Steel Plant

Chapter 3
Steel Plant Refractories

D.H. Hubble, Chief Refractory Engineer, U.S. Steel Corp. (Retired)
3.1 Classification of Refractories

Refractories are the primary materials used by the steel industry in the internal linings of furnaces
for making iron and steel, in vessels for holding and transporting metal and slag, in furnaces for
heating steel before further processing, and in the flues or stacks through which hot gases are con-
ducted. At the risk of oversimplification, they may, therefore, be said to be materials of construc-
tion that are able to withstand temperatures from 260–1850°C (500–3400°F).
As seen in Fig. 3.l, the melting point of refractory materials in the pure state varies from
1815–3315°C (3300°–6000°F). Refractories in service can tolerate only small amounts of melting
6300 6300 – Graphite, C, pure

6000 5430 – Thoria, ThO2, pure sintered
5070 – Magnesia, MgO, pure sintered
5700 4890 – Zirconia, ZrO2, pure sintered
5400 4620 – Beryllia, BeO, pure sintered
5100 4080 – Silicon Carbide, SiC, pure
3980 – Magnesia, 90-95%
4800
Melting temperature (°F)
3960 – Chromite, FeO-Cr2O3

4500 3880 – Chromium Oxide
3720 – Alumina, Al2O3, pure sintered
4200 3580 – Chromite, 38%, Cr2O3
3900 3400 – Alumina, fused Bauxite
3600 3400 – Silicon Carbide, 80-90%
3400 – Fireclay
3300 3300 – Kaolin, Al2O3-SiO2
3000 2802
2700 2651
2400
2100 1981
1800
1500 1220
1200
900 786
621
600 448
300
Refractory
materials
Aluminum
Tin
Iron
Zinc
Lead
Nickel
Copper
Key industrial materials

Fig. 3.1 Refractory and industrial materials melting point chart.
Copyright © 1999, The AISE Steel Foundation, Pittsburgh, PA. All rights reserved. 161
Ironmaking Volume
(1–5%) without loss of their important structural characteristics. Subsequent discussion will show,
however, that the use of many such materials is limited by factors such as cost or instability in cer-
tain atmospheres. Also, fluxes present in the initial impure refractory and/or encountered in ser-
vice can seriously reduce these melting points.
Refractories are expensive, and any failure in the refractories results in a great loss of production
time, equipment, and sometimes the product itself. The type of refractories also will influence
energy consumption and product quality. Therefore, the problem of obtaining refractories best
suited to each application is of supreme importance. Economics greatly influence these problems,
and the refractory best suited for an application is not necessarily the one that lasts the longest, but
rather the one which provides the best balance between initial installed cost and service perfor-
mance. This balance is never fixed, but is constantly shifting as a result of the introduction of new
processes or new types of refractories. History reveals that refractory developments have occurred
largely as the result of the pressure for improvement caused by the persistent search for superior
metallurgical processes. The rapidity with which these ever recurring refractory problems have
been solved has been a large factor in the rate of advancement of the iron and steel industry. To dis-
cuss the many factors involved in these problems and to provide information helpful to their solu-
tion are the objectives of this chapter.
Refractories are also vital in the safe operation of the processes and must not expose personnel to haz-
ardous conditions during their manufacture, installation, use or during disposal following their use.
Refractories may be classified in a number of ways. From the chemical standpoint, refractory sub-
stances, in common with matter in general, are of three classes; namely, acid, basic, and neutral.
Theoretically, acid refractories should not be used in contact with basic slags, gases or fumes
whereas basic refractories can be best used in contact with a basic chemical environment. Actually,
for various reasons, these rules are often violated. Hence, the time honored chemical classification
is largely academic, and of little value as a guide to actual application. Also, the existence of a truly
neutral refractory may be doubted. Classifications by use, such as blast furnace refractories or
refractories for oxygen steelmaking, are generally too broad and are constantly subject to revision.
For our purposes, refractories will be classified with reference to the raw materials used in their
preparation and to the minerals predominating after processing for use. This classification is
believed to offer the best possibility for a clear understanding of the origin and nature of steel plant
refractories.
3.1.1 Magnesia or Magnesia–Lime Group

This group includes all refractories made from synthetic magnesites and dolomite. These consti-
tute the most important group of refractories for the basic steelmaking processes. All these mate-
rials are used primarily as a source of magnesia (MgO).
3.1.1.1 Magnesia
Modern high-purity magnesias are produced in well controlled processes. The principal sources of
magnesias are brines (often deep well type) and seawater. Magnesium hydroxide, Mg(OH)2, is pre-
cipitated from these sources by reaction with calcined dolomite or limestone; one source uses a novel
reactor process. The resultant magnesium hydroxide slurry is filtered to increase its solids content.
The filter cake can then be fed directly to a rotary kiln to produce refractory grade magnesia, but
more commonly now the filter cake is calcined at about 900–1000°C (1650 –1830°F), usually in mul-
tiple-hearth furnaces, to convert the magnesium hydroxide to active magnesia. This calcined magne-
sia is then briquetted or pelletized for firing into dense refractory-grade magnesia, usually in shaft
kilns which reach temperatures around 2000°C (3630°F). The end product is sintered magnesia.
Fused magnesia is produced by melting a refractory grade magnesia or other magnesia precursor
in an electric arc furnace. The molten mass is then removed from the furnace, cooled, and broken
up to begin its path for use in refractories.
162 Copyright © 1999, The AISE Steel Foundation, Pittsburgh, PA. All rights reserved.
The impurities in magnesia are controlled by the composition of the original source of the magne-
sia (brine or seawater), the composition of the calcined dolomite or limestone, and the processing
techniques. In particular the amounts and ratio of CaO and SiO2 are rigorously controlled, and the
B2O3 is held to very low levels. The end results are high-grade refractory magnesias which are
ready for processing into refractory products. Tables 3.1 through 3.4 show the compositions of dif-
ferent grades of magnesia.
Table 3.1 Selected Sintered Magnesias Produced in North America

(high CaO/SiO2 ratio)
Reference Code SM-1 SM-2 SM-3
Chemical Analysis (wt%)

CaO 2.2 2.2 0.8
SiO2 0.7 0.35 0.1
Al2O3 0.1 0.20 0.1
Fe2O3 0.2 0.20 0.1
B2O3 0.015 0.02 0.005
MgO 96.7 96.3 98.8
CaO/SiO2 ratio 3.1 6.3 8.0

Bulk Density (kg/m3) 3400 3420 3400
Average Crystallite
Size (mm) , 80 , 90 , 100
From Ref. 2.
Table 3.2 Selected Sintered Magnesias Produced in Europe and the Middle East
(high CaO/SiO2 ratio)

CaO 1.90 0.65 0.7
SiO2 0.20 0.15 0.03
Al2O3 0.05 0.06 0.03
Fe2O3 0.20 0.50 0.04
B2O3 0.015 0.008 0.005
MgO 97.5 98.5 99.2

Bulk Density (kg/m3) 3440 3450 3430–
3450
Average Crystallite
Size (mm) 150–160 , 100 70-90
From Ref. 2.
Ironmaking Volume
Table 3.3 Selected Fused Magnesia Produced in North America

Reference Code FM-1 FM-2

CaO 0.95 1.8
SiO2 0.30 0.5
Al2O3 0.11 0.2
Fe2O3 0.16 0.5
B2O3 0.0025 0.01
MgO 98.50 97.0
CaO/SiO2 ratio 3.2 3.6
Bulk Density (kg/m3) 3530 3500
Average Crystallite
Size (mm) , 780 , 800
From Ref. 2.
Table 3.4 Selected Sintered Magnesias Produced in North America

(low CaO/SiO2 ratio)

CaO 0.7 0.80 0.9
SiO2 0.7 0.60 2.1
Al2O3 0.1 0.20 0.15
Fe2O3 0.2 0.20 0.35
B2O3 0.1 0.10 0.23
MgO 98.2 98.05 96.2
Bulk Density (kg/m3) 3330 3340 3260
Average Crystallite
Size (mm) 50-60 , 60 , 40
From Ref. 2.
High purity is quite important because MgO has high refractoriness and good resistance to basic
slags. Minimizing the total impurities content in magnesias is quite important because impurities
affect refractoriness and performance. A high CaO/SiO2 ratio, preferably 2:1 or slightly higher, is
optimum for maintaining high refractoriness in magnesias. As the total impurities are reduced to
about 1% or less, the CaO/SiO2 ratio has less and less significance. High density reduces infiltra-
tion and dissociation of magnesia grain by slag. Large crystallite size (best achieved in fused mag-
nesias) provides less surface area for slag attack. Low B2O3 content ensures high strength at elevated
temperatures for burned brick. Low lime to SiO2 ratios are required in certain products where MgO
is used with other raw materials (such as Cr2O3) or for maximum resistance to hydration.
3.1.1.2 Dolomite
The natural double carbonate dolomite (CaCO3•MgCO3) can be converted to refractory dolomite
(CaO•MgO) by high temperature firing. A limited number of dolomite deposits exists in the world
with satisfactory uniformity, purity, and calcining behavior to be processed into high purity, refrac-
tory dolomite at a reasonable cost. High purity dolomite is greater than 97% CaO + MgO and
0.5–3% impurities. Most high purity dolomite deposits are difficult to calcine and sinter to high
density and usually require special methods to yield acceptable refractory grade dolomite. Silica,
iron oxide and alumina are the most common impurities in high purity dolomite. See Table 3.5.
Table 3.5 Chemical and Physical Properties of High Purity, Dead-Burned Dolomite
Chemicial Analysis (wt%)

_________________________ Physical Properties
____________________
Al2O3 0.45 Bulk Density (kg/m3) 3250
Fe2O3 0.90 Closed pore volume (%) 1.3
SiO2 0.70
MgO 41.20
CaO 56.70
From Ref. 2.
Dolomite has excellent refractoriness and is thermodynamically very stable in contact with steel or
steelmaking slags. Note in Table 3.6 that CaO is the most stable of the common refractory oxides
at steelmaking temperatures.
The free lime portion of the dead burned
Table 3.6 The Free Energies of dolomite can react with atmospheric moisture
Formation for Various which causes the material to powder and
Refractory Oxides crumble. The degree of hydration under set
conditions of time, temperature, and relative
DG @ 1600°C
humidity is dependent upon the proportion of
1 2
Kcal
}} lime and impurities contained in the material
mole O2
2 Ca + O2 = CaO –205 and upon the density of the grain achieved
4
⁄3 Al + O2 = ⁄3 Al2O3
2
–175 during the dead burning process.
2 Mg + O2 = 2 MgO –170 In practice, with modern packaging materials
Si + O2 = SiO2 –140 and techniques together with other means of
4
⁄3 Cr + O2 = 2⁄3 Cr2O3 –110 protecting the products, the storage of dolo-
From Ref. 2. mite products can be extended.
3.1.2 Magnesia–Chrome Group

Chrome ores, or chromites, often called chrome enriched spinels, are naturally occurring members
of the spinel mineral group. These materials are all characterized by relatively high melting points,
good temperature stability (particularly in thermal union with magnesite) and moderate thermal
expansion characteristics.
Chrome ores are covered by the formula (Mg•Fe+2) (Cr,Al,Fe+3)2O4ʹ where magnesium can substi-
tute for iron, and aluminum for chromium. Accessory minerals, called gangue, are often associated
with these ores. Gangue minerals contribute residual silica, lime and additional magnesia to the
chrome ore. Control, placement and quantity of gangue minerals is as important to finished refrac-
tory properties as the basic chemical composition of the individual ore. Ores are often upgraded
by mechanical separation techniques to minimize impurities. Thus, chrome concentrates with low
gangue impurities, as shown in Table 3.7, are often the article of commerce used in refractories.
Pure Cr2O3 chemically separated from chromite can be used in combination with lower cost oxides
to produce specific refractory properties.
In addition to basic refractory raw materials composed of dead burned magnesia and chrome ore
as starting materials, other magnesite-chrome combinations are also a part of this series. A mag-
nesite-chrome group of raw materials exists, including co-sintered magnesite-chrome, fused mag-
nesite-chrome and synthetic picrochromite—a combination of magnesite and chromic oxide.
Typical magnesite-chrome combinations appear in Table 3.8.
Ironmaking Volume
Table 3.7 Examples of Refractory-Grade Chrome Ores (typical data)

Philippines South
Source –10 concentrates Africa Turkey
Typical Chemistry
Silica (SiO2) 3.20% 0.70% 3-4%
Alumina (Al2O3) 28.10 14.80 20-21
Iron Oxide (Fe2O3) 15.20 28.60 15-16
Lime (CaO) 0.28 0.05 0.30
Magnesia (MgO) 17.80 10.80 18.00
Chromic Oxide (Cr2O3) 35.00 47.10 41.42
Size 12% min. +14 mesh –20+150 –10+100

20% max. – 65 mesh or Lump
Structure Hard, Massive Friable
From Ref. 3.
Table 3.8 Examples of Magnesite-Chrome Raw Materials (typical data)

South Africa Canada/U.S.
Source U. Kingdom Fused Fused
Co-Sintered Mag. Chrome Mag. Chrome
Typical Chemistry
Silica (SiO2) 1.15% 1.60% 1.2%
Alumina (Al2O3) 5.70 7.20 6.1
Titania (TiO2) 0.20 0.20 0.3
Iron Oxide (Fe2O3) 12.50 10.70 13.0
Lime (CaO) 0.80 0.60 0.4
Magnesia (MgO) 61.50 59.90 60.1
Chromic Oxide (Cr2O3) 18.00 18.80 18.6
Bulk Specific Gravity (g/cm3) 3.53 3.75 3.8

From Ref. 3.
Table 3.9 Examples of Refractory-Grade Silicas 3.1.3 Siliceous Group

(typical data) 3.1.3.1 Natural
Natural Silica Fused Natural silica occurs primarily as the
Type (Quartzite) Silica mineral quartz. Heat treatment of sil-
Typical Chemistry
ica in the manufacture of refractory
products can result in the formation
Silica (SiO2) 99.70% 99.60%
of a number of different crystalline
Alumina (Al2O3) 0.09 0.20 forms. Depending on the thermal his-
Titania (TiO2) 0.01 — tory applied, silica refractories may
Iron Oxide (Fe2O3) 0.09 0.03 contain various mineral assemblages
Lime (CaO) 0.03 of quartz, cristobalite, and tridymite.
0.04
Magnesia (MgO) 0.01 Silica raw materials used in the man-
Total Alkalies 0.02+ 0.01 ufacture of conventional silica refrac-
tories, as shown in Table 3.9, must
Bulk Specific Gravity (g/cm3) 2.33 2.20 contain high silica (99% or above
SiO2) and low impurity levels, partic-
Major Minerals
ularly alumina and alkalies, which
Quartz X can act as fluxes during firing of the
Vitreous or Glassy X refractory and reduce overall refrac-
From Ref. 3. toriness of the end product.
3.1.3.2 Fused Silica

Fused silica is produced by actual fusion of specially selected, very high grade silica sands in elec-
tric arc, electrical resistance, or other furnace procedures. Crystalline raw material is converted into
an amorphous glass, or fused silica. Properties of this fused raw material vary considerably from
those of the original quartz sand, in particular fused silica has very low thermal expansion. Fused sil-
ica products exhibit low thermal conductivity, high purity and excellent resistance to thermal shock.
3.1.3.3 Silicon Carbide

Commercial silicon carbide (SiC) used as a refractory raw material is manufactured by abrasive
grain producers in electric furnaces from a mixture of coke and silica sand. The finished material
is extremely hard (9.1 MOH’s scale) with high thermal conductivity and good strength at elevated
temperatures, as well as very good resistance to thermal shock. Silicon carbide dissociates at
2185°C (3965°F) and oxidizes slowly in air, but is relatively stable under reducing conditions. The
material is serviceable at 1535–1650°C (2800–3000°F) for many applications. Table 3.10 shows
some analysis of refractory grade silicon carbides.
Table 3.10 Examples of Refractory-Grade Silicon Carbides
Coarse, 6/F 200/F 10/F

Type Low Iron Low Iron
Typical Chemistry
(Dry Basis)
Silica (SiO2) 3.50% 1.90% 2.40%
Alumina (Al2O3) 0.50 0.40 0.76
Titania (TiO2) 0.05 0.06 0.07
Iron Oxide (Fe2O3) 0.24 0.45 0.93
Lime (CaO) 0.12 0.20 0.49
Magnesia (MgO) 0.02 0.03 0.29
Alkalies <0.04 <0.04 —
Silicon Carbide (SiC) 94.90 96.80 91.70
Total Carbon 29.10 29.20 30.70

Free Carbon 0.60 0.30 3 20
From Ref. 3.
3.1.3.4 Zircon/Zirconia
Zircon, or zirconium silicate (ZrO2•SiO2), is a naturally occurring raw material having excellent
refractoriness. Specific gravity (4.5–4.6 g/cm3) is unusually high compared to most refractory
materials.
Major zircon sources include the natural sands of western Australia, eastern Florida, South Africa’s
northern Natal east coast, the European Economic Community countries and the Peoples Repub-
lic of China. Zircon usually is found with other heavy mineral sands, most notably titania miner-
als. Table 3.11 shows properties of zircon.
Zirconium oxide (ZrO2) is produced commercially as the naturally occurring mineral baddeleyite.
The refractory industry has been a major growth area for zirconia. The relatively high melting point
of baddeleyite, along with superior resistance to corrosion and erosion, make zirconia an ideal
component for several refractory systems. Zirconia in the natural state occurs in the monoclinic
crystal phase. When heated, zirconia undergoes a phase change (to tetragonal) and a volume
shrinkage of about 5%; with additional increases in temperature, the cubic form is stable at 2350°C
(4260°F). To counteract these deleterious phase changes, zirconia may be stabilized to the cubic
phase with small amounts of calcia, magnesia, or yttria, with the result that stability, thermal shock
Ironmaking Volume
Table 3.11 Refractory-Grade Zircons (typical data)

Zircon Zircon Zircon
Type Sand –200 Mesh –400 Mesh
Typical Chemistry
(Calcined Basis)
Silica (SiO2) 32.60% 33.60% 33.70%
Alumina (Al2O3) 0.80 0.50 0.50
Titania (TiO2) 0.20 0.10 0.20
Iron Oxide (Fe2O3) 0.03 0.02 0.01
Lime (CaO) 0.03 0.03 0.03
Magnesia (MgO) 0.03 0.02 0.02
Zirconia (ZrO2) 66.20 65.60 65.60
Screen Analysis
+200 Mesh — 2 0
+325 Mesh — 11 1
+400 Mesh — 7 5
–400 Mesh — 79 95
From Ref. 3.
resistance, and hot load properties are

Table 3.12 Refractory-Grade Zirconia
enhanced in the final product. Zirconia
(typical data)
is available in the natural, stabilized, or
Baddelevite Zirconia fused state, the latter often as a mixture
(natural mineral Fused with alumina, silica, or other compati-
Type concentrate) (lime stabilized) ble oxides. Table 3.12 shows typical
Typical Chemistry properties of zirconia.
(Calcined Basis)
Silica (SiO2) 0.36% 0.37%
Alumina (Al2O3) 0.01 0.50
3.1.4 Clay and High-Alumina
Titania (TiO2) 0.30 0.31 Group
Iron Oxide (Fe2O3) 0.20 0.09 3.1.4.1 Clays
Lime (CaO) 0.01 3.98
Magnesia (MgO) 0.05 0.27 Although clays were among the first
Zirconia (ZrO2) 99.20 94.46 raw materials used to make refracto-
Soda (Na2O) 0.02 0.02 ries, their usage has diminished as
Potash (K2O) <0.01 <0.01 demands placed on modern refracto-
ries have necessitated better perform-
Lithia (Li2O) <0.01 <0.01
ing materials to replace them. Never-
Loss on Ignition 0.11 —
theless, clays are still an important
material in the industry.
From Ref. 3.
Clays may be used as binders, plasticizers, or as aggregates for producing refractories. The prop-
erties of some domestic clays are shown in Table 3.13.
3.1.4.2 Bauxitic Kaolins

Several other types of 50–70% alumina raw materials are also used in refractories. Bauxitic
kaolins, or bauxite clay combinations, represent another class of natural aluminum silicates used
in refractory manufacture. Major U.S. commercial deposits of aluminous refractory raw materials
are found in southeastern Alabama, and in east-central Georgia. These materials, shown in Table
3.14, are mined, blended and fired in rotary kilns to yield a versatile line of calcined aluminum sil-
icates, ranging from about 50% to over 70% alumina content.
Table 3.13 Some Examples of Refractory Clays (typical data)

Locale Missouri Missouri Kentucky Georgia
Type Flint Plastic Mine Run Kaolin Ball Clay Bentonite
Typical Chemistry
(Calcined Basis)
Silica (SiO2) 50.0% 59.5% 51.4% 52.1% 62.5% 67.7%
Alumina (Al2O3) 45.3 32.3 42.5 44.5 31.4 20.4
Titania (TiO2) 2.4 1.6 2.6 1.8 1.5 0.2
Iron Oxide (Fe2O3) 0.8 2.2 1.8 0.6 2.5 4.5
Lime (CaO) 0.2 0.4 0.2 0.3 0.5 1.1
Magnesia (MgO) 0.2 0.8 0.5 0.1 0.7 2.3
Alkalies (total) 0.6 3.1 1.4 0.3 0.7 3.5
L.O.I. (dry basis) — 8.3 12.7 13.7 10.3 5.4

B.S.G. (g/cm3) 2.57–2.62
Bulk Density (Pcf) 148 130 143 — — —
Linear Change,
after Cone 15 (in/ft.) 0.5–1.1 0.75–1.25 0.75–1.0 — — —
PCE (Cone) 34+ 29–31 321⁄2–33 34+ 311⁄2 13+
From Ref. 3.
Table 3.14 Domestic U.S. Bauxitic Kaolins
Typical Chemistry Ucal 50 Ucal 60 Ucal 70

Silica (SiO2) 47.30% 36.90% 25.70%
Alumina (Al2O3) 49.20 59.20 70.00
Titania (TiO2) 2.40 2.60 3.20
Iron Oxide (Fe2O3) 1.00 1.10 1.00
Lime (CaO) 0.02 0.02 0.02
Magnesia (MgO) 0.04 0.05 0.04
Alkalies (total) 0.08 0.11 0.07
Specific Gravity (g/cm3) 2.60 2.70 2.78
Pyrometric Cone Equivalent 35–36 37–38 38–39

From Ref. 3.
3.1.4.3 Sillimanite
Andalusite, sillimanite and kyanite comprise the water-free, natural aluminum silicate varieties of
minerals known as the sillimanite group. Andalusite and kyanite are the more common commer-
cial materials. These minerals are normally about 60% alumina, with the balance composed pri-
marily of silica with minor iron and titania impurities, Table 3.15. Andalusite and sillimanite have
several important characteristics; when heated at high temperatures, the refractory mineral mullite
(3AL2O3•2SiO2) is formed. Complete mullite occurs at 1300–1400°C (2372–2552°F). This min-
eral is a key component of many high-alumina materials.
3.1.4.4 Bauxite
Bauxite in the crude state is a naturally occurring group of minerals composed primarily of either
gibbsite (Al2O3•3H2O), diaspore, or boehmite [AlO(OH)], and various types of accessory clays.
Refractory grade calcined bauxites are a specific form as found in Table 3.16. These are produced
from low iron, low silica materials in rotary kiln calcining operations or down-draft kilns. Calcin-
ing temperatures are in the 1400–1800°C (2550–3275°F) range. Crude bauxite is converted to the
Ironmaking Volume
Table 3.15 Examples of Andalusite and Kyanites
Andalusite Kyanite
Source S. Africa A S. Africa B French A United States
Typical Chemistry
Silica (SiO2) 38.00% 38.10% 43.80% 39.90%
Alumina (Al2O3) 60.30 61.90 53.00 56.00
Titania (TiO2) 0.17 0.16 0.25 1.80
Iron Oxide (Fe2O3) 0.72 0.59 1.20 1.60
Lime (CaO) 0.11 0.06 0.20 0.04
Magnesia (MgO) 0.09 0.11 0.15 0.04
Soda (Na2O) 0.06 0.06 0.20
Potash (K2O) 0.22 0.13 0.20 0.16 total
Lithia (Li2O) 0.02 0.01 —
L.O.I. (dry Basis) 0.44 0.28 0.08 0.90
Screen Analysis
+6 Mesh 17% 77% 35 mesh or
–6 +10 Mesh 34 14 100 mesh 3
–10 +16 Mesh 44 9 down
–16 Mesh 5 —
From Ref. 3.
Table 3.16 Refractory-Grade Calcined Bauxites (typical data)
Guyana, United States

Source S.A. China Brazil (Diaspore)
Typical Chemistry
(Calcined Basis)
Silica (SiO2) 6.50% 5.60% 9.0–10.0% 16.8%
Alumina (Al2O3) 88.00 87.50 85.0–87.0 75.7
Titania (TiO2) 3.25 3.64 1.9–2.3 3.8
Iron Oxide (Fe2O3) 2.00 1.56 1.6–2.1 0.9
Lime (CaO) 0.02 0.06 <0.2 0.3
Magnesia (MgO) 0.02 0.17 <0.2 0.5
Alkalies (total) 0.01 0.62 <0.1 2.3
Loss in Ignition 0.25 trace <0.1 trace
Bulk Specific Gravity (g/cm3) 3.10 min. 3.20 3.15–3.20 2.74
From Ref. 3.
minerals corundum (Al2O3) and mullite (3Al2O3•2SiO2) — both very refractory components.
Important features of bauxites are maximum alumina values (85% or more desired), maximum
bulk specific gravity, and minimum impurities such as iron oxide, titania, alkalies (Na2O,K2O and
Li2O) and alkaline earths (CaO and MgO). Major sources for calcined, refractory grade bauxite are
Guyana, South America, and the Peoples Republic of China.
3.1.5 Processed Alumina Group

Several types of chemically and thermally processed aluminas are used in refractories. These
include calcined, tabular, and fused alumina all from the Bayer process as illustrated in Fig. 3.2 and
Fig. 3.3.
In general, calcined aluminas are used to promote refractory binding during manufacture or use
whereas tabular or fused products form very stable aggregates. Tabular alumina is formed by cal-
cination at 1925°C (3500°F), whereas fused alumina is more dense after total melting and rapid
solidification. Tables 3.17 through 3.19 show properties of these alumina grades.
Bauxite
Size reduction
Make-up
NaOH
Strong Extraction Impurity

of separation
liquor aluminum hydroxides and washing
Residue
Concentration Precipitation
by of
evaporation crystalline Al(OH)3
Seed
Weak Product
separation
liquor and washing
Al(OH)3
Dehydration
by
calcination
Alumina
(Al2O3)
Fig. 3.2 Schematic of Bayer process.
Ironmaking Volume
Primary Secondary Market Application

process process
Bayer process
refinery
Aluminum Drying Industrial chemicals Glass and ceramics
hydroxide
Calcination
Auto catalyst substrates

Wear-resistant ceramic
Grinding Ceramics, polishing Whitewares
and refractories Spark plugs
Porcelain insulators
Substrates
Polishing
Refractories
Kiln furniture
Pelletizing
Ceramic tile
Glass
Translucent ceramic
High-purity Cutting tools
Sintering aluminas Biomedical ceramic
Synthetic gems
Calcined Lens polishing
aluminas
Catalyst bed support
(Spheres) Industrial chemicals Filter media
Heat exchange media
Crushing
Refractories Kiln furniture
Refractories
Sizing
Catalyst bed support
Industrial chemicals
Abrasives
(Grain)
Fillers and coatings Electrode coatings

Resin filler
Ruby abrasive grain

Abrasives Pink abrasive grain
Fusion
White abrasive grain
Primary Refractories Fused cast
aluminum Fibers
smelters
Fig. 3.3 Products of Bayer Process.
3.1.6 Carbon Group

Various carbon forms are used to an ever increasing extent in refractories. For example, modern
refractories use various graphite forms in combination with oxides to impart special properties.
The graphite may be synthetic in nature as produced by heating calcined petroleum coke to 3000°C
(5400°F) or may be natural graphite(s) from China, Mexico, Canada etc. Some all-carbon or all-
graphite refractories may be produced for applications in highly reducing atmospheres.
Table 3.17 Calcined Aluminas (typical data)
Hydrated Intermediate Intermediate Low

Soda Soda Soda
Fine Ground Super Ground
Source Calcined A Calcined B Calcined C
Typical Chemistry
Silica (SiO2) 0.01% 0.03% 0.05% 0.04%
Alumina (Al2O3) 65.00 99.40 99.80 99.70
Titania (TiO2) — 0.03 0.01 <0.01
Iron Oxide (Fe2O3) 0.004 0.04 0.02 0.04
Lime (CaO) — 0.06 0.20 0.04
Magnesia (MgO) — 0.02 0.03 0.01
Total Alkalies (Na2O) 0.15 0.19 0.23 0.08
Loss on Ignition (LOI) 34.50 0.21 0.10 0.03
Ultimate Crystal Size,(mm) — 3 to 5 — 1.50
Surface Area (m2/g) 0.15 0.50 — 1.00
From Ref. 3.
Table 3.18 Tabular Alumina (typical data)

Source United States United States
Typical Chemistry
Silica (SiO2) 0.04% 0.04%
Alumina (Al2O3) 99.70+ 99.80+
Titania (TiO2) 0.01 0.01
Iron Oxide (Fe2O3) 0.06 0.03
Lime (CaO) 0.04 0.01
Magnesia (MgO) <0.001 0.02
Boron Oxide (B2O3) <0.001 —
Soda (Na2O) 0.16 0.05
Bulk Specific Gravity (g/cm3) 3.55 3.54

From Ref. 3.
Table 3.19 Fused Aluminas (typical data)
Typical Chemistry Brown1 White2 White3

Silica (SiO2) 0.40% 0.04% 0.14%
Alumina (Al2O3) 96.70 99.50 99.41
Titania (TiO2) 2.52 — <0.01
Iron Oxide (Fe2O3) 0.10 0.10 0.02
Lime (CaO) 0.05 0.05 0.02
Magnesia (MgO) 0.18 — 0.02
Soda (Na2O) 0.10 0.30 0.39
Potash (K2O) — — 0.01
Bulk Specific Gravity (g/cm3) 3.75 3.77 3.50*

1 General Abrasives – Treibacher, Inc. 2 Washington Mills 3 C.E. Minerals * bulk density
From Ref. 3.
Ironmaking Volume
Generally, graphites are used in refractories in order to reduce the wetting characteristics of the
refractory material with respect to slag corrosion and to increase the thermal conductivity which
will result in better thermal shock resistance. In oxide-carbon refractories, the carbon content may
range anywhere from as low as 4–5% up to as high as 30–35%. Note that as the graphitic content
increases, the thermal conductivity of the refractory increases, but the density of the refractory
decreases. This result is primarily due to the fact that the density of graphite is much less than the
density of the other refractory materials being used. There are other contrasting differences in the
morphology of the graphite as compared to the other refractory materials. The graphite materials,
which are used in refractories, are commonly of a flaky structure; therefore, these flakes do not
lend themselves to the same particle packing phenomena as do granular particles.
Table 3.20 shows the properties of several types of graphite. Flake graphite is commonly purified to
extremely high carbon contents of 99 wt% carbon or higher. This purification utilizes both a chemi-
cal process and a thermal process; these steps have a significant effect on the price of the materials.
Table 3.20 Typical Properties of Graphite

Amorphous Flake High crystal- Primary Secondary
line artificial artificial
Carbon (wt%) 81.00 90.00 96.70 99.90 99.00
Sulfur (wt%) 0.10 0.10 0.70 0.001 0.01
True Density (kg/m3) 2310 2290 2260 2250 2240
Graphite Content (wt%) 28.0 99.9 100.0 99.9 92.3
d-Spacing (002) (nm) 33.61 33.55 33.54 33.55 33.59
Ash True Density (kg/m3) 2680 2910 2890 2650 2680
Resistivity (ohm-m) 0.00091 0.00031 0.00029 0.00035 0.00042
Morphology Granular Flaky Plates Granular Granular
Needles
Granular
From Ref. 2.
3.2 Preparation of Refractories

3.2.1 Refractory Forms
Refractories are produced in two basic forms: preshaped objects and unformed compositions in
granulated, plastic forms or spray mixes. The preformed products are called bricks and shapes. The
unformed products, depending on composition and application, are categorized as specialties or
monolithics. Refractory shapes, found in Fig. 3.4, may range from simple nine inch brick to
tapered 30 inch brick, or complex tubes or rods.
Specialties or monolithics are refractories that cure to form a monolithic, integral structure after
application. These include products known as plastics, ramming mixes, castables, pumpables,
pourables, spray mixes, gunning mixes and shotcreting. The mortars used to install brick and
shapes constitute a category of the specialty refractory classification.
Mortars are available in compositions that either approximate the brick they are bonding together,
or are chosen so that their thermal expansion will be similar to the brick with which they are used.
The goal is to achieve a lining that comes as close as possible to being a monolithic and continu-
ous refractory structure.
As the name implies, plastic refractories are ready-to-use materials that are installed by tamping or
ramming. After drying, either the heat from firing the equipment or a chemical binder converts the
plastic material to a solid, monolithic structure. Plastic refractories are available in both clay and
nonclay compositions.
Al2O3-C
Al2O3-C or
Slagline ZrO2-C or
sleeve MgO-C
A. B.
Fig. 3.4 A. Bricks and shapes; B. more complex tubes and rods.
Castables, or refractory concretes, are predominately dry, granular refractory mixes designed to be
mixed on site with water and capable of curing to a stable dimensional form through hydraulic or
chemical setting. Castables are particularly suited to the molding of special shapes and parts at the
installation site. They can be used for forming complete furnace linings, and other unique shapes.
They can be applied by pouring, pumping, troweling, gunning, and shotcreting. They have the
advantage of being readily usable at the operating temperature of the equipment after hydraulic or
chemical setting and removal of all moisture has taken place.
Spray mixes are made from a variety of refractory compositions. The common feature of these mate-
rials is that they contain sufficient water for transport via pumps and for spray application onto a fur-
nace wall or ceiling. A set accelerator is added in sufficient quantity such that the mix sets rapidly.
Gunning mixes comprise a variety of specialty refractory compositions that develop a solid shape
by air drying, hydraulic setting, or heat curing. The principal requirements are that they can be
blown into position by air pressure through a lance or nozzle, but must adhere on impact and build
up to the desired lining thickness. They are used for patch-type repairs, especially inside empty, hot
furnaces.
Fig. 3.5 is a greatly simplified flowsheet illustrating the various methods of refractory manufac-
ture and the resultant products, and classifies raw materials as either calcined, uncalcined (raw), or
binders. Calcined materials have been fired to remove moisture and volatiles and to densify the
material to minimize subsequent in-service shrinkage and reaction. The calcining temperature will
range from 1093–3315°C (2000–6000°F). Raw or uncalcined materials are cheaper to use than cal-
cined materials and are used to impart desirable characteristics such as plasticity or volume expan-
sion to certain refractories. Binders are used to impart strength to the refractory during
manufacture or in service.
3.2.2 Binder Types

3.2.2.1 Temporary Binder
Temporary binders include paper byproducts, sugar, or certain clays. Their function is to improve
handling strength during manufacture.
176
Raw Raw Binders
materials, materials, (solid or liquid-
unfired fired hot or cold)
Crushing Crushing
Ironmaking Volume
and and
screening screening
Mixing - blending
(hot or cold - wet or dry)
Pressing Pressing Vibration, pneu-

matic ramming Extrusion cold Extrusion hot Fusion Packaging
cold hot - mix
hand forming
Fig 3.5 Simplified flowsheet for refractory materials.

Crushing and
Drying Drying and Tempering screening
firing
Pressing
Impregnation Steel cladding
Firing
Blowing, spinn-
ing, drawing
Copyright © 1999, The AISE Steel Foundation, Pittsburgh, PA. All rights reserved.
Resin Pitch Impreg- Steel- Plastic Taphole Anhydrous Fused Fiber Mortars Castables
Burned Temp- plastic or Rebonded and Granular
bonded ered brick bonded nated clad or plastic trough taphole cast insulating sealing and gunn-
brick brick materials
brick brick brick brick block materials material brick materials materials ing mixes
Table 3.21 Dry Binder Ingredients

3.2.2.2 Chemical Binders
for Ramming Mixes, Chemical binders impart strength during manufac-
Plastics and Patching ture, after manufacture, or on installation as a
Plasters monolithic material. Tables 3.21 and 3.22 show
some of the binders which might be used for mono-
Boric Acid lithic materials depending on the application and
Colloidal Aluminas properties desired.
Colloidal Silicas
Goulac 3.2.2.3 Cement Binders
Hydrated Aluminas
Methyl Cellulose Cement binders set hydraulically when mixed with
Mono-Aluminum Phosphates water. The primary binders of this type used in
Petroleum Pitches refractories are the calcium-aluminate cements
Resins which set rapidly and are able to retain some of
Silica Fume their bonding strength to intermediate tempera-
Sodium Aluminum Phosphates tures. The cement(s) used in refractories are
Sodium Poly Phosphates CaO•Al2O3 type cements which develop strength
Sodium Silicate more rapidly than portland cements (near maxi-
Starches mum strength in one day as opposed to 7–10 days)
Synthetic Clays and are capable of retaining strength to higher tem-
Waxes peratures than Portland cements. Table 3.23 shows
several grades of refractory cements.
3.2.2.4 Organic Binders

Table 3.22 Liquid Binder Ingredients for
Ramming Plastics and Patching Mixes Organic binders include tars,
pitches, or resins for use in reduc-
Aqueous Non-Aqueous ing atmospheres where the carbon
Alginates Coal Tars residuals impart bonding strength
Chromic Acid Oils or act to inhibit alteration.
Colloidal Silica Solution Petroleum Tars
The coal tar pitches, derived as
Glutrin Resins
by-products of coke oven opera-
Mono Aluminum Phosphates
tion, were the preferred binders
Molasses
for refractories up to the mid
Phosphoric Acids
1970s. Around 1978, transition
Potassium Silicates
was made from coal tar to petro-
Resins
leum pitches. Today most organic
Sodium Aluminum Phosphates
binders are resin based so they can
Sodium Silicates
be processed cold to minimize
Water
environmental problems.
Table 3.23 Chemical Composition of Calcium-Aluminate Cements

A B C D E
Al2O3 39.0 47.0 58.5 72.5 79.0
TiO2 + Fe2O3 17.8 7.4 1.7 0.25 0.3
CaO 38.5 34.3 33.5 26.0 18.0
SiO2 4.5 7.9 5.2 0.35 0.2
MgO 1.0 0.9 0.1 — 0.4
Alkalies + SO3 0.25 2.1 0.4 0.4 0.5
Ironmaking Volume
Phenolic resins, the most important synthetic binders for refractory materials, are condensation
products of phenol and formaldehyde. These resins are differentiated according to whether they
represent novolacs or self curing resoles. The versatility of phenolic resins is derived from the var-
ious bonding functions that they can provide. Among other things phenolic resins can provide an
intermediate bonding function associated with a thermosetting range of properties during produc-
tion of refractory bricks and, at a later stage, are capable of forming (polymeric) carbon by
pyrolytic decomposition. In the form of bonding carbon, these results contribute to the quality of
the products.
3.2.3 Processing
A knowledge of the various steps in refractory processing is very important in understanding the
behavior of steelplant refractories. As shown in Figure 3.5, all refractories use crushing, sizing and
mixing or blending as the first steps in their manufacture. These steps produce the proper particle
sizings necessary for the desired product density and strength. In sizing a mix to produce dense
brick, for example, the raw materials are crushed and screened to produce some desired particle
size range as illustrated below:
Screen Size The crushing and screening tech-

niques used are increasingly com-
mm Tyler Mesh Size wt% of Mix plex, including vibratory screening
4.6 through 1.65 4 through 10 20 equipment and air classification
1.65 through 1.17 10 through 14 30 techniques.
1.17 through 0.83 14 through 20 5
Mixing and blending steps range
from the simple addition of water
to clay, to hot mixing of preheated
aggregates with selected resin or
0.07 through 0.04 200 through 325 10
other anhydrous binders. Special
Below 0.04 Below 325 20
sequences of combining raw mate-
rials, the time of mixing, and the
use of high energy mixing equipment are used to obtain uniform mixing and equal distribution of
additives.
Fig. 3.5 illustrates some (but not all) of the forming methods that are used after mixing and blend-
ing. The most widely used manufacturing method involves cold pressing of the grain-sized and
blended mix to produce a dense refractory shape. Power or dry pressing of the mix into a shape is
done on hydraulic or mechanical presses capable of forming the moist material (2–5% water) at
pressures of 34.5 to 103.4 MPa (5 to 15 ksi). The degree of compaction obtained in this pressure
range depends on plasticity and particle sizing, but most high quality brick are pressed to the point
where further pressure would produce laminations or internal cracking. The pressing chamber or
mold may be evacuated or de-aired to increase density and prevent laminations resulting from
entrapped air. Dry pressing lends itself to a wide variety of materials and can produce a wide range
of properties. Certain products may be pressed hot and these materials are usually plasticized with
liquid pitch.
Shapes may also be formed by applying pressure by other means such as vibration, pneumatic ram-
ming, hand molding, or isostatic pressing. Many brick and special shapes in fire clay compositions
are also formed by extrusion followed by low pressure pressing (the stiff mud repress process). In
this process, more plastic mixes (10–15% water) are forced through a die by a power driven auger,
cut into slugs, and then pressed to shape. This process usually involves de-airing during extrusion.
Hot extrusion of pitch or other anhydrous bonded materials may also be used; however, this is
mainly for monolithic materials.
In limited cases, raw materials are fused in very high temperature electric furnaces and cast into
larger ingots in graphite molds. These ingots can subsequently be cut to the desired shape, or may
be broken and crushed into a refractory raw material for use in conventional powder pressed brick
or for use in monoliths. In still another process, molten refractory may be blown, drawn, or spun
into fibers for subsequent use in forming mats, blankets, or boards.
Many refractory materials are used in bulk form. Sized, granular refractories may be used in dry
form or mixed with water at the plant site before installation by casting or gunning. Wet extruded
material may be packaged to avoid drying and shipped to the plant site ready for application by
ramming into place as a large monolithic structure. Wet bonding mortars may be shipped in sealed
containers ready for use.
As shown in Fig. 3.5 many products are prefired before shipment. The purpose of firing is to pro-
duce dimensionally stable products having specific properties. Firing in modern refractory plants
is accomplished in continuous or tunnel kilns.
In tunnel kiln firing, which is usually preceded by tunnel drying, the unfired brick loaded on small
cars are passed slowly through a long tunnel shaped refractory lined structure, divided successively
into preheating, firing and cooling zones, generally taking three to five days for the trip. This time
will vary widely, however, with the product being fired. Products of combustion from the fuel
burned in the firing zone pass into the preheating zone (countercurrent to the direction of travel of
the cars onto which the brick are stacked) and give up their heat to the oncoming loads of brick.
Some refractories are also fired in batch or periodic kilns where two to four week cycles are used
for heating, cooling, and loading and unloading kilns.
Temperatures of firing are important regardless of the type of kiln used, because both the quality and
properties of the brick may be affected. The final properties and behavior of most brick can be mod-
ified by firing them in an oxidizing or a reducing atmosphere. By controlling the rate of heating and
the maximum soaking temperature and soaking time, change in the crystalline structure can be
effected, which in turn can also affect the service performance of the brick. In general, the objectives
in firing are to (a) drive off hygroscopic, combined water, and CO2; (b) bring about desired chemi-
cal changes such as oxidizing iron and sulfur compounds, and organic matter, etc; (c) effect trans-
formations of the mineral constituents and convert them to the most stable forms; and (d) effect
necessary combinations and vitrification of bonding agents. Firing temperatures vary from as low as
1093°C (2000°F) for certain fireclay materials to over 1770°C (3200°F) for some basic products.
Certain refractories with carbon binders or containing oxidizable constituents may be indirectly fired
inside muffles to prevent oxidation or may be packed in coke or graphite during firing for the same
purpose. One grade of carbon refractory is hot pressed by electrically heating it during pressing. This
accomplishes the forming and thermal treatment of the refractory in a single step. Nitrogen or other
special atmospheres may be used to impart special binding phases such as silicon nitride.
Manufacturing processes for making lightweight or insulating brick aim for high porosity, prefer-
ably with a fine pore structure. This is accomplished by mixing a bulky combustible substance, like
sawdust or ground cork, or volatile solid, such as napthalene, with the wet batch, by forcing air into
the wet plastic mass, or by mixing into the batch reagents which will react chemically to form a
gas and a product not injurious to the brick. In firing such brick, the combustible or volatile mate-
rial is eliminated and the remaining refractory structure is rigidized. Low density, pre-expanded
aggregate may also be used to make products by conventional brickmaking methods.
Some processing after the fired brick are produced may also be performed. For example the brick may
be steelcased for use in applications where oxidization of the steel case between brick serves to weld
or hold the brick together. Many brick types are also impregnated by placing the fired brick into vac-
uum tanks and introducing liquid pitch or resins into the brick pore structure. This treatment results
in formation of a carbon phase in service which has highly beneficial effects in some applications.
3.2.4 Products
With the large number of raw material types, refractory forms, and manufacturing techniques a
multitude of refractory products are produced. A significant number of them are currently used in
the iron and steel industry. This chapter and the one that follows will present considerable more
detail regarding the specific uses of refractories.
Ironmaking Volume
3.3 Chemical and Physical Characteristics of Refractories and

their Relation to Service Conditions
The foregoing discussions have indicated that there is a wide variety of refractories from the stand-
point of raw materials, overall composition, and method of manufacture. The requirements for
refractories are equally diverse. Analysis of service conditions in iron and steelmaking in general
shows that refractories are required to withstand:
1. A wide range of temperature, up to 2200°C (4000°F).
2. Sudden changes in temperature; high tensile stresses accompanying these rapid
temperature changes cause thermal shock which result in cracking or fracturing.
3. Low levels of compressive stresses at both high and low temperatures.
4. Abrasive forces at both high and low temperatures.
5. The corrosive action of slags, ranging from acidic to basic in character.
6. The action of molten metals, always at high temperatures and capable of exerting
great pressures and buoyant forces.
7. The action of gasses, including CO, SO2, Cl, CH4, H2O, and volatile oxides and salts
of metals. All are capable of penetrating and reacting with the refractory.
8. As a refractory is being subjected to one or more of the previously stated conditions,
it usually functions as a highly effective insulator, or may also be required to be a
conductor or absorber of heat depending on its application. The refractory also must
perform without exposing workers or environments to unsafe or unhealthy condi-
tions at all times.
As any particular service environment usually involves more than one of the above factors, prede-
termining the life of a refractory is a complex process involving information on physical and ther-
mal properties as determined by laboratory testing, analysis of the effect of service or process
conditions and media on the refractory, and a knowledge of the fundamental reactions between
refractories and the various contaminants encountered in service. In this section the physical and
chemical characteristics of selected refractories as measured in a variety of laboratory tests will be
described as a general guide to understanding the complex nature of these materials in relation to
their service environments.
3.3.1 Chemical Composition

As described in the section on refractory classification, the raw materials used in making refracto-
ries differ appreciably and result in materials with a wide range of compositions. It must be empha-
sized that these are unaltered refractories before use and not refractories that have been chemically
changed in service. Refractories have the unique ability to withstand alteration by penetration,
contamination, and/or reaction in service and still function as reliable engineering materials. Sec-
tion 3.4 will describe the reactions between refractories and their environments. As a general rule,
however, recent trends in refractory development require refractories with the minimum content of
impurities, and these impurities are deliberately decreased during raw material or product process-
ing. The following describes the
Refractory Type Undesirable Impurities undesirable constituents (originally
Silica Al2O2 alkali TiO2 present or from contamination) in
Fireclay—all types alkali, iron oxide, CaO, MgO several types of refractories. It
High alumina—all types SiO2, iron oxide, CaO, MgO should be noted that many other
Magnesia-chrome—all types SiO2, iron oxide impurities (such as PbO, ZnO,
Magnesia—all types SiO2, Al2O3, iron oxide B2O3, etc.) which are undesirable in
Carbon alkali, iron oxide all refractories because of their low
melting points have not been shown.
The importance of composition will be described in Section 3.4 where phase diagrams will be used
to indicate the reactions in refractories and their environment at elevated temperatures.
3.3.2 Density and Porosity

Refractory density and porosity are among the most misunderstood and yet useful characteristics
of steelplant refractories. It must be appreciated that most refractories are not fully dense but delib-
erately contain both open and closed pores. Fully dense refractory materials with their inherently
low tensile strengths can not resist the temperature gradients in normal service at constant tem-
peratures (for example across a wall of a furnace or a tube with steel inside and ambient conditions
outside). Refractory materials are also commonly made as heterogeneous structures to promote
resistance to thermal cracking. The volume of pores are measured by immersion of a refractory
specimen in a liquid or in vacuum. (The measurement of total volume is based on Archimedes’
principle that states that a body submerged in a fluid will weigh less than its actual weight by an
amount equal to the weight of the displaced fluid. Knowing that difference in weight and the den-
sity of the fluid, the volume of the submerged body is easily calculated. The volume of open pores
is measured by the amount of liquid absorbed by the sample.)
Fig. 3.6 illustrates the difference between apparent and closed porosity and Fig. 3.7 shows the steps
in determining these properties using a vacuum pressure technique. These simple measurements
are somewhat useful in comparing like refractory products but most useful as quality control mea-
surements for the consistency of manufacture for a given refractory brand.
Closed pore
Open pore Solid material
(a) Open pores versus (b) True porosity: the total (c) Apparent porosity:
closed pores volume of open and closed pores the volume of open pores
Fig. 3.6 Schematic representation of: (a) open pore, and closed pore, and sample volume for (b) true poros-
ity, and (c) apparent porosity. From Ref. 2.
The size and quantity of pores in solid materials are found by measuring the quantity of mercury
that can be forced into the pores of the material under study at various pressures. Since mercury
does not usually wet the materials, the mercury will not penetrate the openings among the cluster
of particles unless force is applied. Thus, the mercury will penetrate into pores in strict relation-
ship with pressure.
Pore size not only defines the size distribution of the pores of a refractory material, but also
defines the size of the bonds formed in the porous matrix and between the porous matrix and dense
coarse particles. The size of each bonding area has an important effect on the critical tradeoff in
refractories properties. A large number of smaller bond areas can blunt crack propagation and
increase crack propagation resistance, thereby improving thermal-shock resistance. However, a
smaller number of bonds with a larger cross-sectional area are slower to dissolve and can enhance
resistance to slag corrosion.
Fig. 3.8 shows typical pore size distribution results for several refractories. Note the small pore
sizes and differences between these products.
Ironmaking Volume
Step No. 7
1,128.70 Step No. 1 Remove samples from
Obtain dry weights tank and submerse
in liquid
Step No. 2
Load samples into 790.60
vacuum pressure tank

Step No. 8
Obtain
Step No. 3 suspended
Pull vacuum and hold weights
for 30 min.
1,177.70 Step No. 9
Obtain
saturated weights
Step No. 4
Cover samples
with liquid Step No. 10
Calculate results
Porosity and related figures
Step No.5
Restore vacuum and Sample weights Data
hold for 5 min. Sample Dry Sus. Sat. Sp. Gr.Abs.% Den. Por.%
number D S W D W–D D W–D
D–S D W–S W–S
1-7 1,128.70 790.60 1,177.70 3.34 4.34 2.92 12.66
1-8 1,129.90 792.50 1,179.30 3.35 4.37 2.92 12.77
1-9 1,135.90 795.30 1,185.50 3.33 4.37 2.91 12.71
Step No. 6 1-10 1,147.80 803.80 1,197.50 3.34 4.33 2.92 12.62
Pressurize tank 1-11 1,145.50 802.97 1,194.50 3.34 4.28 2.93 12.52
and hold for 15 min. Average 3.34 4.34 2.92 12.66
Fig. 3.7 Vacuum pressure technique for measuring porosity in ten steps. From Ref. 2.
Pressure (psia)
10 102 103 104
1.2
1.0
Penetration volume (cm3/g)
1
0.8
0.6
0.4 2
3
0.2
4
0.0
10.0 1.0 0.1 0.01
Void or pore radius (mµ)
Fig. 3.8 Typical results of pore size distribution in

shaped bricks. From Ref. 2.
3.3.3 Refractoriness
The use of the classic PCE, pyrometric cone equivalent test, has little usefulness for today’s refrac-
tories. (This test measures a relative softening point.) The relative refractoriness of modern refrac-
tories is more meaningfully measured using tests for deformation under load and /or creep. Table
3.24 shows comparable results in a short time load test.
Table 3.24 Typical Results of Load Test as Indicated by 24 lb/in2 Load Testing
Type of brick Results of load testing
Fireclay
Superduty 1.0–3.0% subsidence after heating at 2640°F
High-duty Withstands load to <2640°F
Low-duty Withstands load to <2640°F
High-alumina
60% class 0.1–0.5% subsidence after heating at 2640°F
Corundum class 0.1–1.0% subsidence after heating at 2900°F
Silica (superduty) Withstands load to 3060°F
Basic
Magnesite, fired Withstands load to >3200°F
Manesite-chrome, fired Withstands load to 2700°F
Magnesite-chrome, unburned Withstands load to 2950°F
Chrome, fired Withstands load to 2800°F
Chrome-magnesite, fired Withstands load to 3020°F
Chrome-magnesite, unburned Withstands load to 3020°F
Silicon carbide Withstands load to 2800°F
Zircon 0.1–0.8% subsidence after heating at 2900°F
From Ref. 2.
Like most structural materials, refractories

experience creep behavior, Fig. 3.9, when
exposed to high temperatures (0.5 × melting Region I- Creep rate decrease
temperature). Most refractories show two Region II- Creep rate is constant
characteristic stages of creep. In the first Region III- Creep rate increase
stage, called primary creep, the rate of subsi-
dence declines gradually with time. In the Tertiary
secondary stage, called steady state, the rate Primary
creep creep
of subsidence is constant. At very high tem- Secondary
I creep III
peratures, steady state creep is sometimes fol-
lowed by tertiary creep region where the rate II Fracture
of subsidence accelerates and leads to cata-
Time (hours)
strophic failure or creep rupture. Primary
creep is generally short in duration, while sec- Fig. 3.9 Conventional creep curve. From Ref. 2.
ondary creep can occur over a long term.
Therefore, secondary creep usually provides a more meaningful comparison of refractories. Sec-
ondary creep is the parameter determined by the method described above.
Fig. 3.10 shows creep test results on several high alumina brick. Note that the low-alkali (low
impurity) 60% Al2O3 brick has superior creep resistance to high alumina brick with more alkali.
Such load and creep results are useful in many refractory applications. For example in a blast fur-
nace stove checker setting where loads and temperatures are known, Fig. 3.11, such information
can be readily used for design purposes.
Ironmaking Volume
4
2
Linear subsidence ( %)
0 60% Alumina-low alkali
Fig. 3.10 Creep measurement of various –2

high-alumina refractories under 25 psi load
at 1425°C (2600°F) for 0-100hrs. Note the –4 70% Alumina
excellent creep resistance of 60% alumina 85% Alumina
low alkali brick. From Ref. 3. –6
–8 50% Alumina
–10
0 10 20 30 40 50 60 70 80 90 100
Time (hrs.)
1300 0
1200 50
Temperature 100
Temperature (°C)
1100 150
Static load (kPa)

1000 200
250
900 300
Fig. 3.11 Temperature and load distribution 800 350
in checker setting for 1315°C (2400°F) top Load
700 400
temperature. 450
600 500
500 550
400 600
650
300 700
200 750
800
0
5 10 15 20 25 30 35
Checker height (m)
3.3.4 Strength
The cold strength(s) of refractories are
commonly measured in either compres-
sion, Fig. 3.12 or transversely in a modu- Spherical
lus of rupture test, Fig. 3.13. seat Spherical
bearing block
These cold strength values give some Center of Test specimen
measure of the consistency of the refrac- spherical surface Bedding material
tory product and its ability to function
during shipment and installation.
Hot strength values are also determined
by similar procedures with the refractory Load
specimen contained in a furnace which C.C.S. (psi) = load(lbs)/area(in2)*
may also provide for a protective atmos- * Area is cross section area (average) of sample
phere surrounding the specimen. Fig.
3.14 shows the simplest hot modulus of Fig. 3.12 Cold crushing strength test apparatus.
rupture apparatus. From Ref. 2.
Test specimen Fixed bearing edge
T
Fig. 3.13 Cold modulus of rupture test
Pivoted W apparatus. From Ref. 2.
L Pivot
bearing
edge
CMOR (psi) = 3 x P x L/2 wt2
Upper bearing edge
Fig. 3.14 Cross section of a typical hot

Specimen modulus of rupture apparatus.
From Ref. 2.
Loading means
Furnace
Lower bearing edge
Table 3.25 Physical Properties of Refractory Brick

Apparent Cold crushing Modulus of
Density porosity strength rupture
Type of brick (lb/ft3) (%) (lb/in2) (lb/in2)
Fireclay
Superduty 144–148 11.0–14.0 1800–3000 700–1000
High-duty 132–136 15.0–19.0 4000–6000 1500–2200
Low-duty 130–136 10–25 2000–6000 1800–2500
High-alumina
60% class 156–160 12–16 7000–10,000 2300–3300
70% class 157–161 15–19 6000–9000 1700–2400
85% class 176–181 18–22 8000–13,000 1600–2400
90% class 181–185 14–18 9000–14,000 2500–3000
Corundum class 185–190 18–22 7000–10,000 2500–3500
Silica (superduty) 111–115 20–24 4000–6000 600–1000
Basic
Magnesite, fired 177–181 15.5–19.0 5000–8000 2600–3400
Magnesite chrome, fired 175–179 17.0–22.0 4000–7000 600–800
Magnesite-chrome, unburned 185–191 — 3000–5000 800–1500
Chrome, fired 195–200 15.0–19.00 5000–8000 2500–3400
Chrome-magnesite, fired 189–194 19.0–22.0 3500–4500 1900–2300
Chrome-magnesite, unburned 200–205 — 4000–6000 800–1500
Magnesite-carbon 170–192 9.0–13.0 — 1000–2500
Dolomite 165–192 5.0–20.0 1500–3500 500–2500
Fused cast magnesite-chrome 205–245 1.0–15.0 900–1400 6000–8000
Silicon carbide 160–166 13.0–17.0 9000–12,000 3000–5000
Zircon 225–232 19.5–23.5 7000–11.000 2300–3300
From Ref. 3.
Ironmaking Volume
Table 3.25 shows some typical cold properties for several types of refractories and the range of
properties between and within these groups are obvious. As previously stated, these cold proper-
ties are most useful in rating the consistency of quality of particular refractory products.
The hot strengths of refractories will vary significantly with temperature and with other parame-
ters such as furnace atmosphere. Fig. 3.15 and 3.16 illustrate the complex nature of refractory hot
strength for two types of refractories. Fig. 3.15 shows the strength of periclase-chrome refractories
as it varies with temperature and is affected by both composition and initial firing temperature dur-
ing manufacture. Fig. 3.16 shows the strength at a single test temperature as it is influenced by
small changes in the impurity (SiO2) level and the ratio of CaO to SiO2 in the refractory.
Burn
Composition % 1540°C 1705°C
60-40 Periclase - chrome
10,000 80-20 Periclase - chrome
60-40 Periclase - chrome with
9000 finer chrome
8000
Temperature (kPa)
7000
Fig. 3.15 Strength-tem-
perature relationship for 6000
refractories of the indicated 5000
compositions.
4000
3000
2000
1000
0 250 500 750 1000 1250 1500 1750
Temperature (°C)
Al2O3 0.2%
100 Fe2O3 0.15%
B2O3 < 0.01%
High purity Magnesia (99.85% MgO)
90
Modulus of rupture strength (MN/m2)
80 0.85 wt%
SiO2
70
60
Fig. 3.16 Effects of lime/silica ratio on the hot modulus
of rupture at 1500°C (2732°F) of burned magnesia bricks 50
containing a different silica level. From Ref. 2.
0.5 wt%
40 SiO2
30
0.7 wt%
20 SiO2
10
1.75 2.00 2.25 2.50 2.75 3.00

Lime/silica mole ratio
3.3.5 Stress-Strain Behavior

When a refractory is subjected to a mechanical load, it will compress. This behavior may be quan-
tified by the following equation:
s
ε 5 }} (3.3.1)
E
where:
ε 5 strain (dimensionless),
σ 5 stress psi (MPa),
E 5 Modulus of elasticity, psi (MPa).
Strain is equal to the amount of compression divided by the original length.
DL
ε 5 }} (3.3.2)
L
Stress is the force applied per unit area.
F
s 5 }} (3.3.3)
A
The modulus of elasticity, or Young’s Modulus, of a refractory is constant for a given material and
temperature.
The stress-strain behavior of a refractory material is determined by a method which is similar to that
used to measure hot crushing strength. A cylindrical sample is heated uniformly to the test temper-
ature and compressed using a mechanical testing machine. While loading the sample, the change in
its height is monitored by an electrical transducer which is connected by sapphire sensing rods with
the top and bottom of the sample. The data are used to create a stress versus strain curve.
10
9
E
MO
8
pe
Slo
7 Maximum
stress Crack forms
Stress x 10–3 (psi)
6 Magnesite
spinel
brick
5
4
Yield point
3
1
Elastic Plastic
range range
0
0 5 10 15 20 25 30 35 40 45 50
Strain x 103
Fig. 3.17 Example of stress-strain curve for magnesite-spinel brick. From Ref. 3.
Ironmaking Volume
Table 3.26 Static Modulus of Elasticity Fig. 3.17 illustrates a typical stress-strain
(approximate values) curve. The initial portion of the curve is
usually linear and is called the elastic range.
Modulus of Elasticity The slope of the curve over this range gives
Type of brick at 2000°F (psi)
the modulus of elasticity. The point at
Silica 9.0 x 105 which the stress-strain curve becomes non-
50% alumina 6.9 x 105 linear is called the yield point. Straining the
60% alumina 17.0 x 105 material beyond this point results in perma-
Magnesite-alumina-spinel 14.0 x 105 nent deformation; this portion of the curve
Magnesite-chrome 12.0 x 105
is known as the plastic range. Further
Magnesite 9.3 x 105
straining of the material brings failure.
Magnesite-tar impregnated 21.0 x 105
Some materials do not show plastic behav-
Magnesite-tar bonded 14.0 x 105
Magnesite-carbon (20%) 4.4 x 105
ior at certain temperatures, but instead fail
Magnesite-carbonn (15%) 4.0 x 105
after elastic straining. Modulus of elasticity
values of a number of refractories are
From Ref. 3. shown in Table 3.26.
An understanding of the stress-strain behavior of refractories at elevated temperatures is important
in nearly all applications. The shell or superstructure of a vessel usually acts to restrain the thermal
expansion of the lining. Proper lining design may require gaps in the lining (thermal expansion
allowance) during installation to prevent catastrophic stress buildup in the lining during heating.
However, the expansion allowance must also be designed to prevent instability of the lining. The
engineer, too, must be concerned with the stresses which the expanding refractories induce in the
vessel shell or superstructure. Section 3.5 on the selection of refractories will further describe
analysis techniques for refractory behavior under stress.
3.3.6 Specific Heat

The specific heat of a refractory material indicates the amount of heat necessary to change its tem-
perature by a given amount. The specific heat is the amount of heat in British Thermal Units (BTU’s)
which must be absorbed by one pound of material to raise its temperature by one degree Fahrenheit.
Table 3.27 gives the specific heats of several types of refractory brick and refractory minerals.
Table 3.27 Specific Heats of Refractory Brick and Minerals

Mean Specific Heats between 32°F, and T°F, in BTU per pound per °F
Fireclay Silica Magnesite Chrome Forsterite Mullite Cristobalite Periclase Corundum
T°F Brick Brick Brick Brick Brick T°F 3Al2O3•2SiO2 SiO2 MgO Al2O3
32 0.193 0.169 0.208 0.170 0.180 32 0.184 0.165 0.208 0.171
200 0.199 0.188 0.219 0.176 0.200 200 0.192 0.183 0.227 0.196
400 0.206 0.211 0.232 0.182 0.216 400 0.214 0.204 0.240 0.214
600 0.212 0.229 0.242 0.188 0.230 600 0.223 0.239 0.251 0.226
800 0.220 0.238 0.251 0.194 0.240 800 0.229 0.251 0.257 0.235
1000 0.227 0.246 0.258 0.199 0.246 1000 0.233 0.258 0.262 0.242
1200 0.234 0.252 0.263 0.204 0.250 1200 0.237 0.264 0.267 0.248
1400 0.241 0.256 0.268 0.208 0.254 1400 0.240 0.268 0.270 0.252
1600 0.248 0.260 0.273 0.212 0.258 1600 0.242 0.271 0.274 0.257
1800 0.253 0.264 0.278 0.216 0.262 1800 0.245 0.273 0.277 0.260
2000 0.258 0.268 0.283 0.220 0.266 2000 0.247 0.275 0.280 0.264
2200 0.262 0.272 0.288 0.222 0.270 2200 0.249 0.277 0.282 0.267
2400 0.266 0.276 0.293 0.224 0.274 2400 0.251 0.278 0.285 0.270
2600 0.269 0.279 0.297 0.226 0.278 2600 0.253 0.279 0.288 0.273
2800 0.255 0.280 0.290 0.276
3000 0.256 0.281 0.292 0.279
3200 0.258 — 0.294 0.282
From Ref. 3.
The specific heat values for refractory materials are important in many applications because the
amount of heat stored in the lining during heating is often significant. In furnaces which are cycled,
large amounts of heat are alternately stored during heating stages and lost to the surroundings dur-
ing cooling stages. Applications for which high specific heat is desirable include blast furnace
stoves. Stoves are specifically designed to absorb and store heat from hot waste gases. The stored
heat is subsequently used to preheat combustion air.
3.3.7 Emissivity
Emissivity is the relative power of a surface to emit heat by radiation. It is expressed as a fraction
of the emissivity of an ideal black body. Such black body radiation is the maximum possible, but
it is never achieved by actual materials. A good radiator is an equally good absorber of heat. A
good reflector, obviously, is a poor absorber, and consequently a poor radiator. A perfect reflec-
tor, which also does not exist, would have an emissivity of zero. Materials do not radiate equally
well at all wavelengths. The ability to radiate at a particular wavelength is referred to as the mono-
chromatic emissivity. Total emissivity refers to heat radiation over the entire spectrum of wave-
lengths. The total emissivity of most refractories decreases somewhat with an increase in
temperature.
Emissivities dictate the amount of heat which is radiated across a gap in a refractory structure. The
equation used to calculate this quantity is
1 2
2 T24
T14
}}
Qr 5 s l 1 (3.3.4)
}} 1 }}
ε1 ε2 2 1
where:
Qr 5 heat radiated across gap, Wm–2 (BTU hr–1 ft–2),
σ 5 Stefan-Boltzmann constant, W/m–2 K–4 (BTU hr–1 ft–2 °R–4),
T1 5 temperature of hotter surface of gap, K (°R),
T2 5 temperature of cooler surface of gap, K (°R),
ε1 5 emissivity of hotter surface of gap, dimensionless,
ε2 5 emissivity of cooler surface of gap, dimensionless.
In the above equation, the temperatures must be expressed using an absolute temperature scale. The
temperature in Kelvins is K = °C + 273; the temperature in degrees Rankin is °R = °F + 460. The
Stefan-Boltzmann constant is 5.670 × 10–9 Wm–2 K–4 (1.714 × 10 –9 BTU hr –1 ft –2 R–4)
The emissivity of the outermost component of a vessel determines, to a large extent, the amount
of heat radiated to the surroundings. The loss to the environment by unobstructed radiation is
given by:
Qr 5 εs (T14 2 T24) (3.3.5)
where:
Qr 5 heat radiated to surroundings, Wm–2 (BTU hr–1 ft–2),
σ 5 Stefan-Boltzmann constant, W/m–2 K–4 (BTU hr–1 ft–2 °R–4),
ε 5 emissivity of outer surface, dimensionless,
T1 5 temperature of outer surface, K (°R),
T2 5 temperature of surroundings, K (°R).
In most applications, the outer surface is a metallic shell such as carbon steel plate. the emissivity
of the shell, in this case, is affected by the degree of oxidation. The emissivity tends to increase as
the amount of oxidation increases.
Emissivity values for a number of refractories and materials commonly used for shells are listed in
Table 3.28
Ironmaking Volume
Table 3.28 Representative Values of Total Emissivity
Temperature °F
_________________________________________________________
Material 200 400 800 1600 2000 2400 2800
Fireclay Brick .90 (.90) (.90) .81 .76 .72 .68
Silica Brick (.90) — — .82–.65 .78–.60 .74–.57 .67–.52
Chrome-Magnesite Brick — — — .87 .82 .75 .67
Chrome Brick (.90) — — .97 .975 — —
High-Alumina Brick .90 .85 .79 (.50) (.44) — —
Mullite Brick — — — .53 .53 .62 .63
Silicon Carbide Brick — — — .92 .89 .87 .86
Carbon Steel .80 — — — — — —
Aluminum .20 — — — — — —
(Figures in parentheses were obtained by extrapolation of test data.)
From Ref. 3.
3.3.8 Thermal Expansion

All refractories expand on heating in a manner related to their compositions. Fig. 3.18 shows clas-
sic examples of the expansion of divergent refractory types It is obvious that the shape of the
expansion curve with temperature and the maximum magnitude of the expansions differ apprecia-
bly. Compressible materials or voids are used for relieving such thermal expansion in any refrac-
tory construction because growth of several inches can obviously be experienced. Refractories with
linear or near linear thermal expansion generally require considerably less care than those with
non-linear expansion characteristics. A classic example of a refractory that requires care during
heatup is silica brick, which after firing consists of a carefully controlled balance of various min-
eral forms of the compound SiO. Fig. 3.19 shows the thermal expansion of the various silica min-
eral forms which make slow heatup through various critical ranges necessary. As a result of the
complex mineral makeup of silica brick, large constructions using these brick (such as coke ovens)
are heated and cooled at very slow rates and may require several weeks to reach operating tem-
peratures. Once the critical temperature ranges are passed, however, silica brick exhibit low and
uniform thermal expansion and may be thermally cycled without damage between 1000–1538°C
(1800–2800°F)
2.00
1.80
Magnesite
1.60
Linear expansion (%)
1.40
Chrome and chrome-magnesite
1.20 Conventional silica
100% Alumina
1.00 60% to 70% Alumina
0.80 Semi-silica
0.60
High-duty and
0.40 super-duty fireclay
0.20
Fused silica
0
0 200 400 600 800 1000 1200 1400
Temperature (°C)
Fig. 3.18 Typical curves of linear expansion of various types of refractories.
In general, the thermal expansion behaviors

of unfired refractories are more complex
1.5 balite than those of their fired counterparts. Dur-
β-Cristo
β-Quartz
Expansion (%) ing initial heating, dramatic expansions or
te
contractions may occur in an unfired mate-
1.0 mi rial as a result of changes in bonding struc-
T r idy
β 2- rtz
ture, changes in mineralogy, and sintering
Q ua effects.
0.5 β1 α-
The thermal expansion characteristics of a
number of cement-bonded refractories dur-
α Quartz glass ing initial heat-up are shown in Fig. 3.20
α
0 200 400 600 800 1000 These materials show shrinkage over the
Temperature (°C) temperature range 205–315°C (400–600°F)
which is associated with thermal decompo-
Fig. 3.19 Thermal expansion of silica minerals. sition of the cement. The amount of shrink-
age is determined by the quality and
amount of cement. At temperatures above 980 to 1090°C (1800 to 2000°F), additional shrinkage
occurs as a result of sintering. The underlying thermal expansion is determined by the characteris-
tics of the aggregate. The shrinkages which take place during initial heat-up to 1430°C (2600°F)
are permanent in nature and commonly are on the order of 0.2 to 1.5%.
1.00
0.80 90%
Fireclay Alumina
0.60 castable castable 1/16
Linear expansion percent (%)
0.40
0.20
0
0.00
-0.20
-0.40 1/16
-0.60
Low cement
-0.80 fireclay 1/8
-1.00 castable
-1.20
-1.40
0 400 800 1200 1600 2000 2400 2800 3200
Temperature (°F)
Fig. 3.20 Thermal expansion of various refractory castables. From Ref. 3.
Thermal expansion curves for a number of magnesite-carbon brick are shown in Fig. 3.21. Through
the temperature range 370–540°C (700–1000°F), the formation of a glassy carbon bond tends to
densify the structure and shrinkage results. In materials which contain metals, expansive reactions
take place at higher temperatures. Brick which contain magnesium show a dramatic thermal expan-
sion over the range 540–705°C (1000–1300°F). Materials containing aluminum show gradual
acceleration of expansion above about 760°C (1400°F). Thermal expansion at high temperatures
increases significantly with increases in metals content. After heating to 1430°C (2600°F), mag-
nesite-carbon brick show permanent expansion of 0.1 to 1.0%.
Ironmaking Volume
Linear expansion (%)

Magnesite-carbon brick
2 (silicon & aluminum metals)
Fig. 3.21 Thermal expansion (aluminum & magnesium metals)
of magnesite-carbon brick.
From Ref. 3.
1
Carbon brick
(aluminum metal)
0
0 250 500 750 1000 1250 1500 1750 2000 2250 2500 2750
Temperature (°F)
3.3.9 Thermal Conductivity and Heat

Transfer 8
56
As insulators, refractory materials have always
Thermal conductivity (BTU ft−2 hr−1 (°F)−1 in.−1)

been used to conserve heat, and their resistance to 7 48
heat flow is a prime selection factor in many appli-
cations. Fig. 3.22 shows thermal conductivity
Thermal conductivity (Wm−1 K−1)
curves for several refractory types ranging from 6

40
dense refractories to insulating brick. Some refrac-
tories (for example, carbon or silicon carbide) have
Ma
appreciably higher conductivities (up to 43.26 5

gn
es
Wm–1K–1 or 300 BTU hr –1ft–2°F in–1) whereas oth- 32

ite
ers are available with conductivities lower than 0.14

4
Wm–1K–1 or 1 BTU hr –1ft–2°F in–1 (for example,
block insulation or refractory fiber forms). 24
Using measured conductivity values, heat transfer 3

100% Al2O3
losses through single or multiple component refrac-
90% Al O
tory walls can be calculated using the general for- Magnesite - 2 3 16
2
mula: chrome
O
70% Al 2 3
Q T1 2 T2 Super dut y 8
}} 5 }}
L L2 Ln (3.3.6) 1
Silic
a
A 1
}} 1 }} 1 …}} ting
K1 K2 Kn 1650°C Insula
ting
1260°C Insula
where: 0 200 400 600 800 1000 1200 1400
Q/A 5 heat loss expressed as Wm2 Temperature (°C)
(BTU hr –1ft –2),
T1 5 temperature of the hotter surface, Fig. 3.22 Typical thermal conductivity curves
°C (°F), for various refractory brick.
T2 5 temperature of the cooler surface,
°C (°F),
L1, L2...Ln 5 thickness of each material, (in.),
K1, K2...Kn 5 thermal conductivity of each material, Wm–1K–1 (BTU in. ft–2hr–1°F–1).
Such calculations are now rapidly made using computer simulations where calculated heat trans-
fer rates can be balanced with loss from the outer refractory surface by radiation and natural or
forced convection. Moreover, where once it was a slow mathematical process to determine even the
steady-state heat times, the advent of computers has made the rapid determination of the heat trans-
fer data for even transient conditions routine. Although it might seem that every construction
should be designed for minimum heat losses, this is not always the case, and care must be taken in
some situations including the following:
(1) The hot face refractories in a particular application must be able to withstand the
higher temperatures that will result when layers of highly insulating backup
materials are added.
(2) Other refractory properties must be suitable for the environment. For example,
most insulating materials will not stand direct exposure to metal or slag, and
backup materials may be subject to attack by vaporized process components
(alkali, sulfur compounds, acids) or their condensates. Gas channeling through
permeable materials must also be considered to prevent hot spots on shells.
(3) Insulation increases the depth of penetration and chemical attack on the hot-face
layer.
A wide variety of very insulating fiber based refractories are available in bulk, blanket, board, or
shaped forms. In such products, thermal conductivity is highly dependent on product density as
shown in Fig. 3.23.
DECREASING
SHOT
0.4
Apparant thermal conductivity
Cooler
0.3 6 PCF
(Wm-1 K-1)
4 PCF
0.2 Graphite
6 PCF Semi-
(NO SHOT) graphite
0.1
8 PCF
0
250 500 750 1000
Main temperature (°C)
Fig. 3.23 Thermal conductivity as a function of Fig. 3.24 Use of ultra-conductivity materials in a typ-
temperature for various fiber materials. ical bosh construction.
In certain applications, very high conductivity refractories are used to cool the refractory lining and
stabilize it against further refractory wear. Fig. 3.24 for example shows a blast furnace bosh using
graphite and semi-graphite materials with conductivities of 70–80 and 30–35 Wm–1K–1, respec-
tively, at 1000°C. (Note these values in comparison to the more conventional refractories shown in
Fig. 3.22). In such designs, heat extracted through the copper coolers enables the refractories to last
many times longer than uncooled linings. Similarly, a composite construction with silicon carbide
in a blast furnace stack as shown in Fig. 3.25 controls shell temperatures and allows the formation
of a stable lining. In this case the cooling plate density is desirable as varying from none to very
close or dense.
Ironmaking Volume
27 8 1
131/2 131/2 SiC
FC Steel
(PPK) 1 in. SPC
2800 2800 2800 (S) (O)

2000 2000 2000
(M)
Temp. (°F)
Temp. (°F)
Temp. (°F)
1500 1500 1500
(O) (D)
1000 (O)
1000 1000
(M) (S)
500 (S) 500 (M) 500
(D) (D)
0 0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 0 5 10 15
Inches Inches Inches
Heat Max Heat Max Heat Max
Flux Shell Flux Shell Flux Shell
Cooling BTU/Hr-ft2 Temp. (°F) Cooling STU/Hr-ft2 Temp. (°F) Cooling BTU/Hr-ft2 Temp. (°F)
Non (O) 1104 303 Non (O) 1481 376 Non (O) 3027 600
Sparse (S) 6202 160 Sparse (S) 13641 218 Sparse (S) 28476 541
Medium (M) 8993 124 Medium (M) 46232 496
Medium (M) 21771 165
Dense (D) 15090 89 Dense (D) 97693 360
Dense (D) 44591 104
Fig. 3.25 Temperature profile for lining B.
Argon stir Transport Ladle additions

Slag LMF
OBM skimmer Caster
Ladle Temperature
additions modification
Argon stir
Transport Pour steel
Pit
procedures
Transport
Empty
Ladle on Ladle w/ cover
Vertical w/o cover
Preheater upright
Transport
Empty
Either operation w/o cover
upright
Transport
Clean ladle Empty
w/o cover
Well repair on side
Fig. 3.26 Ladle cycling procedure.
2950
2380 Hot face temp

Mid 1st layer
Temperature (°F)
Interface layers 1/2

1810
Shell temperature
1240
670
100
0 4 8 12 16 20 24 28 32
Time (hrs.)
Fig. 3.27 Transient phenomena in ladle cycling.
Many steelplant refractory applications also never reach thermal equilibrium, and dynamic heat
transfer calculations must be employed in the analysis of such applications. For example, a typical
steel ladle will cycle several times daily as shown in Fig. 3.26. During such cycles, the tempera-
ture at the hot face and to a lesser extent the entire lining, will show cyclic behavior, as displayed
in Fig. 3.27.
Importantly, remember that while heat flow at steady state is governed by thermal conductivity k
(first Fourier’s law), the material property required in the transient case is thermal diffusivity λ
(second Fourier’s law).
Thermal diffusivity λ is the equivalent of a diffusion coefficient in the Fourier equation for heat
flow and is a measure of how fast a heat pulse is transmitted through a solid. Thermal diffusivity
is related to both thermal conductivity and heat capacity through the relationship:
k
l 5 }} (3.3.7)
ρC p
where:
k 5 thermal conductivity,
ρ 5 density
Cp 5 heat capacity,
λ 5 m2s–1.
Thermal diffusivity is most commonly determined by the heat-pulse method where a heat pulse is
applied to one side of a disc or short cylinder and the time period for the temperature of the oppo-
site face to reach one-half its final value is recorded. The heat pulse is commonly provided by a
power laser flash.
Table 3.29 shows some diffusivity values for selected products. The relationship among density,
heat capacity, thermal conductivity and thermal diffusivity should be borne in mind, so if one prop-
erty is unkown, the others can be reqadily calculated.
Ironmaking Volume
Table 3.29 Thermal Diffusivity of Different Products, in 10 –6 m2s–1
25°C 500°C 1000°C 1500°C

Alumina > 99.5% 10.7 2.6 1.8 1.3
Alumina, 90% 6.2 1.7 1.2 0.9
Magnesia, 30% porous , 12 ,4 , 1.5 —
Spinel ,6 ,2 , 1.2 —
Zirconia (cubic) 0.7 0.6 0.5 0.6
Silicon carbide (dense) 40–50 15–18 11–13 7–9
Graphite _l c 250–1000 50–150 30–75 —
Graphite l c 2.8–5.4 0.4–0.8 0.2–0.5 —
From Ref. 2
3.3.10 Thermal Shock

Thermal shock resistance is a complex issue when investigating refractory system design. Thermal
shock or spalling is caused by thermal stresses which develop from uneven rates of expansion and
contraction within the refractory, caused by rapid temperature changes or high inherent tempera-
ture gradients in a refractory. A qualitative prediction of the resistance of materials to fracture by
thermal shock can be expressed by the factor:
ks
}} (3.3.8)
αE
where:
k 5 thermal conductivity,
s 5 tensile strength,
α 5 coefficient of thermal expansion,
E 5 modulus of elasticity.
The higher the value of the factor in Eq. 3.3.8, the higher the predicted thermal shock resistance of
the material. Some measure of shock resistance can be obtained using the work of fracture measure-
ments at elevated temperatures to predict the energy to propagate a slow moving crack. Other ther-
mal shock measurement techniques usually involve thermal cycle tests as described in Section 3.5.
3.4 Reactions at Elevated Temperatures

The foregoing discussion of the high temperature behavior of refractories emphasizes physical fac-
tors, but refractory behavior also depends greatly on high temperature reactions occurring not only
within the the refractories themselves but between refractories and contaminants encountered in
service.
Phase equilibrium diagrams have proved to be invaluable guides to understanding service reactions
and the influence of composition on refractory properties. As excellent compilations of diagrams
have been published (see references at end of chapter), only a few are reproduced here. However,
it should be recognized that these diagrams are not without limitations in their use to predict or
explain refractory behavior. For example, the various systems have been explored using simple
combinations of pure oxides and represent equilibrium conditions, while refractories are rarely
pure and seldom in equilibrium, either as manufactured or in service. Because of this complex
chemical nature of refractories, information is often needed on reactions involving so many oxides
that the usefulness of phase equilibrium information on systems involving three or four oxides is
minimal. The diagrams also give no information on such significant matters as viscosity of the liq-
uids formed or the rates at which reactions proceed.
~2020°C Fig. 3.28 shows the Al2O3–

SiO2 system which applies to
Corundum silica, fireclay, and high alu-
Liquid +
Liquid mina refractories. It will be
1900 noted that the lowest temper-
~1850°C 1840°C ature at which any liquid is
Temperature (°C)
Cristobalite
Liquid developed in the system is
1590°C (2894°F), while those
1723°C compositions more alumi-
1700 Mullite + Liquid nous than mullite (3Al2O3•
Corundum 2SiO2) or above 71.8% Al2O3
Mullite - SS
+ develop liquids at quite high
1590°C Mullite temperatures. It is obvious
that useful refractories can be
Mullite + Cristobalite made of the pure oxides Al2O3
1500 or SiO2. The pronounced effect
0 10 20 30 40 90 100 of the impurities present in
SiO2 3Al2O3.2SiO2 Al2O3 most commercial refractories
in this system can be appreci-
Weight %
ated by comparing the tem-
peratures at which they de-
Fig. 3.28 Phase diagram of the Al2O3–SiO2 system. form initially under load,
1150–1200°C (2100–2200°F),
SiO2 with the initial liquid temper-
atures of 1540 or 1840°C
(2874 or 3344°F) indicated
on the diagram. The oxide
Fig. 3.29 The FeO–Al2O3–SiO2 10 90
system.
impurities largely responsi-
ble for lowering the refrac-
20 80 toriness of fireclay refracto-
ids
iqu
ries are CaO, MgO, FeO,

lite
ol
Na 2 O, and K2O. As previ-

Tw
ba
30 70
ously described, mining and
isto
beneficiation techniques are

Cr
40 60 now widely used to obtain

Mullite raw materials with minimum
m ite impurity levels.
50 idy 50
Tr
Iron cordierite
60 2FeO.2Al2O3.5SiO2 40
Fayalite
30
lite
2FeO.SiO2
Mullite
ya
3Al2O3.2SiO2
ite
Fa
yn
Corundum
rc
80 20
He
90 10
ite
¸st
W
FeO 10 20 30 40 50 Hercynite 70 80 90 Al2O3

FeO.Al2O3
Weight %
Ironmaking Volume
Fig. 3.29 shows the FeO–Al2O3–SiO2 system. Here it is seen that the formation of some liquid can
be expected even below about 1095°C (2000°F) with Al2O3–SiO2 refractories and that very dam-
aging amounts will be formed at the higher temperatures common to iron and steel processes. This
is particularly true as iron oxide bearing liquids are characteristically very fluid.
Fig. 3.30 is the diagram of the CaO–Al2O3–SiO2 system, which is most applicable to reactions of
fireclay refractories with blast furnace slags and indicates superior resistance for higher Al2O3
products in such environments. This system has also been useful in predicting behavior of silica
brick, which will be discussed later.
The ternary phase equilibrium diagrams for K2O or Na2O reactions with Al2O3 and SiO2 are repro-
duced in Fig. 3.31 and Fig. 3.32. It is evident that the refractoriness of alumina-silica refractories
will be seriously affected by very small amounts of Na2O, less than 1% being sufficient to lower
the temperature of initial liquid formation to less than about 1095°C (2000°F), while approxi-
mately 10% is sufficient to completely liquefy the more siliceous alumina-silica compositions at
about 1205°C (2200°F). Potassium oxide (K2O) has a similar effect in amounts up to 10%.
SiO2
1713°
1593°
10 90
°
00
Two liquids
17
Cristo- 1470°
20 balite 80
150000° °
°
00
16
1713° °
16 00° 00
°
14
0
15
0
30 70
7
ite
1
1471° Tridym 1165°

0°
1438° 130
°
°
00
°
1400
40
°
Mullite 60
1300
00
13
CaO.SiO2
14
1548° Pseudo-
Wolla-
°
0°
50 50
00
Anorthite 1515°
180
stonite
15
1463°
1500°
3CaO.2SiO 2 1832°
1310° 1265°
1467°
60 1336° 1557° 40
2CaO.SiO2 1400°
2066° 1382°
β
0°
15
1600°
00
0
α-2CaO.SiO
1800°
1700°
1500°
21
2066° °
70 2 30
3Al2O3.2SiO2
°
00
° Gehlenite
20
00
3C 19 ° 1480°
Corundum
80 00 20
aO 18 1595°
0°
1900°
Lime S. i 17
0
O
°
00
3C
15
°
2 00 1505°
16
00
2200
aO
90 ° 10
2100
°
200
15
00
230
. 5A
19
β
°
°
240
14
2000°
18 °
00
00
17 0°
00
°
00
0°
°
0°
l 2O
13
0
16
14
16 °
00
35
250
15
0°
°
15
0
14
3
00 05
0°
0°
°
°
CaO 10 20 30 40 50 60 70 80 90 Al2O3
2571° 1539° 1397° 1402° 1596° 1715° 2038°
3CaO.Al2O3 CaO.Al2O3
5 CaO.3Al2O3
Fig. 3.30 The CaO–Al2O3–SiO2 system.
SiO2
1713°
Cristobalite 1600
90 10
1470° 1400
Tridymite
1200
Quartz 80 1000
20
867° 800
Sodium 789°
Albite F = Albite (Na2O.Al2O3.6SiO2)
disilicate 70
80
F 30
0
Na2O.2SiO2 W = Carnegieite, Nephelite
Mu
874° (Na2O.Al2O3.2SiO2)
llit
837°
e
60 80
0 40
Sodium
00
10
metasilicate
00
12
Na O.SiO
00
2 2 Nephelite 50
14
1089°
1022° 1810°
Co
1120°
W
ru
40 120
0 60
nd
0
140
um
2Na2O.SiO2 Sodium 0
160
orthosilicate 70
30
Carnegieite 3Al2O3.2SiO2
20 80
10 90
Na2O 10 20 30 40 50 Na2O.Al2O3 70 80 90 Al2O3

2050°
Fig. 3.31 The K2O–Al2O3–SiO2 system.
Low levels of impurities are far more critical in silica brick than fireclay brick, and alkalies are the
worst offenders. The amounts of Na2O, K2O, and Al2O3 required to lower the melting point of pure
silica from about 1725°C to about 1675°C (about 3140°F to about 3050°F) are, respectively 1.4,
1.9, and 3.1%.
Another very deleterious fluxing agent for fireclay refractories is MnO, as can ben seen in the dia-
gram of the MnO–Al2O3–SiO2 system in Fig. 3.33. Note that this system is quite similar to the
FeO–Al2O3–SiO2 system.
Because the raw materials for silica brick lack both a natural bond and a high melting point, care
must be taken that the required bonding addition has the minimum effect on refractoriness. Fig.
3.34 shows the CaO–SiO2 system and explains why lime is universally used for this purpose. With
additions of CaO to SiO2, the melting temperature remains unchanged between 1 and 27.5% CaO,
due to the formation of two immiscible liquids. No such phenomenon occurs in the Al2O3–SiO2
system, and by referring again to the CaO–Al2O3–SiO2 system, it is found that only a small amount
of Al2O3 is required to destroy the CaO–SiO2 immiscibility. In fact the effect of minor increments
Ironmaking Volume
SiO2
1713°
Cristobalite 1600
90 10
1470° 1400
Tridymite
1200
Quartz 80 1000
20
867° 800
Sodium 789°
Albite F = Albite (Na2O.Al2O3.6SiO2)
disilicate 70
80
F 30
0
Na2O.2SiO2 W = Carnegieite, Nephelite
Mu
874° (Na2O.Al2O3.2SiO2)
llit
837°
e
60 80
0 40
Sodium
00
10
metasilicate
00
12
Na O.SiO
00
2 2 Nephelite 50
14
1089°
1022° 1810°
Co
1120°
W
ru
40 120
0 60
nd
0
140
um
2Na2O.SiO2 Sodium 0
160
orthosilicate 70
30
Carnegieite 3Al2O3.2SiO2
20 80
10 90
Na2O 10 20 30 40 50 Na2O.Al2O3 70 80 90 Al2O3

2050°
Fig. 3.32 The Na2O–Al2O3–SiO2 system.
of Al2O3 on the liquid development of silica brick is such that the temperature of failure under a
load of 172 kPa (25 psi) will decrease approximately 5°C (10°F) for each 0.1% increase in Al2O3
in the 0.3 to 1.2% range of Al2O3 between super duty and conventional silica brick. Fig. 3.35 of
the FeO–SiO2 system shows that FeO, like CaO, also forms two immiscible liquids when added to
SiO2, thus greatly increasing the tolerance of silica brick for FeO. Furthermore, as with CaO, a
small amount of Al2O3 can eliminate this immiscibility.
As atmospheric conditions in steelplant furnaces may range from highly reducing to highly oxi-
dizing, the form of iron oxides present may vary from FeO to Fe2O3. Accordingly, Fig. 3.36 shows
the system FeO–Fe2O3–SiO2 and is of considerable importance in understanding the behavior of
silica brick in service. Thus, it is seen that the lowest-melting liquids occur from reaction of FeO
and SiO2, and that at temperatures in the range of 1455–1665°C (2650–3030°F) less liquid, and a
less siliceous liquid, will be produced with either FeO•Fe2O3, or Fe2O3 than with FeO, due to the
greater extent of the two liquid region under oxidizing conditions.
The two principal refractory oxides considered to be basic are magnesia (MgO) and calcia (CaO).
Magnesia is noted for its tolerance to iron oxides. As shown in Fig. 3.37, MgO and FeO form a
SiO2
10 90
20 80
30 70
Cristobalite 60
40
°
00
14
Tridymite 300
°
1140° 1
1140° 1120° 1200° Mullite
50 1175° 50
Rhodonite ° 2MnO.2Al2O35SiO2
MnO.SiO2 11
50
sartite
nite Spes 1162°
Rhodo 1175°
1165°
60 3MnO.Al2O3 40
13 00°
°
Corundum
00
12
12
3SiO2
00
°
1200°
1160° 1190°
70 Tephroite Galaxite 30
Tephroite Mullite
2 MnO.SiO2 3 Al2 O3.2SiO2
80 20
90 10
MnO 10 20 30 40 50 60 70 80 90 Al2O3
Weight % Galaxite MnO.Al2O3
Fig. 3.33 The MnO–Al2O3–SiO2 system.
continuous series of solid solutions which have high refractoriness even with very high FeO con-
tents. Under oxidizing conditions, magnesia is even more tolerant to iron oxide. Magnesia and iron
oxide form the refractory compound magnesioferrite (MgO·Fe2O3) which contains 80 wt% Fe2O3.
Magnesioferrite forms solid solutions with magnetite (FeO·Fe2O3) at higher iron oxide contents
and with magnesia at lower iron oxide contents. On the other hand, calcia is more reactive with iron
oxide, forming low melting calcium ferrites such as dicalcium ferrite (2CaO·Fe2O3) that melts
incongruently at ~1440°C (~2620°F). Also, calcia is subject to hydration and disruptive disinte-
gration on exposure to atmospheric conditions and cannot be used in refractory shapes made by
conventional procedures. It is evident, therefore, that magnesia is the more useful basic refractory
oxide and forms the base for all types of basic refractories including those made from magnesite,
olivine, dead burned dolomite, and magnesite and chrome ore.
Magnesia-bearing refractories, regardless of type, contain accessory refractory oxides and
encounter other refractory oxides in service which exert an important influence on their perfor-
mance. Fig. 3.38 shows the reactions and phase assemblages in the MgO–CaO–SiO2 system. In the
high MgO portion of this system, the principal mineral is of course always periclase. The accessory
Ironmaking Volume
2600
2500
2400
2CaO.SiO2 + 3CaO.SiO2
2300 Lime
2200 2CaO.SiO2 + Liquid +
Liquid
2100 α−CaO.SiO2
3CaO.SiO2 Lime
2000 +
2 CaO.SiO2
Temperature (C°)
1900 Liquid "A"
1800 Two
Liquids
1700
Fig. 3.34 The CaO–SiO2 system. Liquid "B"
Cristobalite
1600 +
Liquid "B" Lime
1500
+
1400 3CaO.SiO2
Tridymite
1300 +
Tridymite + Liquid "B" Pseudo-
1200 Wollastonite
(α−CaO.SiO ) 2
1100 Tridymite + Wolla-
stonite (β−CaO.SiO2)
1000
SiO2 10 20 30 40 50 60 70 80 90 100
CaO.SiO2 2CaO.SiO2
3CaO.2SiO 3CaO.SiO2 2
CaO (%)
1800
Liquid A Two liquids
1700 Liquid B
Wüstite
Cristobalite and liquid
Temperature (°C)
1600 and liquid B

Fayalite
1500 and
Wüstite
1400
Tridymite
and liquid B Fayalite
1300 and
liquid
1200
Tridymite and fayalite
1100
SiO2 10 20 30 40 50 60 70 80 90 100
.
2 FeO SiO2
FeO (%)
Fig. 3.35 The FeO–SiO2 system.
SiO2
17
00
°
10 90
20 80
30 70
40 60
1691°
Christobalite
50 50
160
0°
150
0°
60 1 40
30 1
117 0° 400°
8°1 120
14 0°
0°
Fayalite 30
2FeO.SiO2 1666°
0°
°
1200
120
1177°
0°
130
80 12001°147°
20
°
00
Magnetite
14
1300°
90 Wüstite Hematite 10
°
00
n
15
Iro
0°
140
FeO 1580° 20 30 40 50 60 FeO.Fe2O3 80 90 Fe2O3
Weight %
Fig. 3.36 The FeO–Fe2O3–SiO2 system.
2800°C silicate bonding minerals, however, will vary

considerably depending on the ratio of CaO to
SiO2. Table 3.30 presents a summary of the
compounds present with periclase as affected
by the CaO/SiO2 weight ratio and their
approximate melting points. The type of
refractory bonding preferred will depend on
the intended application.The fosterite bond,
which occurs at CaO/SiO2 ratios less than
1371°C
0.93, is desired in some basic brick to prevent
excessive formation of monticellite and mer-
MgO 10 20 30 40 50 60 70 80 90 100 winite. These minerals form low temperature
FeO % liquids and therefore have poor high temper-
ature load carrying ability. Refractory bond-
Fig. 3.37 The MgO–FeO system. ing phases with high melting points can also
Ironmaking Volume
SiO2
1728°
1713° 1709°
10 90
Two Liquids
20 80
1713°
30 70
Cristobalite 1709°
1471° 1471° 1547°
1438° Tridymite 1562°
1336° 1320° Pyroxenes
βCaO.SiO2 40 CaO.MgO.2SiO2 + + + + + + + + + + + + + 60
Solid solutions ++++++
++ + MgO.SiO2
αCaO .SiO2 1391°
CaO.SiO2 + + + + 1360° 1350°
1544° 50 50
1357° Forsterite
1463° . .
1467° 1378° 2CaO MgO 2SiO2 1438°
1402° 1454°
3CaO.2SiO2 1382° 1438° 1505° 2MgO.SiO2
60 40 1910°
1580° ++
2CaO.SiO2 CaO.MgO.SiO2
1501° 1871°
2065°
3CaO.MgO.2SiO2
70 30
3CaO.SiO2
80 20
90 Lime Periclase 10
2300°
CaO 10 20 30 40 50 60 70 80 90 MgO
2577° 2799°
Fig. 3.38 The MgO–CaO–SiO2 system.
be obtained at CaO/SiO2 ratios above 1.86 where refractory dicalcium or tricalcium silicates are
present. In recent years, in addition to the control of CaO/SiO2 ratio, increased emphasis has been
placed on decreasing the amounts of lime and silica present to obtain the maximum advantage of
the properties of nearly pure magnesia. This has been accomplished largely through the use of the
improved synthetic magnesites.
The high temperature reactions of refractories made from magnesia and chrome ore are under con-
stant study. The properties of refractories made from these two raw materials are excellent because
of the tendency of each to minimize the major weaknesses of the other constituent. Chrome ore
consists of a solid solution of chrome spinels (Mg,Fe)O·(Cr,Al,Fe)O3 with appreciable amounts of
gangue silicates. At high temperatures, the gangue silicate in chrome ores is responsible for poor
resistance to deformation under load and the iron oxides, when alternately oxidized and reduced,
cause expansion and contraction, often causing disintegration. Also, chrome spinel shows consid-
erable growth or bursting when reacted with iron oxide at high temperatures as a result of the for-
mation of solid solutions of magnetite (FeO·Fe2O3) and other spinels. With additions of magnesia
Table 3.30 Mineral Phases in Equilibrium with Periclase (MgO) in the

MgO–CaO–SiO2 System
Weight Ratio Approximate Melting

CaO/SiO2 Minerals Present Composition Temperature (°C)
less than 0.93 Forsterite 2MgO•SiO2 1900
Monticellite CaO•MgO•SiO2 1490*
0.93 Monticellite CaO•MgO•SiO2 1490*
0.93 to 1.40 Monticellite CaO•MgO•SiO2 1490*
Merwinite 3CaO•MgO•2SiO2 1575*
1.40 Merwinite 3CaO•MgO•2SiO2 1575*
1.40 to 1.86 Merwinite 3CaO•MgO•2SiO2 1575*
Dicalcium silicate 2CaO•SiO2 2130
1.86 Dicalcium silicate 2CaO•SiO2 2130
1.86 to 2.80 Dicalcium silicate 2CaO•SiO2 2130
Tricalcium silicate 3CaO•SiO2 1900**
2.80 Tricalcium silicate 3CaO•SiO2 1900**
More than 2.80 Tricalcium silicate 3CaO•SiO2 1900**
Lime CaO 2565
*Incongruent melting
**Stable only between 1900° and 1250°C.
Dissociation below and above these temperature into 2CaO•SiO2 and CaO.
to chrome ore, however, the gangue silicates are converted on firing during manufacture or in ser-
vice to the more refractory phases such as forsterite or dicalcium silicate, and the iron oxide to the
spinel MgO·Fe2O3 by co-diffusion of Fe2O3 and MgO between the magnesia and chrome spinel.
MgO·Fe2O3 is more resistant to deterioration in cyclic oxidizing and reducing conditions than the
iron oxides in the original chrome ore. The addition of still greater amounts of magnesia to chrome
ore improves significantly the resistance of the refractory to iron oxide bursting because of the
greater affinity of magnesia for iron oxide as compared with Cr2O3. The addition of chrome ore to
magnesia on the other hand improves the resistance of magnesia to thermal spalling through an
apparent stress relief in an otherwise rigid structure.
As in magnesia refractories, the CaO/SiO2 ratio exerts an important influence on the phases pre-
sent in composite refractories of magnesia and chrome ore. At CaO/SiO2 ratios less than 1.86, the
primary phases between MgO, CaO, and SiO2 are the same as that previously discussed with the
sesquioxides Cr2O3, Al2O3, and Fe2O3 combined with MgO and FeO to form spinel solid solutions.
At higher CaO/SiO2 ratios, the sesquioxides form low melting compounds with CaO.
As in magnesia refractories, efforts have been made to replace the silicate bond in refractories
made from magnesia and chrome ore. The most noteworthy of these efforts to date is the use of fir-
ing temperatures above those normally employed to produce a low silica burned basic refractory
with so-called direct bonding of the magnesia. Fig. 3.39 shows photomicrographs of the conven-
tional silicate and these direct bonds, respectively. Direct-bonded brick have usually high hot
strengths, several times that of conventional brick of similar composition.
Although the phase assemblages previously discussed indicate the general combinations of iron
oxides in basic refractories, it is equally important to consider the influence of furnace temperature
and atmosphere on the oxidation state of the iron present in various compositions including the
oxides MgO, Cr2O3, Al2O3, FeO, and Fe2O3. It has been shown, for example, that in mixtures con-
sisting originally of Fe2O3 and MgO·Fe2O3, dissociation of the Fe2O3 is accompanied by solution
of the dissociation product magnetite (FeO·Fe2O3) in MgO·Fe2O3 until a single spinel phase is
Ironmaking Volume
Fig. 3.39 Photomicrographs showing the structure of basic brick with different bonds. Photomicrograph A is
a magnesite-chrome ore composition in which the bond between the large white chrome spinel grains and
rounded magnesia crystals is principally the dark silicate. Photomicrograph B shows a magnesite-chrome
composition having less silicate and a direct bond between white chrome spinel and periclase (MgO) con-
taining multiple white magnesioferrite inclusions. Reflected light, 200X.
formed. In this same region, it has been shown that the spinel to sesquioxide transition tempera-
ture decreases as the magnesia content increases, but at the same time magnesia stabilizes Fe2O3
at higher temperatures. Chrome oxide (Cr2O3), on the other hand, has an opposite effect to mag-
nesia in that it increases the sesquioxide to spinel transition temperature and lowers the degree of
dissociation of Fe2O3 at higher temperatures. In mixtures consisting initially of MgO·Fe2O3 and
MgO, dissociation of Fe2O3 proceeds with a decrease in the amount of spinel and solution of iron
oxide in periclase.
With the advent of carbon bonded refractories and oxide carbon refractories, the study of oxide
reducing, gas-solid reactions has become an important concern for refractory technologists. Three
main subjects are important: the straight or the direct oxidation of carbon (and graphite) by air; the
reduction of magnesia (and other oxides) by carbon (graphite); and the role of antioxidants.
The rate of the magnesia-carbon reaction below is mainly affected, in addition to temperature, by
the purity and the crystal size of the magnesia grain.
MgO(s) 1 C(s) →Mg(g) 1 CO(g) (3.4.1)
This reaction reduces the magnesia to Mg vapor, which migrates to the surface, oxidizes, and forms
a dense MgO zone or is incorporated in the furnace atmosphere. The dense zone may have a pos-
itive effect on the resistance to slag attack and oxidation resistance, but, on the other hand, the loss
of refractory ingredients due to the MgO–C reaction results in a porous matrix behind the hot face
which can be easily penetrated by slag at a later stage.
Antioxidants as metals (Mg, Al, Si) or carbides (SiC, B3C) are often used to improve refractory
properties such as oxidation resistance. Metals are added as antioxidants due to the fact that the
55 affinity of some metals to oxygen, at the

50 critical temperature range, is stronger
Oxidized area (%)
45 than that of carbon. Each metal, or metal
40
35
combination, has a different oxidation
30 resistance characteristic as shown in Fig.
25 3.40. By protecting the graphite and the
20 bond network from oxidation, and
15 through the formation of carbides and
10 nitrites, the metals improve the elevated
5 temperature strength, in air, and under
0 reducing atmospheres; see Fig. 3.41.
No addition Metal A Metal B Metal C
Moreover, once the graphite is oxidized,
Metal(s) oxidized at 1600°C there is no barrier preventing the magne-
sia grain from being washed into the
Fig. 3.40 Effect of metal(s) addition on resistance to oxi- slag; therefore, by protecting the graph-
dation. From Ref. 2.
ite from oxidation, the metals also im-
prove the resistance to slag attack.
22
20
Flexural strength (MPa)
18
16
14
12
10
8
6
4
2
0
No addition Metal A Metal B Metal C
Metal(s)
Coked at 1200°C Coked at 1400°C
Fig. 3.41 Effect of metal(s) addition on coked strength. From Ref. 2.
Results of thermodynamic calculations of magnesia carbon refractories are summarized in Table

3.31. In this table Zone I is at the refractory hot face and Zone V is at the cold face. The oxides are
to be found in a decarburized zone at the hot face. In the next zone, metal oxidation should be
occurring together with the formation of secondary carbon from CO present inside of the brick
matrix. This secondary carbon could produce extremely low porosity and a low permeable zone
that precludes gas penetration and slag infiltration into the brick matrix. When CO is completely
used up and only N2 is available, the latter gas will form AlN and Mg3N2, respectively.
Carbides of Al, Si, and perhaps Mg can be formed by direct reaction between the respective metal
and carbon further from the hot face. Thermodynamically, the effectiveness of the metals as antiox-
idants should be in the order Si–Al–Mg, least to most. Experimental data showed, however, that the
order is reversed and does not follow oxygen affinity to the metals.
Ironmaking Volume
Table 3.31 Reaction Products for Three systems: MgO 1 Al 1 C, MgO 1 Mg 1 C,

and MgO 1 Si 1 C at Different Temperatures Throughout a BOF Lining
Zone I (CO2, CO, N2)
MgO 1 MA MgO 1 M2S
MA 5 MgO Al2O3 MgO M2S 5 2MgO SiO2
(Spinel)
Zone II (CO, N2)
. 2080K: . 1800K:
MgO 1 AIN + C MgO 1 M2S + C
, 2080 K: MgO , 1800K:
MgO 1 MA 1 C MgO 1 SiC 1 C
Zone III (N2)
MgO 1 AIN 1 C . 1600K: . 1700K:
MgO 1 Mg 1 C MgO 1 SiC 1 C
, 1600K: , 1700K:
MgO 1 Mg3N2 1 C MgO 1 Si3N4 1 C
Zone IV
(Neutral Atmosphere)
. 1380K: . 1810K:
MgO 1 Mg 1 C MgO 1 M2S 1 Mg 1 C
, 1380 MgO 1 Mg 1 C , 1810K:
MgO 1 Al4C3 1 C MgO 1 SiC 1 C
Zone V
(Unchanged Material)
MgO 1 Al 1 C MgO 1 Mg 1 C MgO 1 Si 1 C
From Ref. 2.
3.5 Testing and Selection of Refractories

3.5.1 Simulated Service Tests
Section 3.3 described various tests for single refractory properties. Because most refractories wear
by a complex combination of mechanisms, many simulated tests have been developed to study
refractory behavior for specific applications.
One simple example includes failure of refractories in a special gaseous environment such as car-
bon monoxide. Carbon monoxide disintegration occurs as carbon from the CO gas deposits around
iron concentrations in a refractory:
2CO→CO2 1 C (3.5.1)
The iron concentrations catalyze this reaction, producing carbon buildups and growth on iron loca-
tions which may fracture the refractory behind the hot face (the above reaction is highly tempera-
ture sensitive). Simulative tests have been designed to predict resistance to carbon monoxide
disintegration in all types of refractories by heating the refractory for long periods in CO gas. Fig.
3.42 shows refractory samples after such tests. Although refractories without iron will not be
affected by this reaction, the total removal of iron from most natural raw material is not economi-
cally practical. A more practical solution involves proper handling of raw materials and processes
to remove ternary iron and minimize iron contamination from grinding. Control of the form of iron
in the fired aggregate in monolithic materials or fired products has led to the development of spe-
cial low iron refractories for blast furnace linings. Grinding equipment improvements and magnetic
Fig. 3.42 Specimen rating after the carbon monoxide disintegration test.
Ironmaking Volume
separation have further been used to minimize contamination. Examples of other simple simulated
service tests include testing for exposure to alkalis, hydration, sulfidation, or oxidation.
Many more complex simulative tests exist where various failure mechanisms may be simulated at
one time. For example, refractory samples may be exposed to slags in a rotating furnace heated
with an oxygen-gas torch as shown in Fig. 3.43 (a), (b), and (c). In this test refractory wear from
chemical reaction with the molten slag, erosion by slag, thermal cycling, and other effects can be
measured for various times, temperatures, and atmospheres.
Similiar slag tests are conducted using a metal induction furnace with the specimen as part of the
lining. In this case, wear by both metal and slag can be obtained on the same refractory specimens,
as illustrated in Fig. 3.44.
The results from such tests can usually be correlated with the overall composition, with specific
additives to the refractory and with the overall physical characteristics of the refractory. These
effects are illustrated in Fig. 3.45 through Fig. 3.47. The slag resistance of various basic brick are
Feeder and loading

mechanism
Slag test
furnace
Insulating 381 mm
sheild (15 in.)
et
P ell
of
el
Trav
Air cylinder
Travel
High velocity A A
air stream 229 mm
(9 in.)
203 mm (8 in.)
(a) Rotary apparatus for determining relative slag 95.3 mm
erosion 43.9 mm (3.75 in.)
(1.73 in.)
44.5 mm
(1.75 in.)
101.6 mm
(4 in.) Mortar joints 76 mm
Burner (3 in.)
end 127 mm
(5 in.) Standard brick
Staggering of mortar joints between A-A

adjacent brick Loose aggregate
44.5 mm Test specimen
(1.75 in.) Eroded area
Castable
Burner Refractory liners
end
25.4 mm (1 in.)
381 mm (15 in.)
(c) Positions at which measurements shall be taken

on section specimen after slag test. (b) Cross-section of rotary slag furnace lining.
Fig. 3.43 (a) rotary apparatus for determining relative slag erosion, (b) cross section of rotary slag furnace
lining, (c) positions at which measurements shall be taken on section specimen after slag test.
Fig. 3.44 Sectioned sample after slag test show-

ing high erosion (1.27 in2).
5 V2X slag
4
Relative slag erosion
3 Fig. 3.45 Relation of rela-

tive slag erosion in rotary
test to MgO content for var-
ious basic brick.
2
Direct-bonded
1
Phosphate-bonded
Rebonded
Fused-cast
0
30 40 50 60 70 80 90 100
MgO content (%)
Ironmaking Volume
V1
5 V2X
Relative slag erosion

4
0
10 12 14 16 18 20 22 24 26
Apparent porosity (%)
Fig. 3.46 Effect of porosity on relative slag erosion in rotary test of various basic brick with indicated slag
compositions.
shown to be related to brick compo-

sition (MgO content) and physical 100
structure (apparent porosity) and the 90
Relative corrosion (%)
slag resistance of a magnesia carbon 80

brick is related to the type of metal 70
antioxidant used. 60
50
Refractory specimens may also be 40
exposed to erosion and thermal 30
shock by dipping or spinning in
20
steel baths or by pouring steel on or
10
through them. For example, Fig.
0
3.48 shows steel pouring tubes after No addition Metal A Metal B Metal C
testing in steel while they were
rotated at various speeds to simulate Metal(s)
steel flowing through the exit ports. Rotary slag test at 1650°C
It should always be remembered Fig. 3.47 Effect of metal(s) addition on corrosion resistance.
that any of the simulated refractory From Ref. 2.
tests fail to some degree to duplicate
service conditions. Fig. 3.49 is an effort to fully characterize the slag-refractory reaction involving
wear by corrosion, penetration and dissolution.
What makes corrosion resistance so difficult to measure is the fact that this characteristic is also a
matter of the composition of the slags and their fluidity or viscosity, surface tension, wettability
and work of adhesion, thermal conductivity, mass conductivity and density.
The extrinsic properties of both the refractory and the slag are also important parameters to define
shape and size of the bricks, lining design, nature of the joints, expansion joints, and method of
refractory installation, on one hand; fluid dynamics, heat and mass transport phenomena at the
Fig. 3.48 One brand of pouring tubes after testing of various speeds for one hour in C 13450 (AISI 1345)
steel.
The Refractory The Slag

1. Intrinsic properties 1. Intrinsic properties
Raw materials The nature of the process
Composition Composition
minerals and their properties Chemical major, minor constituents
glass and impurities basicity ratio
Texture Properties
porosity - aggregate - matrix viscosity, surface tension
size distribution conductivity
density
Interaction
2. Extrinsic properties 2. Extrinsic properties
shape - size fluid dynamics and mode of
lining design heat transfer
joint - nature expansion temperature - pressure
installation cyclic
variations in the process
Physical
Fig. 3.49 Parameters influencing the slag-refratory reaction. Case of pure dissolution. From Ref. 2.
Ironmaking Volume
refractory-slag interface, temperature, pressure, temperature gradients, concentration gradients,

variability, and the cyclic nature of the process, on the other hand.
In porous refractories, corrosion-dissolution is an accompaniment of penetration (chemical inva-
sion) into pores, but also in all openings: at the lining level, in open joints, gaps, and cracks; at the
brick level, in connected networks of voids so the texture of refractories and the method of instal-
lation are of paramount importance.
Both physical penetration and chemical invasion are favored by effective slag-solid wetting and by
low viscosity of the slag. Therefore, the slag chemical composition is also of paramount importance.
Slag corrosion always begins with the dissolution of either the matrix material and/or the aggregates
themselves, or, in other words, by the interaction of some minor or major constituents of the refrac-
tories with the liquid slag. Dissolution means either the solubility of a solid phase into a liquid
phase, or a chemical reaction leading to the disappearance of the original solid phase. Therefore,
knowing whether the product is liquid or solid at the prevailing temperature is very important.
3.5.2 Post-MortemStudies
To improve refractory life, the mechanisms by
which the refractories are consumed must be 25
fully understood. A considerable and continu- Non-pitch-bearing
ing effort has therefore been made to thor-
oughly analyze used refractories and to 20
integrate this information into improved
refractory products. In its simplest form, used
15
Amount (%)
refractories are examined by chemical analy-

ses and by microscopic methods. Fig. 3.50 Iron oxides
shows results using this approach on two types 10
of brick taken from a BOF furnace after ser-
vice. In the magnesite brick without pitch CaO
(burned product), CaO, FeO, and SiO2 from 5
the slag have penetrated the brick to a depth of
25 to 75 mm (l to 3 in.) and altered the CaO2
microstructure and properties of the refrac- 0
tory. Such a refractory may not wear uni-
formly due to structural spalling as well as hot 20 Pitch-bearing
face corrosion and erosion. In a similar mag-
nesite brick made with pitch or having
Amount (%)
retained carbon, penetration has been 15

restricted to an area only a few millimetres
thick by the presence of carbon in the brick.
The carbon minimizes wetting and causes the 10 CaO
formation of a dense impervious layer due to
limited MgO volatilization. A carbon bearing
refractory of this type wears uniformly by cor- 5 C
rosion and erosion. The brick without carbon Iron oxides SiO2
has been severely altered by slag penetration.
diffusion, and reaction with the brick compo- 0
nents, whereas penetration has been held to 1 25 50 75 100 125
to 2 mm in the carbon bearing brick. Unfortu- Distance from hot face (mm)
rately, carbon penetration inhibitors cannot be
used for all applications because the carbon is Fig. 3.50 Influence of carbon on slag penetration into
easily oxidized in some oxidizing processes. magnesite brick in a BOF lining.
B) Silicon character-
A) Photomicrograph istic x-ray image
C) Calcium character- D) Zirconium character-

istic x-ray image istic x-ray image
Fig. 3.51 Photomicrograph of center of cross sectioned sensor button showing predominately large grains
with glassy material at grain boundaries and characteristc x-ray images (600X) showing distribution of elements.
Many more sophisticated techniques can be used as part of post-mortem examinations including
x-ray, scanning electron microscope, or electron probe techniques. Fig. 3.51 illustrated the use of
the electron probe to indicate the distribution of impurities in a zirconia refractory. The concentra-
tion of impurities in the grain boundary area is typical of refractory material and helps to explain
why small quantities of such impurities often produce significant detrimental effects including a
significant loss in refractoriness.
3.5.3 Thermomechanical Behavior

Studies of the thermomechanical behavior of refractories and refractory lining systems are being
used to an increased extent. The availability of larger personal and professional computers and
sophisticated software has resulted in increased use of such techniques in refractory design and
failure analysis. Compressive static stress/strain data along with compressive creep data, both at
realistic stress levels, are required to characterize the mechanical behavior of refractories. Refractory
Ironmaking Volume
lining systems are subjected primarily to strain

320
controlled loadings (restrained thermal expansion)
and to a much lesser degree stress controlled load-
ings. As a result, ultimate compressive strain is a 280
better parameter for ranking and selecting the
Circumferential shell stress (MPa)

strongest refractory and not the ultimate compres- 240
Elastic analysis
sive stress. Methods for measuring strain and creep (ultrasonic modulus)
were briefly described in Section 3.3.5.
200
The refractory criteria selected for use in these
analysis techniques are becoming increasingly
realistic. Fig. 3.52 shows calculated results using 160 Elastic / plastic analysis
(static modulus-no mortar
older elastic modulus data on refractories which joint effects)
gave unrealistically high shell stress values in a 120
steel ladle. These values could be reduced to more Test results
realistic (measured) values by adjusting the mod-
80
ulus of elasticity values.
The latest analysis techniques require static com- Elastic / creep analysis
40
pressive stress-strain data and compressive creep (ultrasonic modulus-reduced
data at reasonably high stress levels. Fig. 3.53 shows coefficient of expansion)
stress-strain behavior of a high alumina refractory 0 2 4 6 8 10 12 14 16
with and without mortar joints, which serve to
Time (hrs)
reduce the stress level for any degree of strain.
Stress analysis can be used to understand and Fig. 3.52 Shell stress versus time. From Ref. 4.
modify observed refractory behavior.
50
1 43 2
40 1,2 20°C
6 3,4 800°C
Stress (MPa)
5 5,6 1100°C
30 7,8 1200°C
20
8
10
7
0
0.000
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.010
Strain
Fig. 3.53 Static compressive stress-strain behavior of a 70% alumina brick with a 1.6 mm brick mortar joint.
From Ref. 2
Fig. 3.54 shows the radial compressive load of the working lining against the safety lining that
results from the thermal expansion of the ladle refractory brick in an oval shaped ladle. Note that
the force imposed by the working lining on the safety lining is greatest in the radial section of the
ladle. This compressive force diminishes in the flatter section of the trunnion. Cooling of the
refractory between heats causes an inverted thermal gradient through the ladle lining with the brick
hot face at a lower temperature than the intermediate temperatures deeper in the lining, as illus-
trated in Fig. 3.55.
Cross Axis
Cross
Trunnnion Axis
+
Axis
Half length
of gap
+ Trunnion axis
Gap width
Fig. 3.55 Finite element analysis of ladle quarter sec-

Fig. 3.54 Illustration of radial compressive load tion after cooling between heats. Note the densification
of working lining against the safety lining due to at the hot face due to heating, opening of joints due to
thermal expansion of the working lining. Note the shrinkage during cooling, and gap formation at the
compressive forces are lowest at the trunnion. trunion. From Ref. 5.
From Ref. 5.
These mechanisms help explain the problems experienced by oval type ladles in the ladle straight
sections. Varying material expansion under load characteristics can reduce such problems.
3.6 General Uses of Refractories

Other chapters of this book cover the uses of refractories in specific applications. This section is
intended to offer only a flavor of refractory uses along with general comments.
3.6.1 Linings
The most common type of installation involves a refractory lining to contain molten metal and slag.
Such linings are made of several types of refractories to resist specific conditions and usually con-
sist of a working or hot face lining and a backup lining of material which may include an insulating
material to reduce shell temperatures. Fig. 3.56 shows a typical lining in a steel ladle used in mod-
ern steelmaking to transport and treat molten steel and slag. Note the use of high alumina castable
(85–95% Al2O3) in the metal portion of the ladle with basic (MgO–chrome or MgO–carbon) brick
in the area which must resist slag. The conditions to which refractories in ladles are subjected are
very severe as the ladle will undergo several process cycles daily and may be used for complex
Ironmaking Volume
13 ft-4 in.
9 in. 9 in.
6 in. Direct or MgO-carbon

bonded brick
brick
70 Brick 12 ft
Fig. 3.56 Typical steel ladle 3 in.
lining. 85 Castable
11/4 in.
Steel
shell 95 Castable
12 ft-8 in.
Conventional bloom
Skimmer Degassed slab OBM
Regular slab
Direct slab
Arc heater
Regular slab Trim station
Conventional bloom
Fig. 3.57 Shop process routes. Degassed slab
Non-
degassed Degasser Degassed slab Direct
bloom slab
Conventional bloom
Bloom caster Slab caster
ladle refining, arc reheating, degassing, or other treatments. Fig. 3.57 shows the different cycles to
which ladles are exposed in one steelmaking shop which produces various slab and bloom heats.
Life of the working lining will range from as little as 20 heats for some slag lines to over 100 heats
for the barrel or metal containing zone. Increased ladle wear will result from factors such as:
• higher tap temperatures,
• localized tap stream erosion,
• increased overall steel holding time,
• corrosion by superheated artificial slags during arc heating,
• thermomechanical cracking associated with ladle movements,
• more thermal cycling.
Table 3.32 shows various changes

Table 3.32 Properties of Various MgO–C Brick for
which might be made in MgO–car-
Ladle Slag Lines
bon brick to improve their life at the
Property Better Resistance Obtained by slag line. The exact brick selected
Arc Resistance More graphite will be based on the most economic
Higher purity graphite materials in balance with the life of
the remaining components of the
Oxidation Resistance Metal(s)
ladle.
Hot Strength Metal(s)
Metals and higher purity graphite
Slag line life will depend strongly
on the extent of arc reheating and
Corrosion Resistance High purity sintered grain the manner in which it is con-
High purity graphite ducted. Fig. 3.58 shows the extent
Fused grain to which slag is superheated de-
Combinations of above pending on the thickness of the slag
layer and the manner in which it is
stirred with lances, porous plugs, or by electro-
400
Slag temperature in excess
Stirred magnetic force. Increased superheat can increase

of steel temperature (°F)
Not stirred refractory erosion 2–4 times the normal rates.

300 35mm
Slag
200 thickness,mm 3.6.2 Metal Containment, Control and
100 Protection
50mm
0 Fig. 3.59 shows a simple concept of the various
refractories used in the modern casting process
to contain, control and protect the steel between
0 1 2 3 4 5 6 the steel ladle (top) and caster.
Time after arc heating (min.) Various plates, rods, tubes, and linings are em-
ployed and their properties vary widely. All of
Fig. 3.58 Effect of slag thickness and stirring on
slag superheat. these refractories must present the ultimate in
Table 3.33 Typical Properties of Sliding Gate Plates—Conventional Type

High
85% 90% Alumina, Alumina/
Type Alumina Alumina . 92% Al2O3 Chrome** MgO
Apparent porosity (%) 16 17 15 16 18
Bulk density (g/cm3) 2.88 3.00 3.15 2.95 2.91
Cold crushing strength (kg/cm2) 1700 1500 1500 1500 ND*
Thermal expansion (% at 1000°C) 0.65 0.67 0.78 0.57 ND*
MOR 2500°F (Pa 3 105) 68 82 82 82 109
Major composition (%)
Al2O3 87 90 95 87
SiO2 11 8 3 9
C*** 2–3 2–3 2–3 2–3
Cr2O3 3
MgO 95
Recommended applications Mild, General Throttling General, Low C,
general tundish severe billet
*ND – No data availlable.
**Replaces part of Al2O3.
***Carbon from tar-impregnated which may be followed by partial coking.
Ironmaking Volume
C BC B
D-1 D-1 F
D D D E
A
A
G B
E
F
D D-1
D
Fig. 3.59 Typical refractory components in the ladle to mold system. A = ladle slide gates, B = ladle to
tundish shrouds, C = Tundish stopper rods, D = tundish to mold shrouds with associated nozzles (D–1), E =
tundish slide gates, F = tundish linings, and G = tundish covers.
properties, such as thermal shock resistance, as they are exposed to instant temperature changes
from ambient to steelmaking conditions. These refractories also must not erode or allow air infil-
tration to prevent deterioration to steel quality.
Tables 3.33 and 3.34 show properties of slide gate refractories.
The materials are designed to function in either ladle and/or tundish gates as generally illustrated
in Fig. 3.60 and Fig. 3.61.
Table 3.34 Properties of Typical Higher Carbon Gate Plates
Alumina– Alumina– Alumina–Zirconia

Property Carbon-A Carbon-B Carbon
Composition (%)
Al2O3 89–92 68–75 70–75
SiO2 4–8 3–6 3–5
C 3–6 10–15 7–12
ZrO2 6–9
Density (g/cm3) 3.15–3.20 2.70–2.80 3.00–3.10
Porosity (%) 10–12 11–13 9–11
Hook
Electromotive
thruster or
hydraulic
cylinder
Hinge bolt
Coil spring
Liner
Spring compressor
Lower nozzle
seating block
Upper nozzle Upper nozzle
brick seating block
Upper plate
brick
Upper frame
Lower plate
brick
Arm
Lower frame Connecting rod

Fork end
Cassette shield
Lower nozzle Slide frame
support ring
Lower nozzle
brick
Fig. 3.60 Typical ladle with system.
Ironmaking Volume
Tundish well
Air Tundish
lining
Air cooling
Tundish
Argon oxy bottom
Mounting
plate
Blank gate Discharge extension

Statinary top plate Permeable plug gate
Nozzle gate Permeable plug
Submerged tube holder Tube holder stop
Fig. 3.61 Tundish valve showing submerged pouring tube in position.
The sliding ladle gate uses air-cooled stainless steel springs to maintain a constant sealing pressure
on the sliding refractory interfaces. A valve mechanism permits rapid opening or closing to replace
or inspect refractory components. Refractories for ladle gates must have excellent resistance to
thermal shock and metal erosion and maintain a smooth surface during use. Depending upon steel
grade, plate life may be single or multiple heats without replacement or inspection. Consistency is
vital because of the severe danger to personnel and equipment.
The tundish nozzle is a critical link in the continuous-casting system because it must deliver a con-
stant and controlled rate of flow of steel to the mold with minimum stream flare to minimize spatter,
spray and atmospheric oxidation. Because the ferrostatic head remains substantially constant in the
tundish throughout casting of much of each heat, the bore of the nozzle must remain at a constant
diameter throughout the cast.
Tundish rods and ladle or tundish shrouds are made from isostatically pressed oxide-carbon com-
binations as illustrated in Table 3.35. These materials must have the ultimate resistance to thermal
shock and erosion from steel and slag in the case of ladle to tundish shrouds or from steel and mold
powders in the case of tundish to mold shrouds. Ladle to tundish shrouds may have lives of single
or multiple heats whereas the tundish to mold shroud may be changed in mid-heat if clogging
occurs from the buildup of deoxidation products such as Al2O3. Argon gas injection is usually sup-
plied in one or more areas of the tundish to mold control system to minimize Al2O3 formation
and/or buildup as shown in Fig. 3.62. This requires that a specific portion of the refractory be pro-
duced with a sufficient permeability to direct the inert gas flow.
Refractory usage in the tundish proper continues to evolve as the tundish is required to remove and
store undesirable materials in the tundish slag or on baffles or filters in the tundish. The refractory
inner linings are thin consumable layers applied between strings of heats of like composition, as in
Fig. 3.63. After each string of heats, the tundish and skull are dumped and a fresh refractory layer
applied. The fresh layer, such as a gunnable MgO layer ~1–2 in. thick, privides a completely clean
material to avoid any possible steel contamination. A monolythic safety lining and insulation pro-
vide for safety and low shell temperatures. The length of the string of heats is largely independent
of the refractory as it depends on allowable slag buildup in the tundish and the need for a particu-
lar steel order.
Table 3.35 Properties of Refractories for Shrouds and Tundish Rods
Bulk
Composition (%) Apparent Density
Area Used Al2O3 SiO2 C ZrO2 MgO Other Porosity (g/cm3)
(%)
Base Al2O3
Ladle-to-tundish shroud, 50–56 14–18 26–33 0–5 15–18 2.30–2.44
tundish-to-mold shroud, 42–44 18–24 24–31 0–3 0–2 12–18 2.20–2.35
rod for tundish.
Fused silica
ladle-to-tundish shroud, # 0.3 $ 99 12–17 1.82–1.90
tundish-to-mold shroud.
ZrO2–C
Slag line or nozzle of 1–2 5–7 7–10 75–80 0–2 12–16 3.57–4.18
tundish-to-mold shroud, 20–25 67–74 0–2 15–18 3.20–3.60
head of rod.
MgO–C
Nozzle of tundish to mold 0–2 5–9 85–92 15–18 2.46–2.51
shroud, head of rod 12-16 10-20 58-78 14-17 2.25-2.50
Porous ring Stopper Micro pore Slit

method method method method
Fig. 3.62 Inert gas bubbling methods used to decrease alumina clogging problems.
Ironmaking Volume
Consumable liner
Monolithic
Well block safety lining
Pour pad
Baffle or
filter
Well
block
Fig. 3.63 Consumable liner in tundish for clean steel practice and to facilitate skull removal.
3.6.3 Refractory Use for Energy Savings

Although refractories always provide a degree of thermal insulation, there are some steel plant
applications where these criteria are more important than containment or steel quality protection.
In slab reheating furnaces, refractories are used over the areas of the skid pipe system not in con-
tact with the slab to decrease energy usage, as illustrated in Fig. 3.64. Special refractory metal
anchoring systems or interlocking refractory shapes are used in this application to resist vibration
and facilitate skid insulation replacement. Fig. 3.65 shows some examples of methods of holding
refractory tile on to a skid pipe system.
Slab travel
Plastic Tandem Plastic

tiles
Fig. 3.64 Identifica-

tion of the major com-
ponents of a skid pipe Skid tile
system. (Elbow)
(Horizontal)
U-shaped Long skid
upright
(Leg) supports Wear surface
(welded) Cross-pipe
and single
Saddle cross-pipe tiles
Straight tine
anchor strap
Wire mesh
embedded
in refractory
Headed
stud anchor Vee-anchor
Randomly oriented
stainless steel
wires embedded
in refractory
Latching pin
Fig. 3.65 Various methods of using metallic enhance-

ment with anchors in skid tiles.
In recent years, increased use has been made of various forms of fiber refractory materials in many
types of reheating furnaces because of their good insulating properties and light weight. Such
materials are not suitable in contact with slag or in areas requiring abrasion resistance, but are suit-
able for many less severe areas. Fig. 3.66 shows a plate mill continuous-annealing furnace lined
with fiber blanket held in place with metallic anchors. Refractory type anchors may be used in such
a system to extend the useful temperature of fibers. Fibers may also be used in stackbonded or
modular form where the fibers are folded or fastened to a steel backup plate attached directly to
the steel super-structure. Veneers of fiber are also being evaluated over existing refractory linings
to reduce energy consumption. In all cases, the economies of fiber materials must be carefully eval-
uated in comparison to other insulating materials including insulating brick, insulating castables
and mineral wool or other insulating blocks.
Fig. 3.66 Plate mill continuous annealing furnace with fiber blanket wallpaper construction.
Ironmaking Volume
Fibers offer added advantages when operating equipment under cyclic conditions because of their
low heat storage characteristics.
3.7 Refractory Consumption, Trends, and Costs

Some general comments on refractory consumption, costs and trends are made here for illustrative
purposes. Because this type of information changes rapidly with changes in the steelmaking
process, changes in refractory technology, changes in economic conditions, and changes in envi-
ronmental factors, care must be taken to gather updated information for any specific use. Also,
comparing such data between countries (or even companies) must be done only after an under-
standing in the wide differences in conditions. Consideration must be given regarding integrated
and nonintegrated shops and between the types of steel products being produced.
The types of refractories used in recent years trend toward more use of monolithic and special
shaped products with less standard brick being used. This trend reflects the drive for both lower
labor costs and the changes in the process to produce higher quality steel products. Tables 3.36 and
3.37 show the distribution of refractory product types for an integrated and nonintegrated plant in
Table 3.36 Breakdown of U.S. Refractory Usage—Integrated Plant

Percentage In Indicated Form
Refractory Type Brick Shapes Monolithic
All 49 15 36
Silica 5 25 †
Fireclay 15 25
High alumina 19 58 30
Al2O3–SiC–C 10 20
Burned basic 22
Magnesia carbon 25* 2
Other MgO 5‡ 23§
Miscellaneous 3 10 2
Total 100 100 100
* All types including tempered, impregnated, graphite bearing. ‡ Mainly tundish boards.
† Mainly fused-silica shrouds and coke oven shapes. § Mainly basic gun materials.
Table 3.37 Breakdown of U.S. Refractory Usage—Nonintegrated Plant

Percentage In Indicated Form
Refractory Type Brick Shapes Monolithic
All 41 15 44
Silica 25 †
Fireclay 11 12
High alumina 30 55 21
Al2O3–SiC–C 5
Burned basic 15
Magnesia carbon 40* 2
Other MgO 5‡ 65§
Miscellaneous 4 8 2
Total 100 100 100
* Mainly graphite bearing. ‡ Mainly tundish boards.
† Mainly fused-silica shrouds § Mainly basic gun materials.
Table 3.38 Major Areas of High-Alumina Refractory Usage in the Steel Plant
Type of High Alumina
Process Area Application Product Used Desirable Refractory Features Future U.S. Directions
Blast furnace 1. Furnace lining 90% Al2O3, Al2O3–Cr2O3 brick, Shock and alkali resistance, hot More nonoxide SiC, graphite-type
60–70% Al2O3 brick, strength brick
often tar-impregnated
2. Stove—hot/air system Creep-resistant brick (60–80) Creep resistance, volume stability
3. Cast house (troughs, runners, Al2O3–SiC–C or high-alumina Volume stability, metal and slag Longer life/high quality Al2O3
spouts) castables, plastics, ram mixes erosion resistance types
4. Hot-metal transfer/treatment Burned and impregnated high–Al2O3 Metal and slag erosion resistance Increased Al2O3–SiC–C types as
(torpedo ladles, mixers, lances, products (70–80) metal treatment increases
stirrers) Al2O3–SiC–C brick for metal Cycle resistance
treatment(s)
Steelmaking 1. Runners EAF Al2O3–SiC–C monolithics Metal and slag resistance Increased use replacing older material
2. Delta sections, smoke rings–EAF Low-moisture high-alumina castables Shock resistance types
Ladles 1. Linings Brick (70–80) Thermal shock, metal and slag Increased castable use. Higher
Low-moisture castables (70–95) resistance alumina with clean-steel practices.
2. Pads–blocks Low-moisture castables preformed Increased use with clean-steel
3. Lances Low-moisture castables preformed Thermal shock and erosion practices. Increased use over brick
resistance or other materials
4. Permeable plugs Fired shapes Permeability, resistance to erosion
Tundish 1. Safety Linings Low-moisture castable Volume stability Increased use over brick or other
2. Covers Low-moisture castable Volume stability materials
Shrouds 1. Ladle-to-tundish Al2O3–C Shock resistance, metal and flux Increased use with decreased fused-
2. Tundish-to-mold Al2O3–C erosion resistance silica use
Slide gates Ladle and tundish Fired impregnated Al2O3 Shock resistance, metal erosion Increased use of alumina-carbon
Al2O3–Cr2O3 resistance types. Repair of plates
Copyright © 1999, The AISE Steel Foundation, Pittsburgh, PA. All rights reserved.
Al2O3–C
Reheating Linings High-alumina plastics, castables, Volume stability, insulating effects Continued emphasis on energy
furnaces preformed shapes, fibers savings and ease of installation
227
Ironmaking Volume
Table 3.39 Projected USA Refractory Usage in Indicated Plant

Consumption (kg/ton)
Integrated Nonintegrated
Area Plant Plant
Coke oven 0.40
Blast furnace 0.45
Cast house 2.55
Hot-metal transfer and treatment 2.25
Steelmaking 3.05 4.00
Steel ladles & ladle metallurgy 3.60 4.05
Gates or rods/tubes 0.65 1.05
Tundish 1.25 1.65
Finishing 1.25 1.25
Total 15.45 12.00
the United States. Each of these product types show complex trends in refractory use and some of
these will be obvious by reading other chapters which highlight refractory uses in specific cases.
Table 3.38, for example, shows some of the trends in the use of high Al 2O3 products in the
steelplant. Iron or steel processing changes have driven such trends as increased use of high Al2O3
materials in blast furnace casthouses which now cast on an almost continuous basis, steel ladle
refractories exposed to more severe conditions during ladle treatment of steels, or the increased use
of stream shrouding devices during casting to prevent oxidation.
The consumption of refractories (in terms of lbs of refractories per ton of steel produced) contin-
ues to decrease both from improved life and because modern processes are ever more productive.
At the same time, the cost of refractories per ton is increased as refractories are improved to meet
the new process conditions. Table 3.39 shows the current refractory consumption in two United
States steelplants. As previously stated, these figures vary widely and comparisons between shops
should not be attempted without a full knowledge of process conditions and parameters. The cost
of refractories produced will also vary widely from $0.20 to over $2.50 per lb depending on com-
position and complexity. The cost of refractories per ton of steel produced for an integrated plant
of the type shown in Table 3.39 would range from $9.00 to $12.00 per ton of steel produced.
It should also be noted that refractory costs are often secondary to other parameters including steel
quality, performance predictability and safety, energy considerations, and environmental criteria.
References
1. W.T. Lankford et al., The Making, Shaping and Treating of Steel, 10th ed. (Pittsburgh: Associa-
tion of Iron and Steel Engineers, 1985), 37-96.
2. M.A. Rigaud and R.A. Landy, eds., Pneumatic Steelmaking, Volume Three, Refractories
(Warrendale, PA: Iron and Steel Society, 1996).
3. Modern Refractory Practice, 5th ed., (Pittsburgh: Harbison-Walker Refractories, 1992).
4. C.A. Schacht, “Improved Mechanical Material Property Definition for Predicting the Thermo-
mechanical Behavior of Refractory Limings of Teeming Ladles”, J. Am. Ceram. Soc., 76 (1),
202-206.
5. J.A. Boggs, et al, “Thermal Mechanical Effects on the Structural Integrity and Service Life of
Ladle Refractory, in Secondary Metallurgical Process”, Iron and Steelmaker, June 1992, 33.

Steel Plant

Uploaded by

Copyright:

Available Formats

Steel Plant

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Steel Plant

Uploaded by

Copyright:

Available Formats

Chapter 3

Steel Plant Refractories

3.1 Classification of Refractories

6300 6300 – Graphite, C, pure

3960 – Chromite, FeO-Cr2O3

Key industrial materials

3.1.1 Magnesia or Magnesia–Lime Group

Table 3.1 Selected Sintered Magnesias Produced in North America

Chemical Analysis (wt%)

CaO/SiO2 ratio 3.1 6.3 8.0

Chemical Analysis (wt%)

CaO/SiO2 ratio 9.5 4.3 23.3

Table 3.3 Selected Fused Magnesia Produced in North America

Chemical Analysis (wt%)

CaO/SiO2 ratio 3.2 3.6

Bulk Density (kg/m3) 3530 3500

Table 3.4 Selected Sintered Magnesias Produced in North America

Chemical Analysis (wt%)

CaO/SiO2 ratio 1.0 1.3 0.4

Bulk Density (kg/m3) 3330 3340 3260

Chemicial Analysis (wt%)

3.1.2 Magnesia–Chrome Group

Table 3.7 Examples of Refractory-Grade Chrome Ores (typical data)

Size 12% min. +14 mesh –20+150 –10+100

Table 3.8 Examples of Magnesite-Chrome Raw Materials (typical data)

Bulk Specific Gravity (g/cm3) 3.53 3.75 3.8

Table 3.9 Examples of Refractory-Grade Silicas 3.1.3 Siliceous Group

3.1.3.2 Fused Silica

3.1.3.3 Silicon Carbide

Table 3.10 Examples of Refractory-Grade Silicon Carbides

Coarse, 6/F 200/F 10/F

Total Carbon 29.10 29.20 30.70

Table 3.11 Refractory-Grade Zircons (typical data)

resistance, and hot load properties are

3.1.4.2 Bauxitic Kaolins

Table 3.13 Some Examples of Refractory Clays (typical data)

L.O.I. (dry basis) — 8.3 12.7 13.7 10.3 5.4

Table 3.14 Domestic U.S. Bauxitic Kaolins

Typical Chemistry Ucal 50 Ucal 60 Ucal 70

Specific Gravity (g/cm3) 2.60 2.70 2.78

Pyrometric Cone Equivalent 35–36 37–38 38–39

Table 3.15 Examples of Andalusite and Kyanites

Table 3.16 Refractory-Grade Calcined Bauxites (typical data)

Guyana, United States

Bulk Specific Gravity (g/cm3) 3.10 min. 3.20 3.15–3.20 2.74

3.1.5 Processed Alumina Group

Strong Extraction Impurity

Fig. 3.2 Schematic of Bayer process.

Primary Secondary Market Application

Auto catalyst substrates

Fillers and coatings Electrode coatings

Ruby abrasive grain

Fig. 3.3 Products of Bayer Process.

3.1.6 Carbon Group

Table 3.17 Calcined Aluminas (typical data)

Hydrated Intermediate Intermediate Low

Table 3.18 Tabular Alumina (typical data)

Bulk Specific Gravity (g/cm3) 3.55 3.54

Table 3.19 Fused Aluminas (typical data)

Typical Chemistry Brown1 White2 White3