Processing of Homogeneous Zirconia-Toughened Alumina Ceramics withHigh Dry-Sliding Wear Resistance

Bas Kerkwijk,† Louis Winnubst,† Elmer J. Mulder,‡ and Henk Verweij*,†

Laboratory of Inorganic Materials Science and Laboratory of Tribology, University of Twente,7500 AE Enschede, The Netherlands

The preparation of dense homogeneous zirconia-toughenedalumina (ZTA) with high dry-sliding wear resistance is de-scribed. These ZTA ceramics are obtained by sinteringgreen compacts, made by colloidal filtration of well-definedZrO 2–Al2O3 particle suspensions, for 2 h at 1400°C. Theoptimum solid and stabilizer concentrations for the filtra-tion process were determined. The sintered ZTA micro-structure consists of a homogeneous distribution of zirconiagrains in an alumina matrix with grain sizes of 0.2 and 0.5µm, respectively. In pin-on-disk tribological measurementsat relatively high initial contact pressures (1130 MPa) andsliding speeds (0.5 m/s) a very low specific wear rate (about5 × 10−8 mm3/(N?m)) and a coefficient of friction of 0.45–0.55 were found. It is shown that, in this case, wear isdominated by abrasion and polishing.

I. Introduction

AT PRESENT, lubricants are often used to reduce wear andfriction in sliding contacts of moving parts in equipment.

The use of these lubricants, however, becomes more and moreunwanted for environmental and other reasons. In practicalunlubricated sliding contacts the touching materials must becarefully chosen such that the specific wear rate is well below10−6 mm3/(N?m) with a coefficient of friction preferably lowerthan 0.2.1,2 The wear mechanism should in any case be mild,meaning that the contacting materials preferably polish eachother, resulting in a very low specific wear rate. The unwantedopposite of mild wear is “severe” wear, in which contact-surface roughening takes places due to breakout of large piecesof material. In many cases a transition takes place from mild tosevere wear if the load is increased.3 In addition, the occur-rence of fatigue at a constant load may trigger a rapid transferfrom mild wear to severe wear after a certain sliding time (ordistance).

In the literature it is suggested that for many cases, the“ideal” dry-sliding couple has not yet been found.4 This may beconnected with the fact that the materials investigated untilnow did not fulfil the requirements mentioned above. One ofthe major problems is the fact that the homogeneity of theavailable ceramic materials is insufficient for the applicationsin mind.5 The homogeneity may be expressed in terms of theoccurrence of, for instance, processing defects, unwanted sec-ond phases, too large grains,6 and spatially correlated porosity.Many investigators have given examples of tribological prop-

erties of oxide ceramics, but not all use microstructural obser-vations to identify the possible differences in wear and friction.

Hsu and Shen7 reported mild wear for dense alumina with5 mm grains at sliding speeds up to 0.2 m/s. This alumina hasa mild-to-severe wear transition pressure of 700 MPa. Using alower sliding speed of <0.01 m/s, mild wear was found atpressures up to 2 GPa. Since the large grain size and lowtoughness of pure dense alumina are less favorable for dry-sliding applications, alternative material compositions shouldbe considered as well. Dense zirconia ceramics can be obtainedwith much smaller grains and higher toughness than alumina.SiC and Si3N4 (ceramics with covalent bonding) are muchharder and SiC has a higher thermal conductivity than theoxides. However, wear studies7 of yttria-stabilized tetragonalzirconia (Y-TZP), SiC, and Si3N4 ceramics showed a transitionfrom mild to severe wear at much lower sliding speed and at alower contact pressure than for single-phase alumina. The mostobvious disadvantages for the use of zirconia are the low hard-ness and low thermal conductivity. The ceramics with covalentbonding are more difficult to process, generally leading to lesshomogeneous microstructures. In addition, the non-oxides SiCand Si3N4 are easily oxidized and form soft SiO2 films at thesurface that rapidly degrade. This process leads to unacceptablehigh wear rates.1 As a consequence, developing high-wear-resistant materials by starting with dense alumina ceramics andreducing its less-favorable properties by proper nonsoluble ad-ditives seems to be a good strategy. Hence for the present studywe have focused on preparation and properties of fine-grainedcomposites with an alumina matrix in which isolated smallzirconia grains are present. Such a composite is well known inthe literature as zirconia-toughened alumina (ZTA). In ZTAwith 8–15 vol% zirconia the favorable properties of aluminacan be maintained since ZTA is nearly as hard and as thermallyconductive as pure alumina. The zirconia may introduce extratoughening8–10 and lead to a smaller (submicrometer) grainsize in the ceramics. The enhanced toughness of the aluminaphase may be caused by a latent possibility of martensitic phasetransformation of the zirconia phase, but toughening due tocrack-deflection is also possible, especially because of thesmall grain sizes used in the present study. A homogeneousdistribution of small zirconia particles also acts as an impedi-ment for alumina grain growth during sintering.9,11

He et al.3 studied dry-sliding wear of ZTA, among others.The authors showed that the value of the mild-to-severe wear-transition load depends on the following:

● The zirconia content: An increase to 12 vol% results inan increase in wear-transition load.

● The alumina matrix grain size:Increasing grain sizeresults in a decreasing wear-transition load.

He et al.8 observed mild wear with a very low wear rate of2 × 10−8 mm3/(N?m) for ZTA under reciprocal dry slidingagainst a stainless steel plate. Test conditions were an initialmean contact pressure of 300 MPa and an average slidingspeed of 0.02 m/s. The wear mechanism observed was mainlyadhesive material transfer of a metallic layer on the ceramicsurface. An increase in wear resistance by the addition of zir-conia to alumina has also been reported by others.12–14Trabelsi

R. O. Scattergood—contributing editor

Manuscript No. 190226. Received April 29, 1998; approved December 28, 1998.Supported by Technology Foundation STW.*Member, American Ceramic Society.†Laboratory of Inorganic Materials Science, Department of Chemical Engineering.‡Laboratory of Tribology, Department of Mechanical Engineering.

J. Am. Ceram. Soc., 82 [8] 2087–93 (1999)Journal

2087

et al.10 stated that the addition of zirconia to alumina results ina material with a lower wear resistance due to the decrease inhardness. Most investigators conclude, however,3,9 that an op-timum amount of zirconia of 10–12 vol% exists, which pro-vides enough improvement in toughness and decreased grainsize to compensate for the lower hardness. Kamiyaet al.15

performed collision tests (erosive wear) on alumina and ZTAceramics with Al2O3 or SiC particles and showed that ZTA hadthe lowest wear rate. Unfortunately, detailed systematic studieson the effect of ZTA microstructure on wear resistance are notyet available, meaning that not much material for comparisonis present.

To obtain fine-grained ZTA ceramics with optimum wearproperties, it is important that (1) the green compact consists ofa homogeneous, close packing of well-defined particles with anarrow size distribution in the submicrometer range, (2) thezirconia phase is homogeneously distributed in the matrix inthe sintered alumina, and (3) the sintered material is defect-poor (low amount of faults and cracks). During the consolida-tion or green-forming step, several defects can be introduced,which remain or can even grow during the sintering step. Es-pecially dry pressing of powders is known as a technique wheredefects are easily introduced because of density variations and(remainders of) powder agglomerates. It is generally ac-cepted16–23 that consolidation techniques based on colloidalprocessing result in a more defect-poor green ceramic body. Toachieve this, it is necessary to prepare a homogeneous, stablesuspension of well-defined nonagglomerated particles with anarrow size distribution. Unstable suspensions are known toresult in an inhomogeneous green phase distribution due toflocculation/segregation effects.17,18,23Colloidal consolidationtechniques such as pressure filtration, slip casting, centrifugalcasting, or electrophoretic deposition may (easily) lead to ho-mogeneous green compacts with a close-packed particle stack-ing. The use of such a high-density homogeneous stackingmay also lower sintering temperatures needed to obtain denseceramics.20

In the present paper the colloidal consolidation and tribo-logical properties of ZTA ceramics are described. The objec-tive of the present study has been investigation of possibilitiesfor improvement of the tribological properties of ZTA materi-als by increasing the homogeneity of the phase distribution.Therefore, the preparation of very well-defined microstructuresis necessary and hence colloidal processing is chosen as theconsolidation method.

II. Experimental Procedure

(1) Powder PreparationUndoped zirconia for colloidal processing was prepared by

precipitation of a diluted ZrCl4 (Merck, Darmstadt, Germany)solution in an excess of ammonia. Further details of thismethod can be found elsewhere.24 After being washed withwater and ethanol and then dried, the zirconium hydroxideprecipitate was calcined at 500°C for 2 h. This resulted in a100% tetragonal zirconia phase as confirmed by XRD, with anaverage crystallite size of about 10 nm, calculated from XRDline broadening using the (111) reflection at 30° 2u. The BETsurface area of this powder was 100 m2/g. The crystallite sizeof the a-alumina powder (AKP-50, Sumitomo Chemical Co.Ltd., Osaka, Japan) was 0.1–0.3mm as specified by the manu-facturer. The BET surface area of this powder was 10 m2/g.

(2) Colloidal ConsolidationVarious single-phase suspensions of alumina and zirconia

were prepared by using HNO3 as a stabilizing agent and ball-milling these systems with zirconia balls (f 4 2 mm) for atleast 16 h. Acid concentrations used were 0.05M, 0.10M, and0.15M, based on the pure water phase. Acid concentrationslower than 0.05M gave unstable suspensions. The aluminasolid concentration was varied between 33 and 75 wt%. The

zirconia solid concentration was kept constant at 33 wt%. Thesingle-phase suspensions were subsequently mixed for 1 h inthe appropriate ratio (85 wt% Al2O3, 15 wt% ZrO2) using theabove-described milling equipment and then filtered over amembrane filter with a pore size of 0.45 mm (ME25, Schlei-cher and Schuell, Dassel, Germany). The suspension surfaceremained in contact with the atmosphere while the drivingforce for the filtration process was created by a vacuum. Thedisk-shaped green compact (f 4 42 mm) could easily bereleased from the membrane filter after 10 h of drying in stag-nant air at room temperature. The green compacts were sin-tered for 2 h at 1350°, 1400°, or 1450°C (with heating andcooling at 2 and 4°C/min, respectively). Microstructural inves-tigation was performed using SEM on polished and thermallyetched samples. Grain sizes were determined using the linearintercept technique.25 Cross-sectional SEM pictures of thesamples were used to study possible segregation effects duringfiltration.

(3) Preparation of Conventional ZTA forComparison Purposes

Results of the tribological measurements of the colloidallyprocessed ceramics were compared with those for dense ce-ramics made with a commercially available, spray-dried 85wt% alumina and 15 wt% zirconia powder mixture. This ma-terial, further indicated as ZTACOM (Zirconia Sales, Surrey,Great Britain), was made by conventional uniaxial preformingof 48 mm f pellets at 30 MPa and isostatic pressing at 400MPa. Sintering was done using an initial heating rate of 1°C/min up to 400°C and 1 h dwell for binder burn-out, followed byheating at 2°C/min to 1600°C and a 2 h dwell followed bycooling down at 4°C/min.

(4) Wear TestingWear tests on ceramic disks were performed at room tem-

perature on a pin-on-disk tribometer (CSEM, Neuchatel, Swit-zerland). Commercially available 10 mm alumina bearing balls(grade 10, Gimex Technical Ceramics BV, Geldermalsen, TheNetherlands) were used as pin material. The geometry of thetest setup is shown in Fig. 1. The sliding velocity in the contactwas set at 0.5 m/s. A constant force,F, of 15.5 N was applied,resulting in an initial mean Hertzian contact pressure of 1130MPa. Dry-sliding wear was measured in a controlled air atmo-sphere using a climate chamber (Heraeus HC4057, Balingen,Germany). The temperature was set at 23°C and the relativehumidity at 40%. The wear volume,Vw, was determined aftera sliding distance,s, of 120 km. Wear is expressed as a specificwear ratekw (in mm3/(N?m)), using

kw =Vw

Fs(1)

The wear volume,Vw, was determined by measuring theweight loss of the sample or by determination of the wearvolume from the wear track depth as measured by interferenceprofiling (ATOS micromap, Pfungstadt, Germany). Both mea-surements gave values of the same order of magnitude. Thespecific wear rate was determined from a minimum of threetests. The friction was measured continuously during the ex-

Fig. 1. Tribological test setup.

2088 Journal of the American Ceramic Society—Kerkwijk et al. Vol. 82, No. 8

periments and is expressed as a coefficient of friction. Thecoefficient of friction is defined as the ratio between the mea-sured friction force and the applied normal force.

III. Preparation of ZTA by Colloidal Filtration

(1) ResultsIn Fig. 2 is shown the microstructure of ZTA obtained by

colloidal filtration and subsequent sintering at 1400°C for 2 h.This sample is 99% dense and the grain size is 0.5 and 0.2mmfor alumina and zirconia, respectively. SEM pictures of severalparts of a cross section of this ceramic showed the same ho-mogeneous microstructure with identical alumina and zirconiaphase concentrations. The sample height of the disk-shapedsamples was up to 4 mm, depending on the solids loadingduring filtration.

The composition of the suspensions used, named ZTA/1–5,and the green and sintered densities of the consolidated com-pacts are given in Table I. The ceramics made from the sus-pensions are indicated with the same designation code as thesuspension. As can be seen in Table I, the alumina solid andnitric acid concentrations have a significant effect on the den-sity of the ceramics after sintering at 1450°C. No consistentchanges are found in the green densities of the consolidatedcakes. However, ZTA/3 and ZTA/5 ceramics are the only onesthat reach nearly full density. In Fig. 3, a SEM picture is shownof the microstructure of the top part of the ZTA/1 ceramic disk.From this picture and in comparison with Fig. 2 it is concludedthat a low concentration of alumina in the starting suspensionresults in a significantly lower zirconia concentration in the toppart of the ceramic disk; so no homogeneous distribution of thesecond phase is obtained. This implies that in ZTA/1 ceramics,there is a gradient in zirconia content in the original filtrationdirection. Similar results were obtained with disks made withZTA/2. Figure 4 shows that a high HNO3 concentration (inZTA/4) results in clustering (by flocculation) of the zirconiaparticles throughout the ceramic disk. In addition, the top layerof the disk-shaped samples pealed off in most cases if a highacid concentration was used. Hence, the optimum solid andnitric acid concentration for obtaining a homogeneous ZTAcompact, within the scope of the used materials and stabiliza-tion route, is regarded to be the composition of ZTA/5, becauseof the higher solids loading and green density compared to

ZTA/3. Details on ceramic microstructure (density, grain size)of ZTA/5 are given in Table II.

A SEM picture of ZTACOM is given in Fig. 5. This mate-rial, made with a commercially available powder, shows arather coarse structure and the homogeneity of the phase dis-tribution is inferior to that of our colloidal processed ceramic.Especially the grain boundaries of the ZTACOM are not well-defined. When the material is sintered at 1450°C for 2 h, adensity of 90% is obtained, so a high sintering temperature isnecessary. A density of 98% is reached only after sintering at1600°C for 2 h, resulting in grain sizes of 3.1 and 0.9mm foralumina and zirconia, respectively.

(2) DiscussionThe alumina powder concentration in the single-phase sus-

pension and the nitric acid concentration prior to mixing influ-ence the zirconia–alumina phase distribution after consolida-tion (Figs. 2 to 4). The influence of these concentrations can beexplained by two phenomena: sedimentation (possibly result-ing in segregation) and flocculation. The influence of the ex-perimental conditions on the occurrence of these phenomena isalso shown in Table I.

In the zirconia–alumina suspensions sedimentation will oc-cur during the consolidation step. In a sufficiently stable sus-pension this sedimentation will be very limited. In a nonstablesuspension, sedimentation occurs more rapidly. This sedimen-tation may be either a collective movement of both types ofparticles (no segregation) or a separate movement of bothphases (segregation). From the results it is found that if thealumina concentration is too low, the zirconia can easily moveindependently from the alumina particles and settle more rap-idly. This means that sedimentation by gravitation takesplace.26,27Concha and Almendra28 describe sedimentation ve-locities of suspensions of spherical particles. They describe thesedimentation velocity of the particles according to Stokes’law, but the velocity is corrected for the influence of the solidconcentration of the suspension. This correction is expressed asthe hindrance factor,h. From their expressions the appropriateexpression for the correction factor,h, for hindered settling isderived,29 which is valid for particle volume fractions,f, up to0.59 and with the Reynolds number limited to 0:

h~f! =~1 − f!~1 − 1.45f!1.83

1 + 0.75f1/3 (2)

If h is close to unity there is no hindrance (particles sedimentaccording to Stokes’ law) and ifh << 0.5 there is a significantamount of mutual particle hindrance.29 In Table I,h values aregiven, calculated from Eq. (2) and the assumption that bothalumina and zirconia phases contribute tof. In Table I anindication is given that the transition from individual particlemovement to collective particle movement takes place atf ≈ 0.2 andh ≈ 0.3; these values are extrapolated from thevalues between 0.38 and 0.24 for the hindrance factor,h. Thiswould correspond to an alumina solid loading of about 40 wt%.As a consequence, a minimum of 50 wt% alumina is used inthe single-phase suspension to minimize zirconia sedimenta-tion after mixing, processing, and sintering. Suspensions withlarge concentrations of alumina (ZTA/5 and ZTA/6) do nothave a visible segregation effect.

The large solid concentration not only slows down the sedi-mentation of zirconia, but also influences the number of colli-sions between the particles in suspension. Because the numberof collisions is higher, reversible flocculation of particles ofboth types will be more pronounced. This flocculation mayprevent sedimentation from occurring.30,31Besides that, Willi-ams et al.31 state that when the particles in suspension arebidisperse, this flocculating effect is enhanced. There are morephenomena that can alter the particle movement in concen-trated suspensions, but for this discussion we have focused onboth the hindrance factor for sedimentation in concentratedsuspensions and the flocculation effect. However, the use ofFig. 2. Microstructure of ZTA/5 after sintering at 1400°C.

August 1999 Processing of Homogeneous Zirconia-Toughened Alumina Ceramics 2089

solid concentrations of 75 wt% in suspension ZTA/6 causedhandling problems during consolidation because the viscosityof the suspension became too high.

The f-value where transition from individual to collectiveparticle behavior is found is significantly lower than the valueof 0.4 reported by Locketet al.32 for a bimodal particle system,and the value of 0.5 reported by Velamakanniet al.17 based oninterparticle potential arguments. The discrepancy is probablyrelated to the fact that the systems investigated in Refs. 17 and32 use powders that have a smaller BET surface area. Espe-cially the zirconia powders used in this study have a very largesurface area, meaning that particle interactions will be verystrong. Thus, flocculation without segregation can take place atlower solid concentrations.

Separate flocculation of zirconia in the sintered ceramics isobserved when too high nitric acid concentrations are used.This flocculation is caused by the classical “salt” effect, ex-plained by the DLVO theory. This means that the stabilizingeffect of the acid is diminished by the increased concentrationof counterions and the particles will flocculate and form clus-ters in the suspension structure. It is not known whether theseclusters are formed during removal of water from the binarysuspension or whether they are already formed in the single-phase suspension. The latter is more likely, because the verysmall zirconia particles show a natural tendency toward rapidagglomeration.

The optimization of processing parameters for colloidal pro-cessing shows that suspensions should be stabilized using0.05M HNO3 with a zirconia concentration of 33 wt% and an

alumina concentration of 66 wt% in the single-phase startingsuspensions (ZTA/5). ZTA/5 is chosen because the materialalready reaches nearly full density at 1400°C. The results showthat by using the described consolidation technique (with theproper experimental parameters), a high green density of 66vol% is obtained, resulting in dense sintered ceramics with ahomogeneous phase distribution. In addition, the grain sizes forboth phases in ZTA/5 are relatively small. Because both par-ticle types tend to impede each other’s grain growth processduring sintering, it can be assumed that a limited grain growthof alumina in the colloidally processed compact is an indicationof an initially homogeneous stacking of the alumina particles inthe matrix. It is found that in the ZTA/5-based ceramics, alu-mina grains grew from 0.23mm slightly agglomerated particlesto 0.51mm grains. In the same way zirconia grew from 10 nmhighly agglomerated particles to 0.16mm grains after sinteringat 1400°C.

The mutual grain-growth impediment effect is clearly dem-onstrated if one looks at the final alumina grain size obtainedwith inhomogeneous green compacts. In ceramics made fromthe suspension with low solid loading and those made with atoo high acid concentration, the alumina grain size is muchlarger: 1.6 and 1.1mm, respectively.

IV. Tribological Properties

(1) ResultsUnder relatively extreme conditions (initial mean contact

pressure 1130 MPa, sliding speed 0.5 m/s), the wear of ZTA/5

Table I. Compositions of the Suspensions Used for Colloidal Filtration of Green Compacts, Their Density in the Green Stateand after Sintering at 1450°C for 2 h, Total Suspension Solid Loading,f, and Hindrance Factor, h, Calculated from Eq. (2)

and the Experimentally Derived Occurrence of Collective Particle Movement and Flocculation

[HNO3] Al 2O3 ZrO2 rgreen r1450°C Collective particleSuspension (mol?L−1) (wt%) (wt%) (%) (%) f h movement Flocculation

ZTA/1 0.05 33 33 61 97 0.15 0.38 No NoZTA/2 0.15 33 33 61 92 0.15 0.38 No YesZTA/3 0.05 50 33 54 >99 0.24 0.24 Yes NoZTA/4 0.10 50 33 59 98 0.24 0.24 Yes YesZTA/5 0.05 66 33 67 >99 0.33 0.13 Yes NoZTA/6 0.05 75 33 59 97 0.38 0.09 Yes No

Fig. 3. Microstructure of ZTA/1 (with low solid content during col-loidal filtration) after sintering at 1450°C.

Fig. 4. Microstructure of ZTA/4 (with high nitric acid concentrationduring colloidal filtration) after sintering at 1450°C.

2090 Journal of the American Ceramic Society—Kerkwijk et al. Vol. 82, No. 8

remains in the mild regime until failure of the alumina testsphere. From all tests, the average specific wear rate,kw, was5 × 10−8 mm3/(N?m) (±0.5 × 10−8 mm3/(N?m)) for the entiremeasurement (using 120 km sliding distance). For ZTACOM,kw was as high as 2 × 10−6 mm3/(N?m) (±0.5 × 10−6 mm3/(N?m)). In all cases the coefficient of friction was between0.45 and 0.55. Microstructural observations were used to gaininsight into the wear mechanisms. Figures 6 and 7 showwear tracks for both materials. In Fig. 6(a) the wear track ofZTACOM reveals a grooved pattern. Figure 6(b) shows asimilar pattern for ZTA/5, but the distance between the groovesin the wear track of ZTACOM is smaller than that forZTA/5. Figure 7(a) shows wear track details for ZTACOM.It is found that material in these grooves is fractured andsmeared in the sliding direction. ZTA/5 (Fig. 7(b)) showsthe same features for wear inside the grooves as ZTACOM,but with a significantly smaller number of grooves. The wearscar on the alumina pin is shown in Fig. 8, revealing large“asperities” of about 10mm. These “asperities” are likely tobe caused by the microstructural inhomogeneities in the alu-mina ball.

(2) DiscussionThe grooves in the wear tracks of the ceramic disks are

caused by the asperities in the wear scars of the alumina pinbecause the sizes of the asperities are the same (≈10mm) as thegroove widths. It is found that in the surface of the materialnext to the grooves, grain pullout takes place for ZTACOM,whereas for ZTA/5 only a mild polishing/deformation effecttakes place. This difference in wear mechanism is the mainreason for the difference in the specific wear rate. The morehomogeneous microstructure of the colloidal processed ZTAcomposite leads to a significant improvement in wear resis-tance compared to that of a commercially available, dry-pressed material.

V. Conclusion

Colloidal processing of zirconia-toughened alumina resultsin defect-poor, dense composites. A homogeneous compositeceramic could be obtained by colloidal filtration of an aqueoussuspension stabilized using 0.05M HNO3. The suspension con-tained a mixture of fine-grained uniformly sized alumina (66wt% solid in single-phase suspension) and zirconia (33 wt%solid in single-phase suspension). The ceramic, formed by col-loidal processing and sintering, showed a more homogeneousphase distribution and also smaller average grain sizes than aceramic processed and sintered from commercial material,used as comparison. The colloidal processed material sinteredat 1400°C had grain sizes as low as 0.5mm for alumina and 0.2mm for zirconia. During sintering, mutual grain-growth inhi-bition of alumina and zirconia grains in the colloidal processedceramic is caused by a uniform distribution of the particles inthe green compact.

The ZTA ceramics investigated show mild wear under dry-sliding testing conditions (initial contact pressure 1130 MPa,

Table II. Microstructural Characteristics of ZTA/5 after Sintering at 1350°, 1400°,and 1450°C

SinteringT(°C)

Density(%)

Grain size (and phase)of Al2O3 (mm)

Grain size (and phase)of ZrO2 (mm)

1350 91 —† (corundum) —†

1400 >99 0.51 (corundum) 0.16 (tetragonal and monoclinic)1450 >99 0.7 (corundum) 0.3 (tetragonal and monoclinic)†Not measured.

Fig. 5. Microstructure of ZTACOM.

Fig. 6. (a) Wear track (overview) of ZTACOM. (b) Wear track(overview) of ZTA/5.

August 1999 Processing of Homogeneous Zirconia-Toughened Alumina Ceramics 2091

velocity 0.5 m/s). The specific wear rate is 5 × 10−8 mm3/(N?m)for the colloidal processed material and 2 × 10−6 mm3/(N?m)for the reference materials made by dry pressing. The differ-ences in wear rates are caused by grain pullout in specificregions of the reference material whereas the colloidal-processed material shows only polishing and deformation inthese regions. It can therefore be concluded that the improvedmicrostructure of the colloidal processed material leads to animproved wear resistance compared to the more coarse-structured and irregular reference material.

Acknowledgments: A. F. Stassen is acknowledged for his contributionto materials processing, and M. A. Smithers for the SEM micrographs. P. M.Biesheuvel is acknowledged for his contribution to describing the processesduring pressure filtration, and H. Metselaar and D. J. Schipper for their usefulcomments on the wear mechanisms.

References1K. Adachi, K. Kato, and N. Chen, “Wear Maps of Ceramics,”Wear, 203–

204, 291–301 (1997).2H. Czichos, D. Klaffke, E. Santner, and M. Woydt, “Advances in Tribology:

The Materials Point of View,”Wear, 190, 155–61 (1995).3C. He, Y. S. Wang, J. S. Wallace, and S. M. Hsu, “Effect of Microstructure

on the Wear Transition of Zirconia-Toughened Alumina,”Wear, 162–164, 314–21 (1993).

4J. D. Sibold, “Wear Applications”; pp. 973–77 inEngineered MaterialsHandbook, Vol. 4, Ceramics and Glasses. Edited by S. J. Schneider. ASMInternational Handbook Committee, 1991.

5S. Sasaki, “Effects of Environment on Friction and Wear of Ceramics,”Bull.Mech. Eng. Lab. Jpn., 58, 1–109 (1992).

6Y. J. He, A. J. A. Winnubst, A. J. Burggraaf, and H. Verweij, “Grain-SizeDependence of Sliding Wear in Tetragonal Zirconia Polycrystals,”J. Am. Ce-ram. Soc., 79 [12] 3090–96 (1996).

7S. M. Hsu and M. C. Shen, “Ceramic Wear Maps,”Wear, 200, 154–75(1996).

8Y. J. He, A. J. A. Winnubst, D. J. Schipper, P. M. V. Bakker, A. J. Burg-graaf, and H. Verweij, “Friction and Wear Behaviour of Ceramic-Hardened Steel Couples under Reciprocating Sliding Motion,”Wear, 184, 33–44 (1995).

9K. J. Konsztowicz and R. Langlois, “Effects of Heteroflocculation of Pow-ders on Mechanical Properties of Zirconia–Alumina Composites,”J. Mater.Sci., 31, 1633–41 (1996).

10R. Trabelsi, D. Treheux, G. Orange, G. Fantozzi, P. Homerin, and F. Thev-enot, “Relationship between Mechanical Properties and Wear Resistanceof Alumina–Zirconia Ceramic Composites,”Tribol. Trans., 32 [1] 77–84(1989).

11F. F. Lange and M. M. Hirlinger, “Hindrance of Grain Growth in Al2O3 byZrO2 Inclusions,”J. Am. Ceram. Soc., 67 [3] 164–68 (1984).

12K. Cherif, B. Gueroult, and M. Rigaud, “Wear Behaviour of AluminaToughened Zirconia Materials,”Wear, 199, 113–21 (1996).

13A. Ravikiran, G. R. Subbanna, and B. N. Pramila Bai, “Effect of InterfaceLayers Formed during Dry Sliding of Zirconia-Toughened Alumina (ZTA) andMonolithic Alumina against Steel,”Wear, 192, 56–65 (1996).

14Y. J. He, A. J. A. Winnubst, D. J. Schipper, A. J. Burggraaf, and H. Ver-weij, “Effects of a Second Phase on the Tribological Properties of Al2O3 andZrO2 Ceramics,”Wear, 210, 178–87 (1997).

15H. Kamiya, M. Sakakibara, Y. Sakuri, G. Jimbo, and S. Wada, “ErosionWear Properties of Tetragonal ZrO2(Y2O3)-Toughened Al2O3 Composites,”J.Am. Ceram. Soc., 77 [3] 666–72 (1994).

16E. M. DeLiso, W. R. Cannon, and A. S. Rao, “Dispersion of Alumina–Zirconia Powder Suspensions”; pp. 335–41 in Advances in Ceramics, Vol. 24,Science and Technology of Zirconia III.Edited by S. Somiya, N. Yamamoto,and H. Yanagida. American Ceramic Society, Westerville, OH, 1988.

17B. V. Velamakanni and F. F. Lange, “Effect of Interparticle Potentials andSedimentation on Particle Packing Density of Bimodal Particle Distributionsduring Pressure Filtration,”J. Am. Ceram. Soc., 74 [1] 166–72 (1991).

18A. Bleier, P. F. Becher, K. B. Alexander, and C. G. Westmoreland, “Effectsof Aqueous Processing Conditions on the Microstructure and TransformationBehaviour in Al2O3–ZrO2(CeO2) Composites,”J. Am. Ceram. Soc., 75 [10]2649–58 (1992).

19J. Cesarano and I. A. Aksay, “Processing of Highly Concentrated Aqueous

Fig. 7. (a) Details of wear track of ZTACOM. (b) Details of wear track of ZTA/5.

Fig. 8. Wear scar of alumina pin.

2092 Journal of the American Ceramic Society—Kerkwijk et al. Vol. 82, No. 8

a-Alumina Suspensions Stabilised with Polyelectrolytes,”J. Am. Ceram. Soc.,71 [12] 1062–67 (1988).

20A. Roosen and H. K. Bowen, “Influence of Various Consolidation Tech-niques on the Green Microstructure and Sintering Behaviour of Alumina Pow-ders,”J. Am. Ceram. Soc., 71 [11] 970–77 (1988).

21F. F. Lange and K. T. Miller, “Pressure Filtration: Consolidation Kineticsand Mechanics,”Am. Ceram. Soc. Bull., 66 [10] 1498–504 (1987).

22A. Bleier and C. G. Westmoreland, “Effects of pH and Particle Size on theProcessing of and the Development of Microstructure in Alumina–ZirconiaComposites,”J. Am. Ceram. Soc., 74 [12] 3100–11 (1991).

23T. Kimura, Y. Kaneko, and T. Yamaguchi, “Consolidation of Alumina–Zirconia Mixtures by a Colloidal Process,”J. Am. Ceram. Soc., 74 [3] 625–32(1991).

24W. F. M. Groot Zevert, A. J. A. Winnubst, G. S. A. M. Theunissen, andA. J. Burggraaf, “Powder Preparation and Compaction Behaviour of Fine-Grained YTZP,”J. Mater. Sci., 25, 3449–55 (1990).

25J. C. Wurst and J. A. Nelson, “Lineal Intercept Technique for MeasuringGrain Size in Two-Phase Polycrystalline Ceramics,”J. Am. Ceram. Soc., 55 [2]109 (1972).

26A. P. Philipse, B. C. Bonekamp, and H. J. Veringa, “Colloidal Filtrationand (Simultaneous) Sedimentation of Alumina and Silica Suspensions: Influ-ence of Aggregates,”J. Am. Ceram. Soc., 73 [9] 2720–27 (1990).

27F. M. Tiller, N. B. Hsyung, and D. Z. Cong, “Role of Porosity in Filtration:XII. Filtration with Sedimentation,”AIChE J., 41 [5] 1153–64 (1995).

28F. Concha and E. R. Almendra, “Settling Velocities of Particulate Systems,2. Settling Velocities of Suspensions of Spherical Particles,”Int. J. Miner.Proc., 6, 31–41 (1979).

29P. M. Biescheuvel, A. Nijmeijer, and H. Verweij, “A Theory of BatchwiseCentrifugal Casting,”AIChE J., 44, 1914–22 (1998).

30J. F. Richardson and F. A. Shabi, “The Determination of ConcentrationDistribution in a Sedimenting Suspension Using Radioactive Solids,”Trans.Inst. Chem. Eng., 38, 33–42 (1960).

31R. A. Williams, W. B. K. Amarasinghe, S. J. R. Simons, and C. G. Xie,“Sedimentation Behaviour of Complex Polydisperse Suspensions,”PowderTechnol., 65, 411–32 (1991).

32M. J. Lockett and H. M. Al-Habbooby, “Differential Settling by Sizeof Two Particle Species in a Liquid,”Trans. Inst. Chem. Eng., 51, 281–92(1973). h

August 1999 Processing of Homogeneous Zirconia-Toughened Alumina Ceramics 2093