"Technical requirements of refractory materials for burner tile applications - the

advantage of a phosphate binder system".

Jens Decker, Tom Atkins, David Mintz – Stellar Materials, Boca Raton, USA
Karel Hoeksma, Roy Ovink – Corus Research, IJmuiden, The Netherlands

1. Introduction
Major industrial processes require combustion in different forms. The burner is the
device that is used to combust fuels with an oxidizer and to control the heat transfer into
a furnace in which these processes take place. An essential part of a burner is the
refractory burner tile either made of ceramic fiber or refractory pre-fabricated shapes.

A burner tile is a complex functional refractory product and inaccurate operation can
cause downtime costs which exceeds the costs for the burner equipment by far.
Environmental aspects and more stringent regulations have changed the philosophy of
burner design in the last 20 years. Burner design is now a compromise between
efficient mixing of fuel and oxidizer and the necessity to reduce the total fuel
consumption and pollutant emissions, which requires moderate low flame temperatures
in order to reduce NOx. These designs have led to the development of tailor made
solutions for certain applications and to highly sophisticated burner designs with tiles of
complex shapes. (Picture 1).

This paper discusses the technical requirements of refractory materials for burner tile
applications and evaluation methods in order to generate thermo- mechanical data as an
engineering tool for the design of burner tiles. Furthermore the influence of two different
binder systems will be described in this context.

2. Burner Tile Design

The design of a burner tile plays an important role in the performance of a burner. The
tile
- forms the desired flame pattern
- avoids flame radiation towards the burner casing
- forms the orifice that controls the flow of air for the combustion reaction
- stabilizes the flame due to the thermal insulating properties of the refractory
material

Regarding the burner design, the combustion engineer has to consider the characteristic
of the ceramic burner tile material in order to obtain stable combustion process
conditions.

The main design problems are:

- The tile is restrained against complete, free, thermal expansion by the external
structural steel casing or anchoring. The designer has to consider the physical
characteristics of the refractory material by definition of the thermal restraint
stresses.
- Sharp inside 90°edges lead to material stress peaks in service.
- The wall thickness of the burner tile is not uniform under isotherm conditions and
stresses may concentrate in the thinner areas. Hence critical internal strain
controlled loads can occur.
 - The design of the tile must be in accordance with the occurring temperature
range and the plastic / elastic response of the refractory material within this
range.
- Part of the tile is cooled by the fuel and oxidizer, another part is heated by the
flame and the furnace. As a consequence the tile has to cope with large
temperature differences. In many burners this may be over 1000C.
- In case the burner is stopped and the furnace is hot, then the tile will heat-up
quickly and damage caused by thermo-shock is often observed.

Picture1.) Flat Flame Burner (courtesy of John Zink Company)

Picture 1.0 shows a low emission flat flame burner of complicated tile shape with sharp
edges and additional fuel jets in order to reduce NOx. The purpose of the geometry is to
control a precise oxidizing process of the fuel (1) The wall thickness is not uniform and
the jets probably keep parts of the tile cold. This is an example for a burner with a critical
tile design.

3. Specific evaluation and testing criteria of refractory materials for burner tiles.

3.1. Burner tile temperature

Essential for the material choice and design of the burner tile is the knowledge of the
temperature profile in operation. Reliable data can be obtained by experiments.
Experiments however are time consuming and dependent on the availability of
appropriate equipment.
In addition thermal boundary conditions used for the finite element model are estimated
and based on the flame temperature. However, adiabatic flame temperatures are not
useful data to estimate the resulting tile temperature. For example in the hydrocarbon
and petrochemical industry most burners are working with excess air. Additional cooling
of the tiles occurs in low NOx burners because introduction of inert flue gases into the
combustion zone of the burner reduces the peak flame temperature and NOx formation.
Therefore the burner type, the composition of fuel and air inlets and a possible flue gas
recirculation must be considered when estimating thermal boundary conditions of the
tile.
A good indication about the flame temperature and pattern can be obtained from burner
modeling by computational fluid dynamics (CFD)

Material testing should also reflect the different dynamic temperature stages to
determine suitable materials. A burner tile is exposed to one or more temperature cycles
during the lifetime. A temperature cycle can be divided in different stages which are
important for the material requirements. These are:

a.) Heat up b.) Steady heat flow at operating temperature c) Thermal cycling
caused by burner control and operation d.) Cool-down.

Hence the temperature dependent stress – strain behavior of the refractory material
must be defined.

3.2. Thermochemical Considerations

External chemical attack on burner tiles occurs predominantly on heavy oil fired burners.
The reason for the attack is the reaction of impurities from oil like vanadium and sodium
with the refractory to low melting silicates. Therefore beside the alkalis, the silica content
determines the chemical stability of the material.

The ISO standard 13705 specifies the required chemical composition of burner tiles for
certain applications:

- Normal burner tile: > 40% alumina

- High intensity combustor: > 85% alumina
- Oil firing tile, <50mg/kg V, Na: > 60% alumina
- Oil firing tile, >50mg/kg V, Na: > 90% alumina

Table 1 shows as an example the chemical analysis of a sample taken from a burner tile
of an oil fired ship boiler burner. The tile failed after 6 months in operation because of
surface spalling after cool down. The table shows the quantity of impurities picked-up
during operation.
Table 1
V2O5 + 2,7%
NiO + 2,0%
Na2O + 1,3%

The mineral composition also plays an important role in the internal thermo-chemical
and mineralogical reactions. A high mullite content is favorable for good thermal shock
properties due to the low reversible expansion. The alkali and alkali earth metal oxide
content has an influence on the modulus of elasticity, the resulting thermal shock
properties and phase transitions that lead to stresses due to expansion or shrinkage.
Furthermore a long term attack occurs from particles of combustion air. Silica and alkali
containing particles melt in the flame, and due to recirculation of combustion products,
attack the tile. Due to this, burner tiles can show severe surface cracking after years in
service.
3.3. Thermomechanical Considerations
During high temperature service, burner tiles are required to conduct heat and they may
be subjected to rapid temperature variations. This takes place during adjustments of the
burner, fuel changes and shut down conditions when cooling down and heating up the
burner again. The non-linear temperature gradient during rapid heat up and cool down
develops thermal stresses or thermal expansion forces that may fracture the refractory.
The refractory structure should resist tensile and compressive strains through a wide
temperature range from a rigid solid state at low temperatures to a more viscous state at
high temperatures. Cracking from these stresses can have an impact on the flame
pattern and in a worst case scenario cause the flame to be in contact with the tile
causing problems like hot spots and total destruction of the tile and burner casing.(2)

One approach to determine the resistance of a refractory material against cracking in

service is the evaluation of different stress parameters. The expressions obtained from
the calculation of the stress parameter R1 – R3 can be transformed in terms of the
maximum temperature difference, heat flux or temperature gradient to which the
refractory can be subjected without exceeding the thermal stress failure.
The “Hasselmann thermal stress resistance parameters R1-R3” are a tool to predict
thermal stresses in materials of brittle - elastic behavior for the purpose of basic
materials ranking (3).

R1 = (1- ) / ( E)
A rapid thermal shock leads to thermal gradient on the surface of a material. R1 is the
maximum allowable temperature difference in body under conditions of steady heat flow.
( T in °K)

R2 = R1
Considering a constant heat transfer on the surface - R2 is the maximum allowable heat
flux through body under conditions of steady heat flow (W/m)

R3 = R1 a
R3 is the maximum allowable rate of surface heating.

= tensile strengths
= Poisson ratio (transversal contraction index)
= linear thermal expansion
E = Young’s modulus (static: determined as V-modulus)
= thermal conductivity
a = / (cp ) = thermal diffusivity

Based on the stress parameter, an ideal thermal shock resistant material should
possess the following thermo-mechanical properties:
- low thermal expansion
- high thermal conductivity
- high strengths under consideration of the stress/strain relationship. (A high / E
ratio leads to good thermal shock properties)

Furthermore the tile design is important:

- the wall thickness should be uniform and reasonably thin
 - the overall dimensions of the tile or pieces of the tile should be small
- the geometry of the tile should avoid sharp angular shapes

4.0 Material choice for burner tile application.

Castable materials have been traditionally used for burner tiles because of the
advantage in minimizing costs when manufacturing complex shapes of limited quantities.
Placement of castables, when compared to pressed- or rammed products, is easier and
the bodies show a more isotropic structure which is advantageous in relation to thermal
shock properties.
One difference between fired products and castable refractories is the higher shrinkage
of castables at elevated temperatures which affects the geometry of the body at service
temperature. Thus the interaction between permanent linear change and thermal
expansion needs to be considered as an important factor for the design of burner tiles.

4.1 Cement bonded materials

In the past 30 years cement bonded materials have evolved from high cement
conventional formulations to low and ultra low cement materials with an improvement in
placement properties and high temperature stability.
This evolution has been assisted by the development of additives and fillers in order to
achieve water reduction to decrease porosity and increase mechanical strengths (4)

4.1.1 Binder characteristic and thermomechanical properties of cement bonded

castables.
The binder characteristic of calcium aluminate bonded castable materials has been
clearly explained by numerous authors.(5,6,7,8) A simplification of the complex setting
procedure can be explained as follows:

High alumina cement CA, CA2, C12A7 etc.) +additives + water calcium aluminate
hydrates (CAH10, C3AH6, C2AH8, etc.) +additives + water + heat + residual water. (9)

Due to the high quantity of water and the low permeability of cement bonded materials, a
rapid temperature increase may result in explosive spalling or steam explosion. Factors
that can contribute to explosive spalling include curing conditions, heating and dry-out
method, and physical properties of the material (10). Due to these factors pre-cast
burner tiles require accurate dry out or pre-firing before the installation.

Usually thermo-mechanical data of cement bonded materials includes refractoriness

under load or hot modulus of rupture data. However, compressive stress/strain data of
castable materials which are necessary to predict strain controlled loads are not often
published.
C.A. Schacht reported that the chemistry of the refractory material has a significant
influence on the stiffness of the castable. The lower the cement content and the higher
the alumina content the higher the stiffness.(11) Due to the high strength of the low-
cement castables the maximum strain is similar to conventional castables

4.2 Phosphate bonded materials

Experience has shown that phosphate bonded burner tiles of complicated shapes can
hold up, in some critical process conditions, much better than conventional burner tiles
made of cement bonded materials. Since not all phosphate bonded castable materials
possess comparable properties it is important to refer to the distinguishing
characteristics.

4.2.1 Binder Characteristic and thermo-mechanical properties

Phosphate bonded refractory castable products are available with various binder
systems which can be differentiated as follows:

- heat setting
- cold setting + setting additive
- one component dry material with water addition
- two component mix with phosphoric acid solution or monoaluminium phosphate
solution + dry component

The chemical compositions of cold setting phosphate bonded castables are based on
the system Al2O3-SiO2-P2O5-MxO (MxO is alkali- or alkali earth oxide). Heat setting
phosphate bonded castables are based on the system Al2O3-SiO2-P2O5. The P2O5
content ranges from approximately 3% – 6% and the MxO content from approximately
0.5% - 6%.
A great number of works have been reported about the reaction between alumina and
phosphoric acid (12,13,14) Alumina reacts with phosphoric acid above 127°C (260°F) to
aluminium phosphate and shows considerable strengths above 350°C (660°F). The
most often applied phosphate binder in the refractory industry is mono-
aluminiumphosphate (MAP) in liquid or spray-dried form.
If setting properties at ambient temperature are desired an additional setting agent is
needed. This can be an alkali- or alkali earth metal oxide component (MxO) like sodium,
magnesia, calcium or compounds thereof.
The disadvantage of a spray dried MAP additive in a single component material is the
poor shelf life due to the strong hygroscopic behavior. Alternatively, phosphoric acid can
be used as a liquid in a two component product. Depending on the reactivity both MAP
and phosphoric acid need to be diluted to slow down the exothermic reactions.
Phosphoric acid is a tri-basic acid and reacts with different metal oxides to form salts.
Some of these salts function as a refractory binder. In an aqueous solution the
phosphoric acid reacts in three stages according the molar ratio MxO / P2O5 of 1:1, 2:1 or
3:1.
Table 2 shows as an example the reactions between magnesia and phosphoric acid and
the resulting three different stages. Magnesia and magnesium hydroxide are widely used
as setting additives for phosphate containing formulations. The reaction type 1 and type
2 are exothermic. However, reaction 1 is much stronger than 2 whereas reaction 3
doesn’t show cold setting properties.

Table 2.

1.) Formation of mono- phosphate: MgO + 2 H3PO4 = Mg (H2PO4)2 + H2O

2.) Formation of di- phosphate: Mg(H2PO4)2 + MgO = 2 MgHPO4 + H2O
3.) Formation of tri- phosphate 2MgHPO4 + MgO = Mg3(PO4)2 + H2O

The phosphates being formed at ambient show dehydration of chemically bonded water
at elevated temperatures. This condensation process is described in Table 3.
Table 3.

1a.) Mg(H2PO4)2 = Mg(PO3)2 Mg- metaphsophate + H2O 150ºC (302ºF)

2a.) 2 MgHPO4 = Mg2P2O7 Mg- phyrophosphate + H2O 213ºC (415ºF)
3a.) Mg3(PO4)2 = Mg3(PO4)2 Mg- orthophosphate 700ºC (1292ºF)

Phosphate bonded products with a binder system of a two component mono- and a
single component di- phosphate are commercially available. The advantage of a two
component mono- phosphate bond is the faster setting and faster heat up time.
Flame impingement experiments with both kinds of materials have shown that mono-
phosphates are less susceptible to explosive spalling than di- phosphate bonded
materials.
Possible explanations for this are:
- The mono-phosphate formation is highly exothermic, which cause dehydration
during setting.
- The quantity of chemically combined water in the mono- phosphate is higher than
in di- phosphates. This can have an influence on the steam pressure in the
capillary system of the material during bake out.
- The water in mono-phosphate evaporates at 150 C with max 5 bar steam
pressure. For di- phosphate the water is released at 213 C and the steam
pressure of 20 bars may exceed the strength of the material.
- Two component products use a molar excess of H3PO4 compared to MxO. The
vapor pressure of phosphoric acid and monoaluminium-phosphate solutions are
lower than water. Thus, they are less explosive than other (water containing)
castable refractories during heat up (15)

It should be noted that the whole dehydration process of phosphate bonded castable
refractories hasn’t been properly evaluated yet.

The thermo-mechanical properties of phosphate bonded materials are different than

those with a cement bond. After setting, additional curing time is not needed. Poly- and
meta -phosphate compounds formed during condensation possess a ring and chain
structure with a stability up to 800°C (1500 ºF). (16) This flexible structure leads to a low
modulus of elasticity and contributes to the thermal shock and abrasion resistance.

Above 900°C (1650ºF) first melting of the MxO- phosphate compounds of the binder
system occurs. Depending on the raw materials, heat setting phosphate bonded
materials behave similar because of a reaction of phosphates with MxO impurities. Thus,
phosphate bonded castables with alkali additions can show lower hot modulus of rupture
and refractoriness under load numbers compared to cement formulations. Depending on
the quantity and kind of alkali additive, this melt phase is of a temporary nature and
further reacts with alumina to melts of higher viscosity. In many burner tile applications a
certain quantity of melt phase at service temperature reduces the stiffness of the
material and improves the thermal shock properties, which is desired.

Between 900°C (1650ºF) and 1600°C (2900ºF) amorphous phosphates and phosphate
melts are being formed. Depending on the kind of MxO- phosphate and quantity thereof,
occurring melts can attack the aggregates of the raw materials at process temperatures
above 1350°C (2460ºF). Resulting volume changes are due to the formation of
additional melt (shrinkage) or formation of new compounds like Mg-spinel and mullite
(expansion). Above 1600°C (2900°F)aluminium-phosphate starts to decompose to
aluminium-oxide and P205 which is volatile at this stage.

4.3 Material testing of phosphate bonded materials

As aforementioned cold setting phosphate bonded materials with MxO- phosphates as a
binder component show partial melting at temperatures between 900°C (1650°F) and
1400°C (2550°F) Table 4 shows some alkali- and alkali earth metal oxides used as a
setting agent.

Table 4. Melting points of alkali phosphates (17)

Chemical formula Name Melting Point °C (°F)

Na4 P2O7 Na pyrophosphate 985 (1805)

K4 P2O7 K pyrophosphate 1109 (2030)
Ca (PO3)2 Ca metaphosphate 975 (1790)
Ca2 P2O7 Ca pyrophosphate 1350 (2460)
Mg(PO3)2 Mg metaphosphate 1165 (2130)
Mg2 P2O7 Mg pyrophosphate 1383 (2520)
Al (PO3)2 Al metaphosphate 1060 (1940)

There is some evidence that occurring phosphate melts are temporary in nature and
change at increased temperatures to different compounds. Some test methods,
however, do not reflect this materials characteristic. Critical tests are:

- Thermal expansion under load > 1100°C (2000°F).

- Hot modulus of rupture > 1100°C (2000°F)
- Refractoriness under load

The results obtained from these tests can be significantly influenced by the occurring
melt phase at test temperature. For instance samples pre fired at 1100°C (2000°F) with
a holding time of 24 hours showed 10% HotMOR increase and with a holding time of
168 hours an increase of 65% at 1100°C (2000°F). Typically, published data references
test results after only a five hour holding time.
Another example is thermal expansion under load. Due to the load, plastic flow and the
coefficient of thermal expansion are mixed and resulting high permanent linear change is
then overstated.
The next example shows the conversion from calcium phosphate to Anorthit.
This reaction is evident from XRD analysis. Calcium-monophosphate as part of the
binder component is being formed at ambient temperatures by the reaction of
phosphoric acid and calcium oxide. During heat up this monophosphate converts into
calcium - metaphosphate as shown in Table 3. Calcium metaphosphate has a melting
point of 975°C (1800°F).
After firing at 1600°C (2900°F) with 10hours holding time the XRD diagram (Figure 1)
shows that all phosphate has been converted to AlPO4. Calcium oxide on the other
hand has been reacted with mullite and converted into Anorthite with a melting point of
1550°C (2800°F)
Figure 1 XRD analysis Formula 6B – 1600 C/10h (2900°F)

Test procedures of phosphate bonded materials should report this operational

characteristic since burner tiles are exposed to elevated temperatures with holding times
up many years. Otherwise a material assessment based on standard test procedures
would lead to the wrong conclusion.

5. Numerical studies

Comparative FEM analysis of two burner tiles made of commercially available cement
bonded- and phosphate bonded castables have been conducted in order to determine
the thermo-mechanical behavior of the different materials during the critical heat up. The
tile used for the modeling is a burner tile of an annealing furnace.
A heat up was considered on this axi-symmetric burner tile as indicated in Figure 2.The
thermal boundary conditions are indicated in Figure 3. A heat up time of 48 hours was
considered.
Figure 2 The axi-symmetric burner geometry.

Figure 3 Thermal boundary conditions used in the finite element model

5.1 Properties of test materials
The material properties used in the simulation are those for cement bonded castable
“HS” and phosphate bonded castable “SB” and are summarized in Table 5.
The cement bonded material is an andalusite based LCC which had been supplied by
the manufacturer in a dried “ready to install” state.
The phosphate bonded material is a mullite based two component mono-phosphate
castable as shown by reaction 1, Table 2, with 61% alumina and 4.5% P2O5 content.
The yield stress under tensile loading conditions is taken to be temperature dependent
and is summarized in Table 6 and 7. The yield stress under compressive loading is
assumed to be a factor of 10 higher.
Table 5 Material properties of test material

Andalusite Phosphate
HS Bonded SB
Density 2450 kg/m3 2360 kg/m3
Thermal expansion 6.6e-6 K-1 5.5e-6 K-1
Thermal conductivity 1.4 W/mK 1.7 W/mK
Specific heat 1000 J/kgK 1015 J/kgK
Poisson's ratio 0.3 0.3

Table 6. The temperature dependent yield stress under tensile loading conditions for LCC Andalusite HS.

Temperature Yield Strain V- modulus

(ºC) stress (10-3) (N/mm2)
(MPa)
20 (70°F) 14 0.5 28 000
400 (750°F) 15 0.7 21 428
800 (1500°F) 16 0.8 20 000
1000 (1800°F) 18 2.5 7 200
1200 (2200°F) 11 13.8 797

Table 7 The temperature dependent yield stress under tensile loading conditions for Phosphate bonded SB

Temperature Yield Strain V- modulus

(ºC) stress (10-3) (N/mm2)
(MPa)
200 (70°F) 5.7 0.6 9618
600 (1100°F) 11.1 1.2 9502
800 (1500°F) 7.9 0.8 10368
1100 (2000°F) 5.1 1.4 3596
6 Results

6.1 FEM Analysis for the Andalusite HS burner tile

From the simulations it turned out that most of the damage occurred within the first hour
of heat up. Figure 4 gives the calculated temperature contours after one hour of heat-
up. Calculated displacements have been scaled by a factor of 50 so that one can clearly
see the trumpet like shape of the burner block. Highest stresses occur in circumferential
direction. Figure 5 gives the calculated stresses in the circumferential direction.
Calculated stresses are close to the yield stress limit for much of the outer burner block
region. The calculated plastic deformation depicted in Figure 6 indicates where cracks
are most likely to be initiated. Notice that crack deformation in reality may be more
severe because of the brittle nature of ceramics in which the calculated regions with
plastic deformation fail to carry any normal stresses.

Figure 4 Temperature contours after one hour of heat up (°K). A displacement-scaling factor of 50 was used.
Figure 5 Calculated stresses in circumferential direction (Pa)

Figure 6 Calculated plastic strain in circumferential direction (non reversible strain, Epsilon
6. 2 FEM analysis for the Phosphate bonded SB

From this simulation it turned out that no damage occurred within the first hour of heat
up. Figure 7 shows the calculated temperature profile after one hour of heat-up. Figure
8 gives the calculated stresses in the circumferential direction. Compared to Figure 5
the stresses are lower than the yield stress stated in table 6.

Figure 7 Temperature contours after one hour of heat up. A displacement-scaling factor of 50 was used.
Figure 8 Calculated stresses in circumferential direction.

6.3 Calculation of Hasselmann parameter

The calculations of the “Hasselmann thermal stress resistance parameters R1-R3” are
shown in Table 8 and 9.
In the critical temperature range up to 850°C (1560°F) R1, R2 and R3 of Phosphate
bonded SB shows better numbers compared to Andalusite HS. Basically this is a
confirmation of the results from FEM analysis and means the phosphate bonded
material allows a higher temperature change; heat flux and temperature increase at the
surface without exceeding the failure stress compared to the cement bonded material.
However, a material ranking based on these results is difficult because above 1100°C
(2000°F) both materials form melted phases and show plastic behavior. Hasselmann
assumes only elastic properties.
Hence the quality of a material choice based on thermal stress resistance parameter
calculations would increase with the data base and comparative FEM analysis.

Table 8 Stress parameter of Phosphate bonded SB

Temperature (°C) R1 (°C) R2 (W/m) R3 (m2°K/s)

200 (390°F) 75 128 5,34 e-5

600 (1100°F) 149 253 9,16 e-5
800 (1500°F) 97 175 5,89 e-5
1100 (2000°F) 178 378 12,10 e-5
Table 9 Stress parameter of Andalusite HS

Temperature (°C) R1 (°C) R2 (W/m) R3 (m2°K/s)

20 (70°F) 54 75 3,1 e-5

400 (750°F) 74 103 4,2 e-5
800 (1500°F) 84 117 4,7 e-5
1000 (1800°F) 268 375 15,3 e-5
1200 (2200°F) 1453 2034 82,9 e-5

Conclusion

Comparative FEM analysis of a burner tile made of low cement bonded andalusite and a
phosphate bonded castable have been conducted.

Under given thermal boundary conditions it is likely that the cement bonded material
would show cracking because the calculated stresses (Figure 5) are close to the yield
stress limit (Table 6). It can be concluded that improved properties can be achieved by
using a cement bonded castable of higher cement content in order to lower the stiffness
in this temperature range.
The results of the phosphate bonded material “SB” show a lower V-modulus, higher
thermal conductivity and a lower coefficient of thermal expansion compared to the low
cement bonded andalusite material. The determining low V- modulus is characteristic for
phosphate bonded material and can be explained with the formation of poly- and meta -
phosphate compounds during the condensation process which leads to flexible ring- and
chain structure with a stability up to 800°C (1500 ºF).
Due to this and the properties explained under 4.2.1 mono-phosphate bonded materials
do not require dry-out and therefore it is possible to manufacture burner tiles “just in
time” with an installation right after setting.

Although this FEM analysis is only focused on the heat up process of the tile it shows
the material behavior at a very critical state of the temperature cycle. This is confirmed
by calculations of the thermal stress resistance parameters R1-R3 which show different
results of the two test materials in this temperature range.

Both FEM analysis and the calculation of stress parameter are practicable tools for
materials ranking. A ranking based on stress parameter calculations however depends
on a comprehensive data base. Traditional material testing procedures have shown to
be inaccurate predictors of the performance of phosphate bonded materials in this
application. One has to consider that tests under load conditions show results which are
probably influenced by temporary melt phases with lower viscosity and plastic flow.
Therefore it is recommended to compare results achieved by using varied and extended
holding times.
1.) Baukal C.E.,”The John Zink Combustion Handbook,” pp 303. 2001
2.) Schacht C.A.,”Thermomechanical Behavior of Refractories” in: “Key Engineering
Materials”, Vol.88. 1993 pp 204-206
3.) Hasselmann D.P.H., “Thermal Stress Resistance parameters for brittle refractory
ceramics,” Bulletin of the Amer. Ceram. Soc. 49. 1970. pp 1033 - 1037
4.) Parr C., Bier T.A., Bunt N.C., Spreafico E.E.,”Calcium Aluminate Cement based
castables for demanding applications,” Lafarge Technical Paper
5.) Parr C., Bier T.A., Mathieu A., Spreafico E.E, “Calcium aluminate cements for
monolithic refractories,” Proceedings of the 1 Monolithic Conference, Teheran, Iran;
1997
6.) Petzold, Ulbricht,”Feuerbeton und betonartige feuerfeste Massen und Materialien,”.
Deutscher Verlag fuer Grundstoffindustrie, Leipzig. pp40- 41
7.) Schulle W. “Feuerfeste Werkstoffe,” Deutscher Verlag fuer Grundstoffindustrie,
Leipzig. pp352 – 357
8.) Plibrico Japan,”Technology of Monolithic Refractories,”. 1999. pp148 – 164
9.) Krebs R.”Unshaped Refractory Materials,” 1997. pp193 in: Routschka G. “Refractory
Materials.” 1997. pp 193
10.) Plibrico Japan,”Technology of Monolithic Refractories.” 1999. pp225 - 225
11.) Schacht. Charles A.,”Refractory Linings”. New York 1995. pp 157 - 192
12.) F.J Gonzales, J.W Halloran “Reaction of Orthophosphoric- Acid with Several Forms
of Aluminum Oxide,” Amer. Ceram. Bulletin Vol. 59 No.7, 1980. pp 727
13.)Bartha P., Lehmann H., Koltermann M.,”Untersuchungen zur Bindung keramischer
Werkstoffe mit H3PO4”, Ber. Dt. Keram. Ges. 48 (1971) H3. pp111
14.) Bothe J.V., Brown P.W.,”Reactivity of Aluminatowards Phosphoric Acid”, J. Am
Ceram Soc. 76 (9) pp 2353-58 (1993)
15.) Plibrico Japan,”Technology of Monolithic Refractories,” 1999. pp 237
16.) Appelt R., “Aluminiumoxid und seine Bindung mit Monoaluminiumphosphat,”
Sprechsaal, Vol.116, No.9, 1983
17.) Plibrico Japan,”Technology of Monolithic Refractories,” 1999. pp 175