/smash/get/diva2:517094/FULLTEXT01.pdf

/smash/get/diva2:517094/FULLTEXT01.pdf
Grain Refinement of High Alloyed Steel
With Cerium Addition
Eivind Strand Dahle
Materials Technology
Submission date: July 2011
Supervisor:
Øystein Grong, IMTE
Norwegian University of Science and Technology
Department of Materials Science and Engineering
2
Preface
The following work has been carried out at the Department of Materials
Technology, Norwegian University of Science and Technology (NTNU)
from February to July 2011. The work has been in collaboration between
NTNU, Elkem, Sintef and Scana Steel Stavanger.
Professor Øystein Grong have been the supervisor, along with Dr.ing.
Casper van der Eijk (SINTEF), Dr.ing. Ole Svein Klevan(Elkem) and Dr.
Fredrik Haakonsen (SINTEF) as co-supervisors.
I would also like to thank Yingda Yu for help with the SEM examination,
and Morten Raanes for executing the EPMA examination.
Trondheim
July 2011
Eivind Strand Dahle
3
4
Abstract
This master thesis has the objective to improve the mechanical properties
of Super Duplex steel by adding Elkem Grain Refiner (EGR). EGR is
commercial grain refiner where the active element is cerium. Cerium is a
strong oxide and sulphide former whose inclusions are to act as sites for
heterogeneous nucleation during casting. The cerium inclusions will form
at low undercooling making it possible to grow equiaxed grains ahead of
the solidifying front, reducing the size of the columnar zone normally
seen in an ingot. By reducing the columnar zone the steel will have a
larger zone of equiaxed, and smaller, grains improving the mechanical
properties of the steel and reducing the segregation throughout the ingot.
The experiment was performed with S4501 Super Duplex steel provided
by Scana Steel Stavanger. The casting was done at Frekhaug stål, Bergen
by adding 0.05, 0.075 and 0.1 wt% cerium to 120 kg casts. There were a
total of 3 parallels, where 1 was cast at 1525ºC and 2 series at 1540ºC.
The as cast structure was significantly refined, the columnar zone was
reduced from 22 mm to being absent in the cast with most cerium added.
The mechanical results show a linear increase in both yield and ultimate
tensile strength with increasing amount of cerium. The elongation
increased somewhat, but the impact toughness decreased linearly with
increasing cerium content. The casting temperature did not seem to have
an effect on the grain refining.
5
6
Table of contents
Table of contents ........................................................................................ 7
1. Introduction ............................................................................................ 9
2. Theoretical background ........................................................................ 12
2.1 Structural zones in casting .............................................................. 12
2.1.1 Chill zone. ................................................................................ 12
2.1.2 Columnar zone. ........................................................................ 13
2.1.3 Equiaxed zone .......................................................................... 13
2.2 Effect of grain refinement............................................................... 14
2.2.1 Homogeneous nucleation ......................................................... 14
2.2.2 Heterogeneous nucleation ........................................................ 15
2.2.3 Constitutional supercooling ..................................................... 17
2.2.4 Cellular dendritic growth ......................................................... 18
2.2.5 Segregation .............................................................................. 19
2.2.6 Shrinkage ................................................................................. 20
2.3 Cerium as grain refiner ................................................................... 20
2.4 Inclusions in a metal ....................................................................... 22
2.5 Super Duplex steel .......................................................................... 23
2.3.1 Composition and properties ..................................................... 24
2.3.2 Heat treatment .......................................................................... 27
3. Experimental procedure ....................................................................... 28
3.1 Casting ............................................................................................ 28
3.2 Mechanical testing .......................................................................... 30
3.3 Chemical analysis ........................................................................... 32
3.4 Macroscopic examination. .............................................................. 32
3.5 Microscopic examination ............................................................... 32
3.6 SEM/EPMA .................................................................................... 32
7
4. Results .................................................................................................. 34
4.1 Chemical analysis ........................................................................... 34
4.2 Macroscopic examination ............................................................... 35
4.4 Light microscopic examination ...................................................... 40
4.5 EPMA ............................................................................................. 43
4.5 Mechanical testing .......................................................................... 53
4.6 SEM ................................................................................................ 56
5. Discussion ............................................................................................ 60
5.1 Experiment...................................................................................... 60
5.2 As cast structure.............................................................................. 60
5.3 Light microscopic examination ...................................................... 61
5.4 Cerium particles in the metal .......................................................... 61
5.5 Mechanical tests ............................................................................. 61
5.6 Effect of casting temperature .......................................................... 63
6. Conclusion ............................................................................................ 64
Proposal for further work ......................................................................... 65
Appendix .................................................................................................. 68
8
1. Introduction
Steels with high chromium and high nickel contents have for years been
used in demanding corrosive environments. The alloys are available in a
wide variety of versions. These steels are mostly used in marine
construction, chemical industries and power plants due to the combination
of mechanical properties, corrosion resistance and availability.
Duplex stainless steels contain roughly 18-26% Cr, 3-8% Ni as the main
alloying elements, causing the steel to have both a ferritic and austenitic
microstructure. The first commercial product was available as early as
1929, however most of the development happened since the 1970s, due to
the advances in refining. [1]. The super duplex alloys came in the 1980s,
with slightly higher alloy content, giving the steels a PREN value of at
least 40, which separates super duplex from duplex.
Compared to fully ferritic or austenitic stainless steel, duplex steel
combines the mechanical properties of the two. Increased ferrite content
in austenitic steel increases its strength and weldability, but reduces the
toughness and elongation. Typical strength for a cast duplex steel is 500
MPa yield, 750 MPa ultimate tensile strength. The goal for this report is
to increase these values by refining the microstructure by adding cerium
in to the melt.
Duplex steel solidifies completely as δ-ferrite, followed by solid state
transformation to ~50 vol% austenite during cooling. This formation will
not change the grain size as low alloyed steel, any grain refinement must
therefore happen during solidification. A cast ingot can usually be
separated into 3 different zones; chill, columnar and equiaxed zone. The
chill zone will be small equiaxed grains located along the mould wall
where solidification is quick, but when the cooling rate decreases the
grains will grow as columns into the ingot. After the columnar growth
there will be equiaxed grains in the centre. An ingot with a large
columnar zone will have large segregations compared with a small or
none columnar zone. The size of the columnar zone is therefore important
regarding the mechanical properties of the ingot.[2]
9
To improve the mechanical properties, particles can be introduced to act
as impurities for nucleation of equiaxed grains ahead of the solidifying
front. For nucleation ahead of the columnar zone to occur, an undercooled
region is required for precipitation of such particles. Such an
undercooling can be achieved when dendrites grow into the melt, pure
metal will solidify first and push solute ahead of the dendrite tip causing
the liquid in front of the dendrite contain higher solute than the rest of the
liquid, thus lowering the liquidus temperature causing a constitutional
supercooling. When stable particles form ahead of the columnar zone new
grain can nucleate much like solidification against the mould wall. A high
number of such nucleation sites can make an ingot consist completely of
equiaxed grains reducing segregation and improving the mechanical
properties.[3]
Cerium based particles have in recent years been tested as nucleation sites
for heterogeneous nucleation for austenitic and duplex steels. The results
show a correlation between reduced columnar zones with added cerium
[4][sintefrapport]. Cerium particles have shown promising properties
regarding nucleation in steel due to requiring low undercooling for
precipitation of cerium sulphides and oxides and the degree of atomic
misfit is low. Elkem have in recent years, in cooperation with Sintef,
developed grain refiners containing cerium. Elkem Grain Refiner (EGR)
is an alloy containing ~10% cerium along with elements present in the
liquid. The grain refiner is added to the tap stream before casting of the
component, making cerium react with impurities as sulphur and oxygen to
precipitate particles.
Scana Steel Stavanger have previously used EGR with super duplex steel
after experienced increased forgeability, but have had problems with large
oxide inclusions with detrimental effect of the products. Therefore Scana
Steel wishes to explore the possibility to add EGR to super duplex
without these problems.[5]
The experiments described in this report were done at Frekhaug stål, with
steel supplied by Scana Steel Stavanger. The experiment consisted of
adding EGR to super duplex steel supplied by Scana Steel in various
10
amounts to 3 parallels. Test blocks were designed to provoke typical ingot
solidification, giving a structure with chill, columnar and equiaxed zone.
The blocks were later examined by macro/microscopy, mechanical testing
and SEM/EPMA.
11
2. Theoretical background
2.1 Structural zones in casting
Generally a casted ingot will have three different zones. Near the mould
wall there will be an outer chill zone of equiaxed crystals, a columnar
zone of elongated grains, and an equiaxed zone in the centre. Often one of
the zones is missing; in stainless steels the structure can be fully
columnar, with no central equiaxed zone and no chill zone. A well grain
refined aluminium alloy can have a structure of completely equiaxed
grains. The occurrence and degree of these zones are known to depend
both on nucleation and on crystal multiplication. [2]
Figure 2.1: The different zones in a cast ingot.[6]
2.1.1 Chill zone.
During casting the liquid metal comes in contact with the cold and is
rapidly cooled below its liquidus temperature. The undercooling will
cause the liquid to form many solid nuclei along the mould wall which
will grow into the liquid, some grains more than others. [3]
12
2.1.2 Columnar zone.
Soon after pouring the temperature gradient of the liquid near the mould
wall decreases and the crystals in the chill zone grow dendritically in
certain crystallographic directions, <100> for cubic metals. Crystals with
the <100> directions parallel to the largest temperature gradient,
perpendicular to the mould wall, will outgrow neighbours with a less
favourable direction. This leads to the formation of the columnar zone, all
with nearly the same orientation. The diameter of the grains can increase
making it possible for some ternary arms to grow ahead of their
neighbours and becoming new primary arms as shown in the figure. This
is simply due to a corresponding decrease in the cooling rate with time
after pouring. The region between the tips of the dendrites and the point
where the last drop of liquid is solidifying is known as the mushy or pasty
zone. The length of this zone depends on the temperature gradient and the
non-equilibrium freezing range of the alloy.[3]
2.1.3 Equiaxed zone
The centre of an ingot often consists of equiaxed grains randomly
oriented. One origin of these grains is thought to be remelted dendrite side
arms, called grain multiplication. The side arms are narrower at the roots,
which makes it possible to isolate the arms if the temperature increases
after the arm is formed. The detached arm can then act as a seed for a new
crystal. Convection currents can transport the detached arm to develop
into an equiaxed grain. Equiaxed grains can also grow from nucleation
sites in the liquid ahead of the columnar zone. Similar to the chill zone,
the grains need a solid wall to grow from, where stable particles already
formed in constitutionally supercooled liquid can act as nucleation sites
for new grains.[3] Equiaxed grains are beneficial in regards of mechanical
properties of a metal. Seeing as the orientation will change when moving
from one grain to another across a grain boundary, there is a region of
disturbed lattice only a few atomic diameters thick. High angle grain
boundaries have high surface energy, causing boundaries to be a source
for solid-state reactions as diffusion, phase transformation and
precipitation reactions. This also means a higher concentration of solute
atoms than in centre of grains. When a polycrystal is under strain, the
grains will try to deform homogeneously while still keeping the
13
boundaries intact will cause differences in the deformation between
neighbouring grains and within grains. If the strain increases and grain
size decreases the deformation becomes more homogeneous. Deformation
bands will be created due to lattice rotations because of different slip
systems in the same grain. More slip systems are in effect near the grain
boundary, therefore a reduced grain size will make the effect of grain
boundaries influence the centre of a grain making the metal stronger. A
general relationship between yield stress and grain size is known as the
Hall-Petch relation given in equation 1 [7]
σ0 = σi + kD-1/2
(1)
Where σ0 is the yield stress, σi is friction stress, k is the locking parameter,
and D is grain diameter.
2.2 Effect of grain refinement
2.2.1 Homogeneous nucleation
For an impurity-free liquid to solidify it requires significant cooling
below its melting temperature for the driving force to be big enough to
solidify the melt. If a liquid at a temperature ΔT below Tm with a free
energy G1 the atoms will cluster together to form a small sphere of solid,
the free energy will change to G2 given by:
(2)
Where VS is the volume of the solid sphere, VL the volume of liquid, ASL
is the area of the solid/liquid interface, GSV and GLV are the free energies
per unit volume of solid and liquid respectively, and γSL the solid/liquid
interfacial free energy. Below Tm, Gv is positive since the free energy
change with formation of a solid has a negative contribution due to lower
free energy of a bulk solid, but the creation of a solid/liquid interface is a
positive contribution. This gives a sphere with radius r with a free energy:
(3)
14
This equation shows a dependency of r, the second term is dominant at
first, but as r increases the first term will cause the product to be negative.
There will be a critical r* which is when ΔGhom is at a maximum. If r < r*
the free energy can be reduced by dissolution of the solid, but if r > r* the
system can reduces if the solid grows, which will is referred to as a
nuclei. [3]
2.2.2 Heterogeneous nucleation
For nucleation to occur at small undercooling, the interfacial energy term
must be reduced. The easiest way to achieve this is to have the nucleus
form in contact with either a particle or the mould wall. Assuming the
solid/liquid interfacial free energy, γSL is isotropic it can be shown that for
a given volume of the solid the total interfacial energy of the system is
minimized if the shape of a spherical cal with a “wetting” angle θ given
by the condition that the interfacial tensions γML, γSM and γSL balance in the
plane of the mould. [3]
(4)
Figure 2.2: Illustration of nucleation from a solid wall.[3]
The vertical component of γSL unbalanced, forcing the mould surface
upwards, giving equation (4) to define the optimum embryo shape when
the mould wall remain planar. The formation of an embryo is associated
with the excess free energy, which can be written in terms of the wetting
angle and cap radius.
15
(5)
Where
(6)
The expression of ΔGhet is the same as for ΔGhom except for the S(θ) factor.
This factor is only dependent on the wetting angle.
(7)
This expression shows a lower activation energy barrier against
heterogeneous nucleation than homogenous due to the shape factor. By
differentiation of equation (5) it can be shown that
(8)
Which shows the critical nucleus radius being unaffected by the mould
wall but and only dependent on undercooling.
Figure 2.3: Schematic illustration of excess free energy for homogeneous and
heterogeneous nucleation compared.[3]
16
2.2.3 Constitutional supercooling
After the first metal near the mould wall have solidified the temperature
gradient will decrease rapidly. As a result of the varying solute
concentration ahead of the solidification front there is also a variation in
the liquidus temperature. The liquidus temperature increases with the
distance from the interface because the lower the solute content the higher
the liquidus temperature, illustrated in fig. 2.4. When the temperature in
front of the interface is below the liquidus temperature the liquid is
constitutionally supercooled, which means the supercooling comes from
change in composition, not temperature. This supercooling results in
instability of the plane front since any bump forming on the interface
would find itself in supercooled liquid and therefore would not disappear.
[2]
Figure 2.4: Illustration of constitutional supercooling.[3]
The concentration gradient of the liquid can be expressed
(9)
Where DE is a turbulent diffusion coefficient and k is a constant, v is the
solidification speed and cl is the liquid concentration. The criterion for
constitutional supercooling is that the slope for the liquidus temperature
(dT1/dz) is bigger than the temperature gradient (dT/dz), which gives the
following expression.
17
(10)
This expression shows that a high solidification velocity is beneficial for
the supercooling, but also a low as possible temperature gradient will
increase the supercooled area. The temperature gradient can be lowered
by reducing the pouring temperature.[8]
2.2.4 Cellular dendritic growth
If the temperature gradient ahead of an initially planar interface is
gradually reduced below the critical value the first stage breaking down
the planar interface. The formation of the first protrusion causes solute to
be rejected laterally and pile up at the root of the protrusion. This lowers
the liquidus temperature causing which trigger the formation of other
protrusions. Assuming a region of constitutionally supercooled liquid in
front of the solidifying interface, the temperature of the tip of a protrusion
will be higher than the surrounding interface. Provided that the tip
remains below the local liquidus temperature solidification is still possible
and the protrusion can develop. But if the temperature gradient ahead of
the interface is steeper than the critical gradient the tip will be raised
above the liquidus temperature and the protrusion will melt again.
Eventually the protrusions develop into long arms or cells growing
parallel to the direction of heat flow. The solute rejected from the
solidifying liquid concentrates into the cell walls which solidify at the
lowest temperatures. The tips of the cells, however, grow into the hottest
liquid and therefore contain the least solute. Even if the solute
concentration is much smaller than maximum solubility the liquid
between the cells may reach the eutectic composition in which case the
cell walls will contain a second phase. Each cell has virtually the same
orientation as its neighbours and together they form a single grain. [3]
Cellular microstructures are only stable for a certain range of temperature
gradients. At sufficiently low temperature gradients the cells, are
observed to develop secondary arms, and at still lower temperature
gradients tertiary arms develop, i.e. dendrites form. Associated with this
change in morphology there is a change in the direction of the primary
arms away from the direction of heat flow into the crystallographically
18
preferred directions such as (100) for cubic metals. In general, the trend to
form dendrites increases as the solidification range increases. Therefore
the effectiveness of different solutes can vary widely. The reason for the
change from cells to dendrites is not fully understood. However it is
probably associated with the creation with the creation of constitutional
supercooling in the liquid between the cells causing interface instabilities
in the transverse direction. Note that for unidirectional solidification there
is approximately no temperature gradient perpendicular to the growth
direction. The cell or dendrite arm spacing developing is probably that
which reduces the constitutional supercooling in the intervening liquid to
a very low level. This would be consistent with the observation that cell
and dendrite arm spacings both decrease with increasing cooling rate:
Higher cooling rates allow less time for lateral diffusion of the rejected
solute and therefore require smaller cell or dendrite arm spacings to avoid
constitutional supercooling.[3]
Figure 2.5: Columnar grains with similar direction of growth.[3]
2.2.5 Segregation
Two types of segregation can be distinguished in solidified structures.
There is macrosegregation: composition changes over distances
comparable to the size of the specimen. There is also microsegregation:
segregation which occurs on the scale of secondary dendrite arm spacing.
It has been known that large differences in composition can arise across
the dendrites due to coring and the formation of non-equilibrium phases
in the last solidifying drops of liquid. Experimentally it is found that
while cooling rate affects the spacing of the dendrites it does not
significantly change the amplitude of the solute concentration provided
the dendrite morphology does not change and that diffusion in the solid is
19
negligible. The mushy zone length is proportional to the non-equilibrium
solidification range, which is usually larger than the equilibrium melting
range.
There are four important factors that can lead to macrosegregation in
ingots: Shrinkage due to solidification and thermal contraction; density
differences in the interdendritic liquid; density differences between the
solid and liquid; and convection currents driven by temperature-induced
density differences in the liquid. All of these factors can induce
macrosegregation by causing mass flow over large distances during
solidification. Interdendritic liquid flow can also be induced by gravity
effects. In general segregation is undesirable as it has marked deleterious
effects on mechanical properties. The effects of microsegregation can be
reduced by subsequent homogenization heat treatment, but diffusion in
the solid is far too slow to be able to remove macrosegregation. [3]
2.2.6 Shrinkage
Most metals shrink on solidification and this has important consequences
for the final ingot structure. In alloys with a narrow freezing range the
mushy zone is also narrow and as the outer shell of solid thickens the
level of the remaining liquid continually decreases until finally when
solidification is complete the ingot contains a deep central cavity or pipe.
In alloys with a wide freezing range the mushy zone can occupy the
whole of the ingot. In this case no central pipe is formed. Instead the
liquid level gradually falls across the width of the ingot as liquid flows
down to compensate for the shrinkage of the dendrites. However, as the
interdendritic channel close up this liquid flow is inhibited so that the last
pools of liquid to solidify leave small voids or pores.
2.3 Cerium as grain refiner
Cerium is a very strong oxide and sulphide former which means it that
cerium will react with available oxygen and sulphur when added to liquid
metal. The inclusions cerium will form are beneficial to the refine the
casting microstructure by acting as sites for heterogeneous nucleation. To
be a suitable nucleation site inclusion must be able to be formed at low
undercooling and have a low atomic misfit with the solidifying metal.
20
Figure 2.7 summarizes the potential of some inclusions to act as
nucleation sites for steel which solidifies as ferrite.
Figure 2.7 Required undercooling versus lattice misfit for different inclusions found
in steel.[9]
(11)
Where (hkl)s = a low-index plane of the substrate, [uvw]s = a low-index
direction in (hkl)s, (hkl)n = a low-index plane in the nucleated solid,
[uvw]n = a low-index direction in (hkl)n, d[uvw]s = the interatomic spacing
along [uvw]s, d[uvw]n = the interatomic spacing along [uvw]n ,γ = the angle
between [uvw]s and [uvw]n. The degree of misfit shown in figure 2.7, and
explained in equation (11) is calculated for ferrite lattice. The density of
Cerium has been used as grain refiner with success; it is often added as
Elkem Grain Refiner (EGR), a master alloy consisting of Ce, Si, Cr and
C. A number of experiments have used cerium as grain refiner;
experiments with low alloy steel have shown a reduction of secondary
21
dendrite arm spacing and increase in area fraction of equaxed grains with
cerium addition[10]. The columnar zone has been greatly reduced in
austenitic manganese steels have and an increase in impact toughness. [4,
11, 12] Stainless steels, ferritic, austenitic and duplex, have been
experimented on in regards of grain refining, where the microstructure
have been refined, but not translated into improved mechanical
properties.[13, 14] There have however been problems with large cerium
oxide clusters during industry scale production; clusters have caused
failure in components.
2.4 Inclusions in a metal
There are generally two kinds of non-metallic inclusions in steel; trapped
unintentionally, or those separated from the metal due to temperature or
composition change. The inclusions trapped unintentionally are normally
obstructed in the melt before casting, like slag. The other inclusions are
formed by reactions in the melt or precipitated when solubility decreases
during cooling, normally; oxides, sulphides or nitrides. During refining
the goal is to reduce the amount of these elements, but they will always be
present in some amount. These inclusions can be damaging to the
properties of steel. Investigation of low alloyed steel has shown that the
amount of inclusions has little or no effect on ultimate tensile strength,
and also a small increase in yield strength. The toughness and fatigue
properties of steel have been reported to decrease, but the overall
conclusion is that the type is more important than the number of
inclusion.[15] Therefore process control is more important; avoiding large
clusters of inclusions will determine the success of particles introduced as
grain refiner.
There are a number of methods for removing unwanted inclusions;
settling, flotation, filtration and stirring are the most used. In an induction
furnace and a ladle without possibility for bubbling, settling is the only
method available. Inclusions are often lighter than the metal, making
gravity a help for removing inclusions. Light inclusions will float up to
the slag, and heavier inclusions will settle to the bottom. The driving
force for settling is the difference of density between the inclusion and
liquid metal expressed in equation (12)
22
Fg = ΔρgV
(12)
Where Fg is the gravity force, g is the gravity acceleration, Δρ is the
difference of density and V is the volume of the inclusion. The
effectiveness of the settling is determined by the relative speed of the
inclusion relative to the liquid which can be expressed by equation 2.X.
(13)
Where ur is the relative speed, ɸ is an empirical friction factor and Ap is
the area of the cross-section transverse to the direction of flow. The
friction factor for a sphere is 12/Re. Inclusions are normally so small that
that Re < 2 giving the following equation for spherical particles, known as
Stoke’s law:
(14)
Where a is the radius of the particle and ν is the kinematic viscosity of the
liquid. The equation is valid for particles with a radius up to 50 µm,
which covers most inclusions, if the difference in density is not too big.
The mixing of the inclusions in the melt is important since inclusions can,
besides nucleation solidification, also influence the mechanical properties
in a negative manner. The holding time is important to let the settling
affect the inclusions in the, but also a too long holding time can make the
cerium inclusion grow, to potentially be mechanically dangerous. A small
scale experiment have found that a holding time of 5 minutes produces
the most amount of inclusions, and a longer time will cause the size of the
inclusions to increase.[16]
2.5 Super Duplex steel
Cast stainless steels are high alloyed steel with superior corrosion
resistance in tough environments due to high contents of chromium and
nickel, also nitrogen molybdenum, copper, silicon and tungsten is present
to control properties. Duplex stainless steels got their name from having a
mix of austenite (fcc) and ferrite (bcc) microstructure, the amount of each
phase is a function of composition and heat treatment. Equal amounts of
23
ferrite and austenite is usually the desired ratio in these steels. Super
Duplex is defined to have the pitting resistance equivalent number (PREN)
above 40, unlike duplex steel. A high PREN number is crucial in high
chloride environments, the equation for calculating PREN is given in eq.
(15).
PREN = %Cr + 3.3(%Mo) + 16(%N)
(15)
2.3.1 Composition and properties
The amount of the different elements determine the amount of each
phase, the Super Duplex alloy used in these experiments are a Scana Steel
Stavanger alloy similar to ASTM A182 F55 but with smaller range in
composition than the standard. The composition of ASTM A182 F55 is
given in table 2.X along the phase each element promotes. Controlling the
amount of each phase is important to achieve the desired properties; both
phases have some wanted and unwanted assets. Ferrite is beneficial to
improve weldability, stress corrosion cracking and strength but have low
toughness compared to the austenitic phase. The resistance to stress
corrosion cracking is increased due to the ferrite pools in the austenite
matrix making it harder for cracks to propagate.
Table 2.1: Composition of ASTM A182 F55 Super Duplex steel, and which phase
the element promotes.
Wt%
C
<0.03
Prom.
Aust
Cr
24.026.0
Fer.
Ni
6.08.0
Aust
Cu
0.51.0
Aust
Mn
<1.0
Aust
Mo
3.04.0
Fer.
N
0.20.3
Aust
P
0.0
3
Fer.
S
0.0
1
-
Si
1.0
W
0.51.0
Fer. Fer.
Chromium is added to improve the local corrosion resistance by forming
a passive oxide-layer, higher chromium content than 25% is normally not
beneficial because of precipitation of intermetallic phases which leads to
reduced ductility, toughness and corrosion properties. Chromium, along
with other elements, promotes ferrite with varying effect. The chromium
equivalent indicates the effect of each element in that regard.
Creq= %Cr + %Mo + 0.7 x %Nb
(16)
24
As seen from equations for PREN and Creq molybdenum is beneficial
against pitting corrosion, as well as crevice corrosion, and stabilising
ferrite. Molybdenum increases pitting resistance by suppressing active
sites by creating an oxy-hydroxide or molybdate ion. [1] Molybdenum
content from 2 to 3% will give increased resistance against chloridebearing environments, crevice corrosion and pitting. Molybdenum content
between 3 and 4% will weaken its resistance to highly oxidizing
environments but increase the resistance against halide-bearing media and
reducing acids. [17]
There are a number of elements that stabilises austenite, the nickel
equivalent shows which elements promotes austenite and their effect,
given in equation (17)
Nieq = %Ni + 35 x %C + 20 x %N + 0.25 x %Cu
(17)
To achieve about equal amounts of ferrite and austenite the ferrite
stabilisers need to be balanced austenite stabilisers, therefore the nickel
content will primarily depend on the chromium content illustrated in
figure 2.X.
25
Figure 2.8: Schematic phase diagram showing increasing nickel effect on phases
present, based on [18]
Too high nickel content will give austenite content above 50% which
increases the chance to transform ferrite to intermetallic phases.
Carbon is present in small numbers, maximum 0.03 wt% due to the
danger of chromium-carbides at grain boundaries, which is detrimental to
pitting and intergranular corrosion. Nitrogen has several effects on
duplex steel, even small content can increase pitting resistance, strength
and austenite content as seen from eq. (15) and (17) Nitrogen has low
solubility in ferrite, which solidifies first, making relative high content
dangerous for pores if segregated in to the small amount of austenite
present during solidification, as seen in [19]. Nitrogen content in the
range 0.3-0.4 wt% will make the steel solidify as austenite-ferrite opposed
to ferrite single-phase, reducing the ferrite content.[19]
There are also some elements with content below 1 wt%, manganese,
copper, silicon and tungsten. Manganese, silicon and tungsten is not
included in any of the equations for PREN/Creq/Nieq, but is added for
increased strength and abrasion; Mn. Crevice corrosion resistance and
26
pitting resistance; W. Silicon gives increased resistance to nitric acid and
high temperature service, but the content is normally limited to 1 wt% due
to promoting sigma phase. Cu addition is limited due to reducing hot
ductility, but also reduces general corrosion.[1]
2.3.2 Heat treatment
When duplex steel solidifies, δ-ferrite dendrites are formed segregating
the nickel in to austenite areas between the ferrite dendrites. This causes
the steel to be heavily segregated, requiring heat treatment [19]. The first
part of heat treatment is there for a homogenization to ensure a homogen
steel, this is normally in the ferrite region since diffusion speed is about
twice in ferrite compared to austenite. If duplex steel is cooled slowly,
chromium-rich carbides form at ferrite-austenite interfaces, these carbides
deplete the surrounding matrix of chromium weakening the corrosion
resistance of the alloy, carbides can also lead to localized pitting.
Therefore solution annealing makes the steel less vulnerable to
intergranular attack. The solution annealing should also be performed at a
temperature in the center of the dual phase area to give the steel equal
amounts of ferrite and austenite. Figure 2.10 shows a schematic view of
the phases and where homogenization and solution annealing will be
done. Solution annealing is completed by water quenching.[1]
Figure 2.9: Schematic phase diagram showing typical heat treatment steps phase to
homogenization and phase balance.
27
3. Experimental procedure
3.1 Casting
Casting of the keel blocks used in this experiment was done at Frekhaug
Stål in Bergen. The blocks were cast by melting pre-alloyed blocks of
S4501 Super Duplex steel supplied by Scana Steel Stavanger, S4501 is
Scana own version of ASTM A182 F55. Scana Steel supplied two blocks
at 790 kg each dimensioned to fit the furnace at Frekhaug. The two blocks
was melted in an induction furnace, then added CaSi to deoxidise the
melt. The 12 blocks which were to be cast was divided in to 3 series, each
with 4 blocks where there was 1 reference and 0.05%, 0.075% and 0.1%
Ce respectively added to the ladle.
Table 3.1: Overview of block cast, addition of Ce and Al and casting temperature
Block #
1
2
3
4
5
6
7
8
9
10
11
12
% Ce
0.05
0.075
0.1
0.05
0.075
0.1
0.05
0.075
0.1
% Al
0.021
0.021
0.021
0.021
0.021
0.021
0.021
0.021
0.021
0.021
0.021
0.021
Casting temp
1525°C
1525°C
1525°C
1525°C
1540°C
1540°C
1540°C
1540°C
1540°C
1540°C
1540°C
1540°C
The aimed casting temperature was 1540°C, but the first series of 4
blocks were cast at 1525°C to examine the effect of casting temperature.
The cerium containing masteralloy (EGR) was added just after FeAl into
the ladle during tapping from the furnace. Each cast consisted of 120kg
liquid in the ladle. The measure the time of each cast is given in figure
3.1. The figure shows the time from start tapping from furnace, to EGR
addition, tapping stop, pouring into the mould start and finish. The
various lengths of time between tapping stop and pouring start was to
reach the desired temperature, varying from 19 seconds in cast #1 to 2
28
minutes, 6 seconds for cast #11, the blocks were cast in the order shown
in the figure.
Figure 3.1: Time used for each cast from start to finish.
The moulds were made at Frekhaug Stål, had dimensions 250 x 120 x 110
mm excluding the feeder, illustrated in figure 3.2. After cooling, a part
with dimensions 50 x 120 x 110 mm was cut to examine the as cast
structure of the cross section of the block. The remaining part was heat
treated at Frekhaug Stål. The heat treatment consisted of a
homogenization treatment for 2 hours at 1130°C followed by 2.5 hours at
1070°C for phase balance followed by water quenching from this
temperature to retain the phase composition and avoid sigma phase
formation.
29
Figure 3.2: Geometry of block, showing cuts for mechanical testing and
macroscopic examination.
3.2 Mechanical testing
The samples for mechanical testing were cut from 20 mm thick slice 10
mm below the feeder. The placements of the test bars are given in figures
3.4 and 3.5. From each block there was cut 2 tensile test specimens, they
were cut side by side from ¼ of width of the block, farthest away from the
inlet. The impact toughness specimens were cut from ¼ of width of the
opposite side from the tensile specimens, cutting all 3 specimens
longitudinal starting from farthest away from the inlet. The dimensions of
the specimens were according to standard ASTM E23. The tensile tests
were performed at room temperature, while the impact toughness tests
were performed at -46°C with liquid CO2 as cooling agent. The hardness
tests were performed on the tensile bars.
30
Figure 3.3 Overview of mechanical testing samples in block.
Figure 3.4 Dimensions of mechanical test samples.
31
3.3 Chemical analysis
The chemical analysis was performed with an optical Emission
Spectrograph at Scana Steel Stavanger, every tensile bar was analysed
two times, making the result for each block an average of 4 analyses.
Cerium and oxygen content were analysed by combustion at D-lab,
Degerfors, Sweden.
3.4 Macroscopic examination.
The keel block was cut in to two parts, one piece was kept in it’s as cast
condition to examine the structure. A 10 mm thick cross section was
etched in concentrated HCl diluted 50-50 with distilled water. The etchant
was heated and kept at 70-80°C, etching the samples for 15 minutes. The
samples got a dark layer during etching which was removed by adding a
few ml of concentrated HNO3 at the end. The samples were photographed
with a Canon D60.
3.5 Microscopic examination
The microscopic examination was performed on samples from charpy
specimen E, which showed the most consistent results. The specimens
was cut 10 mm from the fracture and cast in 25 mm diameter Struers
ClaroCit. The samples were manually grinded in a Struers Knuth-rotor in
mediums 80, 500 and 1200p discs, followed by polishing in a Struers DPU3 with mediums 3 and 1 µm. The samples were etched electrolytic in
40% NaOH at 8 volt potential until visible reaction, which was after about
3 seconds. The samples were photographed with Progress C10 plus digital
camera attached to a Leica MeF4M light microscope. Measuring of ferrite
content was performed by software scan at Scana Steel by scanning 10
different areas of each block to determine volume amount of ferrite.
3.6 SEM/EPMA
The fracture surfaces were also examined in a Zeiss Ultra 55, the samples
were kept in a desiccator to avoid contamination of the surface. The
fracture surfaces were imaged with secondary electrons and visible
particles in the fracture dimples were analyzed with EDS to determine if
cerium was present, or other differences between the samples.
32
The Electron Probe Micro Analysis (EPMA) was performed on the
samples used for light microscopy. Areas with a high number of particles
were chosen to be pictured and subsequently identify particle
composition. Particles containing cerium will be a lot brighter due to
atomic number contrast, areas with bright particles were chosen
deliberately. The instrument used was Jeol JXA-8500F with 5
Wavelength dispersive X-ray spectrometers.
33
4. Results
4.1 Chemical analysis
Chemical compositions are given in table 4.1; some elements which were
present in very small amounts are not included in this table. The variation
between the different casts shows that EGR did not significantly alter the
balance between the elements, but the Si-content is higher in cast 7,8,11
and 12 which was added the most EGR. Traces of cerium are found in all
3 references, which is either residue from the casts added with EGR, or
inaccurate analysis. There was not found any trace of Ca, which was
added as CaSi to the furnace for deoxidation, but the oxygen content
averaged of 430 ppm indicates unsuccessful deoxidation. The amount Si
varies since it was added both as CaSi and as EGR. The references
contain less Si than the blocks with EGR added, and it increases with
more EGR added.
Table 4.1: Chemical composition of all blocks,
C
Si
Mn
Cr
Mo
Ni
Al
W
N
O
Ce
1
0.029
0.370
0.540
25.0
3.65
7.22
0.006
0.585
0.261
0.035
0.007
2
0.027
0.390
0.525
24.9
3.65
7.20
0.004
0.575
0.253
0.039
0.034
3
0.028
0.380
0.533
24.9
3.65
7.21
0.005
0.580
0.257
0.042
0.036
4
0.027
0.385
0.529
24.9
3.65
7.21
0.004
0.578
0.255
0.041
0.064
5
0.027
0.360
0.480
24.8
3.68
7.29
0.004
0.570
0.258
0.040
0.009
6
0.028
0.400
0.460
24.7
3.68
7.33
0.006
0.570
0.264
0.050
0.049
7
0.028
0.430
0.530
25.1
3.65
7.23
0.006
0.570
0.253
0.043
0.052
8
0.028
0.475
0.520
25.1
3.61
7.14
0.004
0.565
0.250
0.047
0.073
9
0.028
0.345
0.510
25.0
3.67
7.29
0.004
0.575
0.254
0.034
0.006
10
0.028
0.395
0.500
25.0
3.65
7.26
0.006
0.570
0.257
0.047
0.035
11
0.028
0.430
0.480
24.9
3.63
7.16
0.007
0.570
0.252
0.051
0.067
12
0.028
0.450
0.460
24.9
3.63
7.22
0.006
0.575
0.255
0.052
0.078
The analysed cerium content is plotted versus added cerium in figure 4.1,
the trend lines show average yield of the two casting temperatures; 78%
34
yield for 1540°C and 60% for blocks cast at 1525°C; while the overall
yield is 72%.
Yield of cerium added
Cerium content [wt%]
0.1
0.08
0.06
0.04
0.02
0
0
0.02
0.04
0.06
0.08
0.1
Cerium added [wt%]
Figure 4.1: Graph of cerium analysed versus cerium added, lines showing average
yield for series cast at 1540 ºC (blue) and 1525ºC (red).
4.2 Macroscopic examination
The images of the cross section of the blocks 1-4 and 9-12 in their as cast
condition is shown in figures 4.2-4.9. The images clearly show reducing
length of columnar zone with increasing amount of cerium, summarized
in table 4.2. The images also show a smaller grain size in the equiaxed
zone with increased cerium content, but this has not been quantified.
35
Figure 4.2: Cross section of the as cast block #1 cast at 1525°C, 0% cerium added.
Figure 4.3: Cross section of the as cast block #2 cast at 1525°C, 0.034% cerium.
36
Figure 4.4: Cross section of the as cast block #3 cast at 1525°C, 0.036% cerium.
Figure 4.5: Cross section of the as cast block #4 cast at 1525°C, 0.064% cerium.
37
Figure 4.6: Cross section of the as cast block #9 cast at 1540°C, 0% cerium added.
Figure 4.7 Cross section of the as cast block #10 cast at 1540°C, 0.037% cerium.
38
Figure 4.8: Cross section of the as cast block #11 cast at 1540°C, 0.067% cerium.
Figure 4.9: Cross section of the as cast block #12 cast at 1540°C, 0.078% cerium.
39
Table 4.2: Length of columnar zone with cerium content.
Block Length of
columnar
zone[mm]
1
22.5
2
22
3
7.5
4
0
5
23
6
18.5
7
13.5
8
0
9
19
10
19
11
16.5
12
0
Cerium
content
[wt%]
0.007
0.034
0.036
0.064
0.009
0.049
0.052
0.073
0.006
0.035
0.067
0.078
4.4 Light microscopic examination
Images acquired from light microscopic examination are given in figures
4.10-4.13, the examination did not reveal much about the microstructure
since it is difficult to determine grain size for the dual phase nature of the
steel. Ferrite content versus cerium content is given in figure 4.14 for the
blocks cast at 1540ºC, which does not show correlation between phase
balance and cerium content. The ferrite content versus Ce content is
plotted in figure 4.37, showing a slight increase of ferrite with increasing
amount of Ce.
40
Figure 4.10: Light microscopic image of block 1, reference cast at 1525ºC.
Figure 4.11: Light microscopic image of block 2, 0.034% Ce cast at 1525 ºC.
41
Figure 4.12: Light microscopic image of block 3, 0.036% Ce cast at 1525 ºC.
Figure 4.13: Light microscopic image of block 4, 0.064% Ce cast at 1525 ºC.
42
Ferrite content, blocks cast at 1540ºC
50
Ferrite content [%]
48
46
44
42
40
0
0.02
0.04
0.06
0.08
0.1
Ce content [Wt%]
Figure 4.14 Ferrite content versus cerium content for blocks with casting
temperature 1540°C.
4.5 EPMA
Figures 4.16-4.26 shows particles analysed with EPMA, some of the
particles were analysed quantitatively with the result given in the
corresponding table where all values are atomic%. Most of the particles
from block with added cerium included 3 different phases; bright, grey
and dark, some had only one or two of the phases large enough to
analyse. An example from block 8 is shown in figure 4.15 where all three
phases are present in the particles. The largest particle in figure 4.18 is
also analysed with backscatter electrons in figure 4.24. The blocks
without cerium addition had only particles with a single phase.
43
Figure 4.15: Image of particles up close with different contrast, from block 8.
44
Figure 4.16: Image of particles from block 5 (reference), composition given in table
4.3
Table 4.3: Composition of particles in figure 4.19, and likely phase, all values in
atomic%.
#
1
2
3
4
5
Al
8.3
9.6
10.5
10.4
10.1
O
56.0
56.3
56.2
56.3
56.1
S
0
0.1
0
0
0.1
C
1.9
2.1
2.1
2.1
2.1
Cr
19.6
17.8
17.2
17. 1
17.8
Mn
12.7
12.7
12.6
12.5
12.5
Phase
(CrMnAl)2O3
(CrMnAl)2O3
(CrMnAl)2O3
(CrMnAl)2O3
(CrMnAl)2O3
45
Figure 4.17: Image of particles in block 6 (0.049% Ce), composition given in table
4.4.
Table 4.4: Composition of particles in figure 4.20, and likely phase, all values in
atomic%.
#
1
1
2
2
2
3
3
3
4
4
Area
Grey
Dark
White
Grey
Dark
White
Grey
Dark
Grey
Dark
Al
7.7
27.7
0.3
8.5
11.1
0.1
8.2
14.9
7.5
10.8
O
61.3
58.5
60.1
61.3
55.0
58.1
61.3
55.4
61.1
56.6
S
0
0.1
0.1
0.1
0
0
0.1
0.1
0
0
C
2.4
2.7
1.9
3.1
3.0
1.9
2.9
2.5
3.4
2.8
Cr
2.0
4.9
1.5
2.7
16.9
1.2
1.8
13.5
2.2
11.7
Si
14.4
0.1
13.4
13.0
0.0
14.0
13.4
0.1
12.6
3.8
Mn
4.1
2.1
2.2
4.0
12.6
3.4
3.1
12.2
3.2
9.2
Ce
6.8
2.9
19.1
6.1
0.1
19.4
8.0
0.1
7.4
1.8
Phase
(AlCeSi)2O3
AlO2
(CeSi)O2
(AlCeSi)2O3
(AlCrMn)2O3
(CeSi)O2
(AlCeSi)2O3
(AlCrMn)2O3
(AlCeSi)2O3
(AlCrMn)2O3
46
Figure 4.18: Image of particles in block 7 (0.052% Ce), composition given in table
4.5
Table 4.5: Composition of particles in figure 4.21, and likely phase, all values in
atomic%.
#
1
2
2
3
3
4
4
4
5
5
5
Area
White
White
Dark
White
Dark
White
Grey
Dark
White
Grey
Dark
Al
0.4
0.0
29.3
0.1
14.1
0.1
10.5
29.4
0.1
8.7
29.4
O
60.2
58.2
57.8
58.3
52.8
58.5
60.0
57.5
58.8
61.0
57.6
S
0
0.1
0.1
0
0.1
0
0
0.1
0.1
0
0
C
1.3
1.6
2.4
1.5
2.7
1.5
2.0
2.4
1.6
2.2
2.5
Cr
0.4
0.9
3.9
1.1
15.4
1.0
2.1
4.2
1.0
2.0
4.1
Si
13.7
14.2
0.1
14.3
0.1
14.1
12.9
0.1
13.8
14.2
0.1
Mn
2.0
3.3
2.4
3.5
11.2
3.2
4.3
2.4
3.2
4.1
2.3
Ce
19.4
19.8
3.1
18.9
0.5
19.4
6.5
3.0
18.9
6.3
2.9
Phase
(CeSi)O2
(CeSi)O2
AlO2
(CeSi)O2
(AlCrMn)2O3
(CeSi)O2
(AlCeSi)2O3
AlO2
(CeSi)O2
(AlCeSi)2O3
AlO2
47
Figure 4.19 Image of particles in block 8 (0.073% Ce), composition given in table
4.6
Table 4.6: Composition of particles in figure 4.22 and likely phase, all values in
atomic%.
#
1
1
1
2
2
2
3
3
3
Area
White
Grey
Dark
White
Grey
Dark
White
Grey
Dark
Al
0.1
8.2
12.0
0.0
7.8
11.6
1.2
7.6
12.8
O
55.4
59.1
55.7
57.8
58.7
56.6
55.9
59.7
55.2
S
0
0
0
0.1
0.1
0
0.1
0
0.1
C
6.0
5.5
4.7
7.5
7.1
5.8
5.6
7.3
5.8
Cr
1.0
1.6
11.3
0.7
1.6
10.0
2.0
1.4
12.3
Si
13.4
13.5
1.2
12.5
13.4
2.0
8.0
12.9
0.2
Mn
3.3
3.9
11.2
2.6
3.7
11.7
9.7
3.7
11.8
Ce
18.3
6.5
2.4
17.0
6.5
0.5
15.7
6.3
0.7
Phase
(CeSi)O2
(AlCeSi)2O3
(AlCrMn)2O3
(CeSi)O2
(AlCeSi)2O3
(AlCrMn)2O3
(CeSi)O2
(AlCeSi)2O3
(AlCrMn)2O3
48
Figure 4.20: Image of particles in block 9 (reference), composition given in table 4.7
Table 4.7: Composition of particles in figure 4.23 and likely phase, all values in
atomic%.
#
1
2
3
4
5
Al
13.5
12.2
11.9
12.7
11.6
O
55.1
54.6
54.7
54.6
54.7
S
0
0.1
0.1
0.1
0
C
3.5
3.5
3.5
3.5
3.5
Cr
14.6
15.7
16.1
15.6
16.6
Mn
12.4
12.4
12.4
12.5
12.4
Phase
(CrMnAl)2O3
(CrMnAl)2O3
(CrMnAl)2O3
(CrMnAl)2O3
(CrMnAl)2O3
49
Figure 4.21: Image of block 10 (0.035% Ce) showing little cerium content in found
particles.
50
Figure 4.22: Image of block 11 (0.067% Ce), composition given in table 4.7.
Table 4.7: Composition of particles in figure 4.25, and likely phase, all values in
atomic%.
#
1
1
2
2
2
3
3
4
4
4
5
5
Area
Grey
Dark
White
Grey
Dark
White
Grey
White
Grey
Dark
White
Grey
Al
7.8
11.5
0.2
8.2
15.3
0.2
7.8
0.1
8.0
12.1
0.1
7.7
O
60.2
54.8
59.5
60.0
55.0
61.5
60.9
61.2
60.6
55.1
60.5
60.5
S
0.1
0.1
0
0
0.1
0
0.1
0
0.1
0
0
0.1
C
4.0
3.4
2.5
3.9
3.6
3.7
3.4
3.9
3.8
3.4
3.4
4.7
Cr
3.2
16.4
1.1
2.1
13.9
0.1
1.9
0.4
2.0
15.7
0.4
1.8
Si
12.8
16.4
13.0
14.0
0.1
15.2
14.3
14.9
14.6
0.1
14.6
14.2
Mn
4.5
12.5
3.2
3.7
10.9
0.2
4.1
0.2
3.9
12.4
1.4
3.7
Ce
6.0
0.1
18.6
6.2
0.2
18.1
6.6
17.6
6.2
0.1
18.3
6.4
Phase
(AlCeSi)2O3
(AlCrMn)2O3
(CeSi)O2
(AlCeSi)2O3
(AlCrMn)2O3
(CeSi)O2
(AlCeSi)2O3
(CeSi)O2
(AlCeSi)2O3
(AlCrMn)2O3
(CeSi)O2
(AlCeSi)2O3
51
Figure 4.23: Image of particles from block 12 (0.078% Ce), showing particles with
different phases.
52
Figure 4.24: Electron backscatter image of particle from block 8, showing all three
phases seen in other particles, bright, grey and dark.
4.5 Mechanical testing
The results from the tensile tests are given in figures 4.25-4.27, where the
two different cast temperatures is given as green and blue for
temperatures 1525ºC and 1540 ºC respectively. Both series show a linear
increase in both yield and ultimate tensile strength, the higher cast
temperature shows a stronger effect of EGR addition, due to the
references is weaker at the higher casting temperature. The series cast at
1540 increases 7.2% and 3.3% in yield and ultimate tensile strength
respectively with most Ce added. The lower casting temperature have its
peak in ultimate tensile strength with 0.034% Ce (0.075% added) but
nearly the same as the block with most cerium. The yield strength result
of lower casting temperature is somewhat different from ultimate tensile
strength: The results here shows an increase from 533 MPa for the
reference to 542 MPa with most cerium (0.064% Ce), while the two
53
blocks with lower Ce content shows slightly decreased yield compared to
the reference. The elongation is increased from 38%, for the reference, to
41% for all three blocks with added cerium. The impact toughness test
shows an overall decrease for the blocks with added cerium, from 139 J
for the reference to 125 J with 0.036% Ce which was the lowest.
The blocks cast at 1540ºC the elongation does increase a small amount,
but the blocks with most cerium have lower elongation than the
references. The impact toughness test shows a linear reduction of
toughness with increasing cerium content. The lowest value is the block
with the lowest cerium content, block 10; 0.035% Ce. However, all
results is clearly above the requirements for cast duplex according to the
standard. Regarding the reducing of impact toughness which decreases, it
is still over 2 times higher than the required energy absorption.
Ultimate tensile strength
Ultimate tensile strength [MPa]
800
790
780
770
760
750
740
730
720
710
700
0
0.02
0.04
0.06
0.08
0.1
Ce content [Wt%]
Figure 4.25: Ultimate tensile strength and cerium content, green show series cast at
1525 ºC, blue cast at 1540ºC. Red line show required strength according to
standard. [20]
54
Yield strength
Yield strength [MPa]
590
570
550
530
510
490
470
450
0
0.02
0.04
0.06
0.08
0.1
Ce content [Wt%]
Figure 4.26: Yield strength versus cerium content, green show series cast at 1525
ºC, blue cast at 1540ºC. Red line show required strength according to standard.[20]
Elongation
44
Elongation [%]
42
40
38
36
34
32
30
0
0.02
0.04
0.06
0.08
0.1
Ce content [Wt%]
Figure 4.27: Elongation versus cerium content, green show series cast at 1525 ºC,
blue cast at 1540ºC. Requirement from standard is 18%.[20]
55
Impact toughness
160
Absorbed enery [J]
140
120
100
80
60
40
20
0
0
0.02
0.04
0.06
0.08
0.1
Ce content [Wt%]
Figure 4.28: Impact toughness at -46 ºC versus cerium content, green show series
cast at 1525 ºC, blue cast at 1540ºC. Red line show required strength according to
standard.[20]
4.6 SEM
The fracture surfaces from the impact toughness test were examined by
EDS in a SEM. Exact composition is not obtained by EDS, but the peaks
of graphs give an indication of the amount of elements present. Figures
4.29-4.32 shows a correlation with the composition given in table 4.4.
Overall the particle found in the dimples in the fracture surfaces
resembles the particles scanned with EPMA, which is mainly ceriumoxide and aluminium-oxides.
56
Figure 4.29: Example of EDS scan from block 6, image acquired with secondary
electrons, analysis done with backscatter electrons. Numbers indicate particles
scanned with results in figures 4.28 – 4.31.
Figure 4.30 (a), (b): EDS scan of particles 1 (a) and 2 (b) from figure 4.28, showing
elements present.
57
Figure 4.31 (a), (b): EDS scan of particles 3 (a) and 4 (b) from figure 4.28, showing
elements present.
Figure 4.32 (a), (b): EDS scan of particles 5 (a) and 6 (b) from figure 4.28, showing
elements present.
58
Figure 4.33: EDS scan of particle 7 from figure 4.29, showing elements present.
59
5. Discussion
5.1 Experiment
The execution of this experiment can influence the results, the yield of
cerium added show this. Figure 4.1 shows the relation between added
cerium and actual content after casting. The overall yield of 72% is higher
than seen in previous trials. The reason for this can be caused by the short
time between addition to the ladle and casting trapping particles that
would float to the surface if given more time. The large oxygen content is
likely to have influenced the yield of cerium, normally the amount of
oxygen is lower than the cerium content, but in this case the ratio is nearly
1:1 between cerium and oxygen. The large content of oxygen is partly due
to unsuccessful deoxidizing, CaSi was added to the furnace for
deoxidizing but this appears to have been insufficient amount, but worked
somewhat since no Ca is found in chemical analysis. The chemical
analysis found Ce in small amounts in all references; this is likely traces
in the ladle after previous cast, or traces from production at Scana since
block 1 which was the first to be cast also contains Ce. The holding time
of each cast in the ladle was generally short, the time from EGR addition
to the ladle was as short as 50 s, which does not give much time for
cerium to react and the oxides to distribute through the liquid. As seen
from the theory, a holding time of up to 5 minutes is beneficial to the
number of inclusions and will give more time for the large inclusions to
float to the slag. Mixing of the ladle may be beneficial; inert gas is often
used and may help remove the largest inclusions seen in the EPMA
analysis.
5.2 As cast structure
The macroscopic examination of the blocks, figure 4.2-4.9, reveals a
strong effect of refining the cast structure by adding EGR. The length of
the columnar zone was reduced from 22 mm (reference) to 0 for the
blocks with the biggest addition of EGR. The grain size of the grains in
the equiaxed zone in the middle of the block is also reduced. These results
are contributed to the nucleation potential of the cerium particles obtained
by adding EGR to the ladle during casting. This in return will affect the
segregation of steel, making it more homogeneous and therefore making
60
the heat treatment more effective in homogenizing the steel. A more
homogen steel can contribute to the increased mechanical properties of
the blocks with 0.075% Ce added; the macro etch reveals a columnar
zone present but the mechanical tests show almost as good results as the
blocks which does not have a columnar zone present at all.
5.3 Light microscopic examination
The light microscopic examination is somewhat insufficient, due to the
lack of quantifying the grain size. It is a difficult task because of the dual
phase composition of the steel. It was suggested that the best technique
for determining a grain size is using EBSD, where each grain will get
different colour due to lattice rotation. The ferrite amount was measured
with light microscope and software at Scana steel. The result is given in
appendix 2, but does not seem to vary significantly with cerium content.
5.4 Cerium particles in the metal
The particles found and identified by EPMA examination fits with the
theory of the strong oxidizing effect of cerium. The type of particles does
not seem to vary with larger Ce addition. Almost all particles found with
cerium content also consist of phases without cerium, while the particles
found in the references contained only a single phase. The particles
analysed did not contain any sulphur which is the most beneficial cerium
inclusion regarding undercooling and lattice misfit.
5.5 Mechanical tests
The tensile tests show a linear increase in strength for increased cerium
content; this is contributed to the refined casting structure. The
mechanical test also show a decreasing impact toughness, this effect are
most likely due to the increased particle content, which also correlates to
the theory [15]. The results show different effect of EGR to elongation;
for the series cast at 1525ºC the elongation increases. The series cast
1540ºC the elongation is increasing except for the 2 blocks with the
largest EGR addition, which causes this series to show no influence to the
elongation overall.
The casting temperature does not seem to affect the mechanical properties
of the grain refined blocks. The reference for the series cast at 1525°C is
61
stronger than its counterpart cast at 1540°C making the increased
mechanical properties obtained by adding EGR smaller for the lower
casting temperature. The only mechanical difference between the series is
elongation, where blocks cast at 1525°C shows a linear increase with
increased Ce content, while blocks cast at 1540°C show no overall effect
on elongation.
During machining blocks 3 and 4 were accidentally cut the wrong way,
causing the tensile bars from these blocks to be cut lower in the block.
This deviation is not considered to have influenced the result significantly
since figure 4.2 and 4.3 shows the same trend as for the series cast at
1540ºC, figure 4.6 and 4.7. From block 6 one of the tensile bars was label
“abnormal” from the test lab, resulting in too high ultimate tensile
strength. If this bar is included; figure 4.6 will look like figure 5.1, which
shows the abnormal bars influence on the graph and does not follow the
linear trend in figure 4.6.
Tensile strength cast at 1540ºC
800
Tensile strength [MPa]
790
780
770
760
750
740
730
720
710
700
0
0.02
0.04
0.06
0.08
0.1
Ce content [Wt%]
Figure 5.1: Ultimate tensile strength versus cerium content, with included abnormal
specimen.
62
5.6 Effect of casting temperature
The two different casting temperatures used in this experiment does not
seem to affect EGR’s potential for grain refining. Neither the as cast
structure, mechanical properties or particles present varies significantly
with change of casting temperature. The only notable difference between
the 2 series were increased strength in reference cast at 1525°C compared
to the references cast at 1540°C.
Mikrostrukturen..
Tid brukt på støpinga, boble argon
Fraværet av svovel
63
6. Conclusion
The conclusions based on this work are the following:
By adding EGR to liquid Super Duplex steel it will react with available
oxygen to form complex particles containing (CeSi)O2, (AlCeSi)2O3,
(AlCrMn)2O3
With the correct amount of EGR added to Super Duplex the as cast
structure is greatly refined, ultimately eliminating the columnar zone
completely.
The refining of Super Duplex with EGR can result in an increase of 7%
and 3% for yield and ultimate tensile strength respectively.
Reducing the cast temperature from 1540 to 1525 for Super Duplex steel
has no effect of EGRs potential for grain refining.
64
Proposal for further work
From the positive results in this experiment by adding EGR to Super
Duplex steel, there is possible to enhance the mechanical properties
considerably. The large oxygen content in the steel is likely to be reduced
by performing the casting by refining in an AOD converter where process
control is better. Also an experiment where the holding time is controlled
may be beneficial to explore optimal conditions for grain refining with
EGR.
65
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Gunn, R., Duplex Stainless Steels. 1997: Abington Publishing.
Flemings, M.C., Solidification Processing. 1974: McGraw Hill
Inc.
Porter, D.A., K.E. Easterling, and M.Y. Sherif, Phase
Transformations in Metals and Alloys. 2009: CRC Press.
Nilsson, M.R., Kornförfining av stål, del 2. 2007, Swecast.
Lidegran, P., Microprobe investigation of A17061 sample received
from Siepmann. 2008.
Verhoeven, J.D., Fundamentals of Physical Metallurgy. 1975:
John Wiley & Sons, Inc.
Dieter, G.E., Mechanical Metallurgy. 1988: McGraw-Hill Book
Company.
Engh, T.A., Principals of Metal Refining. 1992: Oxford Science
Publications.
Akselsen, O.M., C.v.d. Eijk, and Ø. Grong, Grain refinement of
steel castings, O.M. Akselsen, Editor. 1999, SINTEF.
MuxingGUO and H. SUITO, Influence of Dissolved Cerium and
Primary Inclusion Particles Ce203 and CeS on Solidification
Behavior of Fe-0.20 mass% C-0.02 mass%PAlloy. ISIJ
International, 1999. 39(7): p. 722-729.
Haakonsen, F., Optimizing of Strømhard austenitic manganese
steel., in Material Science and Engineering. 2009, Norwegian
University of Science and Technology.
Dahle, E.S., EGR additions to austenitic manganese
steel. SINTEF F18030, 2011.
13.
Eijk, C.v.d., et al., Grain Refinement of Fully Austenitic Stainless
Steels
Using a Fe-Cr-Si-Ce Master Alloy, in 59th Electric Furnace and 19th
Process Technology Conferences,. 2001: Phoenix, AZ USA.
14.
Eijk, C.v.d. and A. Slettan, NYKURS-Project: Grain Refinement
of Stainless Steels by Ce Additions. STF24 F01215, 2001.
15.
Thornton, P.A., The Influence of Nonmetallic Inclusions on the
Mechanical Properties of Steel: A Review. Journal of Material
Science, 1971. 6: p. 347-356.
16.
Guo, M. and H. Suito, Dispersion of Primary inclusions of Ce2O3
and CeS in Fe-O.20 mass% C-0.02 mass%P Alloy. ISIJ
International, 1999. 397(7): p. 678-686.
66
17.
18.
19.
20.
Blair, M., Cast Stainless Steels, Properties and Selection: Irons,
Steels, and High-Performance Alloys. ASM Handbook. Vol. 1.
1990: ASM International.
Solberg, J.K., Teknologiske metaller og legeringer. 2009: Tapir.
Park, Y.-H. and Z.-H. Lee, The effect of nitrogen and heat
treatment on the microstructure and tensile properties of 25Cr–
7Ni–1.5Mo–3W–xN duplex stainless steel castings. Material
Science And Engineer, 2001. A297: p. 78-84.
Norsok, Material data sheets for piping M630. 2004.
67
Appendix
Hardness
240
Hardness [HB]
235
230
225
220
215
0
0,02
0,04
0,06
0,08
0,1
Ce content [Wt%]
Figure: Hardness versus Ce content.
68
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement