Processing and mechanical properties of
nickel(titanium)/alumina and metal/hydroxyapatite composites
Citation for published version (APA):Zhang, X. (1994). Processing and mechanical properties of nickel(titanium)/alumina and metal/hydroxyapatite composites. Technische Universiteit Eindhoven. https://doi.org/10.6100/IR427896
DOI:
10.6100/IR427896
Document status and date: Published: 01/01/1994 Document Version:
Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page numbers.
Link to publication
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:
www.tue.nl/taverne
Take down policy
If you believe that this document breaches copyright please contact us at: openaccess@tue.nl
providing details and we will investigate your claim.
Processing and Mechanical Properties
of Nickel(Titanium)/Alumina and
Metal/Hydroxyapatite Composites
Processing and Mechanical Properties
of Nickel(Titanium)/ Alumina and
Metal/Hydroxyapatite Composites
Proefschrift
ter verkrijging van de · graad van doctor aan de Technische Universiteit Eindhoven, op gezag van de Rector Magnificos, prof. dr. J. H. van Lint, voor een commissie aangewezen door bet College van Dekanen in bet openbaar te verdedigen op
vrijdag 18 november 1994 om 16.00 uur
door
Xi-Ming Zhang
geboren te Henan, China
Dit proefschrift is goedgekeurd door de promotoren
prof. dr. R. Metselaar
en
prof. dr. R.
J.
Brook
CIP-DATA KONINKLIJKE BIBLIOTHEEK, DEN HAAG
Zhang, Xi-Ming
Processing and mechanical properties ofnickel(titanium)lalumina and metal/hydroxyapatite composites
I
Xi-Ming Zhang. - [s.l.:s.n.] Thesis Eindhoven. - With ref. - With summary in Dutch.ISBN 90-386-0244-8
Contents
1 Introduction
1.1 Brittle Nature of Ceramics . 1.2 Strengthening of Ceramics . 1.3 Outline of this Thesis .
References . . 1 3 4 2 Literature Review 5 2.1 Introduction . . . 5 2.2 Ductile Toughening . . . 5 2.2.1 Experimental Study . . . 5 2.2.2 Mechanism . . . . <. . . • . 8
2.2.3 Design for Tough Metal Particulate Reinforced Ceramics . . . 13
2.3 Ni/A}z03 System. . . 16
2.3.1 Wetting Character . . . 16
2.3.2 Chemistry of the Interface . . . 16
2.3.3 Composite Properties . . . 18
2.4 Hydroxyapatite Matrix Ceramics . . . 19
2.4.1 General . . . 19
2.4.2 Monolithic HAp . . . 20
2.4.3 HAp Matrix Composites . . . 22 2.5 Summary . . . . References . . . 3 Experimental Procedure 3.1 Introduction . . . . 3.2 Powder Processing . . . 3.2.1 Raw Materials . 3.2.2 Mixing . . . . . 3.2.3 Green Compact Preparation . 3.3 Sintering . . . . 24 25 31 31 31 31 31 32 33
ii CONTENTS 3.4 Microstructure Investigation . . . 34 3.4.1 X-Ray Diffraction . . . 34 3.4.2 Scanning Electron Microscopy and Transmission Electron Microscopy . 35 3.5 Property Measurement . . . . 3.5.1 Density . . . . 3.5.2 Elastic Modulus . . . . 3.5.3 Fracture Strength . . . . . 3.5.4 Fracture Toughness . . . . References . . . . 36 36 .. . . . 36 37 37 40
4 Investigation on the Processing ofNi{Ti)/AI:z03 Composites 4.1 Introduction . . . .
41
. . . . • . 41
4.1.1 Problem Outline . . . . . 4.1.2 Grinding Technique . . . . 4.2 Grinding of Ni and Ni-Ti Powder . . . .
4.2.1 Influence of Milling Time . . . .
41
42 42 42 4.2.2 Influence of Milling Balls, Liquid and Coarse Alumina Powder. . . . 43 4.2.3 Effect of Dry Attrition Milling of Ni-TI Powder . . . 43 4.2.4 Summary of Powder Preparation
4.3 Character of Green Body . . . . 4.4 Sinteringinvestigation . . . .
4.4.1 Sintering of Monolithic Al203
4.4.2 Sintering of Ni{Ti)JA)z03 Composites . . 4.4.3 Analysis of the Sintering Experiments . . 4.5 Summary and Conclusions . . . .
References . . . .
5 Microstructure and Mechanical Properties of Ni(Ti)/ AI203 Composites 5.1 Introduction . . . . 5.2 Characterization of Microstructure . . . . 5.2.1 Phase Analysis by XRD . . . . . 5.2.2 General Microstructure . . . . . 5.2.3 Interface Structure . . . . 46 46 . . . 47 . . . 47 . . . . . 48 53 54 54 55 55 55 . . • . . 55 56 5.2.4 Thermodynamic Analysis of the Interface Reaction . . . .
59 64 66 66 68 68 72 5.3 Mechanical Properties of Materials . . .
5.3.1 Elastic Modulus and Hardness . 5.3.2 Fracture Strength . . . . 5.3.3 Fracture Toughness . . . .
CONTENTS
5.4.1
5.4.2 5.4.3
Fractography and Fracture Strength . . . . Toughening . . . . Critical Particle Size for Crack Bridging . . . . 5.5 Summary and Conclusions . . ..
References . . . . Appendix . . . .
6 Sllver!Hydroxyapatite Composites 6.1 Introduction . . . .
7
6.2 Materials and Methods . . . 6.2.1 Raw Materials . . . 6.2.2 Powder Processing . 6.2.3 Sintering . . . .
6.2.4 Characterization . . . . 6.3 Sintering and Stability of Monolithic HAp . . . . 6.4 Sintering of Composites . . . . 6.4.1 Dilatometer Investigation . . . . 6.4.2 Sintering of Large Block Samples . .
6.5 Mechanical Properties . . . • . . . . . 6.5.1 Hardness and Elastic Modulus . . . . 6.5.2 Strength . . . . 6.5.3 Fracture Toughness . . . .
6.5.4 Opportunities for further Toughening . . . . . 6.6 Summary and Conclusion
References . . . .
Appendix . . . .
FeCralloy!Hydroxyapatite System 7.1 Introduction
..
. .
7.2 Experiments
7.3 Diffusion Couple Studies
..
.
7.3.1 General. . .
7.3.2 Annealing in Air . 7.3.3 Annealing in Nitrogen . 7.3.4 Annealing in Hydrogen 7.3.5 Discussions 7.4 FeCralloy/HAp Composites.
7.4.1 Sintering Investigation . 7.4.2 Hot-Pressing . .. .
. ..
,,..
. .
.
. . . .
..
. ..
. . . .
..
. . . .
.
iii 72 73 76 80 80 82 83 83 83 83 84 84 86 86 88 88 92 97 98 99 . . 101 . . 103 .. 104 .. 104 . 106 107 . 107 . 107 . 109 109 . 110 . 115 . 116 . 116 . 120 . 121 . 121iv
7.5 Summary and Conclusions . . . . References .
Appendix . . . . 8 General Discussion and Conclusions
8.1 Comparison and Discussions . . . . 8.2 Conclusions . . . . 8.3 Remaining Problems and Future Work . . . . Summary
Samenvatting (Summary in Dutch) List of Symbols and Abbreviations Acknowledgments Curriculum vitae CONTENTS . . 123 . . 124 . . 125 126 . . . . 126 . . . . . 127 . . . 129 130 132 134 137 138
Chapter 1
Introduction
1.1 Brittle Nature of Ceramics
The name ceramics applies to a wide variety of inorganic materials including glasses, single crystals and polycrystalline ceramics. Structural ceramics refer to materials intended to serve as structural parts subjected to mechanical stress and in many cases also high temperatures. Typical examples of structural ceramics are alumina, zirconia, silicon nitride and silicon carbide. These materials possess some excellent properties such as chemical stability, refractory character, wear resistance and high hardness. However, in monolithic form, ceramics are generally very brittle at low and medium temperatures. That is, the fracture of monolithic ceramics is governed by the growth of cracks and very little plastic deformation, which could dissipate strain energy and relieve stresses at crack tips, is involved. This brittle nature has led to a low material reliability as reflected by broad strength distributions and catastrophic fracture behavior, which has strongly limited the applications of ceramics as structural components.
1.2 Strengthening of Ceramics
Over the years a variety of attempts have been made to increase the reliability of ceramic materials. The reliability can be improved by minimizing flaw size and by increasing fracture toughness, as outlined by the Griffith fracture criterion
K1c
q
=
--rj2. Yac(1.1)
where o-is the strength, K1c is the critical stress intensity factor, i.e., a measure of the fracture toughness of the materials, ac the flaw size, and Y a well-documented crack and specimen geometry parameter.
Considerable research has been involved in identifying and eliminating the most detrimental flaws during processing [1]. Today, with the progress in processing, machining and
non-2 Introduction
destructive evaluation, ceramics with very high fracture strengths and narrow strength distribu-tions can be produced. However, the processing requirements are quite strict (high cost) and the materials remain brittle. The high strength will diminish substantially due to damage introduced during service (e.g., because of impact or contact, thermal shock, static load slow crack growth, cyclic fatigue and creep damage). Thus, more recent research aims to create materials that are sufficiently tough that the strength becomes insensitive to the size of flaws and the failure mode changes from a catastrophic brittle one to a more controlled fracture. Obviously, this approach has the advantage that appreciable processing and post-processing damage can be tolerated without compromising the structural reliability [2].
The fracture toughness can be improved by tailoring the monolithic (matrix) microstructure and by incorporating various reinforcements (second phase). The manipulation of monolithic microstructure has been used effectively in silicon nitride materials. Elongated grain microstruc-ture shows much higher fracmicrostruc-ture toughness and strength than the equiaxed microstrucmicrostruc-ture [3]. However, there is no easy method to form controlled, elongated grain microstructure in other systems, even in those as well studied as alumina [4].
The use of various reinforcements to form ceramic matrix composites can in principle be applied to any ceramic materials. Significant improvements in toughness have been shown in many materials due to a number of operating toughening mechanisms associated with the incorpora-tion of reinforcements [2] [4] [5]. Examples are phase transformation toughening in tetragonal Zr02 reinforced ceramics, microcracking toughening in TiB2 reinforced SiC, pull-out effect in SiC whisker reinforced Si3N4 or Al203 fibre reinforced SiC, crack bridging toughening in Co reinforced WC or AI reinforced Ah03, and more recently nanoparticle dispersion toughening in SiC particle (0.2 pm) reinforced A}z03.
As noted, the reinforcement can take many forms (ceramic, metallic or even polymeric if the materials are not intended for application at elevated temperatures)and various shapes (platelet, whisker or continuous fibre). GeneraUy, as the aspect ratio of the second phase increases, so does the mechanical performance, but other problems and drawbacks become evidertt, such as difficulty of fabrication, material anisotropy, etc. From the processing standpoint, particulate reinforcements are preferred because the composites can be fabricated by methods which require only minor modifications to the routes as used for the matrix materials.
In the present work metal particles were chosen for the toughening of ceramics. This has been applied firstly toNi and Ni-Ti alloy reinforced A}z03 ceramics as a model system and then to Ag or FeCralloy (Fe, Cr and AI alloy) reinforced calcium hydroxyapatite bioceramics with a commercial application background.
1.3 Outline of this Thesis 3
1.3 Outline of this Thesis
The objective of the present work is to prepare metal particle toughened ceramics. The process-ing is investigated with the aim of obtainprocess-ing a controlled microstructure with high density by conventional pressureless sintering. The mechanical properties of the composites are compared with toughening models and related to the metal/ceramic interfacial character.
Following this introduction chapter, chapter 2 reviews the literature on toughening by ductile re-inforcements. The Ni/Ah03 interfacial character and the influencing factors are discussed. The fundamental aspects of hydroxyapatite ceramics and the present stage of research on reinforcing hydroxyapatite are also summarized. Chapter 3 describes the general experimental methods including powder preparation, pressureless sintering, property characterization and microstruc-ture analysis. Chapter 4 presents the problems encountered with large sized metal powders and attempts in milling metal powders. The results of pressureless sintering of Ni(Ti)/Ah03
composites with various metal contents starting with different raw powders are also given. Chapter 5 reports the microstructure and mechanical properties of Ni(Ti)/Alz03 composites and their relations. The results are compared with literature and toughening models. Chapter 6 deals with silver toughened hydroxyapatite ceramics. The processing and properties of the composites have been investigated systematically. Chapter 7 discusses the problems relating to the reactions and densifications in FeCralloy reinforced hydroxyapatite. The final chapter summarizes the main results and conclusions of this work.
4 Introduction
References
I. F. F. Lange, "Processing Science and Technology for Increased Reliability",]. Am Ceram. Soc., 72 [I] 3-15 (1989).
2. A. G. Evans, "Perspective on the Development of High-Toughness Ceramics", J. Am. Ceram. Soc., 73 [2] 187-206 (1990).
3. F. F. Lange, "Relation between Strength, Fracture Energy, and Microstructure of Hot-Pressed Silicon Nitride", J. Am. Ceram. Soc., 56 512-22 (1973).
4. M. P. Harmer, H. M. Chan and G. A. Miller, "Unique Opportunities for Microstructural Engineering with Duplex and Laminar Ceramic Composites", J. Am. Ceram. Soc., 75 [7]
1715-28 (1992).
5. P. F. Becher, "Microstructural Design of Toughened Ceramics", J. Am. Ceram Soc., 74 [2] 255-69 (1991).
Chapter2
Literature Review
2.1 Introduction
Brittle ceramics can be toughened by incorporating ductile reinforcements into them. The toughness of the composite derives from the ductility of the reinforcing phase. Utilization of the inherent toughness of the reinforcements may be affected by many factors, incJuding the physical and chemical compatibility of the matrix and reinforcement, the microstructure and the interfacial characteristics. In this chapter, various investigated systems are summarized and their microstructural features, mechanical performance and prOcessing techniques are outlined. Toughening mechanisms relating to ductile reinforcements are reviewed and the literature data are evaluated according to these mechanisms. Some fundamental aspects about Ni/Ah03 and Ag/hydroxyapatite systems are also given.
2.2 Ductile Toughening
2.2.1 Experimental
Study
There has been constant interest in using ductile phases to improve the toughness of brittle ceramics. Various kinds of ductile reinforcements, mostly metals or alloys, have been dispersed in ceramic matrices. Table 2.1 summarizes some of the investigated systems. Very promising toughening has been achieved in some composites.
The composites exhibit different microstructures in the distribution of ductile reinforcements. The reinforcements can be distributed discretely as an isolated phase, e.g., in many particle or short fibre reinforced composites. The reinforcements can also form a continuous, interpene-trating structure, e.g., in Lanxide AI networks interpeneinterpene-trating Ah03• In other cases, structures may include reinforcements with large dimensions such as long fibres or laminates. Materials of this type are not listed in the table, but they have received much interest in recent years. Examples are Pb wire in glass [31], Ni alloy fibres or plates in TiAl (brittle intermetallic) [32] [33], Nb laminates in MoSi;a [34] and Ni laminates in Ah03 [35].
Systems Reinforcement Processing J(lc J(lc,m Remarks Reference Content Size(Jlm) (MPam112) (MPam112)
Nil glass 20vol% 5- 10 HP1 Ni oxidized [1]
W/glass 3-70 HP internal stress [2]
Nil glass 20 HP 0.68 [3]
AI/Glass 20vol% 100- 120 HP 6.5 0.80 oxidized, match of E and a [4]
Kovar/Giass 25 vol% 44-75 HP 1.83 0.70 etched [5]
Kovar/Glass 25 vol% 44-75 HP 1.33 0.70 oxidized [5]
Kovar/Glass 25 vol% 44-75 HP 1.31 0.70 cleaned [5]
Au-41/Giass 35 vol% -44 HP 2.13 0.63 [6]
Au-Pt,Pd/Glass 35 vol% -44 HP 1.84 0.63 [6]
Ni!MgO 5-30vol% 3 -5mm HP fibre, increased toughness [7]
FeCralloy/HAp 20vol% 200 HP 7 1 fibre [8]
Al/B4C 37 vol% infiltration 9.7 3.7 [9]
Co/WC 10wt% HP 13- 17 network of Co [10]
Nifl'ZP(4Y) HP 10.1 6.2 [11]
Ag/YBa2Cu301-x 30wt% sintering 2.4 1.1 [12]
Ag!YBa2Cu301-x 18.6wt% 6 sintering 3.6 1.4 [13]
tv ~ \::1 1:: (") :::. ~
Systems Processing I<tc Ktc,m Remarks Reference (;i 1::
(MPam112) (MPam112) OQ ~
s
strength increase [14]s·
OQ 0.5-4 6.03 5.26 [15] 2-7 7.3 4.1 [16] <20nm 8 4.4 high strength [17] RS 8.5 3.5 [18] [19]oxidation of AI 9.5 3.8 Lanxide, network of metal [20] <fo20 squeeze casting I I 3.0 fibre, large scale bridging [21] [22]
squeeze casting 8 [23] infiltration 5.8 [24] coating,HP
ness
[25] <0.2 sol-gel, RS [26] 1-2 RS [27] 2-7 [28] TiN interlayer [29] 2 HI [30]1 -hot pressing; 2 - Fe-Ni-Co alloy; 3 - reduction/sintering; 4 - hot isostatic pressing.
8 Literature Review The distribution of the ductile reinforcements strongly influences the properties of the compos-ites. Reinforcements in continuous or large dimensional form can impart substantial toughness to the composites. However, the material homogeneity and isotropy, the corrosion and electrical resistance may be sacrificed. Comparatively, the toughening effect is limited if the reinforce-ments are particulates, especially spheres, although high toughness is observed in some systems, e.g. in AI reinforced glass [4]. The common problem is the lack of plastic deformation of the ductile particles, so the inherent toughness of the ductile phase is not utilized adequately.
The distribution of the ductile phase is high! y dependent on the processing techniques. Compos-ites with isolated reinforcements are usually prepared by powder metallurgical methods. The reinforcements can be introduced either as metals directly or as oxides and then reduced to met-als before sintering. The latter method has the advantage in controlling the particle size. When other techniques such as sol-gel wet chemical methods are involved, composites with nanosized metallic particles can be prepared [17] .. These materials possess very promising properties in both strength and toughness. In the powder methods, hot pressing is quite often used for the purpose of densification. Conventional pressureless sintering is used sometimes for particulate reinforced ceramics.
Some innovative processing routes have been developed recently which can produce metal reinforced ceramic composites with unique structure. They include the Lanxide method [20],
pressure infiltration or squeeze casting [21] [23]. The Lanxide method generates composites by the environmental reaction of a molten metal (e.g. Al/Ah03 composite from the oxida-tion of aluminium alloy). Pressure infiltraoxida-tion techniques are applied by infiltrating ceramic preforms with molten metals .. The size and the content of the metal reinforcement can be modified as desired by tailoring the microstructure of the preforms. Composites prepared by these methods have a microstructure with the metal reinforcements as interpenetrating networks.
For the preparation of laminated composites, diffusion bonding (HP or HIP) is the common practise.
2.2.2 Mechanism
Crack BridgingExtensive experimental and theoretical research has been devoted to the mechanisms of tough-ening by a ductile phase [4] [31] [36] [37] [38]. Many individual mechanisms may be involved, including crack deflection, crack trapping, crack bridging, and crack shielding and plastic dis-sipation associated within a plastic zone. Amllysis indicated that the crack bridging is usually the most potent of these mechanisms. Figure 2.1 schematically illustrates the toughening model based on crack bridging [38]. A propagating crack in a brittle matrix is intersected by ductile
2.2 Ductile Toughening 9
{
0
0
0 0
~
I0
0
0
0
0
0
0 0
0
0
u0
0
CRACK u•u0
"""'[
0
0
0
0
0
0
0
0
0
I
0
0
0 0
0
Figure 2.1: Crack bridging of ductile reinforcements as the toughening mechanism.
particles. The particles stretch and fail as the crack opens. The plastic work of the ligament stretching contributes to the toughness of the composite. The increase in toughness ~Gc (in fact the critical strain energy release rate, which is another measure of fracture toughness in addi-tion to the critical stress intensity factor K tc) by the bridging process can be expressed as [31] [38]
~Gc
=
f [•
u(u)du (2.1)where
f
is the area fraction of ductile reinforcement on the fracture surface, u( u) is the stress/strain relationship for the reinforcement, u is the plastic displacement of the reinforcement between the opening crack surfaces and u• is the plastic displacement upon rupture. Thus, the asymptotic toughness G c of the composite has the formGc
=
Gm(l-f) +~Gc= Gm(l-f)+
f
lou• u(u)du (2.2)where Gm is the matrix toughness. Equation 2.1 can be expressed as [38]
(2.3)
and
r·'R
x
=lo
(CI/Cio)d(u/R) (2.4)where uo is the uniaxial yield strength, R is the cross-sectional radius of the reinforcement, and xis a work of rupture parameter which represents the toughening capacity of the reinforcement.
10 Literature Review
0'--~-'-~---'-~-'--~---'-~--'--~-' 0.00 0.40 0.80 1.20 1.80 2.00 2.40
Normalized Stretch, u/R
Figure 2.2: The effect of debonding length on the stress-strain curve (after Ashby et al [31 ]).
Values of
x
can be determined from the area under the non-dimensional stress-strain curve for the reinforcement.The stress-strain relationship has been evaluated using analytical models [38] and experimental measurement [31]. Results indicate that the stress-strain relation is strongly influenced by the constraint of the ductile ligaments as well as by the reinforcement constitutive behaviour such as work hardening coefficient. Figure 2.2 shows the effect of the constraint. For a well-bonded re-inforcement/matrix interface (de bonding length d ~ 0), the stress increases sharply and reaches a peak value ("' 7uo) because of the elastic constraint ofthe matrix. The stress decays rapidly as the crack opens and the rupture occurs at a small plastic stretch (low
u/
R value). In this case,· x
is low. When interfacial debonding occurs, the peak stress falls, but the plastic stretch prior to failure increases substantially. As a result,x
also increases and the toughening becomes more effective. Physically, this can be appreciatedin terms that debonding maximizes the plastic work in the reinforcement. Specifically, debonding results in a large effective gauge length of the reinforcement, i.e., the plastic stretch of the reinforcement is no longer localized to the region between the main crack surfaces. Thus, a large plastically deforming volume of the reinforcement contributes to the toughening.Figure 2.3 shows the experimental data of
x
obtained from some controlled tensile test of the reinforcements constrained in brittle matrices. The value ofx
can vary between "" 1 and "' 6, governed by the plastic stretch, u* / R. In this plot, the effect of debonding is implicit through its influence on the stretch achieved by the reinforcement.2.2 Ductile Toughening 11
o Pb/glaaa Minimum Constraint
><
0 a) 8 • Nb/TIAI ,_, • TINb/TIAI ::J0.
• ~.
Increased ::J 4•
0••
a:
: Debondlng-
0 ~...
2 0·==
0 Maximum Constraint 0 1 2 3Normalized Plastic Stretch
u·/R
Figure 2.3: The work of rupture as a function of plastic stretch (after Ashby et al {31] and Cao etal {32]).
are the parameters controlling the toughening effect. .These parameters are governed by the combined properties of the reinforcements and the matrix. The interface character plays a critical role in toughening. In designing composites with ductile reinforcements, the greatest toughening is obtained from inclusions with a high yield strength, a large diameter, and an appropriate interfacial bonding strength which allows certain debonding but always keeps some interfacial attachment in the process of crack opening. These conclusions have been verified experimentally in Lanxide Al/Ah03 composites [38], Nb fibre or plate toughened TIAI [32] [33], and some other systems.
It is noteworthy that the crack bridging model predicts the toughening effect at steady state. Before reaching the steady state, the toughness may increase systematically as the crack extends if the bridging reinforcements do not fail immediately. This increasing resistance to crack extension as the crack grows longer is known as R-curve behaviour. If significant, the R-curve behaviour can make the strength relatively insensitive to crack length (flaw tolerance), as can be understood from the Griffith relation. R-curve behaviour imparted by ductile reinforcement has been identified in squeeze casted AliAh03 composites [21] [22].
Crack Deflection·
In the process of crack bridging, a crack is attracted to ductile reinforcements. If the crack propagates around reinforcements, either along the reinforcement/matrix interface or within the matrix, the ductility of the reinforcements will not contribute to the toughness of the composites according to the bridging mechanism. In such a case, the operating toughening mechanism may
12 Literature Review I " c:J
....,
c:J 4 R•12 0 0 CD 7 c .c 3 Cl) ::s 0 1-3 CD > 2:;
Sphere "ia:
1 0.00 0.10 0.20o.ao
0.40 Volume Fraction, V,Figure 2.4: Relative toughness predictions from crack deflection model for spherical particles and rod-shaped particles of three different aspect ratios R (redrawn after Faber et al { 40 ]).
be the crack deflection.
Crack deflection as a toughening mechanism has also been established [40]. The increase in toughness arises from the tilt and twist of the crack front, and hence the resulting greater crack plane area. The microstructural features which influence the toughening effect are the morphology (shape and distribution) and the volume fraction of the reinforcement (Figure 2.4). Reinforcements with a small aspect ratio have a very limited toughening effect even with high volume loading. For example, with 20 vol% spherical reinforcements, the increase in toughness (AGc/Gm) is only about 30%. Therefore, crack deflection is not an effective mechanism for toughening.
It should be noted that in the crack deflection model the effect of interface debonding is not included because the model assumes that a crack deflects away from a particle when it intercepts the particle. In reality a crack usually propagates along the interface to some distance. Therefore, the interfacial debonding also has a certain contribution to the toughness. The pull-out of the reinforcements may also occur, but frictional bridging provides little toughening for spherical particles.
Crack deflection is· often observed in particulate reinforced ceramics. Different from large dimensional reinforcement, whose plastic deformation is always involved when the composite fails during crack opening, particulate reinforcements may not be affected by a bypassing crack
2.2Ductile Toughening 13
due to their geometry. As listed in Table 2.1, the toughening effect by ductile particulate reinforcements is very limited in most cases.
2.2.3 Design for Tough Metal Particulate Reinforced Ceramics
As noted earlier, crack bridging by ductile reinforcements can substantially enhance the fracture toughness of brittle ceramics. The operation of the bridging mechanism requires that the crack is locked by the ductile reinforcements which fail after extensive plastic deformation. The interaction of the crack with the reinforcements depends on the material properties and characteristics of the matrix and the reinforcement. Among them, the thermal expansion coefficient and the elastic modulus as well as the interfacial bonding strength. In the design of metal particulate reinforced ceramics, these parameters are of special importance due to the small geometric size of the particulates.
Match of Thermal Expansion Coefficient
According to Selsing [ 41 ], assuming that there is no stress relaxation during cooling, the radial stress around a spherical particle due to thermal expansion mismatch in a composite can be expressed as
(2.5)
where at is the residual thermal stress, Aa( = a0 - am) is the difference in thermal expansion coefficient of ceramic and (metal) particle, !J.T is the cooling range of the temperature. lie, lim, Ec and Em are the Poisson's ratios and elastic moduli of ceramic matrix and reinforcing particles, respectively. Hac
>
am. cooling from the processing temperature places the particles in compression and the matrix in tension. This situation is in favor of the reinforcement/matrix interfacial bonding and thus attraction of a bypassing crack, but it bears the risk of deteriorating the matrix strength due to the tensile stress within the matrix and the probable link-up of radial cracks from individual particles.H ac
<
a,u cooling from the processing temperature places the particles in tension and the matrix in compression. In this situation, an increased local driving force would be necessary for crack propagation to occur through the matrix. However, if the mismatch is too large and/or the reinforcement/matrix interfacial bonding is too weak, then cracks may develop around the particles, essentially debonding the particles from the matrix. This would not be preferred for the toughening by the crack bridging mechanism.In most cases of metal reinforced ceramics, the thermal expansion coefficient of the reinforce-ment is higher than that of the matrix. The tensile thermal stress on the reinforcereinforce-ment weakens
14 Literature Review the interfacial bond and promotes crack propagation along the interface instead of bridging.
Experiment has shown the crucial influence of the internal thermal stress on the path of propagat-ing cracks, and therefore, on the particulate-composite toughness [4]. In Ni/M-glass composites (a9z,.,8
>
aNi), the hydrostatic compressive stress generated within Ni particles facilitated the reaching of particle/matrix interface by a propagating crack. Thereafter, the crack moved easily along the weak Ni 0/Ni interface. In NiiD-glass composites ( a91aos<
aN;), the large hydrostatictensile stress promoted interfacial fracture on cooling. The particles acted as "pseudopores". Both situations prohibit the crack bridging and the utilization of the inherent metal particle toughness.
In the choice of reinforcement-matrix combinations, the match in thermal expansion coefficient is an important factor. This may be difficult for a specific system. However, other methods, e.g. interfacial films or coatings, can be incorporated to diminish the thermal stress acting on the interface. On the other hand, while the match in thermal expansion coefficient is required for isolated particulate-ceramic composites, this may not be the case for composites with interpenetrating or continuous reinforcement. In AI/ Ah03 (Lanxide) composites, the mismatch between AI and Ah03 is quite large, yet this is a very important example of toughening by ductile phase. Furthermore, mismatch in thermal expansion coefficient may lead to microcracking toughening coupled with the crack bridging toughening. Thus, synergism can be achieved. At present this still needs experimental evidence to support.
Match of Elastic Modulus
While the mismatch in thermal expansion coefficient causes internal residual stress, the mis-match in elastic modulus could lead to stress concentration around the particles under an applied load. According to an analysis (42], if Ec
<
Em, the applied stress becomes magnified at the poles of a particle. If Ec>
Em. the applied stress becomes magnified at the equator of the particle, but released at the poles of the particle. As a propagating crack (perpendicular to the direction of the tensile stress) approaches the particle, the stress intensity factor decreases at the interface if Ec<
Em, whereas it intensifies if Ec>
Em [43]. The former case could cause the crack to deflect towards the pole of the particle and confine the crack to the matrix. The latter case could attract the propagating crack to the equator of the reinforcing particles and facilitate the crack bridging toughening.Experimental study on NilS-glass composites (Eglass
<
ENi and aglasa ~ ow;) revealed thatthe crack propagated entirely in the glass matrix. The elastic stress concentration, developed at the poles of the particles during loading, steered the crack around the particles and avoided utilization of the particle ductility.
2.2 Ductile Toughening 15
Interfacial Bonding Strength
The role of interfacial bonding strength is quite clear in the toughening. According to the crack bridging model, certain interfacial debonding is required to achieve a large effective gauge zone for extensive plastic deformation, but complete debonding should be avoided. In the compos-ites with particulate ductile reinforcements, complete interfacial debonding often occurs due to the weak interfacial bond. Therefore, for this specific type of the metal reinforced ceramics, promoting the interfacial bonding is always an important task.
Many factors may influence the interfacial bond, including the crystallographic relationships, the atomic structure and the match between the reinforcement and matrix, the surface rough-ness, and the interfacial reaction. For metal-particulate/polycrystalline-ceramic composites, the crystallographic relationships and the atomic structure are not so important because of various random combinations at the interfaces. The tensile thermal stress weakens the interfacial bond in the case of a larger thermal expansion coefficient of the metal over that of the ceramic matrix, as discussed earlier. In processing, a transition layer (thin film or coating) will be helpful to relax the stress and enhance the interface strength. The surface roughness of the reinforcement or the morphology at the interface influences the mechanical bond at the interface. Corrugated surfaces with deep reentrant corners could result in a mechanical interlocked (zigzagged) interface. Such interlocking interfaces are expected to be much more resistant to debonding. As exemplified in the Kovar/ glass composites (Table 2.1 ), the acid etching of Kovar particles generates extensive, uniform surface roughness and results in a mechanically interlocked particle/glass interface, re-sponsible for the relatively larger increase in toughness compared with the other two composites.
Interfacial reactions are deemed to affect the interfacial bond. An appropriately controlled inter-facial reaction layer could be beneficial. As indicated in the All glass system {Table 2.1), with a good match in both thermal expansion coefficient and elastic modulus, the oxidation of AI further improved the bonding between AI and the glass matrix by the formation of an Ah03 interlayer. As a result, a very large increase in toughness was achieved due to the crack bridging mechanism.
The interfacial reaction layer can be predicted by thermodynamic calculations. In processing, the reaction can be controlled by atmosphere, temperature and the inclusion of other elements, e.g. alloying. The incorporation of interfacial films or coating also modifies the interfacial chemistry as well as the interfacial thermal stress.
'
For the description of the interfacial bond, a very often cited parameter is the so-called work of adhesion,
w ..
d. defined asl¥11d = im
+
ic - imc (2.6)16 Literature Review respectively, lmc represents the energy of the relaxed interface between the metal and the ce-ramic. The quantity Wad is thus the reversible work released per unit area of interface formed by two free surfaces, or the work required to separate a unit area of interface into its component parts. Obviously, Wad is a parameter to evaluate the interfacial bonding strength, although it is not clear what the relationship between Wad and the fracture strength or fracture energy of the metal-ceramic interface is [44]. In fact, it bas been shown that the interfacial fracture energy is much larger than W~d [45].
In practice, Wad cannot be measured directly, but can be deduced by measuring the contact angle (} established by a metal in contact with a ceramic at equilibrium conditions,
Wad= lm(l
+
cos8) (2.7)Improving the wetting behaviour between the metal and the ceramic will increase the work of adhesion and hence the interfacial strength.
2.3 Nil Ah03 System
The Ni/Ah03 system is one of the most extensively studied metal-ceramic couples. There was a lot of research on the wetting behaviour, the interfacial atomic structure, the thermodynamics and kinetics ofNiO and NiAh04 spinel formation at the interface.
2.3.1 Wetting Character
Generally, liquid Ni does not wet Ah03 and the wetting angle is approximately 128" [46]. The wetting can be improved by oxygen solution in Ni [47]. It can also be promoted by alloying Ni. For example, as shown in Figure 2.5, the interfacial energy between Ni and sapphire (or sintered Ah03) decreases sharply by the addition of a small amount ofTi [48]. This is because of the segregation of Ti by Gibbsian absorption and the formation of Th03 at the interface. In addition to oxygen and alloying, the use of a TiN interlayer also helps wetting and improves Ni!Ah03 bonding [29].
2.3.2 Chemistry of the Interface
The probable reactions at the Ni/Ah03 interface are the formation of NiO or NiAh04 spineL The oxidation of Ni depends on the oxygen partial pressure in the atmosphere and can be predicted from thermodynamic calculations. The NiO can further react with Ah03 to form NiA}z04 spinel. However, it is also possible to form spinel without the presence of NiO [49] [50] [51].
2.3 Ni/Al203 System Interfacial Ener 1.50 1.30 1.10 0.90 0.70 0.1 1 Percent Tl • 0 0 0 0
-17 10Figure 2.5: The influence ofTi addition on the Ni/AlzOJ interfacial energy. After Annstrong et al [48].
Figure 2.6 shows qualitatively the phase relationship of the Ni-Al-0 ternary system at the Ni corner. The diffusion paths represent the dissolution of a-A]z03 into Ni and thus the diffusion of AI and 0 away from the interface. 0 diffuses much faster than AI in Ni and Ni/Al203 diffusion paths extend across the 1-Ni solid solution field predominantly parallel to the Ni-Al border. At low oxygen activity (path 1) the Ni/Ah03 interface remains stable. Assuming that
the diffusion coefficients of 0 and AI in dilute solutions in the 1-Ni are essentially independent of concentration, additional oxygen would shift the diffusion path to higher oxygen levels. At some level a NiAh04 spinel interlayer would be formed (path 2). Trumble and RUhle confirmed this prediction and calculated the threshold oxygen level for the formation of spinel [51]. The threshold activity of oxygen dissolved in Ni is less than the solubility limit of oxygen in Ni in equilibrium with NiO. Therefore, spinel formation proceeds not necessarily through a NiO intermediate.
As stated earlier, interfacial reactions have a large influence on the interfacial properties. The reaction products (thermodynamics) and the extent of the reactions (kinetics) all play a critical role here due to the compatibility of the reaction products with both metal and ceramic. Extensive reactions may cause large volume changes and result in interfacial stresses. NiO was found to be beneficial to Nil glass bonding and optimum strength was obtained when Ni was pre-oxidized (at 700 °C for 30 min). However, the Ni/NiO interface was found to be weaker compared with NiO/glass [4]. In diffusion bonding of Ni to Ah03, NiAh04 spinel formation resulted in a stronger interface and some NiO at the Ni surface is beneficial for the reaction [52]. But excess NiO could deteriorate the bonding due to cracking along the Ni/NiO interface [53]. In another study, the Ni/NiAh04 interface was found to be weaker than the NiAh04/ Ah03 interface [51].
18 Literature Review
NiO 1673K
Ni
AI ---;.Figure 2.6: Qualitative phase relationship of the Ni-rich Ni-Al-0 phase diagram as presented by Metselaar and van Loo [50] showing diffusion paths for Ni/Al203 couples.
2.3.3 Composite Properties
Ni particle reinforced Ah03 has been studied before [26] [27] [28] [30]. Some increase in fracture toughness bas been reported for the composites over the monolithic alumina matrix (Table 2.1). However, it seems that the potential for toughening is not fully reached. Crack bridging occurred at some Ni particles, but cracks often propagated around Ni particles, i.e., along the Nil Ah03 interfaces. Therefore, the interfacial bonding between Ni and Ah03 should be strengthened in order to obtain a well-toughened composite. The work of Tuan and Brook showed that oxygen from the atmosphere dissolved in Ni could benefit to the Ni/Ah03 bonding and result in effective toughening [27]. No interfacial phase was formed although the processing started from NiO following the reductioolsintering route. The work of Sun indicated that the interface could be strengthened by the spinetformation as well as by the control of extent of oxygen promoted wetting [54].
In Ni laminate/A}z03 composites, the fracture toughness was profoundly enhanced due to the crack bridging by the large dimensional Ni laminates [35]. · The interfacial bonding strength was controlled by interfacial tortuosity. An important conclusion is that a more tortuous interface favors the fracture strength. However, it is not advantageous for the fracture toughness, consistent with the debonding requirement according to the bridging model.
2.4 Hydroxyapatite Matrix Ceramics 19
2.4 Hydroxyapatite Matrix Ceramics
2.4.1 General
Calcium hydroxyapatite [Ca10(P04)6(0H)a; HAp], along with other calcium phosphate-based ceramics have been in use in medicine and dentistry for many years, as reviewed by de Groot [55], Hench [56], LeGeros [57] and many other authors. HAp is the principal component of bones and teeth. HAp ceramics have become increasingly promising materials for bone substitution applications because of their full biocompatibility and bioactivity. However, the utilization of this kind of materials as load bearing implantations has been limited considerably because of their brittleness. As shown in Table 2.2, the various mechanical strengths of dense sintered HAp ceramics are comparable or even superior to those of natural bones, whereas the fracture toughness of HAp is much lower compared to that of bones. Furthermore, HAp has a much larger elastic modulus than bones. As a result, HAp has very limited fracture energy (as can be estimated from 1 = l(fcf2E) as compared to natural bones. These mechanical drawbacks of HAp have led to two different approaches for its application: as coatings on metallic components or in composites.
HAp Bone Enamel
(dense sintered) (60-70% HAp) (92-97% HAp)
Structure Hexagonal(P6)lm)
Density(glcm3) 3.156 1.5-2.2 2.9-3.0
Mechanical Strength (MPa)
Compressive 270-900 140-300 250-400 Bending 80~250 100-200 Tensile 90-120 20-114 Diametral 35-95 Fracture Toughness(MPam112 ) 0.6- 1.2 2-12 Vickers Hardness(GPa) 3.0-7.0 0.4-0.7 3.4-3.7 Elastic Modulus(GPa) 35- 120 7-25 40-84 Heat Capacity(J/K·g) 0.765 Thermal Conductance(W/m· K) L3 0.6 0.9
Expansion Coefficient(x I0-6/K) IS
Table 2.2: Comparison of mechanical properties of sintered HAp with calcified tissue of verte-brates. Data mostly based on reference [8] [58].
20 Literature Review
2.4.2 · Monolithic HAp
Structure and Stoichiometry
The crystal structure of HAp belongs to the space group P63
fm
in the hexagonal system withthe lattice parameter a= b = 0.9432 nm and c = 0.6881 nm [61]. At stoichiometric composi-tion [Ca10(P04)6( OH)z], HAp has a Ca/P atomic ratio of 1.67 and a HzO content of 1.79 wt%.
Nonstoichiometry often occurs and it is associated with.Ca deficiencies and adsorbed water, as expressed by
The ratio Ca/P is usually used as an index for the nonstoichiometry. Upon heating, nonstoichio-metric HAp decomposes to stoichiononstoichio-metric HAp and tricalcium phosphate [Ca3(P04)z; TCP) according to the following reactions
Cal0-x(HP04)x(P04)6-x( OHh-x · nH20 -+ Ca10-x(Pz01 )x( (P04)6-:ax( OH)z
+nHzO(g) (
<
700 oC) (2.8) Cato-x(Pz07)x((P04)6-2x(OH}z-+ (1-x)Caw(P04)6(0H)z + 3xCa3(P04h+xHzO(g) .. (700- 800 °C) (2.9)
Stability
The stability of HAp and other calcium phosphates in aqueous environment at room temperature has been studied by de Grootet al [62]. Dicalcium phosphate [CaHP04(0H}z] is the most stable
phase at pH
<
4.2 and HAp is the most stable one at pH>
4.2. TCP and tetracalcium phosphate[C~P209; TICP] can transform into HAp in aqueous solution at body temperature.
At high temperatures at which sintering is usually conducted (1000- 1500 °C}, the stability of HAp depends on the water content in the atmosphere. Figure 2.7 shows the phase diagrams of the system CaO-PzOs and the influence of water vapor pressure [63]. In the absence of water, HAp is not stable. If water is added, HAp starts to form. The decomposition temperature increases with increasing water vapor pressure in the atmosphere. Studies on the decomposition process of HAp show two steps, dehydration at relatively low temperature ("" 800 oC) and decomposition at the temperature shown in the phase diagram (T1). For the dehydration and decomposition of HAp, the following reactions can be written
Caw(P04)6(0H)2 -+ Ca,o(P04)6(0Hh-2,9xDx + xHzO(g) (2.10) Cato(P04)6(0H}z-2x0xDx-+ 2Ca3(P04}z
+
C~P209 + (1-x)HzO(g) (2.11)2.4 Hydroxyapatite Matrix Ceramics T 17110 1100 1SDD 1300 CtO+C4P a'C3P +C4P 1<17&'
aCSP +C4P aCSP + Uq.
1~~--~~--~----~~--~~
70 u c.cPeo 6&C3P &0 - - - wt%Ca0 (a) C3P+C4P T1 CSP +Uq. C4P CaO+C4P +Ap CSP T +Ap Tl! ~ CaO+Ap CSP +CIP cao C4P Ap CSP (c) 21 T 1700 1800 1SDD 1<100 1300 C&O+Ap ciCIP+ClP ,~~--~~--~--~~~--~~ 70
l
a E E 2 D..a 1J
0 - - - wt%Ca0 C4P +Uq. 5.20 (b)~
o'CSP +C4P aCSP , +C<IP T1 Tl5.&0 8.00 lAO 8.&0
104/T, K
(d)
Figure 2.7: (a) Phase diagram of the system CaO-P20s at PH2o
=
0 mmHg; (b) PH2o =500 mmHg; (c) at arbitrary PH2o; (d) Influence of water vapor pressure on decomposition temperature. Note, C3P corresponds to TCP, C4P to TTCP and Ap to HAp.
22 zoo 1110 I
..
II I.SO 1.5S+
-TCP•
a 1.1,() HAP+TCP Literature Review a 12.SO"Cl
•
IISO"Cf
,~~1 2 ) 4b
a
J. Ca/P1.65.+
1.70 1.7S---+- .,...__
HAP+CaO HAPFigure 2.8: The strength dependence of the stoichiometry of HAp (after [64 ]).
where D denotes the vacancies at OH lattice sites. In dehydration, HAp loses OH radicals and forms oxyhydroxyapatite (OHA). In decomposition, OHA further transforms to TCP and TTCP.
Mechanical Properties
The mechanical properties of dense sintered monolithic HAp have been listed in Table 2.2. The fracture strength is dependent on the stoichiometry of HAp. Figure 2.8 shows that the highest strengths are obtained at a composition with a Ca/P ratio ranging from 1.60 to 1.66 (HAp
+
TCP) [64]. The influence of the stoichiometry is related to transformation of (3-TCP into a-TCP {at 1180oq
during sintering. The transformation produces a volume increase which introduces compressive stress in the surface and hence strengthens the nonstoichiometric HAp. Another important feature of the strength is its time dependence. Substantial slow crack growth has been observed in HAp, especially in wet environments [59] [65]. This makes it almost impossible for monolithic HAp to be used as load-bearing comp<>nents for a long period.2.4.3 HAp Matrix Composites
Studies on the mechanical properties of ceramics have progressed with the aim of overcoming the inherent brittleness and realizing the practical use of HAp as bioactive load-bearing implant materials. By preparing composite materials, some improvement has been demonstrated in the fracture toughness of HAp ceramics, as listed in Table 2.3. Various reinforcements have been tried. In many cases the composites cannot be successfully prepared and their mechanical properties cannot be characterized.
2.4 Hydroxyapatite Matrix Ceramics 23
Reinforcement
I<tc/
/{lc,m Processing RemarksRdil
C fibres HP cracks in HAp [8
lOOO"C tion
C fibres HP dehydration and £661 1
decomposition of HAp
Ah03 HP cracks in HAp [8]
1000 "C no reaction
A)z03 fibres HP dehydration and [66]
decomposition of HAp
Ah03 fibres sintering dehydration and [67]
1300 "C decomposition of HAp
Ah03 platelet sintering formation of [68]
1300"C ( CaO )m ( Al:z03 )n
SiC platelets sintering low density [69]
SiC fibres sintering reaction [70]
a 3 HP decomposition [71] (lOvol%) 1050"C of HAp Zr02 3 HP decomposition [71] (lOvol%) 1050"C of HAp Zr02 HP stabilization of [8] 1000 "C ZrO:z y -ZrO:z HP stabilization of [66] Zr02 y Zr02
<
1.6 HIP no stabilization of (26.8wt%) 800-1200"C Zr02stainless steel HP severe interface [8]
fibres IOOO"C reaction
FeCralloy fib HP [8]
(20-30 vol%) lOOO"C
Ti sintering severe reaction this work
He or Ar
FeCralloy fibres sintering low density this work
air severe reaction
Ag particles 2-3.5 sintering this work
(20-30 vol%) air
24 Literature Review
I
aoo-
6/K)II
I
a (lo-6JK>I
• HAp 11-14 Stainless Steel 15c
0.6-4.3 Fecralloy llSiC 4.5 Ti 9
ShN4 3.5 Ag 20
Ah03 8-11 Pt 9
Zr02 9-11 Au 14.1
Table 2.4: Thermal expansion coefficient of various reinforcements.
The general problems associated with the fabrication of the HAp based composites are (1) the low density and the presence of defects (cracks); (2) the degradation of the reinforcements due to severe reactions with the HAp matrix; and (3) the dehydration and decomposition of the HAp caused by the sintering atmosphere and the presence of the reinforcements. Some complicated techniques, e.g. hot (iso~tatic) pressing, are used to help the densification and to suppress reactions (due to lower temperature).
The selection of the reinforcements is critical. In the first place, the reinforcement should be biocompatible. As cracking is concerned, the match in thermal expansion coefficient should be considered. Table 2.4 compares the thermal expansion coefficient of the most used reinforce-ments with that of HAp matrix. To toughen HAp, a reinforcement with higher thermal expansion coefficient is preferred. In this way compressive stress will be introduced into the HAp matrix. Since HAp bears very high thermal expansion, metallic reinforcements are preferred. Further, considering the stability of HAp, some noble metals should be used, which allow the sintering in a moist oxygen atmosphere.
Ag fulfills these requirements and is selected in the present work. Biologically, Ag shows anti-bacterial properties in vitro, and HAp adsorbed with Ag (concentration up to 4 p.glmg) exhibits very similar behaviour as the control HAp, with remarkably good osteoconduction and . incorporation in bone in vivo implantation [73] [74]. However, the interaction between Ag and HAp is totally unknown, in contrast to the well defined Ni/Ah03 system.
2.5
Summary
Studies on structural ceramics have been focused on the development of materials with improved mechanical reliability. Among the various approaches, incorporation of ductile reinforcements can significantly enhance the fracture toughness of brittle ceramic matrix. The increase in toughness is mainly due to the crack bridging processes.
2.5 Summary 25 The crack bridging processes take advantage of the ductility of the reinforcement. The param-eters which control the toughening contribution include the microstructure (morphology), size and yield strength of the ductile reinforcement, the interfacial bonding and the match between the reinforcement and the matrix. Optimized toughening requires an appropriate interfacial debonding which allows extensive deformation in a large effective gauge length.
As particulate reinforcement is concerned, due to its shape and size the general situation is a weak interfacial bonding which makes the bridging impossible. Pull-out of reinforcements and crack deflection may occur, but these two processes have only limited toughening effects. The interfacial strength needs to be improved to assure the occurrence of crack bridging and plastic deformation of the reinforcement. This can be achieved by changing the interfacial chemistry. Methods include atmosphere control, interfacial reaction, alloying addition and interlayer in-corporation. Other factors should also be considered in order to have a strong matrix and to attract a propagating crack to the reinforcing particulates. Among them, the two most important criteria are O:m ;?: O:c and Em $ Ec.
In Ni/ Ab03 composites, toughening due to crack bridging by Ni particles has been observed, but in many cases interfacial debonding prevents further toughening. The potential of the Ni duc-tility is not fully utilized. Adding Ti improves the wetting between Ni and Ab03 and promotes
interfacial bonding. Therefore, this provides an opportunity to further enhance toughening.
In calcium hydroxyapatite based composites, Ag and other metal reinforcements have some advantages over ceramic ones. Ag is especially preferred from the biological point of view.
References
1. M. A.· Stett and R. M. Fulrath, "Strengthening by Chemical Bonding in Brittle Matrix Composite", J. Am. Ceram. Soc.-Discussion and Notes, 51 [10] 569- (1968).
2. Y. Nivas and R. M. Fulrath, "Limitation of Griffith Flaws in Glass-Matrix Composites",
J. Am. Ceram. Soc., 53 [4] 188- (1970).
3. D. R. Biswas, "Strength and Fracture toughness oflndented Glass-Nickel Compacts", J. Mater. Sci., 15 1696-1700(1980).
4. V. V. Krstic and P. S. Nicholson, "Toughening of Glasses by Metallic Particles", J. Am. Ceram. Soc., 64 [9] 499-504 (1981).
5. M. H. Moore and S.C. Kunz, "Metal Particle-Toughened Borosilicate Sealing Glass",
Ceram. Eng. Sci. Proc., Vol. 8, No. 7-8,839- (1987).
6. G. Baran, "Fracture Toughness of Metal Reinforced Glass Composites", J. Mater. Sci., 25 4211- (1990).
26 Literature Review
7. P. Hing and G. W. Groves, 'The Strength and Fracture Toughness of Polycrystalline Magnesium Oxide Containing Metallic Particles and Fibres", J. Mater. Sci., 7 427-(1972).
8. G. de With, A. J. Corbijn, "Metal-fibre Reinforced Hydroxy-apatite Ceramics", J. Mater. Sci., 24 3411- (1989).
9. A. J. Pyzik, I. A. Aksay and M. Sarikaya, "Microdesigning of Ceramic-Metal Compos-ites", p45-54 in Ceramic Microstructures '86, Role of Interface, Ed. J. A. Pask and A. G. Evans, (1987).
10. D. Han and J. J. Mecholsky, Jr. "Fracture Analysis of Cobalt-bonded Tungsten Carbide Composites", J. Mater. Sci., 25 4949-4956 (1990).
11. Wenjie Si, "Ceramic Composite Toughened by Phase Transformation and Particulate Dispersion", Thesis ofTsinghua University, China, (1992). ·
12. F. Yeh and K. W. White," Fracture Toughness Behaviour of the YBa2Cu307.;." Supercon-ducting Ceramic with Silver Oxide Additions", J. Appl. Phys., 70 [9] 4989-94 (1991).
13. L. S. Yeou and K. W. White, ''The Development of High Fracture Toughness YBa2Cu301-x./Ag Composites", J. Mater. Res., 7 [1] 1-4 (1992).
14. C. 0. Mchugh, T. J. Whalen and.M. Humenik, Jr. "Dispersion-Strengthened Aluminum Oxide", J. Am. Ceram. Soc., 49 [9] 486-91 (1966).
15. D. T. Rankin, J. J. Stiglich, D. R. Petrak and R. Rub, "Hot-pressing and Mechanical Properties of A)z03 with an Mo-Dispersed Phase", J. Am. Ceram. Soc., 54 (6] 277-81 (1971).
16. X. Sun, P. A. Trusty, J. A. Yeomans, "The Fabrication and Properties of Alumina-Ductile Metal Partic1eComposites", iccm/8;july, (1991).
17. X. Devaux, C. Laurent, M. Brieu et A. Rousset, "Proprietes Microstructurales et Mecan-niques de Nanocomposites a Matrice Ceramique", C. R. Acad. Sci. Paris, t. 312, Serie IT, p. 1425-1430 (1991).
18. W. H. Tuan and W. B. Chou, ''Toughening Alumina with Silver Inclusions", in Euro-Ceramics, Proceeding of the 2nd ECerS, Germany, ( 1991 }.
19. J. Wang, C. B. Ponton and P.M. Marquis, "Silver-Toughened Alumina Ceramics", Br. Ceram. Trans., 92, No. 2, 67-74 (1993).
20. M. K. Aghajanian, N. H. macmillan, C. R. kennedy, S. J. Luszcz and R. Roy, "Properties and Microstructures of Lanxide Ah03-Al Ceramic Composite Materials", J. Mater. Sci.,
l4 658-670 ( 1989}.
21. F. F. Lange, B. V. Velamakanni and A. G. Evans, "Method for Processing Metal-reinforced Ceramic Composites", J. Am. Ceram. Soc., 73 [2] 388-93 (1990).
22. B. D. Flinn, C. S. Lo, F. W. Zok and A. G. Evans, "Fracture Resistance Characteristic of a Metal-Toughened Ceramic", J. Am. Ceram. Soc., 76 [2] 369-75 (1993).
2.5Summary 27
23. S. P. Ray and D. I. Yun, "Squeeze-Cast AhO:/Al Ceramic-Metal Composites", Ceram. Bull. vol. 70, No. 2, (1991).
24. Suxing Wu, A. J. Gesing, N. A. Travitzky and N. Claussen, "Fabrication and Properties of At-infiltrated RBAO-based Composites", J. Eur. Ceram. Soc., 7 277~281 (1991). 25. P. D. Djali and K. R. Linger, "The Fabrication and Properties of Nickel-Alumina Cermets",
Proceedings ofthe British Ceramic Society, pp 113-127 (1978).
26. E. Breval, G. Dodds and C. G. Pantano, "properties and MicrostrUctures of Ni-Alumina Composite Materials Prepared by the Sol-Gel Method", Mat. Res. Bull., VoL 20, pp 1191-1205 (1985).
27. W. H. Tuan and R. J. Brook, "The Toughening of Alumina with Nickel Inclusions", J. Eur. Ceram. Soc., 6 31-37 (1990).
28. X. Sun and J. A. Yeomans, "The Toughening of Alumina Matrices by the Inclusion of Nickel Particles", Special Ceramics, 9, (1990).
29. Shujie Li, "Investigation of the Powder Metallurgical Route for Manufacturing Al203/Ni
Cermets with Interlayers", Thesis of University ofTwente, The Netherlands, (1992). 30. T. Ekstrom, "Alumina Ceramics with Particle Inclusions", J. Eur. Ceram. Soc., 11
487-496 (1993).
31. M. F. Ashby, F. J. Blunt and M. Bannister, "Flow Characteristics of Highly Constrained Metal Wires", Acta Metall., Vol. 37, No. 7, 1847-1857 (1989).
32. H. R. Cao, B. J. Dalgleish, H. E. Deve, C. Elliott, A. G. Evans and R. Mehrabian, "A Test Procedure for Characterizing the Toughening of Brittle Intermetallics by Ductile Reinforcements", Acta Metall., Vol. 37, No. 11, 2969-2977 {1989).
33. H. E. Deve, A. G. Evans, G. R. Odette and R. Mehrabian, "Ductile Reinforcement Toughening of 7-TiAl: Effects of De bonding and Ductility", Acta Metall Mater., Vol. 38, No. 8, 1491-1502 (1990).
34. T. C. Lu, A. G. Evans, R. J. Hecht and R. Mehrabian, ''Toughening ofMosi2 with a Ductile (Niobium) Reinforcement", Acta Metall. Mater., Vol. 39, No. 8, 1853-1862 (1991). 35. Z. Chen and J. J. Mecholsky, Jr., "Control of Strength and Toughness of Ceramic/Metal
Laminates Using Interface Design", J. Mater. Res., Vol8, No. 9, 2362-2369 (1993). 36. J. R. Rice, "Elastic Fracture Mechanics Concept for Interfacial Cracks", Transactions of
the ASME, 98Nol. 55, March, (1988).
37. B. Budiansky, J. C. Amazigo and A. G. Evans, "Small-Scale Crack Bridging and the Fracture Toughness of Particulate-Reinforced Ceramics", J. Mech. Phys. Solids, 36 [2] p167-, (1988).
38. L. S. Sigl, P. A. Mataga, B. J. Dalgleish, R. M. Mcmeeking, "On The Toughness of Brittle Materials Reinforced with a Ductile Phase", Acta Metall., Vol. 36, No. 4, 945-953 (1988). 39. B. D. Flinn, M. RUhle and A. G. Evens, "Toughening in Composites of Al:!03 Reinforced
28
Literature Review 40. K. T. Faber and A. G. Evans, "Crack Deflection Processes -I. Theory", Acta Me tall., Vol.31, No.4, 565-576(1983).
41. Jorgen Selsing, "Internal Stress in Ceramics", J. Am. Ceram. Soc., 44 [8] 419- (1961). 42. J. N. Goodier, "Concentration of Stress Around Spherical and Cylinderical Inclusions and
Flaws", J. Appl. Mech., 1 [1]39-44 (1933).
43. F. Erdogan, ''The Interface Between Inclusions and Cracks", p 245-267 in Fracture Mechanics of Ceramics 1, edited by R. C. Bredt, D. P. H. Hasselman and F. F. Lange, (1974).
44. M. C. Nicholas. "Interactions at Oxide-Metal Interfaces", Mater. Sci. Forum, Vol. 29, 127-150 (1988).
45. G. E1ssner, T. Suga and M. Turwitt, "Fracture of Ceramic to Metal Interfaces", Journal de Physique, C4, Vol46, pp 597-611 ( 1985).
46. M. Humenik and W. D. Kingery, "Metal-Ceramic Interactions: ill, Surface Tension and Wettability of Metal-Ceramic Systems", J. Am. Ceram. Soc., 37 [1] 18-23 (1954). 47. U. V. Naidich, ''The Wettability of Solids by Liquid Metals", Prog. Sur.fi Membr. Sci.,
14 p353- (1981).
48. W. M. Armstrong, A. C. D. Chaklader and J. F. Clarke, "Interface Reactions Between Metals and Ceramics: I, Sapphire-Nickel Alloys", J. Am. Ceram. Soc., 45 [3] 115-118 (1962).
49. M. Backhaus-Ricoult, "Diffusion Process and Interphase Boundary Morphology in Ternary Metal-Ceramic Systems", Ber. Bunsenges. Phys. Chem., 90 684-690 (1986).
50. R. Metselaar and F. J. J. Van Loo, "The Use of Phase Diagrams for the Study of Metal-Ceramic Interdiffusion", Ceramic Developments, ed. C.C. Sorrel and B. Ben-Nissan, Materials Science Forum, Vol. 34-36, pp 413-420 (1988).
51. K. P. Trumble and M. RUhle, ''The Thermodynamics of Spinel Interphase Formation at Diffusion-Bonded Ni/Ah03 Interfaces",Acta Metall. Mater., Vol. 39, No. 8, 1915-1924 (1991).
52. C. A. Calow, I. T. Porter, ''The Solid State Bonding of Nickel to Alumina", J. Mater. Sci., 6 156-163 (1971).
53. F. P. Bailey and W. E. Borbidge, "Surface and Interface in Ceramic and Ceramic Metal Systems", edited by J. Pask and A. Evans, Plenum Press, New York and London, p 252-(1981).
54. X. Sun, "Aspects of the processing and Properties of Ni Particle Toughened Alumina", Thesis of University of Surrey, United Kingdom, ( 1993).
55. K. de Groot, "Biocompatibility of Clinical Implant Materials", Vol. 1, ed. D. F. Williams, CRC Press, Boca Rato, p 199- (1981).
56. L. L. Hench, "Bioceramics: From Concept to Clinic", J. Am. Ceram. Soc., 74 [7] 1487-510 (1991).
