Hans von Storch (ca 1970-1975 Progammierer bei Ramming und Sündermann)
The effect of additives on Al O –SiO –SiC–C ramming ...
Transcript of The effect of additives on Al O –SiO –SiC–C ramming ...
The effect of additives on Al2O3–SiO2–SiC–C
ramming monolithic refractories
E. Karamian, A. Monshi and S.M. Siadati*
Department of Materials Engineering, Isfahan University of Technology (IUT), Isfahan,84156-83111, Iran
*E-mail: [email protected]
ABSTRACT
The aim of this study is to describe the effect of containing additives on increasing cold crushing
strength (CCS) and bulk density (BD) of Al2O3–SiO2–SiC–C monolithic refractories. Two series of
carbon containing monolithics were prepared from Iranian chamotte (Samples A) and Chinese bauxite
(Samples B), as 65 wt.% in each case together with, 15 wt.% SiC-containing material regenerates
(crushed sagger) – 10 wt.% fine coke (a total of 90% aggregate) and 10 wt.% resole (phenol
formaldehyde resin) as a binder. Different types of additives (such as silicon and ferrosilicon
metal) are added to a batch of 100 g of mixture and the physical and mechanical properties (such
as BD, apparent porosity and CCS) are measured after tempering at 200�C for 2 h and firing at
1100�C and 1400�C for 2 h. After low temperature tempering at 200�C, silicon and ferrosilicon
contribute to the formation of stronger cross linking in the resulting structure and provides CCS
values as high as 65 MPa. After high temperature sintering, at 1400�C, SiC whiskers of nano sized
diameter are formed due to the presence of Si and FeSi2 and increases the CCS values of the
refractories as high as 3 – 4 times in sample containing 6 wt.% ferrosilicon metal as additive,
compared to the material without additive. The temperature of 1100�C is a transient temperature,
used in high temperature sintering.
Keywords: silicon additives, ferrosilicon additives, ramming, monolithic refractories, carbon
1. INTRODUCTION
The development of Al2O3–SiO2–SiC–C resin bonded
ramming monolithic refractory for iron making applications
(such as iron and slag runners in blast furnaces, furnace
bottom and electric-furnace spouts) [1] is a very important
recent refractory development which can be considered to
have evolved from the high performance chamotte–carbon
and bauxite–carbon refractories. This class of refractory not
only shows superior slag corrosion and erosion (wear)
resistance, but also excellent thermal shock resistance and
mechanical properties. Monolithic refractories containing
bauxite aggregate has shown higher densification and
mechanical properties due to the presence of some impu-
rities such as iron which help liquid phase sintering, and
consequently, improved mechanical properties [2].
Probably, the high glassy content of the matrix of chamotte
results in high surface tension and higher amounts of
needles of mullite in chamotte, and has caused more
nucleation of SiC whiskers at high temperature, conse-
quently, resulting in improved strength and higher density
[3]. These refractories consist mainly of aluminosilicate and
carbon, which are bonded by carbon bonds formed by
carbonisation of phenol formaldehyde-resin (resole) during
firing of the refractories [1,3].
Carbon does not wet molten metal and will not melt, that
is an advantage excellent for refractory use. Its great
weakness is that it will oxidise at low temperatures,
above 600�C, which creates high porosity, leading to
lower density and reduced strength [1,4]. In order to
improve the oxidation resistance of carbon-containing
refractories, antioxidants (such as aluminum and silicon
metal) have been often added [5]. However, by examining
the behaviour of Al in such refractories, the following
problems have been observed: (1) Low melting point Al
particles, gives decreasing refractoriness; (2) the reaction of
Al with water, therefore Al cannot be used in carbon-
containing monolithic refractories in which water-
containing binders are used. Based on the above considera-
tion, it is necessary to develop new antioxidants with
excellent hydration resistance and high melting point,
such as Si-containing antioxidants. Mostly, metallic
silicon and aluminum are added to the material as a
sintering agent to improve the mechanical strength and
abrasion resistance under oxidation. The effect of metallic
silicon grain size on the properties of alumina–carbon brick
has been studied, and it is reported that the added metallic
silicon reacts under the reducing atmosphere with carbon in
the brick and produces b-SiC bond which contributes to
improvements of the mechanical strength and the abrasion
MATERIALS AT HIGH TEMPERATURES 26(4) 000–000 1# 2009 Science Reviews 2000 Ltddoi: 10.3184/096034009X12570016241884
MAHT 090263
resistance under oxidation. The influence of b-SiC bond on
the mechanical strength has yet to be clarified [6].
Generally, the modulus of rupture vs. firing temperature is
increased at a given temperature (1100, 1200, 1300, 1500�C)
by increasing the grain size of silicon up to more than 100
microns, but the elastic modulus is decreased. The amount of
SiC formed decreases and the remaining Si increases for
coarse particles. It is concluded that the strength and
durability of the slide gate plates can be improved, not by
increasing the amount of silicon metal, but by decreasing the
grain size to enhance silicon reaction. The results of other
workers show that SiC additions in the range from 4 to 16%
are beneficial to the improvement of mechanical properties
as well as the thermal shock resistance [7]. From the
investigation of the thermodynamics and from observation
of the results after use, it is evident that the SiC materials
have a large influence on oxidation prevention. SiC beha-
viour in Al2O3–SiC–C bricks is illustrated in reactions below
[8].
SiCðsÞ þ COðgÞ ¼ SiOðgÞ þ 2CðsÞ ð1Þ
SiOðgÞ þ COðgÞ ¼ SiO2ðsÞ þ CðsÞ ð2Þ
The most successful and commonly used ramming mixes
were based on phenol-formaldehyde resins. The cured
strength of these products was high because of polymerisa-
tion of the resin, which is called a ‘resit’ structure, at about
200�C. At elevated temperatures, however, the products
become weak because of the oxidation in the carbonaceous
bond phase, which creates high porosity, lower density and
reduced strength [1].
Advantages of a resole binder is that it is environmental
friendly, easy to mix, with high residual carbon. A disad-
vantage of a resole binder is the pyrolization of the ‘resist’
structure, that is called polymer carbon, at temperatures
about 300 – 500�C, and the products became weak
(Figure 1). The usage of additives such as silicon and
ferrosilicon is supposed to reinforce the ‘resit’ structure by
increasing the cross linking. The cross linking phenomenon
was discovered by Charles Goodyear in 1839 and involved
heating the raw rubber with sulfur (the Vulcanisation
process). The mechanism appears to involve addition of
sulfur units to the double bonds in the rubber with the
production of cross links between the polymer chains [9].
Cross-linking reactions in general are those that lead to the
formation of insoluble and infusible polymers in which
chains are joined together to form a three dimensional
network structure. A sample cross linking reaction is exem-
plified by polymer chains with several functional groups,
designated A, that are capable of reacting among themselves
to form chemical bonds, A22A, (Figure 2).
If these polymer chains are exposed to conditions such
that the functional groups do react, then all the chains in the
reaction vessel will be tied to each other through A22A
bonds [10]. In principle, the polymer molecules in the
reaction vessel will have formed giant molecules. Cross-
linking changes and generally improves physical properties.
A chemical reaction that leads to formation of a chemical
bond can be used for cross linking of polymers if the
requirements of effective average functionally per molecule
are met. Phenol-formaldehyde resins including resoles and
novolacs are technologically useful cross-linking systems.
Cross linking agents must carry two or more functional
groups per molecule [10].
2. MATERIALS AND METHODS
Two series of compositions were prepared. The first compo-
sition (samples A) containing as aggregates a combination of
The effect of additives on Al2O3-SiO2-SiC-C Ramming Monolithic Refractories: S.M. Siadati et al.
2 MATERIALS AT HIGH TEMPERATURES www.scilet.com
Figure 1 Linear polymer and cross-linked polymer.
linear polymer
cross-linked polymer
���?
Figure 2 Effect of additives on chamotte based cured at 200�C for
2 h. (Samples A).
chamotte (0 – 3 mm, 65 wt.%), crushed sagger which is a
SiC-containing material regenerates (15 wt.%) and fine coke
(10 wt.%) with different amounts of additives such as Silicon
and ferrosilicon metal. The second one (samples B)
containing as aggregates a combination of bauxite
(20 wt.% coarse bauxite (1– 3 mm), 45 wt.% fine bauxite
(0 – 1 mm)), crushed sagger (15 wt.%) and fine coke
(10 wt.%) with different amounts of additives such as
Silicon and ferrosilicon metal. At low temperature of
curing (200�C), three other additives such as sulfur, carbon
black and silica fume are also used. Since silica fume and
carbon black were not so effective in increasing of CCS
values, they were eliminated from higher temperature tests
(1100�C and 1400�C). Chamotte, bauxite and sagger are
used to increase the strength of the refractory. Coke is used
to lower the wettability with the iron or steel melt. Both
series consisted of 10 wt.% resole (phenol formaldehyde
resin) as the binding agent. The chemical compositions and
phase contents of the raw materials are given in Tables 1 and
2 respectively.
Specimens were prepared by mixing 90 wt.% aggregate
and 10 wt.% resole as a binder with different amounts of
additives on top of 100 wt.%. A twin blade mixer was used
for dry mixing and kneading. The raw materials were dry
mixed for 5 min and then the resole liquid was added
gradually. Mixing continued for another 5 min. The
mixture was packed into metal moulds with a dimension of
50650650 mm and then formed by using a hydraulic press
under a low pressure of 4 MPa, to act similar to ramming.
Pressed samples were cured at 200�C for 2 h. Finally, some
of the samples were fired for 2 h at either 1100�C or 1400�Cin a coke environment to avoid an oxidising atmosphere. The
specimens cured or fired were characterised according to
ASTM standards. The following tests were carried out: bulk
density (BD), apparent porosity (AP), cold crushing strength
(CCS), also X-ray diffraction (XRD). Scanning electron
microscopy (SEM) were performed in some cases. CCS
was measured by using a Hydraulic Testing Machine, type
Amsler model D3010y2E.
BD and AP were determined by using the following
equations:
BD ¼ WyWf ð3Þ
AP ¼ ðWs �WÞyWf ð4Þ
where W, Ws and Wf respectively are the weight of dried
sample, saturated weight with water and weight of the
sample in water suspended by a thin thread without
contacting the vessel walls, but fully inside the water and
reading by a digital balance under the vessel. In fact Wf is the
weight reduction of the saturated sample in water, which is
equivalent to its volume.
An XRD instrument (Philips Xpert-MPD System) with Cu
Ka1(l1 ¼ 1:5405 A) and Cu Ka2(l2 ¼ 1:5443 A) and Ni
filter was used for phase analysis. A voltage of 30 kV and
current of 25 mA were employed. In order to study precisely
the phases, the XRD experiments were carried out with a rate
of 0.04 degreesysec between 5 – 80 degrees (2Y).
SEM Philips XL30 was used for microstructure investiga-
tions. X-ray fluorescence (XRF) was also used for elemental
analysis and oxides content calculations (Table 1).
3. RESULTS
3.1 Cold crushing strength and bulk density results
Figures 2 – 5 show the influence of additives on CCS of
resole bond, Al2O3–SiO2–SiC–C refractory. Figures 6 – 8
illustrate BD variations.
The effect of additives on Al2O3-SiO2-SiC-C Ramming Monolithic Refractories: S.M. Siadati et al.
www.scilet.com MATERIALS AT HIGH TEMPERATURES 3
Table 1 The chemical composition of raw materials (wt.%)
Material Al2O3 SiO2 CaO MgO TiO2 SiC C Si Fe S Grain size(mm)
Chinese bauxite 85 6.9 0.15 0.1 3.9 – – – – – 0 – 1,1– 3Iranian chamotte 42 53 0.5 0.5 – – – – – – 0 – 3Crushed sagger 45 15 – – – 38 – – – – 0 –3Fine coke 10 5 – – – – 80 – – – 0 – 0.3Ferrosilicon metal – – – – – – 2 72 23 – 0 – 0.15Silicon metal – – – – – – 1.0 97 1.0 – 0 – 0.15Silica fume – 92 – – – – – – – – 0 – 0.1Carbon black – – – – – – – 98 – – 0 – 0.1Sulfur powder – – – – – – – – – 97 0 – 0.3
Table 2 Phase analysis of raw materials (XRD results)
Material Major phase Medium phase Minor phase(s)
Chinese bauxite aAl2O3 Mullite(3Al2O3.2SiO2) TiO2, SiO2
Iranian chamotte Mullite(3Al2O3.2SiO2) – Cristobalite (SiO2)Crushed sagger aSiC Mullite(3Al2O3.2SiO2) a-Al2O3
Fine coke Graphite – Al2O3, SiO2
Ferrosilicon metal FeSi2 Si –Silicon metal Si – –
The effect of additives on Al2O3-SiO2-SiC-C Ramming Monolithic Refractories: S.M. Siadati et al.
4 MATERIALS AT HIGH TEMPERATURES www.scilet.com
Figure 3 Effect of additives on bauxite based cured at 200�C for
2 h. (Samples B).
Figure 4 Effect of additives on bauxite and chamotte based fired at
1100�C for 2 h.
Figure 5 Effect of additives on bauxite and chamotte based fired at
1400�C for 2 h.
Figure 6 Effect of additives on chamotte and bauxite based cured at
200�C for 2 h.
Figure 7 Effect of additives on chamotte and bauxite based fired at
1100�C for 2 h.
Figure 8 Effect of additives on chamotte and bauxite based fired at
1400�C for 2 h.
Figure 9 Micrograph of chamotte – carbon refractory resole bond
containing 6 wt.% ferrosilicon metal as additive, fired at 1100�C,
where small quantities of SiC whisker are observed.
3.2 Microstructure
Figures 9 and 10 show fired microstructures of 1100�C firing
for 2 h. Figures 11 and 12 show fired microstructures of
1400�C for 2 h.
3.3 Phase compositions
Phase compositions by XRD results are given in Figures 13
to 16.
4. DISCUSSION
The effect of additives on sample A containing Iranian
chamotte as the main aggregate (65 wt.%) is such that
silicon metal and ferrosilicon metal generally increase cold
crushing strength at all heat treatments of 200�C, 1100�Cand 1400�C and also increase bulk density. Sample B based
on Chinese bauxite also shows the CCS increases at 200�Cand 1400�C, but the CCS is reduced at 1100�C. SEM
photomicrographs have shown that SiC whiskers, of nano
sized diameter, have developed in sample A, but not in
sample B at 1100�C. The reason for the increase of strength
in sample A, which has not been clarified in some previous
studies [5], is now evident and is related to the development
of nano-sized whiskers of SiC within the microstructure. The
reason these develop only in sample A, may well depend on
the higher amount of mullite and glassy phase in chamotte,
when compared bauxite. Needles of mullite in a glassy
matrix phase provides good nucleation sites from which to
growth SiC whiskers at the low temperature of 1100�C. At
1400�C, the thermodynamic driving force is high enough for
the production of SiC whiskers in both samples A and B. But
the concentration of whiskers at a comparable level of
6 wt.% ferrosilicon metal (FeSi2) is more for the chamotte
matrix (Figure 11) than for the bauxite matrix (Figure 12). It
seems that these whiskers are formed on chamotte aggregates
as can be seen in Figure 11, but for the bauxite matrix the
whiskers are nucleated on the FeSi2 particles, spherical
particles seen in Figure 12, formed on the outside of the
bauxite particles. These figures support the hypothesis that
SiC whisker formation nucleates on mullite needles, formed
in the chamotte aggregates, due to the surface tension of the
glassy matrix. The reason for the strength increase at high
temperatures is evidently related to the formation of SiC
whiskers. However at 200�C the mechanism for CCS
increase cannot be related to SiC whisker formation but is
related to cross linking formation in the phenol-formalde-
hyde resole structure or resit. For this temperature, 5 different
additions of silicon metal, ferrosilicon metal, sulfur, silica
fume and carbon black were used. The first two were capable
of increasing the cross linking, but the other three were not
so effective (see Figures 2 and 3). The Si has a strong effect,
when in a small amount of 1%, due to the large number of
covalent bonds related to pure silicon. The effect is reduced
for 2% and above, probably due to an interfering effect of the
Si atoms. By considering Figure 1, it is probable that
formaldehyde (CH2O), which acts as cross linking agent of
A for the main monomer of phenol (C6H5OH), is interfered
with when a lot of Si atoms are present. Each is associated
The effect of additives on Al2O3-SiO2-SiC-C Ramming Monolithic Refractories: S.M. Siadati et al.
www.scilet.com MATERIALS AT HIGH TEMPERATURES 5
Figure 10 Micrograph of bauxite –carbon refractory resole bond
containing 5 wt.% silicon metal as additive, fired at 1100�C, which
shows that no SiC whiskers are formed.
Figure 11 Micrograph of chamotte –carbon refractory resole bond
containing 6 wt.% ferrosilicon metal as additive, fired at 1400�C,
showing SiC whiskers of about 75 nanometres in diameter in the
microstructure.
Figure 12 Micrograph of bauxite –carbon refractory resole bond
containing 6 wt.% ferrosilicon metal as additive, fired at 1400�C,
which shows SiC whiskers are developed from ferrosilicon metal
particles in the microstructure.
with covalent bonds, at the expense of, and hence
decreasing, the cross links of the phenol-formaldehyde
monomers. Formaldehyde is very reactive and readily poly-
merises with ionic initiation [11]. Either anionic or cationic
materials can initiate the polymerisation reaction.
�R þ CH2O?RCH2O�; ðR � O� ¼ CH2Þ ð5Þ
þR þ CH2O?RCH2OþðR � Oþ ¼ CH2 or R � O� CHþ2 Þ
ð6Þ
Ferrosilicon is less effective than silicon and shows a
maximum CCS at 2% and reduces the strength for 3% and
above. The mechanism of the effect is the same as for silicon.
Micro silica or silica fume (small amorphous spherical
particles of SiO2) and carbon black, amorphous carbon, do
not contribute to a stronger cross linking, probably due to
them having an amorphous and non crystalline nature and
the submicron size of these amorphous particles. Instead,
they have an interfering effect and reduce CCS. Almost the
same effects are observed when these additives are used in
bauxite based refractories of sample B (Figure 3). Results for
samples A and B provide confirmation to each other. The
role of sulfur, in increasing of CCS, was good to a degree
(Figures 2 and 3), but was not as effective as Si or
ferrosilicon. At high temperature, above 800�C, the resit
structure is destroyed and graphite phase, as a crystalline
carbon state, is formed. XRD results show higher peaks of
the graphite phase at 1400�C firing (Figures 13 and 14) when
compared to 1100�C firing (Figures 15 and 16).
5. CONCLUSIONS
1. The addition of silicon or ferrosilicon metal to alumina-
silicate-carbon ramming monolithics, with resole bond,
for Iron making applications, improves mechanical
properties and develop a high strength and therefore
more wear resistance for applications such as blast
furnace runners and troughs.
2. For low temperature heat-treatments, of 200�C, these
additions contribute to the formation of the resit structure
in resole, by increasing the cross linking, and provides
CCS values as high as about 65 MPa , which is remark-
able for a carbon containing monolithic.
3. There is a limit of 1 wt.% for silicon, and 2 wt.% for
ferrosilicon, to increase the cross linking phenomena and
thus raise the strength of the refractory. Beyond this the
additions interfere with the formaldehyde hence the
effectiveness of cross linking is reduced and the strength
is lowered.
4. Three additives, silica fume, carbon black and sulfur,
were not so effective as compared with silicon and
ferrosilicon metal.
5. SiC whiskers of nano-sized diameter are observed to
form at 1400�C in both samples containing chamotte
and bauxite. At 1100�C these whiskers are only
observed in the chamotte based refractory. It is proposed
that probably the higher glassy content of the matrix of
chamotte has caused a higher surface tension. This,
The effect of additives on Al2O3-SiO2-SiC-C Ramming Monolithic Refractories: S.M. Siadati et al.
6 MATERIALS AT HIGH TEMPERATURES www.scilet.com
Figure 13 XRD results of fired chamotte – carbon refractory, resole
bond. (1400�C, 2 h).
NOTA
S COP
Y SUPPLI
ED
Figure 14 XRD results of fired bauxite – carbon refractory, resole
bond. (1400�C, 2 h).
NOTA
S COPY
SUPPLI
ED
Figure 15 XRD results of fired chamotte – carbon refractory, resole
bond. (1100�C, 2 h).
NOTAS COPY SUPP
LIED
Figure 16 XRD results of fired bauxite – carbon refractory, resole
bond. (1100�C, 2 h).
NOTAS COPY SUPP
LIED
together with the higher amounts of needles of mullite
formed in chamotte when compared to bauxite, has
caused more nucleation of SiC whiskers. The tempera-
ture of 1100�C is a transient temperature, where any
cross linking is destroyed yet SiC whisker formation is
not fully formed.
REFERENCES
[1] Bosall, S.B. and Henry, D.K. (1985) New developments in
monolithic refractories. The American Ceramic Society,
Westerville, OH, Vol. 13, pp. 331 –340.
[2] Zawrah, M.F. (2007) Effect of zircon additions on low and
ultra-low cement alumina and bauxite castables. Ceramics
International, Vol. 33, no. 5, pp. 751 – 759.
[3] Chan, C.F., Aregent, B.B. and Lee, W.E. (1988) Influence of
additives on slag resistance of Al2O–SiO2–SiC–C refractory
bond phase under reducing atmosphere. J. Am. Ceram. Soc.,
No. 81, pp. 3177– 3188.
[4] Parsons, J.R., Forest, P. and Rechter, H.L. (1974) Refractory
composition comprising coarse particles of clay or bauxite and
carbon. U.S. Patent 3 810 768.
[5] Yamaguchi, K., Zhang, S. and Hashmato, J.YU. (1991)
Behavior of antioxidant added to carbon-containing refrac-
tories, J. Ceram. Soc. Japan, pp. 341 – 348.
[6] Akamine, K., Suruga, T., Yoshitomi, J. and Asona, K. (2006)
Effect of metallic silicon size and heat temperature on proper-
ties of Al2O3–C brick. Internationals Feuerfest-kolloquium,
Aachen.
[7] Fangbao, Ye, Rigaud, M., Liu, X. and Zhong, X. (2004)
High temperature mechanical properties of bauxite-based
SiC-containing castable. Ceramics International, Vol. 30,
pp. 801 – 805.
[8] Nishio, H. and Matsuo, A. (1995) Al2O3–SiC–C Bricks for
Torpedo Car, Taikabutsu Overseas, Vol. 15, No. 4, pp. 39 – 46.
[9] Streitwieser, Jr. A. and Heathcock, C.H. (1981) Introduction to
organic chemistry, second edition. Macmillan Publishing Co.
Inc., USA, pp. 595, Chapter 20 (conjugation).
[10] Kroschwitz, J.I. (1990) Concise encyclopedia of polymer
science and engineering. John Wiley & Sons, Inc., Cross
linking, pp. 213 – 219.
[11] Ibid, page 4.
The effect of additives on Al2O3-SiO2-SiC-C Ramming Monolithic Refractories: S.M. Siadati et al.
www.scilet.com MATERIALS AT HIGH TEMPERATURES 7