Plastically deforming amorphous ZrO2-Al2O3

Acta Materialia 51 (2003) 1641–1649 www.actamat-journals.com Plastically deforming amorphous ZrO2-Al2O3 A.S. Gandhi ∗,1, V. Jayaram Department of Metallurgy, Indian Institute of Science, Bangalore 560 012, India Received 19 March 2002; received in revised form 25 November 2002; accepted 28 November 2002 Abstract We report for the first time non-viscous, plastic deformation in an amorphous oxide: ZrO2-Al2O3. Dense samples of amorphous ZrO2-Al2O3 made by hot pressing spray-pyrolysed powder were deformed in uniaxial compression at 600– 700°C at strain rates from 6 × 10⫺5 s⫺1 to 10⫺3 s⫺1. A transition from elastic to plastic deformation occurred at a critical stress ~360 MPa. The onset of plastic deformation was associated with a drop in the stress by 20–25%. Little influence of strain and strain rate on the flow stress was observed. The non-viscous, plastic deformation is related to the open structure of the amorphous phase as indicated by its low true density.  2003 Acta Materialia Inc. Published by Elsevier Science Ltd. All rights reserved. Keywords: Amorphous oxides; Plastic deformation; Shear bands; Compression test 1. Introduction It is well known that glassy or amorphous ceramic materials, i.e. silicate glasses, exhibit homogeneous viscous flow at temperatures above the glass transition temperature, Tg. At room temperature, the oxide glasses are brittle. However, room temperature indentation has been the subject of many investigations undertaken to determine if the deformation is by densification or shear flow. The existence of shear flow has been established by the analysis of spiral flow lines beneath the indentations in soda-lime glasses and silica glass [1–3]. ∗ Corresponding author. Tel.: +1-805-893-8390; fax: +1805-893-8486. E-mail address: [email protected] (A.S. Gandhi). 1 Present address: Materials Department, University of California, Santa Barbara, CA 93106, USA Localised shear flow has also been reported in amorphous silica in uniaxial compressive deformation at temperatures from 950°C to the Tg, 1200°C [4,5]. Although localised shear deformation was seen along planes of maximum shear stress, the macroscopic deformation behaviour below the Tg is non-linear viscous. Refractory crystalline ceramics such as Al2O3, ZrO2, SiC, are hard and brittle at low temperatures and ambient pressure. What little plasticity is exhibited by such strongly bonded ionic or covalent compounds is generally found at high temperatures or under superimposed hydrostatic compression, such as exists deep below the earth’s crust or in special testing apparatus. Single crystals of αAl2O3 deform by dislocation motion at temperatures as low as 200°C, but only in the presence of a constraining pressure ~1.5 GPa [6] and with an applied uniaxial stress ~7.5 GPa. This latter value drops to ~2.5 GPa at 800°C, but in the absence of 1359-6454/03/$30.00  2003 Acta Materialia Inc. Published by Elsevier Science Ltd. All rights reserved. doi:10.1016/S1359-6454(02)00566-9 1642 A.S. Gandhi, V. Jayaram / Acta Materialia 51 (2003) 1641–1649 the constraining pressure, no plasticity is seen until temperatures of ~1400°C [7]. The fracture toughness is so low that brittle fracture supervenes at stresses much lower than those required for initiating plastic deformation, particularly in polycrystalline materials in which grain boundaries serve as further impediments to dislocation motion. At very high temperatures a significant component of deformation is of the time-dependent type and involves diffusion aided mechanisms such as Coble creep or grain boundary sliding [8]. The microstructure of a polycrystalline solid therefore influences the deformation behaviour perhaps more than it influences certain other properties. Conventional oxide glasses, such as those based on silicates, are also brittle at room temperature. No microstructure is present in single phase glassy or amorphous solids in which the mechanical properties under ambient pressures clearly separate into two domains: brittle behaviour at low temperatures and viscous deformation in the vicinity of the glass transition temperature (Tg). It is therefore of great interest to report the unusual behaviour of amorphous ZrO2-Al2O3 in which large, plastic (nonviscous) deformation is found at temperatures as low as 600°C, and by extrapolation from indirect experiments, probably even at 400°C. The deformation occurs in the absence of any hydrostatic constraint and at low stresses that are comparable to those encountered in a few strong metallic alloys. The amorphous oxide phases are metastable below the melting temperature [9,10]. They are made by several routes including rapid solidification processing (RSP) [11,12], mechanical alloying [13], sol-gel synthesis [14], and precursor spray pyrolysis (SP) [15]. Sol-gel and spray pyrolysis are more suitable to oxide synthesis. Owing to the need to prevent the formation of the equilibrium crystalline phases, high rates of heat and mass transport are generally involved in such processes, leading to powders or ribbons with large ratios of the surface area to the volume. What has posed a severe challenge is the ability to densify the powders or ribbons into bulk pieces without losing the metastable amorphous phase, to enable subsequent property measurement. Recently we have shown [16] that amorphous ZrO2-Al2O3 powders may be consolidated at moderately high pressures (500– 750 MPa) and low temperatures (450–650°C). The amorphous phase is retained after densification and relative densities up to 99% have been obtained in compositions with 40, 62 and 80 mol% Al2O3. The pressure and time dependence of densification of amorphous ZrO2-Al2O3 made it unlikely that conventional models that are applied to viscous glasses [17], or sintering of crystalline ceramics or metals, could hold in the present case. For instance, the relative density of the amorphous ZrO2-40% Al2O3 powder at 600°C increased from 43% to 90% as the pressure was increased up to 750 MPa (Fig. 1), whereas it increased further only up to 95% when the pressure of 750 MPa was held constant for 1 h. The time dependent densification was small at all pressures (250, 500 and 750 MPa) and relative densities. This behaviour is in marked contrast with that of a borosilicate glass powder which Fig. 1. Amorphous ZrO2-40% Al2O3 powder was first compacted to a density of 42% at room temperature at a pressure of 50 MPa. It was then heated to 600°C and hot pressed at 750 MPa. The graph shows that the relative density during hot pressing increased rapidly as the pressure was continuously increased up to 750 MPa. However, as the pressure of 750 MPa was held constant, only 4% increase in the relative density was recorded after 5 minutes and only ~1% further increase in the next 55 min. A.S. Gandhi, V. Jayaram / Acta Materialia 51 (2003) 1641–1649 registered large increments in relative density as a function of time in the vicinity of Tg (500–550°C) as the pressure was held constant [18]. Moreover, a model of cold compaction of ductile metal powders [19] was in excellent agreement with the hotpressing data on ZrO2-Al2O3. This report describes the tests conducted to unequivocally demonstrate for the first time the plastic deformation that takes place in these precursor derived amorphous ZrO2Al2O3 materials. 1643 to a displacement transducer. Stress relaxation up to 30 min was also conducted after loading up to a certain compressive stress and then keeping the cross-head stationary. Scanning electron microscopy (SEM, JEOL JSM 840) was performed on a compression sample before and after deformation, with one face polished to ceramographic finish and coated with a gold-palladium alloy. 3. Results and discussion 2. Experimental details Amorphous powder with the composition ZrO240mol% Al2O3 were prepared by spray pyrolysis of an aqueous solution of the metal nitrates on a Teflon coated aluminium pan at 250°C, followed by further thermal decomposition of the powder into an amorphous oxide at 750°C for 1 h. Particles coarser than 15 µm were removed by sedimentation in water and uniaxially hot pressed in a nickel based superalloy die of 5 mm internal diameter. Dense amorphous discs of ZrO2-40% Al2O3 with 5–10% porosity (90–95% relative density) were obtained by hot pressing the powder at 600°C under a pressure of 750 MPa for 15 min. The flat faces of the discs were ground to 400 grit finish. The discs were then cut into samples with approximate cross section of 2 mm × 2 mm and height 1.8 mm such that the compression axis was parallel to the hot pressing axis. It is recognised that the ratio of length to width (1 or less) was smaller than that prescribed by ASTM for compression testing. A small aspect ratio leads to the overestimation of the flow stress of a ductile material because a large fraction of the sample is constrained by the friction between the sample and the platens. However, the purpose of the present deformation experiments, which was to first establish the nature of the stressstrain response of the amorphous ZrO2-Al2O3 material, was amply served by the sample geometry mentioned above. Al2O3 platens were used, without any lubricant between the sample and platen surface. Deformation was carried out at temperatures up to 700°C with constant engineering strain rates. Displacement was monitored with a temperature compensated extensometer attached Fig. 2 shows typical porosity distribution in the hot pressed amorphous ZrO2-40% Al2O3. The hypothesis that plastic deformation occurs in this material at elevated temperatures was confirmed by the results of uniaxial compression experiments. At 700°C, it was first established that small stresses only caused elastic deformation. At a critical value of the stress (yield stress, Fig. 3a), permanent deformation was initiated accompanied by a steep drop in the stress (yield drop). Further plastic deformation occurred at an essentially constant or slowly decreasing stress (the flow stress). Reloading a specimen immediately upon unloading led to the resumption of deformation at about the same stress, and no sharp yield drop was seen (Fig. 3b). Deformation was carried out up to a maximum of 14% strain and since the sample remained unbroken, there is reason to believe that the Fig. 2. Representative optical micrograph showing the porosity distribution in hot pressed amorphous ZrO2-40% Al2O3. This sample is 96% dense. 1644 A.S. Gandhi, V. Jayaram / Acta Materialia 51 (2003) 1641–1649 maximum realisable compressive strain may be higher. In a compression experiment at 700°C at a lower strain rate of 10⫺4 s⫺1 (Fig. 4) the yield stress was lower than at the higher strain rates shown in Fig. 3. However, the scatter in the data due to the variation of porosity from one sample to the next makes this observation unreliable. The same scatter masked the influence of temperature on the yield stress. Tests were performed at strain rates in the range from 6 × 10⫺5 s⫺1 to 10⫺3 s⫺1 at temperatures of 600 and 650°C. However, consecutive tests carried out at increasing strain rates on the same sample showed little influence of the rate of deformation on the flow stress (Fig. 5). The results described above have verified the existence of plastic deformation in amorphous ZrO2-Al2O3 at elevated temperatures. In order to detect any microscopic features of the deformation, a compression test was conducted on a sample with one of its sides polished with 3 µm diamond suspension. The test was conducted at 700°C with a strain rate of 3 × 10⫺4 s⫺1. An SEM micrograph of the polished surface of the undeformed sample is shown in Fig. 6a and that of the deformed sample is shown in Fig. 6b. The undeformed sample shows some surface relief contrast owing to difficulty in obtaining good finish in such a small Fig. 3. (a) Compression test at 700°C on a sample of amorphous ZrO2-40mol% Al2O3 at the indicated strain rates shows elastic deformation followed by plastic deformation up to a large plastic strain of 8.5%. Note the drop in the stress upon the onset of plastic deformation. (b) A sample was reloaded immediately after unloading. No sharp yield point is seen upon reloading. The temperature was 700°C. Fig. 4. A compression test on amorphous ZrO2-40% Al2O3 at a lower strain rate of 10⫺4 s⫺1 than shown in Fig. 3. A.S. Gandhi, V. Jayaram / Acta Materialia 51 (2003) 1641–1649 1645 Fig. 6. (a) SEM micrograph of the polished face of an undeformed sample. (b) The polished face after compressive deformation at 700°C shows shear bands. The sample was tilted to improve topographical contrast, by 30° around an axis approximately top to bottom. The compression axis is also from top to bottom. Such shear bands are not observed in unconstrained deformation of conventional glassy oxides. Fig. 5. Samples were deformed in compression at (a) 600°C and (b) 650°C by loading and unloading at successively higher strain rates as indicated. The lower yield stress was not sensitive to the strain rate. sample. The presence of shear bands is clearly seen in the deformed sample. The bands were seen throughout the length of the specimen. The inclination of these traces to the compression axis was measured from other micrographs to be ~52°. The true inclination of the slip planes can only be determined by measurements on two orthogonal faces. Such shear band formation in an amorphous oxide material has not been reported earlier. Amorphous materials may relax upon annealing. Therefore, a sample already deformed at 600°C was immediately reloaded till plastic deformation resumed and the cross-head was stopped. Fig. 7a shows the stress relaxation over 5 min. The magnitude of the stress relaxation was quite small. The stress reduced by 7% and then remained constant up to the end of the test (30 min). Interestingly, a sample already deformed at 700°C and cooled was reheated to the same temperature and was deformed at 5 × 10⫺4 s⫺1 after 75 min. The load against the strain is plotted in the Fig. 7b. (The 1646 A.S. Gandhi, V. Jayaram / Acta Materialia 51 (2003) 1641–1649 Fig. 7. (a) Stress relaxation in an amorphous ZrO2-40mol% Al2O3 sample was measured after deforming at 600°C. The stress decreased by ~7% in 5 min and then remained constant. No change in the strain was recorded. A conventional oxide glass would relax the stress much more, the relaxation kinetics being determined by its viscosity. (b) A sample deformed at 700°C was reheated to the same temperature for 75 min before another compressive loading. The drop in the stress at the onset of plastic deformation reappeared, indicating that structural relaxation occurring during heating can restore the original sample properties. Note that while not much of stress relaxation occurred in the amorphous ZrO2-Al2O3, annealing in the absence of an applied stress led to partial recovery of the original structure. cross section area could not be determined since all the deformed samples are fragile at room temperature). It is seen that the yield drop was recovered, with the load dropping from 450 to 420 N (6.7% decrease). This result implies that with longer annealing in the absence of an applied stress, the deformed material would relax to its original strength corresponding to the yield stress. A powder X-ray diffraction pattern taken from another sample deformed at 600°C confirmed that it remained amorphous. These observations clearly establish a unique deformation behaviour not seen in an amorphous oxide material so far. The contrast with conventional oxide glasses is remarkable. As discussed below, the deformation characteristics of metallic and polymeric glasses bear some resemblance to that of ZrO2-Al2O3. Metallic and polymeric glasses which exhibit glass transition deform in a viscous manner at temperatures near the Tg and above, whereas they deform plastically at temperatures lower than Tg [20,21]. The elastic–plastic transition occurs at a particular stress and the strain rate effect is far less pronounced than in viscous flow. Homogeneous flow occurs in a polymeric glass at temperatures higher than the glass transition temperature (Tg). Strain hardening may be seen at large strains in a tensile test when the polymeric chains become aligned to the tensile axis. At temperatures lower than the Tg, the permanent deformation becomes localised in the form of shear bands. At temperatures lower than that of shear band formation, the polymers deform by the process of crazing. Crazes are regions of highly oriented fibrils of molecules, separated by voids. These regions appear in planes perpendicular to the direction of the maximum principal stress, i.e. perpendicular to the tensile axis in a tension test. Similarly, in a metallic glass [22–25], at temperatures higher than 0.6 Tg to 0.7 Tg homogeneous flow occurs. The flow stress is highly sensitive to the temperature and the strain rate. At lower temperatures elastic–plastic behaviour is seen, with possible non-linear elasticity before the onset of permanent deformation. The yield stress is not sensitive to temperature and strain rate as much as it is in the high temperature viscous flow range near Tg. The plastic deformation is inhomogeneous A.S. Gandhi, V. Jayaram / Acta Materialia 51 (2003) 1641–1649 and occurs by the formation of localised shear bands. The tensile elongations to failure are very small but plastic strains of a few percent are seen in compression. It is also possible to cold roll metallic glass ribbons. In some cases the yield stress decreases by ~25% after the onset of plastic deformation. The flow stress does not show strong strain hardening or softening behaviour, though small magnitudes of either may be seen depending on the temperature. The plastic deformation of metallic glasses at low temperatures was first explained on the basis of dislocation motion [26,27]. The dislocation in a glass was thought to have a varying Burgers vector along its length. The deformation behaviour of metallic glasses has also been explained in terms of the free-volume theory [23,24] advanced on the basis of the arguments of Polk and Turnbull [28] that the shear bands represent regions of destroyed short range order. It was observed by Pampillo [29] that the shear bands can be preferentially etched, implying that these are regions of high energy associated with excess free volume and low viscosity. Another theory of localised deformation and the formation of shear bands in metallic glasses is based on adiabatic increase in the local temperature [30,31]. It is argued that the vein morphology of the tensile fracture surface of a metallic glass is similar to the separation of a thin film of a fluid between two plates which are pulled apart. This implies local reduction of viscosity, brought about by adiabatic heating. Since fracture is preceded by plastic deformation in a shear band, local temperature rise is also regarded as the mechanism of shear band formation. However, in spite of extensive research, both experimental and theoretical, the mechanism of the inhomogeneous deformation of metallic glasses is not clearly understood. Under a uniaxial stress, even though the maximum shear stress is at an angle of 45° to the direction of loading, the traces of shear bands have been observed at smaller angles to the compressive axis and at angles more than 45°C in tension. Such behaviour is seen in both polymeric glasses [32] and metallic glasses [33,34]. This is due to the contribution of the normal stress to the nucleation of shear bands by enhanced local dilatation. It is suggested that the normal stress and the hydrostatic 1647 stress should be included in the yield criterion for such materials [35] analogous to the Mohr–Coulomb yield criterion for granular materials. Experimental deformation studies in the presence of hydrostatic pressure have been carried out to identify the role of these parameters on the flow and fracture of metallic glasses e.g. on Pd-Cu-Si [36] and Zr-Ti-Ni-Cu-Be glasses [37]. In the amorphous ZrO2-Al2O3 samples in the present investigation, the true inclination of the normal to the slip plane lies between 90° and 38°. If this angle were greater than 45°, it would be consistent with the Mohr– Coulomb yield criterion. Owing to the possible similarity between the structure of amorphous ZrO2-40% Al2O3 and the structure of metallic glasses, the deformation of the amorphous ZrO2-40% Al2O3 is considered more analogous to that of metallic glasses, than that of glassy polymers. Whereas polymeric glasses consist of covalently bonded long chains that are bonded together by van der Waals forces, metallic glasses have short range order of the component atoms, and silicate glasses have networks of tetrahedra modified by the alkali cations. The structural information on amorphous ZrO2-Al2O3 material comes from of 27 Al magic angle spinning nuclear magnetic resonance (MAS-NMR) spectroscopy [38,39]. Amorphous ZrO2-40% Al2O3 was shown to contain about 40% relative amount of Al3+ ions in 5-fold co-ordination with O2⫺, along with ~40% of 6-fold and ~20% of 4-fold co-ordinations. A similar co-ordination persists in the product of crystallisation, which is a tetragonal (t) ZrO2(Al2O3) solid solution. The Raman spectra [40] from the product of crystallisation also match those of t-ZrO2. It is therefore likely that the amorphous material has Zr4+ ion coordination similar to t-ZrO2. The extensive aliovalent substitution of aluminium ions may cause the formation of a large number of stoichiometric vacancies on the oxygen sites. The true density of amorphous ZrO2-40% Al2O3 is only 3.4 gcm⫺3, as measured by pycnometry [16]. This implies that the amorphous phase has an open structure. Although the density of the equilibrium microstructure of mZrO2 and α-Al2O3 is 5 gcm⫺3, the true density of a sample crystallised at 1000°C, consisting of t-ZrO2 solid solution and γ-Al2O3 phases, is 3.8 gcm⫺3. The density of the unpartitioned t-ZrO2(Al2O3) 1648 A.S. Gandhi, V. Jayaram / Acta Materialia 51 (2003) 1641–1649 solid solution may be slightly lower. Therefore, it is argued that the structure of the amorphous phase may consist of cation–anion co-ordination similar to that in the first crystalline product, with short range ordering subject to the condition of local charge neutrality. Such a structure may be compatible with a low true density of the amorphous phase. The density of liquid Al2O3 at the melting point (2054°C) has been measured as 2.8 gcm⫺3 [41]. Assuming an average thermal expansion coefficient for α-Al2O3 of 10 × 10 - 6 °C⫺1, its density would be 3.72 gcm⫺3 at the melting point, implying that the density difference is ~25%, taking the solid as reference. Since such anomalous behaviour has not been reported for zirconia, it appears unlikely that such factors can completely account for the low density of amorphous ZrO2Al2O3. The reason that amorphous Al2O3 expands so much upon melting is that it changes from the 6-fold coordination in α- Al2O3 to a substantial 4fold coordination in the liquid. These considerations indicate that certain additional factors may be responsible for the low density of the amorphous ZrO2-Al2O3. The processing of an amorphous oxide by spray pyrolysis is also likely to influence its structure. Nitrate and hydroxyl radicals removed during thermal decomposition may leave larger free volume in the amorphous material than if it were to be produced by rapid solidification. The densities and the deformation or densification behaviour of rapidly solidified amorphous ZrO2-Al2O3 may therefore be different from that of the spray pyrolysed material. Such pronounced differences in the true densities of precursor derived amorphous phases and the products of their crystallisation are also found in the Si-C-B-N system in which studies of creep show a progressively decreasing creep rate as the structure gradually relaxes to higher true densities [42]. The presence of OH⫺ radicals in the amorphous ZrO2-Al2O3 samples, which may also influence the deformation behaviour, cannot be ruled out although IR spectra [43] did not detect any radicals. The yield stress of the amorphous ZrO2-Al2O3 (~360 MPa) is very low in comparison with the high temperature hardness of the two component oxides, ZrO2 and α-Al2O3. Prabhu and Bourell [44] have measured the hot hardness of yttria stabilised tetragonal zirconia with 150–200 nm grain size. Vickers hardness at 500 g load at 600°C and 700°C were 5.65 and 5.13 GPa, respectively. The uniaxial yield stress (one third of the hardness) then comes to about 1.7–1.9 GPa, which is more than four times that of the amorphous ZrO2-Al2O3. As pointed out at the beginning, α-Al2O3 single crystals deform by prismatic slip, which is considered easier than basal slip, at a normal stress of ~1.3 GPa at 700°C, when the applied hydrostatic stress was 1.5 GPa. It is clear that the amorphous ZrO2Al2O3 has not only a much smaller yield stress than the corresponding crystalline phases but also a much higher toughness in its ability to dissipate energy through plastic deformation as large as 14% in the absence of hydrostatic pressure. Similarly, it has been reported that crystallisation in metallic glasses considerably degrades their ductility and toughness [45] suggesting that common structural reasons may underlie the phenomenon in both classes of materials. Another phenomenon that results from localised deformation in some metallic glasses is nanocrystallisation within the shear bands [46]. Such a process is unlikely in the present instance where the material has been subjected to far more severe deformation without crystallisation during densification. 4. Concluding remarks A new type of deformation behaviour has been reported for the first time in an amorphous oxide, namely ZrO2-40% Al2O3. Compression tests at 600–700°C revealed elastic–plastic deformation. The onset of plasticity is associated with a sharp yield drop and the formation of localised shear bands throughout the sample. No measurable strain hardening or softening occurs. The average value of the yield stress is ~360 MPa whereas the lower, plateau flow stress is ~240 MPa. No effect of strain rate on the flow behaviour was observed in the range 6 × 10⫺5 to 10⫺3 s⫺1. If one relates the distances between coordination polyhedra (from density) to the elastic modulus and the yield stress, then the low density of the amorphous phase (3.4 gcm⫺3) may be responsible for its ability to yield at low stresses. Although the mechanism of defor- A.S. Gandhi, V. Jayaram / Acta Materialia 51 (2003) 1641–1649 mation is not known, it is of interest to explore other oxide systems for similar behaviour. Indeed, densification experiments on other compositions in the Al2O3-ZrO2 system (20 and 35% ZrO2) [43], and also on Al2O3-37.5% Y2O3 [47] have revealed the ability of the amorphous phases to densify. Additionally, the structural features responsible for the unusual deformation behaviour of amorphous ZrO2-Al2O3 is most likely to lead to an entire range of interesting properties, for example, electrical conductivity, thermal conductivity and optical transmittivity. Acknowledgements Financial support for this work was provided by the Department of Science and Technology, India. A.S. Gandhi was supported by the Department of Atomic Energy, India through the Dr. K.S. Krishnan Fellowship. The authors are grateful to Prof. A.H. Chokshi for his suggestions and access to experimental facilities. Gas Turbine Research Establishment, Bangalore supplied the superalloy bar-stock. References [1] Hagan JT. J Mater Sci 1979;14:462. [2] Hagan JT. J Mater Sci 1980;15:1417. [3] Kurkjian CR, Kammlott GW, Chaudhari MM. J Am Ceram Soc 1995;78:737. [4] Donnadieu PJN. J Non-Cryst Solids 1989;111:7. [5] Donnadieu P, Jaoul O, Kléman M. Phil Mag 1985;52:5. [6] Castaing J, Cadoz J, Kirby SH. J Am Ceram Soc 1981;64:504. [7] Cadoz J, Castaing J, Philibert J. Revue de Physique Appliquee 1981;16:135. [8] Cannon RM. In: Kingery WD, editor. Advances in Ceramics Vol. 10, Structure and Properties of MgO and Al2O3 Ceramics. Columbus, OH: The American Ceramic Society, Inc; 1984. p. 818–38. [9] Turnbull D. Metall Trans 1981;12A:695. [10] Levi CG. Acta mater 1998;46:787. [11] Jacobson LA, McKittrick J. Mater Sci Eng R 1994;R11:355. [12] Brockway MC, Wills RR. Rapid solidification of ceramics—a technology assessment, Metals and Ceramics Information Center Report, MCIC 84-49, Battelle, Columbus, OH, USA (1984). [13] Murty BS, Ranganathan S. Intl Mater Rev 1998;43:101. 1649 [14] Brinker CJ, Scherer GW. Sol-Gel Science—The Physics and Chemistry of Sol-Gel Processing. Boston, MA, USA: Academic, 1990. [15] Messing GL, Zhang S-C, Jayanthi GV. J Am Ceram Soc 1993;76:2707. [16] Gandhi AS, Jayaram V, Chokshi AH. J Am Ceram Soc 1999;82:2613. [17] Gandhi AS, Jayaram V, Chokshi AH. Mater Sci Eng 2001;A304-306:785. [18] Gandhi AS, Jayaram V. Acta Mater 2002;50:2137. [19] Akisanya AR, Cocks ACF, Fleck NA. Int J Mech Sci 1997;39:1315. [20] Courtney TH. Mechanical behavior of materials. Singapore: McGraw-Hill Publishing Co, 1990 p, 325. [21] Meyers MA, Chawla KK. Mechanical behavior of materials. Upper Saddle River, NJ, USA: Prentice-Hall Inc, 1999 p.138. [22] Pampillo CA. J Mater Sci 1975;10:1194. [23] Spaepen F. Acta metall 1977;25:407. [24] Argon AS. Acta metall 1979;27:47. [25] Srolovitz D, Vitek V, Egami T. Acta metall 1983;31:335. [26] Gilman JJ. Dislocation dynamics. In: Rosenfield AR, Hahn GT, Bement Jr AR, Jaffee RI, editors. New York, NY, USA: McGraw-Hill Book Inc; 1968. p. 3. [27] Gilman JJ. J Appl Phy 1975;46:1625. [28] Polk DE, Turnbull D. Acta metall 1972;20:493. [29] Pampillo CA. Scripta metall 1972;6:915. [30] Chen HS. Scripta metall 1973;7:931. [31] Flores KM, Dauskardt RH. J Mater Res 1999;14:638. [32] Kramer EJ. Polymer Sci 1975;13:509. [33] Donovan PE. Mater Sci Eng 1988;98:487. [34] Wright WJ, Saha R, Nix WD. Japan Inst Metals Mater Trans 2001;42:642. [35] Li JCM, Wu JBC. J Mater Sci 1976;11:445. [36] Davis LA, Kavesh S. J Mater Sci 1976;10:453. [37] Lowhaphandu P, Montgomery SL, Lewandowski JJ. Scripta mater 1999;41:19. [38] Mishra RS, Jayaram V, Majumdar B, Lesher C, Mukherjee AK. J Mater Res 1999;14:834. [39] Balmer ML, Eckert H, Das N, Lange FF. J Am Ceram Soc 1996;79:321. [40] Balmer ML, Lange FF, Levi CG. J Am Ceram Soc 1994;77:2069. [41] Landron C, Hennet L, Jenkins TE, Greaves GN, Coutures JP, Soper AK. Phys Rev Lett 2001;86:4839. [42] Riedel R, Ruswisch L, An L, Raj R. J Am Ceram Soc 1998;81:3341. [43] Gandhi AS. Ph.D. Thesis 2001; Indian Institute of Science, Bangalore. [44] Prabhu GB, Bourell DL. Scripta metall mater 1995;33:761. [45] Xing LQ, Herlach DM, Cornet M, Bertrand C, Dallas JP, Trichet MF, Chevalier JP. Mater Sci Eng 1997;A226228:874. [46] Chen H, He Y, Shiflet GJ, Poon SJ. Nature 1994;367:541. [47] Choudhury S, Gandhi AS, Jayaram V. J Am Ceram Soc 2002; in press.