three-dimensional investigations of inclusions in ferroalloys

www.steel-research.de

FULL

PA

Three‐Dimensional Investigations of Inclusionsin Ferroalloys

PER

Yanyan Bi,� Andrey Karasev, and Pär G. Jönsson

As the requirements on material properties increase, there has been a demand on anadditional knowledge on the effect of impurities in the ferroalloys on the properties. Thus,the number, morphology, size, and composition of inclusions in four different ferroalloys(FeTi, FeNb, FeSi, and SiMn) were investigated. This was done in three dimensions (3D) byusing scanning electron microscopy in combination with energy dispersive spectroscopyafter electrolytic extraction of the ferroalloy samples. The non‐metallic and metallicinclusions were successfully analyzed on the surface of film filter. Thereafter, the particlesize distribution was plotted for most of the non‐metallic inclusions. The non‐metallicinclusions were found to be REM oxides in FeTi, FeSi, and SiMn, Al2O3, Ti–Nb–S–O oxides inFeNb and silicon oxides in SiMn. Moreover, the intermetallic inclusions were found to be aTi–Fe phase in FeTi, Ca–Si, and Fe–Si–Ti phases in FeSi and a Mn–Si phase in SiMn. Inaddition, the almost pure single metallic phases were found to be Ti in FeTi, Nb in FeNb, andSi in FeSi.

1. Introduction

Ferroalloys are commonly used in the steel industry to

alloy or deoxidize the steel during the secondary steel-

making process before casting. However, the addition of

ferroalloys may cause a supply of deleterious impurities to

the liquid steel. In some cases, these impurities are clearly

related to the raw material and the way of producing a

ferroalloy and are more or less unavoidable.[1] As the

requirements on material properties increase, the effect of

impurities in ferroalloys on the steelmaking process has

been studied more deeply. Jonsson et al.[2] concluded the

future demands on ferroalloys from the view point of

customers. The first kind of ferroalloys is the high purity

ferroalloys that are used for late additions during ladle

refining. The second is the less refined grades of lower

cost, which should be added earlier in the steelmaking

production line. Thus, in order to meet the forthcoming

demands, ferroalloys should be characterized with respect

to the number, morphology, size, and composition of

inclusions.

Some publications about the inclusions in different

kinds of ferroalloys as well as characterization method are

listed in Table 1. SiO2, Al2O3, and some intermetallic phase

[�] Y. Bi, A. Karasev, P. G. JonssonKTH-Royal Institute of Technology, Brinellvagen 23, Stockholm, SE-10044, SwedenEmail: [email protected]

DOI: 10.1002/srin.201300157

� 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

were observed in FeTi alloys.[1,3–6] Inclusions in FeSi have a

direct link to the Al and Ca contents in the ferroalloys.[3,7–9]

Similarly, the types of inclusions in FeMn depend on the

silicon[10,11] and C[1,14] contents in the ferroalloys. Also,

the inclusions in different grades of FeSi alloys have a

connection to the steel cleanliness. However, inclusions in

FeMn only have a temporary influence on the inclusions in

liquid steel according to laboratory and plant trials.[12,13]

Finally, inclusions in FeMo, FeNb, FeP alloys were also

studied by some researchers.[1,3,14]

However, the above mentioned studies have been

concentrated on two dimensional (2D) investigations,

except for the study by Pande et al.[1,14] They used acids

to dissolve steel and to study the impurities in three

dimensions (3D). However, some impurities might dis-

solve during the acid extraction. Moreover, very few

researchers tried to provide information about the particle

size distribution of inclusions in ferroalloys.[2] Therefore,

the present study is concentrated on three dimensional

investigations of inclusion characteristics (such as com-

position, size, number, and morphology) in the following

four ferroalloys: FeTi, FeNb, FeSi, and SiMn. The

electrolytic extraction (EE) method was used in this study.

This dissolution technique leads to less dissolution of

inclusions in comparison to the acid method.[15] More

specifically, the EE method is recommended for the

analyses of REM oxides, Al2O3, and SiO2 inclusions which

are also observed in this study. After separating the

particles on the surface of film filter, the particle size

distribution was determined for most of the major non-

metallic inclusion types found in the samples.

steel research int. 85 (2014) No. 4 659

Ferroalloya) Method Inclusion Ref.

Composition Shape Size

[mm]

Number Other

FeTi (35) Acid (3D) Si/SiO2,

Al–Ti–O

Irregular 1–50 9–9.5% Titanium oxide is not reduced

completely by the aluminum

during the production.

Pande

et al.[1]SEM

EDS

FeTi (35) SEM Iron oxide

magnesium oxide

Irregular – – – Tiekink

et al.[3]AIA

FeTi (35) SEM Al2O3, TiN,

Al4TiO8

– �20 PSD Inclusions in FeTi35

increased the product

rejection. FeTi70 is cleaner

than FeTi35.

Kaushik

et al.[5]

FeTi (70) EDS Pande

et al.[6]

FeTi (70) Acid (3D) Si/SiO2,

Al–Ti–O

Faceted 1–20 1–1.5% Inclusions in the matrix are

in good agreement with the

extracted inclusions.

Pande

et al.[1]

SEM Al2O3,

Fe–Ti–Al2O3

– – – – Gasik

et al.[4]EDS

FeNb (65) Acid (3D) Not any The microstructure of FeNb

reveals two phases with no

apparent inclusions.

Pande

et al.[1]SEM

EDS

FeSi

(65, 75)

FGA SiO2, Al2O3,

(Al,Ca,Si)xOy,

(Al,Mg)xOy

– – PSD FeSi with lower Al content

results in a low oxygen

content and a small quantity

of Al2O3 inclusions.

Grigorovich

et al.[8]

FeSi (75) QTM FeSi2.5, Si2Ca,

Si2Al2Ca, FeSi2Ti,

Fe4Si8Al6Ca

– 2.8–22.4 PSD Impurities are mostly

intermetallic phases and

oxide inclusions are very rare.

The grade of ferrosilicon has

a direct impact on the

inclusions in the liquid steel.

Wijk and

Brabie[7]SEM

EMP

FeSi (75) SEM – – – – High Si phase with Al, Cu and

Ti, high FeSi phase with V and

Fe. Some phases contain Ce.

Tiekink

et al.[3]AIA

FeSi (75) SEM MgO–Al2O3,

SiO2–Al2O3–MnO–CaO

Spherical 1–3 – Al in FeSi enhace the

formation of spinel inclusion.

However, Ca prevents it.

Mizuno

et al.[9]EDS Irregular

HC FeMn

(75)

Acid (3D) C Powder – 0.3–0.5% There is no inclusion due to

the high C content.

Pande

et al.[1,14]

LC FeMn

(80)

Acid (3D) C, Si/SiO2 Powder – 0.2–0.25% –

HC SEM MnS, SiMn oxide,

Mn oxide

– – – The FeMn alloy contains

distinguishable carbide and

nitride phases.

Tiekink

et al.[3]FeMn (80) AIA

Table 1. Continued


660 steel research int. 85 (2014) No. 4 � 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

FULL

PAPER

Ferroalloya) Method Inclusion Ref.

Composition Shape Size

[mm]

Number Other

LC&MC SEM MnO, MnS,

MnO–SiO2–MnS

Dendritic

Rhombic

3–180 PSD The amount of inclusions is

inversely proportional to the

carbon content in FeMn. Low

oxygen FeMn has lower

inclusion mean diameters.

Sjoqvist

et al.[10,12,13]

FeMn (80) PC MIC Spherical

FeMo (70) Acid (3D) Si/SiO2, Al2O3 Spherical 10–50 0.5–0.9% Inclusions in the matrix are

in good agreement with the

extracted inclusions.

Pande

et al.[1]

SEM CaO–SiO2–Al2O3,

SiO2–Al2O3

– – – – Gasik

et al.[4]EDS

FeP (30) Acid (3D) (Fe,P,Mn,Ti)O Angular 10–80 0.3–0.4% The distribution of Ca, Mn,

and Ti is inhomogeneous and

the best addition sequence is

obtained.

Pande

et al.[1,6]SEM

EDS

FeP (35) SEM – – – – High P phase with Si and low

P phase with Mn, Ti, and Cu.

Tiekink

et al.[3]AIA

HC, high carbon; LC, low carbon; AIA, automated inclusions analysis; PSD, particle size distribution; FGA, fractional gas analysis; QTM,Image Analyzer Qantimet; EMP, electron microprobe.a)(), the content of alloying element.

Table 1. Inclusions in different ferroalloys.


FULL

PAPER

2. Experimental

The investigations of inclusions in this study were

carried out by using the following commercially

obtained ferroalloys: FeTi, FeNb, FeSi, and SiMn. The

typical chemical compositions of these ferroalloys are

listed in Table 2. The specimens were cut from a

ferroalloy lump to small pieces (15� 10� 3mm3) before

EE.

In this study, EE was applied for 3D investigations of

inclusion characteristics in the ferroalloys. The EE of

ferroalloys was carried out using a 10% AA (10 v/v%

acetylacetone – 1w/v% tetramethylammonium chloride –

methanol) electrolyte. The current density was set to 30–

40mA cm�2 during the EE. The weight of the dissolved

Ferroalloys C N S P Fe Si

FeTi 0.18 0.23 0.008 0.012 22.426 0.21

FeNb 0.15 – 0.05 0.06 28.296 2.48

FeSi 0.089 – 0.006 0.023 22.386 75.975

SiMn 0.053 – 0.0059 0.046 8.771 29.219

Table 2. Composition of different ferroalloys (wt%).


ferroalloys during EE was 0.08–0.14 g. After extraction, the

solution containing inclusions was filtrated through a

polycarbonate (PC) membrane film filter with an open

pore size of 1 or 3mm. The 3mmfilmfilter was only used to

analyze the large size (>3mm) inclusions in FeTi. The

extracted inclusions were investigated in 3D on a surface

of film filters by using scanning electron microscopy

(SEM). The size of the inclusions dV, is the diameter of

a spherical inclusion. The equivalent size of a non-

spherical inclusions and clusters was calculated from

Equation 1:

dV ¼ Lmax þWmax

2ðfor non� spherical inclusions and clustersÞ

ð1Þ

Al Ca Mn Ti Cr Ni Nb

3.14 – 0.05 71.56 – – –

1 – – 0.82 – – 66.23

1.166 0.224 0.032 0.083 0.0058 0.0039 –

– – 61.471 0.2 0.03 0.03 –



FULL

PAPER
where Lmax and Wmax are the maximum length and width
of the inclusion, respectively.

The number of inclusions per unit volume (NV) was

calculated as follows:

NV ¼ n � Af

Aobs� rmW dis

ð2Þ

where n is the number of inclusions in the appropriate size

interval, Af is the area of the film filter (1396mm2), Aobs is

the total observed area, rm is the density of the ferroalloy

matrix and Wdis is the dissolved weight of the ferroalloy

during extraction.

The composition of inclusions was determined by

energy dispersive spectroscopy (EDS). The total ob-

served area of film filter for different samples was

varied from 0.22 to 2.23mm2. Here, it should be

pointed out that there might some undissolved ferro-

alloy matrix pieces on the surface of the film filter

after filtration. However, these can easily be distin-

guished from the inclusions based on composition and

morphology.

Type Type A Type B

Typical photo

Size range (mm) 6–25 1–8

Composition Ti–Fe Ti–Fe

Percentage (%) 9 75

Table 3. Classification of inclusions in FeTi alloys.

Type Ti Fe Al Si Ni

Type A 45–76 21–49 0–1.5 0–1.3 1–5

Average 61 36 0.9 0.2 2

Type B 45–55 43–47 1–4 0–0.2 1–2

Average 48 45 1.7 0.05 1

Type C 89–97 0–4 0.5–3 – –

Average 92 1 1 – –

Type D – 0–2 0.6–2.0 6–8 –

Average – 0.6 1.4 7 –

Table 4. Composition of different type of inclusion in FeTi alloys (w

662 steel research int. 85 (2014) No. 4

3. Results

3.1. Inclusions in FeTi

The typical photos, size ranges, composition, and per-

centage of the inclusions observed after EE of FeTi alloy are

shown in Table 3. A more detailed composition informa-

tion is listed in Table 4. There are two types of intermetallic

phases in ferrotitanium, a faceted Type A and a flower-like

Type B. The size range of the Type A inclusions is much

larger than that for Type B inclusions. However, the

percentage of Type B inclusions is eight times larger than

the Type A inclusions. The Ti, Fe contents, and the ratios

are different in these two types of intermetallic phases. The

element N is present in both types of intermetallic phases,

as can be seen from the elemental mapping in Figure 1.

However, the light element, such as C and N cannot be

analyzed in an accurate manner by EDS. The average

value of N in these two types of inclusions is 0.008 and

0.1mass%, respectively. Therefore, it is not possible to say

that the titanium carbides and nitrides, which has high

melting points, exist or not in this case. The Type C

Type C Type D

3–15 1–21

Ti–Fe–Al–O REM-Si–Cr–Al–O

10 6

Ce La Pr Nd O N

– – – – 0–1 0–0.5

– – – – 0.7 0.008

– – – – 0–3 0–0.6

– – – – 1.7 0.1

– – – – 3–10 0–1.7

– – – – 5 0.3

27–35 28–32 2–3 6–7 17–22 –

30 30 3 7 20 –

t%).


Figure 1. Elemental mappings of different inclusion types in FeTi alloys.

0 2 4 6 8 10 12 14 160

1000

2000

3000

4000

Nv

(mm

-3)

dv

(µm)

Figure 2. Particle size distribution of Type D inclusions in FeTialloys.


FULL

PAPER

inclusions are almost pure Ti, with small amounts of O, Al,

and Fe. There is also some N in this type of inclusions, as

can be seen fromFigure 1. However, the distribution of N is

not homogenous. Also, the Type D inclusions mostly

consist of REM (Ce, La, Pr, Nd) oxides. These oxides easily

form clusters, which may cause nozzle clogging.[16–18]

According to the literature review, no REM oxides have

been observed in FeTi. The existence of the REM oxides

might be due to the raw material. The particle size

distribution of this type of inclusion is shown in Figure 2.

According to the results reported by Pande et al.,[1,6] the

quantity of inclusions in a FeTi35 alloy ismuch higher than

that in a FeTi70 alloy. Moreover, the size range of

inclusions in a FeTi35 alloy is wider because of the

different manufacturing routes for these alloys. This is in

good agreement with the result reported by Kaushik

et al.,[5] who showed that the sliver index in an FeTi33 alloy

is higher than in an FeTi70 alloy. In addition, Al2O3 and

SiO2 were observed in a FeTi70 alloy.[1,5] However, REM

oxides were detected instead of Al2O3 and SiO2 in the

FeTi70 alloy in this study. The reason for this might be that

the REMmetal has a higher affinity to oxygen compared to

aluminum and silicon.

According to the Fe–Ti–Al phase diagram,[19] the

melting points of Type A, Type B, and Type C inclusions

are about 1200, 1300, 16008C, respectively. The melting

points are lower or approximately equal to a steel

temperature of 16008C. Therefore, it is assumed that these


intermetallic phases will dissolve after the addition to the

liquid steel.

3.2. Inclusions in FeNb

The typical inclusions in a FeNb alloy are shown in Table 5

and the detailed composition information is listed in

Table 6. As can be seen, Type A inclusions aremostlymade


Type Type A (I) Type A (II) Type B Type C

Typical photo

Size range (mm) 2–12 5–27 1–14 2–21

Composition Al–O Al–O Ti–Nb–S–O Nb–Ti–O

Percentage (%) 20 4 17 59

Table 5. Classification of inclusions in FeNb alloys.

Type Fe Nb Al Ti Si O S

Type A 0–0.9 0–2 42–54 – – 45–56 –

Average 0.1 0.4 50 – – 49 –

Type B – 10–26 0–0.7 28–53 0–1.5 8–23 21–32

Average – 15 0.3 44 0.5 12 26

Type C 0–4 80–94 0–2 0–3 0–0.4 2–15 –

Average 0.7 91 0.5 1.2 0.2 7 –

Table 6. Composition of different type of inclusion in FeNb alloys (wt%).


FULL

PAPER

up of pure aluminum oxides. Also, inclusions are present

as single inclusions (Type A (I)) as well as clusters (Type A

(II)). The presence of Al2O3 is linked to the aluminothermic

reduction of FeNb. It is well known that Al2O3 inclusions

can significantly affect the mechanical and fatigue

properties and also cause nozzle clogging problems during

casting.[20] Type B inclusions are Ti–Nb–S–O inclusions.

The sulfur content in a FeNb alloy is 0.05mass% (Table 2).

This is the highest value among the ferroalloys that have

Figure 3. Elemental mapping of Type B inclusions in FeNb alloys.


been analyzed in this study. According to the elemental

mapping of this type of inclusion in Figure 3, the

distribution of Ti, Nb, S, and O is homogenous, which

means that only one phase exists in this type of inclusions.

However, up to now, only the Ti–Nb system[21] and the

Ti–S system[22] were studied. Therefore, the physical and

chemical characteristic of the Ti–Nb–S–O system is not

known. However, nomatter if they will dissolve or not after

the addition to the liquid steel, they will definitely be


0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 300

1000

2000

3000

4000

5000

6000

7000

Type A (I) Type A (II) Type B

Nv

(mm

-3)

dv

(µm)

Figure 4. Particle size distribution of Type A and Type B inclusionsin FeTi alloys.


FULL

PAPER

deleterious to the cleanliness of the steel due to the

presence of S and O. The particle size distribution of

these two type inclusions is shown in Figure 4. The Type C

inclusion is almost a pure Nb phase. The melting

point of the Type C phase is 24008C[23] if only the Nb

and Ti are considered. In addition, the Type C inclusions

Type Type A Type B

Typical photo

Size range (mm) 2–20 5–9

Composition REM-Si–Fe–Ti–O Ca–Si–Al–N

Percentage (%) 36 4

Table 7. Classification of inclusions in FeSi alloys.

Type Si Fe Ti Al Ca

Type A 3–10 0–8 0–5 0–1 –

Average 6 2 2 0.4 –

Type B 35–49 0–3 – 2–6 40–57

Average 44 0.6 – 3 45

Type C 30–35 34–41 24–30 1–3 –

Average 33 36 27 2 –

Type D 80–94 0–2 – 0–2 –

Average 89 0.4 – 0.3 –

Table 8. Composition of different type of inclusion in FeSi alloys (w


represent the majority (59%) of the inclusions in a FeNb

alloy. This is followed by the Type A (I) and Type B

inclusion. Finally, a Type A (II) inclusion is present, which

is an Al2O3 cluster. The size range of the Type A (II) and

Type C inclusions is wider than that for Type A (I) and

Type B inclusions.

No inclusions could be observed in a FeNb alloy in

Pande’s study after dissolution when using strong acids.[1]

However, the characteristics of inclusions in FeNb, such

as the number, morphology, size, and composition can

accurately be investigated on the surface of film filter after

EE. This implies that the EE method is more suitable than

the acids dissolution method for extraction of inclusions

present in FeNb.

3.3. Inclusions in FeSi

Typical inclusions and the information of them in FeSi are

shown in Table 7 and 8. Type A inclusions are made up by

REM oxides with some amounts of Si, Fe, and Ti. Ce oxides

have also been observed by other researchers in FeSi

alloys.[3] The particle size distribution of this type inclusion

is shown in Figure 5. Type B inclusions are CaSi

intermetallic inclusions that contain small amounts of

Type C Type D

2–10 1–26

i–O Fe–Si–Ti–Al–O Si–O

20 40

Ni Mn Ce La Pr Nd O

– 0–1 26–40 21–33 1–3 3–7 15–26

– 0.3 34 27 2 4 20

2–5 – – – – – 4–9

2 – – – – – 5

– 0–2 – – – – 0–4

– 0.6 – – – – 1

– – – – – – 5–20

– – – – – – 10

t%).


0 2 4 6 8 10 12 14 16 18 20 220

1000

2000

3000

4000

5000

6000

7000N

v (m

m-3

)

dv

(µm)

Figure 5. Particle size distribution of Type A inclusions in FeSialloys.


FULL

PAPER

Al and Ni. The melting point of the CaSi intermetallic

phase is about 13008C[24] if only the Ca and Si contents are

considered. Type C inclusions represent a Fe–Si–Ti–Al

intermetallic phase. The phase diagram of Fe–Si–Ti–Al

system calculated by the Therm-Calc software is shown in

Figure 6. As can be seen, the melting point of Type C

inclusions is about 14608C. The Type B and Type C

Figure 6. Phase diagram of Fe–Si–Ti–Al system.


intermetallic compounds were also detected by Wijk and

Brabie.[7] Also, the Type D inclusions contain almost only

pure Si, but with small amounts of O and Al. The melting

point of the pure Si phase is about 14008C.[19] Overall, the

Type D and Type A inclusions are the main inclusion

present in an FeSi alloy. Moreover, the size range of them

are wider than that for Type B and Type C inclusions.

It is reported that Al in FeSi enhances the formation of

MgO · Al2O3 spinel inclusions, and that Ca in FeSi has an

effect to prevent it.[9] However, Ca may react with

aluminum and form calcium aluminates that lead to

nozzle clogging.[25] Therefore, the Al and Ca contents in

FeSi alloys should be accurately determined before being

added to the steel. It should be mentioned that the acid

extraction is not suitable to use for an FeSi alloy due to that

Si is not directly soluble in acids.[1] Therefore, the EE

method is more appropriate.

3.4. Inclusions in SiMn

The characteristics of inclusions in SiMn are shown in

Table 9 and a detailed composition information is listed in

Table 10. As can be seen, Type A inclusions contain REM

oxides with some amounts of Si,Mn, andMg, Al.Moreover,

Type B inclusions are almost pure Al2O3 inclusions. Type C

inclusions are mostly silicon oxide with some amounts of

Ca and Mg. Type D inclusions are also silicon oxide, but


Type Si Mn Fe Mg Al Ca Ce La Pr Nd O

Type A 3–10 0–4 – 0–2 0–1 – 20–40 10–30 1–2 3–6 20–34

Average 8 2 – 1 1 – 32 22 2 4 28

Type B – – – – 42–56 – – – – – 43–60

Average – – – – 50 – – – – – 48

Type C 30–47 – – 0–2 0–1 6–10 – – – – 47–59

Average 37 – – 1 0.3 8 – – – – 53

Type D 36–57 0–3 – – – – – – – – 40–63

Average 45 1 – – – – – – – – 54

Type E 24–38 50–68 0–8 – – – – – – – 6–10

Average 30 59 4 – – – – – – – 7

Table 10. Composition of different type of inclusion in SiMn alloys (wt%).

30000

40000

50000

60000

-3)

Type Type A Type B Type C Type D Type E

Typical photo

Size range (mm) 1–26 2–5 6–12 1–8 3–16

Composition REM-Si–Mn–O Al–O Si–Ca–Mg–O Si–Mn–O Mn–Si–Fe–O

Percentage (%) 56 2 6 8 28

Table 9. Classification of inclusions in SiMn alloys.


FULL

PAPER

with small Mn contents. Type E inclusions are Mn–Si

intermetallic phases with small Fe contents. The melting

point of a Type E inclusion is 12508C, if only Mn and Si are

considered.[26] Overall, REM oxides are the main type of

inclusions found in SiMn alloys. This is followed by Type E

intermetallic phases. The particle size distribution of

Type A inclusions is shown in Figure 7.

0 2 4 6 8 10 12 14 16 18 20 22 24 260

10000

20000Nv

(mm

dv

(µm)

Figure 7. Particle size distribution of Type A inclusions in SiMnalloys.

4. Discussion

4.1. Soluble Impurities

According to the future demands on the ferroalloys from

the stand point of the customers in the steel industry, it is

imperative that more comprehensive investigations of

the dissolved element contents in the ferroalloys will be

� 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim steel research int. 85 (2014) No. 4 667


FULL

PAPER
needed.[2] The ferroalloy quality depends on the specific
use. However, in general the oxygen, carbon, sulfur,

phosphorous, and nitrogen contents are important to

know. The total oxygen content does not only reflect the

total amount of oxides inclusions, but it also represents the

mean diameter of the inclusions in ferroalloys.[10] It is

reported[10] that a similar trend was observed between the

C content and the amount and mean diameter of

inclusions, just in an inverse proportionality. Also, sulfur

and phosphor is known to have a deleterious effect on the

mechanical properties of steels.

Strong deoxidizers like Al, Ca, Mg, Ti, and Ce have a

large effect on the inclusion formation in the steel. These

elements may react with the oxides that exist in the liquid

steel after the addition of a ferroalloy to form complex

inclusions. For example, the TiO2 inclusions change to

Al–Ti–O complex inclusions after the addition of Al to the

steel.[27] Another problem is that large size inclusions may

be formed through the growth of a new layer on the surface

of the existing inclusions. For example, large size REM

oxides with hollow holes were observed after the addition

of REM alloys to the liquid steel.[28] A very small amount of

Al in FeSi enables the formation of MgO · Al2O3 spinel

inclusion in molten steel. Also, calcium in the FeSi alloy

may form Ca-aluminates that lead to severe clogging

problems during casting. Furthermore, Ti in FeSi will

prevent the effective grain growth due to the precipitation

of TiO2 at the grain boundaries. Ti will also react with

C and N and form hard carbon-nitrides that will

negatively affect the material properties of the final steel

product.

Trace elements such as Pb, Sn, Sb, Zn, and Bi should

also be investigated more in-depth. The use of ferroalloys

with lower contents of these tracing elements reduce the

bad quality steel from 30.8 to 4%.[4] Therefore, the exact

quantity of these non-metal impurities, strong deoxidizers

and trace elements in the ferroalloy should be known

before its introduction to the liquid steel. This is especially

important if additions are needed during the final stage of

the secondary refining, since there is only a short time to

promote inclusion separation by optimized stirring.

4.2. Insoluble Impurities

The dissolution kinetics of ferroalloys in the steelmaking

process was studied.[29] The dissolution time for 10–30mm

size ferroalloys is in a range from 2 to 160 s for FeSi,

SiMn, and FeMn at 1873K. However, the dissolution of the

insoluble impurities in the ferroalloys is not mentioned.

4.2.1. Non-Metallic InclusionsThe inclusions in ferroalloys may undergo some physical

or chemical change after the addition to the liquid steel.

Depending on the melting point and the thermodynamic

stability of the specific inclusions at the steelmaking

temperature and composition, the inclusions will dissolve


or stay solid. If the inclusions dissolve, a local high oxygen

or sulfur content may lead to a formation of new

inclusions. If the inclusions stay solid, the first possibility

is that strong deoxidants that dissolved in the liquid steel

may react with them under the formation of complex

inclusion. The second possibility is that these solid

inclusions may act as nucleation sites if the sizes are

small. The third possibility is that these solid inclusions

collide with each other and form clusters, which have a

bad effect on the casting process as well as on the final

product.

The non-metallic inclusions observed in this study are

the REM oxides in FeTi, FeSi, and SiMn, Al2O3 and Ti–Nb–

S–O in FeNb, silicon oxides in SiMn. The REM oxides and

Al2O3 inclusions probably remain solid and form clusters.

These two types of clusters are known as responsible for

the nozzle clogging and to cause decreased mechanical

properties in the final product. Also, silicon oxides might

dissolve in steel and thereafter form new inclusions.

Furthermore, they might react with strong deoxidizers to

form complex or large size inclusions. Overall, it is seems

that further studies of the Ti–Nb–S–O inclusions are

needed in the future.

4.2.2. Metallic InclusionsThe intermetallic phases observed in this study are a Ti–Fe

phase in FeTi alloy, Ca–Si, and Fe–Si–Ti phases in FeSi

alloy and a Mn–Si phase in SiMn alloy. Furthermore, the

mostly pure single metallic phase is pure Ti in FeTi, Nb in

FeNb, and Si in FeSi. According to the phase diagram of

these metallic phases, the majority of the melting points of

them are lower than the steelmaking temperature 16008C,with the exception of the almost pure Nb phase in FeNb.

Therefore, the majority of the metallic phases are assumed

to dissolve after the addition to steel. The effect of an

insolublemetallic phase that themelting points lower than

16008C on the cleanliness of the steel is that they can be

treated as soluble elements. Although the size range of the

metallic inclusions in ferroalloys in this study is large and

the proportion of the number percentage is tremendous,

no further investigation is conducted in this study.

However, the metallic inclusions (the almost pure Nb

phase, 24008C), which have a melting point that is higher

than the steelmaking temperature should be studied

more in depth in the future. In addition, the solubility of

them in steel for steelmaking conditions should also be

considered.

5. Conclusions

The inclusion contents in the following commercially

obtained ferroalloys were studied: FeTi, FeNb, FeSi, and

SiMn. The characteristics (such as morphology, size,

composition, and number) of inclusions were investigated

in 3D after EE. The influence of the soluble and insoluble



FULL

PA

inclusions can be beneficial or detrimental depending on

the specific demands on the steel. The following con-

clusions were obtained.

PER
1.
� 2

The inclusions in ferroalloys (FeTi, FeNb, FeSi, and

SiMn) can be successfully analyzed after a separation of

the inclusions from the matrix using the EE method.

2.
The non-metallic inclusion types observed in this study
are REMoxides in FeTi, FeSi, and SiMn alloys, Al2O3 and

Ti–Nb–S–O in FeNb alloys and silicon oxides in SiMn

alloys.

3.
The following intermetallic inclusions can be found: a
Ti–Fe phase in FeTi alloys, Ca–Si and Fe–Si–Ti phases

in FeSi alloys and a Mn–Si phase in SiMn alloys.

Furthermore, the mostly pure single metallic phases

are Ti in FeTi alloys, Nb in FeNb alloys, and Si in FeSi

alloys.

4.
The effect of the impurities in ferroalloy on steelmaking
process depends on the specific use. The characteristic

of the soluble and insoluble impurities should be

known before its addition to the liquid steel.

Acknowledgements

The China Scholarship Council (CSC) is acknowledged for

financial support to this study.

Received: April 23, 2013;

Published online: August 29, 2013

Keywords: ferroalloy; non-metallic and metallic

inclusions; electrolytic extraction; three-dimensional

investigations

References

[1] M. M. Pande, M. Guo, X. Guo, D. Geysen,

S. Devisscher, B. Blanpain, P. Wollants, Ironmak.

Steelmak. 2010, 37, 7.

[2] P. Jonsson, T. Sjoqvist, O. Grong, INFACON 9,

Quebec, Canada, 2001.

[3] W. Tiekink, A. Overbosch, P. Bras, J. Weerdt, C. Hoek,

R. Kooter, Clean Steel 8, Budapest, 2012.

[4] M. I. Gasik, A. I. Panchenko, A. S. Salnikov,

Electrometallurgy 2011, 3, 1.

013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

[5] P. Kaushik, H. Pielet, H. Yin, Ironmak. Steelmak.

2009, 36, 561.

[6] M. M. Pande, M. Guo, S. Devisscher, B. Blanpain,

Ironmak. Steelmak. 2012, 39, 519.

[7] O. Wijk, V. Brabie, ISIJ Int. 1996, 36, S132.

[8] K. V. Grigorovich, S. S. Shibaev, I. V. Kostenko, The

Twelfth Int. Ferroalloys Congress, Helsinki, Finland,

2010, 929.

[9] K. Mizuno, H. Todoroki, M. Noda, T. Tohge, Steel Res.

Int. 2001, 28, 93.

[10] T. Sjoqvist, PhD thesis, KTH, Stockholm, Sweden,

2001.

[11] T. Sjoqvist, P. Jonsson, O. Grong, Metall. Mater.

Trans. A 2001, 32A, 1049.

[12] T. Sjoqvist, P. Jonsson, 57th Electric Furnace Conf.,

USA, 1999, 383.

[13] T. Sjokvist, M. Goransson, P. Jonsson, P. Cowx,


[14] M. M. Pande, M. Guo, X. Guo, D. Geysen, S. Devisscher,

B. Blanpain, P. Wollants, The Twelfth Int. Ferroalloys

Congress, Helsinki, Finland, 2010, 935.

[15] R. Inoue, K. Kiyokawa, K. Tomoda, S. Ueda,

T. Ariyama, 8th Int. Workshop on Progress in

Analytical Chemistry and Materials Characterization

in the Steel and Metal Industries, Luxembourg, 2011.

[16] A. Katsumata, H. Todoroki, Iron Steelmaker 2002,

29, 51.

[17] N. Kojola, S. Ekerot, M. Andersson, P. G. Jonsson,


[18] A. Memarpour, V. Brabie, P. Jonsson, Ironmak.

Steelmak. 2011, 38, 229.

[19] IRON-Binary Phase Diagrams, 1982, 137.

[20] L. Zhang, B. G. Thomas, ISIJ Int. 2003, 3, 271.

[21] P. S. Rudman, Acta Metall. 1964, 12, 1381.

[22] J. L. Murray, Bull. Alloy Phase Diagrams 1986, 7, 157.

[23] J. L. Murray, Bull. Alloy Phase Diagrams 1981, 2, 55.

[24] P. Manfrinetti, M. L. fornasini, A. Palenzona,

Intermetallics 2000, 8, 223.

[25] M. Burty, L. Peeters, E. Perrin, S. Munzer, P. Colucci,

D. Salvadori, F. Schadow, J.-M. Valcarcel, J. Claes,

Rev. Metall. 2005, 102, 745.

[26] H. Okamoto, J. Phase Equilib. 1991, 12, 505.

[27] M. Sun, I. Jung, H. Lee,Metals Mater. Int. 2008, 6, 791.

[28] Y. Bi, E. Roos, A. Karasev, P. Jonsson, Clean Steel 8,

Budapest, 2012.

[29] Y. E. Lee, H. Berg, B. Jensen, Ironmak. Steelmak. 1995,

22, 486.


three-dimensional investigations of inclusions in ferroalloys

Documents