Liu et al.

SpringerPlus (2016) 5:1028

DOI 10.1186/s40064-016-2289-z

Open Access

RESEARCH

Thermodynamic simulation
onmineralogical composition ofCaOSiO2
Al2O3MgO quaternary slag system
ChaoLiu1, YuzhuZhang1,2*, JieLi2, JunguoLi2 andYueKang2

Abstract
It is necessary to elucidate the crystallization thermodynamic of mineralogical phases during the cooling process of
the molten BFS with different chemical composition, because the high-melting point mineral phase maybe crystal
lized during the fiber forming and thereafter cooling process. Thermodynamic calculation software FactSage6.4 and
the hot remelting experiments were performed to explore the influence of basicity, Al2O3 content and MgO con
tent on the crystallization of mineralogical components and their transformation. The results showed that the main
mineralography of the CaOSiO2Al2O3MgO quaternary slag system was melilite, and a certain amount of anorthite
and calcium metasilicate. The crystallographic temperature of melilite is increased with the increasing of basicity, MgO
and Al2O3 content, which has a significant impact on the utilization performance of the mineral wool prepared with
the hot blast furnace slag directly. With the increasing of basicity, there was a tendency that crystallographic amount
of melilite increased to the summit and then declined, while the amount of anorthite and calcium metasilicate
decreased consistently. Finally, these two mineralogical components could be replaced by magnesium rhodonite and
spinel with the increasing of basicity. When the basicity and MgO content were 1.0 and 9%, the crystallographic mass
ratio of melilite and anorthite increased, while that of calcium silicate declined, and replaced by spinel finally with the
increasing of Al2O3 content. When the basicity and Al2O3 content were 1.0 and 13%, the crystallographic mass ratio of
melilite increased, while that of anorthite and calcium silicate declined, and replaced by pyroxene and spinel with the
increasing of MgO content. To decline fiberization temperature of the melt BFS, the basicity, MgO and Al2O3 content
should be decreased during the modification process of chemical composition, because the crystallization tempera
ture of the primary crystalline phase in the slag system without modification.
Keywords: FactSage 6.4, Thermodynamic simulation, Mineralogical composition, Quaternary slag system,
Hot remelting
Background
Blast furnace slag (BFS) is the main by-product in the
process of smelting the pig iron. The quantity of BFS is
about 330450kg as per ton pig iron produced, which is
mainly composed of CaO, SiO2, Al2O3 and MgO, and a
kind of non-renewable mineral resources (Yi etal. 2012;
Cai etal. 2007). In recent years, the domestic and foreign
researchers have gradually realized the broad prospects
*Correspondence: [email protected]
1
School ofMaterials andMetallurgy, Northeastern University,
Shenyang110819, China
Full list of author information is available at the end of the article

of valorization of BFS to produce mineral wool, and set

out to explore the production technology of mineral wool
direct utilizing the molten BFS (Guo et al. 2012; Hua
and Fan 2008). The advantages of comprehensive utilization of BFS are energy saving, low cost, and recycling
of mineral resource, etc. It has important significance
for the iron and steel and mineral industry. Among the
mineral wool production methods, melt slag on a spinner wheel is used most commonly. Presently, fiber is
formed from the melt film on the spinner wheels is one
of the most important but least understood process phenomena. A great amount of input parameters such as
rotational frequencies of the spinner wheels, melt string

2016 The Author(s). This article is distributed under the terms of the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license,
and indicate if changes were made.

Liu et al. SpringerPlus (2016) 5:1028

impingement point, melt rheological properties, velocity of the blow-in flow, etc. have significant influence on
the fibrosis process. Among these parameters, melt rheological properties were dramatically impacted by the
melt temperature and chemical composition. Once the
fiberization temperature is low or the chemical composition of melt is unsuitable, the high-melting point mineral
phase could be crystallized during the fiberization and
thereafter cooling process, which has notably impact on
the mechanical property, thermal conductivity, stability
and leaching characteristics of the mineral wool. Consequently, application performance of the mineral wool
products should be affected. Therefore, it is necessary to
elucidate the crystallization thermodynamic of mineralogical phases during the cooling process of the molten
BFS with different chemical composition (Wang et al.
2011). In this work, the crystallization thermodynamic
of the quaternary slag system of CaOSiO2Al2O3
MgO has been simulated with thermodynamics software
FactSage 6.4. Simultaneously, the hot remelting test and
X-ray analysis have been carried out to determine the
mineralogical composition of this quaternary slag system
under the temperature of 1250C. Then, the crystallization and transformation processes of the high-melting
point mineralogical phases such as melilite, dicalcium
silicate, anorthite, etc. have been investigated. In view
of chemical composition, the influence of basicity, MgO
content and Al2O3 content on crystallization and transformation of these mineralogical phases have been
explored. The results obtained in this work could provide
experimental and theoretical foundations for modification of BFS and preparation of mineral wool (Gifford and
Gillott 1996; Li etal. 2013).

Methods
Thermodynamic calculation withFactSage 6.4

Thermodynamic simulation was carried out by using

FactSage [version 6.4, Thermfact/CRCT (Montreal) and
GTT technologies (Aachen)], which provides theoretical support for the complex calculation and simulation
of thermodynamic process. FactSage software was established based on the minimum Gibbs principle. Among
the software, the Eqilib module (Li etal. 2013) can calculate all kinds of slag system to reach the equilibrium
state at different temperature in different phases and
quality. Equilib module was used to simulate the thermodynamic equilibrium, and the FToxide database was
selected for this study. All calculations were carried out
under an atmospheric pressure, and the experimental
atmosphere was air. The temperature was selected from
900 to 1600C, and the increment was 5C (Cheng etal.
2013; Cao etal. 2008; Bale etal. 2002; Luz etal. 2011a, b).

Page 2 of 8

Hot remelting experimental andXray analysis

Remelting slag samples were prepared with analytical

reagent of CaO, SiO2, Al2O3 and MgO, which were dried
6 h at 105 C. A certain amounts of analytical reagents
were weighted and mixed according to Table1, and then
300 g remelting sample was transferred into a graphite
crucible (outer diameter 80mm, inner diameter 66mm,
height 100mm). The crucibles were placed in a tube furnace and heated to melt, and all experiments were carried
out under an atmospheric pressure, and the experimental atmosphere was air. The heating and cooling process
was operated as the following: first stage, the heating rate
of the furnace was set at 4C/min and lasted for 50min
to improve the temperature from room temperature to
200 C. Second stage, the heating rate was set at 12 C/
min and lasted for 50 min to increase the temperature
from 200 to 800C. Third stage, the heating rate was set
at 10C/min and lasted for 40min to improve the temperature from 800 to 1200 C. Fourth stage, the heating
rate was set at 6C/min and lasted for 50min to improve
the temperature from 1200 to 1500 C. Fifth phase, the
furnace temperature remained constant under 1500 C
for 120min to remelt the sample completely. Sixth stage,
the cooling rate was set at 5 C/min and lasted for
50min to decline the temperature from 1500 to 1250C.
Seventh stage, the furnace temperature remained constant under 1250 C for 30 min to guarantee thermodynamic equilibrium of the high melting mineral phases.
Finally, the crucible was taken out and quickly immersed
into water to quench the molten slag sample to reserve
the equilibrating phases.
For XRD analysis, the cooling samples were measured on a D/MAX2500PC X-ray diffractometer, used
Cu-K radiation operated at 40kV and 100mA. All the
diffraction profiles were obtained in continuous mode
at the scan speed of 10/min, and the 2 range from
10 to 80. Before analysis, the cooling samples were

Table1Chemical component of the remelting samples

(mass fraction, %)
No.

CaO

SiO2

Al2O3

MgO

R (CaO/SiO2)

39.00

39.00

13

1.0

40.86

37.14

13

1.1

42.55

35.45

13

1.2

44.09

37.14

13

1.3

40.50

40.50

10

1.0

37.50

37.50

16

1.0

40.50

40.50

13

1.0

37.50

37.50

13

12

1.0

Liu et al. SpringerPlus (2016) 5:1028

grinded and screened. The powder samples with particle size<0.048mm were placed into the groove of glass
sheets and smoothed. Then the glass sheets were put
into the X-ray diffraction instrument (Ma 2010; Hu etal.
2004).

Results anddiscussion
Based on original BFS and its modification target, the
chemical compositions of the series of quaternary slag
system selected in this research were listed in Table 1.
As presented in Table 1, the samples from No. 1 to No.
4 could exhibit the variation of slag basicity due to their
basicity being 1.0, 1.1, 1.2 and 1.3, respectively, whilst
MgO and Al2O3 content were 9 and 13 %, respectively.
The samples of No. 1, No. 5 and No. 6 could show the
changing of Al2O3 content, because the Al2O3 content
being 13, 10 and 16 %, respectively, whilst the basicity

Page 3 of 8

and MgO content were 1.0 and 9%, respectively. For the
samples of No. 1, No. 7, and No. 8, the basicity and MgO
content were 1.0 and 13%, respectively, whilst MgO content were 9, 6 and 12 %, which could applied to analyse
the influence of MgO content on the crystallization of
mineralogical phases.
Influence ofbasicity onmineralogical phases

When Al2O3 and MgO content were 13 and 9%, respectively, the mineralogical phases of slag with basicity of
1.0, 1.1, 1.2 and 1.3 under different temperature varied
from 900 to 1600C with an increment of 5C have been
calculated by FactSage 6.4 thermodynamic software. The
changing laws of the mineralogical phases content along
with temperature were illustrated in Fig. 1. As shown
in Fig. 1a for the slag basicity is 1.0, the quantity of liquid phase (denoted with SLAG#1 in Fig. 1) decreased

Fig.1 Changement of the mineralogical phases content along with decreasing of temperature under the condition of slag basicity being 1.0, 1.1,
1.2 and 1.3, respectively. a R=1.0, b R=1.1, c R=1.2 and d R=1.3

Liu et al. SpringerPlus (2016) 5:1028

gradually with the reducing temperature. When the temperature declined to 1387.2 C, the solid phase, namely
melilite (denoted with Mel_ in Fig. 1), start to crystallize. As the temperature declined to 1260.7C, anorthite
(denoted with CaAl2Si2O8 in Fig.1) has been crystallized.
At this point, the mass ratio (/%) of melilite crystallized
in this system reached to 75.9%. When the temperature
declined to 1243.2 C, the liquid melt has transformed
into solid phase completely, and calcium metasilicate
(denoted with CaSiO3 in Fig.1) has been crystallized. At
this temperature, the mass ratio of melilite and anorthite
reached to 84.72 and 11.25%, respectively, and keep constant with the further declining of temperature. Due to
the instability, calcium metasilicate has transformed to
wollastonite (denoted WOLLA in Fig.1). As the temperature declined to 1169.2C, the mass ratio of wollastonite
was 4.03%.
As shown in Fig. 1b for the slag basicity of 1.1, the
mineralogical phases composed of melilite, anorthite,
calcium metasilicate and wollastonite, were the similar
to those in the slag with basicity of 1.0 with the declining of temperature. The crystallization temperature of
melilite and calcium metasilicate increased to 1406.3
and 1247.4 C, respectively, while those of anorthite
and transforming temperature of calcium metasilicate
declined to 1286.8 and 1156.7C, respectively. Compared
with the crystallization quantities of different mineralogical phases in Fig.1a, the mass ratio of melilite increased
to 93.43%, while those of anorthite, calcium metasilicate
and its transformation product, namely wollastonite,
declined observably.
When the basicity of slag increased to 1.2 as shown in
Fig.1c, though the main mineralogical phases were melilite, magnesium rhodonite (denoted with Ca3MgSi2O8
in Fig. 1), dicalcium silicate (denoted with -Ca2SiO4 in
Fig.1) has been emerged and took the place of anorthite
and calcium metasilicate. The crystallization temperature
of melilite increased to 1418.9 C, while those of manganolite rhodonite and dicalcium silicate were 1408.9
and 1393.0 C, respectively. Unstable dicalcium silicate
was transformed to stable rankinite at the temperature
of 1243.6 C. When the temperature of slag declined to
1200C, the mass ratios of melilite and manganolite rhodonite were 87.82 and 10.06%, respectively, while that of
rankinite is only 2.12%.
As shown in Fig. 1d with the slag basicity of 1.3, the
crystallization phases were mainly melilite and magnesium rhodonite, which are similar to Fig. 1c. The
crystallized content of melilite decreased, while that of
magnesium rhodonite increased. Compared with the slag
sample with the basicity of 1.2, the crystallized content
of dicalcium silicate increased. Along with the increasing
of basicity, the crystallization temperature of dicalcium

Page 4 of 8

silicate, melilite and magnesium rhodonite increased to

1495.6, 1425 and 1418.9C, respectively. When the temperature dropped to 1399.6C, a trace amount of spinel
(denoted with SPINA in Fig. 1) was crystallized and its
content is only 0.34 %, while the mass ratios of melilite and magnesium rhodonite were 67.46 and 32.2 %,
respectively.
The accuracy of the simulation results were verified
by remelting experiment. Figure 2 shows the X-ray diffraction pattern of the remelting slag samples, which the
basicity was 1.0, 1.1, 1.2 and 1.3, while Al2O3 and MgO
content were 13 and 9 %, respectively. Those chemical
compositions were identical to that of the sample corresponding to Fig. 1. As illustrated in Fig. 2, the mineral phases of the remelting slag sample whose basicity
was 1.0 were composed of calcium magnesium melilite
(denoted with CaMgSi2O7 in Fig. 2), anorthite (denoted
with CaAl2Si2O8 in Fig. 2) and calcium metasilicate
(denoted with CaSiO3 in Fig. 2). The mineral phases of
the remelting slag sample in Fig. 2, whose basicity was
1.1 were the same as the basicity of 1.0. The mineral
phases of the remelting slag sample whose basicity was
1.2 as illustrated in Fig. 2 were composed of calcium
magnesium melilite (denoted with CaMgSi2O7 in Fig.2)
and dicalcium silicate (denoted with Ca2SiO4 in Fig. 2).
As illustrated in Fig. 2, only calcium magnesium melilite (denoted with CaMgSi2O7 in Fig.2) was found in the
remelting slag sample whose basicity was 1.3. It was possible that the amount of SPINA was too low to be found.
The results were similar to the mineralogical phases
shown in Fig.1 at the temperature of 1250C, and it was
feasible to simulate the mineralogical composition of
quaternary slag system by FactSage 6.4 thermodynamic
software. Accordingly, the water quenching method

Fig.2 X-ray diffraction pattern of the remelting slag sample

Liu et al. SpringerPlus (2016) 5:1028

could keep the mineralogical phases in the slag sample at

high temperature, that is to say, the mineralogical composition obtained at room temperature from the sample
undergone water quenching could reflect that at high
temperature.
In general, the main mineralogical phase of this series
of slag system was melilite. The crystallization temperature of melilite increased with the increasing of basicity,
while the mass ratio of crystallization trends to increase
to summit and then decrease. Meanwhile, basicity of the
slag system has a significant influence on the type and
crystallization temperature of the mineralogical phases.
When the basicity of slag was 1.0, a certain amount of
anorthite and wollastonite could be crystallized with the
declining of temperature. The mass ratios of anorthite
and wollastonite decreased gradually with the increasing of basicity. When the basicity of slag was increased
to 1.2, anorthite and calcium metasilicate were gradually
replaced by magnesium rhodonite and dicalcium silicate, respectively. When the basicity of slag increased to
1.3, the mass ratio of magnesium rhodonite crystallized
increased significantly and finally there was a certain
amount of -Ca2SiO4. The changing law of the dicalcium silicate content with the basicity is consistent with
the results of the Gelfi_M (Gelf etal. 2010), in which the
influence of basicity on dicalcium silicate content and the
leaching behavior of EAF steel slag were investigated.
Influence ofAl2O3 content onmineralogical phases

When the basicity and MgO content were 1.0 and 9 %,

respectively, the mineralogical phases of slag samples
with Al2O3 content of 10 and 16% under different temperature varied from 900 to 1600 C with an increment

Page 5 of 8

of 5 C have been calculated by FactSage 6.4 thermodynamic software. The changing laws of the mineralogical
phases content along with temperature were illustrated in
Fig.3. Combined with Fig.1a with Al2O3 content of 13%,
the influence of Al2O3 content on mineralogical phases of
the slag system could be analyzed.
As shown in Fig. 3a corresponding to the slag sample
with Al2O3 content being 10%, the mineralogical phases
crystallized during cooling process were identical to
those of the slag sample with Al2O3 content being 13 %
illustrated in Fig. 1a. As Al2O3 content was decreased
from 13 to 10%, the crystallization temperature of melilite and anorthite declined from 1387.2 and 1260.7 to
1379.5 and 1240.5 C, respectively. When the temperature dropped to 1301.1 C, calcium metasilicate started
to crystallize from the slag system. At this temperature,
the mass ratio of melilite crystallized reached to 66.57%.
As the temperature declined to 1240.5C, anorthite commenced to crystallize, and the mass ratios of melilite and
calcium metasilicate crystallized have reached to 76.79
and 7.6 %, respectively. When the temperature declined
to 1179.7 C, the unstable compound named calcium
metasilicate has transferred into wollastonite, which content was 7.6 %. At this temperature, the mass ratios of
melilite and anorthite crystallized were 80.93 and 6.56%,
respectively.
Compared Fig. 3b and Fig. 1a with Al2O3 content of
16 and 13%, respectively, the main mineralogical phases
were melilite and anorthite. As shown in Fig.3b, the crystallization temperature of slag sample increased to 1400.7
and 1273.7C, as Al2O3 content increased to 16%. Meanwhile, calcium metasilicate could not be crystallized and
replaced by spinel which crystallized at 1267.4 C. At

Fig.3 Changement of the mineralogical phases content along with decreasing of temperature under the condition of Al2O3 content being 10 and
16%. a (Al2O3)=10%, b (Al2O3)=16%

Liu et al. SpringerPlus (2016) 5:1028

1250C, the mass ratios of melilite and anorthite crystallized were about 84 and 14.27%, respectively, while that
of spinel was only 1.43%.
The accuracy of the simulation results were verified by
remelting experiment. Figure 4 shows the X-ray diffraction pattern of the remelting slag sample with Al2O3 content of 10 and 16%, while the basicity and MgO content
were 1.0 and 9%, respectively, which chemical composition was identical to that of the sample corresponding to
Fig. 3. As illustrated in the Fig. 4, the main mineralogical phases of the remelting slag sample whose Al2O3 was
10% were composed of aluminum melilite (denoted with
Ca2Al2SiO7 in Fig.4), anorthite and calcium metasilicate.
As illustrated in the Fig. 4, aluminum melilite (denoted
with Ca2Al2SiO7 in Fig.4) and spinel were be found in the
remelting slag sample whose Al2O3 was 16%. The results
were similar to the mineralogical phases shown in Fig.3
at the temperature of 1250 C. In addition, due to the
remelting slag samples were prepared with chemical pure
reagents and the heterogeneity of the remelting slag sample during mixing, remelting and cooling process, there
was a trace amount of dicalcium silicate. A little of dicalcium silicate crystallized in the remelting slag sample has
little effect on the result and the mineralogical composition was consistent with that simulated with thermodynamics software.
In general, the main mineralogical phases of the slag
systems were melilite and anorthite, which crystallization
temperature could be improved along with the increasing of Al2O3 content. Furthermore, there was a tendency
that the mass ratios of melilite and anorthite crystallized
increased with Al2O3 content. When the Al2O3 content were 10 and 13 % (shown in Figs. 1a, 3a), a certain

Fig.4 X-ray diffraction pattern of the remelting slag sample

Page 6 of 8

amount of calcium metasilicate could be crystallized.

The mass ratio of calcium metasilicate crystallized trend
to decrease gradually with the increasing of Al2O3 content. When Al2O3 content increased to 16 % (shown in
Fig. 3b), calcium metasilicate could not be crystallized
and substituted by spinel.
Influence ofMgO content onmineralogical phases

When the basicity and Al2O3 content were 1.0 and 13%,
respectively, the mineralogical phases of slag with MgO
content of 6 and 12% under different temperature varied
from 900 to 1600C with an increment of 5C have been
calculated by FactSage 6.4 thermodynamic software. The
changing laws of the mineralogical phases content along
with temperature were illustrated in Fig. 5. Combined
with Fig. 1a with MgO content of 9 %, the influence of
MgO content on mineralogical phases of the slag system
could be analyzed.
As shown in Fig.5a, the mineralogical composition of
the slag sample with MgO content of 6 % was identical
to that the slag sample with MgO content of 9 % illustrated in Fig. 1a. When MgO content decreased from
9 to 6 %, the crystallization temperature of melilite and
anorthite declined from 1387.2 and 1260.7 to 1359.7 and
1249.7C, respectively, while that of calcium metasilicate
increased from 1243.2 to 1291.2 C. It was showed that
the mass ratio of melilite was 40.29% at the temperature
of calcium metasilicate crystallization, which was much
less than crystallization amount as MgO content was 9%.
When the temperature declined to the point of anorthite
crystallization, the mass ratios of melilite and calcium
metasilicate crystallized were 56.02 and 7.6 %, respectively. At 1179.7C, unstable calcium metasilicate should
transform the species of wollastonite, and the mass ratio
of melilite, anorthite and wollastonite were 24.61, 64.67
and 11.04%, respectively.
Comparing the slag sample with MgO content of 6
and 12% shown in Fig.5a, b, respectively, the main mineralogical phase was melilite although MgO content
increased to 12 % shown in Fig. 5b. It was concluded
that calcium metasilicate could not be crystallized or
replaced by spinel as MgO content increased. Meanwhile,
anorthite was disappearred and substituted by pyroxene
(denoted with cPyrA in Fig. 5). The crystallization temperature of melilite, spinel and pyroxene were 1392.3,
1344 and 1247 C, respectively. At 1247 C, the mass
ratios of melilite, spinel and pyroxene crystallized were
84.46, 4.46 and 10.81%.
The accuracy of the simulation results were verified by
remelting experiment. Figure 6 shows the X-ray diffraction pattern of the remelting slag sample with MgO content of 6 and 12%, while the basicity and Al2O3 were 1.0
and 13%, respectively, whose chemical composition was

Liu et al. SpringerPlus (2016) 5:1028

Page 7 of 8

Fig.5 Changement of the mineralogical phases content along with decreasing of temperature under the condition of MgO content being 6 and
12%. a (MgO)=6%, b (MgO)=12%

heterogeneity of the remelting slag sample during mixing, remelting and cooling process maybe the reason to
bring about this phenomena.
In general, the main mineralogical phase of the slag
system was melilite. Both crystallization temperature and
mass ratio of melilite increased with the increasing MgO
content. When MgO content was 6%, a certain amounts
of anorthite, calcium metasilicate and wollastonite
should be crystallized except melilite. With the increasing of MgO content, the mass ratio of anorthite, calcium metasilicate and wollastonite crystallized declined.
Finally, anorthite and calcium metasilicate were replaced
by pyroxene and spinel, respectively, as MgO content
increased to 12%.

Fig.6 X-ray diffraction pattern of the remelting slag

similar to the sample corresponding to Fig. 5. As illustrated in the Fig.4, the main mineralogical phases of the
remelting slag sample whose MgO was 6 % were composed of aluminum melilite (denoted with Ca2Al2SiO7
in Fig. 6), calcium metasilicate and periclase (denoted
with Ca3MgAl4O10 in Fig. 6). As illustrated in Fig. 4,
spinel (denoted with MgAl4O10 in Fig. 6) was found in
the remelting slag sample whose MgO was 12 %. The
results were similar to the mineralogical phases shown
in Fig. 5 at the temperature of 1250 C. Although anorthite was replaced by periclase, as mentioned earlier, the

Conclusion
It is necessary to elucidate the crystallization thermodynamic of mineralogical phases in the melt BFS intent to
fiberization, because crystalline phases have an important influence on application performance of the mineral
wool. Therefore, the crystallization thermodynamic of
the quaternary slag system of CaOSiO2Al2O3MgO
has been simulated with thermodynamics software
FactSage 6.4 in the present study. Simultaneously, the hot
remelting test and X-ray analysis have been carried out
to determine the mineralogical composition of this quaternary slag system under the temperature of 1250C. It
was shown that the results of X-ray analysis were consistent with that simulated with FactSage 6.4. In conclusion, it is feasible to keep the mineralogical phases in the
slag sample at high temperature with the cooling method
of water quenching, and to simulate the mineralogical

Liu et al. SpringerPlus (2016) 5:1028

composition of the quaternary slag system by FactSage

6.4.
Temperature and chemical composition, especially
basicity, MgO and Al2O3 content, have significant influence on crystallization and transformation process of
the high-melting point mineralogical phases. When the
basicity, Al2O3 and MgO content were 1.0, 13 and 9 %,
respectively, the crystallizing phases were composed of
melilite, anorthite and calcium metasilicate. The mass
ratio of melilite crystallized trend to increase to summit and then declined, while that of calcium metasilicate
declined gradually with the increasing of the basicity.
As the basicity increased to 1.2, anorthite was replaced
by magnesium rhodonite. When the basicity and MgO
content were 1.0 and 9%, the mass ratio of melilite and
anorthite crystallized increased, while that of calcium
silicate declined, and replaced by spinel finally, with
the increasing of Al2O3 content. When the basicity and
Al2O3 content were 1.0 and 13%, the mass ratio of melilite crystallized increased, while those of anorthite and
calcium silicate declined, and replaced by pyroxene and
spinel with the increasing of MgO content.
The fiberized temperature is a key parameter for production of the minerla wool, because the high-melting
point mineralogical phases such as melilite, dicalcium
silicate, anorthite, etc. could be crystallized with the
declining of temperature. According to the crystallization thermodynamic results calculated with FactSage 6.4,
the crystallization temperature of the primary crystallize phase such as melilite, anorthite, etc. increased with
the increasing of the basicity, MgO and Al2O3 content
in the slag system. Consequently, to decline the fiberization temperature of the melt BFS, the basicity, MgO and
Al2O3 content should be decreased during the modification process of chemical composition.
Authors contributions
LC participated in hot remelting experimental and thermodynamic calculation
with FactSage 6.4, and drafted the manuscript. LJ participated in hot remelting
experimental. KY participated in thermodynamic calculation with FactSage 6.4
and performed the statistical analysis.ZYZ and LJG conceived and designed
the study, performed the statistical analysis, and modified the manuscript. All
authors read and approved the final manuscript.
Author details
1
School ofMaterials andMetallurgy, Northeastern University, Shen
yang110819, China. 2College ofMetallurgy andEnergy, North China Univer
sity ofScience andTechnology, Tangshan063009, China.

Page 8 of 8

Acknowledgements
This research was supported by Key Projects in the National Science and Tech
nology Pillar Program (2012BAE09B02), National Natural Science Foundation
of China (51474090), State Key Laboratory of Solid Waste Reuse for Building
Materials (SWR-2013-010), and The National Youth Science Fund Project
(51501080).
Competing interests
The authors declare that they have no competing interests.
Received: 26 January 2016 Accepted: 5 May 2016

References
Bale CW, Chartrand P, Degterov SA, Eriksson G, Hack K, Mah-foud RB, Melancon
J, Pelton AD, Petersen S (2002) FactSage thermo-chemical software
and databases. CALPHAD Comput Coupling Phase Diagr Thermochem
26:189228
Cai J, Du T, Chen C, Dong H (2007) Recovery of residual-heat resource in iron &
steel industry and study on key technologies. Iron Steel 06:408416
Cao Z, Song X, Qiao Z (2008) Thermodynamic modeling software and applica
tion. Chin J Rare Met 02:216219
Cheng Z, Guo J, Cheng S (2013) Inclusion composition control during LF refin
ing for SPCC using FactSage combined with industrial trials. J Iron Steel
Res Int 20(03):1420
Gelf M, Cornacchia G, Roberti R (2010) Investigations on leaching behavior of
EAF steel slag. In: Euroslag 2010, University of Bresca, Italy
Gifford PM, Gillott JE (1996) Alkali-silica reaction (ASR) and alkali-carbonate
reaction (ASR) in activated blast furnace slag cement (ABFSC) concrete.
Cem Concr Res 26:2126
Guo Q, Yuan S, Liu J, Li H (2012) Experimental research of slag wool producing
with blast furnace slag. China Metal 8:4649
Hu L, Zhang Q, Shen Y (2004) The experimental method of X-ray diffraction
analysis and application. J Hebei Inst Technol 3:8386
Hua J, Fan J (2008) Development of heat recovery of blast furnace slag in
China. Iron Steel 36:5154
Li J, Zhang Y, Liu W, Xing H, Long Y (2013) Quenched and tempered blast
furnace slag as slag fiber raw materials. Chin J Environ Eng 12:49714977
Luz AP, Braulio MAL, Tomba Martinez AG, Pandolfelli VC (2011a) Thermody
namic simulation models for predicting Al2O3MgO castable chemical
corrosion. Orig Res Artic Ceram Int 37:31093116
Luz AP, vila TA, Bonadia P, Pandolfelli VC (2011b) Thermodynamic simulations
and isothermal solubility diagrams as tools for slag foaming control.
Ceram Int 37:29472950
Ma L (2010) Origin and development of the multi purpose X-ray diffractom
eter (part 1). Phys Test Chem Anal (Part A Phys Test) 46:500509
Wang H, Zhang C, Qi Y (2011) Quantitative analysis of non-crystalline and
crystalline solids in blast furnace slag. J Iron Steel Res Int 18:810
Yi H, Xu G, Cheng H, Wang J, Wan Y, Chen H (2012) An overview of utilization of
steel slag. Procedia Environ Sci 16:791801