Garcia Prieto Ana

FACULTAD DE CIENCIAS
DEPARTAMENTO DE QUMICA INORGNICA
CARACTERIZACIN MECNICA DE
MATERIALES CERMICOS MEDIANTE
ENSAYOS DE FRACTURA ESTABLE
Memoria presentada por
Ana Garca Prieto
Para optar al grado de
DOCTORA EN CIENCIAS QUMICAS
Directora:
Dra. Carmen Baudn de la Lastra
Tutora:
Dra. Ana M Gonzlez Vadillo
INSTITUTO DE CERMICA Y VIDRIO
CONSEJO SUPERIOR DE INVESTIGACIONES

CIENTFICAS
MADRID 2015
Esta tesis ha sido realizada en el Departamento de Cermica del Instituto de
Cermica y Vidrio (ICV) del Consejo Superior de Investigaciones Cientficas
(CSIC). Se enmarca dentro de los proyectos financiados por el Ministerio de
Economa y Competitividad: MCI-MAT2006-13480-C02-01, MCI-MAT2009-
14369-C02-01, MCI-MAT2009-14448-C02-01, CICYT-MAT2010-17753 y
MAT2013-48426-C2-1-R, por la empresa Microtest S.A.: MCI-PET2008-0113 y
por la red CYTED: HOREF312RT0453.
La autora agradece la financiacin econmica otorgada por el Consejo Superior

de Investigaciones Cientficas y el Fondo Social Europeo a travs de la beca
JAPre_2010_00274 para desarrollar esta tesis y para realizar una estancia de seis
meses en Oxford University (Reino Unido). Adems, agradece a la fundacin
JECS TRUST la financiacin econmica para realizar una estancia de tres meses
en Montanuniversitt Leoben (Austria).
RESUMEN
El objetivo principal de este trabajo ha sido el desarrollo de un nuevo dispositivo
experimental y una metodologa de ensayo y anlisis de resultados para determinar la
tenacidad de fractura de materiales cermicos, que puedan ser utilizados como forma de
caracterizacin de rutina en el laboratorio.
Se ha planteado como hiptesis bsica la adecuacin de los ensayos de fractura
estable para la determinacin de la tenacidad de fractura de los materiales.
Se ha propuesto la utilizacin de la apertura de los labios de la grieta como
parmetro de control puesto que, junto con el tamao de la grieta, crece de manera
montona durante todo el ensayo (carga y fractura).
Hasta la fecha de inicio de este trabajo slo se haban reportado ensayos de fractura
estable con control por apertura de los labios de la grieta en el caso de materiales no
frgiles como ytria (Y-PSZ) y hormigones de alta resistencia.
El dispositivo experimental consta de un micrmetro ptico asociado a un
dispositivo de flexin en tres puntos y se ensayan probetas prismticas con entallas
rectas en forma de V (SEVNB).
Con el fin de demostrar la validez del dispositivo experimental y la metodologa de
ensayo y anlisis de resultados desarrollados, se ensayaron materiales cermicos con
comportamientos mecnicos muy diferentes:
Espinela de aluminio-magnesio (MgAl2O4), con comportamiento frgil.
Almina (Al2O3) de grano fino, con comportamiento frgil.
Mullita (3Al2O3.2SiO2), con comportamiento frgil y crecimiento subcrtico de
grietas en aire significativo.
Refractarios comerciales, en los cuales ocurren mecanismos de refuerzo durante la
fractura.
En todos los casos se alcanzaron ensayos de fractura estable demostrndose as la
adecuacin de la utilizacin de la apertura de los labios de la grieta como parmetro de
control.
Se estudi la influencia de las condiciones experimentales en la tenacidad de
fractura. La utilizacin de la apertura de los labios de la grieta como parmetro de
control minimiza el efecto de factores como la inercia de la mquina y permite
identificar el crecimiento subcrtico de grietas.
El valor del factor crtico de intensidad de tensiones en modo I, KIC, obtenido a
partir de ensayos estables es considerablemente inferior al obtenido a partir de ensayos
inestables.
A partir de los ensayos de fractura estable fue posible obtener, adems del KIC, la
energa de fractura, que caracteriza todo el proceso de fractura. No es posible obtener
este parmetro a partir de ensayos inestables.
Una vez demostrada la validez del dispositivo experimental y de la metodologa de
ensayo y anlisis de resultados desarrollados, se prepararon y caracterizaron materiales
nanocompuestos de Al2O3/SiC y se ensayaron segn la metodologa propuesta,
aportando valores de tenacidad de fractura desconocidos hasta la fecha. Estos
resultados permitieron establecer las relaciones microestructura-comportamiento
mecnico en los nanocompuestos de Al2O3/SiC.
Los resultados obtenidos para el sistema modelo Al2O3/SiC permiten inferir que la
metodologa de ensayo y anlisis de resultados desarrollados son adecuados para la
caracterizacin de otros nanocompuestos cermicos.

NOTACIONES
a: Crack length / Longitud de la entalla (Figura 3.3)
B: Beam thickness / Espesor de la probeta (Figura 3.3)
CMCs: Ceramic Matrix Composites / Compuestos de matriz cermica
CMOD: Crack Mouth Opening Displacement / Apertura de los labios de la grieta
CMOS: Complementary Metal Oxide Semiconductor / Semiconductor

complementario de xido metlico
CNB: Chevron Notched Beam / Probetas con entalla triangular Chevron
CT: Compact Tension / Tensin compacta
D: Displacement / Desplazamiento
d50: Average grain size / Tamao de grano medio
DCB: Double Cantilever Beam / Probetas de doble voladizo
DE: Digital Edge-detection / Procesador digital de borde
DT: Double Torsion / Doble torsin
E: Youngs modulus / Mdulo de Young
E: Generalized Youngs modulus / Mdulo de Young generalizado
E.d.: Experimental details / Detalles experimentales
ESIS: European Structural Integrity Society / Sociedad Europea de Integridad

Estructural
G: Energy release rate / Tasa de liberacin de energa
Gc: Critical energy release rate / Tasa crtica de liberacin de energa
GIc: Critical energy release rate in mode I / Tasa crtica de liberacin de

energa en modo I
GF: Specific fracture energy / Energa especfica de fractura
GF/GIC: Toughness, flexibility or apparent ductility ratio
G: Energy release rate in the stationary state of R-curve / Tasa de liberacin

de energa en el estado estacionario de la curva R.
GPSSN: Gas Pressure Sintered Silicon Nitride / Nitruro de silicio sinterizado en
atmsfera gaseosa
HL-CCD: High speed Linear-Charge Coupled Device / Dispositivo de cargas

elctricas interconectadas de alta velocidad lineal.
HPSN: Hot Pressed Silicon Nitride / Nitruro de silicio prensado en caliente
IET: Impulse Excitation Technique / Tcnica del impulso de excitacin
IT: Indentation toughness / Tenacidad de fractura determinada por mtodos

de indentacin
ISB: Indentation Strength in Bending
K: Stress intensity factor / Factor de intensidad de tensiones
KI: Stress intensity factor in mode I / Factor de intensidad de tensiones en

modo I
KIc: Critical stress intensity factor in mode I / Factor crtico de intensidad de

tensiones en modo I
KI0: Crack tip toughness
K: Stress intensity factor in the stationary state of R-curve / Factor de

intensidad de tensiones en el estado estacionario de la curva R
L: Span between loading supports / Distancia entre apoyos inferiores

(Figura 3.3)
LED: Light-Emitting Diode / Diodo emisor de luz
LEFM: Linear Elastic Fracture Mechanics / Mecnica de la Fractura Elstica y

Lineal
m: Weibull parameter / Parmetro de Weibull
MgO-PSZ: Mg-partially stabilized zirconium oxide / Circona parcialmente

estabilizada con magnesia
P: Load / Carga aplicada
Pmax: Maximum load point / Carga mxima
R curve: Resistance to crack growth / Resistencia al crecimiento de grieta
ROR: Ring-on-Ring method
s: Brittleness number / Nmero de fragilidad

SCG: Subcritical Crack Growth / Crecimiento Subcrtico de Grietas
SCF: Surface Crack in Flexure / Flexin con una grieta superficial
S.D.: Standard Deviation / Desviacin estndar
SENB: Single Edge Notched Beam / Probetas con entalla recta
SEVNB: Single Edge Vee-Notched Beam/Probetas con entalla recta en forma de V
SEPB: Single Edge Pre-cracked Beam / Probetas con pre-entalla recta
SSiC: Sintered Silicon Carbide / Carburo de silicio sinterizado
t: Time / Tiempo
tf: Fracture time / Tiempo de fractura
T.D.: Theoretical Density / Densidad terica
TC6: Technical Committee Ceramics of the ESIS in Working Area 6
Y: Adimensional factor / Factor adimensional
YTZP: Yttria-stabilised tetragonal zirconia / Circona tetragonal policristalina

estabilizada con itria.
W: Specimen width / Anchura de la probeta (Figura 3.3)
4-pb: Four-point bending / Flexin en cuatro puntos
3-pb: Three-point bending / Flexin en tres puntos
a: Relative notch depth / Longitud de entalla relativa
ac: Critical relative notch depth / Tamao de entalla crtico relativo
d: Specimens deflection / Deflexin de la probeta
ds: Specimens deflection with crack/ Deflexin de la probeta con grieta
dm: Machine deflection / Deflexin del dispositivo experimental
ef: Unit deformation / Deformacin unitaria
: Fracture energy / Energa de fractura
0: Thermodynamic surface energy / Energa superficial termodinmica
eff : Effective surface energy / Energa superficial efectiva

nbt: Energy for crack initiation / Energa para la iniciacin de la grieta
p: Plastic deformation at the crack tip / Deformacin plstica en el fondo de

entalla
wof: Work of fracture / Trabajo de fractura
l: Adimensional compliance / Flexibilidad adimensional
ls: Adimensional compliance of the specimen / Flexibilidad adimensional

de la probeta
lm: Adimensional compliance of the device / Flexibilidad adimensional del

dispositivo
: Poissons ratio / Coeficiente de Posisson
: Notch tip radius / Radio de curvatura del fondo de entalla
s: Stress or strength / Tensin
sf: Fracture strength / Tensin de fractura
s0: Weibull characteristic strength / Tensin caracterstica de Weibull
En el captulo 6 de materiales refractarios:
S: materiales refractarios de slice
AS: materiales refractarios de almina-slice
AZS: materiales refractarios de almina-zirconia-slice
C: hormigones de alta almina

NDICE DE CONTENIDOS
1. INTRODUCTION 3
1.1. Basic fracture-mechanics parameters for brittle materials 4

1.2. Toughened ceramics 9
1.3. Determination of fracture toughness of ceramics 14
1.3.1. Unstable fracture testing 15
1.3.2. Flexure of Single Edge V Notched Beams 18
1.3.3. Stable fracture testing 21
1.4. Magnesium-aluminate spinel materials 24
1.5. Alumina materials 29
1.6. Mullite materials 33
1.7. Refractory materials 38
1.7.1. General properties 38
1.7.2. Toughness testing of refractories 41
1.8. Nanocomposite ceramic materials 44
1.8.1. General properties 44
1.8.2. Toughness testing of Al2O3/SiC nanocomposites 47
2. OBJETIVOS E HIPTESIS 55
3. DESARROLLO DEL DISPOSITIVO EXPERIMENTAL 59
3.1. Unidad de medida de apertura de los labios de la grieta 60

3.2. Sistema y geometra de carga 63
3.2.1. Cermicas tcnicas avanzadas 65
3.2.2. Materiales refractarios 67
3.3. Dispositivo experimental 69
3.4. Anlisis de estabilidad de ensayos de probetas SENB en control por
posicin 69
3.5. Verificacin del dispositivo experimental y procedimiento de ensayo 75
3.6. Conclusiones 79
Anexo I 81
4. TENACIDAD DE FRACTURA DE CERMICAS FRGILES: ALMINA..91
4.1. Condiciones experimentales 92

4.2. Resultados y discusin 92
Anexo II 97
5. EFECTO DE LAS CONDICIONES EXPERIMENTALES EN LA
TENACIDAD DE FRACTURA DE UN MATERIAL DE MULLITA 107

Anexo III 113
6. INFLUENCIA DE LAS CARACTERSTICAS MICROESTRUCTURALES EN

LA TENACIDAD DE FRACTURA DE MATERIALES REFRACTARIOS 125

Anexo IV 129
7. TENACIDAD DE FRACTURA DE NANOCOMPUESTOS DE Al2O3/SiC 149

Anexo V 157
8. CONCLUSIONES GENERALES 161
BIBLIOGRAFA165
Captulo 1
INTRODUCTION
Captulo 1
1. INTRODUCTION
There are many different definitions for the term ceramics or ceramic materials.
The traditional and most accepted one is that of ceramic materials being non-metallic
inorganic solids obtained through thermal treatments1. This definition would have to be
broadened to include the large series of ceramics that acquire their properties once
installed, such as refractory concretes or magnesia-carbon refractories. Conventional
solid-state reactions between oxides and/or carbonates at high temperatures produce
relatively coarse, impure and agglomerated particles. However, the development of soft
chemistry routes is a promising alternative for the synthesis of highly pure and
homogeneous ceramic materials in controlled particle size and morphology at low
reaction temperatures.
According to their applications ceramic materials may be classified in traditional and
advanced ceramics. Traditional ceramics include porcelains (pottery, floor tiles,
sanitaryware,), refractories and building materials (bricks and roof tiles,...).
Advanced ceramics are materials obtained through substantial refining of the initial
raw materials in order to achieve specific functions for each application1. Advanced
ceramics are usually classified in structural and functional materials. The main aim of
structural materials is to stand mechanical stresses at room or at high temperatures.
Functional ceramics are materials with specific functions such as, biomedical, electric,
electronic and magnetic applications amongst others.
In ASTM C1145:20132 advanced ceramic is defined as a highly engineered, high-
performance, predominately non-metallic, inorganic, ceramic material having specific
functional attributes.
According to ISO 20507:20143 and EN ISO 15732:20054 the use of terms as fine
ceramic, advanced ceramic and advanced technical ceramic is interchangeably accepted
3
Introduction
in business, trade, scientific literature and International Standards. In this work, the
terms advanced ceramics and advanced technical ceramics will be used.
Ceramic materials have a series of advantages in comparison with metallic and
plastic materials, such as chemical stability, resistance to corrosion and wear and the
fact that they maintain high resistance to deformation at temperatures at which other
materials develop generalized flow phenomena.
Main drawback is their brittleness and associated lack of reliability. Therefore, the
definition and measurement of the mechanical properties of ceramic materials is a basic
problem in experimental mechanics and in materials science. The data obtained can be
used in specifying the suitability of structural and functional materials for many
applications and for understanding their mechanical behaviour.
1.1. Basic fracture-mechanics parameters for brittle materials
The study of the mechanical behaviour of ceramic materials has attracted a notable
part of the attention of the scientific community since the early ages of the existence of
these materials as cutting tools or structural components in houses. The wide field of
application of ceramic materials and their use in our routine lives as structural and
functional components make necessary a detailed analysis of the causes that produce
their failure.
Reviews about mechanical behaviour of structural ceramics can be found in
references 5-75-7. The type of fracture of a material subjected to stress is determined by
its capacity to deform plastically. Due to the highly directional covalent-ionic bond in
ceramic materials, they are characterized by the absence of significant plastic
deformation up to relatively high temperatures. Thus, the form of fracture of ceramic
materials is fundamentally brittle. There are three fracture modes: mode I or tensile
opening (tension normal to the crack plane), mode II or in-plane shear (shear loading in
4
Captulo 1
crack direction) and mode III or out-of-plane shear (loading). Ceramics usually fail in
mode I (Figure 1.1).
a b c
Figure 1.1: The three fracture modes8:

a) Mode I: Opening mode.
b) Mode II: In-plane shear mode.
c) Mode III: Out-of-plane shear mode.
Brittle behaviour is characterized because the stress level at which fracture starts is
determined by the presence of defects which act as stress concentrators, producing the
catastrophic failure in the material. Thus, the strength of ceramic materials is an
extremely variable property, which depends on the shape and size of defects present in
their interior and on their distribution and orientation relative to the applied load. Thus,
the strength values are much lower than the theoretical fracture stress, or stress
necessary to separate two atomic planes of the material.
The mechanical resistance of ceramics to catastrophic failure may be described by
the fracture toughness which is defined as the materials resistance to the propagation of
defects, determined by its microstructure and independent of the particular flaws present
in it. Contrarily to the strength, the toughness of brittle materials is an intrinsic property.
There are two basic approaches to the concept of fracture toughness: the global
approach based on an energy criterion and the local approach based on the evaluation of
the distribution of stresses around a flaw, or criterion based on the stress intensity
factor, K.
The energy criterion for fracture was formulated by Griffith who solved the problem
of evaluating the fracture resistance of brittle materials by addressing the two-
5
Introduction
dimensional situation of an infinite sheet with a crack, and considered the energy
involved in the creation of new surfaces9. According to Griffiths energy criterion, the
propagation of a crack is governed by the relationship between the elastic deformation
energy stored in the system which contains the crack and the surface energy required to
form new surfaces, G, which is denominated specific fracture energy or energy release
rate. For mode I failure, when the elastic energy stored equals GIC, the critical energy
release rate in mode I, fracture starts. GIC is the energy released by unit of extension of
the crack front and by unit of body thickness, and is equal to twice the thermodynamic
surface energy (0) for brittle materials, (GIC = 20). Thus GIC is an intrinsic property
which characterizes the materials fracture toughness deriving from the energy criterion.
Davidge10 and Orowan11 extended the Griffiths approach indicating that the surface
energy of fracture in real materials is higher than the thermodynamic surface energy
(0), due to deviations from ideal perfectly brittle behaviour. Consequently, the fracture
of the material is determined by a global energy term, , named effective surface energy
(eff) by these authors10,11, which is defined as the energy necessary to initiate the crack
propagation and is determined by the sum of the contribution of the thermodynamic
surface energy (0) and the plastic deformation at the crack tip (p), ( = 0 + p).
Sakai and Bradt12 proposed that, in addition to the thermodynamic surface energy
(0) and the energy associated to plastic yielding in frontal process zone, energy might
be released by other processes such as crack deflection by interface debonding,
microcracking in frontal process zone, dilatant phase transformation and/or
microcracking in wake region, fibre bridging, grain bridging and interlocking,
viscoelastic bridging and other phenomena, such as heat or sound, which are difficult to
quantify.
6
Captulo 1
The value of is conditioned by grain size and fracture mode, since the energy
consumed during the transgranular fracture is different from the energy of fracture
through the grain boundary10,13 , and the length of the actual crack path would depend
on fracture mode. Because of the diversity of contributions affecting , it is impossible
to quantify theoretically its value in polycrystalline materials, consequently it has to be
determined experimentally.
The approach based on the stress intensity factor was first developed by Irwin14 in
1957, who used Westergards previous research15 as a basis of his work. According to
this approach, in a piece of material subjected to a stress, , the stresses and
deformations at the crack front are related by a universal proportionality factor called
stress intensity factor. For ceramic materials which fail in mode I, the stress intensity
factor is determined by equation 1.1:
K I = Ys a (1.1)
where, KI, is the stress intensity factor in mode I, a, is the crack (or any pre-existing
defect characteristic16) length and, Y, is an adimensional factor depending on the
geometry of the loading system and the crack. The crack is propagated when the stress
intensity factor reaches a critical value, KIC, which depends exclusively on the material,
thus, KIC, is a measure of the materials fracture toughness.
A dimensional analysis, conducted by Irwin17, considered that fracture starts when
the stress, , and the stress intensity factor in mode I, KI, reach their critical values, f
(fracture strength) and KIC (critical stress intensity factor in mode I). Therefore, the
fracture strength, f, of a given ceramic specimen will be given by equation 1.2:
K IC
sf = (1.2)
Y a
7
Introduction
The dependence described by equation 1.2 allows establishing a relationship of
similarity between fracture toughness deriving from the energy criterion and that
deriving from the criterion based on the stress intensity factor and, thus, the equivalence
of the two criteria by equation 1.3:
2
K IC
G IC = (1.3)
E'
where E is the generalized Youngs modulus, E=E/(1-n2) for plane strain and E=E
for plane stress (E is the Youngs modulus and n is the Poissons ratio).
The fracture toughness of brittle materials would be characterized by any of these
parameters KIC or GIC (=2g). These two parameters are defined at the start of
propagation of the crack and are intrinsic properties of the brittle materials, which do
not depend on the load system or on the geometry of the cracks.
The critical parameters, f, GIC and KIC provided the basis of the first theories of
fracture. However, it is now well recognised that failure of ceramics can occur even for
stresses well below the strength of the material. Indeed, KIC (GIC, f) only represent a
critical level for catastrophic crack growth. Ceramic materials are susceptible to slow
crack propagation at KI values under KIC; this phenomenon is often referred as
subcritical crack growth (SCG). KI0 defines a KI threshold below which no crack
propagation occurs. SCG is notable for its extreme sensitivity to applied load and
environment, temperature, and other extraneous variables. That means that under
appropriate conditions, cracks keep on growing for some time until they cause fracture
without warning. Since it is not possible to avoid completely imperfections and
microcracks during the production of sintered ceramic materials, cracks might always
grow in a slow manner before catastrophic failure16,18.
8
Captulo 1
1.2. Toughened ceramics
As discussed above, the relatively low and highly variable strengths of ceramics are
the main drawback for their extensive use in engineering applications involving severe
mechanical solicitations.
In the past, much of the effort to improve the mechanical behaviour of ceramics was
placed in producing the highest degree of homogeneity in bulk monophase materials
with very small flaws. However, in the last three decades new strategies directed
towards the toughening of materials have been developed. Toughening is reached by the
development of new microstructures and/or textures in single phase (i.e. silicon nitride
and zirconia) or composite (monolithic or formed by a combination of layers of
different microstructures) materials.
In particular, the toughening of silicon nitride is usually achieved by a highly
anisotropic microstructure where elongated grains tend to be aligned19. Toughening of
zirconia occurs when the particles are in the metastable tetragonal form20. Monolithic
ceramic composites are being fabricated using hierarchical structures through artificial
methods mimicking natural bio-structures, which significantly overcome the failure
behaviour of the individual components. Besides, materials formed by a combination of
layers of different microstructures and materials fabricated by directional solidification
of compositions close to eutectic ones offer improved behaviour in comparison with the
behaviour of monolithic materials. Another relatively new field of investigation,
initially proposed by Niihara21, is that of ceramic nanocomposites with a submicron
and/or nanometric scale dispersed second phase which show an increase of the strength
and wear performance as compared to that exhibited by the matrix materials21,22. The
nanocomposite ceramics will be described further in section 7.
9
Introduction
Toughening mechanisms reduce the characteristic brittle behaviour of ceramic
materials by raising the inelastic deformation processes during the fracture. Main
toughening mechanisms have been reviewed by different authors7,23-25. A most useful
classification of the toughening mechanisms is the one done attending to their range of
action (Figure 1.2).
Crack bowing
Short-range
Crack deflection
Toughening
mechanisms Phase transformation
Crack-shielding
Microcracking
Long-range
Crack-bridging
Crack-interaction
Debonding and pull-out
Figure 1.2. Scheme of the main toughening mechanisms attending to their range of
action.
Short-range mechanisms are those which radii of action affect a zone of the
microstructure of size smaller or of the same order as the materials microstructural
characteristics. Thus, the interaction of the crack with the microstructure is practically
limited to the crack tip. Such mechanisms cannot be identified if the crack lengths are
longer than the size of the microstructural characteristics and can only produce limited
flaw tolerance. Crack bowing and crack deflection are short-range toughening
mechanisms.
Crack bowing appears when the front of the crack is stopped at several points due to
the presence of inclusions of a second phase or heterogeneities of the matrix itself. The
crack front curves because the part of the crack front which is not held back continues to
advance and, it is necessary to impose higher stress to propagate it. The increase in
10
Captulo 1
toughness produced while the crack is being bowed depends on the volumetric fraction
of obstacles present, their shape and their toughness24.
Crack deflection consists in changes in the direction of crack propagation. Such
changes are determined by the presence of easy cleavage planes and/or residual thermal
stresses in single phase ceramics and by the characteristics of the phases and the
boundaries between different phases in ceramic composites. Crack deflection produces
rough fracture surfaces. The effectiveness of crack deflection as a toughening
mechanism depends on the shape, the density and the distribution of the deflecting
objects but not on their size, and is a maximum for particles with a disk-like shape or
elongated bars24.
Long-range mechanisms are processes which occur around the crack tip and
surrounding the crack wake and can produce extensive flaw tolerance. These
mechanisms can be subdivided into crack-shielding and crack-interaction mechanisms.
There are two kinds of processes that can be responsible for crack-shielding, they are
phase transformation and microcracking23,24. Crack bridging and debonding and pull-
out are mechanisms of crack-interaction.
In the crack-shielding mechanisms irreversible processes lead to microstructural
changes that surround the crack wake as the crack propagates. These processes are
activated in the process zone which is the area surrounding the crack tip. They give rise
to dilatation which is transmitted in the form of compressive forces on the crack lips
while the crack propagates7.
In ceramics, the most representative example of phase transformation with large
increases in toughness is the expansive transformation of zirconia from tetragonal to
monoclinic phase induced by the stress field of a crack and accompanied by an increase
in volume of 3-5%23,24,26-28.
11
Introduction
The opening of microcracks7 under the action of the applied stress also leads to a
volume increase around the crack. Moreover, when microcracking occurs the
microcracked zone presents a lower elastic modulus than the rest of the material which
contributes to the effect of shielding of the main crack because of the reduction of the
elastic energy at the crack tip23. Anisotropy in thermal expansion in monophase
materials, the presence of secondary phases with different coefficient of thermal
expansion from that of the matrix and/or phase transformation may be the origin of the
residual stresses that define the appearance of microcracks in the main cracks stress
field. The increase in toughness produced by crack-shielding mechanisms depends on
the density of elements which produce irreversible deformation, and the size and shape
of the process zone7.
In the other large group of long-range toughening mechanisms, crack interaction
mechanisms, the increase in resistance to the propagation of cracks occurs as a result of
the union of the fracture surfaces by means of microstructural objects, called
ligaments25,26. Thus, additional energy is required to separate the fracture surfaces26,29.
Crack bridging7 occurs when the element acting as a ligament deforms elastically
during the opening of the main crack until it breaks. The increase in toughness is
determined by fracture stress, the elastic modulus and the size of the ligament.
Toughening by crack-bridging is conditioned by the state of residual stresses of the
particle and the matrix, since it determines whether the crack surrounds or draws nearer
to the toughening ligament.
If, in addition to the effect of the union of the crack surfaces, the ligaments are
debonded and pulled-out, the increase in toughness can be much greater. This
contribution increases with the length of the ligaments and can occur at a relatively long
distance from the main crack front30.
12
Captulo 1
In toughened ceramics whose behaviour cannot be regarded as complete brittle, the
parameters deriving from linear elastic mechanics to characterize fracture toughness,
KIC or GIC are not intrinsic properties. Thus, in this type of materials, specific treatment
of fracture parameters is required31-33. One proposal is to extend the principles of linear
elastic fracture to situations where the inelastic deformation occurs prior to fracture, so
that fracture toughness can be determined. Parameters like resistance to crack growth,
R-curve, and J integral arise from this approach.
The R-curve has been extensively used to characterize the fracture toughness of
ceramic materials in which long-range toughening mechanisms are operative. To build
the R-curve, the values of the parameters of linear elastic fracture mechanics, K IC or GIC
are determined in standard tests conducted with long cracks and are represented as a
function of crack size. The R-curve will show a rising part as long as the energy-
consuming processes behind the crack tip (process wake zone) are taking place. Then,
the R-curve will evidence a decline when the process zone ahead the crack tip arrives at
the specimen edge.
Materials where long-range toughening mechanisms act show rising R-curves, as a
result of the activation of the different mechanisms as the crack grows, until toughness
reaches a stationary state (K or G); from this point no new contributions are made to
increase in toughness. The maximum value of toughness reached does not only depend
on the materials, but also on the loading system, the size of the crack and its history
prior to propagation.
When only short-range mechanisms act, toughness is determined by the details of
the microstructure ahead the crack tip. As the crack grows, it intercepts different grains
and grain boundaries, with different orientations from those on the original crack plane.
Thus, toughness increases with the size of the crack until it reaches the value
13
Introduction
corresponding to the polycrystalline material. From this moment, the crack can be
treated macroscopically as if it were propagating in a homogeneous and isotropous
body. Thus, the R-curve rise is extremely step because toughening occurs at the
microstructural level, and it is not detected in standard tests for determining toughness
using long cracks23,26.
Unlike fracture toughness, the specific fracture energy, GF, is defined as the average
value of external work consumed to produce a crack unit during quasi-static fracture.
Experimentally, it is determined on the basis of area under the load-displacement of the
load point curve, obtained during stable tests, where all the work done produces new
fracture surfaces. The work done is divided by the size of the fracture surface, which
means that specific fracture energy is an average value for the entire fracture process.
The advantage of this energy parameter is that it does not require any assumptions about
the constitutive equation of the body with the crack to discuss its propagation34. Thus, it
can be used to describe behaviours which separate from linearity and it is an additive
parameter which makes it possible to quantify the different contributions to energy
dissipation during fracture.
In refractory literature GF/2 is usually called work of fracture, wof. There has been a
lot of research on the definition of wof and the establishing of experimental methods for
its determination, that will be discussed further in the corresponding section (section 6).
1.3. Determination of fracture toughness of ceramics
The high variability in flexural strength values for ceramic materials highlights the
necessity to determine intrinsic fracture parameters, quantifying toughness; such as the
critical stress intensity factor in mode I, KIC, or the fracture energy, g. The development
of techniques to achieve accurate toughness values is critical to allow the use of ceramic
materials in structural and functional applications.
14
Captulo 1
1.3.1. Unstable fracture testing
Most toughness characterisation of ceramics is done in terms of the critical stress
intensity factor in mode I, KIC, using Linear Elastic Fracture Mechanics (LEFM) (Eq.
1.2). Basic assumptions for LEFM are that materials are homogeneous and isotropic
with no discontinuities and analytic equations have been developed for cracks with zero
tip radius.
In conventional KIC testing a specimen with a notch of known geometry and size is
subjected to increasing deformation until catastrophic failure occurs. Consequently K IC
can be measured if the load amplitude, the geometry factor and the length of the crack
associated with the notch can all be determined at the moment of instability.
Additionally, calculation of KIC from the size of cracks emanating from indentations has
extensively used to characterise ceramics (Indentation toughness, IT). This technique is
attractive due to the small amount of volume of material needed; however, it presents
increasing uncertainties as the microstructure of the materials separate from the
continuous. A review on this technique raising the potential problems associated was
published by Quinn and Bradt in 200735. Indentation fracture (IF), is a technique in
which the length of cracks emanating from the corners of a Vickers indentation is
measured, according to EN 14425-136.
Fracture toughness tests of ceramics are usually performed in universal testing
machines by subjecting the specimens located between the loading supports to
increasing deformation by means of the displacement of the loading frame. This
displacement can be controlled directly by imposing a constant rate to the movement of
the frame and, thus, to the loading point (displacement control). Additionally, the
movement of the frame can be indirectly controlled using increasing rates of different
parameters such as the load (load control), the deflection of the specimen directly
15
Introduction
measured at its central point in bending specimens (deflection control) and any
parameter associated to the growth of the crack, such as the crack length or the Crack
Mouth Opening Displacement (CMOD control). The simplest and therefore most
widely used testing conditions involve the control by increasing rates of displacement or
load.
There is not a unique method for K IC determination and standardisation is relatively
recent and practically limited to technical ceramics. All standards deal with beams
tested in flexure. The series EN 14425 1-5:2004 Advanced technical ceramics - Test
methods for determination of fracture toughness of monolithic ceramics contains five
parts: Part 136: Guide to test method selection. Part 237: Single-edge pre-cracked beam
(SEPB) method, equivalent to ISO 15732:200338. Part 339: Chevron notched beam
(CNB) method, EN 14425-3:2010. Part 440: Surface crack in flexure (SCF) method,
equivalent to ISO 18756:200341. Part 542: Single-edge vee-notch beam (SEVNB)
method, equivalent to ISO 23146:201243). When reviewing papers on mechanical
characterization of technical ceramics, it is observed that results for nominally the same
material can differ up to 300% depending on the laboratory44. Such a large dispersion is
partially due to the different microstructures that might present ceramic materials as a
function of processing and to the different testing conditions. However, even specimens
of the same material tested using the same geometry and strain rate, difficulties
associated to specimen machining and crack introduction might lead to variability.
Due to the difficulties involved in machining ceramic parts with special shapes as
well as to the tendency of brittle materials to fail under shear stresses originated by
deficient clamping and/or alignment, the most widely geometry used for KIC
determination of advanced technical ceramics and refractories is that of parallelepiped
bars subjected to bending in three or four points. In addition to the easiness of specimen
16
Captulo 1
machining, bending of bars require simple testing setups7. Moreover, the analysis of
such geometries is simple in comparison with other techniques as, e.g., the double
torsion configuration45.
The generation of suitable cracks is a fundamental requirement for the validity of the
tests. In most cases, the cracks may be starting from indentations -Vickers or Knoop- or
notches - straight or Chevron (triangular)-. The advantage of cracks introduced by
indents is that they have small openings, in agreement with the requirements derived
from LEFM. However, such small openings can be a disadvantage for the precise
measurement of the crack size. An alternative way is to introduce a pop-in crack by
controlled propagation of an existent notch or indentation crack (Single Edge Pre-
cracked Beam, SEPB). This technique is adequate to introduce real cracks in the
specimens; however, it presents experimental problems for brittle materials because of
their proneness to uncontrolled crack growth46. Other technique is the Indentation
Strength in Bending (ISB) which consists in precrack the tensile surface with a Vickers
indenter and then test the beams in 3-point, 4-point or biaxial bending tests. The
residual stress fields developed by indentation are not removed.
Even tough a round robin performed on technical ceramics47 demonstrated the
adequacy of indentation for crack initiation, which is nowadays standardised (SCF, EN
14425-4:2004)40, the analysis of indentation cracks implies some uncertainties. The
shape of cracks starting from indentations might not be well known, resulting in an
uncertain geometry factor. Furthermore, as a consequence of the plastic deformation
zone underneath the indent, undetermined internal stresses which would add to the
applied ones might occur. Thus, sf, Y, and, a, in equation 1.2 cannot be exactly
determined. Thermal treatments to eliminate the residual stresses and grinding the
specimen surface to eliminate the plastically deformed zone are the two methods
17
Introduction
proposed to partially overcome the problems associated with indentation induced
cracking. Annealing or heat treating is not permitted by the standard EN 14425-4:200440
due to the risk of crack tip blunting or crack healing.
The use of beams with chevron notches has also been standardised for technical
ceramics (CNB, EN 14425-3:2010)39. However, there are some practically unsolved
problems regarding this technique. The geometry of cracks produced by loading a
chevron notch is not exactly known and the crack length can not precisely be measured.
Moreover, the introduction of chevron notches involves great difficulty in the control of
the size of the crack generated and to ensure that the beams, made of brittle materials,
do not break when the crack is being introduced48-49.
Parallelepiped beams with straight through notches at their centre are usually called
Single Edge Notched Beams (SENB). When the notch tip has been sharpened,
specimens are called SEVNB, to account for the V shape of the notch. SENB and
SEVNB are the specimen-crack geometries most commonly used to determine KIC in
flexure.
1.3.2. Flexure of Single Edge V Notched Beams
The straight through notches are the simplest to introduce and their geometry is well
defined45.
In general, they are produced by means of cutting tools such as thin grinding wheels
or diamond discs, threads and cutters. The stress concentration at the notch tip might be
weaker than that of a crack for finite notch radius, which might lead to a systematic
overestimation of the fracture toughness, since additional energy is spent for the
formation of a sharp crack from the blunt notch50.
The Technical Committee Ceramics (TC 6) of the European Structural Integrity
Society (ESIS) organized a round robin (1993-1994) on the determination of fracture
18
Captulo 1
toughness of technical ceramic materials at room temperature, results which were
summarized in Primas and Gstrein work51. Five technical ceramics -Al2O3, hot pressed
Si3N4 (HPSN), sintered SiC (SSiC), Mg-partially stabilized ZrO2 (MgO-PSZ) and
Y2O3-tetragonal ZrO2 polycrystal (YTZP)- were tested with five testing methods -
chevron notched beam (CNB) in four point bending, direct measurement of the cracks
emanating from a Vickers indentation (IF), indentation strength in four-point bending
(ISB), single edge precracked beam (SEPB) in four-point bending, and single edge
notched beam (SENB) in four-point bending- by eighteen laboratories. Main conclusion
was that the measured fracture toughness value depends on the testing method and
testing conditions. Regarding the testing methods, the SENB method produced the most
reliable results for notch tip radius, , below a critical one which depended on the tested
material. Variability for SENB KIC values ranged from 6% for Al2O3 (KIC = 3.17 0.19
MPa m1/2) to 16% for HPSN (KIC = 7.48 1.22 MPa m1/2). In the conclusions TC6
raised the problem associated to the experimental determination of the critical for each
material. At the time, notches were produced by sawing and their tips were relatively
large. In this round robin, notches were introduced with diamond blades of width 100
m with grains of 20-30 m and the requirement to consider valid tests (notch width of
75 25m) could not be fulfilled by a number of participants. A critical around 100
m was found for the alumina material which had coarse microstructure (average grain
size of 16 m).
The dependence of KIC on notch tip radius for notches introduced with diamond
blades has been reported for different materials50,52-55. In particular, it is especially
significant for phase transformation toughened ceramics54 (section 1.2) that develop a
crack shielding zone due to notch machining.
19
Introduction
In order to overcome problems due to finite notch tip radii, Nishida et al.55
introduced the technique of sharpen the notch tip by means of a razor blade sprinkled
with diamond paste. A conventional notch is introduced with a blade in the first step
and then, a sharp V-shaped notch is produced by using a razor blade and diamond paste.
These authors claimed that notch tip radii as small as 2 m over the entire width of the
specimen could be produced applying this technique. Following the analysis by Fett56
for V notches to be equivalent to real cracks, they should have tip radius smaller than
3 times the grain size or the largest microstructural characteristic of the material52.
Ultra-short pulsed femtolaser ablation is being lately used also to obtain shallow sharp
notches57.
A round robin on the use of SEVNB in flexure (4pb) was done by the TC6 of
ESIS58. Materials used were coarse (d50 > 10 m, dmax 17 m ) and fine (d50 2m,
dmax 5m) grained aluminas, gas pressure sintered silicon nitride (GPSSN), sintered
silicon carbide (SSiC), and yttria-stabilised tetragonal zirconia (YTZP). Machining of
notches with tip radius between 20 and 30 m was found to be relatively easy and
reproducible while thinner notches presented extremely different quality depending on
the laboratory. No significant effect for tip notches less than 30 m was found for any
material when tests from all laboratories were considered. However, much lower critical
tip radii were found when a single laboratory analysed the problem in a systematic way.
The toughness determination by testing of SEVNB in flexure was proved to be adequate
and robust for materials with grain size or major microstructural feature size over 1m
when using notch tip radius of 10 m. On the basis of this work, the international
standard for SEVNB tested in 4 point bending was established43 (EN 14425-5:200442).
According to this standard, under well-controlled conditions, notch-tip radii in the range
of 1 mm to 20 mm can be prepared depending on the grain size of the test material. For
20
Captulo 1
many materials, this is a close approximation to a sharp crack, and the method has been
found to give fracture toughness values very close to those of other methods such as the
Single Edge Precracked Beam (SEPB) method37 (EN 14425-2:2004) provided that a
sharp crack forms at the root of the notch either during its preparation or during the
subsequent fracture sequence.
1.3.3. Stable fracture testing
Most toughness testing of ceramics, particularly of advanced ceramics, is done
under conditions of unstable fracture. However, when the fracture toughness values are
determined from test configurations that do not allow stable crack growth the calculated
toughness value might be over-evaluated (e.g. for a dense a-SiC, KIC ~ 4 and 3 MPa
m1/2 for unstable and stable tests, respectively)59. Therefore, stable crack growth is
necessary to get reliable and accurate fracture toughness data. Moreover, stable fracture
tests supply much more information about the fracture process than the unstable fracture
ones because, in addition to the conventional fracture toughness for crack initiation,
stable fracture allows the determination of fracture energy.
Stable fracture tests for brittle materials, as most ceramics, are difficult to
accomplish, therefore, they are not usually performed. Since the initial works of
Nakayama et al.60,61 and Tattersall and Tappin62 different authors have developed
innovative specimen designs and tests geometries46,63-67 to attain stable fracture of
materials using displacement controlled loading. Two main problems arise in this kind
of tests. On the one hand, it is difficult to produce notches with well-defined special
geometries in ceramic specimens. On the other hand, unstable crack growth can result
as the amount of energy necessary to create the fracture surfaces is usually small
compared to the elastic energy stored in the test rig unless extremely stiff machines are
used. Unstable crack growth is avoided or limited by careful adaption of the testing
21
Introduction
machine and the specimen geometry to the material to be tested. Such approach has
made it possible to reach stable fracture for materials as brittle as glass34,60 and fine
grained MgO62 using stiff machines.
From the broad spectrum of available fracture tests, bending of parallelepiped
specimens with straight through notches (SENB) is a relatively simple way of testing
and displacement or load controlled bending tests have been widely used for fracture
toughness testing of ceramics. Therefore, different attempts to perform stable tests using
loaded beams have been done.
In principle, four point bending would lead to more accurate toughness data because
it is not necessary to align the crack tip with respect to the central loading roller, as
occurs in three point bending68. However, Baratta and Dunlay69 have shown that
specimens loaded in three point bending are more likely to fracture in a stable manner
than those loaded in four points. Therefore, different attempts to perform stable tests
using three point bending loaded beams using displacement control have been
done34,70-72.
The origin of difficulties related to the attainment of stable fracture when using
SENB specimens in three point bending and displacement as control variable is clear
when the (simplified) graph Load versus Displacement of the loading point of Figure
1.3 is analysed, as done by Sigl73. In terms of stability, the use of the actual deflection of
the specimen as control variable is equivalent to the use of displacement of the load
frame. The curved line is the general stable fracture locus of SENB specimens of a
brittle material tested in bending. This curve is called the Griffith locus and represents
the fracture taking place with constant energy release rate, G, equalling Gc. As the
straight line that represents the loading of the specimen hits the curve, the condition for
crack growth (G=Gc) is satisfied. In order to get stable fracture, G has to be maintained
22
Captulo 1
at its critical value and, thus, decreasing values of the load point displacement (regime I)
followed by increasing values of this parameter (regime II) would be demanded. The
loaddisplacement relationship of regime I is usually called snap back74. Therefore, it
will not be possible to get stable fracture using constant displacement rates in regime I.
The relative weight of the regions corresponding to regimes I and II depends on the
material properties, the specimen and span sizes, the notch depth and the stiffness of the
testing device. For the same material and testing geometry, stiff machines and deep
notches increase the region of regime II. In this case, stable fracture can be reached
controlling by constant displacement rate because increases in displacement after the
maximum load still allow to follow the stable condition for crack growth G=Gc. From
Figure 1.3 it is clear that the control by constant rates of increasing load can never lead
to stable tests because load always decreases after cracking starts.
Regime I
Load
Regime II
Displacement
Figure 1.3. General plot Load Displacement of the loading point for stable fracture
of SENB specimens of brittle materials tested in three point bending. The straight line
represents the loading of the specimen and the curved part corresponds to the Griffith
locus.
Stable fracture for SENB specimens tested in three points bending using
displacement control is relatively easy to attain for materials with R-curve fracture, for
23
Introduction
which the crack resistance (i.e. Gc) increases as the crack propagates. For instance,
stable fracture has been reported for materials with coarse microstructures such as
silicoaluminate and high alumina refractories61 and graphite34,71, for dense alumina with
relatively large grain size (d50 ~ 5.5 mm)72 and for fine grained aluminaaluminium
titanate composites (alumina: d50 ~ 3.23.9 mm, aluminium titanate: d50 ~ 2.2 mm)72. On
the contrary, for brittle materials, very deep cracks and extremely stiff machines would
be needed for stable fracture and thus, it is not possible in practice. In this sense, load
displacement curves showing unstable fracture or sudden load decreases (pop-in)
prior to further stable propagation (i.e. semi-stable fracture) have been reported for fine
grained alumina (d50 ~ 3.5 mm)72 and silicon nitride (d50 ~ 3 mm)70.
Contrarily to the above-discussed parameters, parameters related to crack growth
increase during the whole fracture test, thus, they are suitable to reach stable fracture. In
particular, the Crack Mouth Opening Displacement (CMOD) increases through the
whole fracture test, the loading of the specimen as well as during the crack growth.
Thus, CMOD has been proposed and used as control parameter for stable fracture
testing of high-strength concrete74 and tetragonal zirconia stabilised with 3 mol% of
Y2O375-77 specimens under conditions that would have led to unstable fracture for
displacement controlled tests; both materials present R-curve behaviour. This parameter
has never been used to test brittle specimens.
1.4. Magnesium-aluminate spinel materials
Magnesium-aluminate spinel (MgAl2O4) combines a set of attractive optical,
thermal, electrical and mechanical properties such as transparency, moderate thermal
conductivity (~16W/m.K at 298.15K)78, moderate thermal expansion coefficient (~ 8 10-

6
K-1)78-81, high thermal stability82,83, low electrical losses75,76, high resistance to
chemical attack84,85, relatively high hardness (15-17 GPa)78,86-89, moderate Youngs
24
Captulo 1
modulus (260-280; 295 GPa)78,86-89 and good thermal shock resistance90. Spinel presents
the characteristic low fracture toughness (1.2-3.0 MPa m1/2)78,82-84,86-102 and strength
(185-300 MPa)78,84-89,91,93 of brittle oxide materials.
The most common applications for spinel are high-refractory84,85, electro-
insulating84,85,102, infrared windows/domes78,82,93-95, transparent ceramic
armour82,95,98,103, laser hosts104, lamp envelopes105 and structural material with
application in metallurgical, electrotechnical and radiotechnical industries84,85.
Mechanical properties and transparency are known to depend on powder composition
and subsequent sintering/hot isostatic pressing/spark plasma conditions.
Spinel is transparent in the wavelength range 0.2-6 m and because of the cubic
crystal structure it does not suffer from birefringent light scattering at grain
boundaries78,90,93-95,97,98,106. To sinter highly transparent optical ceramics with high
strength properties is necessary to provide the minimum (nanoscaled) size of the
intergranular boundaries of the sintered ceramics with content of pores and impurities
no more than 10-3 volume percent82,96.
The mechanical behaviour of stoichiometric magnesium-aluminate spinels has been
studied on single crystals99-102,107,108 and polycrystalline samples82-85,90,93-98,100,103.
The mechanical behaviour at room temperature of nano82,95,96, fine83,84,90,93-98,103 and
coarse-grained83,85,91,92,94,95 spinels has been reported. Materials have been prepared by
different sintering routes: conventional sintering84,85,90,95, hot isostatic pressing83,91-

96,98,100,103
and spark-plasma sintering82,97. Sintering additives such as LiF90-92 (generally
1 wt.%) and Y2O390,95 have been used during the sintering processes. The main
microstructural characteristics and the reported mechanical parameters at room
temperature are summarised in table 1.1.
25
Introduction
Table 1.1. Microstructural characteristics and mechanical parameters at room

temperature for polycrystalline spinels.
T.D. d50 (S.D.) KIC (S.D.) s (S.D.) E (S.D.)
Reference
% (mm) (MPa.m1/2) (MPa) (GPa)
E.d. Values E. d. Values E.d. Values
IT+ 1.9 (0.1) s0 169 (3) ROR 260 (5)
IT 1.3 (0.05) m 5 (2) IT 210 (10)
5 ROR
Tokariev et ISB 1.9 (0.2) IET 270
al.93,94, 99 SENB 4-pb 1.8 (0.2)
2013,2012 Critical defect 1.6 (0.2)
+
IT 1.2 (0.05) s0 77 (1) ROR 242 (13)
ROR
60 IT 0.8 (0.05) m 10 (2) IT 225 (10)
SENB 1.0 (0.4) IET 270 (5)
Khasanov et
100 IT+ 2.4 IT 212
al.82, 2013
s0 470
0.345 IT 1.4
Borrero- ROR
Lpez et m 6.2
al.95, 2012 100
s0 364
Mroz et al.96, 2.1 IT 1.4 ROR
2012 m 5.3
s0 160
15 IT 1.4 ROR
m 9.6
Boniecki et
2.1 (1.5) SENB 3-pb 1.9 (0.1) 4-pb sf 111(4)
al.90, 2012
Morita et 3-pb sf 500 (60)

>99.8 0.4 IT 1.5 (0.1)
al.97, 2009 4-pb sf 380 (107)
98
Krell et al. ,
99 1.7 IT 2.0 (0.2) 4-pb sf 225 (25) 275
2009
Baudn et 129 (20) 4-pb 206 (6)

98 1.5 (0.8) SENB 4-pb 3.0 (0.1) 4-pb sf
al.84, 1995 194 (20) IET 258 (2)
White et
>99 75 CNB 2.2 (0.2)
al.92, 1992
SENB 1.46 (0.44)
Ghosh et
99 35 CNB 1.79 (0.11)
al.91, 1991
SCF 1.81 (0.17)
5 1.94 (0.10)
Stewart et 99 12 1.98 (0.14)
SCF IET 258
al.83, 1980 25 1.83 (0.14)
38 1.97 (0.14)
T.D.: Theoretical density (%); d50: Average grain size (m); KIC: Critical stress intensity factor
in mode I (MPa m1/2); s: Strength (MPa); sf: Average strength (MPa); s0: Weibull characteristic
strength (MPa); m: Weibull parameter; E: Youngs modulus (GPa); IT: Indentation Toughness;
ISB: Indentation Strength in Bending; SENB: Single Edge Notched Beam; CNB: Chevron
Notched Beam; SCF: Surface Crack in Flexure; ROR: Ring on Ring; 3 or 4-pb: three or four
point bending; IET: Impulse Excitation Technique; E.d.: Experimental details; (S.D.): Standard
Deviation; +: Niihara equation, : Atkins equation).
26
Captulo 1
Values for dynamic Youngs modulus are similar (260-270 GPa) for sintered84 and
hot pressed materials83,93,94. Static Youngs modulus values84 (200 GPa) are lower than
the dynamic ones and close to those obtained in indentation tests for hot pressed93,94 and
spark plasma sintered (SPS)82 spinels. Ring-on-ring tests93,94 led to much higher values
of Youngs modulus (240-260 GPa) than other static methods for fine ( 5m) and
coarse ( 60m) grained materials93,94.
All Youngs modulus values obtained for polycrystalline spinels are lower than that
one reported by Mitchell108 ( 300 GPa) for MgAl2O4 single crystals using the Voigt
average of the elastic constants. Different authors have demonstrated the anisotropy of
Youngs modulus in spinel single crystals; Values for different cleavage planes are:
165, 198 GPa for <100>109,110 and 282, 308 GPa for <110>107,108.
Reported fracture strength values for spinels with micrometric grain sizes ( 1.7-60
m) determined in 4 points bending84,90,98 or using ring-on-ring bending tests93,94,96,103
are in the range 100-360 MPa. Grain refinement to the nano-scale led to higher
strengths (400-500 MPa) for SPS97 and hot-isostatically pressed95,96 materials. MgAl2O4
single crystals also present anisotropy in strength. Rice et al.102,107 reported values of
269 7 MPa and 200 10 MPa parallel with and perpendicular to the tensile surface
<100> and 210 20 MPa and 165 30 MPa parallel with and perpendicular to the
tensile surface <110>.
Values of fracture toughness have been reported for polycrystalline82-84,90-98 and for
single crystals99-102. Different techniques have been used -SENB in flexure84,90,91,94,
biaxial tests102, Chevron-notched beams in bending91,92, Microcantilever100, Double
cantilever beam101, being the most used the indentation test82,83,93-99,102.
Vickers indentation toughness values in the range 1.8-2.2 MPa m1/2 have been
reported for hot-pressed and hot-isostatically pressed spinels with a wide range of grain
27
Introduction
sizes93,94,98 ( 2-60 m). Spinels with finer microstructures ( 102 nm-1 m) present
slightly lower values ( 1.4-1.5 MPa m1/2) of Vickers toughness95-97. Higher fracture
toughness value, measured by Berkovich nanoindentation, is reported by Khasanov et
al.82 (2.4 MPa m1/2) for a nanostructured spark-plasma sintered spinel.
Values obtained using beams with Knoop or Chevron flaws in bending83,91,92 ( 1.8-
2-2 MPa m1/2) are also independent from grain size for hot pressed spinels with grain
sizes 5-75 m.
There are few papers reporting SENB toughness values for spinel materials84,90,91,94.
Average values between 1.8 and 3 MPa m1/2 have been reported for fine grained ( 1.5-
5 m) materials fabricated by conventional sintering84,90 and hot-pressing94. Contrarily
to Vickers values, a dependence on grain size has been reported, being lower ( 1-1.5
MPa m1/2) the values for coarse grained spinels ( 35-60 m)91,94.
The large difference ( 50%) found for fine grained spinels between the initial work
by Baudn et al.84 and the more recent ones90,94 could be partially attributed to the larger
notch tip radius associated to the notching techniques available at the time.
As for the above described mechanical parameters, toughness of MgAl2O4 single
crystals is anisotropic. Rice et al.102 measured the fracture toughness using different
techniques as double cantilever beams, biaxial bending tests, indentation tests, obtaining
values of ~ 1.0 MPa m1/2, for <100>. Wu et al.101 reported fracture toughness values and
fracture energies of 1.21 MPa m1/2 and 4.4 J/m2, and 1.28 MPa m1/2 and 2.9 J/m2 for
<100> and <110> planes, respectively, using double cantilever beam tests.
Stewart et al.99 reported fracture toughness values 1.18 0.05, 1.54 0.08, 1.90
0.06 MPa m1/2 and fracture energies 3.57, 4.07, 4.85 J/m2 for the different cleavage
planes of single crystals <100>, <110> and <111>, respectively. The lowest fracture
28
Captulo 1
toughness corresponds to <100>, so it is the easier plane for fracture and lower energy
is required to fracture.
1.5. Alumina materials
Alumina (Al2O3) is the ceramic material most extensively studied, both from
theoretical and practical stand points. There exist several reviews about alumina
materials -properties, production, and uses (Briggs, 2007; Doremus, 2008; Drre &
Hbner, 1984; Gitzen, 1970; Hart, 1990; Kingery, 1984; Mc Colm, 1990; Riley,
2009)111-118. An updated review (2014) can be found in chapter 2 of the encyclopaedia
Comprehensive Hard Materials119.
Alumina ceramics are widely applied because of the high hardness (for sapphire H
= 9 in the Mohs scale and Hv is up to 30 GPa depending on the orientation), high
melting point (2050C), good wear resistance, good corrosion resistance and
outstanding mechanical properties at high-temperature. In fact, from all ceramics,
alumina presents the highest thermal stability together with high hardness sustained up
to temperatures over 1200C, therefore alumina is the natural ceramic for wear119.
The excellent wear behaviour of alumina ceramics together with their
biocompatibility and excellent corrosion resistance have carried out to widespread the
use of the alumina to the field of biomaterials as femoral heads in total hip replacements
as an alternative to acetabular cup of polyethylene or to metal-metal bearing devices16.
Because of thermal expansion anisotropy (a25-1000C ~ 9.2 10-6 K-1 and 8.4 10-6 K-1
parallel and perpendicular to c-axis, respectively120) stresses can appear when cooling
alumina ceramics from the sintering temperature121. The stress level depends on the
particular relative orientation of the grain boundaries. For grain sizes above a critical
one, these stresses can lead to fracture. However, for the small grain sized (<30 m)
structural aluminas, the developed stresses remain as residual stresses in the sintered
29
Introduction
materials and are partially responsible for the dependence of properties such as
hardness, fracture toughness, and strength on grain size13,72,122-126.
The mechanical behaviour of alumina materials has constantly improved since the
years 1960s. Efforts have been directed towards the increase of the purity and the
homogenisation and decrease of the grain size of the raw materials, the addition of
dopants to control grain size as well as to the use of advanced sintering techniques. The
development of alumina materials for metal machining and sealing as well as for
prosthetic hip replacement has been the motor of such improvements.
Reported Youngs modulus values for pure (>99%), fine grained (2-6 m) and
dense (98% of theoretical) alumina ceramics range form ~ 380-390 GPa for
conventionally sintered materials72 up to ~ 400-450 GPa for the optimised hot-pressed
and hot-isostatically pressed aluminas126,127.
Strength values determined in bending vary from 350-460 MPa for conventionally
sintered materials with grain sizes ~ 3-6 m to 500-600 MPa for hot-pressed and hot-
isostatically pressed aluminas72,127-129.
Since the initial systematic studies done by the group of Claussen123,130 toughness
values of alumina materials have been recognised as highly dependent on grain size and
test technique. This fact is partially due to toughening mechanisms. Toughening
mechanisms reported to alumina materials are mainly due to the effect of residual
stresses and therefore, highly dependent on grain size. Crack deflection, microcracking,
crack bridging and crack branching, for extremely large grain sizes have been observed
in alumina materials. The action is revealed by the greater roughness of the fracture
surfaces of the larger grain size aluminas, as it is shown in Freiman et al. work131.
Table 1.2, summarises characteristic room temperature toughness values reported
for alumina materials determined in flexure tests performed using the displacement of
30
Captulo 1
the load frame as control parameter. When reported, values obtained using other
techniques in the same study are also included for comparison.
For fine grained materials (d50~2-8 m) KIC values from ~2 to ~6 MPa m1/2 and
specific fracture energy values from ~10 to ~20 J/m2 are found. All KIC values are
higher than those reported for alumina monocrystals by Michalske et al.132 (KIC = 2.0
MPa m1/2).
Bueno et al.72 reported values recorded during semi-stable or stable fracture of
SEVNB tested in flexure using the displacement of the frame load as control parameter,
KIC and values for the two aluminas studied by these authors are very similar to those
determined by Sbaizero et al.127 for aluminas with larger grain sizes using stable fracture
tests.
However, KIC values reported by Bueno et al.72 and Sbaizero et al.127 are lower
(~35%) than those reported for aluminas with similar grain sizes determined from
unstable fracture tests of SENB13,52,123, in agreement to Bar-On et al.70 who, as
discussed in section 1.3.3, concluded that unstable crack extension would result in
apparent increases of fracture toughness values compared to those determined during
semi-stable or stable tests.
Additionally, the larger notch tip radius of SENB as compared to the SEVNB will
also lead to higher results as discussed in previous section 1.3.2.
KIC values reported by Bueno et al.72 and Sbaizero et al.127 are also lower than those
obtained by double torsion (DT)16,128,129,133. It is well known that double torsion tests
always give the highest values as compared to other methods. The specific geometry for
the crack that has to be assumed for the calculations might not be accurate, leading to
erroneous values.
31
Introduction
Table 1.2. Characteristic room temperature toughness values reported for alumina
materials.
T.D. (S.D) d50 (S.D.) KIC (S.D.) GIC (S.D.) g (S.D.)

Reference Method
% (mm) (MPa.m1/2) (J/m2) (J/m2)
Palmero et
>98 <7 DT 5.5 (0.2)
al.129, 2014
Bueno et 98.1 (0.3) 3.5 (0.3) SEVNB 2.9 ; 2.8 20.4 ; 19.6 10.5 ; 9.8
al.72, 2008 98.1 (0.5) 5.5 (0.6) r<20mm 3.2 (0.1) 26.2 (0.7) 20.1 (2.0)
Bueno et 98.1 (0.3) 3.5 (0.3) SEVNB 2.9 (0.2) 14.7 (1.9)
al.136, 2007 98.1 (0.5) 5.5 (0.5) r<30mm 3.0 (0.3) 20.4 (2.8)
De Aza et
~99 1.7 DT 4.2 (0.2)
al.16, 2002
Sbaizero et
99 8 - ~3 20 (5)
al.127, 1998
CT 3.16
1.7
Seidel et ISB 3.6
>99
al.135, 1997 CT 3.75
10.5
ISB 4.1
SEVNB
Damani et <9m 2.8
Low 10
al.52, 1996 ~25m 3.0
~67m 3.5
~92m 3.9
<5 SENB 4-6
Mussler et
>95 6-30 =40-70m 4-4.5
al.123, 1982
2-50 DCB, ISB 3.8-5.8
Claussen et 1 50
SENB
al.130,1982 20 20
SENB 27-36
SEVNB 33-40
Rice et al.13,
3.5-20 DT 38-47
1981
DCB 38-48
DCB* 35-70
4 3.2 ;3.7 17 ; 19
7 3.0 20
Freiman et DT, DCB,
8 3.7 18
al.131, 1973 CNB, SENB
20 4.7 ;4.9 32 ; 35 ; 44
35 4.1 22
in mode I (MPa m1/2); GIC: Critical energy release rate in mode I (J/m2); g: Fracture energy
(J/m2); r: notch tip radius (m); DT: Double Torsion; SEVNB: Single Edge V Notched Beam;
CT: Compact Tension; IT: Indentation Toughness; ISB: Indentation Strength in Bending;
SENB: Single Edge Notched Beam; DCB: Double Cantilever Beam; (S.D.): Standard deviation.
The group of Chevalier16,129 has tested by double torsion alumina materials used as
prostheses in order to detect Subcritical Crack Growth (SCG) in different environments.
The basic mechanism for this subcritical crack growth is the stress assisted corrosion by
32
Captulo 1
environmental species (especially water) at the crack tip bonds, as it was first reported
by Freiman et al.131.
SCG in high purity alumina materials is an extremely slow process in air, as
reported by Chevalier et al.128. These authors found crack growth velocities, v~10-10 m/s
for KIC values (~2.6 MPa m1/2) slightly above than KI0 (~2.5 MPa m1/2).
Values of specific fracture energy for fine grained72 (~ 10 J/m2) and coarse
grained72,127 (~ 20 J/m2) materials are both higher than that reported by Wiederhorn et
al.134 (6 J/m2) for the rombohedral plane which is the preferred cleavage plane in
alumina monocrystals at room temperature. The fracture energies generally determined
for polycrystals are higher than those for monocrystals due to the contribution of
intergranular fracture, in the same way as crack tip toughness values in polycrystals are
higher than those of the easy cleavage planes.
The similarity between the GIC and 2 for the fine grained alumina tested by Bueno
et al.72 revealed the absence of significant crack-size dependent toughening phenomena.
On the contrary, for the coarse grain-sized alumina 2 was much higher (~ 55 %) than
GIC, revealing toughening processes.
1.6. Mullite materials
Mullite (from 3Al2O3.2SiO2 to 2Al2O3.SiO2) is the only stable compound in the
system Al2O3-SiO2 from medium up to high temperature at atmospheric pressure.
Therefore, it is a main constituent of a number of oxide ceramics.
As Schneider et al.137 have described in a review about Structure and properties of
mullite, mullite and mullite ceramics display a large variety of appearances, from
Czochralski-grown single crystals to polycrystalline and polyphase ceramics.
Polycrystalline mullite ceramics may be classified in: monolithic mullite ceramics,
mullite coatings and mullite matrix composites.
33
Introduction
Mullite presents some attractive properties for high-temperature structural
applications such as high melting point (1828 10C)138, low thermal conductivity (6
kcal.m-1.h-1C-1, at 20C)137, low thermal expansion coefficient (25-800C = (4.1 0.1) .
10-6 C-1)139, good chemical and thermal stability, low dielectric constant (e ~ 7), high
deformation resistance at high temperatures (or low creep rate). Most mullite materials
present low characteristic fracture toughness (~ 2.5 MPa m1/2) and moderate
characteristic strength (~ 200 MPa)137. At room temperature, mullite ceramics present
flat R-curve, -no toughening mechanisms have been reported for mullite ceramics-.
Initially used only for high temperature traditional applications (refractories and in
pottery), mullite materials experienced a great development since the seminal work of
Mazdiyasni140 in 1972 where translucent polycrystalline bodies of stoichiometric
mullite with relatively high strength and excellent thermal-shock resistance were
described138,140,141.
Schneider et al.137 described some of the applications of monolithic mullite ceramics
as refractory bricks in glass melting tanks and in furnaces for the temperature treatment
of electronic packaging materials. Optical translucent mullite ceramics are used for
high-temperature furnace windows. Due to the good chemical and thermal stability,
mullite coatings are incorporated to panel for re-entry space vehicles. Also, components
and structures made of mullite fiber-reinforced mullite matrix composites are employed
in gas turbine engines, high duty kiln furnitures and burner tubes.
Reported values for the room temperature mechanical properties of mullite materials
are summarised in table 1.3.
34
Captulo 1
Table 1.3. Microstructural characteristics and room temperature mechanical properties

of mullite materials.
T.D. (S.D) d50 (S.D.) KIC (S.D.) s (S.D.) E (S.D.)

Reference
% (mm) (MPa . m1/2) (MPa) (GPa)
E.d. Values E.d. Values E.d. Values
Sivakumar
et al.149, 98.3 2 IT 2.0 (0.75) 3-pb 441 (34) IET 222
2001
2 2.5 (0.1)
Casellas et 4 2.5 (0.2)
IT
al.153, 1998 1 2.4 (0.2)
1.5 2.2 (0.2)
SENB 4-pb 4-pb 175
Baudn145,
95.2 (0.5) 0.7 (0.5) r=200mm 2.30 (0.09) 4-pb 263
1997 IET 195 (4)
0.05mm/min
SENB 4-pb
r=300mm
Torrecillas 0.005mm/min 1.77

et al.142, High 2 4-pb 200-300 - ~200
0.05mm/min 2.15;2.38
1997
0.2mm/min 2.46;2.61
DT 1.9-2.05
SENB 4-pb
r=200mm
96.5 (0.5) 1.2 (0.9) 0.005mm/min 2.00 (0.31) 4-pb 168 (26) 4-pb 202 (2)
0.05mm/min 2.52 (0.25)
Osendi et 5.0mm/min 3.50 (0.19)
al.148, 1996 SENB 4-pb
r=200mm
95.2 (0.5) 0.7 (0.5) 0.005mm/min 2.00 (0.32) 4-pb 227 (35) 4-pb 195 (4)
0.05mm/min 2.32 (0.08)
5.0mm/min 2.35 (0.10)
Okamoto et
>98 IT 2.7 (0.1) 3-pb 300 (10) IET 224
al.146, 1990
Ismail et
>98 1.04 IT 2.73 3-pb 400 IET 246
al.147, 1987
Mah et
99 6 SCF ~1.7 4-pb 128
al.138, 1983
Mazdiyasni
et al.140, 99.8 (0.2) 2.7 4-pb 269 IET 221
1972
in mode I (MPa m1/2); s: Strength; E: Youngs modulus; : notch tip radius (mm); SCF: Surface
Crack in Flexure; SENB: Single Edge Notched Beam; DT: Double Torsion; IT: Indentation
Toughness; IET: Impulse Excitation Technique; 3-pb: Three-point bending; 4-pb: Four point
bending; (S.D.): Standard deviation.
The availability of fine pure mullite powders and new processing routes has made it
possible to obtain dense polycrystalline mullites with little or nearly no residual glassy
35
Introduction
phase and with higher deformation resistance at higher temperatures than any ceramic
oxide137,141-144. These materials present improved mechanical behaviour as compared to
that of traditional mullite materials. For instance, Kanzaki et al.144 measured bending
strengths of mullite ceramics as high as 360 MPa and fracture toughness of 2.8 MPa
m1/2 at room temperature. The bending strength of these mullite ceramics decreased only
slightly up to 1400C, while other advanced ceramics such as silicon carbide, silicon
nitride, alumina and zirconia display a rather strong reduction of mechanical properties
at elevated temperature, especially in air.
The different processing methods used to obtain the mullite ceramics strongly
influence their mechanical behaviour through the particle size and the homogeneity of
the precursors, the porosity, the existence of impurities and their nature and distribution.
Youngs modulus values between 180 and 250 GPa140,145-148 have been reported for
different mullites. Differences can be explained by different amounts of alumina and
test techniques (static versus dynamic). Contrarily, no dependence on nature and
distribution of impurities on this property has been found on Youngs modulus.
A wide range, 128-440 MPa of strength values has been reported by several
researchers138,140,142,146-151 for mullites with different microstructures. The lowest value
(128 MPa) corresponds to a hot-pressed mullite with average grain size of 6 mm
determined by Mah and Mazdiyasni138 by four-point bending. The authors explained
this low strength as a consequence of large porosity nests which acted as strength-
controlling flaws.
The high mechanical strength (360 MPa) of mullite ceramics published by Kanzaki
et al.144 can be explained by the high purity powders synthesized by spray pyrolysis of
alkoxides used for the processing. Ismail et al.147, Sivakumar et al.149 and Ohira et al.150
reported even higher bending strengths (400 to 440 MPa) starting from commercial high
36
Captulo 1
purity sol-gel mullite powders. The same is true for Itoh et al.151 who published bending
strengths of 415 MPa for mullite prepared from a mixture of kaolinite and aluminium
hydroxide.
The influence of the Al2O3 content of mullite ceramics on the mechanical properties
has systematically been examined by Kumazawa et al.152, who found a significant
increase (~50%) of bend strength with Al2O3 content between 46 to 61 mol.% followed
by slightly decrease for 67 mol.% Al2O3. Multiple factors such as porosity, mullite, a-
Al2O3 and specially glass phase contents may be responsible from this development.
Compared with the changes in bending strength versus Al2O3 content of mullite
ceramics only little variation was observed in the fracture toughness by these authors.
Different researchers have provided fracture toughness values obtained using
different unstable fracture test configurations for mullites with various
microstructures137,138,142,143,145,146,148,153. The range of reported values of the critical
stress intensity factor in mode I, KIC, are around 1.7-3.5 MPa m1/2. In particular, for
fracture testing of SENB in three point bending, 2.0-3.5 MPa m1/2 for load displacement
rates between 0.005 and 5 mm/min have been reported148.
Main differences in SENB toughness values have been attributed to differences in
the rate of displacement of the load frame during the tests142,148. This fact is due to the
proneness of mullite to subcritical crack growth at room temperature141,148,153. For the
same mullite materials, SENB toughness values148 are similar to those obtained by using
the indentation strength four point bending method at equivalent deformation rates153.
Subcritical crack growth (SCG) has been shown to occur in fine grained 3:2
mullites at room temperature142,148,153. Even though SCG occurs in materials presenting
intergranular148 as well as transgranular142,153 fracture mode, it is enhanced by the
presence of a silica glass film at the grain boundaries. This fact was evidenced by results
37
Introduction
obtained for two materials fabricated using the same mullite powder and sintered at
1630C148 and 1700C142. Both materials presented SCG under static loading and double
torsion142 tests. However, in SENB toughness tests, performed at different displacement
rates (0.005-5 mm/min), SCG was only evidenced in the material sintered at the highest
temperature, in which a silica glass film was formed at grain boundaries during
sintering142. Toughness values determined using the slowest (0.005 mm/min) rate
presented relatively high variability (~10 %) for both materials. Slow crack growth in
air is a relatively rapid process in these high purity mullites. Torrecillas et al.142 reported
a crack velocity, v~10-4 m/s, for KIC~1.9-2.05 MPa m1/2.
Up to know no accurate and reliable fracture toughness data have been provided
from stable fracture tests in mullite ceramics. The subcritical crack growth is an
important factor to take in account in this material.
1.7. Refractory materials
1.7.1. General properties
Refractories are essential for all high-temperature industrial processes. They play
the triple role of providing mechanical strength, protection against corrosion and
thermal insulation. Refractory materials, by definition, are supposed to be resistant to
high temperature while exposed to corrosion from solids, liquids and gases, gas
diffusion, different degrees of mechanical stress and strain and mechanical abrasion.
Refractories are mostly used (70%) in metal industry. In petrochemical refining,
although the temperature is much lower than in metal industries the refractories suffer a
high rate of abrasion due to the flow of high-velocity particles at a continuous rate154.
Refractories are broadly divided into two categories: shaped (bricks and cast shapes)
and unshaped (monolithic) refractories154. For shaped refractories the main requirements
are their density and porosity and dimensional tolerance. In recent years, the use of
38
Captulo 1
unshaped refractories has significantly increased reaching ~60-65% of all refractory
use.
Ceramic properties of a refractory material are defined by its nature or reaction when
exposed to high temperature154. For fired bricks like fireclay, high-alumina, magnesia-
chrome-type bricks, the ceramic reactions and bonds have already been instituted by
high-temperature firing, hence, when they are exposed to high temperature, they do not
exhibit any further change. But for unfired refractories, like magnesia or alumina-
carbon bricks and unshaped refractories, the formulations are designed so that the
ceramic properties will be developed at use temperatures. Thus, it is necessary to
understand and predict the ceramic reactions that take place at or before service
temperatures.
Due to the extreme service conditions, the main selection criterion for the
application of refractories is their behaviour at high temperature under specific
corrosive environment, which is mostly done on the basis of their chemistry. However,
microstructure and texture have to be carefully monitored because of the extreme
mechanical solicitations involved in the uses of refractories.
The thermal stress fracture of refractory components, due to temperature cycling
and/or temperature differences through the material, is a widespread problem of
industrial importance. In the same way, mechanical overload as may be originated by
impact during the loading of the process vessel, as occurs in electrical arc furnaces, or
by deformations of the kiln shell, as observed in the cement industry, can lead to
fracture. Fracture as a result of thermal or mechanical shocks could lead to catastrophic
failure of the refractories with strong consequences for the process in which they are
used. However, fracture can also result just in the development of a crack pattern155.
The extension of fracture due to thermal or mechanical strains, i.e: the damage would be
39
Introduction
determined by the ratio between the amount of energy available and the energy needed
to create new crack surfaces or specific fracture energy. Therefore, this ratio will
characterise the resistance of materials to subcritical crack growth and the proneness to
catastrophic failure.
The behaviour of refractories under mechanical loadings is directly related to their
microstructure and texture which, in turn, is determined by the characteristics of the raw
materials (chemical and mineralogical composition and size and shape distribution) and
by the processing or installation procedure. In this regard, it should be pointed out that
the conventional characterisation of the mechanical properties of refractory products is
done in terms of the flexural strength or modulus of rupture, as it is called in the
refractories field. Even though the modulus of rupture is successfully used in industry
for quality control purposes, this parameter does not allow the analysis of the fracture
processes that occur in the material. Therefore, it cannot be used for design and
development purposes.
Refractory products are heterogeneous ceramic materials which fracture exhibits
notable deviations from pure linear elastic and this behaviour has been well
documented155-161. Several of the previously discussed (section 1.2) toughening
mechanisms ahead and behind the crack tip are considered to contribute to this
behaviour. However, these mechanisms present particularities in refractories.
Microcracking is usually accompanied by multiple cracks branching in the frontal
process zone, while the aggregates might act as elastic bridges in the crack wake and/or
debond and be pulled-out. The presence of aggregates in the wake leads to friction of
crack faces when submitted to cycling. As a result, refractories usually present rising R-
curve behaviour, as it was mentioned in section 1.2, in contrast with the flat R-curve
40
Captulo 1
observed for the fine grain size ceramics magnesium-aluminate spinel, alumina and
mullite materials described before in sections 1.4, 1.5 and 1.6, respectively.
The fracture toughness of refractories is usually in the range of about 0.2 to 1.5
MPa m1/2. Refractory castables are at the lower portion of this range and fired bricks at
the higher end155. Youngs modulus of refractories range from about 30 to 70 GPa and
the work of fracture is in the range of about 70-300 J/m2. Therefore, GIC (1-31 J/m2) of
refractories with well designed microstructures is always significantly lower than the
specific fracture energy, GF (2wof). The ratio between the specific fracture energy and
the critical energy release rate in mode I, GF/GIC, has been defined as a toughness,
flexibility or apparent ductility ratio. It will have a large value, higher than 10, for
refractories with high energy consumption for crack propagation relative to crack
initiation. The higher this relation, the higher is the resistance of the material to damage
by thermal or mechanical strains155. An important point to consider when determining
the specific fracture energy (or the work of fracture) of refractories is the potential
influence of the specimen size (size effect) in the obtained values because G F will
increase with increasing fracture surface until the specimen geometry allows the
development of a well developed wake zone.
1.7.2. Toughness testing of refractories
Nakayama, in the 1960s60, was the first researcher that applied the concepts related
to the extension of a single crack to analyse refractory fracture and developed the work-
of-fracture test to characterise toughness of refractories.
Nakayama, Tattersall and Tappin, and Davidge and Tappin accomplished a wide
series of studios on the measurement of work of fracture in the 60s60,62,71. Since then,
many laboratories have conducted work of fracture analysis but a standard test has not
yet been established.
41
Introduction
As discussed before in sections 1.2 and 1.3.3, the concept of work of fracture,
introduced by Nakayama as the mean work per unit of projected fracture area required
to propagate a crack in a stable way, is equivalent to that of specific fracture energy. In
Nakayamas work of fracture test a parallelepiped bar with a triangular (chevron) notch
at its centre is loaded in three point bending. Using this geometry, refractory specimens
fracture in a stable way when a sufficient stiff machine is used. The value of work of
fracture is determined from the total area under the LoadDisplacement curve recorded
during the experimental and the size of the projected fracture surface. Chevron notch
experiments often imply high variability157,158,162-164 (>10%) because coarse aggregates
cause very high work of fracture values as compared to the average for the material
when located close to the apex of the chevron notch. For example, 10-23% variability in
wof has been reported for alumina-spinel castables164. For alumina matrix refractory
compositions (60-100 Al2O3 wt.%158) with fused mullite-zirconia aggregates, and for
high-alumina refractories with 45-99 Al2O3 wt.%162, variability of 3-38% up to 1-26%,
respectively, have been found.
In the 1980s the wedge-splitting fracture test was developed to perform stable
fracture tests and patented by Tschegg165. This test is a special form of the so-called
compact tension test, the specimen with a groove and notch is split in two halves while
monitoring the load and crack mouth opening displacement (CMOD). In this
experimental setup, large specimens of the size of bricks can be tested156,159,166-168.
Several researchers have provided data for high alumina castables and alumina-based
shaped materials166,169-173 using the wedge-splitting test. Recently, a new methodology
was presented using the wedge-splitting test complemented with images obtained
during mechanical loading to determine the crack propagation for a pure alumina and
alumina with titania and zirconia additives refractory compositions171,173. Jin et al.174
42
Captulo 1
have proposed a methodology to estimate the tensile strength, and Youngs modulus of
refractories in addition to the specific fracture energy from wedge splitting test results.
It should be pointed out that variability of data of mechanical properties in the
refractory literature is most of the time not reported. Moreover, in many cases only one
data for each material experimental condition is provided. This is often the case of work
of fracture values determined by the splitting test so, it is not possible to discuss in a
general way the repeatability associated with this technique. Nevertheless, the scarce
data available reveal rather high variability of results. For two commercial alumina
based low cement (2wt.%) castables heat-treated at 1100C, 12 and 18% variability
have been reported in the gwof values169 and variability between 5-22% in gwof has been
reported for basic refractories by Harmuth et al.168.
Despite the fact that there is an evident interest in the wedge splitting test for work
of fracture determination of refractories due to its capability for testing relatively large
specimens in order to average the whole fracture process, it is complicate to establish
this method as a laboratory routine one. Moreover, as mentioned before, variations in
reported ranges of result would mask the effect of microstructural differences on
toughness.
From the extensive spectrum of available fracture toughness tests, bending of
parallelepiped specimens with straight through notches (SENB, Single Edge Notched
Beam) using as control variable the displacement of the load frame (displacement
control) is also a relatively simple way of testing heterogeneous materials, as
refractories. In general, in the refractory field this method is used to determine
toughness parameters describing the initiation of fracture, critical stress intensity factor
in mode I, KIC, and the energy for crack initiation, nbt which is a measure of the critical
energy release rate Gc, Gc = 2nbt. The advantage of using stable fracture tests of SENB
43
Introduction
in three point bending is that both parameters for initiation, nbt, and propagation, wof,
of fracture can be extracted from a single test.
1.8. Nanocomposite ceramic materials
1.8.1. General properties
As described in section 1.1 ceramics are brittle and show catastrophic failure
limiting their applications as structural engineering materials. Therefore, during the last
two decades much effort has been made to develop a new design concept of structural
ceramics throughout the processing improvement and incorporating second phases in
order to overcome the inherent brittleness and the associated low strength and lack of
reliability.
In particular a wide range of monolithic ceramic composites have been developed,
i.e. ceramic matrix composites (CMC), with metallic or ceramic second phases. CMCs
are considered as enabling technology for advanced aeropropulsion, space propulsion,
space power, aerospace vehicles, space structures, ground transportation, as well as
nuclear and chemical industries. They would find applications in advanced aerojet
engines, stationary gas turbines for electrical power generation, heat exchangers, hot gas
filters, radiant burners, heat treatment and materials growth furnaces, nuclear fusion
reactors, automobiles, biological implants, etc. Other applications of CMCs are as
machinery wear parts, cutting and forming tools, valve seals, high precision ball bearing
for corrosive environments, and plungers from chemical pumps175.
According to Niihara21, ceramic matrix composites can be divided into two types:
microcomposites and nanocomposites. In the microcomposites, micro-size second
phases such as particulate, platelet, whisker and fibre are dispersed at the grain
boundaries of the matrix. The main purpose of these composites is to improve the
fracture toughness.
44
Captulo 1
There are three types of nanocomposites: intragranular and intergranular micro-
nanocomposites and nano/nano composites. In the intragranular nanocomposite,
submicrometer and/or nano-size particles of second phase are dispersed mainly within
the matrix grains of micrometer size, whereas in the intergranular nanocomposite, the
second phase particles are dispersed at the grain boundaries of the matrix. Niihara
proposed that the aim of the micro-nano composites was to improve not only the
mechanical properties such as hardness, fracture strength and toughness and reliability
at room temperature but also high-temperature mechanical properties such as hardness,
strength, and creep and fatigue fracture resistances. The nano/nano composites are
composed of the dispersoids and matrix grains with the nanometre-size. The primary
purpose of the nano-nanocomposites, according to Niihara, is to add the new functions
such as machinability and superplasticity like metals to ceramics21.
Originally a number of combinations of ceramic nanocomposites, such as
Si3N4/TiN prepared by chemical vapour deposition and Al2O3/SiC, Al2O3/Si3N4,
Al2O3/TiC, mullite/SiC, B4C/SiC, B4C/TiB2, SiC/amorphous SiC, Si3N4/SiC prepared
by pressureless sintering, hot pressing and hot isostatic pressing, were reported by
Niiharas group21. Strength values at room temperature two to five times higher than
those of monolithic materials were claimed. Moreover, superior high temperature
hardness and creep behaviour as well as resistance to thermal shock fracture were also
reported.
Since the seminal work by Niihara21, the development of ceramic nanocomposites
has generated considerable research activity due to the potential appealing mechanical,
physical and tribological properties. A major challenge in this research field is the
understanding of the basic properties of the materials focusing new developments. In
this sense, several studies investigating the relationships between microstructure and
45
Introduction
strength at room and elevated temperatures, hardness, wear resistance, creep resistance
and toughness have been conducted by various authors in composites ceramic matrix-
ceramic nanoparticles22,72,139,176-254, ceramic matrix-metal nanoparticles127,246,255-260, and
ceramic matrix nanotubes261-264.
Alumina-based nanocomposites have been extensively studied because alumina
exhibits some excellent properties, such as high hardness, low electrical conductivity,
oxidation resistance, good chemical stability and high wear resistance, as it was
mentioned before in section 1.5.
Most researches focus on particle-dispersed alumina composites in order to
improve their mechanical properties. The second phase particles may be
SiC176,178,181,184,188-202,205,206,211,212,214,215,218,220,221,223,226,228,229,231,233,235-241,249-254,265,
ZrO2242,243, TiN/TiC/TiO2/TiCN244,245,248,267, BN268, and metal particle such as Cr255,
W257, Mo258, Ni259, Cu260.
Al2O3/SiC nanocomposites have been the most studied. They are produced by the
incorporation of small (50-200 nm) SiC particles in alumina matrices with typical grain
sizes ~15 mm. Several reviews describing processing, microstructure and mechanical
behaviour of this family of nanocomposites are available22,176,191. To highlight the work
done by the group of the Institute of Scientific and Industrial Research (Osaka,
Japan)175,238,223,205, Leigh University (Pennsylvania, USA)190,195,209,218,219,227,234,269-271,
Leeds University and Oxford University (Great Britain)176,178,179,181,182-184,188,192,193,196-

208,200-202,208,210,211,216,217,222,224,225,230,232,241,247
, Kyoto Institute of Technology209,212,213,272,
Nagoya Institute of technology22,203,214,266. Other groups177, 178,189,190, 199,206,207,215,
220,221,226,229,233,235-237,239,240,246,248,251,254,265,273
.
A dispersion of SiC submicron-sized particles has been proved as beneficial for
some behavioural properties such as creep189,195,214,218,227 and thermal shock182,183. The
46
Captulo 1
most remarkable and reproducible benefits offered by the Al2O3/SiC nanocomposites
are in their tribological properties where great improvements compared with pure
alumina occur in the resistance to severe wear and surface finish following grinding and
polishing as have been reported by different researchers193,176,178,200,180,196,224,225,190

,193,207,218,248,274
.
One of the most studied properties of Al2O3/SiC nanocomposites has been the
flexural strength188,175,209,272,177,181,22,238,182,184,190,192,194,196,198,202,205,207,226,235,236,240. It has
not been possible to reproduce the high values of strength reported by the seminal work
by Niihara21 (from 350 MPa to 1520 MPa for Al2O3/5vol.% SiC composite). Most
authors report moderately higher strength values for the nanocomposites than for single
phase alumina188,175,209,181,192,22,26,196,202,226,235,236,237,240,254.
The addition of one or more components into the base material to form ceramic
matrix nanocomposites has been found to be effective to enhance the fracture toughness
and to improve strength175,240,247,22,248,266,26,275,276. In the nanocomposites the mechanical
properties are determined by many critical factors, such as Al2O3 and SiC particle size
(micrometer or nanosized), composition of powder mixtures (volume fraction of SiC,
content of silica as the product of oxidation of SiC particles), and distribution of SiC
inclusions (intergranular, intragranular or both)226.
1.8.2. Toughness testing of Al2O3/SiC nanocomposites
Main drawback in understanding the mechanical behaviour of nanocomposites is
the lack of accurate and reliable toughness values. Most reported values have been
determined by indentation188-190,192,194,198,199,202,226,235,236,239,251,254,183.
Table 1.4 summarises indentation toughness values reported for Al2O3/SiC
nanocomposites together with their main microstructural characteristics and strength
values.
47
Introduction
Table 1.4. Microstructural characteristics, strength and indentation toughness for

Al2O3/SiC nanocomposites.
vol.% T.D. (S.D.) d50(S.D.) KIC (S.D.) s (S.D.)

Reference
SiC % (mm) (MPa.m1/2) (MPa)
E.d Values
0 98.4 1.6 (0.1) 5.0 (0.3) 300 (125)
5 99.5 11.0 (0.1) 5.6 (0.4) 300 (50)
Parchovianski et
10 99.5 2.0 (0.1) 5.6 (0.4) 4-pb 420 (40)
al.226, 2013
15 99.5 1.4 (0.1) 5.2 (0.6) 420 (40)
20 99.4 0.8 (0.05) 5.1(0.5) 670 (110)
0 99.2 11 3.2 (0.4) 280 (33)
1.6 98.2 4.8 7.6 (0.2) 364 (68)
Shi et al.235,
3.1 97.9 3.4 6.0 (0.4) 3-pb 268 (29)
2010
4.7 96.4 2.3 5.0 (0.4) -
6.2 95.8 1.9 4.1 (0.1) 460 (12)
0 99.2 10 3.2 (0.4) 280 (33.0)
Shi et al.254,
0.3 99.5 4.9 (0.3) 3-pb 516 (20.8)
2010
0.6 99.3 5.1 (0.1) 484 (17.3)
0 2.7(0.4)++
99.3 28.2 (12.8)
3.6 (0.4)+++
Reveron et 1 3.4 (0.6)++
99.2 12.7 (7.9)
al.189, 2010 4.3 (0.6)+++
5 4.0 (0.7)++
97.1 1.4 (0.4)
5.2 (0.7)+++
0 99.2 - 3.7 (0.4) 280.0 (33.0)
0.3 99.8 6.85 4.8 (0.6) 282.2 (50.2)
0.6 99.5 3.41 4.4 (0.3) -
Dong et al.236,
1.6 98.2 1.21 7.4 (0.2) 3-pb 363.8 (67.8)
2009
3.1 97.9 - 5.9 (0.4) -
4.7 96.5 - 5.0 (0.3) -
6.2 95.8 - - 459.7 (11.2)
Galusek et 5 98 1.5 4.5 (02)
al.251, 2007 8 98 1.2 4.8 (0.1)
0 99.6 (0.06) 4.10 (0.94) 2.6 (0.2) 371 (10)
Maensiri et 1 100 (0.9) 6.85 (2.47) 2.3 (0.2) 369 (66)
4-pb
al.183, 2002 2.5 98.1 (1.4) 6.66 (2.48) 2.2 (0.1) 409 (67)
5 99.9 (0.6) 2.82 (0.51) 2.6 (0.1) 417 (56)
0 99.9 3.5 (1.3) 3.2 (0.2) 431 (53)
5 99.8 4.0 (1.1) 4.6 (0.1) 646 (41)
Anya198, 1999 4-pb
10 99.7 2.9 (0.5) 5.2 (0.2) 560 (8)
15 99.6 2.6 (0.3) 5.5 (0.4) 549 (30)
0 99.8 3.25 (0.12) 560 (40)
3 99.3 2.99 (0.15) 887 (67)
Zhang et al.194,
6 99.5 - 3.42 (0.15) 3-pb 853 (53)
1997
12 99.4 4.08 (0.13) 867 (33)
24 99.4 4.63 (0.19) 747 (27)
0 4.20 (0.10)
239
Xu et al.,
6 - 1.7 (0.9) 4.70 (0.11)
1997
12 1.3 (0.7) 4.85 (0.03)
0 0.9 2.1
Hoffman et 5 98.8 2.5 1.9 595 (82)
4-pb
al.277, 1997
48
Captulo 1
vol.% T.D. (S.D.) d50(S.D.) KIC (S.D.) s (S.D.)

Reference
E.d Values
0 99.6 (0.4) 5.0 3.25 (0.27) 491 (63)
Sternitzke et 2.5 99.6 (0.4) 3-5 2.89 (0.34) 475 (35)
4-pb
al.188, 1998 5 99.3 (0.4) 2.1 3.47 (0.61) 689 (85)
10 100 3.5 3.52 (0.41) 539 (56)
3.53 (0.02)H
0 99.6 (0.4) 5.0 491 (63)
3.25 (0.27)
4.23 (0.15)H
97.8 (0.4) 3.5 738 (115)
3.02 (0.21)
Carroll et al.192, 4.00 (0.08)H
98.8 (0.4) 4.0 4-pb 549 (125)
1996 2.76 (0.31)
1.6
3.71 (0.05)H
98.5 (0.4) 2.1 593 (95)
3.04 (0.39)
3.63 (0.05)H
99.3 (0.4) 3.5 689 (80)
3.47 (0.61)
99.5 4.8 2.9 (0.1) 4-pb 559 (51)
0
190 99.3 3.2 2.9 (0.1) 3-pb 569 (72)
Zhao et al. ,
99.9 4.2 3.3 (0.1) 4-pb 1001 (102)
1993
5 99.9 4.2 3.6 (0.1) 4-pb 760 (28)
98.3 5.4 2.7 (0.1) 3-pb 586 (72)
0 ~100 23 3.5 350 (42)
Niihara21, 1991 -
5 ~100 3 4.8 1520 (125)
in mode I (MPa m1/2); s: Strength (MPa); 3 or 4-pb: Three or Four point bending; E.d:
Experimental details; (S.D.): Standard deviation; ++: Anstis equation; +++: Liang equation; H:
Hertzian indentation.
In fact, no conclusive reported data are found for Vickers toughness of Al2O3/SiC
nanocomposites. Considering those studies where limited variability of data is reported,
in general, toughness of the composites are comparable to those of monolithic
alumina183,187,197,247 or relatively low increases in average values are associated with the
presence of SiC nanoparticles (e.g.: 12 vol.%)239.
Table 1.5 summarises the main microstructural characteristics, strength and fracture
toughness for Al2O3/SiC nanocomposites tested by techniques different from IT, such as
CT, SENB/SEVNB, ISB and calculated from the critical defect size and the strength.
Much less toughness testing of nanocomposites has been done with notched beams
in flexure181,22,249,277,278 Davidge et al.181 tested single edge notched beams (SENB) in
flexure and found no statistically significant difference between K IC values for
nanocomposites containing 5, 10 and 15 vol.% SiC. The variability of values reported
49
Introduction
by these authors increased with SiC content. Hoffman and Rdel277 performed Compact
Tension tests to determine crack tip toughness of nanocomposites and found that it was
similar to that of fine grained alumina and equivalent to the fatigue limit of the
nanocomposite. The R-curve determined for the nanocomposite was extremely weak.
As signalled by the authors, main problem of the results was the high variability found
due to the heterogeneous microstructure of the nanocomposite. Choi and Awaji22 used
Single Edge V Notched Beams with notch radius tips lower than 20 m tested in three
point bending, and found a significant increase in toughness from 3.72 MPa m1/2 for
single phase alumina to 5.06 MPa m1/2 for Al2O3/3vol.% SiC. Prez-Rigueiro et al.278
did not find any difference between values obtained for reference aluminas with slightly
higher densities (99% of theoretical) and average grain sizes (2.3-4.6 m) than those
of composites with 5 and 20 vol.% (98% of theoretical and 0.6-1.8 m). Belmonte et
al.249 reported 3 point bending SEVNB toughness results for Al2O3/20vol.% SiC; four
SiC powders were considered. They found an increase (14-37%) in toughness as
compared to those for a reference alumina with slightly higher density (99% of
theoretical) and similar average grain size (0.83 0.52 m) than those of the composites
(98% of theoretical and 0.4-0.8 m). In these works, unstable tests were used and the
reported values might be over-estimated.
Summarising, there is not a well established method for toughness testing of Al2O3/
/SiC nanocomposites and most available data lead to inconsistent conclusions about the
nanocomposite effect. Potential reasons for this fact are that materials with different
degrees of homogeneity have been tested and results have been compared to those
corresponding to single-phase alumina materials with large microstructural differences.
However, the main reason for such a lack of conclusive results is that most data have
been obtained using indentation toughness. This situation for the Al2O3/SiC
50
Captulo 1
nanocomposites, which are the most well known, is still worse for other systems that
have been less studied.
Table 1.5. Microstructural characteristics, strength and fracture toughness for Al2O3/SiC
nanocomposites tested by techniques different from indentation toughness (IT).
vol.% T.D. (S.D) d50 (S.D.) KIC (S.D.) s (S.D.)
Reference
E.d. Values E.d Values
0 99.9 0.83 (0.52) 4.3 (0.1)

Belmonte et
SEVNB
al.249, 2006 99.2 0.86 (0.53) 5.9 (0.3)
99.5 0.80 (0.51) 5.2 (0.4)
20
98.7 0.44 (0.42) 4.9 (0.2)
92.4 0.24 (0.16) 3.0 (0.4)
0 99.85 3.1 345 (55)
2 99.24 15 3.75 380 (55)
3 99.37 11 ISB 4.1 405 (55)
Sun et al.199,
4 99.56 4 + 4.25 3-pb 500 (65)
2005
5 99.67 2.2 3-pb 4.75 610 (65)
6 99.79 2 4.5 530 (55)
7 99.21 1 4.15 495 (55)
Choi et al.22, 0 2 SEVNB 3.72 462
3-pb
2005 3 3-pb 5.06 760
0 >99.5 2-5 3.1 (0.1) 462 (90)
Davidge et 5 >99.5 2-5 3.3 (0.3) 760 (60)
SENB 4-pb
al.181, 1997 10 >99.5 2-5 3.5 (0.2) 806 (33)
20 >99.5 2-5 3.6 (0.4) 793 (55)
Hoffman et 0 >99.5 0.8 2.3 630 (40)
CT 4-pb
al.277, 1997 5 98.8 2.5 2.07 595 (82)
SEVNB
4.0 (0.3)
3-pb
100 4.6 409 (21)
Critical defect
0 4.4
3-pb
SEVNB
99.8 2.3 3.8 390 (20)
3-pb
SEVNB
3.6 (0.2)
3-pb
Perez- 99.0 2.6 492 (60)
Critical defect
Rigueiro et 5 3.1 3-pb
3-pb
al.278, 1997
SEVNB
99.0 3.5 2.5 (0.2) 396 (19)
3-pb
SEVNB
97.6 1.8 3.1 (0.1) 473 (22)
3-pb
SEVNB
20 3.36 (0.04)
3-pb
96.5 0.6 436 (29)
Critical defect
2.7
3-pb
in mode I (MPa m1/2); s: Strength (MPa); SEVNB: Single Edge V Notched Beam; CT:
Compact Tension; SENB: Single Edge Notched Beam; 3 or 4-pb: Three or Four point bending;
E.d: Experimental details; (S.D.): Standard deviation.
51
Introduction
52
Captulo 2
OBJETIVOS E HIPTESIS
Captulo 2
2. OBJETIVOS E HIPTESIS
A partir de todo lo expuesto anteriormente:
Se evidencia la necesidad de contar con un ensayo para la determinacin precisa de
la tenacidad de materiales cermicos, que pueda ser utilizado de manera rutinaria en el
laboratorio.
Por lo tanto, el objetivo principal de este trabajo ha sido el desarrollo de un
dispositivo experimental y una metodologa de ensayo y anlisis de resultados para la
determinacin de la tenacidad de fractura de materiales cermicos.
La hiptesis bsica de este trabajo es la adecuacin de los ensayos de fractura
estable para la determinacin de la tenacidad de fractura de los materiales.
Con objeto de alcanzar fractura estable se propone:
i) La idoneidad de la apertura de los labios de la grieta como parmetro de control
para alcanzar ensayos de fractura estable. Se ha elegido este parmetro
indirecto como variable de control puesto que crece de manera continua y su
deteccin por medios pticos es ms simple que la deteccin directa de la
grieta.
ii) Las altas prestaciones de los equipos electromecnicos modernos van a permitir
el uso de la apertura de los labios de la grieta.
Se han seleccionado ensayos de flexin en tres puntos de probetas SEVNB porque
los medios actuales de introduccin de entallas en forma de V conducen a resultados
reproducibles. El uso de este ensayo est ampliamente extendido.
Con el fin de demostrar la validez del dispositivo experimental y la metodologa de
ensayo y anlisis de resultados desarrollados, se ensayaron materiales cermicos con
comportamientos mecnicos muy diferentes:
55
Objetivos e hiptesis
Espinela de aluminio-magnesio (MgAl2O4), con comportamiento frgil.
Almina (Al2O3) de grano fino, con comportamiento frgil.
Mullita (3Al2O3.2SiO2), con comportamiento frgil y crecimiento subcrtico de
grietas en aire significativo.
Refractarios comerciales, en los cuales ocurren mecanismos de refuerzo durante la
fractura.
Se han propuesto los materiales nanocompuestos de Al2O3/SiC como sistema
modelo para evaluar la validez del dispositivo experimental y de la metodologa de
ensayo y anlisis de resultados desarrollados para la caracterizacin de otros materiales
cermicos avanzados.
56
Captulo 3
DESARROLLO DEL DISPOSITIVO EXPERIMENTAL
Captulo 3
3. DESARROLLO DEL DISPOSITIVO EXPERIMENTAL
El diseo, desarrollo y puesta en servicio del dispositivo experimental y de los
procedimientos de ensayo para la realizacin de ensayos de fractura estable de
materiales cermicos utilizando como variable de control la apertura de los labios de la
grieta, se ha llevado a cabo en el marco de un Proyecto de Investigacin Fundamental
Orientada a la Transmisin de Conocimiento a la Empresa (TRACE), MCI-Microtest
SA:Desarrollo de un equipo para ensayos de fractura estable de materiales, contrato
Microtest-Instituto de Cermica y Vidrio (MCI-PET 2008-0113, 11/2009-12/2011).
Parte de los resultados derivados de esta investigacin dieron lugar a una
publicacin en la revista Journal of Strain Analysis for Engineering Design (Anexo I).
Se ha seleccionado la flexin en tres puntos de probetas en forma de V por su
simplicidad y la repetibilidad de sus resultados (apartado 1.3.2).
Hasta la fecha los parmetros de control de la deformacin de las probetas ms
utilizados han sido carga y desplazamiento del marco de carga, como se explic en el
apartado 1.3.3. En este trabajo se propuso la utilizacin de la apertura de los labios de la
grieta (CMOD) como parmetro de control puesto que aumenta de manera continua
durante la fractura y de esta manera, al menos a nivel terico, sera posible alcanzar
ensayos de fractura estable. El CMOD se plantea adems como parmetro de medida
indirecta como respuesta a la difcil deteccin de grietas por medidas pticas debido a
su pequea apertura.
Teniendo en cuenta la pequea apertura de los labios de la grieta en los materiales
cermicos, se adhirieron unas espigas de 1.5 mm de dimetro y 12 mm de longitud a
ambos lados de la entalla, para poder detectar y medir la apertura de los labios de la
grieta durante el ensayo.
59
Desarrollo del dispositivo experimental
El dispositivo experimental utilizado se muestra en la figura 3.1. Los elementos
necesarios son un emisor que permite la iluminacin de la probeta a ensayar,
delimitando la zona de medida a la distancia existente entre dos espigas adheridas a
ambos lados de la grieta, un receptor, un convertidor y un procesador que permitan
monitorizar y registrar la medida.
Figura 3.1. Equipo para la realizacin de ensayos de fractura estable que consta de
una mquina electromecnica con un dispositivo de flexin en tres puntos, un
micrmetro ptico conectado a un controlador que permite monitorizar y registrar el
CMOD, y el soporte informtico que integra el sistema SCM3000.
3.1. Unidad de medida de apertura de los labios de la grieta
Existen dos grandes grupos de sistemas de medida sin contacto que podran ser
utilizados para medir la apertura de los labios de la grieta: los denominados
micrmetros pticos, basados en la deteccin y anlisis de seales lser (reflexin
lser, interferometra Doppler, difraccin lser, barrido lser, interferometra lser, etc.)
y los basados en visin artificial, mediante cmaras digitales44.
Se plantearon, en concreto, dos opciones para la instrumentacin de medida:
- Micrmetro ptico serie LS-7600
- Visin artificial de 2 megapixeles CV-2600P
60
Captulo 3
Se eligi el micrmetro ptico en funcin del rango de medida y la precisin
necesarios y el tipo de seal proporcionada. Si bien el intervalo exacto de valores de
apertura de los labios de la grieta depender del material ensayado, la geometra
especfica del ensayo- tamao de probeta y distancia interapoyos- y el tamao de la
grieta, una evaluacin preliminar del rango necesario se ha realizado a partir de
resultados previos. Se han considerado ensayos de flexin en tres puntos de probetas
prismticas con entallas en forma de V (SEVNB) de circona tetragonal policristalina
estabilizada con ytria (Y-TZP)76, uno de los materiales cermicos de mayor tenacidad
(KIC~8 MPa m1/2). Asimismo, se han considerado los resultados obtenidos en un
material de almina-mullita279 de muy baja tenacidad (KIC~2-3 MPa m1/2). Para este
material solo se consiguieron ensayos estables o semiestables para tamaos de entalla
relativamente grandes (>0.7), por lo que la mayor parte de las probetas fallaban
durante el mecanizado. Para las geometras utilizadas, Y-TZP76: probetas 50x10x5 mm3,
distancia interaopoyos 40 mm; almina-mullita279: probetas 50x6x4 mm3, distancia
interapoyos 40 mm, la mayor parte de la fractura ocurra en intervalos de CMOD
menores a 25 y 15 m para Y-TZP y el material compuesto, respectivamente76,279. Los
resultados obtenidos han permitido inferir que la precisin requerida estaba por debajo
de 1 m.
El micrmetro ptico seleccionado es de alta precisin del tipo LS7010M (Keyence,
Japn). El principio de medida del sistema es el siguiente (figura 3.2): un LED (Light-
Emitting Diode) verde de alta intensidad GaN irradia luz, que es transformada en un haz
paralelo uniforme a travs de la unidad de difusin especial y de lentes colimadoras, en
el intervalo de medida que va desde 0.04 mm hasta 6 mm. Este haz paralelo "ilumina"
el rea de medida. La imagen de la sombra de la probeta aparece en el HL-CCD (High-
speed Linear Charge Coupled Device) a travs del sistema ptico telecntrico. Con el
61
sistema telecntrico de lentes el tamao de la imagen en el CCD se mantiene, incluso si
se mueve, por lo tanto, se obtiene la misma precisin todo el tiempo. La seal incidente
de salida del HL-CCD es procesada por el procesador DE (Digital Edge-detection) en el
controlador y la CPU. El equipo incorpora una cmara CMOS (Complementary Metal
Oxide Semiconductor) para capturar la imagen en tiempo real.
A continuacin se exponen las caractersticas ms significativas del micrmetro
ptico de alta precisin seleccionado:
El micrmetro ptico seleccionado dispone de un sistema de deteccin basado en un
dispositivo de carga aplicada (HL-CCD) que ofrece una repetibilidad de 0.06 m.
El sistema ptico telecntrico permite la medicin con alta precisin, 0.5 m, muy
superior a la precisin alcanzada mediante el uso de micrmetros de barrido lser ( 2
m) que no sera adecuada a los requerimientos descritos (< 1 m). Para mejorar la
precisin de la medida pueden eliminarse valores anormales que superan un valor
prefijado, por lo que puede evitarse una deteccin incorrecta debida, p.ej., a polvo en el
aire o gotas de agua.
El sensor HL-CCD permite una alta velocidad de muestreo, 2400 datos/segundo,
dos veces ms rpido que los sensores convencionales utilizados en los mtodos de
barrido. Esta alta velocidad es necesaria para permitir la respuesta gil del equipo a
desviaciones de la rampa de incremento de CMOD programada.
El micrmetro ptico convierte las distancias detectadas en seales analgicas que
son procesadas en la unidad de control de la mquina de ensayos donde el software
SCM3000 (Microtest, Espaa) convierte estas seales analgicas en digitales.
62
Captulo 3
Figura 3.2. Principio de medida del micrmetro ptico y sistema SCM3000.
3.2. Sistema y geometra de carga
La mquina de ensayos es electromecnica del tipo EM1/50 (Microtest, Espaa) de
un husillo y doble columna. Las mquinas de ensayo universales electromecnicas se
utilizan sobre todo para ensayos con variaciones montonas de la carga en modo de
traccin o de compresin. Mediante un sistema electrnico se genera la seal de control
que hace que el actuador mueva el puente inferior en direccin ascendente o
descendente. El desplazamiento del puente inferior est controlado y es medido por
medio de un codificador ptico colocado en el eje motor.
63
El controlador electrnico SCM3000 (Microtest, Espaa) incluye canales de carga y
posicin, as como otros canales de posicin auxiliares que permiten la adicin de
canales de deformacin para extensmetros que posibilitan el uso de variables de
control externas a la mquina, no slo carga y posicin, como es el caso de la apertura
de los labios de la grieta (CMOD).
Se ha elegido una mquina de ensayos universal electromecnica porque con ella se
pueden alcanzar variaciones de posicin extremadamente pequeas, como son las
requeridas para ensayos de fractura de materiales frgiles. Si bien tradicionalmente han
sido necesarias las mquinas hidrulicas para asegurar respuestas rpidas del marco de
carga; p.ej., en los ensayos de fatiga, ha sido posible el uso de una mquina
electromecnica debido al gran desarrollo que se ha producido ltimamente en los
sistemas de control y en la mecnica de este tipo de mquinas.
El intervalo de carga mxima para la mquina de ensayos seleccionada es de 50 kN,
tanto a traccin como a compresin. El rango de desplazamiento es de 0-100 mm y la
velocidad mxima es de 100 mm/min. Las cargas reales aplicadas se miden mediante las
clulas de carga extensomtricas montadas en la lnea de aplicacin de la fuerza. Se ha
seleccionado una clula de carga de 5kN para asegurar la alta rigidez del dispositivo
experimental que es ~ 2 108 N/m.
Las probetas a ensayar se colocan en un dispositivo de flexin en tres puntos entre el
marco de carga y el puente inferior mvil.
En la figura 3.3 se muestra un esquema general del dispositivo de flexin en tres
puntos, donde, P, es la carga aplicada, L, es la distancia entre los apoyos inferiores del
marco de carga, B, es el espesor de la probeta, W, es la anchura de la probeta y, a, es la
longitud de la grieta.
64
Captulo 3
P
P
P/2 P/2 P/2 P/2
Figura 3.3. Representacin esquemtica de la geometra de carga para ensayos de

flexin en tres puntos, donde P es la carga aplicada, L es la distancia entre apoyos
inferiores, B es el espesor de la probeta, W es la anchura de la probeta y a es la longitud
de la grieta. Para materiales elstico-lineales, la deflexin de la probeta () puede
igualarse al desplazamiento del marco de carga.
3.2.1. Cermicas tcnicas avanzadas
Para las cermicas tcnicas avanzadas se ha elegido la geometra de la norma clsica
ASTM STP601 puesto que es la que ms se ha utilizado y se dispone de un mayor
nmero de datos sobre materiales cermicos. Esta geometra consiste en probetas de
50x6x4 mm3 y una distancia interapoyos de 40 mm. Los soportes de flexin son rgidos
de acero inoxidable (Figuras 3.4 y 3.5).
La flexibilidad de la mquina, clula de carga y sistema de apoyos se determin
experimentalmente ensayando una barra de almina sin agrietar de 4 mm de espesor
(B), 6 mm de anchura (W) y 50 mm de longitud; el valor obtenido fue 1.5 10-7 m/N
hasta 150 N.
Para la determinacin de la tenacidad de fractura, como se ha mencionado
anteriormente (apartado 1.3.2), se seleccionaron probetas prismticas con entalla en
forma de V (SEVNB). La entalla inicialmente se hizo con un disco diamantado de 300
m de espesor. Usando esta pre-entalla como gua, la parte restante de la entalla se
corrigi con una cuchilla de 150 m de espesor, impregnada con pasta de diamante de
15, 6 y 1 m, de tal manera que se consigui un radio de curvatura de fondo de entalla
por debajo de 10 m, de acuerdo con las recomendaciones derivadas de las conclusiones
del round robin descrito en el apartado 1.3.2 (Figura 3.6).
65
Figura 3.4. Mquina de ensayos Microtest EM1/50 con el dispositivo experimental

para cermicas tcnicas avanzadas.
Figura 3.5. Detalle del dispositivo de flexin en tres puntos con el micrmetro
ptico para cermicas tcnicas avanzadas.
66
Captulo 3
Figura 3.6. Superficie lateral de una probeta del material de almina ensayado.
Microscopia ptica de luz reflejada.
(a) Corte con disco diamantado, pre-entalla.
(b) y (c) Se muestra el fondo de entalla en forma de "V" corregido con una cuchilla
impregnada en pasta de diamante.
3.2.2. Materiales refractarios
Para los materiales refractarios se han seleccionado dos tamaos de probetas
150x25x25 y 200x40x40 mm3, siendo la distancia interapoyos 125 y 180 mm,
respectivamente. Estas dimensiones son las recomendadas para la determinacin del
mdulo de rotura de materiales refractarios conformados densos segn norma UNE-EN
993-6:1995280. El rodillo de carga es articulado mientras que los rodillos de soporte
tienen posibilidad de giro sin articulacin. (Figuras 3.7 y 3.8).
Figura 3.7. Mquina de ensayos Microtest EM1/50 con el dispositivo experimental

para materiales cermicos refractarios conformados densos.
67
Figura 3.8. Detalle del dispositivo de flexin en tres puntos con el micrmetro
ptico para materiales cermicos refractarios conformados densos.
La flexibilidad de la mquina, clula de carga y sistema de apoyos se determin
experimentalmente ensayando una barra de almina densa sin agrietar de 25 mm de
espesor (B), 25 mm de anchura (W) y 150 mm de longitud. El valor obtenido fue 1.5
10-7 m/N hasta 150 N, el mismo valor que el obtenido con el dispositivo de flexin para
la configuracin de cermicas avanzadas.
Para la determinacin de la tenacidad de fractura de los materiales refractarios se
utilizaron probetas de flexin en tres puntos con entallas rectas, SENB. Como el tamao
de agregados en estos materiales heterogneos es ~ 3-5 mm, no es necesario utilizar
entallas con forma en V, y la entalla se realiz con un disco diamantado de 300 m
obtenindose radios de fondo de entalla alrededor de 100-150 m (Figura 3.9).
Figura 3.9. Superficie lateral de una probeta del material refractario silico-aluminoso
ensayado. Microscopia ptica de luz reflejada.
68
Captulo 3
3.3. Dispositivo experimental
Una vez definido el sistema de carga y el micrmetro ptico como unidad de medida
sin contacto de la apertura de los labios de la grieta, ha sido necesario asegurar el
acoplamiento global. Por una parte fue preciso asegurar el ajuste mecnico entre la
mquina de ensayos y el micrmetro ptico. Para evitar interferencias mecnicas y
asegurar una correcta orientacin del haz luminoso con respecto al eje de carga y el
montaje de flexin, el micrmetro ptico se uni al apoyo de carga inferior.
Por otra parte, ha sido preciso desarrollar una interfaz entre la mquina de ensayos
y el micrmetro ptico que permita tanto registrar el CMOD como la realizacin de
ensayos en control de este parmetro.
Para el tamao de probeta de 50x6x4 mm3 se seleccion un rango de medida de
50 m, segn las variaciones de CMOD detectadas en ensayos previos, este rango en
trminos de seal analgica se corresponde con un voltaje de escala de 5 m/V, lo que
se traduce en una resolucin terica de 0.02 m.
Para los tamaos de probetas de 150x25x25 y 200x40x40 mm3 fue necesario
aumentar el rango de escala hasta 5 mm obtenindose entonces una resolucin terica
de 2 m.
Durante la realizacin de los ensayos se registraron los valores de carga (P),
posicin (d), apertura de los labios de la grieta (CMOD) y tiempo (t), de forma
continua.
3.4. Anlisis de estabilidad de ensayos de probetas SENB en control
por posicin
Como se describi en el apartado 1.3.3, es difcil conseguir fractura estable de
materiales frgiles en ensayos de flexin en tres puntos de probetas SENB utilizando el
desplazamiento del marco de carga como parmetro de control. Es posible justificar esta
69
dificultad a partir del anlisis del locus de Griffith que describe las relaciones carga-
deflexin de la probeta durante el proceso de fractura estable.
De acuerdo a los anlisis realizados por Sigl73, se produce crecimiento inestable de
grieta cuando (G/a) (Gc/ac) y crecimiento estable de grieta cuando (G/a) <
(Gc/ac), donde Gc es la tasa crtica de liberacin de energa (energy release rate,
apartado 1.1) y la longitud de entalla relativa (=a/W). Por lo tanto la condicin de
equilibrio que define el locus de Grifftith es G=Gc.
Para representar los locus de Griffith correspondientes a ensayos de fractura estable
de materiales bajo condiciones experimentales dadas, Biolzi y col.74 definieron los
parmetros carga y deflexin de la probeta (equivalente al desplazamiento, figura 3.3)
adimensionales.
Teniendo en cuenta que la relacin entre la carga aplicada (P) y el factor de
intensidad de tensiones en modo I (KI) para ensayos de flexin en tres puntos, viene
dada por la ecuacin 3.1281:
3PL
KI = K b (a )
2 BW 3 / 2 (3.1)
donde P, L, B y W tienen el mismo significado que anteriormente (figura 3.3) y
K() es una funcin geomtrica vlida para cualquier valor de longitud de grieta
relativa (0 1) y relaciones de distancia entre los apoyos-anchura de la probeta ( =
L/W) mayores que 2.5 (2.5 16) (ecuacin 3.2)281:
a {p (a ) + 4 / b [ p 4 (a ) - p (a )]}
K b (a ) =
(1 - a ) 3 / 2 (1 + 3a ) (3.2)
70
Captulo 3
Las funciones p4() y p() definidas en las ecuaciones (3.3) y (3.4) son polinomios
cbicos para = 4 (equivalente a una probeta de referencia con L/W = 4) y =
(formalmente equivalente a flexin pura)281.
p4 (a ) = 1.9 + 0.41a + 0.51a 2 - 0.17a 3 (3.3)
p (a ) = 1.99 + 0.83a - 0.31a 2 + 0.14a 3 (3.4)
La fractura tendr lugar para una carga mxima aplicada (Pmax) tal que el factor de
intensidad de tensiones en modo I (KI) en la punta de la grieta alcance el factor crtico
de intensidad de tensiones en modo I (KIC) del material, ecuacin 3.5.
3Pmax L
K IC = K b (a )
2 BW 3 / 2
(3.5)
Al dividir ambos trminos de la ecuacin 3.5 entre f/W1/2, se obtiene la ecuacin
3.674:
KI 3P L
= K b (a ) = s
s- f W 1/ 2
2s f BW 2 (3.6)
siendo f la tensin de fractura determinada en flexin y s un nmero adimensional
que describe la fragilidad de la muestra.
De esta manera queda descrita la carga adimensional segn la ecuacin (3.7)74:
PL 2s
=
s f BW 2
3K b (a )
(3.7)
La deflexin de la probeta, ds, viene determinada por la deflexin de la probeta sin
grieta (d) y la deflexin debida a la grieta (Dd),ds = d+Dd, ecuaciones 3.8-3.11.

d=
2P
(L 2W )2 L 2W + 2.85 (L 2W ) - 0.42 2 4
EB (L 2W ) (3.8)
18(L 2W )
2 a 2
Dd =
EB
P K b (a ) a [ ]
0 (3.9)
71
a
ds =
2P
EB
( )2 L
L 2W
2.85
+
2W (L 2W ) - 0.42
(L 2W )

[ 2
2 4 + 9 a f (a ) ]
0 (3.10)
P
ds = ls
EB (3.11)
donde, E, es el mdulo de Young, ls, es la flexibilidad adimensional de la probeta y
f(a) es igual a Kb(a)/a1/2.
En una primera aproximacin, el anlisis terico de estabilidad se hace suponiendo
mquinas de ensayo totalmente rgidas, de tal manera que la flexibilidad de la mquina
y dispositivos de ensayo, dm, sea igual a cero y la flexibilidad total, dt=ds+dm, sea
equivalente a la de la probeta.
Al dividir ambos trminos de la ecuacin 3.11 entre f/W1/2, se obtiene la ecuacin
3.12:
ds P
= ls
s fW 12
s f EBW 1 2
(3.12)
Aplicando la ley de Hooke (f = E.ef, donde ef es la deformacin unitaria) y
relacionando la ecuacin 3.12 con la ecuacin 3.7 queda descrita la deflexin
adimensional de la probeta segn la ecuacin 3.13:
dsL PL 2s
= ls = ls
e fW 2
s f BW 2
3K b (a ) (3.13)
De esta manera, se puede representar, para una geometra de carga determinada
(siendo L y W conocidos) la carga adimensional frente a la deflexin adimensional, para
distintos valores de la fragilidad de la muestra (s). De acuerdo con esta aproximacin, la
grieta crecer de manera estable cuando se cumpla que (P/d) es menor o igual a cero.
En la figura 3.10 se representan las curvas caractersticas carga-deflexin
adimensional para la geometra de ensayo utilizada en este trabajo para caracterizar los
72
Captulo 3
materiales avanzados. Se representa el locus de Griffith para distintos nmeros de
fragilidad de la probeta (s). Se observa como primero la deflexin de la probeta
disminuye con la carga (rgimen I) y despus la deflexin aumenta mientras la carga
sigue disminuyendo (rgimen II), para distintos valores de fragilidad de la probeta (s)74.
Conforme aumenta el nmero de fragilidad de la probeta (s) el locus de Griffith se
desplaza hacia la derecha aumentando la regin correspondiente al rgimen II. La curva
P- de la figura 3.10 presenta una derivada positiva para a/W < 0.4 y una derivada
negativa para a/W 0.4. Por lo tanto, para la geometra de ensayo utilizada en este
trabajo para cermicas tcnicas avanzadas, si se utilizara una mquina con rigidez
infinita, se alcanzara crecimiento estable de grieta (rgimen II) a velocidades crecientes
del desplazamiento del marco de carga (control por desplazamiento), a partir de
longitudes de entalla relativas (a/W) superiores a 0.4.
Figura 3.10. Locus de Griffith correspondiente a los valores numricos carga-

deflexin adimensionales para la geometra de ensayo de cermicas tcnicas avanzadas
(apartado 3.2.1). Clculos realizados para flexibilidad de la mquina igual a cero y
probetas con diferentes nmeros de fragilidad, s. La lnea recta corresponde al tamao
relativo de entalla crtico, (c=0.40).
73
En la figura 3.11 se representan las curvas caractersticas carga-deflexin
adimensionales para la geometra de ensayo utilizada para caracterizar los materiales
refractarios, (probetas pequeas 150x25x25 mm3). La curva P- presenta una derivada
positiva para a/W < 0.6 y una derivada negativa para a/W 0.6. Se alcanzara pues
crecimiento estable de grieta (rgimen II) a velocidades crecientes del desplazamiento
del marco de carga (control por desplazamiento), a partir de longitudes de entalla
relativas (a/W) superiores a 0.6. El mismo resultado se obtiene para la geometra
seleccionada para el ensayo de probetas mayores 200x40x40 mm3, figura 3.12.
1.8
1.6
s=0.1
Carga adimensional
1.4 s=0.2
s=0.3
1.2
s=0.4
1.0 s=0.5
a=a/W=0.60 s=0.6
0.8
s=0.7
0.6 s=0.8
s=0.9
0.4
s=1.0
0.2
0.0
0 20 40 60 80 100 120 140
Deflexin adimensional

deflexin adimensionales para la geometra de ensayo de refractarios (probetas
pequeas 150x25x25 mm3, apartado 3.2.2). Clculos realizados para flexibilidad de la
mquina igual a cero y probetas con diferentes nmeros de fragilidad, s. La lnea recta
corresponde al tamao relativo de entalla crtico, (c=0.60).
74
Captulo 3
Los valores de a obtenidos por Biozi y col.74 a partir de los cuales se alcanzan
ensayos de fractura estable, son coincidentes con aquellos obtenidos por Bar-On y col70.
1.8
1.6 s=0.1
s=0.2
1.4 s=0.3
Carga adimensional
1.2 s=0.4
s=0.5
1.0 a=a/W=0.60 s=0.6
s=0.7
0.8
s=0.8
0.6 s=0.9
s=1.0
0.4
0.2
0.0
0 20 40 60 80 100 120 140

deflexin adimensionales para la geometra de ensayo de refractarios (200x40x40 mm3,
apartado 3.2.2). Clculos realizados para flexibilidad de la mquina igual a cero y
probetas con diferentes nmeros de fragilidad, s. La lnea recta corresponde al tamao
relativo de entalla crtico, (c=0.60).
3.5. Verificacin del dispositivo experimental y procedimiento de
ensayo
Para verificar la adecuacin del sistema desarrollado y del CMOD como parmetro
de control del movimiento del marco de carga durante el ensayo, se seleccion una
espinela de magnesio-aluminio que se caracteriza por ser un material cermico frgil
(seccin 1.4).
Las propiedades del material ensayado84 se describieron en el captulo 1, tabla 1.1.
Se trata de una espinela de grano fino (1.5 m) y alta densidad (98% terica) obtenida a
partir de un polvo comercial prensado isostticamente en fro y sinterizado en aire con
75
una velocidad de calentamiento y enfriamiento de 2C/min hasta 1630 C y un tiempo
de permanencia de 2 h.
La geometra de ensayo fue la correspondiente a cermicas tcnicas avanzadas
descrita anteriormente en el apartado 3.2.1. Las caractersticas de las entallas fueron:
a=0.64 y radio de curvatura del fondo de entalla alrededor de 25 m.
En la figura 3.13 se representa el locus de Griffith determinado teniendo en cuenta
la flexibilidad del equipo (ecuacin 3.11) y las propiedades del material de espinela
(s=0.22). Para alcanzar ensayos estables en control por desplazamiento del marco de
carga seran necesarios a0.70, superior al valor obtenido considerando la flexibilidad
de la mquina nula (figura 3.10).
0.40
0.35 s=0.22; lm=155
0.30
Carga adimensional
0.25
0.20
0.15 a=a/W=0.70
0.10
0.05
0.00
0 10 20 30 40 50 60 70 80 90 100

deflexin adimensionales para la geometra de ensayo de cermicas tcnicas avanzadas
(apartado 3.2.1). Clculos realizados para la flexibilidad adimensional, lm,
correspondiente a la flexibilidad del dispositivo utilizado, (dm/P)=1.5 10-7 m/N. La lnea
recta corresponde al tamao relativo de entalla crtico, (c=0.70).
Con el fin de constatar la posibilidad del CMOD como parmetro de control, se
realizaron ensayos a velocidades constantes de CMOD desde 0.05 hasta 1 m/min,
76
Captulo 3
como se observa en la figura 3.14, las velocidades programadas se alcanzaron en todos
los casos con variaciones inferiores a 0.02 m/min.
H M
4.0 1.6
3.5 a 1.4 b
3.0 1.2
CMOD [mm]
CMOD [mm]
2.5 1.0
2.0 0.8
1.5 0.6
L
1.0 0.4
0.5 0.2
0.0 0.0
0 200 400 600 800 1000 0 200 400 600 800 1000
Tiempo [s] Tiempo [s]
Figura 3.14. Valores experimentales CMOD-tiempo para distintas velocidades de

ensayo programadas:
(a) L = 0.05 m/min, M = 0.5 m/min, H = 1 m/min.
(b) 0.1 m/min
En la figura 3.15 se muestra una curva experimental caracterstica carga-tiempo
correspondiente a la espinela ensayada en este trabajo a 0.1 m/min, se observa una
disminucin montona de la carga con el aumento de tiempo despus de alcanzar la
carga mxima, como corresponde a los ensayos de fractura estable.
En la figura 3.16 se muestra la curva experimental caracterstica carga-
desplazamiento correspondiente al ensayo de la figura 3.15, se observa cmo, a partir
del valor de carga mxima, es necesaria una disminucin del desplazamiento para
alcanzar ensayos estables.
77
12
10
Carga [N]
8
6
12
4
2
10
0
820 840 860 880
Tiempo [s]
8
Carga [N]
6
0
0 200 400 600 800 1000
Tiempo [s]
Figura 3.15. Curva experimental caracterstica carga-tiempo correspondiente a la

espinela caracterizada en este trabajo. Se muestra en detalle la zona de fractura donde
se aprecia una disminucin montona de la carga con el tiempo despus de alcanzar la
carga mxima, correspondiente a la fractura estable.
12
10
8
Carga [N]
0
0.000 0.001 0.002 0.003 0.004 0.005
Desplazamiento [mm]
Figura 3.16. Curva experimental caracterstica carga-desplazamiento. Se observa

una disminucin de la posicin tras alcanzar el punto de carga mximo caracterstica de
los ensayos de fractura estable en control por CMOD.
En la tabla 3.1 se muestran los valores de tenacidad de fractura obtenidos en este
trabajo a partir de ensayos estables y los valores obtenidos en ensayos inestables de
probetas con igual geometra y mayor radio de curvatura (apartado 1.4) por Baudn y
78
Captulo 3
col84. Los valores de KIC y GIC son la media de los obtenidos a partir de tres ensayos y
S.D. es la desviacin estndar.
Para calcular KIC se utiliz la ecuacin 3.1. A partir del valor de K IC, el mdulo de
Young y el coeficiente de Poisson (0.294 para espinela de aluminio-magnesio)282 se
calcul la tasa crtica de liberacin de energa en modo I, G IC, de acuerdo con el anlisis
de Irwin para condiciones de deformacin plana, segn la ecuacin 1.3.
El valor de KIC es un 65% inferior al valor reportado previamente para este mismo
material84. Esta diferencia se hace ms patente cuando se considera GIC, que es un orden
de magnitud inferior. El efecto combinado del mayor radio de curvatura del fondo de
entalla (apartado 1.4) y la energa adicional involucrada en la fractura inestable, es
responsable de esta disparidad en los valores.
Tabla 3.1. Parmetros de tenacidad de fractura para el material de espinela.
Baudn y col.84 Este trabajo

KIC (S.D.) GIC (S.D.) KIC (S.D.) GIC (S.D.)
1/2
(MPa m ) (J/m2) (MPa m ) 1/2
(J/m2)
3.0 (0.1) 32 (2) 1.04 (0.06) 3.94 (0.01)
KIC: Factor crtico de intensidad de tensiones en modo I (MPa m1/2); GIC: Tasa
crtica de liberacin de energa en modo I (J/m2); S.D.: Desviacin estndar.
3.6. Conclusiones
- Se ha desarrollado un dispositivo experimental que utiliza la seal analgica
correspondiente a la apertura de los labios de la grieta (CMOD), medida por un
micrmetro ptico de alta precisin, para controlar el movimiento del marco de carga de
las mquinas de ensayos electromecnicas comercializadas por Microtest (Espaa).
- La apertura de los labios de la grieta (CMOD) es un parmetro de control vlido
para alcanzar ensayos de fractura estable de manera rutinaria en laboratorio de probetas
estndar SEVNB en flexin en tres puntos.
79
- Para un material de espinela de grano fino, modelo de material con fractura frgil,
el valor de KIC, obtenido a partir de ensayos estables es sensiblemente inferior (65%) al
obtenido a partir de ensayos inestables utilizando el desplazamiento del marco de carga
como parmetro de control y radios de fondo de entalla superiores a los empleados en
este trabajo.
80
ANEXO I
Este trabajo se present como pster en la Conferencia Ibrica de Fractura e
Integridad Estructural celebrada en Oporto en 2010.
- C. Baudn, A. Garca, J. Hernndez, M. Lpez. Controlled fracture tests of

brittle ceramics. Conferencia Ibrica de Fractura e Integridad Estructural 2010.
Oporto, Portugal. 17-19 de Marzo de 2010.
Los trabajos presentados en esta conferencia aparecen publicados en los libros
colectivos Anales de Mecnica de la Fractura:
- C. Baudn, A. Garca, J. Hernndez, M. Lpez. Controlled fracture tests of

brittle ceramics. Proceedings XXVII Encuentro del Grupo Espaol de Fractura.
Volumen I. p: 291-295. Ed. Secretara del Grupo Espaol de Fractura, Madrid
(2010).
A raz de la presentacin de este trabajo se recibi una invitacin para su
publicacin en la revista Journal of Strain Analysis for Engineering Design:
- A. Garca-Prieto, J. Hernndez, M. Lpez and C. Baudn., "Controlled fracture

test for brittle ceramics", J. Strain Anal. Eng. Des., 46 (1) 27-32 (2011). (DOI:
10.1243/03093247JSA685).
Journal Citation reports 2011 (JCR): IF=1.085, posicin 10 de 32 en la categora

Materials Science, Characterisation and Testing.
81
Mensaje recibido para la invitacin de la publicacin de este trabajo en la revista JSA:
-----Mensaje original-----
De: gustavovictor.guinea@upm.es [mailto:gustavovictor.guinea@upm.es]
Enviado el: mircoles, 21 de abril de 2010 12:43
Para: Baudn de la Lastra Carmen
CC: Paulo de Castro T
Asunto: Invitation to submit your paper to JSA
Dear Mrs. Baudn,
On behalf of the Selection Committee, it is our pleasure to inform you that your work
entitled "CONTROLLED FRACTURE TESTS OF BRITTLE CERAMICS" and co
authored by "C. Baudn, A. Garca, J. Hernndez, M. Lpez" presented at the past
CIFIE 2010 conference held at Porto on March 16-19, 2010, has been pre-selected for
recommended publication in the Journal of Strain Analysis for Engineering Design
(JSA) (http://journals.pepublishing.com/content/119785) following the agreement
between the Spanish Group of Fracture and this journal.
Please let us know at your earliest convenience, and in any case before
April 30, 2010, if you on behalf of the other authors agree to submit your
work to this journal, under the understanding that all the pre-selected works will be
peer-reviewed before publishing.
If your answer is positive, we will contact you again to send detailed

instructions on how to proceed.
Looking forward to hearing from you.
With best regards.
Gustavo Guinea
Paulo de Castro
82
27
Controlled fracture test for brittle ceramics

A Garca-Prieto1, J Hernandez2, M Lopez2, and C Baudn1*
1
Instituto de Ceramica y Vidrio, CSIC-Campus de Cantoblanco, Madrid, Spain
2
Microtest S.A., Madrid, Spain
The manuscript was received on 15 June 2010 and was accepted after revision for publication on 29 July 2010.
DOI: 10.1243/03093247JSA685
Abstract: Controlled fracture tests that ensure the full conversion of supplied energy into crack
surface energy are required for the accurate determination of the toughness parameters of
materials. In this work, a new experimental configuration to perform three-point-bending single-
edge V-notch beam stable fracture tests controlled by the crack mouth opening displacement
(CMOD) is presented. The CMOD is chosen as the control variable because it is the only
parameter that increases throughout the whole fracture process. Results obtained by applying
the proposed approach to a fine-grained test sample of the spinel-structured phase of
magnesium aluminium oxide are presented. The results obtained demonstrate that the proposed
configuration allows stable fracture tests to be performed on extremely brittle ceramics.
Keywords: ceramics, spinel, stable fracture, toughness, work of fracture
1 INTRODUCTION materials that display some level of R-curve behaviour

[3, 4] since there has to be a mechanism to allow the
Stable crack growth conditions are required for load frame displacement to decrease after the peak
reliable and accurate fracture toughness data to be load for stable testing of brittle materials.
obtained. If the fracture toughness values are deter- The crack mouth opening displacement (CMOD) is
mined using test configurations that do not allow the only parameter that increases throughout the
stable crack growth, then the measurements are only whole fracture process, thus, it has been proposed as
valid for the crack initiation process [1]. In such cases, a control variable for stable fracture testing of
the calculated value of the fracture toughness will be ceramics. CMOD-controlled stable fracture tests have
over-evaluated unless the kinetic energy term in the been reported for advanced ceramics that display R-
growth process is ignored [2]. In addition to the curve behaviour such as yttria-partially-stabilized
fracture toughness, stable fracture tests allow the zirconia [5] but not for extremely brittle ceramics
determination of fracture energy and crack-growth such as the spinel-structured phase of magnesium
resistance curves. aluminium oxide (MgAl2O4). Such tests are per-
Controlled fracture tests for brittle materials, such as formed using specific experimental laboratory setups.
the majority of ceramics, are difficult to accomplish In this work, a new experimental configuration to
and therefore they are not usually performed. A perform stable fracture tests on ceramics controlled
chevron notched geometry allows stable fractures to by the CMOD parameter is presented together with
be created in brittle specimens using the displacement results obtained using different experimental condi-
of the loading frame as a control parameter; however, tions for a fine-grained test sample of the spinel-
test specimens are difficult to fabricate especially for structured phase of magnesium aluminium oxide.
fine-grained materials. For other geometries, such as
straight notch beams in flexure and load-frame-
displacement tests, stable fracture is only reached for 2 EQUIPMENT
2.1 Loading device

*Corresponding author: Department of Ceramics, Instituto de
Ceramica y Vidrio, CSIC, Kelsen 5, Madrid, 28049, Spain. The Microtest EM1/50 (Microtest, Madrid, Spain) is
email: cbaudin@icv.csic.es a single-screw dual-column servo-controlled elec-
J. Strain Analysis Vol. 46
83
28 A Garca-Prieto, J Hernandez, M Lopez, and C Baudn
tromechanical test machine (Fig. 1). Electromechan- fore, the control parameters can be not only force
ical or universal testing machines are most com- and displacement but also CMOD.
monly used for static testing in either tensile or The test specimen was placed between the rigid
compression modes within a single frame. The frame and the moving crosshead in a stainless steel
control is performed by an electronic system that three-point bending test fixture with a span (S) of
generates the control signal to make the actuator 40 mm (Figs 1 and 2). A load cell of 5 kN was selected
move the crosshead in an upward or downward to ensure a high stiffness of the loading set-up. The
direction via a drive system. The test samples are compliance of the machine, load cell, and support
placed between the rigid frame (stiffness , 26108 N/ arrangement, in the load range used for testing (up
m) and the moving crosshead. to 150 N) was determined experimentally using an
The maximum load range of the machine is 50 kN, uncracked alumina bar of 4 mm thickness (B), 6 mm
both in tension and compression. The displacement width (W), and 50 mm length; the obtained value
range is 0100 mm and the maximum velocity is was 1.561027 m/N.
100 mm/min. The actual applied loads are measured
by extensometer load cells mounted along the line of
force application. The displacement of the moving 2.2 CMOD measurement unit
crosshead is controlled and measured by means of
For the CMOD measurement and control, a high-pre-
an optical encoder placed in the motor axis.
cision optical micrometer a Keyence LS7010 (Key-
The SCM3000 electronic controller (Microtest,
ence, Osaka, Japan) that incorporates a complemen-
Madrid, Spain) includes load and position channels
tary metal oxide semiconductor (CMOS) camera was
as well as position auxiliary channels with the option
used to capture real-time images of the target (Figs 1
to add additional strain channels for extensometers.
and 2). This optical system provided a measuring
In this case, the signal from the contact-free optical
accuracy of 0.5 mm.
measurement system for CMOD determination is
The principle of measurement of the optical system
directed to one of these auxiliary channels. There-
is as follows (Fig. 3). A high-intensity GaN light-
emitting diode (LED) radiates light which is converted
into a uniform parallel beam by the special diffusion
unit and collimator lens and this beam is used to
illuminate the measurement area. The shadow image
of the target is projected on to the high-speed linear
charge coupled device (HL-CCD) by the telecentric
optical system. The system of lenses ensures that the
size of the image on the CCD does not change even
if it moves, thus, the same precision is maintained.
The output incident signal of the HL-CCD is pro-
cessed by the digital edge-detection (DE) processor
in the controller and central processing unit (CPU).
The controller of the optical system incorporates a
function that eliminates abnormal values in order to
Fig. 1 A Microtest EM1/50 test machine and the Fig. 2 The three-point bending device with the optical
experimental set-up micrometer
84
Controlled fracture test for brittle ceramics 29
Fig. 3 Schematic diagram of the SCM 3000 control system
improve the precision of the measurement, with the separation was checked to be 0.06 mm using a
detection threshold being the adjustable parameter. 1.0 mm diameter round bar located in the centre of
the measuring area. The equipment performed
continuous measurements at a rate of 512 samples
2.3 Set-up per second.
The optical micrometer was attached to the lower The optical micrometer converts the detected
loading support; in this way mechanical obstruc- distances into analogue signals that are used as
tions are avoided and a correct orientation of the input to the control unit of the mechanical testing
light beam with respect to the axis of load and the machine where the software in the SCM3000 con-
bending fixture is ensured. Given the small opening verts them to discrete digital numbers. The whole
displacement of the notch, in order to be able to range of the optical micrometer (6 mm) that corres-
detect and measure its width during the test, pins of ponds to 10 V of analogue signal is converted into
1.5 mm diameter and 12 mm length were adhered to 216 levels (65 536 values), thus, a theoretical resolu-
both sides of the notch ensuring that they were tion of 0.1 mm is obtained for the extreme point of
perpendicular to the light beam. To reach the the range. For the tests, a smaller range of measure-
highest accuracy the tests were performed at 20 uC, ment of the micrometer, 50 mm, was selected to
with a separation between the pins of 1 mm. The detect the CMOD variations. In terms of analogue
repeat accuracy of the optical micrometer for this signal, this range corresponds to a scaling value of
85
5 mm/V and, therefore, with a theoretical resolution

of about 0.02 mm.
3 SPECIMEN PREPARATION
Fine-grained test samples of the spinel-structured

phase of magnesium aluminium oxide, labelled SP,
with a density of 3.491 0.002 g/cm3 and dynamic
Youngs modulus of 258 2 GPa, were prepared
from a commercial MgO?Al2O3 powder and were
sintered in air with a heating rate of 2 uC/min up to
1630 uC, a cooling rate of 2 uC/min and a dwell time
of 2 h [6].
Single-edge V-notch beams of 4 mm thickness (B),
6 mm width (W), and 50 mm length were diamond
machined. The notch was initially cut with a 300 mm Fig. 5 CMODtime plots for bending tests performed
wide diamond wheel. Using this pre-notch as a at different CMOD rates: L 5 0.05 mm/min,
guide, the remaining part of the notch was created M 5 0.5 mm/min, and H 5 1 mm/min
with a 150 mm wide razor blade sprinkled with
diamond pastes of successively 15, 6, and 1 mm grain the geometry of the specimens, and bending device
size. Tip radii of about 25 mm were obtained (Fig. 4). parameters (B 5 4 mm, W 5 6 mm, and S 5 40 mm),
The relative notch depth, a 5 a/W (a 5 notch length), and the analysis according to Bar-On et al. [7], stable
was 0.64. fracture tests could not be obtained under displace-
ment control for this material. However, stable
fracture tests have been performed for specimens
4 TESTS AND DATA ANALYSIS with a relative depth of about 0.64 using CMOD
control, as shown in Fig. 7. In the CMOD-controlled
Figure 5 shows the CMODtime plots corresponding tests performed in this work, continuously decreas-
to tests performed at different constant CMOD rates. ing loads after the maximum load were observed
The programmed CMOD rates with variations lower with increasing time, as shown in Fig. 7, which
than 0.02 mm/min were reached in all cases. The SP correspond to controlled fracture.
specimens were tested at 0.1 mm/min (Fig. 6). Figure 8 shows the loaddisplacement plot corres-
Using the compliance value (1.561027 m/N) de- ponding to the tests of Figs 6 and 7. From the
termined for maximum loads up to 150 N, the
Youngs modulus of the material (E 5 258 2 GPa),
Fig. 4 Final notch in a tested specimen. The black

colouration of the tip is a result of using Fig. 6 CMODtime plot for the selected testing con-
diamond paste. Tip radii of about 25 mm are ditions used to characterize the magnesium
obtained aluminium oxide test material
86
Controlled fracture test for brittle ceramics 31
of the relative notch depth (0 ( a ( 1) and span-to-

depth ratios (b 5 S /W) larger than 2.5 (2.5 ( b ( 16)
pffiffi
afp az4=bp4 a{p ag
Kb a~ 2
1{a3=2 1z3a
The p4(a) and p(a) given by equations (3) and (4)

are cubic polynomials for b 5 4 (equivalent to a
reference beam with fixed S/W 5 4) and b 5
(formally equivalent to pure bending), respectively.
p4 a~1:9z0:41az0:51a2 {0:17a3 3
p a~1:99z0:83a{0:31a2 z0:14a3 4
The onset of crack propagation was considered at

the peak load. The obtained p value
ffiffiffiffi of KIC for this test
Fig. 7 Stable fracture test for a SP test specimen. The sample was 1.04 0.06 MPa m, about one-third of
pffiffiffiffi
insert shows the continuous decrease in load the value determined in unstable tests (, 3.0 MPa m)
with increasing time which corresponds to [6].
controlled fracture The critical energy release rate, GIC, was calculated
from the KIC and Youngs modulus values using the
maximum load, decreases in the displacement were result of Irwin for plane strain conditions
needed to reach stable tests.
The critical stress intensity factor in mode I, KIC, KIC 2
GIC ~ 5
was calculated using the general expression of the E
stress intensity factor for three-point bend speci-
mens (equation (1)) [8] where E9 5 E/(1 2 v2) is the generalized Youngs
modulus for plane strain (E is Youngs modulus and
3SP v is Poissons ratio). The obtained value of GIC is
KIC ~ Kb a 1
2BW 3=2 3.83 0.004 J/m2 using 0.294 as Poissons ratio for SP
[10].
The S, P, B, and W parameters were defined in
previous sections and Kb (a) is a general shape
function (equation (2)), which is valid for any value 5 CONCLUSIONS
The following conclusions can be drawn from the

presented work.
1. It is possible to control an electromechanical

testing machine using an analogue output from
an optical micrometer.
2. Using the CMOD as a control parameter, stable
fracture tests can be performed for extremely
brittle ceramics such as fine-grained samples of
the spinel-structured phase of magnesium alu-
minium oxide.
3. The critical stress intensity factor in mode I
determined for stable fracture is lower than the
value measured for unstable tests. For the fine-
grained test sample measured in this work this
parameter is about one-third of that obtained
Fig. 8 Loaddisplacement plot for a SP test specimen using unstable testing conditions.
87
ACKNOWLEDGEMENTS alumina-aluminium titanate ceramics with second

phase nanoparticles at alumina grain boundaries.
The authors gratefully acknowledge Professor Dr J. Y. J. Eur. Ceram. Soc., 2008, 28, 19611971.
Pastor for helpful discussions and MICINN (Spain) 5 Pastor, J. Y., Planas, J., and Elices, M. Ambient and
for their financial support under grants MAT2006- high-temperature stable fracture tests in ceramics:
13480 C02 and MICINN and Microtest S.A MCI-PET applications to yttria-partially-stabilized zirconia.
J. Am. Ceram. Soc., 1993, 76(11), 29272929.
2008-0113.
6 Baudn, C., Martinez, R., and Pena, P. High-
temperature mechanical behavior of stoichio-
F Authors 2011 metric magnesium spinel. J. Am. Ceram. Soc.,
1995, 78(7), 18571862.
REFERENCES 7 Bar-On, I., Baratta, F. I., and Cho, K. Crack
stability and its effect on fracture toughness of
hot-pressed silicon nitride beam specimens. J. Am.
1 Srensen, B. F., Horsewell, A., Jrgensen, O.,
Ceram. Soc., 1996, 79(9), 23002308.
Kumar, A. N., and Engbk, P. Fracture resistance
measurement method for in situ observation of 8 Guinea, G. V., Pastor, J. Y., Planas, J., and Elices, M.
crack mechanisms. J. Am. Ceram. Soc., 1998, 81(3), Stress intensity factor, compliance and CMOD for
661669. a general three-point-bend beam. Int. J. Fracture,
2 Blackman, B. R. K., Kinloch, A. J., Rodriguez 1998, 89, 103116.
Sanchez, F. S., Teo, W. S., and Williams, J. G. The 9 Irwin, G. R., Fracture, pp. 551589 in Encyclopae-
fracture behaviour of structural adhesives under dia of Physics, 1958, Ed. S. Fluge, (Springer Verlag,
high rates of testing. Eng. Fracture Mech., 2009, 76, Berlin, Germany).
28682889. 10 Shackelford, J. F., Alexander, W., and Park, J. S.
3 Wieninger, H., Kromp, K., and Pabst, R. F. Crack (Eds) CRC materials science and engineering hand-
resistance curves of alumina and zirconia at room book, 1994, pp. 537539 (CRC Press, Boca Raton,
temperature. J. Mater. Sci., 1986, 21, 411418. Florida).
4 Bueno, S., Berger, M. H., Moreno, R., and
Baudn, C. Fracture behaviour of microcrack-free
88
Captulo 4
TENACIDAD DE FRACTURA DE CERMICAS FRGILES: ALMINA
Captulo 4
4. TENACIDAD DE FRACTURA DE CERMICAS FRGILES:
ALMINA
Una vez verificados el dispositivo experimental y el procedimiento de ensayo
desarrollados, se procedi a establecer una metodologa de ensayo y anlisis de
resultados para materiales frgiles. Para ello se seleccion como material modelo una
almina de grano fino que, como se ha descrito en el apartado 1.5, presenta fractura
frgil.
Los resultados derivados de esta investigacin dieron lugar a una publicacin en la
revista Journal of the European Ceramic Society (Anexo II).
Para el anlisis de los resultados experimentales se han propuesto los siguientes
parmetros:
- KIC: factor crtico de intensidad de tensiones en modo I (apartado 1.1), calculado
a partir de la carga mxima registrada en el ensayo (ecuacin 3.1).
- GIC: tasa crtica de liberacin de energa en modo I (apartado 1.1), calculada a
partir de KIC, el mdulo de Young (E) y el coeficiente de Poisson (), (ecuacin
1.3).
- : energa de fractura (apartado 1.1), calculada a partir del rea encerrada bajo la
curva carga-desplazamiento (A) y el rea de las nuevas superficies creadas,
ecuacin 4.1:
A
g = (4.1)
2 B(W - a )
donde B y W son el espesor y la anchura de la probeta, respectivamente, y a es
la longitud de la entalla (Figura 3.3).
Se han comparado tanto las curvas experimentales como los valores de los
parmetros de tenacidad calculados a partir de las curvas carga-desplazamiento
91
Tenacidad de fractura de cermicas frgiles: almina
obtenidas para el mismo material, con igual geometra de ensayo y caractersticas de la
entalla, en ensayos realizados en control de desplazamiento72.
4.1. Condiciones experimentales
Las propiedades del material ensayado72 se describieron en el captulo 1, tabla 1.2.
Se trata de una almina de grano fino (3.5 m) y alta densidad (98% terica). El
material en verde fue obtenido por colaje de una suspensin estable de almina en
moldes de escayola y sinterizado en aire con una velocidad de calentamiento y
enfriamiento de 2 C/min hasta 1450 C y un tiempo de permanencia a la mxima
temperatura de 2 h. En la rampa de calentamiento se incluy un tratamiento isotermo a
1200C durante 4 h.
Para la determinacin de la tenacidad de fractura se utiliz la geometra de ensayo
de cermicas tcnicas avanzadas (apartado 3.2.1). Las caractersticas de la entalla
fueron: =0.5 y ~15 m (figura 3.6).
Se realizaron ensayos a velocidad constante de apertura de los labios de la grieta
(CMOD) de 1.8 m/min.
4.2. Resultados y discusin
Segn el anlisis de estabilidad expuesto en el captulo 3, apartado 3.1.2, para este
material y dispositivo experimental (s=0.08; E~379 GPa; lm=232; C=1.5 10-7 m/N) no
se conseguiran ensayos estables en control por desplazamiento a menos que las grietas
fueran muy profundas, con un a0.75, superior al valor obtenido considerando la
flexibilidad de la mquina nula (Figura 4.1).
Se compararon los resultados obtenidos en este trabajo con aquellos reportados en
ensayos previos72, utilizando el mismo radio de curvatura del fondo de entalla y la
misma longitud de entalla relativa.
92
Captulo 4
En la figura 4.2 se muestran curvas experimentales caractersticas carga-
desplazamiento correspondientes a ensayos realizados en control por CMOD y por
desplazamiento72. Una vez alcanzada la carga mxima, la variacin de la carga con el
desplazamiento depende del tipo de control utilizado.
En la curva correspondiente al control por desplazamiento se observa una cada
brusca de la carga para el desplazamiento fijo, seguida de una disminucin montona
para desplazamientos crecientes. Este tipo de comportamiento es caracterstico de
fractura semiestable72.
En la curva correspondiente al control por CMOD se observa cmo la posicin del
marco de carga disminuye desde el valor de carga mxima hasta un desplazamiento del
marco de carga ~ 0.009 mm (rgimen I, apartado 1.3.3), a partir de dicho valor la
posicin del marco de carga vuelve a alcanzar valores crecientes de desplazamiento
(rgimen II, apartado 1.3.3.). Este tipo de comportamiento es caracterstico de fractura
estable.
0.14
s=0.08; lm=232
0.12
Carga adimensional
0.10
0.08
0.06
a=a/W=0.75
0.04
0.02
0.00
0 5 10 15 20 25 30 35 40 45 50

deflexin adimensionales para la almina (apartado 3.2.1). Clculos realizados para la
flexibilidad adimensional, lm, correspondiente a la flexibilidad del dispositivo utilizado,
(dm/P)=1.5 10-7 m/N. La lnea recta corresponde al tamao relativo de entalla crtico,
(c=0.75).
93
60
50 A1
A3
40
Carga [N] 30
20
10
0
0.000 0.005 0.010 0.015 0.020 0.025
Desplazamiento [mm]
Figura 4.2. Curvas carga-desplazamiento para las probetas ensayadas usando

diferentes parmetros de control.
(A1): control por apertura de los labios de la grieta, CMOD.
(A3): control por desplazamiento.
Los parmetros de tenacidad calculados a partir de las curvas carga-desplazamiento
se recogen en la tabla 4.1. Todos los parmetros determinados en ensayos estables
(CMOD) son inferiores a los determinados en ensayos semiestables (desplazamiento)72:
- KIC es un 10% inferior.
- GIC es un 18-22% inferior.
- es un 29-33% inferior.
Estos resultados se deben a que, como se discuti en el apartado 1.3.3, la energa
suministrada durante los ensayos estables se emplea exclusivamente en la formacin de
nuevas superficies.
94
Captulo 4
Tabla 4.1. Parmetros de tenacidad de fractura para el material de almina.
Bueno y col.72 Este trabajo

KIC (S.D.) GIC g KIC (S.D.) GIC (S.D.) g (S.D.)
1/2
(MPa m ) (J/m2) (J/m2) (MPa m1/2) (J/m2) (J/m2)
2.9;2.8 20.4;19.6 10.5;9.8 2.5 (0.2) 16.4 (2.3) 7.0 (0.3)
crtica de liberacin de energa en modo I (J/m2); : Energa de fractura (J/m2). S.D.:
Desviacin estndar.
4.3. Conclusiones
- Utilizando control por CMOD, se han obtenido ensayos de fractura estable para
un material de almina, en condiciones experimentales en las que los ensayos en control
de desplazamiento son semiestables.
- De acuerdo con la hiptesis bsica de este trabajo, los ensayos de fractura estable
son adecuados para la determinacin precisa de los parmetros de la tenacidad de este
material de almina.
- El valor de GIC obtenido a partir de la carga mxima es coincidente con el valor
correspondiente a dos veces la energa de fractura, como corresponde a un material
frgil.
95
96
ANEXO II
Este trabajo ha dado lugar a las siguientes publicaciones:
- A. Garcia-Prieto y C. Baudn, "Crack mouth opening displacement controlled

fracture tests of brittle ceramics", J. Eur. Ceram. Soc., 30 (16) 3297-3302
(2010). (DOI: 10.1016/j.jeurceramsoc.2010.08.007).

Materials Science, Ceramics.
- A. Garca-Prieto, C. Baudn. Tenacidad y trabajo de fractura de un material

frgil de almina. XXVIII Encuentro del Grupo Espaol de Fractura. Anales
de Mecnica de la Fractura, 28, Vol.1, pp. 125-130. Ed. Secretara del Grupo
Espaol de Fractura, Madrid, Espaa, 2011.
97
98
Available online at www.sciencedirect.com
Journal of the European Ceramic Society 30 (2010) 32973302
Crack mouth opening displacement controlled fracture

tests of brittle ceramics
A. Garca-Prieto, C. Baudn
Instituto de Cermica y Vidrio, CSIC, CSIC-Campus de Cantoblanco, Kelsen 5, 28049 Madrid, Spain
Received 23 June 2010; received in revised form 27 July 2010; accepted 2 August 2010
Abstract
Controlled fracture tests are required for the accurate determination of the toughness parameters of materials in order to assure the full conversion
of the supplied energy into crack surface energy. From the three parameters involved in the test, load, displacement of the load point and crack
mouth opening displacement (CMOD), this latter is the only one that continuously increases as fracture proceeds. Therefore, the CMOD has been
proposed as control variable for the stable fracture tests. In this work, a new equipment to perform stable fracture tests of single edge V-notch beams
(SEVNB) of ceramics in three points bending controlled by the CMOD is presented. The developed equipment allows performing stable fracture
tests of extremely brittle materials. The equipment is presented together with results obtained for fine grained aluminiummagnesium aluminate
and alumina ceramics.
2010 Elsevier Ltd. All rights reserved.
Keywords: Al2 O3 ; Spinel; Toughness; Stable fracture; Work of fracture
1. Introduction the deflection of the specimen directly measured at the central

point, in bending specimens, (deflection control) or the crack
It is well known that stable crack growth is necessary to get mouth opening displacement (CMOD control). The simplest
reliable and accurate fracture toughness data. When the fracture and therefore most widely used testing conditions are the con-
toughness values are determined from test configurations that trol of the specimen deformation by constant rates of increasing
do not allow stable crack growth the calculated toughness value displacement or load.
might be over-evaluated (e.g. for a dense a-SiC, KIC 4 and Controlled fracture tests for brittle materials, as most ceram-
3 MPa m1/2 for unstable and stable tests, respectively).1 More- ics, are difficult to accomplish, therefore, they are not usually
over, controlled fracture tests supply much more information performed. Since the initial works of Nakayama et al.2,3 and
about the fracture process than the fast fracture ones because, Tattersall and Tappin4 different authors have developed innova-
in addition to the conventional fracture toughness for crack ini- tive specimen designs and tests geometries (e.g. 510) to attain
tiation, controlled fracture allows the determination of fracture stable fracture of materials using displacement controlled load-
energy and crack-growth resistance curves. ing. Such approach has made it possible to reach stable fracture
Fracture toughness tests of ceramics are usually performed in for materials as brittle as glass2,11 and fine grained MgO4 using
universal testing machines by subjecting the specimens located hard machines. In general, the specimens required for these tests
between the loading supports to increasing deformation by are difficult to fabricate especially for brittle ceramics.
means of the displacement of the loading frame. The deforma- From the broad spectrum of available fracture tests, bend-
tion of the specimen can be controlled by imposing a constant ing of parallelepiped specimens with straight trough notches
rate to the increase of the displacement of the frame and, thus, to (SENB) is a relatively simple way of testing and displacement
the loading point (displacement control), the load (load control), or load controlled bending tests have been widely used for frac-
ture toughness testing of ceramics. Therefore, different attempts
to perform stable tests using three point bending loaded beams,
Corresponding author. which are generally more stable than the four point bending
E-mail address: cbaudin@icv.csic.es (C. Baudn). ones,12 have been done.
0955-2219/$ see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jeurceramsoc.2010.08.007
99
3298 A. Garca-Prieto, C. Baudn / Journal of the European Ceramic Society 30 (2010) 32973302
Fig. 1. General loadloading point displacement plot for beams of brittle mate-
rials with straight through notches (SENB). The curved part corresponds to the Fig. 2. Loadloading point displacement plot for stable fracture reached using
fracture taking place with constant energy release rate, G, equalling Gc . As constant displacement rate.
the straight line that represents the loading of the specimen hits the curve, the
condition for crack growth (G = Gc ) is satisfied. To keep G = Gc the load point
displacement has to decrease initially (regime I) and then increase (regime II).
size (d50 5.5 mm)16 and for fine grained aluminaaluminium
titanate composites (alumina: d50 3.23.9 mm, aluminium
The schematic representation of Fig. 1 allows discussing titanate: d50 2.2 mm).16 On the contrary, for extremely brit-
some aspects about the stable crack growth in SENB speci- tle materials, very deep cracks and extremely stiff machines
mens of brittle materials, as done by Sigl.13 The curved line would be needed for stable fracture and thus, it is not possi-
is the general stable fracture locus of a material with flat R curve ble in practice. In this sense, loaddisplacement curves showing
(Griffith locus, critical energy release rate, Gc = constant). This unstable fracture or sudden load decreases (pop-in) prior to
curve represents the fracture taking place with constant energy further stable propagation (i.e. semi-stable fracture) have been
release rate, G, equalling Gc . As the straight line that represents reported for fine grained alumina (d50 3.5 mm)16 and silicon
the loading of the specimen hits the curve, the condition for crack nitride (d50 3 mm)17 .
growth (G = Gc ) is satisfied. In order to get stable fracture, G has Contrary to the above-discussed parameters, the CMOD
to be maintained at its critical value and, thus, decreasing values increases through the whole fracture test, the loading of the
of the load point displacement (regime I) followed by increasing specimen as well as during the crack growth. Thus, CMOD has
values of this parameter (regime II) would be demanded. The been proposed and used as control parameter for stable frac-
loaddisplacement relationship of regime I is usually called snap ture testing of high-strength concrete14 and tetragonal zirconia
back.14 Therefore, it will not be possible to get stable fracture stabilised with 3 mol% of Y2 O3 1820 specimens under condi-
using constant displacement rates. The relative weight of the tions that would have led to unstable fracture for displacement
regions corresponding to regimes I and II depend on the mate- controlled tests.
rial properties, the specimen and span sizes, the notch depth and In order to perform in a routine way stable fracture tests
the stiffness of the testing device. For the same material and of ceramics, a new experimental setup to perform three point
testing geometry, stiff machines and deep notches increase the bending stable fracture tests of ceramics controlled by the
region of regime II and situations such as that plotted in Fig. 2 crack mouth opening displacement (CMOD) was developed.21
can occur. In this case, stable fracture can be reached controlling The set up combined an electromechanical universal testing
by constant displacement rate because increases in displacement machine with a high precision optical micrometer. An elec-
after the maximum load still allow to follow the stable condi- tromechanical machine was chosen because the extremely small
tion for crack growth G = Gc . In terms of stability, the use of the displacement variations required for the testing of brittle mate-
actual deflection of the specimen as control variable is qualita- rials can be reached by small turns of the motor in standard
tively the same as the use of displacement. From Figs. 1 and 2 it electromechanical machines whereas they would require spe-
is clear that the control by constant rates of increasing load can cial hydraulic machines with highly precise servo valves. The
never lead to stable tests because load always decreases after use of an optical micrometer permits the measurement of the
cracking starts. CMOD without contact with the specimen. Nowadays, the
Stable fracture for SENB specimens tested in three point high performance of the control systems avoids the necessity
bending using displacement control is relatively easy to attain for of using hydraulic machines to assure rapid responses of the
materials with R-curve fracture, for which the crack resistance load frame and allows the use of control variables external
(i.e. Gc ) increases as the crack propagates. For instance, stable to the machines such as the CMOD. Using this equipment it
fracture has been reported for materials with coarse microstruc- has been possible to test an extremely brittle ceramic such as
tures such as silicoaluminate and high alumina refractories3 fine grained magnesiumaluminium spinel for which a tough-
and graphite,11,15 for dense alumina with relatively large grain ness value (1 MPa m1/2 ) about 66% lower than the previously
100
A. Garca-Prieto, C. Baudn / Journal of the European Ceramic Society 30 (2010) 32973302 3299
CMOD controlled tests is described and compared to that previ-

ously reported for displacement controlled tests for which only
semi-stable fracture could be reached.16
2. Experimental
2.1. Material and specimen preparation
The fabrication and properties of the alumina material tested

are described elsewhere.16 Monophase alumina blocks were
obtained by colloidal filtration in plaster moulds of aqueous alu-
mina (Al2 O3 ) stable suspensions. Sintering of the green blocks
was performed in air in an electrical box furnace (Termiber,
Spain) at heating and cooling rates of 2 C min1 , with 4 h,
dwell at 1200 C during heating and 2 h, dwell at the maxi-
mum temperature 1450 C. The microstructural and mechanical
properties for this material are summarised in Table 1.
Single Edge V-Notch Beams (SEVNB) of 4 mm thickness
(B), 6 mm width (W) and 50 mm length were diamond machined
from the sintered blocks. The notch was initially cut with a
300 mm wide diamond wheel (Fig. 4a). Using this pre-notch as
a guide, the remaining part of the notch was done with a 150 mm
wide razor blade sprinkled with 1 mm diamond paste (Fig. 4b
and c). Tip radii of about 15 mm were obtained (Fig. 4c). The
relative notch depth, a/W (a = notch depth, W = specimen width),
was 0.5.
2.2. Mechanical testing set up
The mechanical tests were performed in a single screw, dual

column and servo-controlled electromechanical universal test-
ing machine with 50 kN load capacity and rigid frame (stiffness
2 108 N/m, Microtest EM1/50, Spain). The displacement of
Fig. 3. Plots corresponding to a stable fracture test for a fine grained the moving crosshead is measured and controlled by means of
magnesiumaluminium spinel.21 (a) Loadtime plot. Monotonous load decrease
with increasing times during fracture corresponding to controlled fracture is
an optical encoder placed in the motor axis. The electronic con-
highlighted in the detail of the fracture part of the plot. (b) Loaddisplacement troller (Microtest SCM3000, Spain) includes load and position
plot. During fracture the displacement has to decrease to reach stable channels as well as position auxiliary ones with the option to add
fracture. additional strain channels for extensometers. The signal from
the system for CMOD determination is directed to one of these
obtained in unstable tests (3 MPa m1/2 )22 was obtained. In auxiliary channels so CMOD can also be a control variable. The
Fig. 3 characteristic plots recorded during CMOD controlled test specimen is placed between the rigid frame and the moving
tests for this material are shown. The shape of the loadtime crosshead in a stainless steel three point bending test fixture with
curve with monotonous load decreases with increasing time dur- a span of 40 mm. A load cell of 5 kN was selected to assure high
ing fracture is characteristic of stable fracture (Fig. 3a). This stiffness of the loading setup. The compliance of the machine,
behaviour is attained using a constant rate of increasing CMOD. load cell, and supports arrangement was determined experimen-
The displacement has to decrease thorough the whole fracture tally using an uncracked alumina bar (4 mm 6 mm 50 mm);
process in order to maintain such rate (Fig. 3b). the obtained value was 1.5 107 m/N up to 150 N.
In this work, after a brief description of the experimental For the CMOD measurement and control, a high preci-
setup, the fracture behaviour of a fine grained alumina using sion optical micrometer Keyence LS7010 (Keyence, Japan) that
Table 1
Microstructural and mechanical parameters for the alumina material tested in this work, A-1450.16 GA = average grain size, = relative density, E = static Youngs
modulus, f = three point bending strength, KIC = critical stress intensity factor in mode I. S.D. = standard deviation.
Material GA (S.D.) (mm) (S.D.) (%theoretical) E (S.D.) (GPa) f (S.D.) (MPa) KIC (MPa m1/2 )
A-1450 3.5 (0.3) 98.1 (0.3) 379 (8) 456 (29) 2.9
2.8
101
work, the sampling frequency of 512 s1 used gave very stable

readings.
The principle of measurement of the optical system is as fol-
lows. A high-intensity GaN green LED (Light-Emitting Diode)
radiates light, which is changed into uniform parallel light
through the special diffusion unit and collimator lens and emitted
to the target in the measuring range. This parallel beam illumi-
nates measurement area. Then the shadow image of the target
appears on the HL-CCD (High-Speed Linear Charge Coupled
Device) through the telecentric optical system. With the tele-
centric system of lenses the size of the image on the CCD does
not change even if it moves, thus, the same accuracy all along
is maintained. The output incident signal of the HL-CCD is
processed by the DE (Digital Edge-detection) processor in the
controller and CPU. The controller of the optical system incor-
porates a function of elimination of abnormal values, to improve
the precision of the measurement, which detection threshold is
an adjustable parameter.
The optical micrometer is attached to the lower loading sup-
port; in this way the mechanical interferences are avoided and a
correct orientation of the light beam with respect to the axis of
load and the bending fixture is assured. Given the small opening
displacement of the notch, in order to be able to detect and mea-
sure its width during the test (the size of detectable minimum
object by the system is of 0.04 mm), pins of 1.5 mm in diameter
and 12 mm length are adhered to both sides of the notch assuring
that they are perpendicular to the light beam. To reach the high-
est accuracy the tests were performed at 20 C, with a separation
between the pins of 1 mm. The repeating accuracy of the opti-
cal micrometer for this separation was checked to be 0.06 mm
using a 1.0 mm diameter round bar located in the centre of the
measuring area.
The optical micrometer converts the distances detected to
analog signals that are input in the controlling unit of the
mechanical testing machine where the software SCM3000
(Microtest, Spain) converts them to discrete digital numbers. The
whole range of the optical micrometer (6 mm) that corresponds
to 10 V of analog signal is converted in 216 levels (65,536
values), thus, a theoretical resolution of 0.1 mm is obtained for
the maximum range. For the tests, a smaller range of measure-
ment of the micrometer, 50 mm, was selected to detect the
CMOD variations. In terms of analog signal, this range corre-
sponds with a scaling value of 5 mm/V and, therefore, with a
theoretical resolution of about 0.02 mm.
2.3. Testing conditions
Fig. 4. Optical micrographs of a lateral surface of an alumina specimen showing

The alumina specimens were tested in the above
the procedure to introduce the V notches. (a) A pre-notch is introduced with described experimental setup using CMOD control at rate of
a thin (300 mm) diamond disc. (b and c) The notch tip is corrected with a razor 1.8 mm min1 . This velocity was chosen in order to use a rate
blade with diamond past (1 mm) to reduce the tip radius below 15 mm. of the deformation of the specimen similar to that previously
used to test this material (0.005 mm min1 ). To determine this
incorporates a CMOS (Complementary Metal Oxide Semicon- velocity, several specimens were tested using different CMOD
ductor) camera to capture real-time image of the target was rates to establish the correspondence between the displacement
used. This optical system provides a measurement accuracy of and CMOD rates from the time and displacement values needed
0.5 mm. The equipment carries out a continuous measurement to reach the maximum loads. The programmed CMOD rate was
averaging up to 2400 samples/s. For the tests performed in this attained in all tests with variations of less than 0.02 mm min1 .
102
Fig. 6. Characteristic loadtime plots for alumina specimens tested using dif-
Fig. 5. Loadcrack mouth opening displacement (CMOD) plots recorded during
ferent control parameters for the deformation of the specimen. (1): Crack mouth
the CMOD controlled tests for two different alumina specimens showing the
opening displacement control; stable fracture is shown. (3) Displacement con-
repeatability of the tests.
trol; semi-stable fracture is shown.16
3. Results and discussion where S is the span, P is the maximum load, B and W are the geo-
metrical parameters defined in Section 2 and K () is a general
In which follows, the results obtained previously using the shape function which is valid for any value of the relative notch
same testing and specimen geometries and displacement con- depth (0 1) and span-to-depth ratios ( = S/W) larger than
trolled tests16 are compared to those obtained in this work using 2.5 (2.5 16) (Eq. (2)):
CMOD control. From calculations17 using the compliance value
of the whole testing system (1.5 107 m/N) and the proper- {p () + 4/[p4 () p ()]}
K () = (2)
ties of the material (Table 1), stable fracture tests could not be (1 )3/2 (1 + 3)
obtained under displacement control for any relative notch depth
value for this alumina material. In fact, only semi-stable fracture The p () and p4 () given by Eqs. (3) and (4) are cubic
was obtained for a limited number of tests of specimens with rel- polynomial for = 4 (equivalent to a reference beam with fixed
ative notch depths of 0.5 in the previously reported study. The S/W = 4) and = (formally equivalent to pure bending).
introduction of larger notches led to the failure of the specimens p () = 1.99 + 0.83 0.312 + 0.143 (3)
during machining.
As it is shown in the loadCMOD curves of Fig. 5, simi- p4 () = 1.9 + 0.41 + 0.512 0.173 (4)
lar results were obtained in this work for different specimens
The value of KIC obtained for CMOD controlled tests was
tested using the same CMOD rate and similar values of a/W the
2.5 0.2 MPa m1/2 , about 10% lower than the value determined
conditions which gave relatively low standard deviations for the
in semi-stable tests.16
fracture toughness parameters.
Fig. 6 shows characteristic loadtime plots recorded using
the two different control parameters. The sudden load decrease
for constant time prior to further monotonous load decrease
observed under displacement control is characteristic of semi-
stable fracture.16 On the contrary, in the CMOD controlled
tests monotonous load decreases with increasing times as
correspond to controlled fracture were always obtained. The
loaddisplacement curves corresponding to the tests of Fig. 6
are plotted in Fig. 7. As discussed in the introduction, a decrease
of displacement was needed to reach stable fracture after the
maximum load which could only be attained by using CMOD
control.
The critical stress intensity factor in mode I, KIC , was calcu-
lated using the general expression of the stress intensity23 and
the value of the maximum load attained during the test (Eq. (1)).
The onset of crack propagation was considered in the peak load.
Fig. 7. Loaddisplacement plots corresponding to the test of Fig. 6. (1) Crack
3SP mouth opening displacement control; a decrease of displacement after the max-
KIC = K () (1)
2BW 3/2 imum load is needed to reach stable fracture tests. (3) Displacement control.16
103
From KIC and Youngs modulus (Table 1), the critical energy References
release rate, GIC , was calculated according to the analysis of
Irwin for plane strain conditions (Eq. (5)): 1. Ghosh A, Jenkins MG, White KW, Kobayashi AS, Bradt RC. Elevated-
temperature fracture resistance of a sintered a-silicon carbide. J Am Ceram
2
KIC Soc 1989;72(2):2427.
GIC = (5) 2. Nakayama J. Direct measurement of fracture energies of brittle heteroge-
E neous materials. J Am Ceram Soc 1965;48(11):5837.
where E = E/(1 2 ) is the generalized Youngs modulus for 3. Nakayama J, Abe H, Bradt RC. Crack stability in the work-of-fracture test:
plane strain (E is the Youngs modulus and v is the Pois- refractory applications. J Am Ceram Soc 1981;64(11):6715.
4. Tattersall HG, Tappin G. The work of fracture and its measurement in
sons ratio). The Poissons ratio for dense and fine grained metals, ceramics and other materials. J Mater Sci 1966;1:296301.
alumina is 0.223 0.004.24 The value of GIC obtained was 5. Calomino AM, Brewer DN. Controlled crack growth specimen for brittle
16.4 2.3 J m2 that is about 20% lower than the value pre- systems. J Am Ceram Soc 1992;75(1):2068.
viously obtained in semi-stable tests.16 6. Srensen BF, Horsewell A, Jrgensen O, Kumar AN, Engbk P. Fracture
The work of fracture, WOF , was calculated by dividing resistance measurement method for in situ observation of crack mecha-
nisms. J Am Ceram Soc 1998;81(3):6619.
the work done on the specimen to propagate the crack, cal- 7. Wan D, Bao Y, Peng J, Zhou Y. Fracture toughness determination of
culated as the integral of the loaddisplacement plot, by the Ti3 Si(Al)C2 and Al2 O3 using a single gradient notched beam (SGNB)
area of the newly created surfaces (Eq. (6)). For parallelepiped method. J Eur Ceram Soc 2009;29:76371.
bars with straight trough notches tested in flexure, this area is 8. Kuszyk JA, Bradt RC. Influence of grain size on effects of thermal expan-
twice the area of the unnotched part of the cross-section of the sion anisotropy in MgTi2 O5 . J Am Ceram Soc 1973;56(8):4203.
9. Srensen BF, Brethe P, Skov-Hansen P. Controlled crack growth in ceram-
specimens. ics: the DCB specimen loaded with pure moments. J Eur Ceram Soc
A 1996;16(9):10215.
WOF = (6) 10. Ebrahimi ME, Chevalier J, Fantozzi G. R-curve evaluation and bridging
2B (W a) stress determination in alumina by compliance analysis. J Eur Ceram Soc
2003;23(6):9439.
where A is the area under the loaddisplacement curves and B, 11. Sakai M, Urashima K, Inagaki M. Energy principle of elasticplastic frac-
W and a were defined in Section 2. ture and its application to the fracture mechanics of a polycrystalline
The value of WOF obtained was 7.0 0.3 J m2 that is graphite. J Am Ceram Soc 1983;66(12):86874.
about 30% lower than the value obtained in semi-stable 12. Baratta FI, William AD. Crack stability in simply supported four-point and
three-point loaded beams of brittle materials. Mech Mater 1990;10:14959.
tests.16 13. Sigl LS. On the stability of cracks in flexure specimens. Int J Fract
1991;51:24154.
14. Biolzi L, Cangiano S, Tognon G, Carpinteri A. Snap-back softening insta-
4. Conclusions bility in high-strength concrete beams. Mater Struct 1989;22:42936.
15. Davidge RW, Tappin G. The effective surface energy of brittle materials. J
An experimental setup to perform stable fracture tests in an Mater Sci 1968;3:16573.
electromechanical machine using the analog output from an opti- 16. Bueno S, Berger MH, Moreno R, Baudn C. Fracture behaviour
of microcrack free aluminaaluminium titanate ceramics with second
cal micrometer was developed. This set up allows using the crack
phase nanoparticles at alumina grain boundaries. J Eur Ceram Soc
mouth opening displacement (CMOD) as a control parameter for 2008;28:196171.
fracture toughness testing. 17. Bar-On I, Baratta FI, Cho K. Crack stability and its effect on fracture
Stable fracture tests can be performed for brittle ceramics by toughness of hot-pressed silicon nitride beam specimens. J Am Ceram Soc
using the CMOD as control parameter and three points bending 1996;79(9):23008.
18. Pastor JY, Planas J, Elices M. Ambient and high-temperature stable fracture
of Single Edge V-Notch Beams as testing geometry.
tests in ceramics: applications to yttria-partially-stabilized zirconia. J Am
Stable fracture tests for fine grained alumina per- Ceram Soc 1993;76(11):29279.
formed using this device have given fracture toughness 19. Pastor JY, Planas J, Elices M. A new technique for fracture charac-
parameters (KIC = 2.5 0.2 MPa m1/2 , GIC = 16.4 2.3 J m2 , terization of ceramics at room and at high temperature. J Test Eval
WOF = 7.0 0.3 J m2 ) lower than those determined in semi- 1995;23(3):20916.
20. Pastor JY, Planas J, Elices M. Ensayos de fractura estables en materiales
stable tests: about 10%, 20% and 30% for critical stress intensity
cermicos. Bol Soc Esp Ceram V 1992;31(4):3225.
factor in mode I, critical energy release rate and work of fracture, 21. Baudn C, Garca A, Hernndez J, Lpez M. Anales de Mecnica de la
respectively. Fractura. In: Proceedings of the Conferencia Ibrica de Fractura e Integri-
dad Estructural 2010, vol. 1. Controlled fracture tests of brittle ceramics,
Secretara del Grupo Espanol de Fractura, Madrid, Spain; 2010. p. 2915.
Acknowledgements 22. Baudn C, Martnez R, Pena P. High-temperature mechanical behavior of
stoichiometric magnesium spinel. J Am Ceram Soc 1995;78(7):185762.
This work has been supported by MCI-MAT2009-14448-C02 23. Guinea GV, Pastor JY, Planas J, Elices M. Stress intensity factor, com-
pliance and CMOD for a general three-point-bend beam. Int J Fract
and PET2008-0113 and Microtest S.A. (Spain). Discussions 1998;89:10316.
with Prof. J.Y. Pastor from ETSI Caminos, Canales y Puertos 24. Burgos-Montes O, Moreno R, Baudn C. Effect of mullite additions on the
(Madrid) are gratefully acknowledged. fracture mode of alumina. J Eur Ceram Soc 2010;30:85763.
104
Captulo 5
EFECTO DE LAS CONDICIONES EXPERIMENTALES EN LA TENACIDAD
DE FRACTURA DE UN MATERIAL DE MULLITA
Captulo 5
5. EFECTO DE LAS CONDICIONES EXPERIMENTALES EN
LA TENACIDAD DE FRACTURA DE UN MATERIAL DE
MULLITA
Una vez demostrada la validez del dispositivo experimental y del procedimiento de
ensayo (captulo 3) y de la metodologa de ensayo y anlisis de resultados (captulo 4)
se ha estudiado la influencia de las variables experimentales, velocidad de
desplazamiento y radio de curvatura del fondo de entalla, en los parmetros de
tenacidad. Para ello se seleccion un material de mullita (Al2O3:SiO2, 3:2) como caso de
estudio porque presenta comportamiento frgil y crecimiento subcrtico de grietas
(SCG) significativo en aire, (apartado 1.6).
Los resultados derivados de esta investigacin dieron lugar a una publicacin en la
revista Journal of the European Ceramic Society (Anexo III).
Las propiedades del material ensayado148 (MB0) se recogen en la tabla 1.3. Se trata
de una mullita de grano fino (0.7m) y alta densidad (95% terica) obtenida a partir de
una mullita comercial Baikowski en polvo y sinterizada en aire a 1630 C durante 4
h148.
de cermicas tcnicas avanzadas (apartado 3.2.1). Las caractersticas de la entalla fueron
=0.64 y ~ 60 y 6 m.
Se realizaron ensayos a velocidades constantes de apertura de los labios de la grieta
(CMOD) de 1 y 18 m/min y a una velocidad constante de desplazamiento del marco de
carga de 0.05 mm/min. Se eligieron estas velocidades porque dan lugar a velocidades de
deformacin de la probeta similares a las utilizadas en ensayos previos para estudiar el
107
Efecto de las condiciones experimentales en la tenacidad de fractura de un material de mullita
efecto del crecimiento subcrtico de grietas para este mismo material en control por
desplazamiento, 0.005 y 0.05 mm/min148.
Segn el anlisis de estabilidad expuesto en el captulo 3, apartado 3.1.2, para este
material y dispositivo experimental (s=0.11; E=195 GPa; m=117; C=1.5 10-7 m/N) no
se conseguiran ensayos estables en control por desplazamiento a menos que las grietas
fueran muy profundas con un 0.7 (Figura 5.1), superior al valor obtenido
considerando la flexibilidad de la mquina nula (figura 3.10). Hasta la fecha de
realizacin de este trabajo no haban sido reportados parmetros de tenacidad
determinados en ensayos de fractura estable para materiales avanzados de mullita.
s =0.11
0.25
0.20
Carga adimensional
0.15
0.10
a/b =0.7
0.05
0.00
0 20 40 60 80 100 120 140 160

deflexin adimensionales para la mullita (apartado 3.2.1). Clculos realizados para la
flexibilidad adimensional, lm, correspondiente a la flexibilidad del dispositivo utilizado,
(dm/P)=1.5 10-7 m/N. La lnea recta corresponde al tamao relativo de entalla crtico,
(c=0.70).
Todos los ensayos realizados en control por CMOD dieron lugar a fractura estable.
En la figura 5.2 se muestran las curvas experimentales caractersticas carga-tiempo para
108
Captulo 5
ensayos realizados a alta velocidad del desplazamiento del marco de carga en control
por CMOD y desplazamiento. Al igual que ocurra en el caso de la almina (figura 4.2)
no se obtienen ensayos estables en control por desplazamiento, en cambio en control
por CMOD s se alcanzan ensayos estables a pesar de la alta velocidad empleada.
16
14 M1
M2
12
Carga [N]
10
0
0 10 20 30 40 50
Tiempo [s]
Figura 5.2. Curvas experimentales caractersticas carga-tiempo para el material de

mullita.
(M1): control por apertura de los labios de la grieta, CMOD.
(M2): control por desplazamiento.
En la figura 5.3 se muestra una curva experimental caracterstica carga-
desplazamiento en control por CMOD del material ensayado de mullita donde se
aprecia, a partir del valor de carga mxima, una disminucin del desplazamiento
(rgimen I) seguida de un aumento del desplazamiento (rgimen II). Este tipo de
comportamiento es caracterstico de fractura estable de materiales frgiles.
Las propiedades mecnicas de este material se recogen en la tabla 5.1. Los valores
de KIC, GIC y g fueron calculados a partir de tres medidas y los errores corresponden a
las desviaciones estndar. El coeficiente de Poisson empleado para esta mullita fue
0.274139.
109
14
12
10
Carga [N] 8
0
0.000 0.002 0.004 0.006 0.008 0.010
Desplazamiento [mm]
Figura 5.3. Curva experimental caracterstica carga-desplazamiento para el material

de mullita ensayado a una velocidad constante de CMOD de 1 m/min y un radio de
curvatura de fondo de entalla igual a 60m.
Tabla 5.1. Parmetros de tenacidad de fractura para el material de mullita.
Osendi y Baudn.148 Este trabajo

Condiciones de KIC (S.D.) Condiciones tf (S.D.)KIC (S.D.) GIC (S.D.) g (S.D.)
ensayo (MPa .m1/2) de ensayo (s) (MPa .m1/2) (J / m2) (J / m2)
v=0.05mm/min 2.32 v=0.018mm/min 0.86 3.5 1.9
30
=200mm (0.08) =6mm (0.06) (0.5) (0.1)
v=0.001mm/min 1.11 5.9 0.9
500
=60mm (0.06) (0.7) (0.2)
v=0.005mm/min 2.0 v=0.001mm/min 0.72 2.5 0.65
500
=200mm (0.3) =6mm (0.07) (0.5) (0.05)
v=0.05mm/min 1.17
- - -
=6mm (0.09)
crtica de liberacin de energa en modo I (J/m2); : Energa de fractura (J/m2); v:
velocidad (mm/min); r: radio de curvatura del fondo de entalla (m); tf: tiempo de
fractura (s); S.D.: Desviacin estndar.
110
Captulo 5
El valor de la tenacidad correspondiente al inicio de fractura (KIC) es muy
dependiente de las caractersticas de la entalla. Se han obtenido valores
significativamente diferentes de KIC calculados a partir de la carga mxima, siendo este
valor un 60% superior para = 60 m con respecto a los obtenidos para = 6 m. Esta
diferencia no se observa en la energa de fractura (), parmetro referente a todo el
proceso de fractura.
A la mxima velocidad de ensayo (v= 18 m/min) los valores de GIC y 2g son
coincidentes. En cambio, para la velocidad lenta (v = 1 m/min) GIC es muy superior a
2g. Este hecho puede explicarse por el fenmeno de crecimiento subcrtico de grieta
(SCG), que para este material es un proceso relativamente rpido, producindose a una
velocidad de 6.10-3 m/s. Esta velocidad es superior a la reportada por Torrecillas y
col.142 para un material de mullita sinterizado a 1700C, con velocidades de crecimiento
de ~10-4 m/s para KIC ~1.9-2.05 MPa m1/2 (apartado 1.6). Este fenmeno no es
detectable para tiempos cortos de ensayo, como corresponde a los tiempos involucrados
en los ensayos realizados a alta velocidad (18 m/min, 30s, figura 5.4 1) o los
necesarios para alcanzar la carga mxima en ensayos lentos (1 mm/min, 100s, figura 5.4
2). El fenmeno de SCG tampoco haba podido ser identificado en este material a partir
de los valores de KIC obtenidos en ensayos inestables. Sin embrago, este fenmeno es
observable para los tiempos correspondientes al crecimiento estable de la grieta (1
mm/min, 400s, figura 5.4 2).
111
12
11
1
10
9
8
Carga [N] 7 2
6
5
4
3
2
1
0
0 100 200 300 400 500
Tiempo [s]
Figura 5.4. Curvas experimentales caractersticas carga-tiempo correspondientes a
ensayos de fractura en control por CMOD para el material de mullita.
1: velocidad constante de CMOD=18 m/min.
2: velocidad constante de CMOD=1 m/min.
5.3. Conclusiones
- Se han alcanzado ensayos de fractura estable para un material de mullita de grano
fino con comportamiento frgil y crecimiento subcrtico de grieta.
- Para este material frgil y que presenta crecimiento subcrtico de grieta
significativo en aire, slo es posible obtener valores reales de los parmetros de
tenacidad en ensayos estables realizados utilizando velocidades altas de apertura de los
labios de la grieta (CMOD, v= 18 m/min).
112
ANEXO III
Este trabajo ha dado lugar a las siguientes publicaciones:
- A. Garca-Prieto, C. Baudn. Influence of experimental variables on fracture

toughness determined on SEVNB in three points bending. Mullite a case study,
J. Eur. Ceram. Soc. 32 (16) 4241-4248 (2012).
(DOI : 10.1016/j.jeurceramsoc.2012.06.018).

Materials Science, Ceramics.
- A. Garca-Prieto, C. Baudn. Efecto de las condiciones experimentales en la

tenacidad de fractura de un material de mullita. XXIX Encuentro del Grupo
Espaol de Fractura. Anales de Mecnica de la Fractura, 29, Vol.2, pp. 617-622.
Ed. Secretara del Grupo Espaol de Fractura, Madrid, Espaa, 2012.
113
114
Influence of experimental variables on fracture toughness determined on

SEVNB in three points bending. Mullite a case study
A. Garca-Prieto, C. Baudn
Instituto de Cermica y Vidrio, CSIC, CSIC-Campus de Cantoblanco, Kelsen 5, 28049 Madrid, Spain
Received 24 April 2012; received in revised form 15 June 2012; accepted 24 June 2012
Available online 20 July 2012
Abstract
The effect of testing variables on toughness of single edge V notched beams (4 mm 6 mm 50 mm, = 0.6) of a fine grained mullite
(d50 = 0.7 0.5 mm) in three points bending (span = 40 mm) is analysed. Mullite was selected as case material because it presents flat R-curve
and subcritical crack propagation. Stable fracture was reached by using the CMOD as control variable (0.001 and 0.018 mm/min). Results for
stable test and unstable displacement (0.05 mm/min) controlled tests are analysed. KIC has been calculated from maximum loads, KICp , and from
the total fracture energy determined in stable tests, KICg . The fact that for materials with flat R-curve both KIC values are coincident has been used
as criterion for adequacy of the test. Stable fracture at high deformation rates is required to fulfil KICp = KICg . Under such conditions, an intrinsic
KIC = 0.86 0.06 MPa m1/2 , less than one half of those previously reported has been obtained.
Keywords: Toughness; Mechanical testing; Crack propagation; Mullite
1. Introduction point (displacement control). Alternatively, other parameters can

be used to indirectly govern the rate of displacement of the frame.
Stable crack growth is necessary to determine reliable and These parameters can be load (load control) or any parameter
accurate fracture toughness data. When the fracture toughness related to crack growth such as the distance between both sides of
values are determined from mechanical testing configurations the notch, usually known as crack mouth opening displacement
that lead to unstable fracture the calculated values might be (CMOD).
over-estimated. For instance, for a dense a-SiC, KIC 4 and Up to now, stable fracture tests for ceramics have been con-
3 MPa m1/2 have been reported for unstable and stable tests, sidered as extremely difficult to realise, thus, they are not usually
respectively.1 Moreover, stable fracture supplies much more performed. Since the initial works of Nakayama et al.2 and Tat-
information about the fracture process than fast fracture because, tersall and Tappin,3 different authors have developed innovative
in addition to the conventional fracture toughness for notch specimen designs and test geometries (e.g. 49) to attain stable
initiation, calculated from the maximum attained load, stable fracture using displacement controlled deformation. The unsta-
fracture allows the determination of fracture energy and crack- ble crack growth is avoided or limited by careful adaptation of
growth resistance curves. the testing machine and the specimen geometry to the material
In fracture toughness testing a specimen with a notch with to be tested. Such approach has made it possible to reach sta-
known geometry is subjected to deformation until failure occurs. ble fracture for materials as brittle as glass10 and fine grained
Controlled increasing deformation is imposed by the displace- MgO.3 Two main problems arise in this kind of tests. On the one
ment of the loading frame. The simplest and therefore most hand, it is difficult to produce notches with well-defined special
widely used means for controlling the test is imposing a constant geometries in ceramic specimens. On the other, unstable crack
rate to the displacement of the frame and, thus, to the loading growth can result as the amount of energy necessary to create the
fracture surfaces is usually small compared to the elastic energy
stored in the test rig unless extremely stiff machines are used.
Corresponding author. In our laboratory a special set up for the routine stable frac-
E-mail address: cbaudin@icv.csic.es (C. Baudn). ture testing of single edge notched beams (SENB) of brittle
0955-2219/$ see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jeurceramsoc.2012.06.018
115
ceramics in three points bending using the CMOD as control

parameter has been developed.11,12 The objective of this work
is to determine the effect of testing variables on the KIC values
calculated from the maximum loads, KICp , determined in stable
and unstable tests, and from the total fracture energy determined
in stable tests, KICg . Traditionally, the total fracture energy in
b < c
Load
ceramics has been named work of fracture, however, the more
standard term fracture energy will be used in this paper. Mullite
has been selected as case material because it is a well known
brittle ceramic used in structural applications which presents
subcritical crack propagation at room temperature. As criterion c
for adequacy of the testing conditions the fact that for materi-
als with flat R-curve both KIC values, KICp and KICg , have to a > c
be coincident has been used. Additionally, this coincidence will Displacement

demonstrate that the time-dependent effects have been avoided.
Fig. 1. General plot load (P)displacement (d) of the loading point for fracture of
1.1. Stability analysis for fracture of SENB in three points SENB brittle specimens in three points bending. The straight lines correspond to
the loading of specimens with different relative notch depths = a/W (a = notch
bending
depth, W = specimen width). The curved part corresponds to the Griffith locus,
i.e.: fracture taking place with constant energy release rate, G, equalling Gc . For
From the broad spectrum of available fracture toughness a critical relative notch depth, c , the derivative, dP/dd, changes from positive to
tests, bending of parallelepiped specimens with straight trough negative. For a > c loading intersects the lower branch of the stability curve,
notches (SENB) is a relatively simple way of testing and dis- thus, the Griffith locus can be followed using constant displacement rates. For
b < c loading intersects the upper branch of the stability curve and any further
placement or load controlled tests have been widely used for
increase in displacement would result in unstable fracture; to keep G = Gc the
ceramics. Different attempts to reach stable fracture under such displacement has to decrease initially and then increase.
configuration have been done.1316
The well known criteria for assessing the stability of cracks
are based on an energy balance (Griffith criterion)17 : the external Therefore, it is not possible to reach stable fracture by using
work performed during crack growth dW must be greater than constant load rates to control the movement of the frame.
(or at least equal to) the associated change in the elastic energy For a critical relative notch depth, c , the slope of the Griffith
dUe stored in the entire testing device (test rig + specimen) and locus (dP/dd) changes from positive to negative; c is calculated
the energy RdA necessary to create the new fracture surface dA: using the condition dP/dd = 0. For specimens containing relative
dW dUe + RdA (R: crack resistance). The work performed and notch depths (a ) larger than the critical one (c ) the straight
the changes in the elastic energy are usually combined in one new line intersects the lower branch of the stability curve, regime
variable, the energy release rate G: G = (dW dUe )/dA. Crack II, thus, the Griffith locus can be followed using constant rates
growth occurs when G R and does not occur when G < R. Sta- of displacement of the loading point. For specimens containing
ble crack growth is then possible when, on the one hand, the relative notch depths (b ) smaller than c the load/displacement
Griffith criterion is met (G R) and, on the other hand, the straight line intersects the upper branch of the stability curve
increase in the energy released during crack growth is lower (regime I) and any further increase in displacement would result
than the increase in the energy necessary for the creation of new in unstable fracture. In regime I, decreasing values of the load
fracture surfaces (dG/dA < dR/dA). point displacement for decreasing loads (snap-back) are needed
The schematic representation of Fig. 1 allows discussing for stable fracture, therefore, it is not possible to get stable frac-
some aspects about the stable crack growth in SENB brittle spec- ture using constant displacement rates. In terms of stability, the
imens in three points bending, as done by Sigl.18 The curved line use of the actual deflection of the specimen (associated to the
is the general stable fracture locus of a material with flat R-curve load point displacement) as control parameter is qualitatively
(critical energy release rate, Gc = constant), called Griffith locus. the same as the use of displacement of the load frame.
This curve represents the fracture taking place with constant The relative weight of the regions corresponding to regimes I
energy release rate, G, equalling Gc . Initially the load increases and II depend on the material properties, the specimen and span
linearly with specimen deflection (in three points bending the sizes, the notch depth and the stiffness of the testing device. For
deflection of the specimen at the central point is associated to the same material and testing geometry, stiff machines and deep
the displacement of the load point), the proportionally constant notches increase the region of regime II.
depending on the relative notch depth ( = a/W, a = notch depth, Stable fracture using displacement control is relatively easy to
W = specimen width) and material properties. In Fig. 1, the linear attain for materials presenting R-curve, for which the crack resis-
relationships that represent the loading of identical specimens tance (R) increases as the crack propagates. For instance, stable
with different are also plotted. As the straight lines hint the fracture has been reported for materials with coarse microstruc-
Griffith locus the condition for crack growth (G = Gc ) is sat- tures such as silicoaluminate and high alumina refractories10
isfied. In order to get stable fracture, G has to be maintained and graphite,13,15 for dense alumina with relatively large grain
at its critical value thus, decreasing loads are always needed. size (d50 5.5 mm16 ) and for fine grained aluminaaluminium
116
Table 1
Microstructural and mechanical properties for the mullite material tested in this work.25,30,32 GA : average grain size; : relative density; E: Youngs modulus; :
Poissons ratio; f : 3-points bending strength; KIC : critical stress intensity factor in mode I. (S.D.): standard deviation.
GA (S.D.) mm (S.D.) % theoretical E (S.D.) GPa (S.D.) f (S.D.) MPa KIC (S.D.) MPa m1/2
0.7 (0.5) 95.0 (0.3) 195 (4) 0.274 (0.006) 227 (24) 2.0 (0.6)
titanate composites (alumina: d50 3.23.9 mm, aluminium particular, for fracture testing of SENB in three points bending
titanate: d50 2.2 mm16 ). On the contrary, for extremely brittle (beams: 4 mm 6 mm 50 mm, span: 40 mm, width notches
materials stable fracture is not possible in practice because very 200 mm, relative notch depth: 0.5), 2.02.4 MPa m1/2 for load
deep notches and stiff machines would be needed, as it was dis- displacement rates between 0.005 and 5 mm/min have been
cussed before. In this sense, loaddisplacement curves showing obtained.24,28,29,32 Up to now no definitive value for intrinsic
unstable fracture or sudden load decreases (pop-in) prior to KIC of fine grained mullite has been provided.
further stable propagation (i.e.: semi-stable fracture) have been In this work, the stable fracture behaviour of a fine grained
reported for fine grained alumina (d50 3.5 mm16 ) and silicon mullite tested using CMOD as control variable at rates 0.001
nitride (d50 3 mm14 ). and 0.018 mm/min and notch tip radii of 6 mm and 60 mm is
Contrary to displacement and load, parameters related to presented and compared to that corresponding to displacement
crack growth increase during the whole fracture test, thus, they controlled tests performed at a rate (0.05 mm/min) equivalent to
are suitable to reach stable fracture. In particular, the crack mouth the largest CMOD one.
opening displacement (CMOD) has been proposed and used as
control parameter for stable fracture testing of high-strength
concrete19 and tetragonal zirconia stabilised with 3 mol% of 2. Experimental procedures
Y2 O3 2022 specimens under conditions that would have led to
unstable fracture for displacement controlled tests; both materi- 2.1. Material and specimen preparation
als present R-curve behaviour. In our laboratory a special set up
for the routine stable fracture testing of brittle ceramics using Mullite specimens have been fabricated from high purity
CMOD as control parameter has been developed; stable fracture (99.99%) powders (193CR, Baikowski Chimie, France) fol-
of materials such as fine grained magnesiumaluminium spinel lowing the procedure described elsewhere.30 Starting powders
and alumina has been attained.11,12 The fracture toughness val- (1.8 mm) were attrition milled with mullite balls down to 0.8 mm,
ues obtained are lower than those previously obtained from the isostatically pressed (200 MPa) and sintered at 1630 C for 4 h.
maximum load values in unstable tests using displacement con- The microstructural and mechanical properties for this mullite
trol. are summarised in Table 1.
Single edge V notched beams (SEVNB) of 4 mm thick-
ness (B), 6 mm width (W) and 50 mm length (S) were diamond
1.2. Mullite machined from the sintered blocks (10 mm 40 mm 65 mm).
Notches were initially machined using a 300 mm wide diamond
Mullite presents some attractive properties for structural wheel. Using the pre-notch as a guide, the remaining part of the
applications such as high melting point (1828 10 C),23 low notch was done with a 150 mm wide razor blade sprinkled with
thermal conductivity (6 kcal m1 h1 C1 )24 and low thermal 15, 6 and 1 mm diamond paste. Tip radii of about 60 and 6 mm
expansion coefficient (25800 C = (4.1 0.1) 106 C1 ).25 respectively, and relative notch depths, 0.6, were obtained
Initially used only for high temperature traditional appli- (Fig. 2a and b). Fig. 2c shows a characteristic fracture surface;
cations, mullite materials experienced a great development straight pre-notch and final notch ( 0.6) tips are observed.
since the seminal work of Mazdiyasni26 where translucent
polycrystalline bodies of stoichiometric mullite with relatively
high strength and excellent thermal-shock resistance were 2.2. Testing conditions
described.23,26,27 The availability of fine pure mullite powders
and new processing routes has made it possible to obtain dense The development and complete description of the exper-
polycrystalline mullites with little or nearly no residual glassy imental setup used is described elsewhere.11,12 It combines
phase and with higher deformation resistance at higher temper- an electromechanical machine with a high-precision optical
atures than any ceramic oxide.24,2729 micrometre. The compliance of the test rig, C (machine, load
Mullite ceramics present flat R-curve and can expe- cell and support arrangement), is 1.5 107 m/N up to 150 N
rience subcritical crack propagation during fracture at and the resolution for the whole system for CMOD recording
room temperature.3032 Different researchers have pro- and control (micrometre + testing device) is 0.02 mm.
vided fracture toughness values obtained using different The CMOD was used as control parameter (rates = 0.001
unstable fracture test configurations and relatively large and 0.018 mm/min) for mullite beams with tip notch radii of
notch tip radii (200300 mm) for mullites with various 60 mm (range from 63 to 56 mm) and 6 mm (range from 10 to
microstructures.24,28,29,32 Reported values of the critical stress 3 mm). Additional displacement controlled tests were performed
intensity factor in mode I, KIC , are around 2.02.8 MPa m1/2 . In using a rate of 0.05 mm/min. This rate was determined from the
117
1.6
1.4
Dimensionless load
1.2
1.0
c = 0.4
0.8
0.6
0.4
s =0.9
0.8
0.7
0.2 0.5 0.6
0.3 0.4
0.1 0.2
0.0
0 20 40 60 80 100 120
Dimensionless displacement
Fig. 3. Griffith locus corresponding to the numerical values of dimensionless

load (PL/ f BW2 ) and dimensionless displacement (dL/f W2 ) for the test geom-
etry of this work (span, L = 40 mm; width, W = 6 mm and thickness, B = 4 mm).
Calculations have been done for a test rig compliance, C = 0, and for speci-
mens with different brittleness numbers, s = KI / f W1/2 . The critical relative notch
depth, c , is the same for all s.
displacement (dL/f W2 ) calculated using the test geometry of

this work (span, L = 40 mm; width, W = 6 mm and thickness,
B = 4 mm) is plotted for specimens with different brittleness
numbers (s = KI / f W1/2 ). Calculations have been done for a
totally stiff machine (compliance, C = 0). The critical relative
notch depth that satisfies dP/dd = 0 and, thus, limits the regimes
I and II of crack growth is c = 0.4 for all values of the brittleness
number.
As described in the introduction, for real test rigs with com-
pliance different from zero, the relative weights of regimes I
and II changes. In these cases, calculations have to take into
account the compliance of the test rig and the characteristics
of the material being tested, through the brittleness number.
For the compliance of the experimental set up and the testing
geometry used in this work and the properties of the material
summarised in Table 1, s = 0.05. The corresponding Griffith
s = 0.05
0.04
Fig. 2. Characteristic notch tips used for testing. (a, b) Optical microscopy
micrographs of as machined lateral surfaces. (c) Scanning electron micro-
graph of fracture surface. (a) Tip radius 60 mm machined with a diamond disc.
Dimensionless load
(b) Tip radius 6 mm machined from notches such as (a) and finished with a
razor blade wrinkled with diamond paste. (c) Low magnification micrograph
showing the process of machining of V notch and initiation of fracture. (1) Pre-
notch done with the diamond disc. (2) Zone machined with the razor blade. The 0.02
end of the notch is straight. (3) Initiation of fracture.
c = 0.7
displacement and times needed to attain the maximum loads in

the CMOD controlled tests to get equivalent deformation rates.
0.00
0 2 4 6 8 10
3. Results Dimensionless displacement
Fig. 4. Griffith locus corresponding to the numerical values of dimensionless

3.1. Stability analysis
load (PL/ f BW2 ) and dimensionless displacement (dL/f W2 ) for the geometry
(span, L = 40 mm; width, W = 6 mm; thickness, B = 4 mm) and test rig compliance
In Fig. 3 the Griffith locus corresponding to the numerical (C = 1.5 107 m/N) used in this work. The critical relative notch depth, c , is
values of dimensionless load (PL/ f BW2 ) and dimensionless larger than for C = 0 (Fig. 3).
118
14 12
2
(a) (b)
12 10
10
8
8
P (N)
P (N)
1 6
6
4
4
2 2
0 0
0 100 200 300 400 500 600 0 10 20 30 40 50
t (s) t (s)
Fig. 5. Characteristic load (P)time (s) experimental curves corresponding to CMOD controlled fracture tests. (a) CMOD rate of 0.001 mm/min for tip radius 6 mm
(1) and 60 mm (2). The maximum load is higher for larger radius. Stable fracture tests were attained. (b)CMOD rate of 0.018 mm/min for tip radius 6 mm.
locus is plotted in Fig. 4; very deep notches (c 0.7) are needed The critical stress intensity factor in mode I, KICp , for the max-
to achieve stable fracture using the displacement as control imum load attained during the test (Pmax ), was calculated as33 :
parameter.
3Pmax L
KICp = K () (1)
3.2. Fracture tests 2BW 3/2
where L is the span (40 mm), B and W are the geometrical param-
The experimental set up permitted to follow the CMOD rate eters defined before and K () is a general shape function which
requirements for the range used (0.001 and 0.018 mm/min). is valid for any value of the relative notch depth (0 1) and
Characteristic plots recorded for CMOD and displacement span-to-depth ratios ( = L/W) larger than 2.5 (2.5 16).33
controlled tests are shown in Figs. 5 and 6. Stable fracture, char- As described in the introduction, this parameter characterises
acterised by a continuous decrease of the load for increasing the initiation of fracture. To evaluate the fracture process, KICg ,
time, was always attained when CMOD was used as control was calculated according to:
parameter (Fig. 5). A decrease of displacement after the maxi-
1/2
mum load was needed to keep stable fracture (Fig. 6a).

2E
As expected for the notch sizes used ( 0.6 < c = 0.7), sta- KIC = (2)
1 2
bility was never achieved in displacement control as a sudden
load drop occurred once fracture initiated (Fig. 6b). where E and are the Youngs modulus and the Poisson ratio
In Table 2 the quantitative parameters of the fracture tests are of the material (Table 1) and is the fracture energy. was cal-
summarised. Two test time values were considered as character- culated as the integral of the loaddisplacement plot divided by
istic, the time to reach maximum loads (tp ) and the time required twice the area of the unnotched cross-section of the specimens,
to complete fracture once the maximum load was attained (t ). which evaluates the newly formed fracture surfaces.
14 28
(a) 2 (b) 4
12 24 3
10 20
P (N)
P (N)
8 16
1
6 12
4 8
2 4
0 0
0.000 0.002 0.004 0.006 0.008 0.000 0.005 0.010 0.015 0.020
d (mm) d (mm)
Fig. 6. Characteristic load (P)displacement (d) experimental curves. (a) CMOD controlled fracture tests using a rate of 0.001 mm/min for tip radius 6 mm (1)
and 60 mm (2). Decreasing of the displacement (snap-back) was needed to reach stability. (b) (3) CMOD controlled fracture test using a rate of 0.018 mm/min for
tip radius 6 mm. Decreasing of the displacement (snap-back) was needed to reach stability. Stable fracture was attained. (4) Displacement controlled fracture test
using a rate of 0.05 mm/min for tip radius 6 mm. Unstable fracture occurred. The actual maximum load is higher than for test (3).
119
Table 2
Fracture toughness values determined in this work. v: deformation rate; : notch tip radius; tp : time required to reach maximum load; t : time required to complete
fracture once the maximum load was attained; KICp : fracture toughness calculated from the maximum load. KICg : fracture toughness calculated from the fracture
energy. (S.D.): standard deviation.
CMOD control Displacement control
v, mm/min 0.001 0.018 0.05

, mm 60 6 6 6
tp , s 100 100 10 10
t , s 500 500 30 0
KICp (S.D.), MPa m1/2 1.11 (0.06) 0.72 (0.07) 0.86 (0.06) 1.17 (0.09)
KICg (S.D.), MPa m1/2 0.66 (0.15) 0.53 (0.02) 0.88 (0.02)
2.5 smallest one (6 mm). On the contrary, KICg values are coin-
0.001 mm/min cident for the range of notch tip radius used, revealing that the
2.0 0.018 mm/min characteristics of the notch are not so critical when calculations
0.005 mm/min are done by averaging the whole fracture process.
KICp ( MPa.m1/2)
0.05 mm/min In Fig. 7, values previously reported for notch tip

1.5
radius 200 mm are also plotted for comparison. A decrease of
KIC values with notch tip radius down to a plateau value, which
1.0
should be considered as the intrinsic value for the material, has
been observed in a number of ceramics3436 and theoretically
0.5 predicted by Fett and Munz.37,38 Essentially, this fact is due
to the non-adequacy of the equations used for calculations of
0.0 KIC from the maximum load values for notches which geome-
0 3 6 9 12 15 try separates from real cracks. When such equations are to be
1/2 1/2
( m ) used, different specimens with different radius should be tested
until the plateau value is reached. However, it is not possible
Fig. 7. Dependence of the KIC values, calculated from the maximum load point
to introduce notches with radius smaller than a few microns
(KICp ), on notch tip radius (). Legends stand for the used rate of the control
variable. Solid symbols correspond to this work (Table 2) and open symbols by standard techniques in fine grained material such as the one
correspond to.30 A decrease of fracture toughness with notch tip radius is shown. studied here, for which a radius 6 mm has been obtained. This
tip notch radius, 6 mm, is lower than the critical radius pre-
For the slow CMOD rate, the times to reach the maximum dicted by Damani et al.,34 15 mm, so we can consider that the
load (100 s) as well as the times required to complete fracture value measured for the 6 mm is independent of the tip notch
(500 s) are similar for both notch tip radii. In both cases, tough- radius and corresponds with the intrinsic fracture toughness
ness values calculated for the initiation of fracture, KICp , are value.
larger (36%) than those which average the whole fracture pro- For fast CMOD controlled tests (rate = 0.018 mm/min),
cess, revealing a time effect due to subcritical crack propagation which imply fracture times about one order of magnitude shorter
in this material that will be discussed latter. than the above discussed (1030 s, Table 2), and small notch tip
For the same CMOD rate (0.001 mm/min, characteristic plots radius (6 mm) there are no differences between KICp and KICg ,
in Figs. 5a and 6a), there are significant differences between KICp which fulfils the criteria of validity of the determined values.
for different notch tip radius, as highlighted in Fig. 7. Values for Moreover, these values are higher than those obtained in slow
the largest radius (60 mm) are about 54% higher than for the tests (Table 2).
Fig. 8. Characteristic fracture surfaces of CMOD controlled tests. Scanning electron micrographs. (a)CMOD rate of 0.018 mm/min for tip radius 6 mm. Mainly
transgranular fracture. (b)CMOD rate of 0.001 mm/min for tip radius 6 mm. Mixed trans/intergranular fracture.
120
As described in the introduction, subcritical crack prop- 4. Calomino AM, Brewer DN. Controlled crack growth specimen for brittle
agation has been reported for this material.30,31 In Fig. 8, systems. J Am Ceram Soc 1992;75(1):2068.
5. Srensen BF, Horsewell A, Jrgensen O, Kumar AN, Engbk P. Fracture
characteristic fracture surfaces for fast and slow CMOD con-
resistance measurement method for in situ observation of crack mechanisms.
trolled tested specimens are shown. For each testing condition, J Am Ceram Soc 1998;81(3):6619.
the same fracture mode was observed through the whole fracture 6. Wan D, Bao Y, Peng J, Zhou Y. Fracture toughness determination of
surface. When 0.018 mm/min was used as CMOD rate, frac- Ti3 Si(Al)C2 and Al2 O3 using a single gradient notched beam (SGNB)
ture was mainly transgranular (Fig. 8a) while the slowly tested method. J Eur Ceram Soc 2009;29:76371.
7. Kuszyk JA, Bradt RC. Influence of grain size on effects of thermal expansion
specimens presented a major proportion of intergranular frac-
anisotropy in MgTi2 O5 . J Am Ceram Soc 1973;56(8):4203.
ture (Fig. 8b). This features would indicate that subcritical crack 8. Srensen BF, Brethe P, Skov-Hansen P. Controlled crack growth in ceram-
propagation is due to the characteristics of the grain boundary, ics: the DCB specimen loaded with pure moments. J Eur Ceram Soc
probably to the presence of a residual silica glassy phase, as 1996;16(9):10215.
proposed in other works.39,40 9. Ebrahimi ME, Chevalier J, Fantozzi G. R-curve evaluation and bridging
stress determination in alumina by compliance analysis. J Eur Ceram Soc
The above discussion demonstrates that relatively fast tests
2003;23(6):9439.
are required to avoid subcritical crack propagation. When this 10. Nakayama J. Direct measurement of fracture energies of brittle heteroge-
process is avoided, KICp calculated from the maximum load for neous materials. J Am Ceram Soc 1965;48(11):5837.
specimens with a notch tip radius of 6 mm will be the intrinsic 11. Garca-Prieto A, Baudn C. Crack mouth opening displacement controlled
value. However, when tests are performed in displacement con- fracture tests of brittle ceramics. J Eur Ceram Soc 2010;30:32973302.
trol using a rate of (0.05 mm/min) equivalent to the fast CMOD 12. Garca-Prieto A, Hernndez J, Lpez M, Baudn C. Controlled fracture test
for brittle ceramics. J Strain Anal 2011;46:2742.
rate (0.018 mm/min), the value of KICp obtained in the unsta- 13. Sakai M, Urashima K, Inagaki M. Energy principle of elasticplastic fracture
ble tests is 36% higher than that obtained for stable fracture and its application to the fracture mechanics of a polycrystalline graphite. J
(Fig. 6b, Table 2). This fact can be attributed to the actual load Am Ceram Soc 1983;66(12):86874.
required to initiate fracture being masked by the effect of the 14. Bar-On I, Baratta FI, Cho K. Crack stability and its effect on fracture
toughness of hot pressed silicon nitride beam specimens. J Am Ceram Soc
inertia of the loading frame.
1996;79(9):23008.
15. Davidge RW, Tappin G. The effective surface energy of brittle materials. J
4. Conclusions Mater Sci 1968;3:16573.
16. Bueno S, Berger MH, Moreno R, Baudn C. Fracture behaviour
of microcrack free aluminaaluminium titanate ceramics with second
The effect of experimental variables on the KIC values deter- phase nanoparticles at alumina grain boundaries. J Eur Ceram Soc
mined in SEVNB in three points bending has been analysed. The 2008;28:196171.
analysis has been possible due to the possibility of comparison 17. Irwin GR. Fracture in encyclopedia of physics. Berlin: Springer Verlag;
of results obtained in stable and unstable fracture tests. 1958. p. 55189.
18. Sigl LS. On the stability of cracks in flexure specimens. Int J Fract
Stable facture tests have been reached for a fine grained mul-
1991;51:24154.
lite by using the CMOD as control variable. Using the same 19. Biolzi L, Cangiano S, Tognon G, Carpinteri A. Snap-back softening insta-
geometrical setup, tests performed using the displacement of bility in high-strength concrete beams. Mater Struct 1989;22:42936.
the frame as control variable led to unstable fracture. 20. Pastor JY, Planas J, Elices M. Ambient and high-temperature stable fracture
In order to determine accurate fracture toughness of materials tests in ceramics: applications to yttriapartially-stabilized zirconia. J Am
Ceram Soc 1993;76(11):29279.
with flat R-curve and that experience subcritical crack propaga-
21. Pastor JY, Planas J, Elices M. A new technique for fracture characterization
tion stable fracture tests at high deformation rates are required. of ceramics at room and at high temperature. J Test Eval 1995;23(3):20916.
Under such conditions, an intrinsic KIC = 0.86 0.06 MPa m1/2 , 22. Pastor JY, Planas J, Elices M. Ensayos de fractura estables en materiales
less than one half of those previously reported has been obtained cermicos. Bol Soc Esp Ceram V 1992;31(4):3225.
for the studied mullite with average grain size 0.7 0.5 mm. 23. Mah T-I, Mazdiyasni KS. Mechanical Properties of Mullite. J Am Ceram
Soc 1983;66:699703.
24. Schneider H, Schreuer J, Hildmann B. Structure and properties of mullite:
Acknowledgements a review. J Eur Ceram Soc 2008;28:32944.
25. Burgos-Montes O, Moreno R, Baudn C. Effect of mullite additions on the
fracture mode of alumina. J Eur Ceram Soc 2010;30:85763.
The authors acknowledge the financial support by projects: 26. Mazdiyasni KS, Synthesis Brown LM. Mechanical properties of stoichio-
MAT2009-14448-C02 and MCI-PET2008-0113 and Microtest metric aluminum silicate (Mullite). J Am Ceram Soc 1972;55:54852.
S. A. (Spain). Garca-Prieto acknowledges the financial support 27. Dokko PC, Pask JA, Mazdiyasni KS. High-temperature mechanical proper-
of the JAE-CSIC fellowship program (Spain). ties of mullite under compression. J Am Ceram Soc 1977;60:1505.
28. Torrecillas R, Fantozzi G, de Aza S, Moya JS. Thermomechanical behaviour
of mullite. Acta Mater 1997;45:897906.
References 29. Fernandez E, Baudn C. Creep damage in different 3Al2 O3 2SiO2
mullites tested in 4-point bending. J Eur Ceram Soc 2001;21:
1. Ghosh A, Jenkins MG, White KW, Kobayashi AS, Bradt RC. Elevated- 224351.
temperature fracture resistance of a sintered a-silicon carbide. J Am Ceram 30. Osendi MI, Baudn C. Mechanical properties of mullite materials. J Eur
Soc 1989;72(2):2427. Ceram Soc 1996;16:21724.
2. Nakayama J, Abe H, Bradt RC. Crack stability in the work-of-fracture test: 31. Casellas D, Baudn C, Osendi MI, Llanes L, Anglada M. Fracture resistance
refractory applications. J Am Ceram Soc 1981;64(11):6715. of mullite under static and cyclic loads. Scripta Mater 1998;38:3944.
3. Tattersall HG, Tappin G. The work of fracture and its measurement in metals, 32. Baudn C. Fracture mechanisms in a stoichiometric 3Al2 O3 2SiO2 mullite.
ceramics and other materials. J Mater Sci 1966;1:296301. J Mater Sci 1997;32:207786.
121
33. Guinea GV, Pastor JY, Planas J, Elices M. Stress intensity factor, com- advanced ceramics, materials and structures-B: ceramic engineering and
pliance and CMOD for a general three-point-bend beam. Int J Fract science proceedings, vol. 18. 1997. p. 15562.
1998;89:10316. 37. Fett T. Influence of a finite notch root radius on fracture toughness. J Eur
34. Damani R, Gstrein R, Danzer R. Critical notch-root radius effect in SENB-S Ceram Soc 2005;25:5437.
fracture toughness testing. J Eur Ceram Soc 1996;16:695702. 38. Fett T, Munz D. Influence of narrow starter notches on the initial crack
35. Nishida T, Hanaki Y, Pezzoti G. Effect of notch-root radius on the frac- growth resistance curve of ceramics. Arch Ap Mech 2006;76:66779.
ture toughness of a fine-grained alumina. J Am Ceram Soc 1994;77: 39. Michalske TA, Freiman SW. A molecular mechanism for stresscorrosion
6068. in vitreous silica. J Am Ceram Soc 1983;66:2848.
36. Kbler J. Fracture Toughness of ceramics using the SEVNB method: prelim- 40. Thomson R. The molecular wedge in a brittle crack: a simulation of mica
inary results. In: Proceedings of the 21st annual conference on composites, water. J Mater Res 1990;5:52434.
122
Captulo 6
INFLUENCIA DE LAS CARACTERSTICAS MICROESTRUCTURALES EN
LA TENACIDAD DE FRACTURA DE MATERIALES REFRACTARIOS
Captulo 6
6. INFLUENCIA DE LAS CARACTERSTICAS
MICROESTRUCTURALES EN LA TENACIDAD DE FRACTURA
DE MATERIALES REFRACTARIOS.
Una vez determinada la validez del dispositivo experimental, el procedimiento de
ensayo y la metodologa de ensayo y anlisis de resultados para materiales frgiles
(captulos 3-5), se ha estudiado la adecuacin y potencial necesidad de los ensayos
controlados por CMOD para la caracterizacin de materiales en los que actan
mecanismos de refuerzo durante la fractura. Para ello se seleccionaron materiales
refractarios comerciales que, a pesar de su importancia tecnolgica, no suelen ser objeto
de estudios sistemticos debido a la dificultad para controlar su composicin qumica
y/o su microestructura a escala industrial (apartado 1.7).
Se caracterizaron un amplio rango de materiales refractarios comerciales, tales
como materiales refractarios de slice (S), almina-slice (AS), almina-zirconia-slice
(AZS) y hormigones de alta almina (C). Se estudiaron las relaciones entre las
caractersticas qumicas, estructurales y microestructurales de estos materiales con su
comportamiento mecnico usando los parmetros de tenacidad descritos en los
apartados 1.2 y 1.7. Como se indic en el captulo 1 estos materiales son heterogneos y
poseen porosidades del orden del 20-60% frente a un 2% para las cermicas tcnicas
avanzadas analizadas en los captulos anteriores.
Parte de este trabajo se englob en el marco de un proyecto final de Carrera titulado
Puesta a punto de un mtodo de ensayo de fractura estable para la determinacin de la
tenacidad de fractura y el trabajo de fractura de materiales refractarios utilizando la
apertura de la grieta como variable de control, realizado por Manuel Dos Ramos
Lotito, estudiante del Departamento de Ciencia de Materiales de la Universidad Simn
Bolvar, Caracas, Venezuela, durante su estancia en el Instituto de Cermica y Vidrio.
125
Influencia de las caractersticas microestructurales en la tenacidad de fractura de materiales
refractarios
Las investigaciones derivadas de este trabajo han dado lugar a una estrecha
colaboracin entre el Instituto de Cermica y Vidrio y la Universidad Simn Bolvar,
quedando reflejados dichos resultados en la publicacin de un artculo en la revista
Journal of the European Ceramic Society (Anexo IV).

Se emplearon tcnicas de anlisis qumico por fluorescencia de rayos-X, difraccin
de rayos-X, microscopa ptica de luz reflejada y microscopa electrnica de barrido
con el fin de realizar una completa caracterizacin microestructural de los materiales.
Para la determinacin de la tenacidad de fractura se utilizaron las geometras de
ensayo para materiales refractarios descritas en el apartado 3.2.2. Se alcanzaron fondos
de entalla con un radio de curvatura de aproximadamente 130 m.
Se realizaron ensayos en control por CMOD a 70 m/min y en control por
desplazamiento a 0.2 mm/min, se seleccionaron estas velocidades porque son
equivalentes a la velocidad de incremento de la tensin, 0.15 MPa/s, segn norma
UNE-EN 993-6: Mtodos de ensayo para productos refractarios conformados densos286.

Los anlisis realizados permitieron caracterizar completamente la microestructura y
la textura de los materiales seleccionados.
Se alcanzaron ensayos de fractura estable tanto en control por desplazamiento como
en control por CMOD para todos los materiales caracterizados en este trabajo para las
dos geometras y longitudes de entalla relativas (a) utilizadas. Este hecho demuestra
que para materiales cermicos en los que tienen lugar mecanismos de refuerzo durante
la fractura, si se emplean mquinas suficientemente rgidas, no es necesario el uso del
CMOD como parmetro de control.
126
Captulo 6
No se encontraron diferencias significativas entre los parmetros de tenacidad de
fractura determinados para un mismo material utilizando distintos parmetros de control
y diferentes geometras de probetas y longitudes de entalla.
En la figura 6.1 se muestran los parmetros de tenacidad de fractura definidos en el
captulo 1, KIC, gwof, gnbt. Se observa como estos valores varan en funcin de las
diferentes caractersticas microestructurales de estos materiales. Se encontraron
diferencias significativas entre los tres parmetros de tenacidad correspondientes tanto
al inicio como a la propagacin de la fractura para materiales diferentes.
1.6
1.4 25
1.2
KIC [MPa m1/2]
20
1.0
0.8 15
gnbt [J/m ]
2
0.6
10
0.4
0.2
5
0.0
AZS1 AZS2 AS ASI S C
0
130
120
110
100
90
80
gwof [J/m2]
70
60
50
40
30
20
10
0
Figura 6.1. Parmetros de tenacidad de fractura para los materiales refractarios.

(a) Factor crtico de intensidad de tensiones en modo I, KIC.
(b) Energa para el inicio de la grieta, gnbt.
(c) Trabajo de fractura, gwof.
127
Influencia de las caractersticas microestructurales en la tenacidad de fractura de materiales
refractarios
6.3. Conclusiones
- Es posible alcanzar de manera rutinaria a nivel de laboratorio ensayos de fractura
estable utilizando el desplazamiento del marco de carga como parmetro de control, si
se utiliza un dispositivo de alta rigidez.
- Se ha demostrado la adecuacin de los ensayos de fractura estable de probetas
SENB en flexin en tres puntos para la caracterizacin de la tenaciad de materiales
refractarios.
- Incluso utilizando el menor de los tamaos estandarizados para la caracterizacin
de estos materiales (150x25x25 mm3) y con una longitud de entalla relativa, a=0.5, se
han encontrado diferencias significativas entre los parmetros de tenacidad en funcin
de la microestructura de los diferentes materiales refractarios caracterizados.
- Las caractersticas microestructurales que determinan la resistencia al inicio de la
fractura son diferentes a las que regulan el proceso de propagacin de la grieta. Las
caractersticas de los agregados determinan la tenacidad en el inicio de la fractura,
mientras que, para alcanzar altos valores del trabajo de fractura, es preciso la presencia
de elementos microestructurales que originen los mecanismos de refuerzo.
128
ANEXO IV
Este trabajo ha dado lugar a la siguiente publicacin:
- Ana Garca-Prieto, Manuel Dos Ramos-Lotito, Delia Gutirrez- Campos, Pilar
Pena, Carmen Baudn. Influence of microstructural characteristics on fracture
toughness of refractory materials. J. Eur. Ceram. Soc. 35 (6) 1955-1970 (2015).
(DOI:10.1016/j.jeurceramsoc.2014.12.020).
129
130
ScienceDirect
Influence of microstructural characteristics on fracture toughness of

refractory materials
Ana Garca-Prieto a , Manuel Dos Ramos-Lotito a,b , Delia Gutirrez-Campos b ,
Pilar Pena a , Carmen Baudn a,
a Instituto de Cermica y Vidrio, CSIC, Kelsen 5, Madrid 28049, Spain
b Universidad Simn Bolvar, Dept. Ciencia de los Materiales, Valle de Sartenejas, Caracas 1080, Venezuela
Received 28 July 2014; received in revised form 15 December 2014; accepted 16 December 2014
Available online 13 January 2015
Abstract
Basic relationships between the microstructure and the texture of refractories and their toughness have been established. A series of commercial
materials has been chosen in order to highlight the influence of microstructural characteristics on fracture behaviour and associated toughness.
Silica, silicaalumina and silicaaluminazirconia based shaped refractories and a calcium aluminate cement bonded concrete have been analysed.
Extensive microstructural characterisation has been performed using a combination of techniques, including chemical analysis by X-ray fluores-
cence, X-ray diffraction, reflected light optical microscopy and scanning electron microscopy with analysis by dispersive energies. Fracture has
been characterised using stable fracture tests of SENB tested in 3 point bending. Stability was reached in displacement and crack mouth opening
displacement controlled tests. Size effect has been analysed by using two different specimen sizes and relative notch lengths. For the range of
microstructures studied, the obtained results have allowed to characterise toughness and establish the relationships toughness-microstructure and
texture.
Keywords: Alumina; Castables; Microstructure; Refractories; Toughness
1. Introduction the processing or installation procedure. Thus, the study of com-

mercial products imposes the necessity of characterising them
The study of commercial refractories is quite uncommon in a comprehensive manner. It is particularly important for the
due to the challenges generated by the impossibility of con- understanding of their behaviour in terms of the basic material
trolling the chemical composition and/or the microstructure of science relationships between their composition, microstructure
the material. Nevertheless, tackling this task has the advantage and properties. By using several complementary techniques like
of avoiding problems linked to reproducing the manufacturing chemical analysis, mineralogical studies, and microstructural
process at a laboratory scale, which is not a trivial issue due and textural analysis, it is possible to obtain useful data for
to inherent differences between laboratory and refractory plants generating a detailed description of the material.
processes. Refractories are applied in processes involving mechanical
The performance of refractories in use is directly related to strains, variable high temperatures and aggressive environments
their microstructure and texture which, in turn, is determined including corrosion and erosion from solids, liquids and gases
by the characteristics of the raw materials (chemical and min- in movement. In particular, the thermal stress fracture of refrac-
eralogical composition and size and shape distribution) and by tory components, due to temperature cycling and/or temperature
differences through the material, is a widespread problem of
industrial importance. In the same way, mechanical overload as
Corresponding author. Tel.: +34 917355840. may be originated by impact during the loading of the process
E-mail address: cbaudin@icv.csic.es (C. Baudn). vessel, as occurs in electrical arc furnaces, or by deformations
http://dx.doi.org/10.1016/j.jeurceramsoc.2014.12.020
0955-2219/ 2015 Elsevier Ltd. All rights reserved.
131
1956 A. Garca-Prieto et al. / Journal of the European Ceramic Society 35 (2015) 19551970
of the kiln shell, as observed in the cement industry, can lead of fracture in the 60s,1315 and since then, many laboratories
to fracture. Fracture as a result of thermal or mechanical shocks have conducted such analyses but a standard test has not yet
could lead to a catastrophic failure of the refractories with strong been established. The concept of work of fracture introduced
consequences for the process in which they are used. How- by Nakayama4,13 is defined as the mean work per unit of pro-
ever, fracture can also result just in the development of a crack jected fracture area required to propagate a crack in a stable
pattern.1 In this regard, it should be pointed out that the conven- way. In Nakayamas test a parallelepiped bar with a triangular
tional characterisation of the mechanical properties of refractory (chevron) notch at its centre is loaded in three point bend-
products is done in terms of two properties: crushing strength ing (3-pb). Using this geometry, refractory specimens fracture
and modulus of rupture. Even though such properties might be in a stable way when a sufficient stiff machine is used. The
adequate for quality control purposes, the results achieved do not value of work of fracture is determined from the total area
allow the evaluation or study of the fracture processes, and more under the LoadDisplacement curve recorded during the exper-
basic mechanical properties are needed for characterisation. iment and the size of the projected fracture surface. Chevron
The extension of fracture due to thermal or mechanical notch experiments often imply high variability (>10%)6,7,1618
strains, i.e.: the damage would be determined by the ratio because coarse aggregates at the apex of the chevron notch give
between the amount of energy available and the energy needed very high values as compared to the average for the material.
to create new crack surfaces or specific fracture energy. There- For example, 1023% variability in wof has been reported
fore, this ratio will characterise the resistance of materials for alumina-spinel castables,18 and for high alumina refracto-
to subcritical crack growth and the proneness to catastrophic ries (45100 wt.% alumina) variability up to 38%7,16 has been
failure.2,3 found.
Refractory products are heterogeneous ceramic materials In the 1980s the wedge-splitting fracture test was developed
which fracture exhibits notable deviations from pure lin- to perform stable fracture tests and patented by Tschegg.19 This
ear elastic.1,410 Several energy-consuming processes ahead test is a special form of the so-called compact tension test,
(process zone) and behind (process wake) the crack tip are the specimen with a groove and notch is split in two halves
considered to contribute to this behaviour. Microcracking and while monitoring the load and crack mouth opening displace-
multiple crack branching are usually observed in the frontal pro- ment (CMOD). In this experimental setup, large specimens of
cess zone, while grain bridging and friction of the crack faces are the size of bricks can be tested.5,8,2030 Most data produced using
able to consume energy in the process wake zone. As a result, the the wedge-splitting test are reported for pure magnesia, magne-
parameters that evaluate toughness of refractories are no longer sia spinel and magnesia carbon refractories,5,8,9,2427 which are
material constants but they increase for increasing crack exten- out of the scope of this paper. A relatively low number of stud-
sion. In the field of advanced ceramics, this fracture pattern is ies provide data for high alumina castables and alumina-based
called rising R-curve behaviour, with R representing toughness; shaped materials.20,22,23,2830
in contrast with the brittle flat R-curve observed for glass or for Ribeiro and Rodrigues22 applied the wedge splitting method
fine grain size ceramic specimens. to characterise fracture energy of two high-alumina refractory
The R-curve concept is not used for the characterisation of castables. Miyaji et al.29 analysed five different castable for-
refractories because of the experimental difficulties associated mulations and introduced a figure of merit derived from the
to follow a single crack in such heterogeneous microstructures. LoadDisplacement curve to evaluate the thermal shock damage
Unlike the R-curve, the work of fracture, wof , has been success- resistance. More recently, a methodology was presented using
fully used to describe the fracture of refractories.6 The advantage the wedge-splitting test complemented with images obtained
of this energy parameter is that it does not require any assump- during mechanical loading to determine the crack propagation
tions about the constitutive equation of the body with the crack for a pure alumina and alumina with titania and zirconia addi-
to discuss its propagation.11,12 In terms of energy, the criti- tives refractory compositions.28,30 Jin et al.27 have proposed a
cal energy release rate, Gc , of refractories with well designed methodology to estimate the tensile strength, and Youngs mod-
microstructures is always significantly lower than the specific ulus of refractories in addition to the specific fracture energy
fracture energy, GF (2 wof ). The ratio between the specific frac- from wedge splitting test results.
ture energy and the energy release rate, GF /Gc , has been defined It should be pointed out that variability of data of mechan-
as a toughness, flexibility or apparent ductility ratio.1 The higher ical properties in the refractory literature is most of the time
this relation, the higher is the resistance of the material to damage not reported, and, in many cases, only one data for each mate-
by thermal or mechanical strains. rial experimental condition is provided. This is often the case
An important point to consider when determining the specific of work of fracture values determined by the splitting test
fracture energy (or the work of fracture) of refractories is the so, it is not possible to discuss in a general way the repeat-
potential influence of the specimen size (size effect) in the ability associated with this technique. Nevertheless, the scarce
obtained values because GF increases with increasing fracture data available reveal rather high dispersion of wof results; for
surface until the specimen geometry allows the development of two commercial alumina based low cement (2 wt.% alumina)
a well developed wake zone. castables heat-treated at 1100 C, 12 and 18% variability have
Nakayama, Tattersall and Tappin, and Davidge and Tappin been reported22 and variability between 5 and 22% has been
accomplished the first studios on the determination of work reported for basic refractories.24
132
A. Garca-Prieto et al. / Journal of the European Ceramic Society 35 (2015) 19551970 1957
Despite the fact that there is an evident interest in the wedge intensity factor in mode I, KIC , has been calculated from the
splitting test for work of fracture determination of refractories maximum load according to Guinea et al.40 :
due to its capability for testing relatively large specimens in
order to average the whole fracture process, it is complicate 3PL
KIC = K () (1)
to establish this method as a laboratory routine one. More- 2BW 3/2
over, as mentioned before, variations in reported ranges of where P is the applied load, L is the span, B and W are the
result would mask the effect of microstructural differences on geometrical parameters of thickness and width of the specimen,
toughness. respectively. K () is a general shape function which is valid
From the extensive spectrum of available fracture tough- for any value of the relative notch length (0 1) and
ness tests, bending of parallelepiped specimens with straight span-to-depth ratios ( = L/W) larger than 2.5 (2.5 16).40
through notches (SENB, Single Edge Notch Beam) using as From KIC and Youngs modulus, the energy for crack initiation,
control variable the displacement of the load frame (displace- nbt (= Gc /2) was calculated according to the analysis of Irwin
ment control) is a relatively simple way of testing materials for plane strain conditions41 :
with R-curve fracture, like refractory products. In general, in
the refractory field this method is used to determine toughness 2 (1 2 )
KIC
parameters describing the initiation of fracture, critical stress GC = (2)
E
intensity factor in mode I, KIC , and the energy for crack initia-
tion, nbt which is a measure of the critical energy release rate where KIC is the critical stress intensity factor in mode I, is
Gc , Gc = 2 nbt .4,15,3133 the Poissons ratio and E is the Youngs modulus.
The advantage of using stable fracture tests of SENB in 3- To evaluate the fracture process, wof has been calculated
pb is that both parameters for initiation, nbt , and propagation, from the area under the LoadDisplacement curve and the
wof , of fracture can be extracted from a single test. The size projection of the fracture surface following the procedure of
of the specimen can be readily adjusted to ensure that the lig- Nakayama.4,13
ament is large enough to encompass the fracture process zone; In order to estimate the inelastic energy contribution to frac-
in this way, the results will be statistically valid. As a term of ture, the ratio between the specific fracture energy and the energy
comparison, a representative volume 34 times the largest aggre- release rate, GF /Gc , has been calculated using the experimental
gate size was determined by Romero and Masad34 and Wagoner wof and nbt values.
et al.35 for SENB testing of asphalt concrete. However, there The chemical, structural and microstructural characteristics
are two main experimental problems to solve with regard to of the six studied refractory materials were correlated with
this test.1 On the one hand, the attainment of stability for this their fracture behaviour using the different toughness parameters
geometry is more difficult than for the chevron specimens and, above described.
on the other hand, straight through notches are more prone to
lead to the wandering of the propagating crack from the initial 2. Experimental
plane.
When displacement of the load frame is used as control Six different types of commercial refractory materials were
variable for SENB in 3-pb, a general requirement to reach sta- studied: two aluminasilicazirconia (AZS), a superduty fire-
bility is to use a high stiffness machine.3639 The crack mouth clay, one group 28 insulating firebrick, one standard silica brick
opening displacement (CMOD) has been proposed and used and a high-alumina regular castable heat treated at the use tem-
as control parameter for stable fracture testing under condi- perature (1100 C). They were labelled as follows: AZS1, AZS2,
tions that would have led to unstable fracture for displacement AS, ASI, S and C, respectively. Fig. 1 shows the macroscopic
controlled tests. In this regard, detailed procedure and theoret- aspect of these six refractories.
ical considerations for performing CMOD controlled fracture Chemical analysis was carried out with a Philips (Holland) X-
tests of brittle materials have been previously reported by the ray fluorescence equipment, model MagiX PW 2424. Samples
authors.3739 were prepared with the standard procedure of forming a fused
In this work six different types of commercial refractory pellet. Li2 B4 O7 was added to the ground powder and prepared
materials with distinctive specifications in chemical compo- capsules were heat treated at 1000 C.
sition and microstructural characteristics were tested. The Bulk and true densities and apparent and true porosity were
products were evaluated using CMOD and displacement determined following the procedures described in two standards:
of the load frame as control parameters using equivalent EN 993-142 and EN 993-2.43 He picnometry was done using a
rates and two relative notch lengths (0.25 and 0.50). Once Quantachrome (USA) apparatus.
assured the significance of the obtained data, the important Determination of crystalline phases was performed on ground
challenge in the study has been to correlate the micro- samples in a Bruker (Germany) X-ray diffractometer, model D8
structural features of the refractory materials with the fracture Advance, with copper anode (CuKa1 = 0.15418 nm) working
behaviour. at 40 kV and 40 mA. Scans were performed in continous mode
From the stable fracture tests different toughness parameters with steps of 0.05 at a rate of 153 s per step. The analysis of
have been evaluated; the usual terminology used in refractory the XRD patterns was accomplished using the EVA 6.0 Diffrac
practice has been assumed for reporting. The critical stress plus software (Bruker, Germany). The experimental diffraction
133
Fig. 1. Macroscopic aspect of the six commercial refractory materials studied. Scanned images of polished cross sections of tested specimens (25 mm 25 mm). (a)
Aluminasilicazirconia brick, AZS1. (b) Aluminasilicazirconia brick, AZS2. (c) Superduty fireclay brick, AS. (d) Group 28 insulating brick, ASI. (e) Standard
silica brick, S. (f) High-alumina regular castable, C.
patterns were compared to the files of the International Centre following the procedure of EN 993-6 standard,45 using a uni-
for Diffraction Data (ICDD).44 versal testing machine (Instron 1114, USA). Reported values
Standard specimens (EN 993-6)45 for mechanical charac- are the average of 3 determinations and errors are the standard
terisation (150 mm 25 mm 25 mm and 200 mm 40 mm deviations.
40 mm) were diamond machined from the received bricks and Notches for toughness testing were done using a Buehler
concrete pieces. (USA) sawing machine model IsoMet 4000 with diamond disc
The size distributions of the aggregates were evaluated from of 300 mm width to reach notches with tip radius around 100 mm.
scanned (HP Scanjet 5370 C, USA) images of the lateral surfaces Specimens with relative notch lengths = a/W = 0.25 and 0.50
(150 mm 25 mm 25 mm) of the specimens used for mechan- (a = notch length, W = specimen width) were prepared and tested
ical testing with the Leica Qwin software (UK). The equivalent in three point bending using spans of 125 mm and 180 mm
diameter was calculated from the surface of the particles assum- for the small and large specimens, respectively. All tests were
ing spherical shape. A minimum of 560 particles was analysed conducted in universal testing machine (model EM1/50/FR,
for each material. This study was not performed for the insulat- Microtest, Spain) with capability of crack mouth opening dis-
ing brick because its largest microstructural features were the placement (CMOD) recording and controlling. This equipment
pores which all had similar size (1200 mm). has been described elsewhere.37,38 Tests were performed using
Specimens for microstructural evaluation were embedded the CMOD and the displacement of the frame load as control
in resin in vacuum environment to assure the penetration of parameters. Rates of 70 mm/min and 0.02 mm/min for CMOD
the resin in the pores. Then, microstructural characterisation and displacement, respectively, were applied. Three tests were
of the materials was carried out using a reflected-light opti- performed for each testing condition; reported values of the
cal microscopy, RLOM, with a Zeiss (Axiophot, Germany) toughness parameters are the average of the three determinations
microscope and field emission scanning electron FE-SEM with and errors are the standard deviations.
analysis by energy dispersive X-ray spectroscopy (EDS) micro-
scope (Hytachi S-4700 type I, Japan).
Youngs modulus of the refractory materials was deter- 3. Results
mined from the resonance of the parallelepiped bars
(150 mm 25 mm 25 mm) tested in flexure by impact 3.1. Physico-chemical characterisation
(Grindosonic, Belgium). Calculations were performed using the
common value of Poissons ratio for refractory materials (0.17). Table 1 shows the complete chemical analyses for the stud-
Given values are the average of 6 determinations and errors are ied refractories. As expected, major constituents of both AZS
the standard deviations. materials (AZS1 and AZS2) are Al2 O3 , SiO2 and ZrO2 . SiO2
Room temperature modulus of rupture (MOR) was deter- content is more than double in AZS2 than in AZS1 while Al2 O3
mined by three point bending, 3-pb, (span 125 mm; 0.5 mm/min) and ZrO2 contents are close for both products. Main differences
134
Table 1
Chemical analysis of the studied materials (wt.%).
wt% AZS 1 AZS 2 AS ASI S C
Al2 O3 68.8 0.4 57.7 0.4 44.6 0.3 48.9 0.3 0.71 0.5 74.2 0.4
SiO2 10.8 0.3 22.8 0.2 48.9 0.4 45.7 0.3 95.2 0.3 13.2 0.2
Fe2 O3 0.20 0.04 0.16 0.04 3.08 0.03 1.09 0.03 0.22 0.04 0.82 0.03
K2 O 0.038 0.005 0.61 0.05 1.32 0.05 0.16 0.06 0.11 0.05
MgO <0.001 0.31 0.05 0.39 0.05 0.17 0.04 0.10 0.04
CaO 0.096 0.006 0.093 0.006 0.48 0.08 0.69 0.08 3.3 0.07 9.36 0.04
TiO2 0.48 0.01 0.20 0.01 1.67 0.01 1.41 0.01 0.16 0.01 1.89 0.01
P2 O5 0.050 0.002 0.13 0.002 0.19 0.002 0.02 0.004 0.058 0.002
Na2 O 0.19 0.05 0.091 0.02 0.096 0.04 0.20 0.05 <0.08
ZrO2 15.9 0.1 18.0 0.1 0.084 0.04 0.071 0.04 0.036 0.04 0.094 0.04
Y2 O3 0.25 0.02 0.25 0.02
HfO2 0.36 0.02 0.39 0.02
WO3 0.05
Loss of ignition 0.28 0.03
Table 2 Apparent porosity constitutes about 6674% and 88% of the

Density and porosity values of the studied materials. b : bulk density; t : true total porosity for the four conventional bricks and the unshaped
density; t : true porosity; a : apparent porosity.
material high-alumina castable (C), respectively.
Material b (g/cm3 ) t (g/cm3 ) t (vol.%) a (vol.%) The crystalline phases identified in the materials are summ-
AZS 1 3.11 0.03 3.944 0.002 21.0 0.8 13.8 0.2 arised in Table 3 along with the ICDD (International Centre
AZS 2 2.65 0.03 3.74 0.01 29.1 0.8 19.9 0.4 for Diffraction Data) files used for identification.44 The minor
AS 2.12 0.06 2.85 0.01 26 2 17.7 0.5 phases correspond to impurities that are usually present in the
ASI 0.9 0.1 2.84 0.01 68 3 66 9 raw materials of these products.
S 1.84 0.01 2.34 0.02 21.3 0.5 15.8 0.1
C 2.67 0.02 2.93 0.01 19.6 0.7 17.3 0.7
Zircon (ZrSiO4 ) is one of the major phases in both AZS
materials being corundum (a-Al2 O3 ) and mullite (Al6 Si2 O13 )
the second phases in AZS1 and AZS2, respectively. A signifi-
cant amount of baddeleyite (monoclinic zirconia; m-ZrO2 ) was
in the compositions of the dense fireclay (AS) and the insulating also detected in AZS2. The presence of AlPO4 in AZS1 was
firebrick (ASI) are the higher Al2 O3 and lower impurity (TiO2 , identified taking into account the chemical analysis (Table 1).
Fe2 O3 ) contents in the latter. The analysis of the silica brick (S) The dense fireclay and the insulating bricks (AS and ASI) are
corresponds to a standard one with around 1 wt.% total amount constituted mainly by mullite (Al6 Si2 O13 ) and minor amounts
of characteristic impurities (Al2 O3 , TiO2 and alkalis). The com- of quartz, corundum and cristobalite. This last silica phase was
position of the castable (C) is that of a regular high-alumina one more abundant in the dense fireclay than in the insulating brick.
with high cement content (>8 wt.% CaO). The diffractogram of AlPO4 overlaps with those of the silica
Table 2 summarises porosity and density values. For four of polymorphs so its presence could not be conclusively established
the shaped materials (AZS 1 and 2, AS and S) and the castable only by this technique. But, from the XRF analysis (Table 1),
(C), porosity values range between 15 and 30% whereas material both materials probably had phosphoric additives in their com-
ASI presents the high porosity characteristic of insulating bricks. position.
Table 3
Crystalline phases detected by X-ray diffraction in the studied materials.
Zircon Baddeleyite Mullite Corundum Quartz Tridymite Cristobalite Fluorite CaAl2 O4 Other minors
(ZrSiO4 ) (m-ZrO2 ) (Al6 Si2 O13 ) (a-Al2 O3 ) (SiO2 ) (SiO2 ) (SiO2 ) (CaF2 )
ICDD file44 81-0589 37-1484 15-0776 46-1212 46-1045 71-00261 82-0512 35-0816 70-134
AZS 1 Major Major (AlPO4 )
11-500
AZS 2 Major Abundant Major Minor
AS Major Minor Abundant Major
ASI Major Minor Abundant Minor
S Minor Abundant Major (pWCaSiO3 )
74-0874
C Abundant Major Abundant Abundant (CaAl4 O7 )
23-1037
135
Fig. 2. Characteristic microstructural features of the aluminasilicazirconia materials. Polished surfaces. (a) AZS1. Porous alumina aggregates (grey), zircon
sand (white) and fine alumina (grey) are observed. Reflected light optical microscopy micrograph. (b) AZS1. Detail of a porous aggregate. Reflected light optical
microscopy micrograph. (c) AZS2. Electrofused mullite aggregates surrounded by a mullite matrix (grey) with dispersed zircon flour (white). Reflected light optical
microscopy micrograph. (d) AZS2. Detail of an electrofused mullite aggregate with alumina particles embedded. Reflected light optical microscopy micrograph.
(e) AZS2. Detail of the matrix showing partially decomposed zircon flour (grey particles surrounded by nanometric white zirconia particles). Scanning electron
microscopy micrograph.
The silica brick (S) is composed of the silica polymorphs of this parameter corresponds to the dense fireclay material (AS,
cristobalite and tridymite with minor amount remnant quartz; 4000 mm) while for AZS1 it is intermediate (4400 mm). The
traces of wollastonite (CaSiO3 ) are also present in this brick. silica refractory, S, has the widest distribution from 5400 mm
The major phase in the conventional high-alumina castable down to about 100 mm. Aggregates for the AS material are
(C) is corundum, followed by mullite. The hydraulic phases of
these types of refractories (CaAl2 O4 and CaAl4 O7 ) were also
clearly identified in the XRD pattern. Fluorite (CaF2 ) was also
Table 4
detected in this material.
Parameters of the distributions of aggregate size in the studied materials.
3.2. Microstructure Grain size (mm) AZS 1 AZS 2 AS S C
Average 1709 1591 1227 1736 1393

Condensed information of quantitative data for aggregate Standard deviation 597 694 539 923 732
size distributions of the five dense materials is presented in Maximum 4396 5178 3901 5370 5212
Minimum 634 623 457 91 315
Table 4. Maximum aggregate sizes for S, C and AZS2 compo- No. of particles analysed 643 729 567 563 943
sitions are similar and the largest (5000 mm). Minimum value
136
Fig. 3. Characteristic microstructural features of the fireclay-based bricks. Reflected light optical micrographs of polished surfaces. (a) Superduty fireclay brick, AS.
Mullite chamote aggregates in a mullite matrix. (b) Detail of the aggregate- matrix interface showing bonding. (c) Group 28 insulating brick, ASI. Spherical pores
are observed. (d) Detail of the mullite matrix in figure (c).
smaller than those of the AZS being the distribution centred of similar colour and pores (Fig. 3a and b). According to the XRD
around 1200 mm. (Table 3) and the chemical analyses (Table 1), the aggregates
Micrographs summarising the characteristic microstructural (4000500 mm, Table 4) present in this material are a mullite
features of the studied refractory materials are shown in chamote with low alumina content (typically wt.% 65% mul-
Figs. 25. lite, 15% cristobalite, 20% glassy phase). The matrix would have
Microstructure of AZS1 is constituted by grey rounded similar composition as that of the aggregates. In this material, a
porous aggregates (4400600 mm, Table 4, Fig. 2a) in a matrix good bonding between the aggregates and the matrix is observed.
of fine and medium porous grey particles and dense rounded No aggregates are present in the insulating brick (ASI)
white particles (180 mm) (Fig. 2b). Significant porosity is (Fig. 3c and d) and its microstructure is formed by relatively
also observed in the matrix. According to the crystalline phases small particles and round pores (diameter 1200 mm), which,
detected by XRD (Table 3) and the morphology and colour of the as described before, are the largest microstructural features.
particles in RLOM, the grey particles were identified as alumina Considering the XRD (Table 3) and the chemical analysis
and the dense white particles as zircon sand. (Table 1), the composition of this material is similar to that of
The microstructure of AZS2 (Fig. 2ce) is dominated by the matrix of AS.
dense grey aggregates (5000600 mm, Table 4) constituted by Brick S presents the standard microstructure of a silica
columnar grains and significant amounts of glassy phase at the brick (Fig. 4a and b), made of relatively large (1005400 mm,
boundaries. Clearer particles are observed in the interior of Table 4) aggregates with well-defined grain boundaries in
some of the aggregates. Such features, together with the XRD RLOM (Fig. 4a). These aggregates with the typical fish-scale
(Table 3) and chemical analyses (Table 1), allowed identifying substructure are constituted by the smaller cristobalite particles
these aggregates as electrofused mullite, with corundum and formed during the transformation of quartzite, in agreement with
glass as secondary phases. Fine (<50 mm) white angular par- the XRD (Table 3). Two different phases are observed in the
ticles and medium and small grey particles are present in the porous matrix (Fig. 4b). The clearest one- in which additives,
matrix (Fig. 2d). Taking into account the XRD (Table 3) and the CaO and main impurities were detected by EDS- is a Ca-rich sil-
chemical analyses (Table 1) the grey ones could be identified as icate glass in which the conversion of cristobalite into tridymite
mullite and the white ones as flour zircon particles. These latter is favoured.46 The darker phase in contact with the glass should
were partially decomposed at the grain boundaries (Fig. 2e) in be the tridymite detected by XRD (Table 3).
agreement with m-ZrO2 being detected by XRD (Table 3). Material C (Fig. 5a and b) was constituted by rounded
The dense fireclay material (AS) was formed by coarse and aggregates (5200300 mm, Table 4) embedded in a matrix with
medium dense (down to 200 mm) aggregates bonded by a matrix medium sized angular particles. The aggregates presented a
137
Fig. 5. Characteristic microstructural features of the high-alumina regular

castable, C. Polished surfaces. Reflected light optical microscopy micrographs.
(a) General aspect of the microstructure. Coarse bauxite aggregates and medium
brown corundum particles (angular) are observed. (b) Detail showing the sub-
structure of the bauxite aggregates constituted by mullite (grey) and corundum
(white).
3.3. Mechanical characterisation
Youngs modulus (E) and modulus of rupture (MOR) values

are summarised in Table 5. Both parameters follow the same
trend, being the largest for the AZS2 material and the high-
alumina castable, C, and the lowest for the insulating material,
ASI. Intermediate range of values was obtained for AZS1, AS
Fig. 4. Characteristic microstructural features of the standard silica brick, S.
Polished surfaces. (a) General aspect of the microstructure. Reflected light opti-
and S.
cal microscopy micrograph. (b) Detail showing the highly transformed silica
grains (cristobalite) and the tridymite grains embedded in the CaO-rich glassy
Table 5
phase formed in the matrix of the refractory. Scanning electron microscopy
Youngs modulus (E and E0 ) and modulus of rupture (MOR) of the studied mate-
micrograph. (c) Reflected light optical micrograph showing the grain boundary
rials. E tested in flexure by impact. E0 calculated from the MSA model.51 MOR
cracking.
determined from 3-pb tests. Main crystalline phases detected by X-ray diffrac-
tion are also shown: ZS = zircon, A = corundum, M = mullite, m-Z = baddeleyite,
Q = quartz, C = cristobalite, CAC = calcium aluminates.
substructure formed by two main phases (Fig. 5b) typical of Material E (GPa) E0 (GPa) MOR (MPa) Crystalline phases
bauxite aggregates, in agreement with the crystalline phases AZS 1 29 3 54 9.2 0.6 ZS + A
observed by XRD (Table 3). They were formed by corundum, AZS 2 46 2 110 14 2 M + ZS + m-Z
mullite and secondary Fe and Ti containing phases. Medium AS 26 1 57 8.0 0.9 M
fraction was constituted of brown corundum. According to ASI 4 1 30 1.0 0.2 M
S 17 1 33 8.2 0.5 Q+C+A
the XRD (Table 3), the matrix was constituted by calcium
C 50 4 91 13 2 A + CAC + others
aluminates.
138
Fig. 6. Characteristic Load (P)Displacement (d) curves. The dimensions of the beams tested were 150 mm 25mm 25 mm for (a) and (b) and
200mm 40mm 40 mm for (c). The relative notch lengths were 0.25 for (a) and (c) and 0.50 for (b). (a) Group 28 insulating brick, ASI. For the same material,
similar curves were obtained for different tests performed under the same conditions. (b) Aluminasilicazirconia brick, AZS2. Similar results were obtained for
both control parameters. (c) High-alumina regular castable, C. Notice the apparent snap-back for specimens tested using the crack mouth opening displacement as
control parameter.
Characteristic LoadDisplacement curves recorded during control when the LoadDisplacement loop associated with the
the fracture tests are plotted in Fig. 6 and values for the toughness apparent snap-back present in the latter was relieved.
parameters are summarised in Tables 6 and 7. LoadDisplacement curves obtained for the same material
A monotonous decrease of load with time was found in all and experimental conditions were similar (Fig. 6a) and, con-
tests indicating that stable fracture was reached both in displace- sequently, variability of the calculated toughness parameters
ment and in CMOD controlled tests. All LoadDisplacement was relatively low (10% in most cases, Tables 6 and 7). For
curves showed a linear elastic region followed by a long tail the same material, specimen and span dimensions and relative
of monotonous decreasing load for increasing displacement notch lengths, curves obtained using both control parameters
from the maximum load. Just before reaching the maxi- were similar (Fig. 6b).
mum load, some of the curves showed a moderate non-linear As shown in Tables 6 and 7, for each material there are no
region. significant differences between the toughness parameters deter-
The only case in which it was not possible to reach stabil- mined using different specimen and span dimensions, relative
ity using CMOD control at the experimental rate used in this notch lengths and control parameters. Therefore, the average
work was the combination of the largest specimen size and the values obtained for relative notch lengths, = 0.5, in CMOD
smallest notch ( = 0.25) (Fig. 6c). In this case, an apparent snap- controlled tests (Table 6) are plotted in Fig. 7 to facilitate compar-
back was observed in the LoadDisplacement curves, which ison. For all materials, values of the energy for crack initiation,
would indicate that the material presented brittle fracture. How- nbt , are lower than those of the work of fracture, wof . The
ever, fracture in displacement control was stable and, in fact, the largest value of the toughness ratio corresponds to material
obtained curves were coincident with those obtained in CMOD AZS1, followed by C.
139
Fig. 7. Average toughness parameters for the studied materials calculated from the values recorded during stable tests ( = 0.50, crack mouth opening displacement
controlled tests, Table 6). (a) Critical stress intensity factor in mode I, KIC . (b) Work of fracture, wof . (c) Energy for crack initiation, nbt . (d) Toughness ratio,
wof / nbt .
Materials AZS2 and C present the maximum KIC 4. Discussion

values (1 MPa m1/2 ), followed by AS, AZS1 and S
(0.50.7 MPa m1/2 ); the lowest value corresponds to the insu- 4.1. Microstructure
lating firebrick (ASI) (Fig. 7a). The work of fracture does not
follow the same trend. The high-alumina castable (C), presents The nature and estimated relative amounts of phases identi-
the highest value, followed by AZS1. AZS2 and AS present fied by XRD (Table 3) were consistent with the main constituents
similar values, which are higher (40%) than those of the silica of the refractories according to their chemical and micro-
material, S. The lowest value is again found for the insulating structural analysis (Table 1, Figs. 25).
brick, ASI (Fig. 7b). In the AZS refractories appeared the typical zircon impurities,
Characteristic fracture surfaces are shown in Fig. 8. Both AZS Fe2 O3 and TiO2 in the form of ilmenite and rutile, and Y2 O3 and
materials presented tortuous fracture even though the aggregates HfO2 in solid solution. The presence of significant impurities
behaved differently. Most aggregates in AZS1 were traversed of Na2 O + K2 O in AZS1 material is attributed to the porous
by the crack (Fig. 8a) whereas most aggregates of AZS2 were alumina aggregates obtained by Bayer process. Very little alkali
surrounded by the main crack (Fig. 8b). Fracture in the matrix content was detected in AZS2 which had electrofused mullite
of the dense fireclay and insulating materials (AS and ASI) was aggregates (Tables 1 and 3, Fig. 2).
rather flat and the main crack traversed the aggregates in the Both dense fireclay and insulating materials, AS and ASI,
former (Fig. 8c and d). The fracture surface of the silica material contained significant amounts of the typical impurities for alumi-
(S, Fig. 8e) was the flattest with the main crack traversing all nosilicate natural raw materials: iron oxides (assumed as Fe2 O3
particles. The fracture surface of the high-alumina castable (C, for the chemical analyses), TiO2 , CaO and MgO. The presence of
Fig. 8f), presented opposite features, as it was tortuous and most relatively large amounts of K2 O compared to other alkalis sug-
aggregates were surrounded by the main crack. gests that these materials were manufactured with raw materials
140
Fig. 8. Characteristic fracture surfaces. Optical images for tested specimens (150mm 25 mm 25 mm). (a) Aluminasilicazirconia brick, AZS1. (b)
Aluminasilicazirconia brick, AZS2. (c) Superduty fireclay brick, AS. (d) Group 28 insulating brick, ASI. (e) Standard silica brick, S. (f) High-alumina regular
castable, C.
containing K-feldspar (K,Na,Ca,Ba)(Si,Al)4 O8 (Tables 1 and 3, mullite field of the ternary system and in the binary system
Fig. 3). Al6 Si2 O13 ZrO2 . Taking into account the average composition
Material S was a conventional silica brick (9597 wt.% of these refractories both should form stable liquid phases at
SiO2 , 2.53.5 wt.% CaO), well converted into cristobalite and temperatures higher than 1750 C (close to the temperatures of
tridymite. Wollastonite (CaSiO3 ) probably comes from the reac- the ZrO2 Al2 O3 Al6 Si2 O13 and ZrO2 Al6 Si2 O13 eutectics of
tion of lime with silica.46 It should be pointed out that a low the ternary and binary systems). However, temperature for first
content of CaO (typically 24 wt.%) may be used as binder liquid formation in the matrix would be lower (1555 C) due
in silica bricks without loss of refractoriness (Tables 1 and 3, to the formation of transient liquid phases at the zircon-mullite
Fig. 4). interfaces (temperature of the ZrSiO4 Al6 Si2 O13 SiO2 eutectic
The common impurities of bauxites and brown corundum,
iron oxides (Fe2 O3 in the chemical analyses) and TiO2 , were
present in the high-alumina castable, C, due to the nature of the
aggregates (Tables 1 and 3, Fig. 5).
The phosphorous (assumed as P2 O5 for calculations in the
chemical analyses) present in most materials is attributed to the
use of H3 PO4 (phosphoric acid), Al(H2 PO4 )3 (aluminium dihy-
drogen phosphate) or sodium polyphosphate as additives. In fact
AlPO4 was identified in AZS1, AS and ASI.
As a first approach for analysing the expected phases in
complex materials as a function of temperature, simplified
average compositions considering only the three major com-
ponents can be plotted in ternary phase equilibrium diagrams.
Fig. 9 shows simplified average compositions of both AZS
materials (AZS1 and AZS2) inside the ternary phase diagram
Al2 O3 ZrO2 SiO2 .47 AZS1 lies in the Al2 O3 primary field
of the ternary system while AZS2 is located in the primary Fig. 9. Phase equilibrium diagram of the system Al2 O3 ZrO2 SiO2 .47
141
1966
Table 6
Toughness parameters determined using CMOD controlled tests. KIC is the stress intensity factor in mode I calculated from the maximum load point, in MPa m1/2 ; nbt is the energy for crack initiation in J/m2 ; wof
is the work of fracture in J/m2 and wof / nbt is the toughness ratio.
150 mm 25 mm 25 mm 200 mm 40 mm 40 mm
= 0.25 = 0.50 = 0.25
KIC nbt wof wof / nbt KIC nbt wof wof / nbt KIC nbt wof wof / nbt
A. Garca-Prieto et al. / Journal of the European Ceramic Society 35 (2015) 19551970

AZS1 0.69 0.05 8 3 76 7 9 2 0.57 0.09 6 1 72 10 14 4
AZS2 1.0 0.1 10 5 60 7 6.0 0.5 1.0 0.2 10 3 52 4 6 3
AS 0.74 0.05 10 1 64 8 6.3 0.2 0.73 0.03 10 1 52 3 5.1 0.1 0.80 0.07 12 2 72 7 6.1 0.5
ASI 0.11 0.01 1.4 0.2 8 1 5.7 0.8 0.17 0.04 4 1 9 1 3 1 0.081 0.007 0.8 0.1 50.0 0.6 6.3 0.4
S 0.49 0.02 7 0.5 33 4 4.7 0.7
C 1.1 0.2 12 4 112 17 92 1.2 0.3 14 4 110 13 9 3 1.3 0.14 18 3 130 13 7.3 0.9
Table 7
Toughness parameters determined in displacement controlled tests. KIC is the stress intensity factor in mode I calculated from the maximum load point, in MPa m1/2 ; nbt is the energy for crack initiation in J/m2 ;
wof is the work of fracture in J/m2 and wof / nbt is the toughness ratio.
150 mm 25 mm 25 mm 200 mm 40 mm 40 mm
= 0.25 = 0.50 = 0.25
KIC nbt wof wof / nbt KIC nbt wof wof / nbt KIC nbt wof wof / nbt
AZS1 0.61 0.03 61 75 12 11 1 0.62 0.03 6.5 0.5 75 14 11 1

AZS2 1.1 12 66 5.24 1.05 0.08 11 2 62 4 5.44 0.87
AS 0.68 0.05 81 65 8 7.52 0.65 0.69 0.04 91 69 7 7.7 0.1
ASI
S 0.72 15 42 2.8 0.53 0.05 71 35 2 4.37 0.85
C 1.3 0.2 16 4 126 18 7.9 0.5 1.2 0.1 13.5 2.5 136 15 10 0.6
142
several processes for pore generation are well reported in the

literature.50 As described above, the average aggregate size in
the dense fireclay material (1200 mm, Table 4) was compara-
ble to the diameters of the main macrostructural characteristics
of the insulating firebrick which are the pores (Fig. 3c).
4.2. Mechanical behaviour
In order to correlate the microstructural characteristics of the

refractory materials with the Youngs modulus, E, obtained in
the present study, values for the fully dense materials, E0 , have
been calculated using the exponential equation derived from the
minimum solid area models (MSA)51 using the experimental E
values (Table 5) and the true porosities (Table 2).
Calculated E0 values for all materials (Table 5) are lower than
those expected for crystalline bonded zero porosity materials
constituted by the major crystalline phases detected (E0 220,
Fig. 10. Phase equilibrium diagram of the system Al2 O3 SiO2 TiO2 .49 400, 280, 7090 GPa for mullite,52 corundum,53 zircon53 and
silica polymorphs,54 respectively). This is a typical character-
point of the ternary system). The presence of impurities as TiO2 istic of refractories due to the weak bonding between the large
or iron could lower this temperature (1450 C for iron,48 1450 and inert aggregates and the fine matrix, as discussed below.
or 1500 C for TiO2 ).47 E0 for the silica brick is about one half of what could be
Considering the chemical (Table 1), the mineralogical analy- expected for a mixture of silica polymorphs (7090 GPa).54
sis (Table 3) and the microstructure (Fig. 2a and b) of AZS1, This fact can be explained considering the microstructure of
this material was formulated with aggregates of porous cal- this material, formed by large cristobalite particles surrounded
cined alumina (73 wt.%) and the source of ZrO2 was a pure by a transformation zone of tridymite + glass (Fig. 4). The large
ZrSiO4 sand (27 wt.%). No evidence of reaction between thermal expansion mismatch between the two crystalline phases
corundum and zircon was observed. The porosity of the brick ( 10 and 21 106 C1 for cristobalite and tridymite,
and the little or no reaction of ZrSiO4 with the finest parti- respectively)54 would lead to grain boundary cracking during
cles of Al2 O3 present in the matrix, suggest a temperature of cooling from the fabrication temperature. In fact, in the RLOM,
sintering lower than 1450 C (the lowest invariant point of the the boundaries present the typical dark colour of cracks (Fig. 4c).
Al2 O3 ZrO2 SiO2 TiO2 system).47 No cracks were observed in the AZS2 brick constituted
AZS2 (Tables 1 and 3, Fig. 2ce) was fabricated with by phases with similar thermal expansion coefficients ( 4.1
70 wt.% of electrofused mullite. Electrofusion originates mul- and 4.5 106 C1 , for mullite52 and zircon,55 respectively).
lite of composition 2:1, in agreement with the presence of some Even though there is a significant thermal expansion mis-
remnant corundum inside the aggregates (Fig. 2d). From the match between the major phases present in AZS1 (zircon
average composition and the microstructure of this material and and corundum-alumina, 4.5 55 and 8.4 106 C1 52 ) no
the composition of the aggregates (mullite + alumina), it can be cracks were observed in AZS1 specimens (Fig. 2a and b), which
concluded that the source of ZrO2 was 30 wt.% of angular par- can be explained by the low stiffness of the porous aggregates
ticles of zircon flour of sizes smaller than 50 mm. The presence and weakly bonded to the matrix. These two latter characteris-
of glass at the zircon/alumina interfaces (Fig. 2e) would indicate tics will be responsible for the extremely low E0 of AZS1 as
a sintering temperature higher than 1450 C47 (Fig. 9). compared to AZS2, contrary to what could be expected from
The simplified average compositions of the fireclay and the the major crystalline phases present in the materials (AZS1:
insulating materials (AS and ASI, Table 1) are displayed inside corundum + zircon, AZS2: mullite + zircon + baddeleyite).
the ternary phase diagram Al2 O3 SiO2 TiO2 in Fig. 10.49 The There are also differences between E0 for the dense fireclay
compositions of both materials lie in the primary field of mul- and the insulating materials, which presented similar chemi-
lite in the binary system Al6 Si2 O13 SiO2 (eutectic at 1595 C) cal composition and crystalline phases, which can be attributed
and in the ternary system Al6 Si2 O13 SiO2 Al2 TiO5 (eutectic at to the presence of stiff aggregates in the dense material. The
1480 C). Therefore, both materials would form liquid phases stiffest constituents have a determining influence on the Youngs
at temperatures higher than 1480 C. Taking into account the modulus values obtained from the response of specimens to the
microstructures and compositions of these materials, both were small and instantaneous deformations associated with the impact
formulated by mixing a kaolinitic chamote (aggregates) with test.56
plastic refractory clay. As the matrix of the fireclay brick had The relative performances of the materials in terms of mod-
similar composition as that of the chamote aggregates, a good ulus of rupture (MOR) and Youngs modulus (E) were similar
bonding between the aggregates and the matrix was observed. (Table 5). This behaviour is characteristic of refractories with
The porosity of the insulating firebrick (ASI) was due to aggregate sizes within the typical ranges (30006000 mm). Dif-
the addition of pore generators to the fireclay formulation; ferences in the MOR of such materials are determined by E
143
because they present only slightly differences in KIC , and the a consequence, the toughness ratio was much higher for
effect of the critical defect size is masked by the dispersion of AZS1.
results associated with the statistical character of strength. (2) AS and ASI had matrices of similar fireclay composition but
The characteristics of the loaddisplacement curves with very different microstructural features of similar sizes. The
monotonous decrease of the load for increasing displacement major microstructural characteristics of these two materials
during fracture are those corresponding to materials with ris- were chamote aggregates firmly bonded to the matrix in AS
ing R-curve during fracture. This material characteristic allows and pores in ASI (Figs. 3ad and 7ad).
obtaining stable fracture in displacement control for sufficiently AS presented nbt and wof values similar to the clas-
stiff and performing machines, as observed (Fig. 6b and c). sical ones reported by Nakayama for fireclay refractories4
Such fracture behaviour is reflected in the fact that values of ( nbt = 10.6 J/m2 , wof = 60 J/m2 ), however the nbt values
the energy for crack initiation, nbt , are lower than those averag- for AS were lower than those for mullite-rich laboratory
ing the whole fracture process, wof , as occurs for all the studied mixes31 ( nbt = 1825 J/m2 ) and lower than those reported
materials (Fig. 7b and c), and is highlighted by the toughness for similar compositions with improved matrix due to the
ratio values higher than 1 (Fig. 7d). addition of medium particles33 ( nbt = 44 J/m2 ).
The creation or not of the fully developed process zone during All toughness parameters, KIC , nbt and wof , of the
testing depends on the scale of the microstructural features of the porous material (ASI) were much lower (1520%) than
material tested and the experimental set up. The fact that there those of the dense one (AS), as expected. The pores in ASI
are no significant differences between the values of the work of act as stress intensity sinks during fracture which leads to
fracture obtained using different notch sizes for any material, an increased fracture energy as compared to that for crack
indicates that, even for the smallest specimens with the largest initiation. Therefore, the toughness ratio for this material
notches used in this work (specimens of 25 mm 25 mm section was about 60% of that of AS.
and = 0.50), the fracture area is sufficiently large to generate the (3) Composition and microstructure of S and C materials
steady state of fracture. As discussed in the introduction, rather were totally different. S was formed by particles of equal
small specimen sizes relative to the microstructural features have composition whereas C presented a well differentiated
been found adequate for SENB in 3-pb testing heterogeneous cementitious matrix and sintered alumina aggregates. The
materials as asphalt concrete.34,35 large differences in the toughness ratios of these materi-
In order to highlight the influence of the microstructural char- als suggest that the inelastic crack propagation processes
acteristics on the mechanical parameters summarised in Fig. 7, of these two refractory types are markedly different,6 as
the six materials have been compared in terms of their mechan- corresponds to the large differences in the microstructural
ical behaviour and their microstructural features, according to features. The crack path was straight, traversing all grains,
the following considerations: in the single-composition material S and highly tortuous in
the high-alumina castable C. In this latter, the aggregates
(1) AZS1 and AZS2 were two AZS materials with aggregates were surrounded by the cracks because they were tougher
with similar size distributions (Table 4) and extremely dif- than the matrix.
ferent physicochemical and microstructural characteristics
of the constituents. Aggregates were porous alumina for In general, work of fracture values for oxide refractories run
AZS1 and dense mullite for AZS2. The matrix in AZS1 from about 30 J/m2 for the most brittle ones to 100120 J/m2 for
was formed by fine alumina particles and represented a small those with well-designed microstructures, as the castable stud-
part of the total volume of the material in which a significant ied here.1 This material (C) presented wof values similar to
amount of medium size zircon particles was present (Fig. 2a- those reported for a series of high alumina castables (7090
b). In AZS2 there was a high amount of matrix composed alumina wt.%, 110120 J/m2 , 7 115 J/m2 )22 with designed
by mullite, zircon and zirconia particles (Fig. 2c-e). Aggre- microstructures for thermal shock. Those materials included
gates were weakly linked to the matrix in AZS1 and strongly high strength zirconia mullite aggregates which conferred the
bonded to the matrix in AZS2. AZS1 presented KIC and wof castables extremely high values and, consequently, lower tough-
values similar to those reported for alumina-zircon refracto- ness ratios ( wof / nbt 24)7 than the value obtained here for
ries with similar phase composition32 (KIC = 0.62 MPa m1/2 material C.
and wof = 66 J/m2 ).The fracture path in the materials is From the above discussion, it is clear that the characteristics
determined by the characteristics of the aggregates, the of the aggregates determine the toughness values for fracture
matrix and the matrix-aggregate bonds. In this way, the weak initiation, KIC and nbt (Fig. 7a and c). For similar crystalline
porous aggregates present in AZS1 were easily traversed by phase composition (Table 3), alumina, the two extreme cases
the crack whereas the dense ones in AZS2 were surrounded. would be that of the castable, with well sintered alumina aggre-
The dense aggregates with higher strength present in AZS2 gates (Fig. 5a and b) and material AZS1 (Fig. 2a and b), in which
impeded the initiation of fracture leading to KIC and nbt the aggregates presented high levels of porosity. The insulat-
values for this material higher than those for AZS1. The ing material, in which the aggregates were substituted by pores,
medium size zircon particles in AZS1 were responsible for would be the lowest limit for this trend.
crack arrest and deflection during propagation leading to Differently than in the case of fracture initiation, the
higher values of work of fracture than that of AZS2. As presence of microstructural features capable for crack arrest
144
and deflection is needed for high resistance to crack propa- 7. Salvini VR, Pandolfelli VC, Bradt RC. Extension of Hasselmans thermal
gation, wof (Fig. 7b). The microstructure that most clearly shock theory for crack/microstructure interactions in refractories. Ceram Int
2012;38(7):536975.
demonstrates this fact is that of material AZS1 (Fig. 2a and
8. Harmuth H, Tscheg EK. A fracture mechanics approach for the development
b), with medium-sized dense particles dispersed in the matrix of refractory materials with reduced brittleness. Fatig Fract Eng Mater Struct
and the highest toughness ratio (Fig. 7d). Materials AS and 1997;20(11):1585603.
S are examples of the opposite trend. These materials, with 9. Harmuth H, Bradt RC. Investigation of refractory brittleness by fracture
homogeneous composition through the microstructure and mechanical and fractographic methods. Refract Man 2010:610.
10. Palmer GB, Baker G. A load-cycling technique for R-curve behaviour:
well bonded microstructural constituents presented the lowest
application to a low cement refractory. Int J Fract 1993;62(3):23344.
toughness ratios of the dense materials. 11. Bueno S, Baudn C. Mechanical behaviour of structural ceramics. Bol Soc
The effectiveness of the microstructural elements for tough- Esp Ceram V 2007;46(3):10318.
ening is determined, not only by their nature, but also by the 12. Sakai M, Urashima K, Inagaki M. Energy principle of elastic-plastic fracture
characteristics of their bonding to the matrix. Material C is the and its application to the fracture mechanics of a polycrystalline graphite. J
Am Ceram Soc 1983;66(12):86874.
typical case of strong aggregates weakly bonded to the matrix
13. Nakayama J. Direct measurement of fracture energies of brittle heteroge-
and presents the highest wof of all studied materials (Fig. 7b). neous materials. J Am Ceram Soc 1965;48(11):5837.
14. Tattersall HG, Tappin G. The work of fracture and its measurement in metals,
ceramics and other materials. J Mater Sci 1966;1(3):296301.
5. Conclusions 15. Davidge RW, Tappin G. The effective surface energy of brittle materials. J
Mater Sci 1968;3(2):16573.
The capability of stable fracture tests of SENB tested in 3 16. Larson DR, Coppola JA, Hasselman DPH, Bradt RC. Fracture tough-
ness and spalling behavior of high-Al2 O3 refractories. J Am Ceram Soc
point bending to characterise toughness of refractories has been
1974;57(10):41721.
demonstrated. When carefully performed using the high stiff- 17. Chlup Z, Boccaccini DN, Leonelli C, Romagnoli M, Boccaccini AR.
ness and performing machines nowadays available it is possible Fracture behaviour of refractory ceramics after cyclic thermal shock. Ceram-
to establish displacement controlled tests as routine laboratory Silikaty 2006;50(4):24550.
tests for stable fracture. For the typical microstructural char- 18. Ko YC. Influence of the total fines content on the thermal shock damage
resistance of Al2 O3 -spinel castables. Ceram Int 2001;27(5):5017.
acteristics of commercial refractories studied here, standard
19. Tschegg EK, Austria Patent Specification AT 390,328 (1986).
size specimens (150 mm 25 mm 25 mm) tested with span 20. Brhwiler E, Wittmann FH. The wedge splitting test, a new method
125 mm and relative notch length ( = 0.5) give differentiated of performing stable fracture mechanics tests. Eng Fract Mec
toughness values for different microstructures. 1990;35(1/2/3):11725.
The main microstructural features that influence the resis- 21. Harmuth H. Stability of crack propagation associated with fracture
energy determined by wedge splitting specimen. Theor Appl Fract Mec
tance of materials to initiation of fracture are different from
1995;23(1):1038.
those that regulate crack propagation. The characteristics of the 22. Ribeiro S, Rodrigues JA. The influence of microstructure on the maxi-
aggregates determine toughness for crack initiation while, for mum load and fracture energy of refractory castables. Ceram Int 2010;36:
high values of work of fracture the presence of microstructural 26374.
features capable for crack arresting and deflection are needed. 23. Ribeiro S, Ribeiro DC, Souza Dias MB, Ribeiro Garcia GC, Bento dos
Santos EM. Study of the fracture behavior of mortar and concretes with
crushed rock or pebble aggregates. J Mater Res 2011;14(1):4652.
24. Harmuth H, Manhart C. Fractographic investigations of ordinary
Acknowledgements
ceramic refractory materials with reduced brittleness. Key Eng Mater
2009;409:20915.
The authors acknowledge the financial support of CYTED 25. Huilin L, Guifang D. Application of wedge splitting test in refractories.
through the network ref. 312RT0453 and project MAT2013- Chinas Refract 2006;15(2):2730.
48426-C2-1R; and the supplying of materials by INSERTEC 26. Brochen E, Dannert C. Thermo-mechanical characterisation of magnesia-
carbon refractories by means of wedge splitting test under controlled
S.A. (Spain). Ana Garca-Prieto acknowledges the financial sup-
atmosphere at high-temperature. In: Proceedings of 13th biennial worldwide
port of the JAE-CSIC fellowship program JAEPre 2010 00274. conference on refractories, UNITECR 2013. 2013.
27. Jin S, Gruber D, Harmuth H. Determination of Youngs modulus, fracture
energy and tensile strength of refractories by inverse estimation of a wedge
References splitting procedure. Eng Frac Mec 2014;116:22836.
28. Skiera E, Malzbender J, Mnch J, Dudczig S, Aneziris CG, Steinbrech
1. Bradt RC. Fracture of refractories. In: Schacht CA, editor. Refractories RW. Controlled crack propagation experiments with a novel alumina-based
handbook. New York: Marcel Dekker Inc; 2004. p. 1138. refractory. Adv Eng Mater 2012;14(4):24854.
2. Hasselman DPH. Figures-of-merit for the thermal stress resistance of high- 29. Miyaji DY, Otofuji CZ, de Anchieta Rodrigues J. The loaddisplacement
temperature brittle materials: a review. Ceram Int 1978;4(4):14750. curve of steady crack propagation: an interesting source of information for
3. Hasselman DPH. Thermal stress resistance of engineering ceramics. Mater predicting the thermal shock damage of refractories. In: Proceedings of 13th
Sci Eng 1985;71:25164. biennial worldwide conference on refractories, UNITECR 2013. 2013.
4. Nakayama J, Abe H, Bradt RC. Crack stability in the work-of-fracture test: 30. Malzbender J, Skiera E, Mnch J. A simple methodology to visualize crack
refractory applications. J Am Ceram Soc 1981;64(11):6715. propagation for ceramic materials. J Mater Sci 2014;49:4036.
5. Harmuth H, Rieder K, Krobath M, Tschegg E. Investigation of the nonlinear 31. Aksel C. The effect of mullite on the mechanical properties and ther-
fracture behaviour of ordinary ceramic refractory materials. Mater Sci Eng mal shock behaviour of aluminamullite refractory materials. Ceram Int
A 1996;214:5361. 2003;29:1838.
6. Homeny J, Darroudi T, Bradt RC. J-Integral measurements of the fracture 32. Rendtorff NM, Aglietti E. Mechanical and thermal shock behavior of refrac-
of 50% alumina refractories. J Am Ceram Soc 1980;63(56):32631. tory materials for glass feeders. Mater Sci Eng A 2010;527:38407.
145
33. Aksel C. The role of fine alumina and mullite particles on the thermo- 44. International Centre for Diffraction Data (ICDD). Newton Square, USA:
mechanical behaviour of aluminamullite refractory materials. Mater Lett Powder Diffraction File Database; 1998.
2002;57:70814. 45. EN 993-6. Methods of test for dense shaped refractory products. Part 6:
34. Romero P, Masad E. Relationship between the representative volume ele- Determination of modulus of rupture at ambient temperature; 1995.
ment and mechanical properties of asphalt concrete. ASCE J Mater Civ Eng 46. Lee WE, Rainforth WM. Ceramic microstructures property control by
2001;13:7784. processing. In: Refractories. London: Chapman & Hall; 1994. p. 452507.
35. Wagoner MP, Buttlar WG, Paulino GH. Development of a single-edge 47. Pena P, De Aza S. The zircon thermal behaviour: effect of impurities. Part
notched beam test for asphalt concrete mixtures. J Test Eval 2005;33:19. 1. J Mat Sci 1984;19:13542.
36. Peret CM, Rodrigues JA. Stability of crack propagation during bending tests 48. Jones TS, Kimura S, Muan A. Phase relations in the system
on brittle materials. Cermica 2008;54:3827. FeOFe2 O3 ZrO2 SiO2 . J Am Ceram Soc 1967;50(3):13742.
37. Garca-Prieto A, Baudn C. Crack mouth opening displacement controlled 49. Agamawi YM, White J. System Al2 O3 SiO2 TiO2 ; high silica region. Trans
fracture tests of brittle ceramics. J Eur Ceram Soc 2010;30:3297302. Br Ceram Soc 1952;51(5):293325.
38. Garca-Prieto A, Hernndez J, Lpez M, Baudn C. Controlled fracture test 50. Carniglia SC, Barna GL. Principles, types, properties and applications. In:
for brittle ceramics. J Strain Anal Eng 2011;46(1):2732. Handbook of industrial refractories technology. Noyes Publications; 1992.
39. Garca-Prieto A, Baudn C. Influence of experimental variables on fracture p. 3167.
toughness determined on SEVNB in three points bending. Mullite a case 51. Rice RW. Porosity of ceramics. Marcel Dekker, Inc; 1998. p. 102.
study. J Eur Ceram Soc 2012;32(16):42418. 52. Burgos-Montes O, Moreno R, Baudn C. Effect of mullite additions on the
40. Guinea GV, Pastor JY, Planas J, Elices M. Stress intensity factor, com- fracture mode of alumina. J Eur Ceram Soc 2010;30:85763.
pliance and CMOD for a general three-point-bend beam. Int J Fract 53. Montel JM, Giot R. Fracturing around radioactive minerals: elastic model
1998;89:10316. and applications. Phys Chem Miner 2013;40:63545.
41. Irwin GR. Fracture. In: Encyclopedia of physics. Berlin: Springer Verlag; 54. Pabst W, Gregorov E. Elastic properties of silica polymorphs a review.
1958. p. 55169. Ceram-Silikty 2013;57(3):16784.
42. EN 993-1. Methods of test for dense shaped refractory products. Part 1: 55. Bayer G. Thermal expansion of ABO-compounds with zircon and scheelite
Determination of bulk density, apparent porosity and true density; 1995. structures. J Less-Common Metals J 1972;26:25562.
43. EN 993-2. Methods of test for dense shaped refractory products. Part 2: 56. Lawn BR. Fracture of brittle solids. Cambridge: Cambridge University
Determination of true porosity; 1995. Press; 1993.
146
Captulo 7
TENACIDAD DE FRACTURA DE NANOCOMPUESTOS DE Al2O3/SiC
Captulo 7
7. TENACIDAD DE FRACTURA DE NANOCOMPUESTOS DE
Al2O3/SiC
Una vez demostrada la validez del dispositivo experimental, la metodologa de
ensayo y anlisis de resultados para la determinacin precisa de la tenacidad de fractura
de materiales cermicos frgiles (captulos 3-5) y de materiales en los que tienen lugar
mecanismos de refuerzo (captulo 6), se procedi a su aplicacin para la caracterizacin
de materiales cermicos avanzados de alto inters tecnolgico como son los
nanocompuestos de Al2O3/SiC. Como se mencion en el captulo 1, apartado 1.8, son
uno de los sistemas ms estudiados, pero, sin embargo, no se han reportado valores de
tenacidad de fractura concluyentes puesto que la mayora de los valores se han
determinado a partir de ensayos inestables o semiestables mediante tcnicas de
indentacin (IT).
Con objeto de comparar los valores obtenidos en este trabajo con los reportados por
diferentes autores, se calcul KIC, a partir de la ecuacin 7.1:
1/ 2
2 Eg
K I Cg =
1 - n
2
(7.1)
donde E es el mdulo de Young, g es la energa de fractura, n es el coeficiente de
Poisson.
El procesamiento de estos materiales se llevo a cabo durante una estancia de seis
meses otorgada por la beca JAE_Pre_2010_00274 en el laboratorio del Departamento
de Materiales de la Universidad de Oxford, bajo la direccin del Prof. Dr. Richard I.
Todd.
En la actualidad se est elaborando una publicacin conjunta con la Universidad de
Oxford.
149
Tenacidad de fractura de nanocompuestos de Al2O3/SiC
Los nanocompuestos de Al2O3/SiC fueron obtenidos a partir de polvos comerciales
de a-almina y a-SiC mediante molienda de bolas en medio acuoso seguido de una
liofilizacin y una calcinacin a 400C. Los polvos resultantes se prensaron en caliente
a 1650C durante 30 minutos en atmsfera de argon a 25 MPa. Se obtuvieron materiales
con un porcentaje de SiC de 2, 5 y 10 % en volumen. Las caractersticas
microestructurales de los mismos aparecen reflejadas en la tabla 7.1 donde adems se
han aadido los valores pertenecientes a la almina caracterizada en el captulo 4. El
tamao de grano de esta almina es similar al de la matriz de estos nanocompuestos.
Tabla 7.1. Caractersticas microestructurales y propiedades mecnicas de los

nanocompuestos de Al2O3/SiC caracterizados en este trabajo y de la almina de
referencia (captulo 4).
T.D.(S.D.) E (S.D.) sf (S.D.) s0 m
(%) (GPa) (MPa) (MPa)
Al2O3 98.1 (0.3) 387 (3) 456 (29) - -
Al2O3/2vol.% SiC 99.7 (0.3) 401(5) 561.2 (35.8) 577.2 18.3
Al2O3/5vol.% SiC 99.0 (0.3) 395 (4) 531.6 (63.9) 558.5 9.8
Al2O3/10vol.% SiC 98.8 (0.3) 390 (3) 522.6 (52.2) 545.2 11.9
T.D.: Densidad terica (%); E: Mdulo de Young (GPa); sf: Tensin de rotura
(MPa); s0: Tensin caracterstica (MPa); m: Parmetro de Weibull. (S.D.): Desviacin
estndar.
de cermicas tcnicas avanzadas (apartado 3.2.1). Las caractersticas de las entallas
fueron a=0.5 y <10 m. Una vez mecanizadas las entallas, las probetas fueron
sumergidas en aceite de silicona para evitar as el crecimiento subcrtico de grietas por
efecto de la humedad del ambiente.
Se realizaron ensayos a velocidades constantes de apertura de los labios de la grieta
(CMOD) de 1.8 m/min.
150
Captulo 7
7.2. Resultados y disusin
En la figura 7.1 se muestran las curvas experimentales caractersticas carga-tiempo
de los ensayos realizados utilizando el CMOD como variable de control. Se aprecia
cmo, una vez alcanzado el punto de carga mxima, ocurre una disminucin montona
de la carga con el tiempo conforme la grieta va creciendo. Como ya se ha discutido, este
tipo de comportamiento es caracterstico de ensayos de fractura estable.
60 60
Al2O3/5 vol.% SiC 7
Al2O3/2 vol.% SiC 2 Al2O3/5 vol.% SiC 5
50 50
Al2O3/2 vol.% SiC 3
40 40
Carga [N]
Carga [N]
30 30
20 20
10 10
0 0
0 100 200 300 400 500 0 100 200 300 400 500
Tiempo [s] Tiempo [s]
60
Al2O3/10 vol.% SiC 5
50
40
Carga [N]
30
20
10
0
0 100 200 300 400 500
Tiempo [s]
Figura 7.1. Curvas experimentales caractersticas carga-tiempo para las probetas de

los nanocompuestos de Al2O3/SiC:
(a) Al2O3/ 2 vol.%SiC
(b) Al2O3/ 5 vol.%SiC
(c) Al2O3/ 10 vol.%SiC
La figura 7.2 representa las curvas experimentales caractersticas carga-
desplazamiento de los materiales nanocompuestos. Se observa cmo, a partir del valor
151
de carga mxima, tiene lugar una disminucin del desplazamiento (rgimen I) seguida
de un aumento del desplazamiento (rgimen II). Este tipo de comportamiento es
caracterstico de fractura estable de materiales frgiles.
60 Al2O3/2 vol.% SiC 2 60 Al2O3/5 vol.% SiC 7

Al2O3/2 vol.% SiC 3 Al2O3/5 vol.% SiC 5
50 50
40 40
Carga [N]
Carga [N]
30 30
20 20
10 10
0 0
0.000 0.005 0.010 0.015 0.020 0.000 0.005 0.010 0.015 0.020
Desplazamiento [mm] Desplazamiento [mm]
60
50 Al2O3/10 vol.% SiC 4
40
Carga [N]
30
20
10
0
0.000 0.005 0.010 0.015 0.020
Desplazamiento [mm]
Figura 7.2. Curvas experimentales caractersticas carga-desplazamiento para las

probetas de los nanocompuestos de Al2O3/SiC:
(a) Al2O3/ 2 vol.%SiC
(b) Al2O3/ 5 vol.%SiC
(c) Al2O3/ 10 vol.%SiC
En las figuras 7.1 y 7.2 se observan curvas similares para probetas diferentes.
En la figura 7.3 se comparan los valores de KIC obtenidos en este trabajo mediante
ensayos de fractura estable de probetas SEVNB con los valores reportados por
diferentes autores utilizando mtodos experimentales diferentes a IT (apartado 1.8, tabla
1.5). Se observa cmo los valores de KIC son superiores y presentan mayor variabilidad
152
Captulo 7
para ensayos inestables de probetas SENB con grandes radios de curvatura181 que para
los ensayos estables de probetas SEVNB realizados en este trabajo.
Con objeto de comparar los valores de la tenacidad asociados al proceso de
fractura, se ha definido KICg como el calculado a partir de los valores de GF (=2g)
utilizando la ecuacin 7.1.
Los valores de KICg determinados a partir de ensayos de probetas SEVNB, en
ausencia de aceite de silicona, coinciden con los obtenidos para probetas CT277.
Demostrando la existencia de SCG durante la fractura de estos materiales en aire. Por lo
tanto, es preciso analizar los resultados de ensayos en aceite de silicona. Los valores de
KICg son inferiores a los de KIC, hecho no explicable en ausencia de fatiga esttica.
Para explicar este hecho es preciso considerar el efecto de las entallas utilizadas en
estos materiales, que presentan fractura transgranular, como se discute a continuacin.
4.0 Davidge, 1997

(SENB)181
KIC[MPam ] / KICg[MPam ]
1/2
3.5
(SEVNB) con aceite de silicona

3.0
(SEVNB)
1/2
2.5
2.0 Hoffman and Rdel 1997 (CT)277
1.5
0 3 6 9 12
% vol. SiC
Figura 7.3. Representacin de los valores de tenacidad correspondientes al inicio de

fractura, KIC, y a todo el proceso de fractura, KICg, para las probetas SEVNB de
nanocompuestos de Al2O3/SiC caracterizados en este trabajo. Asimismo, se recogen
datos correspondientes a ensayos SENB181 y CT277 (tabla 1.5).
153
En el round robin de SEVNB58 (apartado 1.3.2) se estudiaron cermicas tcnicas
como Si3N4 y Al2O3, con un tamao de grano superior a 2 mm. Para estos materiales se
observ fractura intergranular y que los bordes de grano actuaban como fisuras, este
comportamiento se asemeja al modelo de Fett56 donde el radio del fondo de entalla era
tan pequeo que se poda atribuir el inicio de la fractura a una fisura en la punta de la
entalla (apartado 1.3.2) y es muy diferente al comportamiento observado para los
nanocompuestos de Al2O3/SiC caracterizados en este trabajo.
En los nanocompuestos se produjo una fractura transgranular desde la punta de la
entalla, donde no hay fisuras (figura 1.4), por lo que el inicio de la fractura a partir de la
entalla en estos materiales no es comparable con el modelo de Fett56 (apartado 1.3.2). En
cambio para la almina de referencia el modo de fractura fue mixto inter-transgranular
donde los bordes de grano s actuaran como fisuras (figura 7.5).
Figura 7.4. Fractura transgranular caracterstica de los nanocompuestos de

Al2O3/SiC. Esta micrografa corresponde a una microscopa electrnica de barrido de
Al2O3/10 vol.%SiC.
154
Captulo 7
Figura 7.5. Fractura mixta inter-transgranular caracterstica de la almina.

Microscopa electrnica de barrido.
7.3. Conclusiones
- Se han ensayado materiales homogneos nanocompuestos de Al2O3/SiC en
ausencia de fatiga esttica.
- Se ha alcanzado fractura estable en ensayos de flexin en tres puntos de probetas
SEVNB de nanocompuestos de Al2O3/SiC.
- En los nanocompuestos el modo de fractura es transgranular, existiendo un efecto
de la influencia del radio de curvatura del fondo de entalla en la iniciacin de la
fractura, incluso para probetas SEVNB.
- Por lo tanto, para la determinacin precisa de la tenacidad es necesario el clculo
de la tenacidad a partir de todo el proceso de fractura (KIC), siendo pues necesaria la
realizacin de ensayos de fractura estable.
- Se ha detectado una dependencia no montona de la tenacidad con el contenido en
nanopartculas, mnima para Al2O3/5vol.%SiC.
155
156
ANEXO V
Este trabajo ha sido presentado en los siguientes congresos como conferencia

invitada:
- A. Garca-Prieto, T. Rodriguez-Suarez, R. I. Todd and C. Baudn. Toughness of
nanocomposites. XIII Conference of the European Ceramic Society. Limoges, Francia,
23-27 Junio 2013.
- A. Garca-Prieto, T. Rodriguez-Suarez, R. I. Todd y C. Baudn. Comportamiento

mecnico de nanocomposites cermicos. VII Congreso Internacional de Materiales. 29
Octubre-1 Noviembre 2013, Medelln, Colombia.
- A. Garca-Prieto, C. Baudn. Toughness of Al2O3/SiC Nanocomposites.

Conference for Young Scientists - The Tenth Students Meeting. November 6-9, 2013.
Novi Sad, Serbia.
Y aparece publicado en los libros colectivos Anales de Mecnica de la Fractura de

2013, que recogen los artculos que corresponden a las comunicaciones al XXX
Encuentro del Grupo Espaol de Fractura:
- A. Garca-Prieto, T. Rodrguez-Suarez, R.I. Todd, C. Baudn. Stable fracture
testing for toughness characterization of Al2O3/SiC nanocomposites. Anales de
Mecnica de la Fractura, 30, Vol.3, pp. 590-596. Ed. Secretara del Grupo Espaol de
Fractura, Madrid, Espaa, 2013.
157
158
Captulo 8
CONCLUSIONES GENERALES
Captulo 8
8. CONCLUSIONES GENERALES
Durante el desarrollo de este trabajo se ha llegado a las siguientes conclusiones
generales que surgen de las conclusiones parciales presentadas en los captulos 3-7:
- Se han desarrollado un nuevo dispositivo experimental y una metodologa de
ensayo y anlisis de resultados que permiten caracterizar la tenacidad de una
amplia gama de materiales cermicos, utilizando la apertura de los labios de la
grieta (CMOD) como parmetro de control.
- Se ha demostrado la posibilidad de alcanzar ensayos de fractura estable de
flexin en tres puntos de probetas prismticas de materiales cermicos con
entallas rectas en forma de V (SEVNB).
- Se ha demostrado la necesidad de los ensayos de fractura estable para la
determinacin de los parmetros de tenacidad de los siguientes materiales
cermicos:
- Espinela de aluminio-magnesio (MgAl2O4) y almina (Al2O3) de grano fino,
cuyo comportamiento en fractura es frgil (captulos 3 y 4).
- Mullita (3Al2O3.2SiO2), con comportamiento frgil y en la que tiene lugar
crecimiento subcrtico de grietas en aire significativo (captulo 5).
- Refractarios comerciales, en los cuales ocurren mecanismos de refuerzo
(captulo 6).
- Nanocompuestos de Al2O3/SiC (captulo 7), que se consideran como modelo
de los nuevos nanocompuestos cermicos.
- La utilizacin de la apertura de los labios de la grieta (CMOD) como parmetro
de control es necesaria para alcanzar ensayos estables en los materiales frgiles
de espinela de aluminio-magnesio, almina de grano fino, mullita con SCG
significativo en aire y nanocompuestos de Al2O3/SiC (captulos 3, 4, 5 y 7). Para
161
Conclusiones generales
los materiales refractarios comerciales, en los que ocurren mecanismos de
refuerzo (captulo 6), se han alcanzado ensayos estables utilizando el
desplazamiento del marco de carga como parmetro de control.
- En la tabla 8.1. se muestran los parmetros intrnsecos de la tenacidad (KIC, GIC,
GF=2g) determinados, por primera vez, en este trabajo para los materiales
cermicos avanzados caracterizados.
Tabla 8.1. Parmetros de tenacidad de fractura de los materiales cermicos

avanzados.
KIC (S.D.) GIC (S.D.) GF (S.D.)

Material
(MPa m1/2) (J/m2) (J/m2)
MgAl2O4 1.04 (0.06) 3.94 (0.01) -
Al2O3 2.5 (0.2) 16.4 (2.3) 14.0 (0.6)
3 Al2O3.2SiO2 0.86 (0.06) 3.5 (0.5) 3.8 (0.2)
Al2O3/2vol.%SiC 2.87 (0.14) 20 (1) 15 (1)
Al2O3/5vol.%SiC 2.89 (0.22) 17.99 (0.05) 13.0 (0.2)
Al2O3/10vol.%SiC 3.06 (0.02) 22.59 (0.05) 20.4 (0.4)
KIC: Factor de intensidad de tensiones en modo I (MPa m1/2); GIC: Tasa crtica de
liberacin de energa en modo I (J/m2); GF: Energa especfica de fractura.
162
Bibliografa
Bibliografa
BIBLIOGRAFA
1
R.J. Brook. Advanced Ceramic Materials: An Overview, p. 1-8 in Concise
Encyclopaedia of Advanced Ceramic Materials. Eds. R. J. Brook, R. W. Cahn, M. B.
Bever, Pergamon Press, Oxford (UK) (1991).
2
ASTM C1145:2013. Standard Terminology of Advanced Ceramics. ASTM
International, West Conshohocken (PA) (2013).
3
ISO 20507:2014. Fine ceramics (advanced ceramics, advanced technical ceramics).
Vocabulary.
4
EN ISO 15732:2005. Fine ceramics (advanced ceramics, advanced technical
ceramics). Test method for fracture toughness of monolithic ceramics at room
temperature by single edge precracked beam (SEPB).
5
R.W. Davidge. Fundamental of fracture mechanics, p. 31-50 in Mechanical Behaviour
of ceramics. Eds. R.W. Cahn, M.W. Thompson, I.M. Ward, Ed. Cambridge University
Press, Cambridge (UK) (1979).
6
G. de Portu, G.N. Balbini. Mechanical properties of ceramics, p. 21-49 in Designing
with structural ceramics. Eds. R.W. Davidge, M.H. Van de Voorde, Elsevier Science
Publishing, Essex (UK) (1991).
7
S. Bueno y C. Baudn. Mechanical behaviour of structural ceramics. Boletn de la
Sociedad Espaola de Cermica y Vidrio, 46 [3] 103-118 (2007).
165
Bibliografa
8
G.V. Guinea. http://www.mater.upm.es/-> Docencia-> Ingeniero de Materiales->
Propiedades Mecnicas: Comportamiento en Fractura.
9
A.A. Griffith. The Phenomena of Rupture and Flow in Solids. Philosophical
Transactions of the Royal Society of London A: Mathematical, Physical and
Engineering Sciences, 221 163-198 (1921).
10
R.W. Davidge. The Fracture Strength of Ceramics, p. 75-103 in Mechanical
Behaviour of Ceramics. Ed. R.W. Cahn, M.W. Thompson, I.M. Ward, Cambridge
University Press, Cambridge (UK) (1979).
11
E. Orowan. Energy criteria of fracture. Welding Research Suplement 34 157-160
(1955).
12
M. Sakai, R.C. Bradt. Fracture toughness testing of brittle materials. International
Materials Reviews 38 53-78 (1993).
13
R.W. Rice, S.W. Freiman, P.F. Becher. Grain Size Dependence of Fracture Energy in
Ceramics: I, Experiment. Journal of the American Ceramic Society 64 [6] 345-350
(1981).
14
G. Irwin. Analysis of Stresses and Strains Near the End of a Crack Traversing a Plate.
Journal of Applied Mechanics 24 361-364 (1957).
166
Bibliografa
15
H. Westergard. Bearing Pressures and Cracks. Journal of Applied Mechanics 6 49-53
(1939).
16
A.H. De Aza, J. Chevalier, G. Fantozzi, M. Schehl, R. Torrecillas. Crack growth
resistance of alumina, zirconia and zirconia toughened alumina ceramics for joint
prostheses. Biomaterials 23 937945 (2002).
17
G.R. Irwin. Fracture, p. 551-589 in Encyclopaedia of Physics. Ed. S. Flge, Springer
Verlag, Berlin (Germany) (1958).
18
D. Casellas, M.M. Nagl, M. Vlez, M. Anglada. Determinacin de la presencia de
fatiga mecnica en materiales cermicos. Boletn de la Sociedad Espaola de Cermica
y Vidrio 38 101-110 (1999).
19
N. Kondo, T. Ohji, F. Wakai. Strengthening and Toughening of Silicon Nitride by
Superplastic Deformation. Journal of the American Ceramic Society 81 [3] 71316
(1998).
20
J. Chevalier L. Gremillard, A.V. Virkar, D.R. Clarke. The tetragonal-monoclinic
transformation in zirconia: Lessons learned and future trends. Journal of the American
Ceramic Society 92 1901-1920 (2009).
21
K. Niihara. New Design Concept of Structural Ceramics. Ceramic Nanocomposites.
Journal of the Ceramic Society of Japan 99 974-982 (1991).
167
Bibliografa
22
S.M. Choi, H. Awaji. Nanocomposites-a new material design concept. Science and
Technology of Advanced Materials 6 2-10 (2005).
23
G. Evans. Perspective on the Development of High-Toughness Ceramics. Journal of
the American Ceramic Society 73 [2] 187-206 (1990).
24
K.T. Faber. Toughening mechanisms for ceramics in automotive applications.
Ceramic Engineering and Science Proceedings 5 [5-6] 408-439 (1984).
25
P.F. Becher. Microstructural design of toughened ceramics. Journal of the American
Ceramic Society 74 [2] 255-269 (1991).
26
R.W. Steinbrech. Toughening Mechanisms for Ceramic Materials. Journal of the
European Ceramic Society 10 131-142 (1992).
27
M.I. Osendi, J.S. Moya. Fundamentos de la transformacin martenstica. Su
importancia en el diseo de futuros materiales cermicos. Boletn de la Sociedad
Espaola de Cermica y Vidrio 21 [1] 33-40 (1982).
28
P.M. Kelly, L.R.F. Rose. The martensitic transformation in ceramics its role in
transformation toughening. Progress in Materials Science 47 [5] 463-557 (2002).
29
J. Rdel. Crack Closure Forces in Ceramics: Characterization and Formation. Journal
of the European Ceramic Society 9 323-334 (1992).
168
Bibliografa
30
P. Miranzo, J.S. Moya. Reforzamiento de materiales cermicos y vtreos por fibras.
Boletn de la Sociedad Espaola de Cermica y Vidrio, 27 [3] 145-151 (1988).
31
R.W. Steinbrech, A. Reichl, W. Schaarwchter. R-Curve Behavior of Long Cracks in
Alumina. Journal of the American Ceramic Society 73 [7] 2009-2015 (1990).
32
H.L. Ewalds, R.J.H. Wanhill. Stable Crack Growth and the R-Curve Concept, p. 86-
94 in Fracture Mechanics. Edward Arnold (Publishers) Ltd., London (UK) (1984).
33
M. Sakai, J. Yoshimura, Y. Goto, M. Inagaki. R-Curve Behaviour of a Polycrystalline
Graphite: Microcracking and Grain Bridging in the Wake Region. Journal of the
American Ceramic Society 71 [8] 609-16 (1988).
34
M. Sakai, K. Urashima, M. Inagaki. Energy Principle of Elastic-Plastic Fracture and
Its Application to the Fracture Mechanics of a Polycrystalline Graphite. Journal of the
35
G.D. Quinn, R.C. Bradt. On the Vickers indentation fracture toughness test. Journal
of the American Ceramic Society 90 673-680 (2007).
36
14425-1:2004 (E). Advanced technical ceramics - Test methods for determination of
fracture toughness of monolithic ceramics contains five parts: Part 1: Guide to test
method selection.
169
Bibliografa
37
fracture toughness of monolithic ceramics contains five parts: Part 2: Single-edge pre-
cracked beam (SEPB) method.
38
ISO 15732:2003 . Fine ceramics (advanced ceramics, advanced technical ceramics) -
Test method for fracture toughness of monolithic ceramics at room temperature by
single edge precracked beam (SEPB) method.
39
EN14425-3:2010. Advanced technical ceramics - Test methods for determination of
fracture toughness of monolithic ceramics contains five parts: Part 3: Chevron notched
beam (CNB) method.
40
fracture toughness of monolithic ceramics contains five parts: Part 4: Surface crack in
flexure (SCF) method.
41
ISO 18756:2003 Fine ceramics (advanced ceramics, advanced technical ceramics) -
Determination of fracture toughness of monolithic ceramics at room temperature by the
surface crack in flexure (SC F) method (ISO 18756:2003).
42
fracture toughness of monolithic ceramics contains five parts: Part 5: Single-edge vee-
notch beam (SEVNB) method.
170
Bibliografa
43
ISO 23146:2012. Fine ceramics (advanced ceramics, advanced technical ceramics) -
Test methods for fracture toughness of monolithic ceramics - Single-edge V-notch beam
(SEVNB) method.
44
J.Y. Pastor. Fractura de materiales cermicos estructurales avanzados, p.37-46. Tesis
doctoral. ETSI de Caminos, UPM, Madrid, 1993.
45
A. H. de Aza, J. Chevalier. Revisin de la tcnica de doble torsin y del mtodo
experimental en materiales cermicos. Boletn de la Sociedad Espaola de Cermica y
Vidrio 40 [2] 93-99 (2001).
46
D. Wan, Y. Bao, J. Peng, Y. Zhou. Fracture toughness determination of Ti3Si(Al)C2
and Al2O3 using a single gradient notched beam (SGNB) method. Journal of the
47
G.D. Quinn, J.J. Kbler, R.J. Gettings. Fracture toughness of advanced ceramics by
the surface crack in flexure (SCF) method. A Vamas Round Robin. Technical working
area 3 (1994).
48
K.E. Amin. Toughness, Hardness and Wear, Engineered Materials Handbook:
Volume 4, Ceramics and Glasses, ASM International, pp. 599-609 (1991).
49
T. Nishida, T. Shiono, A. Nagai, T. Nishikawa. Evaluation of fracture toughness for
structural ceramics using SEPB specimens. Journal of the Ceramic Society of Japan 96
608-612 (1988).
171
Bibliografa
50
R.J. Damani, R. Danzer. A method for fracture toughness testing of ceramics-Ready
for standardisation. 12th Biennial Conference on Fracture of the European-Structural-
Integrity-Society (ECF 12)-Fracture from defects 491-496 (1998).
51
ESIS TC6 Round Robin on fracture toughness. Fatigue Fracture Engineering
Materials Structure 20 513-532 (1997).
52
R. Damani, R. Gstrein, R. Danzer. Critical notch-root radius effect in SENB-S
fracture toughness testing. Journal of the European Ceramic Society 16 695-702 (1996).
53
S.S. Scherrer, I.L. Denry, H.W.A. Wiskot. Comparison of three fracture toughness
testing techniques using a dental glass and a dental ceramic. Dental Materials 14 246
255 (1998).
54
J. Wang, W.M. Rainforth, I. Wadsworth, R. Stevens. The effects of notch width on
the SENB toughness for oxide ceramics. Journal of the European Ceramic Society 10
21-31 (1992).
55
T. Nishida, Y. Hanaki, G. Pezzotti. Effect of notch-root radius on the fracture
toughness of a fine-grained alumina. Journal of the American Ceramic Society 77 606-
608 (1994).
56
T. Fett. Estimated stress intensity factors for semi-elliptical cracks in front of narrow
circular notches. Engineering Fracture Mechanics 64 35762 (1999).
172
Bibliografa
57
M. Turon-Vinas, M. Anglada. Fracture toughness of zirconia from a shallow notch
produced by ultra-short pulsed laser ablation. Journal of the European Ceramic Society
34 38653870 (2014).
58
ESIS Technical Ceramics TC6. VAMAS Technical Working Area 3. Fracture
toughness of ceramics using the SEVNB method: round robin (1999).
59
A. Ghosh, M.G. Jenkins, K.W. White, A.S. Kobayashi, R.C. Bradt. Elevated-
temperature fracture resistance of sintered alpha-silicon carbide. Journal of the
American Ceramic Society 72 242-247 (1989).
60
J. Nakayama. Direct Measurement of Fracture Energies of Brittle Heterogeneous
Materials. Journal of the American Ceramic Society 48 [11] 583-587 (1965).
61
J. Nakayama, H. Abe, R.C. Bradt. Crack Stability in the Work-of-Fracture Test:
Refractory Applications. Journal of the American Ceramic Society 64 [11] 671-675
(1981).
62
H.G. Tattersall, G. Tappin. The Work of Fracture and its Measurement in Metals,
Ceramics and other Materials. Journal of Materials Science 1 296-301 (1966).
63
A.M. Calomino, D.N. Brewer. Controlled crack growth specimen for brittle systems.
Journal of the American Ceramic Society 75 [1] 206-208 (1992).
173
Bibliografa
64
B.F. Sorensen, A. Horsewell, O. Jrgensen, A.N. Kumar, P. Engbk. Fracture
resistance measurement method for in situ observation of crack mechanisms. Journal of
the American Ceramic Society 81 [3] 661-669 (1998).
65
J.A. Kuszyk, R.C. Bradt, Influence of Grain Size on Effects of Thermal Expansion
Anisotropy in MgTi2O5. Journal of the American Ceramic Society 56 [8] 420-423
(1973).
66
B.F. Sorensen, P. Brethe, P. Skov-Hansen. Controlled crack growth in ceramics: The
DCB specimen loaded with pure moments. Journal of the European Ceramic Society 16
1021-1025 (1996).
67
M.E. Ebrahimi, J. Chevalier, G. Fantozzi. R-curve evaluation and bridging stress
determination in alumina by compliance analysis. Journal of the European Ceramic
Society 23 943-949 (2003).
68
S. Grendahl, R. Bert, K. Cho, I. Bar-On. Effects of Residual Stress and Loading
Geometry on Single-Edge Precracked-Beam (SEPB) Fracture-Toughness Test Results.
Journal of the American Ceramic Society 83 [10] 262527 (2000).
69
F.I. Baratta, W.A. Dunlay. Crack Stability in Simply Supported Four-Point and
Three-Point Loaded Beams of Brittle Materials. Mechanics of Materials 10 149-159
(1990).
174
Bibliografa
70
I. Bar-On, F.I. Baratta, K. Cho. Crack Stability and its Effect on Fracture Toughness
of Hot Pressed Silicon Nitride Beam Specimens. Journal of the American Ceramic
Society 79 [9] 2300-2308 (1996).
71
R.W. Davidge, G. Tappin. The effective Surface Energy of Brittle Materials. Journal
of Materials Science 3 165-173 (1968).
72
S. Bueno, M.H. Berger, R. Moreno, C. Baudn. Fracture behaviour of microcrack-free
alumina-aluminium titanate ceramics with second phase nanoparticles at alumina grain
boundaries. Journal of the European Ceramic Society 28 1961-1971 (2008).
73
LS. Sigl. On the stability of cracks in flexure specimens. International Journal of
Fracture 51 241-254 (1991).
74
L. Biolzi, S. Cangiano, G. Tognon, A. Carpinteri. Snap-back softening instability in
high-strength concrete beams. Materials and Structures 22 429-436 (1989).
75
J.Y. Pastor, J. Planas, M. Elices. Ambient and high-temperature stable fracture tests
in ceramics: Applications to Yttria-Partially-Stabilized Zirconia. Journal of the
76
J.Y. Pastor, J. Planas and M. Elices. A new technique for fracture characterization of
ceramics at room and at high temperature. Journal of Testing and Evaluation 23 [3] 209-
216 (1995).
175
Bibliografa
77
J.Y. Pastor, J. Planas, M. Elices. Ensayos de fractura estables en materiales
cermicos. Boletn de la Sociedad Espaola de Cermica y Vidrio 31[4] 322-325
(1992).
78
D. C. Harris. Materials for Infrared Windows and Domes, SPIE Press, Bellingham,
WA, (1999).
79
J.F. Shackelford, W. Alexander, J.S. Park, ed., CRC Materials Science and
Engineering Handbook. CRC Press, Boca Raton, Florida, (1994).
80
S.J. Burnett. Properties of Refractory Materials, UKAEA Research Group Report,
Harwell (1969).
81
C. Aksel, R.W. Davidge, P.D. Warren, F.L. Riley. Investigation of thermal shock
resistance in model magnesia-spinel refractory materials. In IV Ceramic Congress,
Proceedings Book, Part 1, Eskisehir, Turkey, pp. 19399, (1998).
82
O. Khasanov, E. Dvilis, A. Khasanov, E. Polisadova, A. Kachaev. Optical and
mechanical properties of transparent polycrystalline MgAl2O4 spinel depending on SPS
conditions. Physica Status Solidi C 10 [6] 918920 (2013).
83
R.L. Stewart, R.C. Bradt. Fracture of polycrystalline MgAl2O4. Journal of the
176
Bibliografa
84
C. Baudn, R. Martnez, P. Pena. High-temperature mechanical behavior of
stoichiometric magnesium spinel. Journal of the American Ceramic Society 78 1857-
1862 (1995).
85
C. Baudn, P. Pena. Influence of Stoichiometry on Fracture Behaviour of Magnesium
Aluminate Spinels at 1200C. Journal of the European Ceramic Society 17 1501-1511
(1997).
86
C. Aksel, B. Rand, F.L. Riley, P.D. Warren. Mechanical properties of magnesia-
spinel composites. Journal of the European Ceramic Society 22 745754 (2002).
87
C.B. Huang, T.C. Lu, L.B. Lin, M.Y. Lei, C.X. Huang. A study on toughening and
strengthening of MgAl spinel transparent ceramics. Key Engineering Materials 336
338 12071210 (2007).
88
S. Bhaduri, S.B. Bhaduri. Microstructural and mechanical properties of
nanocrystalline spinel and related composites. Ceramics International 28 153158
(2002).
89
A. Krell , T. Hutzler, J. Klimke. Transmission physics and consequences for materials
selection, manufacturing, and applications. Journal of the European Ceramic Society 29
207221 (2009).
90
M. Boniecki, Z. Librant, A. Wajler, W. Wesoowski, H. Weglarz. Fracture toughness,
strength and creep of transparent ceramics at high temperature. Ceramics International
38 45174524 (2012).
177
Bibliografa
91
A. Ghosh, K.W. White, M.G. Jenkins, R.C. Bradt. Fracture resistance of a transparent
MgAl2O4. Journal of the American Ceramic Society 74 1624-1630 (1991).
92
K.W. White, G.P. Kelkar. Fracture mechanisms of a coarse-grained, transparent
MgAl2O4 at elevated temperatures. Journal of the American Ceramic Society 75 3440-
3444 (1992).
93
O. Tokariev, R. W. Steinbrech, L. Schnetter, J. Malzbender. Micro- and macro-
mechanical testing of transparent MgAl2O4 spinel. Journal of Materials Science 47
48214826 (2012).
94
O. Tokariev, L. Schnetter, T. Beck, J. Malzbender. Grain size effect on the
mechanical properties of transparent spinel ceramics. Journal of the European Ceramic
Society 33 749757 (2013).
95
O. Borrero-Lpez, A.L. Ortiz, A.D. Gledhill, F. Guiberteau, T. Mroz, L.M.
Goldmanc, N.P. Padture. Microstructural effects on the sliding wear of transparent
magnesium-aluminate spinel. Journal of the European Ceramic Society 32 31433149
(2012).
96
T. Mroz, L.M. Goldman, A.D. Gledhill, D. Li, N.P. Padture. Nanostructured, infra-
red transparent magnesium-aluminate spinel with superior mechanical properties.
International Journal of Applied Ceramic Technology 9 83-90 (2012).
178
Bibliografa
97
K. Morita, B-N. Kim, K. Hiraga, H. Yoshida. Fabrication of high-strength
transparent MgAl2O4 spinel polycrystals by optimizing spark-plasma-sintering
conditions. Journal of Materials Research 24 [9] 2863-2872 (2009).
98
A. Krell, J. Klimke, T. Hutzler. Advanced spinel and sub-m Al2O3 for transparent
armour applications. Journal of the European Ceramic Society 29 275281 (2009).
99
R.L. Stewart, R.C. Bradt. Fracture of single crystal MgAl2O4. Journal of Materials
Science 15 67-72 (1980).
100
W. Cao, A. Kundu, Z. Yu, M.P. Harmer, R.P. Vinci. Direct correlations between
fracture toughness and grain boundary segregation behavior in ytterbium-doped
magnesium aluminate spinel. Scripta Materialia 69 8184 (2013).
101
C.C.M. Wu, K.R. McKinney, R.W. Rice. Zig-zag crack propagation in MgAl2O4
crystals. Journal of Materials Science Letters 14 474-477 (1995).
102
R.W. Rice, J.J. Mecholsky, Jr.P.F. Becher. Fracture and fracture toughness of
stoichiometric MgAl2O4 crystals at room temperature. Journal of Materials Science 31
1353-1360 (1996).
103
J. W. McCauley, P. Patel, M. Chen, G. Gilde, E. Strassburger, B. Paliwal, K.T.
Ramesh, D.P. Dandekar. AlON: A Brief History of Its Emergence and Evolution.
Journal of the European Ceramic Society, 29 [2] 223236 (2009).
179
Bibliografa
104
A. Ikesue, T. Kinoshita, K. Kamata, K. Yoshida. Fabrication and Optical Properties
of High-Performance Polycrystalline Nd: YAG Ceramics for Solid-State Lasers. Journal
of the American Ceramic Society 78 [4] 10331040 (1995).
105
G.C. Wei. Transparent Ceramic Lamp Envelope Materials. Journal of Physics D:
Applied Physics 38 30573065 (2005).
106
O. Tokariev, T. Van Gestel, M. Bram, J. Malzbender. Strength enhancement of
transparent spinel ceramics. Materials Letters 107 364366 (2013).
107
R.W. Rice, J.J. Mecholsky, Jr.P.F. Becher. The effect of grinding direction on flaw
caracter and strength of single crystal and polycrystalline ceramics. Journal of Materials
Science 16 853-862 (1981).
108
T.E. Mitchell. Dislocations and Mechanical Properties of MgOAl2O3 Spinel Single
Crystals. Journal of the American Ceramic Society 82 [12] 330516 (1999).
109
D.H. Chung and W.R. Buessem, in Anisotropy in single crystal refractory
compounds, Volume 2, edited by F.W. Valhldrik and S.A. Mersol. Plenum Press,
New York, p. 217 (1968).
110
G. Simmons, H. Wang, in Single crystal elastic constants and calculated aggregate
properties: A handbook. MIT Press, Cambridge, Massachusetts, (1971).
111
J. Briggs. Engineering ceramics in Europe and the USA. Worcester, UK: Menwith
Wood (2007).
180
Bibliografa
112
R.H. Doremus. Alumina. In J. F. Shackelford, & R. H. Doremus (Eds.), Ceramic and
glass materials structure, properties and processing. USA: Springer (2008).
113
E. Drre, H. Hbner. Alumina. Processing, properties and applications. Berlin,
Germany: Springer-Verlag (1984).
114
W.H. Gitzen. Alumina as a ceramic material. In American ceramic society special
publication 4. Ohio, USA: The American Ceramic Society (1970).
115
L.D. Hart. Alumina chemicals: Science and technology handbook. USA: The
American Ceramic Society (1990).
116
W.D. Kingery. Structure and properties of MgO and Al2O3 ceramics. Ohio, USA:
The American Ceramic Society (1984).
117
I. J. McColm. Ceramic hardness. New York, USA: Plenum Press (1990).
118
F.L. Riley. Structural ceramics. Fundamentals and case studies. Cambridge, UK:
Cambridge University Press (2009).
119
C. Baudn. Processing of alumina and corresponding composites in Comprehensive
Hard Materials edited by Vinod K. Sarin, Daniele Mari, Luis Llanes, Christoph E.
Nebel. Volume 231-72 (2014).
120
D. Taylor. Thermal expansion data: III. Sesquioxides, M2O3 with the corundum and
the A-, B- and C-M2O3. British Ceramic Transactions and Journal 83 9298 (1984).
181
Bibliografa
121
J.E. Blendell, R.L. Coble. Measurement of stress due to thermal expansion
anisotropy in Al2O3. Journal of the American Ceramic Society 65 174178 (1982).
122
S. Bueno, C. Baudn. Instrumented Vickers microindentation of alumina-based
materials. Journal of Material Research, 21, 161173 (2006).
123
B. Mussler, M.V. Swain, N. Claussen. Dependence of Fracture Toughness of
Alumina on Grain Size and Test Technique. Journal of the American Ceramic Society
65 566-572 (1982).
124
P.L. Swanson, C.J. Fairbanks, B.R. Lawn, Y-W. Mai, B.J. Hockey. Crack-Interface
Grain Bridging as a Fracture Resistance Mechanism in Ceramics: I, Experimental Study
on Alumina. Journal of the American Ceramic Society 70 [4] 279-89 (1987).
125
Y-W. Mai, B.R. Lawn. Crack-interface grain bridging as a fracture resistance
mechanism in ceramics: II, Theoretical fracture mechanics model. Journal of the
126
M.N. Rahaman, A. Yao, B.S. Bal, J.P. Garino, M.D. Ries. Ceramics for prosthetic
hip and knee joint replacement. Journal of the American Ceramic Society 90 [7] 1965-
1988 (2007).
127
O. Sbaizero, G. Pezzotti and T.Nishida. Fracture energy and R-curve behavior of
Al2O3/Mo composites. Acta Materialia, 46 [2] 681-687 (1998).
182
Bibliografa
128
J. Chevalier, L. Gremillard. Ceramics for medical applications: A picture for the
next 20 years. Journal of the European Ceramic Society 29 12451255 (2009).
129
P. Palmero, L. Montanaro, H. Reveron, J. Chevalier. Review: Surface Coating of
Oxide Powders: A New Synthesis Method to Process Biomedical Grade Nano-
Composites. Materials 7 5012-5037 (2014).
130
N. Claussen, B. Mussler, M.V. Swain. Grain-Size Dependence of Fracture Energy in
Ceramics. Journal of the American Ceramic Society C14-C16 (1982).
131
S.W. Freiman, K.R. McKinney, and H.L. Smith, Slow crack growth in
polycrystalline ceramics in Fracture Mechanics of Ceramics, edited by R.C. Bradt,
D.P.H. Hasselman, and F.F. Lang (Plenum Press, New York, London), Volume 2, pp.
659676 (1973).
132
T.A. Michalske, B.C. Bunker, and S.W. Freiman, Journal of the American Ceramic
Society 69 721 (1986).
133
M.E. Ebrahimi, J. Chevalier, G. Fantozzi. Slow crack-growth behavior of alumina
ceramics. Journal of Materials Research 15 142-147 (2000).
134
S.M. Wiederhorn. Fracture of Sapphire. Journal of the American Ceramic Society 52
[9] 485-91 (1969).
183
Bibliografa
135
J. Seidel, J. Rdel. Measurement of crack tip toughness in alumina as a function of
grain size. Journal of the American Ceramic Society 80 [2] 433-38 (1997).
136
S. Bueno, C. Baudn. Layered materials with high strength and flaw tolerance based
on alumina and aluminium titanate. Journal of the European Ceramic Society 27 1455-
1462 (2007).
137
H. Schneider, J. Schreuer, B. Hildmann. Structure and properties of mullite-A
review. Journal of the European Ceramic Society 28 329-344 (2008).
138
T-I. Mah, K.S. Mazdiyasni. Mechanical properties of mullite. Journal of the
139
O. Burgos-Montes, R. Moreno, C. Baudn. Effect of mullite additions on the fracture
mode of alumina. Journal of the European Ceramic Society 30 857-63 (2010).
140
K.S. Mazdiyasni, L.M. Brown. Synthesis and mechanical properties of
stoicchiometric aluminum silicate (mullite). Journal of the American Ceramic Society
55 548-552 (1972).
141
P.C. Dokko, J.A. Pask, K.S. Mazdiyasni. High-Temperature mechanical properties
of mullite under compression. Journal of the American Ceramic Society 60 150-155
(1977).
142
R. Torrecillas, G. Fantozzi, S. de Aza, J.S. Moya. Thermomechanical behaviour of
mullite. Acta Materialia 45 897-906 (1997).
184
Bibliografa
143
E. Fernandez, C. Baudn. Creep damage in different 3Al2O3.2SiO2 mullites tested in
4-point bending. Journal of the European Ceramic Society 21 2243-2251 (2001).
144
S. Kanzaki, H. Tabata, T. Kumazawa, S. Ohta. Sintering and mechanical properties
of stoichiometric mullite. Journal of the American Ceramic Society 68 C6-C7 (1985).
145
C. Baudn. Fracture mechanisms in a stoichiometric 3Al2O3.2SiO2 mullite. Journal of
Materials Science 32 2077-2086 (1997).
146
Y. Okamoto, H. Fukudome, K. Hayashi, T. Nishikawa. Creep Deformation of
Polycrystalline Mullite. Journal of the European Ceramic Society 6 161-168 (1990).
147
M.G.M.U. Ismail, Z. Nakai, S. Somiya. Microstructure and mechanical properties of
mullite prepared by the sol-gel method. Journal of the American Ceramic Society 70
C7-C8 (1987).
148
M.I. Osendi, C. Baudn. Mechanical properties of mullite materials. Journal of the
149
R. Sivakumar, D. Doni Jayaseelan, T. Nishikawa, S. Honda, H. Awaji. Influence of
MgO on microstructure and properties of mullite-Mo composites fabricated by pulse
electric current sintering. Ceramics International 27 537-541 (2001).
185
Bibliografa
150
H. Ohira, M.G.M.U. Ismail, Y. Yamamoto, T. Akiba, S. Somiya. Mechanical
Properties of High Purity Mullite at Elevated Temperatures. Journal of the European
Ceramic Sociefy 16 225-229 (1996).
151
M. Itoh, K. Hamano, S. Okada. Preparation of mullite ceramics from kaolin and
aluminum hydroxide. In: Abstracts of the 3rd Autumn Symposium on Ceramics, Paper
6-2 A13. Ceramics Society of Japan, Tokyo.
152
Y. Kumazawa, S. Kanzaki, S. Otha, H. Tabata. Influence of chemical composition
on the mechanical properties of SiO2-Al2O3 ceramics. Journal of the Ceramic Society of
Japan 96 85-91 (1988).
153
D. Casellas, C. Baudn, M. Osendi, L. Llanes, M. Anglada. Fracture resistance of
mullite under static and cyclic loads. Scripta Materialia 38 39-44 (1998).
154
B. Subrata. Properties of Refractories in Refractories handbook edited by Charles A.
Schacht 1-10 (2004).
155
R. C. Bradt. Fracture of Refractories in Refractories handbook edited by Charles A.
Schacht 10-38 (2004).
156
H. Harmuth, K. Rieder, M. Krobath, E. Tschegg. Investigation of the nonlinear
fracture behaviour of ordinary ceramic refractory materials. Materials Science and
Engineering A214 53-61 (1996).
186
Bibliografa
157
J. Homeny, T. Darroudi, R.C. Bradt. J-Integral measurements of the fracture of 50%
alumina refractories. Journal of the American Ceramic Society 63 [5-6] 326-331
(1980).
158
V.R. Salvini, V.C. Pandolfelli, R.C. Bradt. Extension of Hasselmans thermal shock
theory for crack/microstructure interactions in refractories. Ceramics International 38
[7] 53695375 (2012).
159
H. Harmuth, E. K. Tschegg. A fracture mechanics approach for the development of
refractory materials with reduced brittleness. Fatigue and Fracture of Engineering
Materials and Structures 20 [11] 1585-1603 (1997).
160
H. Harmuth, R.C. Bradt. Investigation of refractory brittleness by fracture
mechanical and fractographic methods. Refractories Manual 6-10 (2010).
161
G.B. Palmer and G. Baker. A load-cycling technique for R-curve behaviour:
application to a low cement refractory. International Journal of Fracture 62 233-244
(1993).
162
D. R. Larson, J. A. Coppola, D. P. H. Hasselman, R.C. Bradt. Fracture Toughness
and Spalling Behavior of High-Al2O3. Journal of the American Ceramic Society 57 417-
421 (1974).
163
Z. Chlup, D.N. Boccaccini, C. Leonelli, M. Romagnoli, A.R. Boccaccini. Fracture
behaviour of refractory ceramics after cyclic thermal shock. Ceram-Silikaty 50 [4] 245-
250 (2006).
187
Bibliografa
164
Y.C. Ko. Influence of the total fines content on the thermal shock damage resistance
of Al2O3-spinel castables. Ceramics International 27 [5] 501-507 (2001).
165
E.K. Tschegg. Austria Patent Specification AT 390 328 (1986).
166
E. Brhwiler, F.H. Wittmann. The wedge splitting test, a new method of performing
stable fracture mechanics tests. Engineering Fracture Mechanics 35 [1/2/3] 117-25
(1990).
167
H. Harmuth. Stability of crack propagation associated with fracture energy
determined by wedge splitting specimen. Theoretical and Applied Fracture Mechanics
23 [1] 103-108 (1995).
168
H. Harmuth, C. Manhart. Fractographic investigations of ordinary ceramic refractory
materials with reduced brittleness. Key Engineering Materials 409 209-215 (2009).
169
S. Ribeiro, J.A. Rodrigues. The influence of microstructure on the maximum load
and fracture energy of refractory castables. Ceramics International 36 263274 (2010).
170
S. Ribeiro, D. de Campos Ribeiro, M.B. de Souza Dias, G.C. Ribeiro Garcia, E.M.
Bento dos Santos. Study of the Fracture Behavior of Mortar and Concretes with
Crushed Rock or Pebble Aggregates. Materials Research 14 [1] 46-52 (2011).
188
Bibliografa
171
E. Skiera, J. Malzbender, J. Mnch, S. Dudczig, C.G. Aneziris, R.W. Steinbrech.
Controlled crack propagation experiments with a novel alumina-based refractory.
Advanced Engineering Materials 14 [4] 248-254 (2012).
172
D.Y. Miyaji, C.Z. Otofuji, J. de Anchieta Rodrigues. The load-displacement curve of
steady crack propagation: an interesting source of information for predicting the thermal
shock damage of refractories, Proceedings of 13th Biennial Worldwide Conference on
Refractories, UNITECR 2013, September 10 to 13, Victoria, BC, Canada (2013).
173
J. Malzbender, E. Skiera, J. Mnch. A simple methodology to visualize crack
propagation for ceramic materials. Journal of Materials Science 49 403406 (2014).
174
S. Jin, D. Gruber, H. Harmuth. Determination of Youngs modulus, fracture energy
and tensile strength of refractories by inverse estimation of a wedge splitting procedure.
Engineering Fracture Mechanics 116 228236 (2014).
175
Handbook of ceramic composites. Edited by Narottam P. Bansal. Kluwer Academic
Publishers (preface) (2005).
176
M. Sternitzke. Review: structural ceramic nanocomposites. Journal of the European
Ceramic Society 1061-1082 (1997).
177
X. Teng, H. Liu, C. Huang. Effect of Al2O3 particle size on the mechanical
properties of alumina-based ceramics. Materials Science and Engineering A 452453
545551 (2007).
189
Bibliografa
178
F.C. Zhang, H.H. Luo, T.S. Wang, S.G. Roberts, R.I. Todd. Influence factors on
wear resistance of two alumina matrix composites. Wear 265 2733 (2008).
179
R.I. Todd, A. Limpichaipanit, Microstructure-property relationships in wear resistant
alumina/SiC nanocomposites. Advances in Science and Technology 45 555-563
(2006).
180
A. Mukhopadhyay, R. I. Todd. Microstructure and mechanical properties of Al2O3
matrix nanocomposites produced by solid state precipitation. Journal of the European
Ceramic Society 1359-1372 (2010).
181
R.W. Davidge, R.J. Brook, F. Cambier, M. Poorteman, A. Leriche, D. OSullivan, S.
Hampshire, T. Kennedy. Fabrication, properties, and modelling of engineering ceramics
reinforced with nanoparticles of silicon carbide. British Ceramic Transactions 121-127
(1997).
182
S. Maensiri, S.G. Roberts. Thermal shock resistance of sintered alumina/silicon
carbide nanocomposites evaluated by indentation techniques. Journal of the American
Ceramic Society 85 [8] 1971-1978 (2001).
183
S. Maensiri, S.G. Roberts. Thermal shock of ground and polished alumina and
Al2O3/SiC nanocomposites. Journal of the European Ceramic Society 22 2945-2956
(2002).
190
Bibliografa
184
C.E. Borsa, N.M.R. Jones, R.J. Brook, R.I. Todd. Influence of processing on the
microstructural development and flexure strength of Al2O3/SiC nanocomposites. Journal
185
M. Yoshimura, T. Ohji, K. Niihara. Oxidation-induced toughening and strengthening
of Y2O3/SiC nanocomposites. Journal of the American Ceramic Society 80 [3] 797-99
(1997).
186
A. Rendtel, H. Hbner, M. Herrmann, C. Schubert. Silicon nitride/silicon carbide
nanocomposite materials: II, Hot strength, creep and oxidation resistance. Journal of the
187
H. Park, H-E. Kim, K. Niihara. Microstructure and high-temperature strength of
Si3N4-SiC nanocomposite. Journal of the European Ceramic Society 18 [7] 907-914
(1998).
188
M. Sternitzke, B. Derby, R.J. Brook. Alumina/silicon carbide nanocomposites by
hybrid polymer/powder processing: microstructures and mechanical properties. Journal
of the American Ceramic Society 81 [1] 41-48 (1998).
189
H. Reveron, O. Zaafrani, G. Fantozzi. Microstructure development, hardness,
toughness and creep behaviour of pressureless sintered alumina/SiC micro-
nanocomposites obtained by slip-casting. Journal of the European Ceramic Society, 30
[6] 1351-1357 (2010).
191
Bibliografa
190
J. Zhao, L.C. Stearns, M.P. Harmer, H.M. Chan, G.A. Miller. Mechanical behavior
of aluminasilicon carbide nanocomposites. Journal of the American Ceramic Society
76 [2] 50310 (1993).
191
J.D. Kuntz, G-D. Zhan, A.K. Mukherjee. Nanocrystalline-Matrix Ceramic
Composites for Improved Fracture Toughness. MRS Bulletin 29 [1] 22-27 (2004).
192
L. Carroll, M. Sternitzke, B. Derby. Silicon carbide particle size effects in alumina-
based nanocomposites. Acta Materialia 44 [11] 4543-52 (1996).
193
R.W. Davidge, P.C. Twigg, F.L. Riley. Effects of silicon carbide nano-phase on the
wet erosive wear of polycrystalline alumina. Journal of the European Ceramic Society
16 [7] 799-802 (1996).
194
D. Zhang, H. Yang, R. Yu, W. Weng. Mechanical properties of alumina composites
containing various sizes and fractions of fine SiC particles. Journal of Materials Science
Letter 16 877-879 (1997).
195
A.M. Thompson, H.M. Chan, M.P. Harmer. Tensile creep of alumina-silicon carbide
nanocomposites. Journal of the American Ceramic Society 80 [9] 2221-28, 1997.
196
M. Sternitzke, E. Dupas, P. Twigg, B. Derby. Surface mechanical properties of
alumina matrix nanocomposites. Acta Materialia 45 3963-3973 (1997).
192
Bibliografa
197
S. Jiao, M.L. Jenkins, R.W. Davidge. Interfacial fracture energy- mechanical
behaviour relationship in Al2O3/SiC and Al2O3/TiN nanocomposites. Acta Materialia 45
[1] 149-156 (1997).
198
C.C. Anya. Microstructural nature of strengthening and toughening in Al2O3-SiC(p)
nanocomposites. Journal of Materials Science 34 [22] 5557-5567 (1999).
199
X. Sun, J-G. Li, Z. Xiu, K. Duan, X. Zhi Hu. Intragranular particle residual stress
strengthening of Al2O3-SiC nanocomposites. Journal of the American Ceramic Society
88 [6] 1536-1543 (2005).
200
J.L. Ortiz-Merino, RI. Todd. Relationship between wear rate, surface pullout and
microstructure during abrasive wear of alumina and alumina/SiC nanocomposites. Acta
Materialia 53 [12] 334557 (2005).
201
A. Limpichaipanit, R.I. Todd. The relationship between microstructure, fracture and
abrasive wear in Al2O3/SiC nanocomposites and microcomposites. Journal of the
European Ceramic Society 29 [13] 2841-2848 (2009).
202
B. Derby. Ceramic Nanocomposites: mechanical properties. Current Opinion in
Solid State and Materials Science 3 490-495 (1998).
203
H. Awaji, C-H. Chen, N. Kishi, S-M. Choi. Estimation of critical frontal process
zone size in ceramics. ECF15. Stockolm 2004.
193
Bibliografa
204
W.Z. Zhu, J.H. Gao, Z.S. Ding. Microstructure and mechanical properties of a
Si3N4/Al2O3 nanocomposite. Journal of Materials Science 32 537-542 (1997).
205
Y-K. Jeong, A. Nakahira, K. Niihara. Effects of additives on Microstructure and
Properties of alumina-silicon carbide nanocomposites. Journal of the American Ceramic
Society 82 [12] 3609-3612 (1999).
206
U.N.S. Hareesh, M. Sternitzke, R. Janssen, N. Claussen. Processing and properties of
sol-gel derived alumina/silicon carbide nanocomposites. Journal of the American
Ceramic Society 87 [6] 1024-1030 (2004).
207
L.P Ferroni, G. Pezzotti. Evidence for bulk residual stress strengthening in
Al2O3/SiC nanocomposites. Journal of the American Ceramic Society 85 [8] 2033-2038
(2002).
208
J.L. Ortiz Merino, R.I. Todd. Thermal microstress measurements in Al2O3/SiC
nanocomposites by Cr3+ fluorescence microscopy. Journal of the European Ceramic
Society 23 1779-1783 (2003).
209
I.A. Chou, H.M. Chan, M.P. Harmer. Machining-induced surface residual stress
behaviour in Al2O3-SiC nanocomposites. Journal of the American Ceramic Society 79
[9] 2403-2409 (1996).
210
H.Z. Wu, C.W. Lawrence, S.G. Roberts, B.Derby. The strength of Al2O3/SiC
nanocomposites after grinding and annealing. Acta Materialia 46 [11] 3839-3848
(1998).
194
Bibliografa
211
C.C. Anya, S.G. Roberts. Indentation fracture toughness and surface flaw analysis
of sintered alumina/SiC nanocomposites. Journal of the European Ceramic Society 16
1107-1114 (1996).
212
G. Pezzotti, T. Nishida, M. Sakai. Physical limitations of the inherent toughness and
strength in ceramic-ceramic and ceramic-metal nanocomposites. Journal of the Ceramic
Society of Japan 103 [9] 901-909 (1995).
213
G. Pezzotti, V. Sergo, K. Ota, O. Sbaizero, N. Muraki, T. Nishida, M. Sakai.
Residual Stresses and apparent strengthening in ceramic-matrix nanocomposites.
Journal of the Ceramic Society of Japan 104 [6] 497-503 (1996).
214
T. Ohji, A. Nakahira, T. Hirano, K. Niihara, Tensile creep behavior of
alumina/silicon carbide nanocomposite. Journal of the American Ceramic Society 77
[12] 3259-3262 (1994).
215
H. Tomaszewski, M. Boniecki, H. Weglarz. Toughness-curve behaviour of alumina-
SiC and ZTA-SiC composites. Journal of the European Ceramic Society 20 1215-1224
(2000).
216
R.I. Todd, M.A.M. Bourke, C.E. Borsa, R.J. Brook. Neutron diffraction
measurements of residual stresses in alumina/SiC nanocomposites. Acta Materialia 45 4
1791-1800 (1997).
195
Bibliografa
217
S. Jiao, M.L Jenkins, R.W. Davidge. Electron microscopy of crack/particle
interactions in Al2O3/SiC nanocomposites. Journal of Microscopy 18 2 259-264 (1997).
218
T. Ohji, T. Hirano, A. Nakahira, K. Niihara. Particle/matrix interface and its role in
creep inhibition in alumina/silicon carbide nanocomposites. Journal of the American
Ceramic Society 79 1 33-45 (1996).
219
L.C. Stearns, J. Zhao, M.P. Harmer. Processing and microstructure development in
Al2O3-SiC nanocomposites. Journal of the European Ceramic Society 10 473-477
(1992).
220
J. Prez-Rigueiro, J.Y. Pastor, J. Llorca, M. Elices, P. Miranzo, S. Moya. Revisiting
the mechanical behavior of alumina/silicon carbide nanocomposites. Acta Materialia 46
5399-5411 (1998).
221
L. Gao, H.Z. Wang, J.S. Hong, H. Miyamoto, K. Miyamoto, Y. Nishikawa, D. de la
Torre. Mechanical properties and microstructures of nano-SiC-Al2O3 composites
densified by spark plasma sintering. Journal of the European Ceramic Society 19 609-
613 (1999).
222
C.C. Anya, S.G. Roberts. Pressureless sintering and elastic constants of A1203-SiC
nanocomposites. Journal of the European Ceramic Society 17 565-573 (1997).
223
Y.K. Jeong, K. Niihara. Microstructure and mechanical properties of pressureless-
sintered Al2O3/SiC nanocomposites. Nanostructured Materials 9 193-196 (1997).
196
Bibliografa
224
H. Kara, S.G. Roberts. Wet erosion behaviour of low SiC alumina-SiC
nanocomposites. Journal of Materials Science 37 2421-2426 (2002).
225
H. Kara, S.G. Roberts. Polishing behavior and surface quality of alumina and
alumina/silicon carbide nanocomposites. Journal of the American Ceramic Society 83 5
1219-1225 (2000).
226
M. Parchoviansky, D. Galusek, J. Sedlcek, P. Svancrek, M. Kasiarov, J. Duszac,
P. Sajgalk. Microstructure and mechanical properties of hot pressed Al2O3/SiC
nanocomposites. Journal of the EuropeanCeramic Society 33 22912298 (2013).
227
M. Jaafar, H. Reveron, C. Esnouf and G. Fantozzia. Highly creep-resistant alumina
SiC nanocomposites processed by spark plasma sintering. Scripta Materialia 68 134
137 (2013).
228
R. Mohammad-Rahimi, H.R. Rezaie, A. Nemati. Sintering of Al2O3SiC composite
from solgel method with MgO, TiO2 and Y2O3 addition. Ceramics International 37
16811688 (2011).
229
M. Jaafar, G. Bonnefont, G. Fantozzi, H. Reveron. Intergranular aluminaSiC
micro-nanocomposites sintered by spark plasma sintering. Materials Chemistry and
Physics 124 377379 (2010).
230
I.P. Shapiro, R.I. Todd, J.M. Titchmarsh, S.G. Roberts. Effects of Y2O3 additives and
powder purity on the densication and grain boundary composition of Al2O3/SiC
nanocomposites. Journal of the European Ceramic Society 29 16131624 (2009).
197
Bibliografa
231
H. Luo, F. Zhang, T. Wang. Prediction of fracture characteristic of particle
reinforced alumina-based composites. Science in China Series E: Technological
Sciences 52 864-870 (2009).
232
A.M. Cock, I.P. Shapiro, RI. Todd, S.G. Roberts. Effects of Yttrium on the Sintering
and Microstructure of AluminaSilicon Carbide Nanocomposites. Journal of the
American Ceramic Society 88 [9] 23542361 (2005).
233
D. Sciti, J. Vicens, A. Bellosi. Microstructure and properties of alumina-SiC
nanocomposites prepared from ultrane powders. Journal of Materials Science 37 3747-
3758 (2002).
234
J. Fang, M. P. Harmer, H. M. Chan. Evaluation of subgrain formation in Al2O3SiC
235
X.L. Shi, F.M. Xu, Z.J. Zhang, Y.L. Dong, Y. Tan, L. Wang, J.M. Yang. Mechanical
properties of hot-pressed Al2O3/SiC composites. Materials Science and Engineering A
527 46464649 (2010).
236
Y.L. Dong, F.M. Xu, X.L. Shi, C. Zhang, Z.J. Zhang, J.M. Yang, Y. Tan.
Fabrication and mechanical properties of nano-/micro-sized Al2O3/SiC composites.
Materials Science and Engineering A 504 4954 (2009).
237
A.R. Yazdi, H. Baharvandi, H. Abdizadeh, J. Purasad, A. Fathi, H. Ahmadi. Effect
of sintering temperature and siliconcarbide fraction on density, mechanical properties
198
Bibliografa
and fracture mode of aluminasilicon carbide micro/nanocomposites. Materials and
Design 37 251255 (2012).
238
Y-K. Jeong, A. Nakahira, P.E.D. Morgan, K. Niihara. Effect of Milling Conditions
on the Strength of AluminaSilicon Carbide Nanocomposites. Journal of the American
Ceramic Society 80 [5] 1307309 (1997).
239
Y. Xu, A. Zangvil, A. Kerber. SiC Nanoparticle-Reinforced Al2O3, Matrix
Composites: Role of Intra- and Intergranular Particles. Journal of the European Ceramic
Society 17 92 l-928 (1997).
240
Z-Y. Deng, J-L. Shi, Y-F. Zhang, Da-Y Jiang, J-K Guo. Pinning Effect of SiC
Particles on Mechanical Properties of Al2O3-SiC Ceramic Matrix Composites. Journal
241
H. Wu, S. G. Roberts and B. Derby. Residual stress and subsurface damage in
machined alumina and alumina/silicon carbide nanocomposite ceramics. Acta
Materialia 49 507517 (2001).
242
G.J. Liu, H.B. Qiu, R.I. Todd , R.J. Brook, and J.K. Guo. Processing and mechanical
behaviour of Al2O3/ZrO2 nanocomposites. Materials Research Bulletin 33 [2] 281288
(1998).
243
W.H. Tuan, R.Z. Chen, T.C. Wang, C.H. Cheng, P.S. Kuo. Mechanical properties of
Al2O3/ZrO2 composites. Journal of the European Ceramic Society 22 28272833
(2002).
199
Bibliografa
244
J. Li, Y-S. Yin, R-X. Shi, L-P. Ma, J. Li. Microstructure and mechanical properties
of Al2O3TiC4vol.%Co composites prepared from cobalt coated powders. Surface &
Coatings Technology 200 3705 3712 (2006).
245
J. Li, L. Gao, J. Guo. Mechanical properties and electrical conductivity of TiN
Al2O3 nanocomposites. Journal of the European Ceramic Society 23 6974 (2003).
246
J. Luo, R. Stevens. The Role of Residual Stress on the Mechanical Properties of
Al2O3-5vol.% SiC Nano-Composites. Journal of the European Ceramic Society I7
1565-1512 (1991).
247
R.I. Todd, B. Derby. Thermal stress induced microcracking in alumina20% SiCp
composites. Acta Materialia 52 16211629 (2004).
248
J. Sedlacek, D. Galusek, P. Svancarek, R. Riedel, A. Atkinson, X. Wang. Abrasive
wear of Al2O3SiC and Al2O3(SiC)C composites with micrometer- and
submicrometer-sized alumina matrix grains. Journal of the European Ceramic Society
28 29832993 (2008).
249
M. Belmonte, M.I. Nieto, M.I. Osendi, P. Miranzo. Inuence of the SiC grain size on
the wear behaviour of Al2O3/SiC composites. Journal of the European Ceramic Society
26 12731279 (2006).
250
Y.M. Ko, W.T. Kwon, Y-W. Kim. Development of Al2O3SiC composite tool for
machining application. Ceramics International 30 20812086 (2004).
200
Bibliografa
251
D. Galusek, J. Sedlacek, R. Riedel. Al2O3SiC composites prepared by warm
pressing and sintering of an organosilicon polymer-coated alumina powder. Journal of
the European Ceramic Society 27 23852392 (2007).
252
Y-H. Cho, A. Nakahira, K. Niihara. Microstructure and mechanical properties of
SiC-platelet reinforced Al2O3/SiC-particle hybrid composites. Journal of Materials
Science 35 3143-3149 (2000).
253
J. Wang, S.Y. Lim, S.C. Ng, C.H. Chew, L.M. Gan. Dramatic effect of a small
amount of MgO addition on the sintering of Al2O3-5vol.% SiC nanocomposite.
Materials Letters 33 273-277 (1998).
254
X. Shi, Y. Tan, F. Xu, J. Li, Y. Dong, L. Wang. Hot-pressed A12O3SiC(C)
composites with polycarbosilane as the precursor. Materials and Design 31 16281632
(2010).
255
Y. Ji, and A. Yeomans. Processing and mechanical properties of Al2O3-5 vol.% Cr
nanocomposites. Journal of the European Ceramic Society 22 1927-1936 (2002).
256
T. Sekino, K. Niihara, Microstructural characteristics and mechanical properties for
Al2O3/metal nanocomposites. Nanostructured Materials 6 663-666 (1995).
257
T. Sekino, K. Niihara. Fabrication and mechanical properties of fine-tungsten-
dispersed alumina-based composites. Journal of Materials Science 32 3943-3949
(1997).
201
Bibliografa
258
M. Nawa, T. Sekino, K. Niihara. Fabrication and mechanical behaviour of Al2O3/Mo
259
T. Sekino, T. Nakajima, S. Ueda, K. Niihara. Reduction and sintering of a nickel-
dispersed-alumina composite and its properties. Journal of the American Ceramic
Society 80 [5] 1139-1148 (1997).
260
S-T. Oh, T. Sekino, K. Niihara. Fabrication and mechanical properties of 5 vol.%
copper dispersed alumina nanocomposite. Journal of the European Ceramic Society 18
31-37 (1998).
261
G.D. Zhan, J.D. Kuntz, J. Wan, A.K. Mukherjee. Single-wall carbon nanotubes as
attractive toughening agents in alumina-based nanocomposites. Nature materials 2 38-
42 (2003).
262
K.E. Thomson, D. Jiang, W. Yao, R.O. Ritchie, A.K. Mukherjee. Characterization
and mechanical testing of alumina-based nanocomposites reinforced with niobium
and/or carbon nanotubes fabricated by spark plasma sintering. Acta Materialia 60 622-
632 (2012).
263
R.W. Siegel, B.J. Chang, A.J. Stone, P.M. Ajayan, R.W. Doremus. Mechanical
behaviour of polymer and ceramic matrix nanocomposites. Scripta Materialia 44 2061-
2064 (2001).
202
Bibliografa
264
E. Flahaut, A. Peigney, Ch. Laurent, Ch..Marlire, F. Chastel, A. Rousset. Carbon
nanotube-metal-oxide nanocomposites: microstructure, electrical conductivity and
mechanical properties. Acta Materialia 48 3803-3812 (2000).
265
I. Levin, W.D. Kaplan, D.G. Brandon, A.A. Layyous. Effect of SiC submicrometer
particle size and content on fracture toughness of alumina-SiC nanocomposites.
Journal of the American Ceramic Society 78 [1] 254-256 (1995).
266
H. Awaji, S-M. Choi, E. Yagi. Mechanisms of toughening and strengthening in
ceramic-based nanocomposites. Mechanics of Materials 34 411-422 (2002).
267
H. Yang, F. Shang, L. Gao. Microstructure and mechanical properties of gas pressure
sintered Al2O3/TiCN composite. Ceramics International 33 15211524 (2007).
268
Z. Shi, J. Wang, G. Qiao, Z. Jin. Effects of weak boundary phases (WBP) on the
microstructure and mechanical properties of pressureless sintered Al2O3/h-BN
machinable composites. Materials Science and Engineering A 492 2934 (2008).
269
J. Fang, H.M. Chan, M.P. Harmer. Residual stress relaxation behavior in Al2O3-SiC
nanocomposites. Materials Science and Engineeriong A 195 163-167 (1995).
270
I.A. Chou, , H.M. Chan, M.P. Harmer. Effect of annealing environment on the crack
healing and mechanical behaviour of silicon carbide-reinforced alumina
nanocomposites. Journal of the American Ceramic Society 81 [5] 1203-1208 (1998).
203
Bibliografa
271
A.M. Thompson, H.M. Chan, M.P. Harmer. Crack healing and stress relaxation in
Al2O3-SiC nanocomposites. Journal of the American Ceramic Society 78 3 567-571
(1995).
272
G. Pezzotti, W.H. Mller. Strengthening mechanisms in Al2O3/SiC nanocomposites.
Computational Materials Science 22 155-168 (2001).
273
H.Z. Wu, S.G. Roberts, B. Derby. Residual stress distributions around indentations
and scratches in polycrystalline Al2O3 and Al2O3/SiC nanocomposites measured using
uorescence probes. Acta Materialia 56 140149 (2008).
274
M.G. Pontin, M.P. Rao, A.J. Snchez-Herencia, F. Lange. Laminar Ceramics
Utilizing the Zirconia Tetragonal-to-Monoclinic Phase Transformation to Obtain a
Threshold Strength. Journal of the American Ceramic Society 85 [12] 304148 (2002).
275
HZ. Wang, L. Gao, L.H. Gui, and J.K. Guo. Preparation and properties of
intragranular Al2O3-SiC nanocomposites. Nanostructured Materials 10 947-953 (1998).
276
C. Xu, D. Sun. Formation of intragranular nano-structures in micro-sized ceramic
composite materials. Materials Science and Engineering A 491 338342 (2008).
277
M. Hoffman, J. Rdel. Suggestion for mechanism of strengthening of
nanotoughened ceramics. Journal of the Ceramic Society of Japan 105 [12] 1086-1090
(1997).
204
Bibliografa
278
J. Prez-Rigueiro, J.Y. Pastor, J. Llorca, M. Elices, J.S. Moya, P.Miranzo, M.
Belmonte. Propiedades mecnicas de nanocompuestos de Al2O3-SiC. Anales de
Mecnica de la Fractura 14 367-371 (1997).
279
S. Mezquita. Caracterizacin de Materiales de almina-mullita para su uso como
herramientas de corte y mecanizado de metales. Proyecto final de carrera ETS
Ingeniera de Materiales, UPM.
280
UNE-EN 993-6:1995. Materiales cermicos densos conformados. Parte 6:
Determinacin del modulo de rotura a temperatura ambiente.
281
G.V. Guinea, J.Y. Pastor, J. Planas and M. Elices. Stress intensity factor, compliance
and CMOD for a general three-point-bend beam. International Journal of Fracture 89
103116 (1998).
282
J.F. Shackelford, W. Alexander, J.S. Park. CRC Materials Science and Engineering
Handbook. CRC Press, Boca Raton, Florida. 537-539 (1994).
205

Garcia Prieto Ana

Cargado por

Copyright:

Formatos disponibles

Garcia Prieto Ana

Cargado por

Información del documento

Derechos de autor

Formatos disponibles

Compartir este documento

Compartir o incrustar documentos

Opciones para compartir

¿Le pareció útil este documento?

¿Este contenido es inapropiado?

Copyright:

Formatos disponibles

Garcia Prieto Ana

Cargado por

Copyright:

Formatos disponibles

FACULTAD DE CIENCIAS

DEPARTAMENTO DE QUMICA INORGNICA

Memoria presentada por

Ana Garca Prieto

Para optar al grado de

DOCTORA EN CIENCIAS QUMICAS

INSTITUTO DE CERMICA Y VIDRIO

CONSEJO SUPERIOR DE INVESTIGACIONES

La autora agradece la financiacin econmica otorgada por el Consejo Superior

El objetivo principal de este trabajo ha sido el desarrollo de un nuevo dispositivo

experimental y una metodologa de ensayo y anlisis de resultados para determinar la

caracterizacin de rutina en el laboratorio.

Se ha planteado como hiptesis bsica la adecuacin de los ensayos de fractura

estable para la determinacin de la tenacidad de fractura de los materiales.

Se ha propuesto la utilizacin de la apertura de los labios de la grieta como

montona durante todo el ensayo (carga y fractura).

frgiles como ytria (Y-PSZ) y hormigones de alta resistencia.

El dispositivo experimental consta de un micrmetro ptico asociado a un

dispositivo de flexin en tres puntos y se ensayan probetas prismticas con entallas

rectas en forma de V (SEVNB).

Con el fin de demostrar la validez del dispositivo experimental y la metodologa de

ensayo y anlisis de resultados desarrollados, se ensayaron materiales cermicos con

comportamientos mecnicos muy diferentes:

Espinela de aluminio-magnesio (MgAl2O4), con comportamiento frgil.

Almina (Al2O3) de grano fino, con comportamiento frgil.

Mullita (3Al2O3.2SiO2), con comportamiento frgil y crecimiento subcrtico de

grietas en aire significativo.

Refractarios comerciales, en los cuales ocurren mecanismos de refuerzo durante la

adecuacin de la utilizacin de la apertura de los labios de la grieta como parmetro de

Se estudi la influencia de las condiciones experimentales en la tenacidad de

fractura. La utilizacin de la apertura de los labios de la grieta como parmetro de

control minimiza el efecto de factores como la inercia de la mquina y permite

identificar el crecimiento subcrtico de grietas.

El valor del factor crtico de intensidad de tensiones en modo I, KIC, obtenido a

partir de ensayos estables es considerablemente inferior al obtenido a partir de ensayos

energa de fractura, que caracteriza todo el proceso de fractura. No es posible obtener

este parmetro a partir de ensayos inestables.

Una vez demostrada la validez del dispositivo experimental y de la metodologa de

ensayo y anlisis de resultados desarrollados, se prepararon y caracterizaron materiales

nanocompuestos de Al2O3/SiC y se ensayaron segn la metodologa propuesta,

aportando valores de tenacidad de fractura desconocidos hasta la fecha. Estos

resultados permitieron establecer las relaciones microestructura-comportamiento

mecnico en los nanocompuestos de Al2O3/SiC.

metodologa de ensayo y anlisis de resultados desarrollados son adecuados para la

caracterizacin de otros nanocompuestos cermicos.

B: Beam thickness / Espesor de la probeta (Figura 3.3)

CMCs: Ceramic Matrix Composites / Compuestos de matriz cermica

CMOD: Crack Mouth Opening Displacement / Apertura de los labios de la grieta

CMOS: Complementary Metal Oxide Semiconductor / Semiconductor

CNB: Chevron Notched Beam / Probetas con entalla triangular Chevron

CT: Compact Tension / Tensin compacta

d50: Average grain size / Tamao de grano medio

DCB: Double Cantilever Beam / Probetas de doble voladizo

DE: Digital Edge-detection / Procesador digital de borde

DT: Double Torsion / Doble torsin

E: Youngs modulus / Mdulo de Young

E: Generalized Youngs modulus / Mdulo de Young generalizado

E.d.: Experimental details / Detalles experimentales

ESIS: European Structural Integrity Society / Sociedad Europea de Integridad

G: Energy release rate / Tasa de liberacin de energa