See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/333746980

Laboratory Testing Methods up to 1600 °C

Chapter · June 2019

CITATION READS

1 3,305

2 authors:

Peter Quirmbach Almuth Sax

University of Koblenz and Landau University of Koblenz and Landau
61 PUBLICATIONS 138 CITATIONS 21 PUBLICATIONS 36 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Peter Quirmbach on 18 July 2019.

The user has requested enhancement of the downloaded file.

1. Testing methods for Corrosion
1.1 Laboratory Testing Methods
1.1.1 Testing up to 1600°C Peter Quirmbach, Almuth Sax

Introduction
Besides erosion and mechanical abrasion, corrosion is one of the major reasons for the
disastrous failure of refractory materials and/or components. At least one of the aforementioned
categories of wear on refractories occurs during the use of these materials and components in
industrial applications, but in most cases even several of these wear types come about at the
same time.
To describe the corrosion behavior of refractory materials or components, these are subject to
test methods to differentiate recently developed types by means of their characteristics, or to
test and compare several materials at the same time with the aim to detect their respective
differences. For this purpose, test substances are regarded as a “system” in the physico-chemical
sense, and the different corrosive mediums (melts, slags, salts, gases) are classified as the
“environment”. Consequently, all interactive reactions can be described thermodynamically
and kinetically.
Corrosion provoked by oxide melts and metal melts
Test procedure in lab scale:
Procedures for testing the aggressive potential of inorganic or metallic melts are described in
the German standard DIN51069 or in the European draft standard CEN/TS 15418 as well as in
various ASTM standards. Both European standards include various static and dynamic test
procedures to study corrosion caused by melts under laboratory conditions.
 Pill test and disc test
 Crucible test
 Induction furnace test
 Test with rotating sample - Finger test by dipping
 Test with rotating slag - Rotary furnace test

 Pill test and disc test

The first method operates as a comparative test while a drop or “pill” of a slag sample is placed
on the surface of a refractory material or in a cavity drilled in the refractory materials surface.
After heating up this configuration the depth of penetration of the slag inside the refractory
material as well as the spread of the molten slag can be examined in means of corrosion
interaction. Another kind of interaction can be achieved using the “disc test” where powders of
refractory sample material and solid salt with a maximum particle size of 0.5 mm each are
mixed at the rate of 7 : 3. Then the mixture is pressed with a force of 80 MPa to discs with a
diameter of 50 mm and a thickness of 10 mm. Commonly temperatures of 1100  T  1300°C
and dwell times 5  t  50 h are used. As result changes in sample diameter, weight and length
as well as impacts of corrosion reactions like cracks, decomposition effects and color changes
can be observed [1].
 Crucible test

The standard [2] requires cubic crucibles of 80 x 80 mm² edge length, and a height of 65-76 mm.
The hole to insert the inorganic or metallic metal melt is drilled centrically into the top face. It
has a diameter of 40    42 mm and a depth of 35  h  55 mm. After having filled the
crucible with the corrosive material, the crucible has to be covered with a lid, which must have
the same material as the crucible itself. Fig.1 shows the dimensions of the crucible in a
schematic drawing. Temperature and duration of the test are defined precisely to describe the
future real application conditions as accurately as possible. Usually, T = 1500°C is the upper
temperature limit. According to standard, a constant heating rate of 1 to 4 K/min is
recommended, and a free cooling of the furnace and crucible after the end of the test.
After the cooling process, the crucible is cut along the vertical axis of the hole and diagonally
through the top face of the cube. Documentation of the corrosive attack is carried out by
measuring the material loss at different points along the vertical axis of the hole, the dimensions
of the zone with a visible penetration of the microstructure and, in some cases, the dimensions
of the zone with discolorations, e.g. provoked by iron or other coloring ions. Fig.2 shows an
example of a cut crucible after the test.

crucible (top view)

80
(100)

41  1 (50)

80 (100)

lid 10

65
(100)

crucible (side view)

Fig.1 - Schematic drawing Fig.2 - Example of a cross section of the crucible after
of the crucible and testing
lid dimensions
In addition to standard evaluations of the dimensions of those zones that are influenced by
corrosion, the microscopic analysis (light microscopy LM and/or Scanning Electron
Microscopy SEM) of the microstructure gives detailed information on the chemical interaction
between the treated material and the corrosive mediums, providing evidence of the dominant
corrosion mechanism as well as the kind of diffusion.
 Induction furnace test

For testing the corrosion resistance of refractories against metal melts, e.g. steel, aluminum or
copper alloys, the induction furnace test is the favored method. The melt is generated via
induction of the metallic feed material. The refractory test material can either be installed as a
lining of the vessel, or in the form of a crucible made of a certain refractory material, or in the
form of a crucible composed of different prisms (as in Fig.6). Furthermore, it can be dipped
into the melt in the form of one or more fingers. Additionally, a covering slag can be involved
for testing also the corrosion resistance against slags, and at the three-phase-boundary between
the refractory material, the metal and the slag. Fig.3 shows examples for both installation
situations during the induction furnace test. Owing to the use of an inductive field for heating,
the melt is continuously moved during the test. As a consequence, both installation methods
permit dynamic corrosion testing with a certain relative velocity between melt and sample.

Fig.3 - Induction furnace test. Left : refractory material is installed as the lining of the
vessel, right : finger test installation with four fingers.

Main advantage of the test version with refractory lining is the temperature gradient along the
refractory wall thickness, which simulates real thermal conditions during the test. The furnace
vessel should have a volumetric capacity of at least 2 liters melt, corresponding to appro-
ximately 15 kg steel and 1-2 kg slag. Usual dimensions of the vessel are  = 220 mm and
h = 250 mm ( = diameter; h = height). In the case of a crucible composed of prisms, the
refractory samples were cut from standard bricks of l = 230 mm length; the dimensions of cold
and hot faces were similar to those of the rotary furnace test, i.e. a length of l = 78 mm (longer,
cold face), and about l = 45 mm (smaller, hot face), respectively, for an installation with 8
samples. Fig.6 shows a schematic top view of an arrangement with 8 prismatic samples in the
induction furnace. A possible disadvantage of this arrangement is the risk of interactions
between the neighboring samples, if incompatible materials are placed side by side.
 Fig.4 - setup for atmosphere control during
induction furnace test with
left : inert gas inlet (i.e. Argon) and
right: ZrO2–pO2-measuring probe

Fig.5 - periodically extraction of samples

of slag during test runs to follow kine-
tics of corrosion

The duration of the test depends on the requirements of future operational conditions. As
recommended by the standard CEN/TS 15418 [3], metal and slag samples should be taken at
the beginning of the test, and after every 30 minutes to determine the change in chemical
composition, and a possible saturation of the corrosive medium on account of dissolved
refractory components.
To analyze the corrosive attack, the samples are cut along their vertical axis as shown in Fig.7
for two examples: one with and the other without adhering metal melt residues. The corrosion
resistance was determined according to standard instructions: detecting the loss in volume at
different points along the vertical axis of the sample. Additionally, the extent of penetrated
volume and discolored microstructure were determined. As shown in the examples in Fig.7, the
maximum loss of material is to be expected in the area of the three-phase boundary between
refractory lining – metal – atmosphere, or refractory – metal – slag as a result of the forced melt
flow due to the Marangoni convection in this area [4].

8 2

7 137.61 3

57
6 4

5
Fig.6 - Schematic top view of arrangement with 8 prismatic
samples in the induction furnace.

57
25

77.71
 Fig.7 - Refractory samples after

Hi
the induction furnace test
– cross section along the
left : vertical axis of the
Ti prisms with adhering metal
melt, and
right : with a particularly
strong extent of erosion in
the three-phase boundary
without adhering melt

The second possibility to conduct the induction furnace test is to introduce the refractory
samples in the form of fingers, which are dipped into the melt. The dimensions and the capacity
of the furnace vessel were the same as in the first test. Although this arrangement has the
disadvantage of a homogeneous temperature profile across the sample, the possibility of testing
a few samples at identical test conditions without the risk of an undesired interaction of the
materials is the main benefit of this method. A forced melt movement can be generated through
rotation of the refractory fingers during the corrosion test. Furthermore, cyclic and vertical
movements (lifts) of the sample during this test allow simulating a superimposed thermal shock.

Fig.8 - Finger sample after induction furnace test

Fig.8 shows a finger sample after an induction furnace test. For evaluation purposes, the change
in dimension, the volume after corrosion and the extent of infiltrated microstructure are usually
determined.
 Test with rotating sample - Finger test by dipping

An alternative test procedure to detect the corrosion of inorganic melts (especially glass melts)
is the test with a long, cylindrical sample (“finger”), which is dipped into a crucible filled with
melt of an inert material, e.g. platinum. The finger is dipped vertically into the melt under
isothermal conditions; temperature and duration are chosen in dependence on the requirements
of later applications. A typical duration is 24 hours or 48 hours. There are two different kinds
of test performance: stationary or dynamic. Either the fingers or the crucible rotate. Fig.9 shows
a test setup with two fingers in a dynamic arrangement with rotating crucible.
 Fig.9 - Setup of a finger test with two fingers in dynamic arrangement with rotating
crucible

To evaluate the corrosion resistance of the material, a visual inspection of the remaining sample
volume is to be carried out first. After having cut the sample along the vertical axis, the residual
diameter of the finger is to be measured.

Fig.10 - Samples ofter glass corrosion tests prepared for measuring materials loss

In most cases, slag residues will adhere to the sample surface. Therefore, a binocular loupe
should be used. The remaining sample volume or, respectively, the volume after corrosion can
be determined, if a clear separation of the sample material and the adhering slag is possible.
 Test with rotating slag - Rotary furnace test

Main disadvantage of the described test procedures in the previous chapters are the isothermal
conditions, where the sample material has a homogeneous temperature profile along the
complete cross section. The rotary furnace test allows studying the corrosion resistance under
more operational conditions: a hot face in contact with melt, and a temperature gradient in
direction of the cold outer face of the sample. The test procedure is consistent with the European
standard CEN/TS 15418 [3] and the standard ASTM C 874 [5].
For this test, prismatic samples are installed in a cylindrical chamber, which is heated by a gas
burner to Tmax. = 1700°C. The temperature control is accomplished by 3 thermocouples, 2 inside
the prisms and 1 in the bottom of the reaction vessel. The furnace chamber rotates around its
horizontal axis (max. 5 revs/min.) and enables, in this way, a dynamic movement of the melt.
For this installation, prismatic cutted samples are placed into the chamber as illustrated in
Fig.11. The usual furnace chamber dimensions are: 400    450 mm diameter and
500  l  600 mm length. The samples are cut from standard bricks of l = 230 mm length, and
the prisms have a longer (cold) face of l = 78 mm, and a smaller (hot) face of l = 45 mm (see
example in Fig.6 with 8 segments).
The rotary furnace test is commonly applied for inorganic melts like slag, so-called “secondary
fuels”, or cement clinker, which is introduced as a fine granulate before the beginning of the
heating process. The atmosphere during the test depends on the used burner and the target
temperature. Usually, a propane/air burner is employed. Although the use of the burner initially
leads to an oxidizing atmosphere, it is possible to achieve a desired reducing character of the
atmosphere. This can be attained either by lowering the air volume in the burner or, if high
temperatures are required, by simultaneously adding carbon, in the form of coke or anthracite
granules, to the furnace chamber (see Fig.12).

Fig.11 - Arrangement of 8 pris- Fig.12 - Rotary furnaces with gas burner in operation
matic samples in the rota-
ry furnace test with dark
melt residues at the inner
hot face

The analysis of specimen prisms is carried out accordingly to the test procedure with the
Induction furnace test (see Fig.6 and Fig.7).
Corrosion provoked by salts
The corrosion of refractories caused by salts is a crucial factor in applications, where alkali
compounds are in contact with the refractory material. In most applications, alkalis are present
in the atmosphere in the form of gaseous compounds like chlorides or sulfates, which can react
after their condensation with formed melts that have a low melting point. In the worst case,
solid alkali aluminosilicate phases like nepheline with a comparable high specific volume are
formed. The crystallization of theses phases is associated with a high crystallization pressure,
thus leading to a strong crack formation or even complete destruction of the refractory. The
following equation describes an example of the ideal stoichiometric formation of nepheline
from mullite and sodium oxide, which is thermodynamically favored at temperatures below
T  1256°C and coupled with a volume increase of V  18%.

3 Al2O3  2 SiO2 + Na2O  Na2O  Al2O3  2 SiO2 + 2 Al2O3 (1)

A 100% simulation of operational conditions is hardly feasible. Great efforts are required to
generate appropriate alkali concentrations in the gas phase. Moreover, strong corrosion of the
test equipment has also to be taken into account.
For this reason, crucible tests (see chap. Crucible test) are commonly used in laboratory scale
to evaluate the alkali resistance of refractories. These tests are mainly carried out with solid
alkali carbonates to represent an alkali containing corrosive medium. Although the test
conditions are far away from those conditions in industrial applications, it is at least possible to
compare the alkali resistance of different refractory materials. Additionally, the destructive
effect of an alkali attack under operational conditions is accelerated by using highly
concentrated solid alkali compounds. Therefore, the tests can be conducted in an appropriate
time in comparison to the real reaction period.
A standard test procedure is recommended by ASTM C 454 [6] to test the alkali resistance of
carbon refractories. This standard is also accepted for non-carbon containing refractories.
Crucibles were cut from standard bricks with cube dimensions that are either according to
standard ASTM C454: l = 51 mm edge length, a hole diameter of  = 22 mm and a depth of
d = 25 mm or, alternatively, according to standard CEN/TS 15418, which recommends as
follows: an edge length of l = 80 x 80 mm² and a height h = 65-76 mm, a hole diameter of
 = 40-42 mm and a depth d = 35-55 mm. ASTM C454 recommends to use K2CO3 as a
corrosive alkali compound. Depending on the requirements of future application conditions,
also Na2CO3 or mixtures of K2CO3 and Na2CO3 are used. In order to avoid a considerable alkali
loss via evaporation, crucibles must be covered with a lid made of the sample refractory.
The test temperature is about T = 1000°C, the recommendation of ASTM C 454 is T = 955°C
due to the fact that formation of the alkali compounds preferentially proceeds in this
temperature range. A test duration of t = 5 hours is recommended.
After cooling, the alkali resistance is evaluated by means of an optical inspection of the crucible.
Fig.11 shows two examples of refractory materials with low alkali resistance after testing. A
strong swelling of the upper part of the crucible and a massive crack formation can be observed.
It is difficult to determine the direct specific values that characterize the alkali resistance.
Reasons for this are the specific damage symptoms. Evaluation of the alkali resistance is,
therefore, done by classifying the materials into four different damage groups: U (unaffected),
LC (light affected), C (cracked) and D (disintegrated).
 Fig.13 - Examples of refractory crucibles after alkali tests

Corrosion provoked by gases

The interaction of a corrosive ambient atmosphere with the refractory, in general, leads to a
change in weight of the material. Mainly the formation of gaseous reaction products causes a
loss in weight, which implies active corrosion. Example for active corrosion is: the reduction
of oxide refractory in strongly reducing atmospheres like hydrogen or carbon monoxide.
Processes that cause an increase in weight are denoted passive corrosion, whereby mostly solid
or partly liquid reaction products develop. The prime example is the oxidation of silicon carbide
with the formation of a solid silica layer on the SiC surface. The silica layer passivates the
material and acts as a protective layer, which generally makes possible to use silicon carbide
for high temperature applications. However, also the formation of the protective silica layer
consumes the refractory SiC, and at a certain point too strong oxidation leads to a shorter service
life of the lining.
Third, corrosion caused by gas can emerge when new solid phases form in the bulk of the bricks,
since these phases cause damage due to their crystallization pressure. The most important
example is the carbon-induced bursting of refractories, when they are exposed to carbon
monoxide atmosphere in applications, where a cyclic temperature change occurs within the
temperature range of about T = 500°C. At high temperatures, carbon monoxide is
thermodynamically stable. When the temperature decreases below approximately T  500°C,
carbon dioxide becomes the stable gaseous species, and solid carbon will be deposited,
according to Boudouard´s equilibrium :

CO2(g) + C(g)  2 CO(g) (2)

The transformation of CO to CO2 and, thus, deposition of solid carbon is strongly kinetically
inhibited. It can only proceed in the presence of a catalyst, especially iron oxide. If no catalyst
is available, carbon deposition normally does not take place or is very limited.
Test procedure in lab scale:
The most common method to study gas corrosion phenomena is to carry out thermogravimetric
measurements in a thermobalance (Fig.14). These allow detecting a change in weight easily at
different temperatures in dependence on time. Major disadvantage of this method is the strongly
limited sample volume: only some milligrams of almost every available material can be used.
An alternative is the alteration of cylindrical or prismatic samples with a representative volume
in a closed furnace that is purged with the corrosive gas, and to measure the weight change after
certain intervals. Besides the enormous effort to be made when applying this method, periodic
cooling and heating of the samples can influence the result.

Fig.14 - combined TG-DSC measuring device for various atmospheres up to T = 1600°C

To reduce the experimental effort, it is also possible to determine the corrosion resistance of
refractories against gaseous attack by using other special methods, which focus on specific
corrosion problems. Some examples of standard and non-standard test procedures for important
applications are as follows:
 steam test for silicon carbide according to ASTM C863 [7],
 resistance against carbon monoxide according to ASTM C288 [8] and as per European
standard EN ISO 12676 [9], and
 resistance against hydrogen corrosion (non-standard).

 Corrosion caused by oxidative gases – oxygen and steam in contact with silicon carbide

The oxidation resistance of silicon carbide refractories is one major criterion for the use of
linings in plant operation at elevated temperatures, for instance in waste incineration plants. On
the one hand, oxidation leads to passivation of the SiC surface and, therefore, is a desired effect
to some extent. On the other hand, oxidation of SiC leads to an increase in volume and can
cause serious damage to the lining. The standard ASTM C863 recommends a test method,
which provides the volume expansion of silicon carbide bricks in an oxidizing environment.
The use of steam in the test procedure increases the oxygen attack, in contrast to the use of pure
oxygen gas. Since hydroxide ions will be incorporated into the glassy network of the oxide
protective layer on the SiC surface, the diffusion rate will increase significantly. This leads to
an amplified oxidative attack on the material generating test results with an acceptable
expenditure of time.
The method according to ASTM C863 refers to prismatic samples with dimensions
25 x 25 x 150 mm³ (prism for cold bending strength test). The samples are placed into a closed
furnace chamber, which is purged with a constant steam flow of minimum 32 kg H2O/m³. The
test temperature is 800°C  T  1200°C depending on the requirements during subsequent
applications of the bricks. The test duration should be t = 200 hours, according to standard
recommendations. After cooling, the change in volume is measured. Usually, the maximum
acceptable volume increase depends on the requirements of future applications. Additionally,
microstructural analysis can help to interpret the test results. Fig.15 shows samples that are
unaffected and those who are strong corroded (SiC brick with low steam resistance).

Fig.15 - results from steam treatment of SiC samples left : unaffected, right : destroyed

 Carbon monoxide corrosion

The corrosive attack on refractories caused by carbon monoxide happens on account of solid
carbon deposition at temperatures T  500°C. Since carbon deposition is possible only in the
presence of an iron oxide catalyst, the initial recommendation is to restrict the iron oxide
concentration in refractories for their use in CO-atmospheres, and during possible operation
conditions of T  500°C. However, also very low iron oxide concentrations can cause bursting
of the material, since even single iron containing particles in the bulk of the brick can be the
cause of damage.
The test method for disintegration of refractories in carbon monoxide atmosphere is satisfying
the requirements of ASTM C288 and the European standard EN ISO 12676. The samples have
the dimension of half a standard brick, i.e. 230 x 64(54) x 64 mm³, and they are stored in a
chamber furnace, which is purged with 100% CO atmosphere at T = 500°C. Test duration is set
to t = 200 hours, according to standard recommendations, although in most cases carbon
deposition occurs during the first 24 hours.
 Fig.16 - Refractory castable samples after Fig.17 - Micrograph of an included iron par-
experimental CO-Test ticle (white) with surrounding car-
bon deposition (black) in a alumina-
rich refractory concrete [10]

The carbon monoxide resistance is evaluated by macroscopic inspection of the samples, and
classified into damage categories according to ASTM C288:
Ⓐ: no visible defects, no carbon deposition, no spalling or cracks;
Ⓑ: visibility of some surface defects: spalling size is smaller than 10 mm;
Ⓒ: cracks are visible, spalling size is larger than 10 mm;
Ⓓ: sample is broken into pieces (two or more), or it can be broken manually;

A similar classification is used in the European standard EN ISO 12676 providing damage
classes in serial numbers from 1 to 4.
 Hydrogen corrosion

Corrosion of refractory materials caused by hydrogen attack occurs, for instance, in heat
treatment of steel or other metal products, as well as in operational plants of chemical or
petrochemical industries. Especially oxide refractories that contain SiO2 undergo a serious
attack due to active corrosion, depending upon temperature and flow velocity of the gas stream.
Main mechanism of damage is the evaporation of the SiO2 constituents by formation of the
gaseous SiO(g) suboxide according to the following equation:

SiO2(s) + H2(g)  SiO(g) + H2O(g) (3)

As follows from the above equation, the concentration of residual steam in the hydrogen gas
also influences the degree of attack. According to Le Chatelier´s principle, the formation of the
gaseous SiO will decrease with increasing hydrogen concentration H2O(g). On the contrary,
hydrogen attack will become stronger, if dry hydrogen with low residual H2O is used.
The laboratory test procedure to estimate the hydrogen resistance of refractories has not been
previous standard. The prismatic samples with dimensions 25 x 25 x 150 mm³ (prism for cold
bending strength test) are placed into a closed furnace at a temperature range of
1000°C  T  1600°C in a pure hydrogen atmosphere. The number of samples should be
limited to 5 samples/10 liters furnace volume to avoid saturation of the gas atmosphere with
gaseous corrosion products during testing. Fig.18 shows the samples arranged in a chamber
furnace with electric heating to carry out the hydrogen test. Duration of the test should be at
least t = 200 hours, since the selective evaporation of refractory constituents provoked by active
corrosion is rather slow, and the purging rate used under laboratory conditions is limited. The
use of dry hydrogen (below 100 ppm residual steam) is recommended for oxide refractories.
Special studies of SiC materials could also be carried out in wet hydrogen (100 - 500 ppm
residual steam), on account of a weak corrosion resistance of SiC under a reducing, steam
containing atmosphere.
Corrosion progress is monitored during the test by measuring: the change in weight, the bulk
density and the dynamic modulus of elasticity (interruption of heating). As an example, Fig.19
represents chamotte samples after the hydrogen test and the respective change in weight, in
dependence on test duration. The grey discoloration is a typical symptom of a reducing attack
on oxide refractories, in general. After the corrosion test, destructive mechanical methods are
applied to characterize the corrosive attack: e.g. cold bending strength, static modulus of
elasticity and cold crushing strength.
The microscopic and energy dispersive X-ray spectroscopy (EXD) analysis after testing helps
to identify critical microstructural constituents and the weak point in microstructural design of
the material. Fig.20 and Fig.21 display the micrographs of the used chamotte samples shown in
Fig.19.

Fig.18 - Sample arrangement in the Fig.19 - Chamotte samples after the hydrogen test
furnace for hydrogen testing at 1300°C/200 hours in dry hydrogen,
corresponding graph of change in weight
in dependence on time.
 Fig.20 - Micrograph of chamotte sample’s Fig.21 - Element mapping (by EDX) of Sili-
near-surface microstructure after con and Aluminum in the near-sur-
hydrogen testing at T = 1300°C/200 face microstructure shown in Fig.20
hours in dry hydrogen – Silicon is depleted due to hydrogen
attack in the first app. 100 µm of the
surface

Outlook
In addition to the standard testing methods described in this paper, there is a wide variety of
further, specifically developed procedures to measure corrosion. These latter procedures
display their own technological conditions, depending on each individual application field,
particularly when selecting the corrosive mediums. Further measurement methods are currently
under development. Amongst others, these methods permit the superimposition of corrosive
and, for example, mechanical stresses at the same time. Consequently, they serve to simulate
interactions under real operational conditions. A further individual scientific field is the CAT
(Computer aided thermochemistry). For this field, the software (e.g. FACTSAGE) [11] is
applied to calculate the thermochemical reactions between the system and the environment,
which in turn allows comprehensive simulations.

References
[1] C. G. Aneziris, U. Fischer, E. Schlegel, The use of secondary fuels (waste) and the
Corrosion problems of Refractories, in: Proc. XVI. Int. Conf. Refractories, Prague, 2008:
177-184
[2] DIN 51069-2(1972) Testing of ceramic materials; comparing test of the resistance of
refractory bricks to the attack of solid and liquid materials at high temperature, crucible
method
[3] DIN CEN/TS 15418:2006-09 Methods of test for dense refractory products – Guidelines
for testing the corrosion of refractories caused by liquids
[4] K.G. Nickel, P. Quirmbach, J. Pötschke, High Temperature Corrosion of Ceramics and
Refractory Materials, in: Tony Richardson et. al. eds., Shreir´s Corrosion, 1st edition
Amsterdam: Elsevier B.V.; 2009: 668-690
 [5] ASTM C874-11a Standard Test Method for Rotary Slag Testing of Refractory Materials
[6] ASTM C454-10(2017) Standard Test Method for Disintegration of Carbon Refractories
by Alkali
[7] ASTM C863-00(2016) Standard Test Method for Evaluating Oxidation Resistance of
Silicon Carbide Refractories at Elevated Temperatures
[8] ASTM C288-87(2014) Standard Test Method for Disintegration of Refractories in an
Atmosphere of Carbon Monoxide
[9] DIN EN ISO 12676:2003-05 Refractory products – Determination of resistance to carbon
monoxide
[10] Principles for the wear of refractory materials especially castables, caused by CO-
atmosphere, AiF-Report N 14161, Forschungsgemeinschaft Feuerfest e.V.
[11] FactSage 7.0, http://www.gtt-technologies.de/factsage

Authors
Quirmbach, Peter, Prof. Dr. rer. nat. Dr. h.c., DIFK Deutsches Institut für Feuerfest und
Keramik GmbH, Rheinstr. 58, 56203 Höhr-Grenzhausen, GERMANY
Sax, Almuth, Dr.-Ing., University Koblenz-Landau, Chair for Technical Chemistry and
Corrosion Science, Universitätsstr.1, 56070 Koblenz, GERMANY

View publication stats