i | P a g e

UNDERSTAND CLOGGING IN IF GRADE AND

SUGGEST COUNTERMEASURES TO MINIMIZE IT
A Project Report Submitted in Fulfilment of the Requirements for Summer Training




By

ABHIJEET DASH (VT20132141)
Metallurgical and Materials Engineering
National Institute of Technology, Rourkela




Under the Guidance of


Mrs Shainu Suresh
Sr. Manager, Steel Technology and Casting
Flat Product Technology Group,
Tata Steel Ltd. Jamshedpur
ii | P a g e



DECLARATION

I hereby declare that the project work entitled Understand clogging in IF grade and
suggest countermeasures to minimize it is an authentic record of my own work
carried out at Tata Steel Ltd., Jamshedpur as requirement of short term
industrial/research experience for the award of degree of B.Tech Metallurgy,
National Institute of Technology, Rourkela, under the guidance of Mrs Shainu Suresh
during 14
th
May to 9
th
July, 2013


Place: Jamshedpur Abhijeet Dash
Date: 8
th
July,2013 VT20132141

Certified that the above statement made by the student is correct to the best of my
knowledge and belief.

Mrs Shainu Suresh
Sr.Manager, Steel Technology and Casting
Flat Product Technology Group,
Tata Steel Ltd. Jamshedpur.



iii | P a g e


iv | P a g e

ACKNOWLEDGEMENT
I would like to thank the Management of Tata Steel Ltd. for giving me this wonderful
opportunity to work with highly knowledgeable people on a project of great
importance. I owe my deepest gratitude to my guide Mrs. Shainu Suresh, Sr.Manager,
Steel Making and Casting Technology, FPTG and my industry coordinator Dr. T.K.
Roy, Head, Steel Technology, FPTG, whose encouragement, guidance and support
from the initial to the final level enabled me to develop an understanding of the
subject. I would also like to acknowledge the valuable inputs from Mrs. Akasmita
Biswal, Manager,FPTG-TS at each and every step of the project.. I convey my deep
appreciation to all the technical and administrative staffs at LD#2&SC,FPTG team
and SNTI for their help and support during the tenure of my project work.
I am really indebted to NIT, Rourkela for such an opportunity to undertake this
project work as a part of Summer Training.
I feel honoured to avail the opportunity to express my sincere gratitude to Dr. Sunil
K. Sarangi, Director (NIT Rourkela),Prof. B.C. Ray, HOD, Metallurgical And
Materials Engineering and all the faculty members of Metallurgical And Materials
Department of NIT Rourkela for their encouraging support and guidance in carrying
out this project.

v | P a g e

ABSTRACT
In the casting of Ti-bearing ULC steels, clogging of the submerged entry nozzle
(SEN) affects both the productivity of the caster and the quality of cast products. The
plant experience shows a negative influence of increasing Ti and P contents on the
clogging tendency of ULC steel grades. Analytical investigations on samples from
the process and laboratory experiments have been carried out in order to gain a better
understanding of the influence of Ti and P on the clogging phenomenon of ULC steel
grades. After casting Ti and P alloyed ULC steels, the deposits from clogged SEN
have been analysed. The results indicate that the particles consist mainly of alumina
(Al
2
O
3
> 90 mass%), while the Ti content makes up only a very small portion (0.58
3.58 mass%). An automated inclusion analysis by OES-PDA shows that the size of
oxide inclusions decreases with increasing Ti content. The determination of the
influence of Ti and P solutes on the wetting in the liquid Fe/Al
2
O
3
system indicates
that both elements lower the wetting angle and reduce the melt surface tension.
Concerning Ti, the decrease of the wetting angle occurs only in the case of reactive
wetting. The effects of Ti and P on clogging are found to be significantly different: Ti
lowers the size of Al
2
O
3
inclusions only if an interfacial wettable surface layer forms.
Small alumina inclusions enhance the clogging proneness. The time between Al
deoxidation and Ti addition is thus an essential process variable in order to prevent
clogging. As for P, it increases the Ti activity if both solutes are added to the melt.
Moreover, it lowers the melt viscosity, thus increasing the wettability in the Fe-Ti-P
system. The negative effect of P on clogging is, therefore, less pronounced than that
of Ti and presumably harder to influence.

Key words: continuous casting, interstitial free steel, non-metallic inclusions, nozzle
clogging, ultra low carbon steel, wetting




vi | P a g e

CONTENT
Page No.
Declaration i
Acknowledgement ii
Abstract iii
1. CHAPTER I: Introduction 1-3
1.1. Tata Steel 1
1.2. Interstitial free steel 3
1.3. Objectives 3
2. CHAPTER II: Literature review 4-35
2.1. Production practice for IF grades 4
2.1.1 Steelmaking 4
2.1.1.1 Carbon Removal 4
2.1.1.2 Nitrogen Removal 5
2.1.1.3 Oxygen Control 5
2.1.2 Vacuum Degassing 6
2.1.2.1 Carbon Control 7
2.1.2.2 Nitrogen Control 8
2.1.2.3 Oxygen Control 8
2.1.2.4 Interstitial Stabilization via Precipitation 8
2.1.3 Casting 10
2.2. Nozzle Clogging 12
2.2.1. Morphology 14
2.2.2. Mechanisms of inclusion formation 20
2.2.2.1 Thermally Induced Inclusions 21
2.2.2.2 Diffusion of Oxygen through Nozzle Refractory 22
2.2.2.3 Flow Transported Inclusions 22
vii | P a g e

2.2.2.4 Thermochemically formed Inclusions 24
2.2.2.5 Discussion of Inclusion Mechanisms 26
2.2.3. Countermeasures 26
2.2.3.1 Oversized Nozzles 27
2.2.3.2 Argon Injection 27
2.2.3.3 Calcium Treatment 29
2.2.3.4 Lime Bearing Nozzle Materials 30
2.2.3.5 Other Nozzle Materials 31
2.2.3.6 Nozzle Modifications 32
3. CHAPTER III: Experimental Methods and Data Analysis 36-37
3.1. Materials Studied 36
3.2. Morphology Characterization 37
4. CHAPTER IV: Results and Discussion 38-48
4.1. Visual Observation 38
4.2. Data Analysis 41
5. CHAPTER V: Conclusion and Recommendation 49
5.1. Conclusion 49
5.2. Scope of Future Work 49
6. REFERENCES 50-51

1 | P a g e

CHAPTER 1
1. Introduction
1.1 TATA Steel
Established in 1907, Tata Steel is among the top ten global steel companies with an
annual crude steel capacity of over 28 million tonnes per annum (mtpa). It is now one
of the world's most geographically-diversified steel producers, with operations in 26
countries and a commercial presence in over 50 countries.
Tata Steels larger production facilities include those in India, the UK, the
Netherlands, Thailand, Singapore, China and Australia. Operating companies within
the Group include Tata Steel Limited (India), Tata Steel Europe Limited (formerly
Corus), NatSteel, and Tata Steel Thailand (formerly Millennium Steel).
Tata Steels plant in Jamshedpur, Jharkhand with production of 6.855 million tonnes
of crude steel in 2010-2011 exceeded its installed capacity of 6.8 MTPA. The
company produced 6.691 million tonnes of saleable steel during last fiscal.
The flat product division of Tata steel, Jamshedpur supplies automobile skin panels to
almost every automobile manufacturing company in India. The flat product division
also achieved the strategic milestone of more than 1 million tonnes auto sales, an
increase of 18% from last year.
The flat product division relies on the LD#2 and Slab Caster department for
providing quality slabs for rolling into Hot Rolled and Cold Rolled coils.
LD#2 and Slab Caster
LD#2 and Slab Caster consists of:
1. 3 Nos. Hot metal Desulphurization units
2. 3 Nos. 160 ton LD Convertors for Steel making
3. 2 Nos. 160 ton Ladle Furnace
4. 1 Nos. RH Degassing unit
5. 3 Nos. Single Strand Continuous slab casters
Caster-A is vertical mould caster and Caster-B & C are curved mould casters.


2 | P a g e


Figure 1: Process flow at LD#2 and Slab Caster
Process Overview:
Hot metal from G and H blast furnaces is brought in torpedo cars for steelmaking at
LD#2 & SC.
1. Desulphurization: Hot metal from Blast Furnace has high amount of S in it. Since
subsequent LD vessels cannot maintain conditions which can eliminate Sulphur
efficiently, it is externally desulphurized at this station. A mixture of Mg and CaC
2

powder are injected pneumatically to the DS ladle through a lance.
2. LD Convertor: LD convertor derives its name from the name of two towns in
Austria viz. Linz and Downawtiz where it was developed. In LD convertor,
impurities like C, Si, Mn, and P in the hot metal are removed by forming their
oxides using oxygen lance.
3. On-line Purging: Its purpose is to homogenize steel with regard to chemistry and
temperature by purging with Ar. Deoxidizer (Al) in the form of wire can be fed
into liquid steel for killing. Also alloying additions are possible at this station.
4. Ladle Furnace: In this, the ladle containing liquid steel is slid below a lid fitted
with electrodes. So it acts as an arc furnace for heating the metal. Various Ferro
alloys and fluxes are added to the ladle while Argon is purged for
homogenization.
5. RH Degasser: The RH unit is used to produce Ultra Low carbon steels. The
process involves vacuum decarburization and subsequent alloying. After
decarburization is completed, the balance oxygen is deoxidized either by
aluminium as per the requirement. Other Ferro alloys may be added after
deoxidation.
Steel from Ladle furnaces, RH degasser or directly from On-line purging station is
sent to the Slab casters.


3 | P a g e

1.2 Interstitial Free Steel
Interstitial free steels have a nearly complete fixation of all interstitial dissolved
nitrogen and carbon atoms, i.e neutralisation of N and C. The complete killing of the
steel is performed by the micro alloying titanium and /or niobium. By this method the
N and C atoms are arrested and all the negative effects described by the Cottrell-
Cloud theory are very much reduced or even abolished .These are the upper and
lower yield point ( yield point effect, discontinuous yielding), strain aging and blue
brittleness.
In other words steel produced with very low amounts of interstitial elements
(primarily carbon and nitrogen) with small amounts of Titanium or Niobium added to
tie up the remaining interstitial atoms. Without free interstitial elements, these steels
are very ductile and soft, will not age or bake harden and will not form strain
(Luders) lines during forming due to the absence of yield point elongation.

Interstitial free grades have specifically used as automotive parts such outer body
applications because of
Cleanliness of the grade
Excellent drawability due to Ti stabilization
High strength and good elongation
Excellent formability
1.3 Objectives
To find a statistical correlation between the different parameters coming during
the different stages of production of interstitial free grades of steel to minimize
clogging
To suggest relevant metallurgical methods to avoid clogging







4 | P a g e

CHAPTER-2
2.0 LITERATURE REVIEW
2.1Production practice for IF grades
2.1.1 Steelmaking
IF steel grades are generally produced by integrated steel producers and the process
involves primary refining in any type of converter, followed by vacuum degassing
and finally continuous casting. The various primary refining vessels range from top
blown BOF to the bottom blown Q-BOP vessel. The bottom blown vessels vary
according to the type of bottom blown gas used or the bottom gas blow rate. Inspite
of the different configuration the primary refining vessels serve the same purpose of
carbon removal and provide sufficient temperature for subsequent processing.

Figure 2 Tapping of hot metal into LD converter
2.1.1.1 Carbon Removal
Carbon removal in the steelmaking vessel occurs by reacting oxygen which is blown
into the bath with the carbon dissolved in the liquid steel to produce carbon monoxide
and carbon dioxide gas. the reaction are as follows:

5 | P a g e

C+0.5O
2
= CO
C+O
2
=CO
2
where the rate of decarburization is determined by the rate of oxygen supplied
through lance. For normal oxygen blow rates ,after silicon has been removed ,the rate
of decarburization decreases linearly with blowing time until about 0.3% carbon
when the rate of decarburization is controlled by mass transfer of carbon from the
liquid to the reaction interface
Rate of decarburization is
Log[(%C)t-(%C)t]=-k(t-t
o
)
Vessels with intense bottom blowing are best in producing lower carbon content as
they increase the rate of transport of carbon to the reaction interface
The achievable end point carbon content form steelmaking is important as it directly
controls the time required for the degasser to achieve the final aim carbon content.
Operation of primary refining vessel thus becomes a compromise between producing
as low as carbon content as possible and minimizing the slag iron oxide production
2.1.1.2 Nitrogen Removal
Nitrogen levels in primary refining vessel are controlled by carbon monoxide gas
evolution which is the primary decarburization product during most of the main
oxygen blow. The vigorous stirring due to the evolution of the carbon monoxide
serves to flush out most of the nitrogen in solution in the hot metal coming from the
blast furnace .As the decarburization reaction tails off towards the end of the main
blow ,air is able to infiltrate into the refining vessel and renitrogenization occurs.
Renitrogenization may also be due in part to late melting scrap which is high in
nitrogen content. Air infiltration and nitrogen pickup, can be controlled by using a
high foaming slag practice at blow end which serves as physical barrier to air
infiltration into the furnace. Carbon Dioxide when used as a replacement for argon as
bottom stirring gas accelerates nitrogen removal.
Another source of nitrogen pickup is during tap. The rate of nitrogen pickup in liquid
steels is greatly enhanced by if oxygen and sulphur levels in steels are very low.
Oxygen and Sulphur are surface active elements and segregate to the surface of the
liquid iron. Their presence on the surface acts as a rate retarder since they occupy the
reaction sites for nitrogen dissociation thereby greatly reducing the effective number
of possible reaction sites and reducing the overall reaction rate.
2.1.1.3 Oxygen Control
For clean steelmaking practices great care must be taken to ensure that the minimum
possible amount of furnace slag is carried over to the ladle. Carried over furnace slag

6 | P a g e

results in unstable secondary steelmaking operations mainly because of the reducible
oxides FeO and MnO in the slag. These oxides usually result in substantial
fluctuations in alloy recovery.
Techniques available for limiting furnace slag carryover include floating a refractory
ball, dart of other shaped piece , which has density intermediate between that of slag
and steel, over the furnace tap hole during tap. Another technique for slag control
form the furnace is to place electromagnetic coils in the tap hole of the vessel which
detect the onset of slag flow and indicates the operator. In addition to the control of
dissolved oxygen it is important to avoid entrainment of any of slags used in type
furnace and on the ladle as entrainment of these slags can lead to a significant
exogenous inclusion content in the steel which can be difficult to remove in
subsequent processing
2.1.2 Vacuum Degassing
Vacuum degassing of the steel is necessary to get very low levels of carbon required
for IF steels i.e.0.005% C. The most common type of degasser used is the circulation
type degasser. In this type of degasser, two refractory tubes or snorkels are immersed
into liquid steel in ladle , vacuum is applied in the chamber above the snorkels, and
argon gas is injected into the steel through a number of injection points.

Figure 3 RH Degasser in operation

7 | P a g e


2.1.2.1 Carbon Control
Decarburization occurs by the reaction
C + O = CO
gas

Decarburization is continued until sufficient time has elapsed to produce the desired
carbon content and also on the degassing rate achievable with the degasser. In RH
degasser the decarburization rate increases with increasing steel circulation rate and
with injected argon gas flow rate. The ability to increase argon gas flow rate is
limited by the amount of splashing that occurs in the vacuum vessel, too much
splashing results in skull build up. Also metal circulation rate can be increased by
increasing the snorkel diameter but it decreases the refractory thickness leading to
snorkel wear. Thus a balance has to be struck between refractory wear and circulation
rate.

Figure 4 Rate of removal of carbon
In general, decarburization rate follows exponential law
C
t
= C
o
exp(k*xt)
Oxygen blowing is introduced into the liquid steel bath in the lower portion of the
vacuum vessel through gas shielded annular tuyeres to promote faster rate of
decarburization. Once decarburization is complete, the steel is deoxidised and any
requires alloy additions are made. The Titanium or Niobium necessary to stabilize the
remaining carbon and nitrogen in the product are added at this time. Carbon pickup
may be from the contamination due to vessel skulls.

8 | P a g e

Q = 7.4310
3
D
4/3
G
1/3
(ln p / p )
1/3
Q- circulation flowrate ,kg/min
G-lift gas flowrate ,m
3
/min
D-snorkel diameter ,cm
P1-atomosphere pressure,torr
P2-vacuum vessel pressure ,torr
2.1.2.2 Nitrogen Control
Vacuum degasser is not effective in removing nitrogen form liquid steel since the rate
of nitrogen removal is usually controlled by a slow chemical reaction on the liquid
iron surface.
2.1.2.3 Oxygen Control
Steel cleanliness in the degasser is controlled primarily by the amount of deoxidiser
added and the subsequent stirring done for inclusion flotation and removal.
Reoxidation sources such as oxidised ladle slag, oxidized degasser vessel skulls and
degasser leaks must also be minimized to ensure steels with low oxide inclusion
contents.

2.1.2.4 Interstitial Stabilization via Precipitation
The control of nitrogen, carbon, and sulfur content during steelmaking and casting is
the first step in production of interstitial-free steels. The next critical metallurgical
process is the removal of solute nitrogen and carbon from solution via precipitation
reactions. Solute nitrogen is effectively removed from solid solution via TiN or AlN
precipitation. The stabilization of carbon, however, may differ in a variety of
interstitial-free steels and has been the subject of many studies (17-34). Based on the
information published in the literature, several possible mechanisms for the
stabilization of carbon in interstitial-free steels have emerged. In the early days of
research into interstitial-free steels, it was assumed that interstitial-free steels could
be considered as dilute microalloyed (HSLA) steels. That is, the precipitation
behaviour in these two types of steels was thought to be similar, except for the
amount of precipitates formed.

9 | P a g e


Figure 5 Stability Diagram
Based on this view, the products of the precipitation reactions were assumed to be as
follows: AlN, TiN, TiS, TiC, and Nb(C,N). It is now well known that there are
several types of interstitial-free steels where the precipitation behaviour might be
different depending on subtle differences in steel compositions.
Niobium additions are typically made to stabilize interstitial species; however, there
are cases where solute niobium will be desired for property development. As will be
discussed in subsequent sections, niobium additions to these grades may be made to
enhance other properties. This solute niobium has been shown to segregate to grain
boundaries, to sub grain boundaries, and to free surfaces. The benefits of solute
niobium may be substantial. Hook et al. (39, 40) showed a strong relationship
between the solute niobium level and both the final texture and formability. In
addition, the segregation of niobium to the grain boundaries will become important in
the microstructural development of galvanneal coatings as well as cold work
embrittlement resistance, as discussed in more detail in later sections. In these grades,
interstitial stabilization primarily occurs through precipitation with titanium, similar
to that found in Ti-stabilized steels. The traditionally accepted precipitation sequence
in Ti-stabilized steels with atomic ratio S:C:N. 1:1:1, is TiN, TiS*, Ti4C2S2 and TiC,
as the temperature decreases. The amount of Ti required to fully stabilize interstitial-
free steels was expressed as (wt%):
Tistabilize = 3.42N + 1.5S + 4C

10 | P a g e


Figure 6 Ti and Carbon stability diagram
Log K
sp
, TiC = log {[Ti][C]} = 9.0 10330/T
where T is the temperature in Kelvin, [Ti] and [C] are solute concentration in weight
percent. This solubility product can be used to predict the precipitation temperature in
a Ti-stabilized steel. For steels whose compositions are above the stoichiometric line
(Ti*/C = 4, Ti* = Titotal 3.42N 1.5S + Nb/1.94),
there is sufficient titanium to stabilize carbon. However, whether complete
stabilization can actually be obtained is also dependent on the kinetics of the reaction.
Hence, similar steels may exhibit different levels of stabilization (or different
precipitates) depending upon the thermo mechanical processing schedules employed
in the hot strip mill.
2.1.3 Casting
Prevention of contamination of the steel during continuous casting is the major
problem faced by the steel makers. There are no major solidification problems
associated with the casting of IF steels as long as there is a significant superheat and
these steels are not prone to major cracking problems. Since these steels are very
pure, there is only a small liquidus-solidus gap and care must be taken to ensure
transport of type steel to the mold.

11 | P a g e


Figure 7 Continuous casting-ladle on turret


Figure 8 Continuous casting Torch cutter

12 | P a g e

But Ti stabilized IF steel grades experience a major nozzle clogging problem during
casting leading to downgrading of slabs meant for automotive applications.
2.2 Nozzle Clogging
Nozzle clogging is a build-up of inclusions on the internal bore of refractory nozzles
used in the continuous casting of steel. The purpose of the refractory nozzles is to
protect the steel stream from reoxidation and control its flow as the steel moves from
the tundish to the mold of a continuous caster. The inclusions eventually build up to a
level that either completely blocks the flow of steel or disrupts the flow of steel to the
extent that casting must be aborted. Nozzle clogging in the continuous casting of
aluminium-killed steels has been a problem for nearly 30 years. Snow and Shea first
reported observing alumina and steel nodule inclusion formations on the walls of
ladle nozzles associated with the teeming of fine grained aluminium killed steels in
1949[1]. This observation was made while studying ladle nozzle erosion. No effort
was made to understand the problem because at that time, steel was teemed into ingot
molds and clogging was not considered a serious problem. Ladle nozzles could easily
be changed or opened with an oxygen lance after casting a heat of steel. It was not
until the advent of continuous casting that nozzle clogging would be recognized as a
major hindrance to productivity. Continuous billet casters in the late 1960s and early
1970 have utilized metered nozzles. Nozzle blockage with these casters caused the
casting speed to be reduced to keep the mold level constant, which hindered
production. Eventually, oversized nozzles were introduced with stopper rods to
control the flow of steel to the mold. The oversized nozzles were a means to counter
clogging by allowing more time before clogging became so severe that the casting
rate would have to be slowed. Shrouding is particularly important with the use of
stopper rods because the steel stream is considerably disturbed so that reoxidation
becomes a serious problem[2]. Clogging continues to be a problem with modern slab
and thin slab casters. Several methods to counter clogging have been introduced such
as calcium treatment and argon injection, however clogging often continues to limit
productivity in the steel industry.

13 | P a g e


Figure 9 (a) Negligible deposition in the barrel of a SEN (after 6 h of use). (b) Transverse section of the SEN, cut
at about 10 cm above the port. (c)Bulk and typical shape of the deposit removed from the port of the same SEN





14 | P a g e

2.2.1 MORPHOLOGY
Most researchers have focused on inclusions associated with low or ultra-low-carbon,
aluminium-killed steels that have not been calcium treated. Therefore, inclusion
morphologies with these types of steels will be discussed first. Then, key factors that
have been reported to effect morphology will be presented. The morphology of
inclusions reported by different researchers varies somewhat. Ogibayashi pointed out
that the differences are likely due to variations in the cleanliness of steel and casting
conditions[1]. However, in all cases, the inclusions consist of powdery, friable
alumina and/or metal nodules containing clusters of alumina inclusions. Many
researchers have pointed out that the morphology of the alumina inclusions in the
steel nodules are the same as the alumina built up on the refractory wall[4,5]. Most
commonly, the alumina clusters consist of individual, round alumina particles
sintered together. The clusters have a coral-like appearance.
Ogibayashi and Poirier et al., have separately described the inclusion formations as
consisting of two layers [3.6]. The first layer, at the refractory-inclusion interface, is
densely packed alumina with small particles of metal contained within the alumina
network. The metal particles in this layer are reported to contain relatively high
concentrations of silicon[3]. This layer has also been reported to contain a small
amount of a vitreous phase consisting of Al2O3, SiO2, Na2O, and K2O[6]. The
maximum thickness of this layer was reported as 300 m. The next layer, the build-
up layer, consists of either alumina clusters with metal nodules containing clusters of
alumina or mostly metal nodules containing clusters of alumina inclusions. Poirier et
al. have also reported observing traces of spinels (MgO-Al
2
O
3
and MgOFe
2
O
3
) in
build-up layer[6]. Many researchers have also reported the formation of a
decarburized layer in refractory materials containing graphite. The decarburized layer
is adjacent to the inclusion and has been reported to have a depth of approximately
400 m[6]. Many factors have been reported to effect inclusion morphology. Among
these are: location of the build-up in the pouring system, steel chemistry, deoxidation
method, ladle treatment, and composition of the refractory. Location seems to play an
important role in the morphology of inclusions in submerged entry nozzles (SENs).
Kasai et al. observed that with various grades of aluminium and aluminium-silicon-
killed steel, deposits in the SEN above the meniscus were composed mainly of
solidified metal and alumina, while below the meniscus, the deposits consisted of
mainly alumina[7]. Poirier et al. similarly reported that the upper portion of the
nozzle (above the meniscus) characteristically contains approximately 80% iron and

15 | P a g e

iron oxide[8]. The lower portion of the nozzle contains approximately 80 to 90%
alumina

Figure 10 Schematic Diagram of Accretion Morphology
.



Figure 11 Schematic diagram of inclusions in a submerged entry nozzle

16 | P a g e




Figure 12 Schematic of a SEN showing inclusion morphology and specimen locations



17 | P a g e


Figure 13 Sketch of the top portion of the SEN showing characteristic structures.

Figure 14 Schematic of the bottom portion of the SEN showing characteristic structure.

18 | P a g e

Steel chemistry has an important effect on inclusion morphology. In particular,
calcium treatment and deoxidation method effect inclusion morphology. For
example, McKague et al. have shown that aluminium-killed grades that are calcium
treated will form calcium aluminates and, if sulphur levels are high enough, calcium
sulphide inclusions[9]. These investigators have also shown that the composition of
the inclusions is determined by the ratio of Al2O3 to CaO when the metal stream
reaches the nozzle. These observations are also true for non-calcium treated steel cast
with nozzles containing lime-bearing materials.
Researchers studying titanium-killed or titanium-added steels have also noted
differences in morphologies of inclusions in the steel and inclusions on nozzles. In
the case of titanium-added, aluminium-killed steels, coarsening of the steel and
alumina nodules that make up the inclusions, have been reported. The individual
particles making up the inclusions and inclusions have a granular shape. The
composition of the inclusions consists predominantly of alumina with titanium oxide
rarely being observed.Inclusions with titanium-killed steels, on the other hand,
typically consist of titanium oxide with approximately 6 to 7% alumina. These
inclusions have an angular shape, in contrast to the round, coral shape typically
reported for aluminium-killed steels.The ladle treatment of steel also seems to affect
nozzle inclusion morphology. Ogibayashi et al. observed that when a heat of steel
was subjected to DH degassing, the inclusion material consisted mostly of metal that
contained clusters of alumina. Those heats that were only subjected to argon bubbling
had inclusion formations that consisted mainly of alumina[10] .The difference in
morphology was reported to be due to the difference in the cleanliness of the steel.
The DH degassed steel is cleaner than steel that only receives argon stirring.
Therefore, its accompanying inclusions contain more solidified steel and less alumina

Figure 15 Typical structure of alumina inclusion. 5000X

19 | P a g e


Figure 16 (A) and (B) RL and CL micrographs of loosely held alumina build-up (bright red CL), (C) and (D) RL
and CL micrographs showing spinel+CA
6
build-up (bright green and intense green CL), and (E) and (F) RL and
CL micrographs of refractory-build-up interface showing relatively wellsintered alumina deposits(red to brown
CL) and minor amounts of green CL CA
6
.



20 | P a g e




Figure 17 CL microstructure of loosely held spinel + CA6 deposits, (B) CL microstructure of dense alumina
layer on the top of refractories

Figure 18 RL and CL microstructures of alumina graphite refractories.

2.2.2 MECHANISMS OF INCLUSION FORMATION
The most prominent theories of nozzle inclusion formation are listed below and are
discussed in more detail in the following sections.
1. The temperature drop that is inherent when steel flows through submerged entry
nozzles causes a resulting decrease in the solubility of oxygen in steel. This drives the
Al-O-Al2O3 equilibrium towards the alumina side, forming alumina inclusions at the
wall of the SEN nozzle[2,11].
2. The flow of steel through the SEN causes low pressure to form inside the nozzle.
This drives the diffusion of atmospheric oxygen through the nozzle wall. Aluminium
in the melt is oxidized to form alumina at the nozzle wall either directly by the

21 | P a g e

oxygen or indirectly from carbon monoxide formed by the reaction of oxygen with
carbon in the refractory[12].
3. Indigenous alumina particles are driven to the nozzle wall due to fluid flow
characteristics of the steel stream. Turbulence or separated flow around disturbances
in the molten steel stream transports alumina to the nozzle where they attach due to a
reduction in the overall surface energy of the system[5,13,14].
4. Thermochemical reactions in the nozzle material transport oxygen either in the
form of sub oxides or carbon monoxide to the nozzle-steel interface where they react
with aluminum in the steel to form alumina at the nozzle wall[6].
2.2.2.1 Thermally Induced Inclusions
Duderstadt et al. observed that in clogged tundish nozzles, steel froze in a directional,
dendritic structure, which indicated that heat was being extracted from the steel
stream by the nozzle refractory material. The inclusion within tundish nozzles was
found to contain alumina suspended within a MnAlSi-oxide matrix. It should be
noted that the steel cast in these experiments was killed with ferrosilicon and
aluminum. Duderstadt et al. concluded that the mechanism of inclusion formation
was the precipitation of aluminum within a MnAlSi-oxide matrix that was further
strengthened by solidifying steel.
Farrell and Hilty further proposed that nozzle inclusion is due to a reduction in
temperature of steel as it passes through a nozzle[11]. It was reasoned that any
deoxidation product that is solid and refractory at steel making temperatures has the
potential to precipitate onto the nozzle wall. The precipitation of solid oxides is due
to the temperature decrease in the steel as it enters the nozzle; even though a nozzle
may be preheated, its temperature is still much lower than steel. This cooling of the
steel causes a decrease in the solubility of the oxide (more likely the solubility of
oxygen) in the steel and causes the oxide to precipitate onto the nozzle wall. At some
point, the rate of precipitation will begin to accelerate as the flow of steel decreases
due to the inclusion formation, which results in a decrease in the flow of steel
through the nozzle and a resulting temperature drop. With a set of casting simulation
experiments, Andersson and Wijk showed that except for the solidification of steel
within the nozzle, temperature is not a significant factor in inclusion formation within
nozzles.[14] The nozzle used in the casting experiments was heated to temperatures
at or above the casting temperature of the molten steel. There was a noticeable
increase in the teeming rate of steel when the temperature of the nozzle was equal to
or up to 50C greater than the steel temperature. The explanation given for the
increased teeming rate was that steel was prevented from solidifying because of the

22 | P a g e

high nozzle temperature. This conclusion was reached because up to 90% of the clog
matrix in the nozzles was composed of steel. It was also observed that the alumina
inclusion within the nozzle was not substantial enough to be the sole cause of nozzle
blockage. As the flow of metal decreases through the nozzle, the rate of heat influx
into the nozzle decreases and the transfer of heat out of the steel becomes large
enough to cause solidification within the nozzle. Therefore, it was concluded that
both alumina accumulation and steel solidification affect nozzle clogging.
Dawson also pointed out that with a temperature drop of 10C in the steel as it passes
through a nozzle, the resultant decrease in solubility of oxygen is less than 1ppm with
Aluminum-killed steels. That is because approximately 99% of the oxygen has
already been consumed in the deoxidation process, leaving only a very small amount
of oxygen to further react.
2.2.2.2 Diffusion of Oxygen through Nozzle Refractory
While the diffusion of oxygen through refractories is commonly mentioned in the
literature as one of the possible mechanisms of nozzle inclusion, no papers
specifically addressing the problem were located. Poirier et al. mentioned that
increased permeability of SEN refractories has been noticed when the refractory
glaze contained defects[8] .They then reported that when the permeability of the
refractory increases, not only is aluminum in the steel oxidized but the steel itself and
then the carbon in the refractory are also oxidized. When the carbon becomes
oxidized, corrosion of the refractory takes place more rapidly because of a breakdown
in the structure of the refractory.
2.2.2.3 Flow Transported Inclusions
Singh proposed that nozzle inclusion formations are produced by deoxidation
products already present in the steel passing through the nozzle[5]. A boundary layer
exists adjacent to the nozzle wall in which the flow of steel is essentially zero. Any
inclusions present in this boundary layer will adhere to the nozzle wall because the
total surface area of the particles in contact with the steel decreases, resulting in an
overall reduction of surface energy. Singh reports that conditions are favourable for a
particle to adhere to the nozzle wall when:

pr

pr
>
pm

pm
+
mr

mr
(1)
= specific energy
= change in surface energy
p = particle

23 | P a g e

r = refractory
m = metal
The work of adhesion of a particle is given by:
W =
pm
(1 Cos h) (2)
Where h = contact angle between the particle and the metal
The contact angle between steel and alumina is reported as being between 134 to 136
degrees. Therefore, there is a significant force acting to adhere a particle of alumina
to another substance with a similar surface energy such as another alumina particle or
a refractory surface. Singh reported that inclusions are transported to the boundary
layer by eddy currents, which throw the inclusions toward the boundary layer where
they attach to inclusions that have already begun to form on the nozzle wall[5]. Once
the inclusions attach themselves together and to the nozzle wall, the particles quickly
begin sintering. Singh calculated that for 2 m diameter alumina particles, a
significant neck forms within one minute[5].
Dawson demonstrated through a series of water modeling experiments that areas of
separated flow occur adjacent to nozzle surfaces whenever the flow of steel is
disrupted[13]. The flow patterns within the separated flow region were demonstrated
to be random. The direction of flow in these regions changes constantly and at times
is directed at the nozzle wall. Areas that were demonstrated to have separated flow
included the inlet of tundish nozzles and the SEN, downstream of tundish slide gates.
These areas of separated flow correspond well with the areas of preferential inclusion
formation observed in industrial nozzles. It was postulated that nozzle clogging
occurred because of the separated flow in these areas.
Dawson went on to test this theory by using a nozzle that was angled at 60 from
horizontal in an a casting simulation. Within an angled nozzle such as this, steel
flows in a streamlined pattern against one side of the nozzle while the other side of
the nozzle has a separated flow region adjacent to the wall. Analysis of the tundish
nozzle revealed that no inclusion formation occurred on the side of streamlined flow
while an inclusion layer formed on the side of the nozzle that experienced separated
flow.
Dawson therefore concluded that separated flow within a nozzle acts to transport
inclusions to the nozzle wall. Once the inclusion is deposited on the nozzle wall, it is
protected from the bulk flow of steel because it lies within the area of separated flow.
It was shown that within the separated flow region, the shear stress adjacent to the
wall is zero. A particle deposited on the nozzle wall in the separated flow region will
not be subject to shear stress and therefore will not be removed from the wall[13].

24 | P a g e

Andersson and Wijk concluded after a set of casting experiments that the rate of
nozzle inclusion formation is dependent on the amount of alumina present in the
steel.[14] In these experiments, the amount of aluminium in the melt was varied. The
total oxygen was assumed to be constant. The equilibrium oxygen in solution in the
steel was then calculated using:
Al
2
O
3
= 2Al + 3O (3)
Logk=64000/T+20.57 (4)

The amount of alumina precipitated in the melt was then calculated from:


It was observed that the rate of clogging was directly proportional to the amount of
alumina precipitated in the melt. It was postulated that a breakdown of the viscous
sub layer below a critical value allowed the transport of alumina particles to the
nozzle wall. However, it was acknowledged that a region of separated flow within the
nozzle might be the rate limiting mechanism of inclusion formation
2.2.2.4 Thermochemically Formed Inclusions
Several researchers have proposed various inclusion formation mechanisms involving
thermochemical reactions between the steel and the nozzle refractory. Some theories
of inclusion formation as a result of these thermochemical reactions are presented
below.
Poirier et al. have proposed a thermochemical mechanism for inclusion formation in
nozzles during the continuous casting of aluminum killed steel[6,8].Their proposal
for the mechanism of inclusion formation involves three steps: the dissolution of
graphite into the steel, the formation of a first layer deposit that consists of alumina
and a vitreous phase formed by the volatilization and oxidation of some of the
refractory species, and finally alumina formation by the oxidation of aluminium in
the melt by carbon monoxide. In the first step, carbon dissolves into the molten steel
from the refractory material.The increased carbon concentration near the nozzle wall

25 | P a g e

increases the activity of aluminium in the melt. This increase in activity of aluminium
causes alumina to form and deposit as a thin film on the nozzle refractory[4,6].
The second step involves a series of volatilization/oxidation reactions.Through a set
of volatilization/condensation experiments it was demonstrated that SiO2, Na2O, and
K2O can be volatilized at steel making temperatures. The mechanism proposed to
volatilize these species is the carbothermal reduction by graphite in the refractory
according to the following reactions:
SiO
2
(s) + C (s) = SiO (g) + CO (g) (6)
Na
2
O (s) + C (s) = Na (g) + CO (g) (7)
K
2
O (s) + C (s) = 2K (g) + CO (g) (8)
A negative pressure caused by the flow of steel through the nozzle drives the gasses
toward the refractory-steel interface.[6,8] At the refractory steel interface, the gaseous
species react with oxygen in the steel, which is reported to have a partial pressure of
10
-11
atmospheres. This produces a low melting temperature liquid phase that
dissolves alumina from the steel or refractory to produce albite according to the
following reactions:
2 Na (g) + 6SiO (g) + Al
2
O
3
+ 7[O]
Fe
=Na
2
O-Al
2
O
3
-6SiO
2
(albite) (9)
2Na (g) + 6SiO (g) + 2[Al]
Fe
+ 10[O]
Fe
=Na
2
O-Al
2
O
3
-6SiO
2
(albite) (10)
Lastly, carbon monoxide oxidizes aluminium in the melt to form Al2O3, which is
precipitated at the nozzle wall.
Sasai and Mizukami performed a set of experiments to investigate thermochemical
reactions as an inclusion formation mechanism[15]. These researchers concluded that
the following reaction occurs when alumina graphite refractories are in contact with
aluminium-killed steels at steelmaking temperatures:
3SiO
2
(s) + 3C (s) + 4[Al]
Fe
=2Al
2
O
3
(s) + 3[Si]
Fe
+ 3C (11)
As an intermediate step to the above reaction, silica and alumina are reduced to sub
oxide gasses by carbon in the refractory. The Al2O gas reacts with additional silica
and reforms solid alumina. This last assumption was drawn after calculating the
equilibrium partial pressures of the sub oxide gasses versus temperature. SiO is the
most stable sub oxide gas at steelmaking temperatures.
Fukuda et al. proposed an inclusion formation mechanism whereby alumina in the
refractory is carbothermally reduced to a sub oxide gas[16].This gas then diffuses to
the refractory-steel interface where it is oxidized to alumina as a precipitate on the
nozzle wall. This mechanism was proposed after observing alumina inclusion
formation on an alumina graphite specimen submersed in iron in the absence of
aluminum. This mechanism occurs by the following reactions:
SiO
2
+ C = SiO (g) + CO (g) (12)

26 | P a g e

Al
2
O
3
+ 2 C =Al
2
O (g) + 2 CO (g) (13)
The following reactions were proposed to occur when the gasses are transported to
the refractory-steel interface.
SiO (g) =[Si]
Fe
+ [O]
Fe
(14)
Al
2
O (g) =2 [Al]
Fe
+ [O]
Fe
(15)
3 Al
2
O (g) =4 [Al]
Fe
+ Al
2
O
3
(16)
The dissolved aluminum metal from the above reaction then reacts with oxygen in the
steel to form alumina at the nozzle wall.
2.2.2.5 Discussion of Inclusion Mechanisms
It is possible that all of these mechanisms play a role in nozzle inclusions. However,
theories 1 (thermally induced inclusions) and 2 (diffusion of oxygen through
refractory) are not likely to play a significant role in nozzle clogging[5] .Severe
clogging is often present in areas of the nozzle that are submerged in molten steel
such as the discharge area of an SEN. This area of the nozzle is not likely to have a
substantial temperature drop, and diffusion of air through the refractory is impossible
because it is surrounded by molten steel. One area where these conditions exist is the
body of the SEN above the meniscus, and in that area significant clogging rarely
exists and theory 4 (thermochemically formed inclusions) does not occur based on his
observations of clogged nozzles in pilot-scale and bench-scale tests. However, it
should be noted that he used zirconia and alumina nozzles for his experiments[5].
There is no indication as to whether or not the nozzles contained graphite and silica,
which Poirier et al.,have indicated facilitate nozzle reactions. One likely theory
presented by Ogibayashi indicates that the dense layer of alumina that is formed on
the refractory surface is the result of thermochemical nozzle refractory reactions[3].
He added that the second, less-dense layer is the result of turbulent flow transporting
the inclusions to the nozzle wall. The particles then adhere to the nozzle wall due to a
reduction in interfacial energy of the system. Finally, the alumina inclusion sinters to
the dense alumina layer.
2.2.3 COUNTERMEASURES.
The most obvious means to reduce clogging is to decrease the concentration of
deoxidation products and the formation of reoxidation products . Means to achieve
this increase in steel cleanliness have been reviewed by Byrne et al. and Szekeres .
The important aspects of clean steelmaking include:

27 | P a g e

Ladle Refining Practice. A vacuum degassing treatment yields better cleanliness
than does argon bubbling.
Reoxidation Prevention. Submerged ladle-to-tundish pouring, shielded tundish
surface, and leak-tight refractory joints will reduce exposure of the steel to oxygen
and thereby improve cleanliness
Deoxidation Product Removal. Optimal tundish flow patterns as well as filtration
and electromagnetic techniques can remove deoxidation products from the melt.
Flux Entrainment Prevention. Submerged ladle-to-tundish pouring and avoidance of
ladle slag carryover will reduce the amount of exogenous inclusions in the melt.

Several other countermeasures have been proposed to eliminate nozzle clogging such
as oversized nozzles, argon injection, calcium treatment, and lime additions to nozzle
refractories. These countermeasures are discussed individually in the following
sections.
2.2.3.1 Oversized Nozzles
Duderstadt et al. proposed one of the first strategies to counter nozzle clogging in the
continuous casting of billet.[2] The proposed strategy was the use of an oversized
nozzle along with a stopper rod mechanism to control the flow of steel through the
tundish nozzle. The oversized nozzle provides longer casting times before inclusion
formation becomes significant enough to slow the casting rate. Duderstadt et al.
found that the use of a stopper rod disturbed the flow of steel from the tundish to the
mold. Therefore, to make the oversized nozzle/stopper rod system functional, a
shrouding system was necessary to prevent reoxidation of the steel due to the
inherently turbulent stream.

2.2.3.2 Argon Injection
Argon injection is used extensively in the steel industry. Depending on each
companys operating practice, argon may be injected into the tundish well nozzle, top
plate, tube holder, and submerged entry nozzle. Argon is bubbled either through
pores in the refractory itself or through pierced holes in the refractory. The injected
argon forms a protective gas layer that prevents the adherence or precipitation of
inclusions and provides turbulent flow to dislodge inclusions from the refractory
wall. It should be noted that improvements to ladle nozzle argon systems have been
made that are similar to tundish well nozzle and SEN improvements. The primary

28 | P a g e

focus of this research is on tundish to mold systems, therefore improvements to ladle
nozzles are not specifically considered.
Several researchers have made improvements to argon injection that have resulted in
a reduction of nozzle clogging. Many of the improvements in argon injection have
focused on preventing cracking of refractory materials, gas leakage, and obtaining
better argon bubble distribution[17-22].
Cameron and Alavanja et al. have reported cracking of the tundish well nozzle
refractory, which leads to non-uniform argon distribution[17]. When refractories
crack, argon is bubbled from one or several concentrated areas rather than in a fine,
evenly dispersed pattern. Steel can also infiltrate the cracks and fill the argon delivery
system. Under these conditions, clogging occurs quickly. Nozzle cracking problems
have been reduced by carefully selecting refractory materials and effectively
preheating nozzles.
Tundish well nozzle pore size and distribution has been shown to be a very important
factor in the effectiveness of argon injection. Alavanja et al. have reported that
variability in the size of pierced holes in nozzles affects clogging[18]. It was
established that when pierced holes are too large, steel can penetrate into the holes,
reducing the effectiveness of argon injection. Schmidt, et al. found that using two
argon injection ports and refractory with smaller and more evenly distributed pores
resulted in significant decreases in the occurrence of tundish well nozzle
clogging[21].
Likewise, Takasugi, et al., found that the pore size and distribution of porous nozzles
could be better controlled by incorporating spherical grains in the refractory
material[9]. Changes in argon back pressure using these nozzles were very small
from the first to the last heats cast through the nozzles, indicating better clogging
prevention performance. This was confirmed by measuring deposits on spent nozzles.
Through water modeling experiments, it was determined that the optimum pore size
for low throughput casting (2 tons/min.) is 40 m from the standpoint of slab quality
and nozzle clogging. Larger gas bubbles are produced with this pore size, which have
less opportunity to become entrapped in the slab and cause pinholes
Leaks from the steel canisters which encase tundish well nozzles have been reported
to decrease the effectiveness of argon injection. Gas leakage from the steel canisters
has been corrected by designing a better seal between the canister and the nozzle
refractory material.
Dofasco has designed tundish well nozzles to form inclusions at specific locations of
the tundish well nozzle. Argon is injected in these areas of highest deposition. This
strategy has worked well, with a 26% reduction in the thickness of deposits for the

29 | P a g e

same length of casting time. Dofasco has also experimented with argon-purged
stopper rods. However, these were found to be ineffective, and when the argon
purging system in the tundish well nozzle performed properly, there was no need for
the added argon shielding.
Ando, et al. have found that the performance of argon-injected submerged entry
nozzles can be improved by reducing the amount of SiO
2
in the refractory. It was
reported that SiO
2
in the refractory is reduced by carbon to form SiO gas. This
changes the size and distribution of pores in the refractory, which leads to an increase
in argon back pressure over time and increased clogging. By reducing the amount of
SiO
2
in the refractory, the reduction in argon back pressure during operation is
reduced, therefore giving greater resistance to clogging.
2.2.3.3 Calcium Treatment
Calcium treatment is a common method utilized to prevent nozzle clogging. Calcium
is added to steel melts to react with the alumina present in the melt so that liquid
calcium aluminate compounds can be formed. The ratio of CaO/Al
2
O
3
must be such
that the calcium aluminate is liquid at steelmaking temperatures. This is achieved by
approaching the composition of 12CaO.7Al
2
O
3
(see Figure 4). While casting, the
liquid calcium aluminates pass through the nozzles and do not attach to the refractory
wall, preventing inclusion formation[23].
Several problems can occur when calcium treating steels. To be effective, the total
oxygen of the steel should be known in order to make the correct amount of calcium
additions. Faulring et al. have shown that with small calcium additions, the
occurrence of clogging increases up to a threshold amount[24] .McKague et al. have
also demonstrated that too much calcium can also cause nozzle clogging. In either
case, the calcium aluminates that form are solid at steel making temperatures and
promote clogging.
If sulfur levels in the steel are too high, solid CaS particles will form with calcium
treatment, which will promote clogging. Bannenberg has shown that there is a small
compositional window of calcium additions that will produce liquid inclusions at
steel making temperatures.This window is related to the total oxygen and sulfur
content of the steel. For constant total oxygen content and increasing sulphur
contents, the amount of calcium needed to precipitate calcium sulfide particles
decreases. Therefore, to be successful, the sulfur content of the steel should be low
before calcium treatment. McKague et al. report that calcium treatment is not
attempted with steels containing greater than 0.012% sulfur to avoid the formation of
CaS.

30 | P a g e

Another problem that can occur with calcium treatment is the reoxidation of steel. If
steel reoxidizes before it is cast, the excess alumina formed can change the
composition of calcium aluminates to be alumina rich, which are solid at steel
making temperatures. McKague et al. reported problems with reoxidation of steel
when filling the tundish.After changes were made to use less aluminum at the ladle
metallurgy station and to reduce reoxidation during tundish fill, the practice of adding
calcium wire to the tundish during start-ups was employed. This practice has reduced
the occurrence of clogging, but has not eliminated it
2.2.3.4 Lime Bearing Nozzle Materials
Lime additions to nozzle refractory materials have also been used to prevent nozzle
plugging. Typically, the refractory material consists of lime stabilized zirconia. Lime
in the nozzle reacts to form low melting temperature calcium aluminates. As reported
by Ogibayashi et al., the liquid calcium aluminates then provide a low energy
interface so that subsequent attachment of alumina particles is avoided.
Ogibayashi et al. have also reported that nozzle replacements due to clogging have
dropped to 14% of the previous frequency with lime bearing nozzles as compared to
alumina graphite nozzles. A decrease of approximately 50% in the number of
inclusions greater than 30 m in the slab, as measured under the microscope, has also
been reported.
It should be noted that Benson et al. have made the observation that good results with
lime bearing refractories are typically observed in steel plants that have only marginal
clogging problems while those that have major problems with clogging do not find
any benefit. In the case of plants that have significant clogging problems, the
transport of lime to the steel refractory interface is considered to be the limiting step
because when analysing spent lime bearing refractories, lime is still available in the
refractory while the alumina build-up on the nozzle surface has not reacted to form
calcium aluminates. Tsujino and colleagues also reported that in laboratory trials,
dirty steel formed greater inclusions on zirconia-lime-graphite refractories than
standard alumina graphite.
Attempts have been made by researchers to provide greater amounts of lime at the
refractory surface to react with alumina in order to avoid these problems. In 1990,
TYK reported the development of a nozzle material with SiO
2
additions that were
intended to accelerate the reaction of lime with alumina. Then in 1991, TYK reported
the development of a nozzle containing additional non-hydrating CaO compounds
that increase the transport of CaO to the steel-refractory interface, thereby improving
the clogging resistance of these nozzles.

31 | P a g e

Likewise, in 1993, Vesuvius reported that a nozzle material was developed that
contains a destabilizer which increases the transport of lime to the nozzle wall,
contributes to the formation of low melting point substances, and provides for
controlled erosion of the nozzle wall so that new refractory material becomes
exposed to steel over time.
Despite the success that some researchers have reported with lime bearing nozzles,
several problems that have also been reported. Problems similar to those experienced
with calcium treatment are also often encountered with lime bearing nozzles. For
example, Ramacciotti and Marino reported that clogging occurs with CaS inclusions
with lime bearing nozzles if the sulfur level of the steel is above 0.020%.
Another problem that is mentioned with lime bearing refractories is the hydration of
lime in the nozzle material. Pure lime cannot be added to refractory materials because
of this hydration. Instead, the lime is added in the form of lime stabilized
zirconia.However, even in stabilized materials, hydration has been observed. In tests
performed by Oguri et al., nozzles that were thought to contain non-hydrating lime
also showed weight increases over time.The weight gain was thought to be caused by
the presence of hydrating lime in the refractory material.
If the concentration of CaO in lime bearing nozzles is too high, erosion can become a
problem. Oguri et al. reported such problems with nozzles containing 29% non-
hydrating CaO compound.Aoki and colleagues reported similar results with
increasing erosion from 20% CaO to complete erosion with 30% CaO additions.
2.2.3.5 Other Nozzle Materials
Boron nitride and zirconia boride have been investigated by Ishii et al. for potential
use as nozzle materials. In laboratory simulations, it was found that BN and ZrB
2

materialsexhibited excellent resistance to alumina deposition. Industrial trials with
nozzles composed of BN based materials showed no alumina adhesion. The clogging
resistance performance of the C-23 BN and ZrB
2
materials is reported to be due to
their non-wetting characteristics with steel. In addition, it is reported that ZrB
2

materials react with oxygen in the steel to form B
2
O
3
gas that forms a protective layer
to prevent alumina deposition. Vesuvius has patented SiAlON, a material that is
reported to have good resistance to alumina adhesion. SiAlON forms a vitreous layer
at the surface of the nozzle which provides for controlled erosion of the nozzle
refractory. This controlled erosion prevents the nucleation and adherence of alumina
to the nozzle wall. The authors of the patent indicate that the material exhibits
superior erosion resistance, similar to alumina graphite. However, Benson et al.
reported that SiAlON exhibits uncertain erosion characteristics.

32 | P a g e

2.2.3.6 Nozzle Modifications
In an effort to reduce the effect of clogging, oversized nozzle bores and replaceable
submerged entry nozzles are widely employed. To reduce the degree of clogging, the
following have been investigated:
Improved joint sealing. Strengthening the steelwork that holds the nozzle in place
was found to reduce air aspiration and thereby reduce clogging
Rounded nozzle entrance. Incorporating a rounded entrance (in lieu of a sharp
corner) to the tundish nozzle and ensuring proper vertical alignment can reduce
clogging at the nozzle entrance by eliminating separated flow
Internal step. A five millimeter annular step incorporated at the mid-height of the
submerged entry nozzle has been found to decrease alumina buildup in the lower part
of the nozzle as well as decreasing flow impingement on the mold wide face
Varying nozzle internal diameter. Increasing the nozzle internal diameter just below
the stopper rod seating surface has reduced clogging
Flat bottomed nozzle. Decreased port clogging was observed when the elevation of
the nozzle internal bottom and port bottom were coincident (i.e., no nozzle well)
Insulation around nozzle. Insulation, as well as preheat and heating, around the
clogging location may reduce clogging
Nozzle design modifications that have been proposed which include annular
step nozzles, an electrochemical cell nozzle, and a bubble curtain nozzle

Figure 19 Lime-alumina phase diagram

33 | P a g e


Figure 20 Compositional window in which inclusions are liquid

Figure 21 Annular Nozzle

Annular Step Nozzles
Flogates Limited has a patent for an annular step nozzle as shown in Figure 6.The
annular nozzle uses a series of steps to produce eddies inside the nozzle. These eddies
are used to preferentially deposit non-metallic inclusions within the step, where they
will not hinder the flow of steel through the nozzle. Using the same type of nozzle,
Tsukamoto et al. have proposed that nozzle clogging can be reduced by preventing
flow deviation downstream of the slide gate. Using water modelling, it was shown
that the flow of steel downstream of the slide gate is severely asymmetric. A
separated flow pattern is generated in a standard SEN downstream from the slide
gate, on the closing side of the gate. The purpose of the annular steps in this case is to
redistribute the flow of steel so that a uniform stream is achieved within the nozzle.
In addition, the vortexes created around the annular steps uniformly redistribute

34 | P a g e

argon gas bubbles within the steel stream. It was shown that in a traditional SEN,
there is a significant difference in the flow rate of steel discharged from one port of
the SEN to the other. The flow of steel leaving the SEN ports was reported as being
equalized with the annular step nozzle. Use of a double annular step nozzle in an
actual caster has given better performance against clogging as well as a 50% decrease
in pinhole defects in the slab.



Electrochemical Cell Nozzle
Electro-Nite Co. holds a U.S. patent for an electrochemical cell that prevents the
deposition of alumina on the nozzle wall (see Figure 7). A zirconia jet or liner is
used with this nozzle. An electric current is passed through the molten metal stream
via a graphite electrode submersed in the molten metal bath and another electrode in
contact with the zirconia liner. No details are given in the patent as to the actual
mechanism employed to prevent the deposition of alumina on the nozzle wall.



Figure 22 Electrochemical Nozzle
Bubble Curtain Nozzle
Yokoya et al. have recently reported on a submerged entry nozzle that employs a
bubble curtain to remove inclusions from the steel stream and prevent their
deposition on the nozzle wall. A vortex is created in the molten steel stream with

35 | P a g e

swirl blades as it enters the SEN. Argon is injected just downstream of the entrance
to the SEN through several holes in the nozzle wall. As the bubbles travel with the
steel stream through the nozzle, the bubbles migrate to the center of the stream,
creating a conical shaped bubble curtain as shown in Figure 8. Inclusions in the
steel stream that might otherwise adhere to the nozzle wall are attached to the
bubbles and transported to the center of the steel stream.
Water modeling and computational fluid dynamic modeling were used to
demonstrate the effectiveness of this nozzle. No actual casting results are reported.


SEN Changing Systems
While the changing of spent submerged entry nozzles during casting does not
actually improve nozzle clogging, productivity improvements can be achieved by
using a quick change system as proposed by Nishio et al.With such a system, the
tundish does not have to be raised and lowered to change the nozzle, and the time it
takes to change a nozzle is significantly reduced. This has also been reported to
eliminate scrap generated during conventional SEN changes.



Figure 23 Schematic diagram of bubble curtain

36 | P a g e

CHAPTER 3
3.0 Experimental Methods and Data Analysis
3.1 Materials Studied
The scope of the present work was basically to study the different parameters
involved directly or indirectly with the production route of SC 104 grade. The
different parameters studied under different category based on literature and
experiences are
RH Parameters
RH in Temp
RH in O
2

Al added for heating and the no. of times
Al for deoxidation
Total Al addition
Immersion Depth
Treatment time after Al addition & FeTi addition
Treatment time between Al addition & FeTi addition
O
2
blow( volume, duration & No. of times)
Min Vacuum
Decarburization Time
Treatment stop time
RH out temperature
Caster Parameters
Average stopper position
Start stopper position
Last stopper position
Average casting speed
Maximum casting speed
Minimum casting speed
Tundish weight and ladle open time





37 | P a g e

Chemistry Parameters
RH out( C,N,Ti,Si,Al)
Tundish(C,N,Ti,Si,Al)
Al/Ti ratio in tundish
C,N pickup & Ti,Al,Si fading
Slag Parameters
Slag Fe
Slag CaO
Slag Al
2
O
3

C/A ratio
OLP Parameters
OLP addition(lime)
OLP out temperature
OLP out O
2

Vessel Parameters
Reblow amount
Quality Parameters
Slabs obtained
Slabs achieved
A total of 13 heat sequences from RH degasser were studied and a SEN sample used
in a sequence was collected and sent to the R&D department of TATA STEEL for the
evaluation and validation of the morphology and composition of clog material. Data
analysis and correlation between the different parameters were established using
STATISTICA and MICROSOFT EXCEL software.
3.2 Morphology Characterisation
The methodology adopted in the present investigation included:
(a) Collection of inclusion deposit from the submerged entry nozzle (SEN) and
corresponding liquid steel and slag samples from ladle and tundish,
(b) Characterization of inclusions in liquid steel samples using SEM-EDX,
(c) Identification of the sources of clogging agents.
. Physical Characteristics of SEN deposits were visually examined. For identification
of various clogging agents in the liquid steel, samples were collected from different
layers of clog and examined in the scanning electron microscope (SEM) and energy
dispersive spectroscopy (EDX) for their influence on nozzle clogging.

38 | P a g e

CHAPTER 4
4.0 RESULTS AND DISCUSSION
4.1 Visual Observation
Majority of the deposits were observed at the SEN bottom, typically in the form of a
conical pile of 80100mm height. This was observed systematically in all SENs,
though the quantity and dimension varied. The extent of deposition was much less
inside the barrel of the SEN compared to the port area. The total volume of deposits
observed in the case of Ti-free grades was relatively less than in the case of Ti-
stabilized grades. The inclusion deposits appeared whitish and were radially
symmetric inside the SEN.


Figure 24 Clogged Nozzle sample taken



39 | P a g e

Table 1 Observations of SEN Deposits
Volume of deposits More in volume
Shape Conical
Density Heavier
Colour Greyish white
Friability Tough
Presence of entrapped steel Large amount of frozen steel throughout
the deposit
Principal clog constituent Coarse Al
2
O
3
& spinel inclusions
Presence of spinel inclusions Up to 15-20% vol spinel present along
with alumina clusters
Presence of Ti bearing inclusions Fine Ti bearing inclusions
Al
2
O
3
inclusions size Mixed size distribution of alumina grains

Samples collected from different layers of SEN deposits were characterised using
SEM-EDX. The nature of clog was found to be highly heterogeneous in all the
samples studied. However, the bulk of the deposits consisted of coral shaped clusters
of pure alumina (Al
2
O
3
). Minor constituents included spinel (MgOAl
2
O
3
) inclusions
and occasional presence of spherical droplets of frozen steel. The area percent of
spinel in the inclusion deposits determined from the image analyser was found to
vary from 5 to 10%. In addition to these constituents, the deposits in the case of Ti-
stabilized ULC steel were observed to contain Ti-bearing alumina (TiO
x
Al
2
O
3
)
inclusions and large amounts of frozen steel throughout the deposits. A typical SEM
micrograph of alumina cluster along with its X-ray elemental map is shown . It can
be seen clearly from Fig. 5 that entrapment of steel in the nozzle deposits occurred in
immediate vicinity of TiO
x
Al
2
O
3
deposits, in the case of Ti-bearing steel grades.

40 | P a g e


Figure 25 SEM micrograph of constituent elements in alumina cluster



Figure 26 SEM microgarph of constituent elemnts in inclusions and frozen metal

41 | P a g e


4.2 Data Analysis
Data for all the parameters were collected for the month of June and on finding the
regression coefficient was found taking sum of abnormalities per heat as the
dependent variable and all other as the independent variables.
It was found that that variables such as
Al/Ti ratio in tundish
Time gap between Al and Ti addition
Time between RH out and cast start
C/A ratio
Reblow Time
Slag Fe

Table 2 Correlation Table of the parameters considered
All these variables are correlating with the sum of abnormalities in slabs with a
confidence limit of more than 90%.
Al/Ti ratio in tundish among all these variables has emerged as prominent variable
with a greater extent of effect on the sum of abnormalities in slabs. Further analysis
of Al/Ti ratio shows that when the Al/Ti ratio was maintained at 4, such that Al
2
O
3
is
the only thermodynamically stable oxide, the results show that transient titanium

42 | P a g e

containing oxides exist temporarily after titanium addition, but with time, the
predominant inclusion was Al
2
O
3
, which would generate little shape change
and produce transient stage inclusions with less titanium contents. When the Al/Ti
ratio was kept at 1/1 at the Al
2
O
3
stable region, it was revealed that transient stage
inclusions, which are titanium containing in nature, were profusely formed after
titanium addition. It was also demonstrated that such a chemistry change was
always accompanied by a corresponding morphology evolution from a spherical-
dominant scenario to an irregular-prevailing one. When the Ti/Al ratio in the melt
was such that Al
2
TiO
5
and Ti
3
O
5
are the thermodynamically predicted stable phases,
the inclusions, which were Al
2
O
3
after aluminum killing, evolved after titanium
addition toward TiO
x
inclusions, accompanied by a shape change from spherical to
irregular.
Lower Al/Ti or higher Ti/Al ratio results in the decrease in the size of the inclusions
and the number of inclusions making it difficult for the removal of inclusions.

Figure 27 Variation of inclusion size as function of Ti/Al ratio

43 | P a g e


Figure 28 Influence of Ti/Al ratio on microinclusions
Moreover it was also found that the Al/Ti ratio was dependent on the %Al in the
tundish (Ti remaining unchanged).

Figure 29 Relation between Al/Ti ratio and tundish Ti


44 | P a g e


Figure 30 Relation between Al/Ti ratio and tundish Al

Secondly the time between Al and FeTi addition should be higher to make sure low
total oxygen in the melt and prevent formation of titania-alumina inclusions. This
also makes sure that the inclusions get time to be entrapped by the slag.

Figure 31 Histogram of time gap between Al & Ti addition

45 | P a g e

The time gap between RH out and cast start should be increased for better inclusion
flotation and hence cleaner steel.

Figure 32 Scatter plot between sum of abnormalities and time gap between RH out and cast start

Figure 33 Histogram of RH out and cast start time gap
Higher C/A ratio and lower slag Fe content decrease the sum of abnormalities in
slabs. Hence a higher C/A ratio of around 1 has to be maintained and slag Fe content
has to be decreased from the present high levels.
R = 0.0425
0
2
4
6
8
10
12
14
16
18
9 14 19 24 29
s
u
m

o
f

a
b
n
o
r
m
a
l
i
t
i
e
s

difference of time between RH out and cast start

46 | P a g e


Figure 34 Histogram of C/A ratio and Slag Fe




Figure 35 Relation between C/A ratio and Slag CaO
R = 0.6646
0
5
10
15
20
25
30
35
40
45
0 0.2 0.4 0.6 0.8 1 1.2
S
l
a
g

C
a
O

C/A ratio

47 | P a g e


Figure 36 Relation between C/A ratio and slag Al
2
O
3

The above two figures show that to increase C/A ratio the slag CaO content has to be
increased and the slag Al
2
O
3
content has to be decreased this indirectly reduces the
alumina content in the inclusions and hence less clogging occurs
At higher temperature inclusions tend to melt, so to attain a higher temperature in RH
aluminium is used in chemical heating by blowing oxygen. This introduces both
aluminium (principal constituent of inclusions) and oxygen responsible for oxide
formation into the system. This reduces the cleanliness of steel.
The figures below show the heats where aluminium has been used for chemical
heating. This was due to the lower temperature supplied by the OLP station.
Aluminium addition for heating increases with lower RH in temperature.
R = 0.6523
0
10
20
30
40
50
60
0 0.2 0.4 0.6 0.8 1 1.2
S
l
a
g

A
l
2
O
3

C/A ratio

48 | P a g e



Figure 37 Scatter plot of Al used for heating against RH in temp and its dependency on OLP out temp



Figure 38 Scatter plot of O
2
blow against RH in O
2
with dependency on OLP out oxygen

The figures above show that for chemical heating oxygen is required and when RH in
oxygen is low more oxygen has to be blown. This introduces oxygen into the system.


49 | P a g e


CHAPTER 5
5.0 CONCLUSION AND RECOMMENDATION
Various parameters likely to be involved in the production of interstitial free grades
were studied and it was found that a number of them are inter-related with varying
degree of influences on each other.
The following recommendations were made based on different treatment stations
OLP
o Higher OLP out temperature
o Addition of FeAl chips at OLP
RH
o Higher RH in temperature to be maintained.
o Higher RH in O
2
to be maintained.
o Treatment time between Al and Ti addition should be increased
o Treatment time after Ti addition should be increased.
o Aim higher aluminum at RH to get higher Al/Ti ratio in tundish
SLAG
o Slag C/A ratio to be maintained at around 1
o Lower slag Fe content
OTHERS
o Higher Al/Ti ratio to be maintained.
o Time from RH-Out to cast start to be maximized

5.1 SCOPE OF FURTHER WORK
Further data should be done to find better correlation between the parameters.Effect
of SEN preheating time can be taken into account in further projects.Argon flow in
shroud and Argon flow in stopper can be studied upon.



50 | P a g e

REFERENCES
1. Snow, R.B.; Shea, J.A. Mechanism of Erosion of Nozzles in Open-Hearth Ladles.
The Journal of the American Ceramic Society. Vol. 32 No. 6. pp. 187-194 (1949)
2. Duderstadt, G.C.; Iyengar, R.K.; Matesa, J.M. Tundish Nozzle Blockage in
Continuous Casting. Journal of Metals. Vol. 20 No.4. pp. 89-94 (1968)
3. Ogibayashi, S. Mechanism and Countermeasure of Alumina Buildup on
Submerged Nozzle in Continuous Casting. Taikabutsu Overseas. Vol. 15, No. 1. pp. 3-
14 (1994)
4. Hiraga, Y.; Yashima, Y.; Fujii, K. Behavior of Alumina-Deposition on Tundish
Nozzles in Continuous Casting of Ti-Added Al Killed Steel. Taikabutsu Overseas.
Vol. 15, No. 1. pp. 22-27 (1995)
5. Singh, S.N. Mechanism of Alumina Buildup in Tundish Nozzles During
Continuous Casting of Aluminum-Killed Steels. Metallurgical Transactions. Vol. 5.
pp. 2165-2178 (1974)
6. Poirier, J.; Thillou, B.; Guiban, M.A.; Provost, G. Mechanisms and
Countermeasures of Alumina Clogging in Submerged Nozzles. Steelmaking
Conference Proceedings. Vol. 78. pp. 451-456 (1995)
7. Kasai, N.; Kawasaki, M.; Hayashi, Y.; Kawai, H. Explanation of Clogging
Phenomena of CC Submerged Nozzle. Taikabutsu Overseas. Vol. 11, No. 1. pp. 22-33
(1990)
8. Poirier, J.; Verrelle, D.; Thillou, B.; Provost, G.; Taffin, C.; Tssot, P. Study of
Clogging Phenomena in Continuous Casting Submerged Nozzles. UNITECR 1991. pp.
226-229 (1991)
9. McKague, A.R.; Engel, R.; Suer, M.A.; Wolf, D.J.; Garbowsky, M.D.; Reed, L.F.
Resolution of Clogging at Armcos Mansfield Slab Caster. Iron and Steelmaker. Vol.
25, No. 11. pp. 35-41 (1998)
10. Ogibayashi, S.; Uchimura, M.; Maruki, Y.; Mizukoshi, D; Tanizawa, K.
Mechanism and Countermeasure of Alumina Buildup on Submerged Nozzle in
Continuous Casting. Steelmaking Conference Proceedings. pp. 337-344 (1992)
11. Farrell, J.W.; Hilty, D.C. Steel Flow Through Nozzles: Influence of Deoxidizers.
Electric Furnace Proceedings. Vol. 29. pp. 31-45. (1971)
12. Benson, P.M.; Robinson, Q.K.; Dumazeau, C. New Technique for the Prevention
of Alumina Build-Up in Submerged Entry Nozzles for Continuous Casting.
UNITECR 1993. pp. 1087-1096 (1993)
13. Dawson, S. Tundish Nozzle Blockage During the Continuous Casting of
Aluminum-Killed Steel. Steelmaking Conference Proceedings. pp. 15-31 (1990)
14. Andersson, M.; Wijk, O. A study on Tundish Nozzle Blockage During Casting of
Aluminum Deoxidized Steel. Sixth International Conference on Refining Process.
pp. 175-209 (1992) C-33
15. Sasai, K.; Mizukami, Y. Reaction Mechanism Between Alumina Graphite
Immersion Nozzle and Low Carbon Steel. ISIJ International. pp. 802-809 (1994)
16. Fukuda, Y.; Ueshima, Y.; Mizoguchi, S. Mechanism of Alumina Deposition on
Alumina Graphite Immersion Nozzle in Continuous Caster. ISIJ International. pp.
164-168 (1992)

51 | P a g e

17. Cameron, S.R. The Reduction of Tundish Nozzle Clogging During Continuous
Casting at Dofasco. Steelmaking Conference Proceedings. Vol. 75. pp. 327-332 (1992)
18. Alavanja, M.; Gass, R.T.; Kittridge, R.W.; Tsai, H.T. Continuous Improvement of
Practices to Reduce Tundish Nozzle Clogging. Steelmaking Conference Proceedings.
Vol. 78. pp. 415-426 (1995)
19. Takasugi, H.; Masaoka, T.; Shirayama, A.; Mori, T.; Hiroshi, M. Prevention of
Alumina Build-Up in Submerged Entry Nozzles for Continuous Casting of Low
Carbon Aluminum- Killed Steel. The International Iron and Steel Congress. Vol. 3.
pp. 462-469 (1990)
20. Ando, M.; Takahashi, S.; Okumura, H.; Yamaguchi, K. Slit Type Immersion
Nozzle of Alumina-Graphite Material. UNITECR 1995. Vol. 2. pp. 73-78 (1995)
21. Schmidt, M.; Russo, T.J.; Bederka, D.J. Steel Shrouding and Tundish Flow
Control to Improve Cleanliness and Reduce Plugging. Steelmaking Conference
Proceedings. pp. 451-460 (1990)
22. Kobayashi, S.; Ishige, T.; Ogata, M. Prevention for Nozzle Clogging of SLCC at
NKK Keihin Works. Taikabutsu Overseas. Vol. 15, No.1. pp. 33-37 (1994)
23. Fruehan, R.J. The Making, Shaping and Treating of Steel. 11th Ed. The AISE
Steel Foundation. Pittsburgh, PA. pp. 687-693 (1998)
24. Faulring, G.M.; Farrell, J.W.; Hilty, D.C. Steel Flow Through Nozzles: Influence
of Calcium. Continuous Casting, Vol. 1, Chemical and Physical Interactions During
Transfer Operations. Iron and Steel Society. Warrendale, PA. pp. 57-66. (1985)
25. Bannenberg, N. Inclusion Modification to Prevent Nozzle Clogging. Steelmaking
Conference Proceedings. Vol. 78. pp. 456-463 (1995)
26. Benson, P.M.; Robinson, Q.K.; Park, H.K. Evaluation of Lime-Containing Sub-
Entry Shroud Liners to Prevent Alumina Clogging. Steelmaking Conference
Proceedings. pp. 533-539 (1993)
27. Tsujino, R.; Tanaka, A.; Imamura, A.; Takahashi, D.; Mizoguchi, S. Mechanism
of Deposition of Inclusion and Metal in ZrO2-CaO-C Immersion Nozzle of
Continuous Casting. ISIJ International. pp. 853-858 (1994)
28. Nakamura, T.; Aoki, T.; Okumura, H.; Kondo, Y. Immersion Nozzle for
Prevention of Alumina Clogging. Taikabutsu Overseas, Vol. 11, No. 1. pp. 39-39
(1990)
29. Oguri, K.; Ando, M.; Muroi, T.; Aoki, T.; Okumura, H. Alumina Clogging
Resistant Materials for Immersion Nozzle. UNITECR 1993. pp. 1119-1127 (1993)
C-34
30. Aoki, T.; Nakamura, T.; Ozeki, H.; Elksnitis, A. Alumina Clogging Resistant
Materials for Tundish Shrouds. Steelmaking Conference Proceedings. Vol. 78. pp.
357-360 (1991)
31. Investigating the influence of Ti and P on the clogging of ULC steels in the
continuous casting process C. Bernhard, G. Xia, A. Karasangabo *, M. Egger, and A.
Pissenberger
32. Nozzle Clogging Behaviour of Ti-bearing Al-killed Ultra Low Carbon Steel(ISIJ
International Vol 44[2004])