Cw = 4178 '(heat capacity of water, in J/kg.K)
k = 29 '(heat conductivity of steel, in W/ra.K)
p = 7854 '(density of steel, in kg/m'3)
Rw = 999 '(desity of water, in kg/m‘ 3)
Vw = .0367 '(water rate of mould cooling, in m*3/sec.)
Rc = .004 '(contact line between rods and the strand surace, in m)
hs = 160000 '(coefficient of spray cooling heat transfer)
Trs = 573 '(surface temperature of the ro lls, in K)
Tin = 310.5 '(temperature of the mould water inlet, in K)
Tout = 316 '(temperature of the mould water outlet, in K)
Ta = 313 '(surrounding temperature, in K)
LI = .6 '(distance between meniscus and the mould bottom, in m)
L2 = .755 '(distance between meniscus andthe end of spray-ring, in m)
L3 = 1.329 '(distance between meniscus andthe end of zone-lA, in fit)
L4 = 1.935 '(distance between meniscus andthe end of zone-lB, in m)
L5 = 4.81 '(distance between meniscus andthe end of zone-2, in m)
L6 = 9.195 '(distance between meniscus andthe end of zone-3, in a)
L7 = 12.075 '(distance between meniscus andthe end of zone-4, in m)
L8 = 15.237 '(distance between meniscus and the straightener, in m)
hczlB = 300 '(heat trans. coef. between ro lls zone-lB and strand su rf., in W/mA2.K)
hcz2 = 300 '(heat trans. coef. between ro lls zone-2 andstrand su rf., in W/mA2.K)
hcz3 = 600 '(heat trans. coef. between ro lls zone-3 and strand su rf., in W/mA2.K)
hcz4 = 900 '(heat trans. coef. between ro lls zone-4 and strand su rf., in W/mA2.K)
hewd = 1200 '(heat trans. coef. between withdrawal ro lls and strand su rf., in W/m‘ 2.K)
Appendix A

NrzlB = 4 '(number of ro lls zone-lB).

Nrz2 = 10 '(number of ro lls at zone-2)
Nrz3 = 12 '(number of ro lls at zone-3)
Nrz4 = 8 '(number of ro lls at zone-4 )
Nrwd = 8 '(number of ro lls at withdrawalregion)
RzlB = .18 '(diameter of ro lls zone-IB, in m)
Rz2 =.245 '(diameter of ro lls zone-2, in m)
Rz3 =.31 '(diameter of ro lls zone-3, in m)
Rz4 =.31 '(diameter of ro lls zone-4, in m)
Rwd =.39 '(diameter of rods withdrawal, in m)

RpzlB = .215 '(ro ll pitch zone-lB, in m)

Rpz2 = .291 '(ro ll pitch zone-2, in m)
Rpz3 = .363 '(ro d pitch zone-3, in m)
Rpz4 = .363 '(ro ll pitch zone-4, in m)
Rpwd = .45 '(ro ll pitch withdrawal, in m)

INPUT DATA FOR RUNNING THE MATHEMATICAL MODEL OF STRAND COOLING

INPUT " SLAB WIDTH (mm) : Sw

INPUT “ TUNDISH TEMPERATURE (C) : Tt
T't = Tt + 273
INPUT 8 LIQUIDUS TEMPERATURE (C)= TL
TL = TL + 273
INPUT “ CASTING SPEED (m/min.) : Vc
Vc = Vc / 60
INPUT 8 CASTING TIME (min.) : tc
INPUT “ SPRAY-RING (1/min.) = Vsr
INPUT “ ZONE-1A (1/min.) : VzlA
INPUT 11 ZONE-1B (1/min.) : VzlB
INPUT 11 ZONE-2 (1/min.) : ■; Vz2
INPUT 8 ZONE-3 (1/min.) =11; Vz3
INPUT 8 ZONE-4 (1/min.) = Vz4

' NODES (x) AND TIMES INTERVAL(ts)

i _____ _____________ _
N = 10
M = 300
x = .2 / (2 * N)
ts = tc * 60 / M
DIM T(N, M)
Sw = Sw / 1000
 115

Appendix A

'INITIAL AND BOUNDARY CONDITIONS

A= 1
DO
T(A, 1) = Tt
A= A+ 1
IF A > N THEN
EXIT DO
END IF
LOOP

'CALCULATING THE STRAND TEMPERATURE

i ________________________________________

J =2
DO
Tb = (J - 1) i ts 1
1=2
DO
PI = T (I, J - 1)
P2 = ts / (x * 2 * p * c)
P3 = T(I + 1, J - 1)
P4 = T(I - 1, J - 1)
T ( I, J) = PI + P2 * (k * (P3 + P4 - 2 * PI) + B / 4 * (P3 ‘ 2 - 2 * P3 * P4 + P4 * 2))

1 = 1 + 1
IF I > (N - 1) THEN
EXIT DO
END IF

LOOP
IF T (I, J - 1) > TL THEN
c = 747
ELSEIF 1674 < T (I, J - 1) AND T ( I f J - 1) <= TL THEN
c = 784.5
ELSEIF 1181 < T(I, J - 1) AND T ( I f J - 1) <= 1674 THEN
c = 137.47 + 348 t N T -3 * T (I, J - 1)
ELSEIF 1033 < T (I, J - 1) AND T (I, J - 1) <= 1181 THEN
c = 672.4
ELSEIF 273 < T (I, J - 1) AND T (I, J - 1) <= 1033 THEN
c = 312 + 442.3 * 10 * -3 * T (I, J - 1)
END IF '
T ( l, J) = T (1, J - 1) + ts / (x ' 2 * p * c) * ( (k * (2 * T(2, J - 1) - 2 * T (l, J - 1))))
IF Tb < (LI / Vc) THEN
Qo = Rw * Cw * Vw * (Tout - Tin) /(Sw * LI)
ELSEIF (LI / Vc) <=Tb AND Tb <(L2/ Vc) THEN
Qo = hs * (Vsr / (2 * 60 * Sw *(L2- LI))) ‘ .75
ELSEIF (L2 / Vc) <= Tb AND Tb <(L3/ Vc) THEN
Qo = hs * (VzlA 1 (2 * 60 * Sw * (L3 - L2))) “ .75
 116

Appendix A

ELSEIF (L3 / Vc) <= Tb AND Tb < (L4 / Vc) THEN

Qsp = hs * (VzlB / (2 * 60 * Sw * (L4 - L3))) * .75
Qrc = NrzlB * Rc * hczlB * (T(N, J - 1) - Trs)
Qrad = (NrzlB * 2 * (RpzlB - RzlB) / (L3 - L4)) * G * E * (T(N, J - 1) * 4 - Ta ' 4)
Qo = Qsp + Qrc + Qrad
ELSEIF (L4 / Vc) <= Tb AND Tb < (L5 / Vc) THEN
Qsp = hs * (Vz2 / (2 * 60 * Sw * (L5 - L4))) * .75
Qrc = Nrz2 t Rc * hcz2 * (T(N, J - 1) - Trs) / (L5 - L4)
Qrad = (Nrz2 * 2 * (Rpz2 - Rz2) / (L5 - L4)) * G * E * (T(N, J - 1) “ 4 - Ta * 4)
Qo = Qsp + Qrc t Qrad
ELSEIF (L5 / Vc) <= Tb AND Tb < (L6 / Vc) THEN
Qsp = hs * (Vz3 / (60 * Sw * (L6 - L5))) ' .75
Qrc = Nrz3 * Rc * hcz3 * (T(N, J - 1) - Trs) / (L6 - L5)
Qrad = (Nrz3 * 2 * (Rpz3 - Rz3) / (L6 - L5)) * G * E * (T(N, J - 1) * 4 - Ta ‘ 4)
Qo = Qsp + Qrc + Qrad
ELSEIF (L6 / Vc) <= Tb AND Tb < (L7 / Vc) THEN
Qsp = hs * (Vz4 / (60 * Sw * (L7 - L6))) * .75
Qrc = Nrz4 * Rc * hcz4 * (T(N, J - 1) - Trs) / (L7 - L6)
Qrad = (Nrz4 * 2 * (Rpz4 - Rz3) / (L7 - L6)) * G * E * (T(N, J - 1) ‘ 4 - Ta ‘ 4)
Qo = Qsp + Qrc + Qrad
ELSEIF (L7 / Vc) <= Tb AND Tb < (L8 / Vc) THEN
Qrc = Nrwd * Rc * hwd * (T(N, J - 1) - Trs) / (L8 - L7)
Qrad = (Nrwd * 2 * (Rpwd - Rwd) / (L8 - L7)) * G * E * (T(N, J - 1) ‘ 4 - Ta * 4)
Qo = Qrc + Qrad
END IF
T(N, J) = T(N, J - 1) + 2 * ts / (x * 2 * p * c) * (k * (T(N - 1, J - 1) - T(N, J - 1)) - x * Qo)

J =J + 1
IF J > M THEN !
EXIT DO
END IF
LOOP .

LPRINT " RESULTS OF THE PROGRAM“

LPRINT " ======================"
J =2
LPRINT
LPRINT ‘ CASTING PARAMETERS“
LPRINT 1 .............................." -
LPRINT * 1. SLAB WIDTH ="; Sw; V
LPRINT * 2. LIQUIDUS TEMPERATURE TL; T
LPRINT " 3. TUNDISH TEMPERATURE =”; Tt; “K *
LPRINT “ 4. CASTING SPEED ="; Vc; “m/second"
LPRINT " 5. CASTING TIME =»; tc; “ min."
LPRINT " 6. WATER RATE:*
LPRINT " (a) SPRAY RING Vsr; "1/mn."
LPRINT “ (b) ZONE-1A =■; VzlA; “l/niin.B
 117

Appendix A

LPRIRT ' (e) IONE-1B ='; Vx 1B; ' 1/iin.1 1

LPRIRT ' (d) IONB-2 ='; Vx2; M/»io.'
LPRIRT ' (e) IONE-3 =’ ; Vs3; 'l/iln.'
LPRINT ' (f) IONE-« ='; Vii; M/iin.'
LPRIRT
LPRIRT
LPRIRT
LPRINT ' RODE TEMPERATURES (CALCULATED AS A FDNCTION OF TIME)
LPRIRT '
LPRINT
LPRIRT
LPRIRT 'Scc. •; ' 1'; ' 2'; ' 3'; * I'; • 5'; ' 6'; ' 7'; ' 8'; ' 9*; ' IO’; ' CORRER'
LPRIRT
Tb * 0
LPRIRT DSIRG ' U H » ; IRT(Tb); SPC(2);
N*l
DO
LPRINT INT(T(V, 1)) - 273;
V :V t 1
IF V > 1Q TBEX
EIIT DO
END ir
LOOP
LPRINT IRT(T(9, 1)) - 273
110 DO
Tb = (J - 1) » ts
Tbl *ABS(LI / Ve - Tb)
Tb2 =ABS(L2 / Ve - Tb)
Tb3 =ABS(L3 / Ve - Tb)
Tbl =ABS(L4 / Ve - Tb)
Tb5 =ABS(L5 / Ve - Tb)
Tb6 sABS(L6 I Ve - Tb)
Tb7 =ABS(L7 / Ve - Tb)
Tb8 =ABS(L8 / Ve - Tb)
IF Tb = 0 OR Tbl < 5 OR Tb2 < 5 OR Ib3 < 5 OR Tbl < 5 OR Tb5 < 5 OR Tb6 < 5 OR Tb7 < 5 OR Tb8 < 5 TBER GOTO 120 ELSE GOTO 130
120 LPRINT OSING ,| | f f I R T ( T b ) ; SPC(3); INT(T(1, J * 1)) - 273; SPC{2);
,1*2
DO r
LPRINT USING '1118'; INT(T(1, J - 1)) - 273; SPC{2); .
1=1*1
IF I > (R - 1) TBEN
EIIT DO
END IF
LOOP •
LPRINT DSIRG ' U H ' ; 1RT(T{N, J)) - 273; SPC{2); INT(T(N, J)) - ( 27 3 ♦ 90)
130 J = J ♦ 1
IF J ) « TBEX
EIIT DO
END IF
LOOP .
END
Appendix A

RESULTS OF THE PROGRAM (NORMAL WATER RATE)

CASTING PARAMETERS

1. SLAB WIDTH 1.2 i

2. LIQUIDUS TEMPERATURE 1789 K
3. TUNDISH TEMPERATURE 1817 K
4. CASTING SPEED .015 m/second
5. CASTING TIME = 17.9 min.
6. WATER RATE:
(a) SPRAY RING = 89 1/min.
(b) ZONE-1A = 380 1/min.
(c) ZONE-IB = 206 1/min.
(d) ZONE-2 = 284 1/min.
(e) ZONE-3 = 206 1/min.
(f) ZONE-4 = 129 1/min.

NODE TEMPERATURES (CALCULATED AS A FUNCTION OF TIME)

Sec. 1 2 3 4 5 6 7 8 9 10 CORNER

0 1544 1544 1544 1544 1544 1544 1544 1544 1544 1544 1544 1 r
36 1544 1543 1543 1543 1542 1535 1507 1421 1231 877 787 Mould
39 1543 1543 1543 1543 1541 1530 1493 1392 1189 841 751 1
43 1543 1543 1543 1543 1539 1524 1476 1362 1146 900 810 “
47 1543 1543 1543 1542 1537 1517 1462 1336 1115 923 833
50 1543 1543 1543 1541 1534 1510 1446 1311 1107 936 846
54 1543 1543 1543 1540 1530 1502 1430 1292 1103 933 843
86 1540 1538 1530 1511 1473 1408 1312 1187 1043 880 790 Z o n e - IA / IB
89 1540 1537 1528 1507 1467 1401 1304 1180 1036 905 815
93 1539 1536 1526 1503 1462 1394 1296 1172 1029 918 828
125 1521 1515 1495 1459 1404 1329 1237 1137 1039 949 859
129 1518 1511 1491 1454 1399 1323 1233 1135 1039 966 876 -
132 1515 1508 1487 1449 1393 1318 1229 1133 1038 974 884 Zone-2
319 1337 1331 1313 1285 1248 1205 1157 1105 1052 996 906
322 1334 1328 1311 1283 1246 1203 1155 1104 i 1051 996 906 Zo ne-3
609 1166 1162 1151 1133 1109 1078 1042 1002 957 908 818
612 1164 1160 1149 1132 1107 1077 1041 1001 956 907 817
616 1162 1159 1148 1130 1105 1075 1040 999 955 906 816 "
Zo ne -4
802 1091 1087 1077 1061 1039 1011 977 939 897 851 761
805 1089 1086 1076 1060 1037 1009 976 938 896 850 760 -
809 1088 1085 1075 1059 1036 1008 975 937 895 864 774
1013 1019 1018 1013 1005 996 985 973 960 949 938 848 W ith d ra w a l
1017 1018 1017 1012 1005 995 984 973 961 949 938 848
1020 1018 1016 1011 1004 995 984 972 961 949 938 848
Appendix A

RESULTS OF THE PROGRAM (REDUCED WATER RATE)

CASTING PARAMETERS

1. SLAB WIDTH = 1.2 m

2. LIQÜIDUS TEMPERATURE = 1789 K
3. TUNDISH TEMPERATURE = 1817 K
4. CASTING SPEED = . 015 m / s e c o n d
5. CASTING TIME = 17.9 min.
6. WATER RATE:
(a) SPRAY RING = 89 1 / m i n .
(b) ZONE-11 = 380 1 / i i n .
(c) ZONE-1B = 206 1/ m i n .
(«) ZONE-2 = 2 8 4 1/ m i n .
(e) ZONE-3 = 183 1 / m i n .
(f) ZONE-4 = 119 1 / m i n .

NODE TEMPERATURES (CALCULATED AS A FUNCTION OF TIME)

Sec. 1 2 3 4 5 6 7 8 9 10 CORNER

0 1544 1544 1544 1544 1544 1544 1544 1544 1544 1544 1544
36 1544 1543 1543 1543 1542 1535 1507 1421 1231 877 787 Mould
39 1543 1543 1543 1543 1541 1530 1493 1392 1189 841 751
43 1543 1543 1543 1543 1539 1524 1476 1362 1146 900 810
47 1543 1543 1543 1542 1537 1517 1462 1336 1115 923 833
50 1543 1543 1543 1541 1534 1510 1446 1311 1107 936 846
54 1543 1543 1543 1540 1530 1502 1430 1292 1103 933 843
86 1540 1538 1530 1511 1473 1408 1312 1187 1043 880 790 Z o n e - 1 4 / lB
89 1540 1537 1528 1507 1467 1401 1304 1180 1036 905 815
93 1539 1536 1526 1503 1462 1394 1296 1172 1029 918 828
1'
125 1521 1515 1495 1459 1404 1329 1237 1137 1039 949 859
129 1518 1511 1491 1454 1399 1323 1233 1135 1039 966 876
132 1515 1508 1487 1449 1393 1318 1229 1133 1038 974 884 Zo ne-3
319 1337 1331 1313 1285 1248 1205 1157 1105 1052 996 906
322 1334 1328 1311 1283 1246 1203 1155 1104 1051 998 908 ~ r
609 1170 1166 1156 1139 1115 1086 1052 1014 972 925 835 Zo ne -3
612 1168 1165 1154 1137 1114 1085 1051 1013 971 924 834
616 1166 1163 1152 1135 1112 1084 1050 1011 969 922 832
Zo ne -4
802 1095 1092 1082 1067 1045 1019 987 950 910 865 775
805 1094 1090 1081 1065 1044 1017 986 949 909 864 774
809 1092 1089 1080 1064 1043 1016 985 948 908 878 788
1013 1026 1024 1020 1013 1003 992 981 969 957 946 856 Wi th d ra w a l
1017 1025 1024 1019 1012 1003 992 981 969 957 946 856
1020 1024 1023 1018 1011 1002 992 980 969 957 946 856
 120

Appendix B

INDEX OF TRANSVERSE CRACKS AND GRAIN SIZE OF THE CAST STRUCTURES

OF STEELS A, B, C AND D HAVING VARIUOUS TITANIUM CONTENTS, CAST
USING WATER RATES 0.756 l/kg AND 0.736 l/kg

S te e ls Ti W a te r r a te N r. o f A STM Mr., (%) D e f e c t s (% )

(wt. %) .... d/kg) s la b s 6 7 8 9

At -
0.756
A2 -
0.756 17 6.5 39.1 36.9 17.4 58.8
A3 -
0.756

B1 0.007 0.756
B2 0.009 0.756 17 0 36.6 46.3 17.0 52.9
B3 0.006 0.756

Cl 0.017 0.756
C2 0.017 0.756 17 0 18.6 55,8 25.5 41.2
C3 0.018 0.756

D1 0.026 0.756
D2 0.028 0.756 16 0 19.6 58.7 21.7 37.5
D3 0.025 0.756

A4 -
0.736
A5 -
0,736 16 4.5 40.9 36.4 18.2 31.3
A6 -
0.736

B4 0.006 0.736
B5 0.006 0.736 17 0 37.2 44.1 18.6 23.5
B6 0.009 0.736

C4 0.018 0.736
C5 0.016 0,736 17 0 18.8 57.9 23.7 17,6
C6 0.017 0.736

D4 0.026 0.736
D5 0.028 0.736 16 0 19.5 58.5 22.0 18.8
06 0.029 0.736
 121

Appendix C

MAIN SPECIFICATION OF CONTINUOUS CASTER

Type of machine: Bow-type Slab machine

Fabricator: SM S Schloemann-Siemag Aktienge-

sellscaft

Casting radius: 9,700 mm with one straightening

point

Mould material: Unplated copper-based alloy

M oud tapper: 1.1%

Casting slab size: Ranges Experiments

- thickness 200 mm 200 mm
- width [950 -1 630 mm 1200 mm
- length 6500-10500 mm 9500 mm

Mould Oscillations :
- frequency 105 cycles/min.
- stroke 6 mm

Secondary cooling: water cooling

Roll diameters: Nozzle type:
- spray ring - Flat jet
- zone-l A /l B 180 mm Full cone/Flat jet
- zone-2 245 mm Slot type
- zone-3 310 mm Slot type
- zone-4 310 mm Slot type
- withdrawal 390 mm -
 122

A ppendix D

SPECIFICATION OF THE MOULD POWDER

Trade Mark: Accutherm ST-C39/4PS-110

Chemical Composition:
CaO 34.5%
Si02 31.50%
A I 2O 3 0.60
Na20 6.00%
F 5.75%
MgO 0.60%
Fixed C 5.75%
C 02 i 0.80%
Basicity 1.10

Properties:
Softening point , 1110 °C
Melting point 1135 0C
Fluid point 1150°C