A Morales-Rodriguez et al /Journal of the European Ceramic Sociery 29(2009)1625-1630 1 mm Fig. 1. SEM micrographs of Al] O3/ZTA multilayer composites (a) Low magnification(dark layers are Al2O3, bright layers are ZTA).(b) Detail of an interface between the Al2O3 layer and the ZTA layer; the latter one is formed by zirconia grains(bright phase)and alumina grains(dark phase) at 80C for 30 min at 30 MPa. The structure was designed to of ZTA (bright phase)215-Am thick; the corresponding volume have alumina layers as outer surfaces. The laminated samples fractions are fAl=0.40 and zTA =0.60. The two types of layers were finally sintered at 1550C for I h in air. After sintering, are well defined with very straight interfaces(Fig. 1(a)and(b)) the density of the samples, measured by Archimedes'method, Alumina grains in the Al2O3 layers have an average grain size d was close to the theoretical density. (taken as the spatial grain size)of 1.2 um. The ZTA layer shows For mechanical testing, rectangular specimens of about the typical equiaxed duplex microstructure of such composites 4 mm x 2 mm x 2 mm were cut from the laminated compacts (Fig. 1(b)), formed by alumina grains(dark phase)of d=0.6 um with the largest dimension(the loading axis) parallel to the layer and slightly smaller zirconia grains of d=0. 4 um. The smaller interfaces. In such configuration, isostrain and thus isostrain-rate size of the Al2O3 grains in the ZTA layer, in comparison to those conditions will apply through the deformation process. Com- in the Al2O3 layer, is due to the presence of the second dis- pressive tests were carried out at constant cross-head speeds persed phase, which produces a fine-grained microstructure that between 5 and 50 um/min(corresponding to initial strain rates is remarkably resistant to coarsening at elevated temperatures strain rates Eo between 2 x 10-and 2 x 10-4s-)and at con- This enhanced microstructural stability of ZTA com stant load(5-120MPa)in air at temperatures between 1400 and been widely exploited to achieve superplasticity 12-1]osites has 1500oC. The recorded data at constant strain rate. load F ver Fig 2 displays ao-Ecurve at 1500C showing several strain sus time t, and at constant load, instantaneous specimen length rates changes which allow the determination of the stress expo l(t) versus time t, were plotted, respectively, as a-E and log nent by using Eq(1). Reasonable steady-state stresses were E-Ecurves, where o= F/So(with So the initial cross-sectional achieved at every rate change, except at the higher strain rate area of the sample)is the nominal stress, E=-In((t)/o(with imposed where the specimen started to fail. At lower temper- lo the initial length of the sample) is the true strain and e the atures, the composites failed at correspondingly lower strain strain rate. Mechanical data were analyzed using the standard high-temperature power law for steady-state deformation l20 E= Ag"d Q where A is a parameter depending on the deformation mecha nism, n the stress exponent, p the grain size exponent, Q the activation energy for flow and RT has the usual meaning The microstructural characterization of the as-received and deformed laminates was carried out using scanning electron 3F0/46×10510x1042.5x104 microscopy(Microscopy Service, University of Sevilla, Spain) To this end, perpendicular sections to the layer interfaces were 40A290 cut from the samples, mechanically polished and then thermally etched at 1350C for 30 min in air to reveal grain boundaries The morphological parameters of the various phases were char- T=1500° acterized by using a semiautomatic image analyzer. 3. Experimental result I(a) shows a low magnification SEM micrograph of (points). Several determinations of the stress exponent n by the as-received laminated composite cross-section, consisting of strain rat re shown. The curve of high-purity monolithic alumina at seven layers of alumina(dark phase)125-um thick and six layers the initia1626 A. Morales-Rodríguez et al. / Journal of the European Ceramic Society 29 (2009) 1625–1630 Fig. 1. SEM micrographs of Al2O3/ZTA multilayer composites. (a) Low magnification (dark layers are Al2O3, bright layers are ZTA). (b) Detail of an interface between the Al2O3 layer and the ZTA layer; the latter one is formed by zirconia grains (bright phase) and alumina grains (dark phase). at 80 ◦C for 30 min at 30 MPa. The structure was designed to have alumina layers as outer surfaces. The laminated samples were finally sintered at 1550 ◦C for 1 h in air. After sintering, the density of the samples, measured by Archimedes’ method, was close to the theoretical density. For mechanical testing, rectangular specimens of about 4 mm × 2 mm × 2 mm were cut from the laminated compacts with the largest dimension (the loading axis) parallel to the layer interfaces. In such configuration, isostrain and thus isostrain-rate conditions will apply through the deformation process. Com￾pressive tests were carried out at constant cross-head speeds between 5 and 50m/min (corresponding to initial strain rates strain rates ε˙o between 2 × 10−5 and 2 × 10−4 s−1) and at con￾stant load (5–120 MPa) in air at temperatures between 1400 and 1500 ◦C. The recorded data at constant strain rate, load F ver￾sus time t, and at constant load, instantaneous specimen length l(t) versus time t, were plotted, respectively, as σ − ε and log ε˙ − ε curves, where σ = F/S0 (with S0 the initial cross-sectional area of the sample) is the nominal stress, ε = −ln [l(t)/l0] (with l0 the initial length of the sample) is the true strain and ε˙ the strain rate. Mechanical data were analyzed using the standard high-temperature power law for steady-state deformation: ε˙ = Aσnd−p exp − Q RT (1) where A is a parameter depending on the deformation mecha￾nism, n the stress exponent, p the grain size exponent, Q the activation energy for flow and RT has the usual meaning. The microstructural characterization of the as-received and deformed laminates was carried out using scanning electron microscopy (Microscopy Service, University of Sevilla, Spain). To this end, perpendicular sections to the layer interfaces were cut from the samples, mechanically polished and then thermally etched at 1350 ◦C for 30 min in air to reveal grain boundaries. The morphological parameters of the various phases were char￾acterized by using a semiautomatic image analyzer. 3. Experimental results Fig. 1(a) shows a low magnification SEM micrograph of the as-received laminated composite cross-section, consisting of seven layers of alumina (dark phase) 125-m thick and six layers of ZTA (bright phase) 215-m thick; the corresponding volume fractions are fAl = 0.40 and fZTA = 0.60. The two types of layers are well defined with very straight interfaces (Fig. 1(a) and (b)). Alumina grains in the Al2O3 layers have an average grain size d (taken as the spatial grain size) of 1.2 m. The ZTA layer shows the typical equiaxed duplex microstructure of such composites (Fig. 1(b)), formed by alumina grains (dark phase) of d = 0.6m and slightly smaller zirconia grains of d = 0.4m. The smaller size of the Al2O3 grains in the ZTA layer, in comparison to those in the Al2O3 layer, is due to the presence of the second dis￾persed phase, which produces a fine-grained microstructure that is remarkably resistant to coarsening at elevated temperatures.11 This enhanced microstructural stability of ZTA composites has been widely exploited to achieve superplasticity.12–14 Fig. 2 displays a σ − ε curve at 1500 ◦C showing several strain rates changes which allow the determination of the stress expo￾nent by using Eq. (1). Reasonable steady-state stresses were achieved at every rate change, except at the higher strain rate imposed where the specimen started to fail. At lower temper￾atures, the composites failed at correspondingly lower strain Fig. 2. Stress–strain curve at 1500 ◦C of laminated composite deformed under isostrain condition (points). Several determinations of the stress exponent n by strain rate changes are shown. The curve of high-purity monolithic alumina at the initial strain rate of ε˙o = 2 × 10−5 s−1 is also shown.