Академический Документы
Профессиональный Документы
Культура Документы
Introduction
Sintering is a thermal treatment for bonding particles into a coherent and
predominantly solid structure via various mass transfer events that generally occur at
the atomic scale. This bonding leads to an improved strength, and a lower system
energy [1]. In the iron ore sintering process, fine particles bond to larger particles,
which, under heat, fuse together in different stages, as shown in Figure 1. This forms
a porous matrix that is required in the blast furnace during the steel making process.
The primary materials required for iron ore sintering are ore, coke, various fluxes, and
recycled material from the plant, i.e. return fines. Initially, these materials must be
blended together in a rotating drum, allowing the smaller particles to adhere to the
larger ones. This material is the placed onto a sintering machine, where the surface is
ignited. Air is sucked down through the sinter bed, thus allowing the agglomeration of
the material, the permeability of which will depend on the size of the heated material.
The heat is supplied by the combustion of the coke in the mixture. The heat profile
generated through the sinter bed has a significant effect on both the type of sinter
produced, as well as
its physical and its
reduction
[2].
properties
Other factors
sinter
quality
proportions
In order to ensure that the correct amount of energy is supplied to the sinter mix
(i.e. the amount of coke), the thermal properties of the iron ores and fluxes must be
known. This will also benefit not only the operation of the sinter plant, but also the
associated costs and economics.
Experimental Details
The specific heats of iron ores were obtained during continuous heating from room
temperature (~25C) up to 1000C at a rate of 10C/min. The apparatus used during
this experimental program was a modified Infrared Gold Image Furnace as shown in
Figure 2.
Six iron ore samples were selected for this study as given in Table 1. Each sample
was vacuum dried, ground to -212m and packed inside a 10.2 mm inside diameter
glass silica tube, to a length of 30mm. Ceramic insulators were also placed into the
end of the tubes. To ensure uniform emissivity, the glass tube was coated in soot
produced from the combustion of acetylene gas. K-type thermocouples monitored the
centre temperature of the sample, and the temperature at the surface of the sample.
Heat was provided to the sample primarily by radiation from a surrounding graphite
tube. A third thermocouple was used to control both the temperature and heating rate
of this tube. Argon gas at 5ml/min was passed over the sample to keep it from
oxidising, and to aid in the removal of volatiles. The density value for Iron Ore 1
(5.7 g/cm3) appears to be quite high with respect to pure magnetite having a density of
approximately 5.2 g/cm3, however these measurements were repeated numerous
times, and this value was considered to be acceptable.
Table 1. Iron ore characteristics.
Iron
Ore
Origin
Density
(g/cm3)*
Australia
5.7
Australia
4.9
Australia
4.5
Australia
5.1
South
America
5.4
Australia
5.4
The ability to evenly transfer heat from the surface of the sinter compact
throughout the material is extremely important, and is also highly dependent on the
process conditions. In the experimental process, radiation was the main form of heat
transfer. The radiant heat transfer depends upon the temperatures of the heat source
and the sample, as well as surface roughness and the viewing angle [1]. The radiative
heat flux can be determined by:
Q = F12 Tg Ts
4
(1)
where Q is the heat flux; F1-2 is the radiation shape factor; is the Stefan Boltzmann
constant; Tg is the graphite temperature (K); and Ts is the sample temperature (K). The
radiation shape factor is defined as a fraction of diffusely distributed radiation leaving
the surface A1 that reacts with A2. It is also a function of the emissivities of both the
glass and graphite tubes, as well as their surface areas.
The radiation shape factor (F1-2) for these experiments can be defined as:
F12 =
A1
1 A1 1
+ 1
1 A2 2
(2)
Cp
dT
2T
=k 2
dt
x
(3)
The calibration results of the experimental procedure are detailed elsewhere [3].
Results
Upon solving Equation (3) the following data for the specific heats were generated as
shown in Figure 3. In most cases, these data clearly show five areas of thermal
importance where either endothermic peaks or exothermic troughs are seen. To
quantify each heat associated with the reaction, a baseline was interpolated by
assuming that a straight line connected the start and end temperature of each reaction.
The heat of reaction was calculated by integrating the area between the apparent
specific heat curve and the assumed baseline in each reaction temperature range.
X-ray diffraction (XRD) was used to determine the mineralogical composition of
the samples. For XRD analysis to take place, the samples had to be prepared
according to the results given in Figure 3. Four temperatures were chosen for
investigation so as to cover the areas before and after each peak or trough. For Iron
Ore 1, these temperatures were 25C, 475C, 700C, and 900C. The XRD results for
Iron Ore 1 are shown in Figure 4.
IRON ORE 1
7.0E+06
Peak
Temp.
(oC)
122
End
Temp.
(oC)
157
(MJ/m3)
65
67
5.0E+06
Start
Temp.
(oC)
87
4.5E+06
3.5E+06
II
III
270
529
338
593
415
626
3.0E+06
IV
634
678
693
2.5E+06
801
840
865
-15
Start
Temp.
(oC)
121
264
549
650
-
Peak
Temp.
(oC)
136
347
584
682
-
End
Temp.
(oC)
151
431
618
709
-
Start
Temp.
(oC)
121
254
550
-
Peak
Temp.
(oC)
140
303
580
-
End
Temp.
(oC)
157
390
591
-
6.5E+06
Region
6.0E+06
5.5E+06
4.0E+06
2.0E+06
0
100
200
300
400
500
600
700
800
H
10
900 1000
Temperature (oC)
IRON ORE 2
Apparent Specific Heat (J/m3K)
7.0E+06
6.5E+06
Region
6.0E+06
5.5E+06
I
II
III
IV
V
5.0E+06
4.5E+06
4.0E+06
3.5E+06
3.0E+06
2.5E+06
H
(MJ/m3)
1.6
218
11
3
-
2.0E+06
0
100
200
300
400
500
600
700
800
900 1000
Temperature (oC)
IRON ORE 3
7.0E+06
6.5E+06
Region
6.0E+06
5.5E+06
5.0E+06
I
II
III
IV
V
4.5E+06
4.0E+06
3.5E+06
3.0E+06
2.5E+06
2.0E+06
0
100
200
300
400
500
600
700
Temperature (oC)
800
900 1000
H
(MJ/m3)
5.3
343
4.3
-
IRON ORE 4
Apparent Specific Heat (J/m3K)
7.0E+06
6.5E+06
Start
Temp.
(oC)
114
270
527
625
822
Peak
Temp.
(oC)
133
325
588
679
853
End
Temp.
(oC)
150
408
617
693
871
I
II
Start
Temp.
(oC)
99
270
End
Temp.
(oC)
160
401
III
IV
V
603
-
Peak
Temp.
(oC)
126
322,
379
679
-
694
-
13.4
-
Start
Temp.
(oC)
Peak
Temp.
(oC)
End
Temp.
(oC)
H
(MJ/m3)
Region
6.0E+06
5.5E+06
I
II
III
IV
V
5.0E+06
4.5E+06
4.0E+06
3.5E+06
3.0E+06
2.5E+06
H
(MJ/m3)
2.4
62
43
12
-5.5
2.0E+06
0
100
200
300
400
500
600
700
800
900 1000
Temperature (oC)
IRON ORE 5
7.0E+06
Region
6.5E+06
6.0E+06
5.5E+06
5.0E+06
4.5E+06
4.0E+06
3.5E+06
3.0E+06
2.5E+06
2.0E+06
0
100
200
300
400
500
600
700
800
H
(MJ/m3)
2.2
46
900 1000
Temperature (oC)
IRON ORE 6
7.0E+06
Region
6.5E+06
6.0E+06
5.5E+06
5.0E+06
101
128
160
3.7
4.5E+06
II
293
329
381
7.5
4.0E+06
3.5E+06
III
520
595
647
39
3.0E+06
IV
603c
678
697
13.4
823
850
868
-4.5
2.5E+06
2.0E+06
0
100
200
300
400
500
600
700
800
900 1000
Temperature (oC)
Figure 3 continued.
7
4000
4000
3500
Fe2O3
FeOOH.nH2O
3000
Fe2O
3000
C o u n ts
1500
Al2Si2O5(OH)4
SiO2
Fe2O3
2500
Fe2O3
Fe2O3
FeOOH.nH2O
2000
Fe2O3
C o u n ts
Al2Si2O5(OH)4
2500
Fe2O3
FeOOH.nH2O
SiO2
Fe2O3
Al2Si2O5(OH)4
2000
1500
1000
1000
500
500
Fe2O3
Al2Si2O5(OH)4
Fe2O3
SiO2
Fe2O3
SiO2
Fe2O3
10
20
30
40
50
60
70
80
2 (o )
4000
10
20
30
40
50
2 (o )
60
70
80
4000
3500
3500
3000
3000
Fe2O3
Fe2O3
2500
C o u n ts
C o u n ts
2500
Fe2O3
2000
Fe2O3
SiO2
SiO2
Fe2O3
1500
Fe2O3
1500
Fe2O3
1000
1000
500
500
Fe3O4
Fe2O3
SiO2
Fe3O4
Fe2O3
Fe2O3
SiO2
Fe3O4
Fe3O4
Fe3O4
Fe2O3
2000
Fe2O3
Fe2O3
Fe3O4
10
20
30
40
50
60
70
80
10
2( )
20
30
40
2(o)
50
60
70
80
Figure 4. XRD results for Iron Ore 1 at 25C, 475C, 700C, and 950C.
From Figures 3 and 4, it is clear that the peaks associated with region I were the
result of the free water being removed from the sample.
Region II was due to the bound water being released from the iron ores. This is
evidenced by Figures 4(a) and (b). Figure 4(a) shows that goethite (FeOOH.nH2O)
was present at 25C, however at 475C, the goethite was no longer present. Cornell
and Schwertmann [4] found that under heat stress, goethite proceeds directly to
hematite without any intermediate phase, as given by: 2 FeOOH Fe2O3 + H2O
It was evident that goethite had a strong influence on the amount of energy
required to heat the iron ores. Depending upon the amount of goethite present in each
sample, the endothermic reaction of region II ranged from 7.5 up to 343 MJ/m3. The
latter was found in Iron Ore 3, showing the sample contained predominantly goethite,
with the bulk of its mass loss occurring in this temperature region and exhibiting
negligible changes at the elevated temperatures.
8
The thermal decomposition of goethite exhibited a double peak when iron ores 4
and 5 were heated as shown in Figure 3. It is still uncertain why this could occur, with
literature sources showing a number of discrepancies [5-7]. It is possible that the
double peaks are caused by the initial nH2O being driven off the sample at lower
temperatures, followed by the decomposition of the hydroxide bonds in the remaining
ferric oxide. However, it has been shown that goethite particle size and impurities in
the sample can also have an effect on the reactions [5, 6].
The origin of the endothermic regions III was identified by the XRD analysis,
showing that kaolinite (Al2Si2O5(OH)4) was present in the majority of samples. To
confirm this, a sample of kaolinite was heated in the infrared furnace in the same way
as the iron ores. The thermal properties of kaolinite were measured and analysed. The
thermal analysis showed a specific heat increase at approximately 600C, thus
confirming the presence of kaolinite in the iron ore samples.
As reported by Deer et al. [8], there is little or no surface water present in
kaolinite, and most of the dehydration (loss of constitutional OH) occurs between
400C and 525C, thus producing the endothermic peak at approximately 550C.
Region IV was due to the magnetic transformation of Fe3O4 at the Curie
temperature at 678oC.
Upon analysis of Figures 4(c) and (d), it was determined that region V was due to
the partial reduction of hematite (Fe2O3) to magnetite (Fe3O4), which occurs at
approximately 850C. This was evidenced by the large number of magnetite peaks
detected by the XRD at 950C. It is possible that this solid-state reduction was caused
by hematite reacting with the carbon present in the iron ore samples.
Scanning Electron Microscopy (SEM) and Energy Dispersive Spectroscopy (EDS)
analysis of the iron ore samples was undertaken in order to investigate the structural
changes of the ores during heating. This was performed on raw samples, as well as
the samples heated to 1000C. The samples were prepared in resin and later polished.
This gave some indication of the chemical and physical changes that occurred as a
result of the thermal processing.
Figure 5 shows the SEM images taken at 1000x magnification of both raw and
heated samples of Iron Ore 1.
(a)
(b)
Figure 5. (a) Magnified image (1000x) of raw Iron Ore 1, and (b) magnified image
(1000x) for Iron Ore 1 after being heated to 1000C.
Figure 5 clearly shows the physical changes that occur within the iron ore samples
due to the thermal processing. Prior to the heating, the particles were clearly
separated. Upon heating (see Figure 5b), sufficient bonding occurred so that the
separate iron ore particles structurally bonded in the same way as shown in Figure 1,
i.e. the particles had fused, forming a porous matrix of iron ore sinter.
A line-map across the sintered sample was taken. This involved selecting a line on
the SEM that appeared to pass through all visible phases present. Every point in this
line was essentially analysed for composition. This gave indications of chemical
composition throughout the various phases. The line map is shown in Figure 6. The
curves below the image are related to the concentrations of each of the elements
present. Thus the concentrations differ according to the nature of the phase.
To further obtain detailed physical descriptions of the effects of the thermal
processing on the iron ore samples, fractured sample SEM images were taken (see
Figure 7). These images required a sample of sintered iron ore to be placed directly
into the SEM, with no prior sample preparation such as for those shown in Figure 5.
The images gave a clear indication of how the different elements behaved during
10
heating. The images were analysed and a number of phases were found which were
later analysed for compositions, and the results are given in Table 2. P1, 2, 3 and 5 in
Table 2 correspond to iron oxide particles with some small traces of Al2O3 and SiO2.
The iron oxide particles have chemistries consistent with hematite (Fe2O3), i.e.
69.94% iron and 30.06% oxygen [4]. P6 and P7 are phases with increased levels of
silica and alumina, while also containing traces of magnesium oxide. These
crystalline phases appear to be the major bonding phase throughout the sintered
sample, as evidenced by their abundance in Figure 7.
Chemical analysis of P4 revealed a widely different chemistry from the bulk
sample. This phase also appears responsible for bonding, however it is much less
abundant. The chemistry of P4 is similar to that of P6 and P7, with traces of sulphur,
calcium, sodium, potassium, chlorine and magnesium also being identified. These
trace elements significantly lowered the melting temperature, hence the particle
appears glassy and globular, which is detrimental to the sinter strength.
P1
P2
P3
P4
P5
P6
P7
34.84
37.67
30.45
42.77
30.55
40.99
33.90
Fe
64.00
60.26
68.05
43.50
68.85
50.69
60.10
Al
0.50
1.29
1.14
1.01
0.59
2.58
1.74
Si
0.66
0.78
0.36
3.27
5.41
4.13
Mg
0.27
0.13
1.23
0.15
Ca
0.63
Na
1.52
Cl
4.58
1.23
0.19
11
(a)
(b)
(c)
Figure 7. Fractured sample SEM images for sintered Iron Ore 1.
13
Conclusions
The thermal properties of six iron ores were investigated using both experimental and
numerical techniques. The iron ores were heated to 1000C in an infrared furnace,
and the areas of thermal importance were identified and analysed. The energies
associated with each of these thermal regions were calculated and could be applied in
determining the amount of energy, i.e. coke rate, required to heat a certain iron ore
during the sintering process. Thus not only do the thermal properties of the iron ores
give some indication of their components, but they could also be used to increase the
efficiency and reduce the running costs of a sinter plant.
The free moisture present was removed at temperatures around 120-140C. The
decomposition of goethite was found to occur between 250-350C. This endothermic
reaction was found to be extremely significant with the energy required to heat an
iron ore being highly dependent upon the amount of goethite present in the sample.
Kaolinite was endothermically decomposed at around 600C. The magnetic
transformation of magnetite was found to occur at approximately 680C, while the
partial reduction of hematite to magnetite occurred at around 850C.
References
1.
German, R.M. 1986. Sintering Theory and Practice; John Wiley and Sons, Inc., New York.
2.
Dawson, P.R., Ostwald, J., and Hayes, K.M. 1993. The Influence of the Sintering Temperature Profile
on the Mineralogy and Properties of Iron Ore Sinters; Australian Institute of Mining & Metallurgy.
3.
Strezov, V., Lucas, J.A. and Strezov, L. 2000. Metallurgical & Materials Trans. B, 31B, 1125-1131
4.
Cornell, R.M., and Schwertmann, U. 1996. The Iron Oxides - Structure, Properties, Reactions,
Occurrence and Usage, VCH, Weinheim, Germany.
5.
Derie, R., Ghodsi, M., and Calvo-Roche, C. 1976. DTA Study of the Dehydration of Synthetic
Goethite -FeOOH, J. Thermal Analysis, 9, 435-440.
6.
Naono, H., Nakai, K., Sueyoshi, T., and Yagi, H. 1987. Porous Texture in Hematite Derived from
Goethite: Mechanism of Thermal Decomposition of Goethite, J. Colloid Interface Sci., 120-2, 439-450.
7.
8.
Deer, W.A., Howie, R.A., and Zussman, J. 1966. An Introduction to the Rock-Forming Minerals, 2nd
edn, Longman Publishers, UK.
14