JAm. Ceram.Soc,84]7677402001) journal Toughened Oxide Composites Based on Porous Alumina-Platelet Interphases Sang. Jin Lee*T and Waltraud M. Kriven** Department of Materials Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801 A novel mechanism for debonding at a weak interphase in an More recently, the formation of transformation-weakened inter- all-oxide composite is introduced. This methodology involves phases that result from a negative volume change in the crystal the use of alumina platelets that have a diameter of 10-15 or structures has been demonstrated to be a viable interphase debond 5-10 um and a thickness of 1 um. The platelets induce Ing mechanism. 25-2 constrained sintering of the ceramie powder, which results in The use of porous coatings has been proposed as a universal and permanent porosity. For room-temperature properties, onl simple way to obtain weak interphases. , 4 However, several minor additions(0-3 vol %)of matrix powder yield sufficiently weak debonding interphases. The platelets lie in random, eventual closure of pores that accompanies continued sintering three-dimensional orientations and provide a de bonding mech- during prolonged operation at high temperatures. Porous mullite anism that is independent of temperature, in chemically com (3AlO3'2SiO,) and alumina(Al,O3) matrices that have been patible matrixes. Laminated composites with two types of matrixes-mullite and alumina-have been fabricated with reinforced with uncoated Al,O3 fibers show promising results for modified fibrous monoliths of alumina in a triple-layer"core/ porous composites that have relatively simple fabrication require ments. However, the extensive matrix porosity limits the overall terphase/matrix"arrangement. In the laminated systems, mechanical strength of the material and, hence, its use in load the intimate mixing of strong versus tough microstructures h been tailored by alternating va Is matrix: interphase thick- bearing applications Thus, this work addresses the need for a tough, flaw-tolerant, ness ratios. Preliminary load-displacement curves clearly all-oxide composite that is relatively dense and is capable of demonstrate characteristics of"graceful failure"and notabl improvements in the work of fracture. Scanning electron sustained performance in an oxidizing environment. The intended microscopic observation of the crack paths confirms the approach uses the phenomenon of"constrained sintering"-to viability of platelets for producing permanently porous, produce permanent porosity in an interphase that should still be debondable interphases at elevated temperatures in air. sufficiently weak to deflect a propagating crack. In this method- ology, there are competing effects of grain growth versus densifi- cation that limit the complete densification of a two-phase mixture, lting in a rigid bu orous body. In recent studies laboratory, 37-39 AL,, platelets of various ranges of size an B RITTLENESS and unreliability each still are difficult problems in volume fraction were distributed in 3AL, 0 2SiO, and zirconia of ceramics. Attempts to impart" graceful failure (ZrO2) matrixes. The resulting microstructures confirmed the which is analogous to ductility in metalshave been partially constrained-sintering effect, with the formation of a stable, uni successful with the use of composites. I Toughening is now well formly porous material in the case of the 3A1203 2SiO, matrix established to result from debonding at an interface (or within an In this paper, we propose that a suitable mixture of Al2O interphase) between a matrix and a reinforcement, or between platelets and ceramic powder, which constitutes an interphase, will laminates in a composite. -Alternatively, in the absence of be weak and porous enough to deflect a crack along the interphase reinforcements, different geometric configurations that incorporate in a composite, as illustrated schematically in Fig. 1. This weak interphase and crack-deflecting or crack-energy-dissipatin henomenon will result in crack blunting, macroscopic-crac systems have been produced. 0-2 These systems include lami deflection and frictional work to be done. which leads to overall nated composites such as silicon carbide-graphite(SiC-C)o, I toughening and flaw tolerance of the composite. This mechanism and silicon nitride-silicon nitride-silicon nitride whiskers(Si, N- has an advantage in that it is independent of temperature, thus, it (Si N(30-vol%Si N, whiskers). 2 Fibrous monoliths such as can function, for example, at temperatures up to 1700C in an Sic-C and silicon nitride/boron nitride(si3Na-BN)have also beer AL2O -rich mullite(2Al O, SiO2 ),4 or up to -1900C in a fabricated; the latter system has been studied extensively. itable matrix such as yttrium aluminate garnet(Y3AlsO12,or Other oxide systems that are based on monazite (LaPO)8-20 and YAG), yttria-stabilized zirconia(3-mol%-Y2O3-ZrO2, or 3Y related xenotime(YPO4 )2-24 interphases also have emerged. TZP), or nickel aluminate(NiAl, O4). The other salient feature of this design is that the porosity is relatively permanent, as opposed to the transient porosity that is achieved in current, fiber- reinforced, porous interphases or matrices, which are made using P. Padture--contributing editor a fugitive graphite phase. Because of the interphase porosity, is anticipated that residual-stress issues that are due to thermal mismatch between the matrix and the ase will be Manuscript No 189060 Received October ed October 31. 2000 he concept of a porous interphase already has been demon- platelets or inclusions is known to retard densification. However, the combination of these two effects to produce a weak, perma nently porous, debondable interphase is a novel approach for National University, Chyunggye-myun, Muan-gun, Chonnam, Republic of Korea. producing tough, high-temperature oxide composites
Toughened Oxide Composites Based on Porous Alumina-Platelet Interphases Sang-Jin Lee* ,† and Waltraud M. Kriven** Department of Materials Science and Engineering, University of Illinois at Urbana–Champaign, Urbana, Illinois 61801 A novel mechanism for debonding at a weak interphase in an all-oxide composite is introduced. This methodology involves the use of alumina platelets that have a diameter of 10–15 or 5–10 mm and a thickness of 1 mm. The platelets induce constrained sintering of the ceramic powder, which results in permanent porosity. For room-temperature properties, only minor additions (0–3 vol%) of matrix powder yield sufficiently weak debonding interphases. The platelets lie in random, three-dimensional orientations and provide a debonding mechanism that is independent of temperature, in chemically compatible matrixes. Laminated composites with two types of matrixes—mullite and alumina—have been fabricated with modified fibrous monoliths of alumina in a triple-layer “core/ interphase/matrix” arrangement. In the laminated systems, the intimate mixing of strong versus tough microstructures has been tailored by alternating various matrix:interphase thickness ratios. Preliminary load–displacement curves clearly demonstrate characteristics of “graceful failure” and notable improvements in the work of fracture. Scanning electron microscopic observation of the crack paths confirms the viability of platelets for producing permanently porous, debondable interphases at elevated temperatures in air. I. Introduction BRITTLENESS and unreliability each still are difficult problems in the use of ceramics. Attempts to impart “graceful failure”— which is analogous to ductility in metals—have been partially successful with the use of composites.1 Toughening is now well established to result from debonding at an interface (or within an interphase) between a matrix and a reinforcement, or between laminates in a composite.2–9 Alternatively, in the absence of reinforcements, different geometric configurations that incorporate a weak interphase and crack-deflecting or crack-energy-dissipating systems have been produced.10–23 These systems include laminated composites such as silicon carbide–graphite (SiC–C)10,11 and silicon nitride–silicon nitride–silicon nitride whiskers (Si3N4– (Si3N4–(30-vol% Si3N4 whiskers)).12 Fibrous monoliths such as SiC–C and silicon nitride/boron nitride (Si3N4–BN) have also been fabricated; the latter system has been studied extensively.13–17 Other oxide systems that are based on monazite (LaPO4) 18–20 and related xenotime (YPO4) 21–24 interphases also have emerged. More recently, the formation of transformation-weakened interphases that result from a negative volume change in the crystal structures has been demonstrated to be a viable interphase debonding mechanism.25–27 The use of porous coatings has been proposed as a universal and simple way to obtain weak interphases.28,29 However, several problems still remain in fiber-reinforced composites, such as the eventual closure of pores that accompanies continued sintering during prolonged operation at high temperatures. Porous mullite (3Al2O3z2SiO2) and alumina (Al2O3) matrices that have been reinforced with uncoated Al2O3 fibers show promising results for porous composites that have relatively simple fabrication requirements.30 However, the extensive matrix porosity limits the overall mechanical strength of the material and, hence, its use in loadbearing applications. Thus, this work addresses the need for a tough, flaw-tolerant, all-oxide composite that is relatively dense and is capable of sustained performance in an oxidizing environment. The intended approach uses the phenomenon of “constrained sintering”31–38 to produce permanent porosity in an interphase that should still be sufficiently weak to deflect a propagating crack. In this methodology, there are competing effects of grain growth versus densification that limit the complete densification of a two-phase mixture, resulting in a rigid but porous body. In recent studies in our laboratory,37–39 Al2O3 platelets of various ranges of size and volume fraction were distributed in 3Al2O3z2SiO2 and zirconia (ZrO2) matrixes. The resulting microstructures confirmed the constrained-sintering effect, with the formation of a stable, uniformly porous material in the case of the 3Al2O3z2SiO2 matrix. In this paper, we propose that a suitable mixture of Al2O3 platelets and ceramic powder, which constitutes an interphase, will be weak and porous enough to deflect a crack along the interphase in a composite, as illustrated schematically in Fig. 1. This phenomenon will result in crack blunting, macroscopic-crack deflection, and frictional work to be done, which leads to overall toughening and flaw tolerance of the composite. This mechanism has an advantage in that it is independent of temperature; thus, it can function, for example, at temperatures up to 1700°C in an Al2O3-rich mullite (2Al2O3zSiO2) 40,41 or up to ;1900°C in a suitable matrix such as yttrium aluminate garnet (Y3Al5O12, or YAG), yttria-stabilized zirconia (3-mol%-Y2O3–ZrO2, or 3YTZP), or nickel aluminate (NiAl2O4). The other salient feature of this design is that the porosity is relatively permanent, as opposed to the transient porosity that is achieved in current, fiberreinforced, porous interphases or matrices, which are made using a fugitive graphite phase.30 Because of the interphase porosity, it is anticipated that residual-stress issues that are due to thermalexpansion mismatch between the matrix and the interphase will be minimized. The concept of a porous interphase already has been demonstrated to be a viable debonding mechanism, and the use of platelets or inclusions is known to retard densification. However, the combination of these two effects to produce a weak, permanently porous, debondable interphase is a novel approach for producing tough, high-temperature oxide composites. N. P. Padture—contributing editor Manuscript No. 189060. Received October 25, 1999; approved October 31, 2000. This work was supported by the Argonne National Laboratory, with funding from the Defense Advanced Research Projects Agency (DARPA), through a Department of Energy Interagency Agreement, under Contract W-31-109-Eng-38. *Member, American Ceramic Society. **Fellow, American Ceramic Society. † Present address: Department of Materials Science and Engineering, Mokpo National University, Chyunggye-myun, Muan-gun, Chonnam, Republic of Korea. J. Am. Ceram. Soc., 84 [4] 767–74 (2001) 767 journal
Journal of the American Ceramic Society-Lee and Kriven Vol 84. No 4 of 1C/min. The laminated green bodies were cold isostatically pressed(CIPed) at a pressure of 270 MPa for 5 min and then POROUS INTERPHASE pressureless-sintered at 1600C for 10 h(for the 3Al2O3. 2SiO matrix)or 2 h(for the Al,O, matrix)to densify the samples After the densified laminates were co-fired, they were cut int ALUMINA PLATELETS the form of bend bars The cutt on was along the longitudinal axis of the specimens in the plane of the lamination The bend bars, which were 30 mm long, 4.0 mm thick, and 3.0 mm wide, were tested in three-point flexure, without any surface polishing of the bend-bar specimens (2) Fabrication of Fibrous Monoliths The conventional fibrous monolithic forming method was adopted" but with some modification: a specific burnout process, RACK PROPAGATION as well as a post-burnout CIP step, was included. The first step in the fabrication of the fibrous composites was to batch the respec. tive polymer and ceramic powder in a high-shear, twin-roll mixer (C. w. Brabender Instruments, Inc, South Hackensack, NJ). In the Fig. 1. preparation of the fibrous ceramics, a paste(which consisted of matic illustration of proposed crack-deflection mechanism in 30-40 vol% polymer and 60-70 vol% ceramic powder)was posite containing porous alumina platelets and weak interphases Synergistic, energy-dissipating mechanisms of crack deflection, crack prepared from ethylene vinyl acetate(Elvax 470TM, which is a blunting, and grain bridging within pores are in operation. co-polymer thermoplastic that is manufactured by DuPont(Or- ange, TX) with a softening temperature of 120C and a density of -1 The mechanism will be demonstrated in the two matrices- powders(Al2O, for the matrix and Al2O, platelets for the Al,O3 and stoichiometric mullite(3A1,03 2SiO,hto compare interphase). The ceramic powders and the polymers were loaded their strengths. Conventional tape casting and co-extrusion tech into the shearing chamber, which was electrically heated to 150C niques will be used to engineer a series of composites in laminated When the high-shear mixing began, the viscosity of the resulting and fibrous monolithic+-configurations, respectively. The effects paste was monitored as a function of time of innovative modifications in design on the mechanical properties The AlzO3-platelet paste was warm-pressed into relatively thi of the composites also will be investigated. These properties weak layers, for use as cell boundaries in the fibrous composite include a bimodal variation of matrix interphase thickness ratio in AlO3-matrix sheets also were warm-pressed( Carver Press, Fred the laminated composites. In the fibrous monoliths, a triple- layer S. Carver, Inc., Menomonee Falls, wi) to a predetermined thick- repeating unit that has a"core/interphase/matrix"construction will be fabricated; this assembly ensures that no continuous weak path A2 O3 matrix and the Al2Or-platelet interphase. The pastes wer through the composite has been designed into the microstructure pressed at a temperature of 140.C under a pressure of 130 MPa for min. Hardened paste pieces of Al2O3 also were loaded into a cylinder assembly that had an inner diameter of 16 mm. Formation IL. Experimental Procedure of the Al,O, feeder rod via compression molding in a universal (I Fabrication of laminates testing machine (Model 4502, Instron Corp, Canton, MA)was Laminates of mullite(3AI03 2SiO2)and alumina(Al203)were conducted in the pressure range of 1-3 MPa; the ram spee bricated using the tape-casting process. These experiments used onitored at 20 mm/s, and the extrusion temperature was set at 3AL2032SiO, powder(KM Mullite-101, Kyoritsu, Inc, Nagoya, Japan), Al,O, powder(Al6SG, Alcoa, New Milford, CT), and After the specimens were formed, the Al2O3-platelet cell Al2O3 platelets(Atochem, Pierre-Benite, France). The slurries for boundary and Al,,-matrix layers were incorporated into the tape casting had the following composition: - 30 vol% oxide mIcrostructure. These layers were applied by wrapping the warm pressed sheet around the Al,, feeder rod; first the Al,O3-platelet butyral)(PVB)(0.5 wt%)(Monsanto, Inc. St. Louis, Mo) was layer, then the AlzO -matrix layer. Atter the green body was added to the slurries as a dispersant. The solvent was composed of tightly wrapped, it was warm-extruded through an orifice(2 mm in mixtures of toluene, n-butyl alcohol, and ethanol(all manufactured diameter) at a ram rate of 5 mm/s, with the extrusion temperature by Aldrich Chemical Co., Milwaukee, WI). The approximate set at 140C. As the green, layered filament came through the mixing ratio of toluene, n-butyl alcohol, and ethanol was 20: 20: 60 orifice of the spinneret, it was cut and then tightly packed into the (by vol%). The organics included PVB, which was used as a extrusion block. Then, a second warm-extrusion pass binder, and polyethylene glycol(PEG 2000) and dioctyl phthalate formed through the same spinneret. The exiting (Aldrich Chemical Co. ) each was used as a plasticizer. The collected and cut into segments 51 mm long, to be organics ratio was 50 wt% binder 50 wt% plasticizer. After th rectangular mold, and then warm-pressed at 140C slurry was pulverized, the binder and plasticizers were added and pressure of 27 MPa into a billet. ball-milled for 24 h. In the case of the Al,O3-platelet slurry, the The organic additives were removed by heating to 700C for 5h mixing was performed by just in an air atmosphere, with a slow heating schedule(0.1C/min) the working vIscosity. a acuI Of he& for 24 h, instead of ball platelet morphology. The within the range of 120%-200oC. After the burnout process, the slurries were stirred in a va to remove any bubbles and adjust samples wer of 270 MPa for 5 min and fter the slurries were aged for I d, they pressureless-sintered at 1600.C for 2 h. The sintered samples were ere tape-cast, using a doctor-blade opening of 130-300 um, to cut and tested in flexure in the same way as that mentioned for the obtain green sheets 50-150 um thick. The cast tapes were dried lamination process under a saturated solvent atmosphere for l d The green laminate composites had mm X 51 mm after the green sheets were exude Bend-bar samples were used to per- arrangement. Thermocompression was per three-point testing at room temperature in the 10 MPa for 10 min at a temperature of 80C, which was the previously mentioned universal testing machine, using a span oftening point of the organics. Subsequently, the organic addi- ength of 20 mm and a crosshead speed of 0.01 mm/min. The tives were removed by heating to 600%C in air, using a heating rate apparent work of fracture (WOF) was obtained by dividing the
The mechanism will be demonstrated in the two matrices— Al2O3 and stoichiometric mullite (3Al2O3z2SiO2)—to compare their strengths. Conventional tape casting and co-extrusion techniques will be used to engineer a series of composites in laminated and fibrous monolithic42 configurations, respectively. The effects of innovative modifications in design on the mechanical properties of the composites also will be investigated. These properties include a bimodal variation of matrix:interphase thickness ratio in the laminated composites. In the fibrous monoliths, a triple-layer repeating unit that has a “core/interphase/matrix” construction will be fabricated; this assembly ensures that no continuous weak path through the composite has been designed into the microstructure. II. Experimental Procedure (1) Fabrication of Laminates Laminates of mullite (3Al2O3z2SiO2) and alumina (Al2O3) were fabricated using the tape-casting process. These experiments used 3Al2O3z2SiO2 powder (KM Mullite-101, Kyoritsu, Inc., Nagoya, Japan), Al2O3 powder (A16SG, Alcoa, New Milford, CT), and Al2O3 platelets (Atochem, Pierre-Be´nite´, France). The slurries for tape casting had the following composition: ;30 vol% oxide powders, ;55 vol% solvent, and ;15 vol% organics. Poly(vinyl butyral) (PVB) (0.5 wt%) (Monsanto, Inc., St. Louis, MO) was added to the slurries as a dispersant. The solvent was composed of mixtures of toluene, n-butyl alcohol, and ethanol (all manufactured by Aldrich Chemical Co., Milwaukee, WI). The approximate mixing ratio of toluene, n-butyl alcohol, and ethanol was 20:20:60 (by vol%). The organics included PVB, which was used as a binder, and polyethylene glycol (PEG 2000) and dioctyl phthalate (Aldrich Chemical Co.); each was used as a plasticizer. The organics ratio was 50 wt% binder:50 wt% plasticizer. After the slurry was pulverized, the binder and plasticizers were added and ball-milled for 24 h. In the case of the Al2O3-platelet slurry, the mixing was performed by just stirring for 24 h, instead of ball milling, to prevent breakage of the platelet morphology. The slurries were stirred in a vacuum to remove any bubbles and adjust the working viscosity. After the slurries were aged for 1 d, they were tape-cast, using a doctor-blade opening of 130–300 mm, to obtain green sheets 50–150 mm thick. The cast tapes were dried under a saturated solvent atmosphere for 1 d. The green laminate composites had area dimensions of 25 mm 3 51 mm after the green sheets were stacked in an alternating arrangement. Thermocompression was performed under a load of 10 MPa for 10 min at a temperature of 80°C, which was the softening point of the organics. Subsequently, the organic additives were removed by heating to 600°C in air, using a heating rate of 1°C/min. The laminated green bodies were cold isostatically pressed (CIPed) at a pressure of 270 MPa for 5 min and then pressureless-sintered at 1600°C for 10 h (for the 3Al2O3z2SiO2 matrix) or 2 h (for the Al2O3 matrix) to densify the samples. After the densified laminates were co-fired, they were cut into the form of bend bars. The cutting direction was along the longitudinal axis of the specimens in the plane of the lamination. The bend bars, which were 30 mm long, 4.0 mm thick, and 3.0 mm wide, were tested in three-point flexure, without any surface polishing of the bend-bar specimens. (2) Fabrication of Fibrous Monoliths The conventional fibrous monolithic forming method was adopted40 but with some modification: a specific burnout process, as well as a post-burnout CIP step, was included. The first step in the fabrication of the fibrous composites was to batch the respective polymer and ceramic powder in a high-shear, twin-roll mixer (C. W. Brabender Instruments, Inc., South Hackensack, NJ). In the preparation of the fibrous ceramics, a paste (which consisted of 30–40 vol% polymer and 60–70 vol% ceramic powder) was prepared from ethylene vinyl acetate (Elvax 470™, which is a co-polymer thermoplastic that is manufactured by DuPont (Orange, TX) with a softening temperature of 120°C and a density of 0.94 g/cm3 ), drops of PEG 2000 (as a plasticizer), and the ceramic powders (Al2O3 for the matrix and Al2O3 platelets for the interphase). The ceramic powders and the polymers were loaded into the shearing chamber, which was electrically heated to 150°C. When the high-shear mixing began, the viscosity of the resulting paste was monitored as a function of time. The Al2O3-platelet paste was warm-pressed into relatively thin, weak layers, for use as cell boundaries in the fibrous composite. Al2O3-matrix sheets also were warm-pressed (Carver Press, Fred S. Carver, Inc., Menomonee Falls, WI) to a predetermined thickness that was based on the desired thickness ratio between the Al2O3 matrix and the Al2O3-platelet interphase. The pastes were pressed at a temperature of 140°C under a pressure of 130 MPa for 1 min. Hardened paste pieces of Al2O3 also were loaded into a cylinder assembly that had an inner diameter of 16 mm. Formation of the Al2O3 feeder rod via compression molding in a universal testing machine (Model 4502, Instron Corp., Canton, MA) was conducted in the pressure range of 1–3 MPa; the ram speed was monitored at 20 mm/s, and the extrusion temperature was set at 145°C. After the specimens were formed, the Al2O3-platelet cell boundary and Al2O3-matrix layers were incorporated into the microstructure. These layers were applied by wrapping the warmpressed sheet around the Al2O3 feeder rod: first the Al2O3-platelet layer, then the Al2O3-matrix layer. After the green body was tightly wrapped, it was warm-extruded through an orifice (2 mm in diameter) at a ram rate of 5 mm/s, with the extrusion temperature set at 140°C. As the green, layered filament came through the orifice of the spinneret, it was cut and then tightly packed into the extrusion block. Then, a second warm-extrusion pass was performed through the same spinneret. The exiting filament was collected and cut into segments 51 mm long, to be laid up in a rectangular mold, and then warm-pressed at 140°C under a pressure of 27 MPa into a billet. The organic additives were removed by heating to 700°C for 5 h in an air atmosphere, with a slow heating schedule (0.1°C/min) within the range of 120°–200°C. After the burnout process, the samples were CIPed at a pressure of 270 MPa for 5 min and then pressureless-sintered at 1600°C for 2 h. The sintered samples were cut and tested in flexure in the same way as that mentioned for the lamination process. (3) Characterization (A) Flexural Testing: Bend-bar samples were used to perform three-point flexural testing at room temperature in the previously mentioned universal testing machine, using a span length of 20 mm and a crosshead speed of 0.01 mm/min. The apparent work of fracture (WOF) was obtained by dividing the Fig. 1. Schematic illustration of proposed crack-deflection mechanism in a composite containing porous alumina platelets and weak interphases. Synergistic, energy-dissipating mechanisms of crack deflection, crack blunting, and grain bridging within pores are in operation. 768 Journal of the American Ceramic Society—Lee and Kriven Vol. 84, No. 4
pril 2001 Toughened Oxide Composites Based on Porous Alumina-Platelet Interphases area under the load-displacement curve by the cross-sectional area along the porous region at the center of the interphase is observed of the sample. The relative WoF values can be compared,(see Fig. 2(b)) because the specimens had essentially the same dimensions. The exceptions were some specimens that had different dimensions, as indicated at the bottom of the respective tables in this paper, which Fracture Behavior of the Mullite Laminates, Relative to the Mullite Content in the Interphase arize the mechanical-property data (B) Microstructure Characterization: The microstructure of The results of flexural testing for the 3Al,O3 2SiO2-matrix laminates, which have a variable 3Al2O3 2SiO, content in the the platelet powders, the surfaces of the sintered samples, and their A1,O3-platelet interphases, are listed in Table I. The strength side views and fracture surfaces after bend testing were observed increased as the 3A1,0,. 2SiO, content in the interphase increased sing scanning electron microscopy(SEM)(Model DS-130, Inter this result was attributed to a denser, albeit still porous, micro national Scientific Instruments, Santa Clara, CA) structure that formed during sintering. Load-displacement curve for the 3Al,O3 2SiO, laminates, as a function of the 3Al2O3 2SiO2 content in the interphase, are shown in Fig 3. The bend bars with I. Results 3Al20, 2SiO2 contents of 5, 10, and 20 vol% produced fracture (I Microstructure of the Platelets and the interphase curves with only a few steps before failure(see Fig. 3(a)). The The typical morphology of the Al2O3 platelets was hexagonal stepwise load drops are characteristic of "graceful failure"and indicate crack deflection and debonding at the interphase. Slightly crystals -1 um thick with diameters in relatively narrow size- more steps were observed in the laminates that had a distribution ranges: 3-5, 5-10, 10-15, and 20-25 Hm. In this 3Al20, 2SiO2 content of 5 vol%(see Fig 3(a)). However,the surfaces. Figure 2 shows representative micrographs of a lami- of the laminate that did not contain any 3Al,O3 2SiO2 nated 3A12O3 2SiO2 matrix with an Al,O3-platelet interphase. The platelets are randomly oriented and clearly exhibit a nonsintere in the interphase was rather low. The WoF at room rature was less for 3Al2O3 2SiO2 additions of >5 vol% region between the 3A10, 2Si0, matrix and the Al,. -platelet (despite relate e ye thic ker interphases compared to the mat phase is relatively dense. After fracture testing, crack deflection interphase)than in laminates that had I and 3 vol%of 3Al2O3, 2SiO2 in the interphase. This observation implies that, at room temperature, interphases with a 3Al2O32SiO2 content of >5 (a) vol% were not weak enough for crack deflection to occur. Micrographs of the tensile-side(failed-side) view of the lami- nates with 3Al,O3 2SiO, contents of 5 and 3 vol% are shown in Figs. 4(a)and (b), respectively. In the 3-vol%0-3Al2O3 2SiO interphase specimen, the crack deflection noticeably proceeded along the weak Al,O3-platelet interphase, despite the interphase Interphase layers being relatively thin(-80 um). This observation implies that the crack deflection is strongly dependent on the microstruc- ure of the porous interphase, rather than its thickness ( Fracture Behavior of the Laminates, Relative to the Matrix: Interphase Thickness Ratio aminates with 3A1,O3, or Al,O, matrixes in different matrix interphase thickness ratios are presented in Tables ll and Ill. The strength and Matrix WoF each increased as the matrix thickness increased In partic ular. the laminates that had a"bimodal" design showed noticeabl 899225K58um higher strength and wOF. In 3Al,O3, 2SiO -matrix laminates, the sample with a 4: I thickness ratio had a higher strength and WoF than did that of the laminate that had a 6: I thickness ratio, despite (b) a thicker interphase. This result is attributed to the effect of the interphase. The effects of thickness ratio and 3Al2O3'2SiO Crack content were optimized in the bimodal 3Al,O, ', laminate that had alternating matrix layers, with matrix interphase ratios of 3 Table I. Variation in Strength and work of fracture for Mullite -Matrix Laminates, According to the Mullite Content Interphase in the Alumina-Platelet(10-15 um)Interphases Thickness Flexural strength Work of fracture, WOF Matrix 889625KV588u 246222 Fig. 2. SEM micrographs of the ite with alumina-platelet- TDensified matrix interphase thickness ratio, ' For specimens 30 mm(length)x4.0 weakened interphases ((a) cross section and (b) crack profile)
area under the load–displacement curve by the cross-sectional area of the sample.43 The relative WOF values can be compared, because the specimens had essentially the same dimensions. The exceptions were some specimens that had different dimensions, as indicated at the bottom of the respective tables in this paper, which summarize the mechanical-property data. (B) Microstructure Characterization: The microstructure of the platelet powders, the surfaces of the sintered samples, and their side views and fracture surfaces after bend testing were observed using scanning electron microscopy (SEM) (Model DS-130, International Scientific Instruments, Santa Clara, CA). III. Results (1) Microstructure of the Platelets and the Interphase The typical morphology of the Al2O3 platelets was hexagonal crystals ;1 mm thick with diameters in relatively narrow sizedistribution ranges: 3–5, 5–10, 10–15, and 20–25 mm. In this study, the 5–10 mm and 10–15 mm size ranges were determined to be the most useful. The surfaces of the platelets were sometimes pitted in the middle but were relatively free of pits on the outer surfaces. Figure 2 shows representative micrographs of a laminated 3Al2O3z2SiO2 matrix with an Al2O3-platelet interphase. The platelets are randomly oriented and clearly exhibit a nonsintered, porous microstructure (see Fig. 2(a)). However, the interfacial region between the 3Al2O3z2SiO2 matrix and the Al2O3-platelet phase is relatively dense. After fracture testing, crack deflection along the porous region at the center of the interphase is observed (see Fig. 2(b)). (2) Fracture Behavior of the Mullite Laminates, Relative to the Mullite Content in the Interphase The results of flexural testing for the 3Al2O3z2SiO2-matrix laminates, which have a variable 3Al2O3z2SiO2 content in the Al2O3-platelet interphases, are listed in Table I. The strength increased as the 3Al2O3z2SiO2 content in the interphase increased; this result was attributed to a denser, albeit still porous, microstructure that formed during sintering. Load–displacement curves for the 3Al2O3z2SiO2 laminates, as a function of the 3Al2O3z2SiO2 content in the interphase, are shown in Fig. 3. The bend bars with 3Al2O3z2SiO2 contents of 5, 10, and 20 vol% produced fracture curves with only a few steps before failure (see Fig. 3(a)). The stepwise load drops are characteristic of “graceful failure” and indicate crack deflection and debonding at the interphase. Slightly more steps were observed in the laminates that had a 3Al2O3z2SiO2 content of ,5 vol% in the interphase (see Fig. 3(b)). The laminate that had no 3Al2O3z2SiO2 had the most steps in its load–displacement curve, in contrast to the laminates that had a 3Al2O3z2SiO2 content of .5 vol% (see Fig. 3(a)). However, the strength of the laminate that did not contain any 3Al2O3z2SiO2 powder in the interphase was rather low. The WOF at room temperature was less for 3Al2O3z2SiO2 additions of .5 vol% (despite relatively thicker interphases, compared to the matrix thickness, and, hence, more likely crack deflection along the interphase) than in laminates that had 1 and 3 vol% of 3Al2O3z2SiO2 in the interphase. This observation implies that, at room temperature, interphases with a 3Al2O3z2SiO2 content of .5 vol% were not weak enough for crack deflection to occur. Micrographs of the tensile-side (failed-side) view of the laminates with 3Al2O3z2SiO2 contents of 5 and 3 vol% are shown in Figs. 4(a) and (b), respectively. In the 3-vol%-3Al2O3z2SiO2 interphase specimen, the crack deflection noticeably proceeded along the weak Al2O3-platelet interphase, despite the interphase layers being relatively thin (;80 mm). This observation implies that the crack deflection is strongly dependent on the microstructure of the porous interphase, rather than its thickness. (3) Fracture Behavior of the Laminates, Relative to the Matrix:Interphase Thickness Ratio The results of the flexural testing of laminates with 3Al2O3z2SiO2 or Al2O3 matrixes in different matrix:interphase thickness ratios are presented in Tables II and III. The strength and WOF each increased as the matrix thickness increased. In particular, the laminates that had a “bimodal” design showed noticeably higher strength and WOF. In 3Al2O3z2SiO2-matrix laminates, the sample with a 4:1 thickness ratio had a higher strength and WOF than did that of the laminate that had a 6:1 thickness ratio, despite a thicker interphase. This result is attributed to the effect of the addition of 1 vol% of 3Al2O3z2SiO2 powder to the platelet interphase. The effects of thickness ratio and 3Al2O3z2SiO2 content were optimized in the bimodal 3Al2O3z2SiO2 laminate that had alternating matrix layers, with matrix:interphase ratios of 3:1 Fig. 2. SEM micrographs of the composite with alumina-plateletweakened interphases ((a) cross section and (b) crack profile). Table I. Variation in Strength and Work of Fracture for Mullite-Matrix Laminates, According to the Mullite Content in the Alumina-Platelet (10–15 mm) Interphases Mullite content (vol%) Thickness ratio† Flexural strength (MPa) Work of fracture, WOF‡ (kJ/m2 ) 0 2:1 70 0.4 1 4:1 77 0.6 3 6:1 88 0.6 5 2:1 86 0.4 10 2:1 97 0.4 20 2:1 128 0.4 † Densified matrix:interphase thickness ratio. ‡ For specimens 30 mm (length) 3 4.0 mm (thickness) 3 3.0 mm (width). April 2001 Toughened Oxide Composites Based on Porous Alumina-Platelet Interphases 769
770 Journal of the American Ceramic Sociery-Lee and Riven Vol 84. No 4 0.2 Mullite 20 vol% Crack Matrix Mullite 5 vol% Only platelets 0.05 oUtset+Ky serum 0.15 Displacement (mm) Crack Matrix Mullite 3 vol% 星015F如m WOF=0.6 KJ/m2 0.05 Fig. 4. Failure-side view SEM micrographs of the laminate composites with mullite contents of (a)5 and (b)3 vol% in the interphases. In the latter image, crack deflection is evident along the porous platelet interphase 00.050.10.150.2025 Table l. Variation in Strength and Work of Fracture for Load-displacement curves of mullite-matrix laminate com- function of the mullite content in the weak interphases Laminates, According to the Thickness Ratio between the Mullite Matrix and the Alumina-Platelet (10-15 um) ase thickness ratios of (a)2: 1 and(b) 4: 1 and 6: 1 are Table D) Thickness llite content Flexural strength Work of fracture. WOF (vo9) 2:1 and 9: 1. Graceful-failure characteristics were observed in the corresponding load-displacement curve To increase the overall strength of the composite, the Al,O atrix was chosen in a bimodal sequence of alternating 12: 1 an Densified matrix interphase thickness ratio. "Bimodal"denotes an alternat 5: 1 ratios, where the alternating regions of 12: 1 and 5: I ratios qual thickness. When up to 3 vol% of 3Al,O3 2SiO, was added 4.0 mm x 3.0 mm .Mullite content in the alt the interphase, the WOF notably increased. The highest strength nd WOF values(112 MPa and 2.1 kJ/m", respectively )were bserved. Again, graceful-failure characteristics were observed in the load-displacement curves for Al2O,(see Fig. 5(a) and Table the matrix and the interphase, even though they consisted of the Ill). Figure 5(b) shows an SEM micrograph that illustrates the same material crack profile that corresponds to the optimized load-displacement curve shown in Fig. S(a). The deflected crack passed through the (4) Mechanical Behavior of the Laminates, Relative to the center of the porous interphase Size of the Alumina Platelets The Al,O3 platelets provided an easy crack-deflection rout The results of the flexural testing for the laminates, according to even in a thin platelet interphase with an Al,O,- matrix interphase the size of the alumina platelets in the pure-Al2O3 interphase, are ickness ratio of 15: 1. This result confirmed that the interphase listed in Tables IV and v. Correspondingly, the load-displacement was much weaker than the matrix. This finding was consistent with curves of the specimens that had platelets 5-10 um in size in the the expectation that no reaction or densification occurred between interphases are shown in Fig. 6. In the 3Al,O3 2SiO2-matrix
and 9:1. Graceful-failure characteristics were observed in the corresponding load–displacement curves. To increase the overall strength of the composite, the Al2O3 matrix was chosen in a bimodal sequence of alternating 12:1 and 5:1 ratios, where the alternating regions of 12:1 and 5: 1 ratios had equal thickness. When up to 3 vol% of 3Al2O3z2SiO2 was added to the interphase, the WOF notably increased. The highest strength and WOF values (112 MPa and 2.1 kJ/m2 , respectively) were observed. Again, graceful-failure characteristics were observed in the load–displacement curves for Al2O3 (see Fig. 5(a) and Table III). Figure 5(b) shows an SEM micrograph that illustrates the crack profile that corresponds to the optimized load–displacement curve shown in Fig. 5(a). The deflected crack passed through the center of the porous interphase. The Al2O3 platelets provided an easy crack-deflection route, even in a thin platelet interphase with an Al2O3-matrix:interphase thickness ratio of 15:1. This result confirmed that the interphase was much weaker than the matrix. This finding was consistent with the expectation that no reaction or densification occurred between the matrix and the interphase, even though they consisted of the same material. (4) Mechanical Behavior of the Laminates, Relative to the Size of the Alumina Platelets The results of the flexural testing for the laminates, according to the size of the alumina platelets in the pure-Al2O3 interphase, are listed in Tables IV and V. Correspondingly, the load–displacement curves of the specimens that had platelets 5–10 mm in size in the interphases are shown in Fig. 6. In the 3Al2O3z2SiO2-matrix Fig. 3. (a) Load–displacement curves of mullite-matrix laminate composites, as a function of the mullite content in the weak interphases. Alumina platelets 10–15 mm in size are present in the interphases, and matrix:interphase thickness ratios of (a) 2:1 and (b) 4:1 and 6:1 are observed (see Table I). Fig. 4. Failure-side view SEM micrographs of the laminate composites with mullite contents of (a) 5 and (b) 3 vol% in the interphases. In the latter image, crack deflection is evident along the porous platelet interphase. Table II. Variation in Strength and Work of Fracture for Laminates, According to the Thickness Ratio between the Mullite Matrix and the Alumina-Platelet (10–15 mm) Interphases Thickness ratio† Mullite content‡ (vol%) Flexural strength (MPa) Work of fracture, WOF (kJ/m2 ) 2:1 0 70 0.5 6:1 0 75 0.6 4:1 1 77 0.6 Bimodal 1 102 1.1 † Densified matrix:interphase thickness ratio. “Bimodal” denotes an alternative combination of thickness ratios of 3:1 and 9:1 and a specimen size of 30 mm (length) 3 4.5 mm (thickness) 3 3.0 mm (width), rather than the normal 30 mm 3 4.0 mm 3 3.0 mm. ‡ Mullite content in the alumina-platelet interphase. 770 Journal of the American Ceramic Society—Lee and Kriven Vol. 84, No. 4
April 2001 Toughened Oxide Composites Based on Porous Alumina-Platelet Interphases Table Ill. Variation in Strength and work of Fracture for Table Iv. variation in Strength and Work of Fracture for Laminates, According to the Thickness ratio between the Mullite-Matrix Laminates, According to Platelet Size in the Alumina Matrix and the Alumina-Platelet(10-15 pm Pure Alumina-Platelet Interphase Interphases Platelet size Thickness Flexural strength Work of fracture, wol Thickness Mullite content Flexural strength Work of fracture, WOF (vol% 0-15 Densified matrix: interphase thickness ratio th)x 4.5 mm( thickness)x 3.0 mm (width), rather than the normal 30 mm X Table V. Variation in Strength and work of fracture for 3.0 Mullite content in the alumina-platelet interphase. Alumina-Matrix Laminates, According to Platelet Size in the Pure Alumina-Platelet Interphase Platelet size Thickness Flexural strengt Work of fracture, WOF Thickness ratio 15: 1 Thickness ratio bimodal g=123 MPa 0=112 MPa 0.25WOF=1.5kJ/m WOF= 2.1 kJ/m 1.0 Densified matrix interphase thickness ratio. 0.2 1 0.25 0.15 =109 MPa 0.2 I Mullite matrix σ=84MI WOF=0.2 kJ/m2 0.15 0.3 Displacement (mm) Matrix L Crack () Mechanical Behavior of the Fibrous Ceramic Compo To make tough, flaw-tolerant, fibrous ceramic composites, the results of the mechanical testing of the laminate composites were g composite. The microstructures of the as-sintered, fibrous ceramic Fig.5.(a)Load-displacement curves of the alumina-matrix laminate composites are shown in Fig. 7. The discontinuous, Al2O3-platelet composites, as a function of the matrix: interphase thickness ratio(see cell boundaries, which defined the matrix and reinforcing regions Table III), the bimodal microstructure consisted of alternating layers of 2: 1 and 5: I matrix: interphase thickness ratios.(b) Failure-side view SEM of Al,O3, are clearly visible. The side view of the as-sintered micrographs of the alumina-matrix laminate composite; the laminate has a fibrous ceramic composite showed the degree of uniformity in the mullite content of 3 vol% in the interphase and a bimodal thickness ratio alignment of the as-extruded filaments(see Fig. 7(b)). The patterns are similar to the side view of the bimodally designed laminated composite. This feature is one of the main advantages that are associated with this forming technique, i.e., the ability to create a laminate, which had Al,O3 platelets 5-10 um in size, notable heterogeneous microstructure with uniform cell-boundary thick rack deflection did not occur at the interphase, and a lower WoF nan that in the 10-15 um platelet 3Al2O3 2SiO, composite was Plots of flexural load versus displacement are shown bserved(Table IV). In contrast, the Al,O,matrix laminate that The strength and woF increased slightly in comparison hat ad Al,O, platelets 5-10 um in size showed improved strength of the AlO3-matrix laminate that had a bimodal thickness ratio. In and woF, in comparison with the 10-15 um Al,O3-platelet ontrast to the fracture curves of the laminates. the fibrous com ceramics exhibited unusual plasticlike behavior, they retained
laminate, which had Al2O3 platelets 5–10 mm in size, notable crack deflection did not occur at the interphase, and a lower WOF than that in the 10–15 mm platelet 3Al2O3z2SiO2 composite was observed (Table IV). In contrast, the Al2O3-matrix laminate that had Al2O3 platelets 5–10 mm in size showed improved strength and WOF, in comparison with the 10–15 mm Al2O3-platelet composite. (5) Mechanical Behavior of the Fibrous Ceramic Composite To make tough, flaw-tolerant, fibrous ceramic composites, the results of the mechanical testing of the laminate composites were applied to the fibrous design. An Al2O3 matrix and Al2O3 platelets 5–10 mm in size were used as the materials of the fibrous composite. The microstructures of the as-sintered, fibrous ceramic composites are shown in Fig. 7. The discontinuous, Al2O3-platelet cell boundaries, which defined the matrix and reinforcing regions of Al2O3, are clearly visible. The side view of the as-sintered fibrous ceramic composite showed the degree of uniformity in the alignment of the as-extruded filaments (see Fig. 7(b)). The patterns are similar to the side view of the bimodally designed laminated composite. This feature is one of the main advantages that are associated with this forming technique, i.e., the ability to create a heterogeneous microstructure with uniform cell-boundary thicknesses. Plots of flexural load versus displacement are shown in Fig. 8. The strength and WOF increased slightly in comparison with that of the Al2O3-matrix laminate that had a bimodal thickness ratio. In contrast to the fracture curves of the laminates, the fibrous ceramics exhibited unusual plasticlike behavior; they retained Table III. Variation in Strength and Work of Fracture for Laminates, According to the Thickness Ratio between the Alumina Matrix and the Alumina-Platelet (10–15 mm) Interphases Thickness ratio† Mullite content‡ (vol%) Flexural strength (MPa) Work of fracture, WOF (kJ/m2 ) 6:1 1 105 1.1 Bimodal 3 112 2.1 15:1 2 123 1.5 † Densified matrix:interphase thickness ratio. “Bimodal” denotes an alternative combination of thickness ratios of 5:1 and 12:1 and a specimen size of 30 mm (length) 3 4.5 mm (thickness) 3 3.0 mm (width), rather than the normal 30 mm 3 4.0 mm 3 3.0 mm. ‡ Mullite content in the alumina-platelet interphase. Fig. 5. (a) Load–displacement curves of the alumina-matrix laminate composites, as a function of the matrix:interphase thickness ratio (see Table III); the bimodal microstructure consisted of alternating layers of 12:1 and 5:1 matrix:interphase thickness ratios. (b) Failure-side view SEM micrographs of the alumina-matrix laminate composite; the laminate has a mullite content of 3 vol% in the interphase and a bimodal thickness ratio. Table IV. Variation in Strength and Work of Fracture for Mullite-Matrix Laminates, According to Platelet Size in the Pure Alumina-Platelet Interphase Platelet size (mm) Thickness ratio† Flexural strength (MPa) Work of fracture, WOF (kJ/m2 ) 5–10 4:1 84 0.2 10–15 6:1 75 0.3 † Densified matrix:interphase thickness ratio. Table V. Variation in Strength and Work of Fracture for Alumina-Matrix Laminates, According to Platelet Size in the Pure Alumina-Platelet Interphase Platelet size (mm) Thickness ratio† Flexural strength (MPa) Work of fracture, WOF (kJ/m2 ) 5–10 6:1 109 1.2 10–15 6:1 105 1.0 † Densified matrix:interphase thickness ratio. Fig. 6. Load–displacement curves of (– – –) mullite-matrix and (—) alumina-matrix laminate composites with a platelet size of 5–10 mm in the interphases. April 2001 Toughened Oxide Composites Based on Porous Alumina-Platelet Interphases 771
Journal of the American Ceramic Society-Lee and Riven Vol 84. No 4 0.3 Alumina/alumina platelet fibrous ceramic composite =119 MPa 0.25 WOF= 1.7 k/ 0.2 0.15 0.1 09885K上 0.3 Displacement (mi Fig. 8. Load-displacement curve of an alumina-alumina-platelet fibrous (a) 24 25Ky 500um Fig. 7.(a) SEM micrographs of the as-sintered fibrous ceramic compos- ite(a)cross section, viewed perpendicular to the fiber orientation, and(b) surface, viewed parallel to the fiber orientation) significant load-bearing capab the initial stepwise load 89625KV drops. The failed specimen is the SEM micrograph in Fig. 9. SEM micrograph of the fracture surface of the fibrous cerami Fig. 9. A slight curvature composite, as viewed from the fracture-surface side deformation, and a nonbrittle fracture surface with a woodlike fibrous cores"that were surrounded by the Al,Ox-platelet inter not exhibit a been at en decrease in strength after the ultimate tensile AL-O, core resulted from the fracture behavior, similar to the strength has tained. Instead, a notable WoF (the area under fiber-pullout effect in fiber-reinforced ceramic composites the curve)is obtained from such a composite. The correspondi SEM micrograph in Fig. 9 suggests a woodlike-fracture mecha- nism that also could operate at elevated temperatures IV. Discussion The measured mechanical properties indicated that, at room temperature, interphases of suitable weak debonding strength were The work presented here is essentially at a preliminary stage achieved using only platelets with only a minimal amount(1-3 and th p vol%)of matrix-powder additions. However, one could speculate comparative guide for further development of the microstructural that, at high temperatures, where transient creep may become an design. However, the concept of interphase debonding by a porous important issue, stronger and more-rigid interphases might be region that consists of nonsinterable platelets has been investi- required. Then, the addition of matrix powders to the interphase ed Optimization of the processing parameters can be improved may be beneficial to the overall long-term, high-temperature ignificantly. Improved presintering compaction of both the lam mechanical properties, to improve creep resistance. inates and the fibrous monoliths should increase the overall The purpose of using a bimodal microstructure was to mix trength of the composite. The matrix interphase ratio in both intimately, on a microstructural level, regions of high strength configurations also must be optimized (high matrix interphase ratio) with regions of lower strength but The fibrous-monolith configuration is versatile for uniform high toughness (low matrix interphase ratio). This concept is response to oncoming cracks perpendicular to the fibrous direc suggested in Fig. 4, where the strong matrix and interphase(which on. The optimum core interphase: matrix thickness ratio for each contained 20 vol% of 3Al203 2SiO2 powder) had high strength but of the Al,O, and 3Al2O3 2SiO2 systems could be determined low toughness (a lack of graceful-failure characteristics). The However, the preliminary load-displacement data(Fig. 8)does composite that contained only platelets in the interphase had low
significant load-bearing capability after the initial stepwise load drops. The failed specimen is shown in the SEM micrograph in Fig. 9. A slight curvature, which corresponded to permanent deformation, and a nonbrittle fracture surface with a woodlike texture were observed in the fibrous specimen. Some Al2O3 “fibrous cores” that were surrounded by the Al2O3-platelet interphase were detected in the fractured specimen. The pullout of the Al2O3 core resulted from the fracture behavior, similar to the fiber-pullout effect in fiber-reinforced ceramic composites.44 IV. Discussion The work presented here is essentially at a preliminary stage, and the mechanical data are useful primarily as a qualitative, comparative guide for further development of the microstructural design. However, the concept of interphase debonding by a porous region that consists of nonsinterable platelets has been investigated. Optimization of the processing parameters can be improved significantly. Improved presintering compaction of both the laminates and the fibrous monoliths should increase the overall strength of the composite. The matrix:interphase ratio in both configurations also must be optimized. The fibrous-monolith configuration is versatile for uniform response to oncoming cracks perpendicular to the fibrous direction. The optimum core:interphase:matrix thickness ratio for each of the Al2O3 and 3Al2O3z2SiO2 systems could be determined. However, the preliminary load–displacement data (Fig. 8) does not exhibit a sudden decrease in strength after the ultimate tensile strength has been attained. Instead, a notable WOF (the area under the curve) is obtained from such a composite. The corresponding SEM micrograph in Fig. 9 suggests a woodlike-fracture mechanism that also could operate at elevated temperatures. The measured mechanical properties indicated that, at room temperature, interphases of suitable weak debonding strength were achieved using only platelets with only a minimal amount (1–3 vol%) of matrix-powder additions. However, one could speculate that, at high temperatures, where transient creep may become an important issue, stronger and more-rigid interphases might be required. Then, the addition of matrix powders to the interphase may be beneficial to the overall long-term, high-temperature mechanical properties, to improve creep resistance. The purpose of using a bimodal microstructure was to mix intimately, on a microstructural level, regions of high strength (high matrix:interphase ratio) with regions of lower strength but high toughness (low matrix:interphase ratio). This concept is suggested in Fig. 4, where the strong matrix and interphase (which contained 20 vol% of 3Al2O3z2SiO2 powder) had high strength but low toughness (a lack of graceful-failure characteristics). The composite that contained only platelets in the interphase had low Fig. 7. (a) SEM micrographs of the as-sintered fibrous ceramic composite ((a) cross section, viewed perpendicular to the fiber orientation, and (b) surface, viewed parallel to the fiber orientation). Fig. 8. Load–displacement curve of an alumina–alumina-platelet fibrous monolithic composite. Fig. 9. SEM micrograph of the fracture surface of the fibrous ceramic composite, as viewed from the fracture-surface side. 772 Journal of the American Ceramic Society—Lee and Kriven Vol. 84, No. 4
April 2001 Toughened Oxide Composites Based on Porous Alumina-Platelet Interphases overall strength, however, the curve was characteristic of graceful References failure and significant crack diversion along the interphase. This expectation was consistent with the sEM micrographs that are IA shown in Fig 4: the platelet interphase with 5 vol% of powder Soc,73[2】187 2D. B. Marshall, B. N. Cox, and A G. Evans, " The Mechanics of Matrix Cracking exhibited brittle failure behavior(Fig. 4(a), whereas the platelet o 5g. Wogavidge, "Fiber-Reinforced Ceramics, "Composites(Guildford, UK), 18(2) terphase with additions of 3 vol% of 3Al203 2SiO2 produced noticeably more crack deflection along the interphase(Fig. 4(b). MY. He and J nson,“ Crack Det at an Interface Between Again, Fig 9 demonstrates more-extensive crack deflection along Dissimilar elas t.J. Solids struct,25{91053-67(1989) the platelet interphase that contains only 3 vol% of 3Al 2O3 2SiO .A. G. Evans, J. w. Hutchinson, "Interface Debonding and Fiber racking in brittle mposites,J.Am. Ceram Soc., 72[121 2300-303 Following the concept of a bimodal microstructure, it is W. Hutchinson and H. M. Jensen, "Models of Fiber Debonding and Pullout in postulated that regions with one layer of high matrix: interphase Brittle Composites with Friction," Mech. Mater, 9, 139-63(1990) ratio(high strength)should be alternated with regions of low R.J. Kerans, R. S Hay, N.J. Pagano, and T. A. Parthasarathy, "The Role of the matrix:interphase ratio(high toughness), which consist of several 429-42(1989). iber-Matrix Interface in Ceramic Composites, Am. Ceram. Soc. Bull, 68 [21 such thin layers---not just a couple of layers, as was fabricated A. G. Evans, "The Mechanical Performance of Fiber-Reinforced Ceramic Matrix ” Mater.Sci.Eng,A,A107,227-39(1989 has been hypothesized to have high kinetic energy. The role of the F be rein forced Band e m i (ck, isetesew of the. hy sizes. a8s gc h. of several layers with a low matrix: interphase ratio is to slow the low.J. Clegg, K. Kendall, N M. Alford, D. Birchall, and T W.Button, "A Simple crack, by causing it to deflect along a tortuous Way to Make Tough Ceramics, Natire (london), 347, 455-57(1990) where some of its energy is dissipated, hence imparting toughnes on and Failure of Laminar Ceramic Composites to the composite. Further work investigating such a mechanism of AM1,4013085-93(92 Y. Shigegaki, M. E. Brito, K. Hirao, M. Toriyama, and S. Kanzak intimate mixing of strength and toughness, on a microstructural of a Novel Multilayered Silicon Nitride, J. Am. Ceram. Soc., 79 [8] 2197-200 preliminary data-such as, for example, that illustrated in Figs. 8 Pat. n: 477 524, sept 20. 1988. tnc ceramic and menod for oucton, and 9(for alternating sequence of just one layer with a 12: 1 ratio, followed by layers with a 5: 1 ratio(Table IShow the Ceramics: 1, Fabrication, Microstructure, and Indentation Behavior, "J.Am. Ceram potential for such a system to exhibit sustained toughness and flaw Soc. 7619)2209-16(1993) Baskaran, S. D. Nunn, D. Popovic, and J. W. Halloran, "Fibrous Monolithic tolerance. as well as a significant If this mechanism is Ceramic cs: Il, Flexural Strength and Fracture Behavior of the Silicon Carbide/Graphite oupled with an intrinsically strong matrix (e.g 3Y-1ZP), the Systsm Bas aman. Cer m oH a6 9)22ib-234(Mo3) Additional work to investigate these speculations and hypothes and Oxidation Behavior of the Silicon Carbide/Boron Nitride System, "J. Am. Ceram. Soc,77[51249-55(1994 is currently underway. In addition, the high-temperature stability D Kovar, B H. King, R. W. Trice, and J. w. Halloran,"Fibrous Monolithic of the pores and the high-temperature mechanical properties will Ceramics, J. Am Cera Soc., 80[10] 2471-87(1997) be investigated and described in a future publication. P. E. D. Morgan and D. B. Marshall."Functional Interfaces for Oxide/Oxide Composites," Mater Sci Eng, A, A162, 15-25(Is P. E. D. Morgan, D. B. Marshall, and R, M. Housley, "High Temp Stability of Monazite-Alumina Composites, " Mater. Sci. Eng A, A195, 215-22 V. Conclusions P: E.D. Morgan and t. B, hssa erams composites of Monazite and A novel mechanism of interphase debonding in an all-oxide Soc. 78(9)2574(199 Kriven, "A Strong and Damage-Tolerant Oxide Laminate, composite system has been introduced. This concept is based on J. Am. Ceram Soc, 80(9)2421-24(1997). the engineering of a suitably weak interphase, through the use of D. H Kuo and W. M. Kriven, "Fracture of Multilayer Oxide Composites, "Mater. relatively unsinterable alumina platelets 10-15 um(or 5-10 um) Sci Eng- A, A241, 241-50(1998). in diameter and I um thick. The room-temperature strength of the interphase can be adjusted using minor additions of matrix powders(on the order of 1-3 vol%). Laminated composites of both t, C. B. Cart PA,1997 mullite and alumina have been fabricated. Modified fibrous 2w. M. Kriven and D H. Kuo, "High-Strength, Flaw-Tolerant, Oxide Cerami monoliths of alumina that consist of a triple layer "core/interphase/ Composite,'VS. Pat No. 5948516, Sept, 7, I of the processing and microstructural-design parameters has not 305-16(1998) Cristobalite Transformation Weakened Interphases, " Ceram. Eng. Sci. Proc., 19 [3 fully conducted, preliminary mechanical data and scanning elec- w. M. Kriven and S J. Lee, " Toughening of Mullite/Cordierite Laminates by on microscopic observation of the crack profiles nevertheless ansformation Weakening of B-Cristobalite Interphases, to be submitted to.Am. ram.Soc. demonstrate this procedure to be a viable high-temperature, 27W. M. Kriven, C M. Huang, D. Zhu, and Y. Xu, "Toughening of Titania by oxidation-resistant, toughening mechanism in chemically compat Transformation Weakening of Enstatite(MgSiO)Interphases, submitted to Acta ble oxide composites. In terms of chemistry, the platelets provid and effective mechanism for debonding in air, indepen- Fiber Coating Concepts for Brittle-Matrix Composites, J. Am. Ceram Soc., 76 [51 dent of temperature, up to the melting point of the alumina or the 12492. May, K Keller, T. A Parthasarathy, and J. Guth, "Fugitive Interface Coating matrix A mechanism that intimately mixes strength and toughness, on 922-30(1993) in Oxide-Oxide Composites: A Viability Study, " Ceram. Eng. Sci. Proc, 14 [9-101 a microstructural scale, through an optimally tailored bimodal C G. Levi, J. Y. Yang, B. J Dalgleish, F. W. Zok, and A G. Evans, "Processing microstructure has been introduced. It has been postulated that and Performance of an Al-Oxide Ceramic Composite,J.Am. Ceram Soc., 81 181 strength)could be alternated with regions of low matrix interphase of Composite Powders, "J. Mater Res, 2[1]59-65(1987). atio(high toughness), and these regions may consist of several uch thin layers. After the crack passes through a thick matrix Bsmm,时 yer, it is believed to have high kinetic energy. The role of the several layers with a low matrix interphase ratio is to slow the 34R.K. and G. W. Scherer, On Constrained Sintering-IIl: Rigid crack, by causing it to deflect along a tortuous interphase path, Inclusions, "Acta Metall. Mater, 36[912411-16(1988). 350. Sudre and F. F. Lange."Effect of Inclusions on Densification: 1. Microstruc- where some of its energy is dissipated and imparts toughness to the tural Development in an Al20, Matrix Containing a High Volume Fraction of ZrO2 composite Am. Ceram.Soe.,753l519-24(1992)
overall strength; however, the curve was characteristic of graceful failure and significant crack diversion along the interphase. This expectation was consistent with the SEM micrographs that are shown in Fig. 4: the platelet interphase with 5 vol% of powder exhibited brittle failure behavior (Fig. 4(a)), whereas the platelet interphase with additions of 3 vol% of 3Al2O3z2SiO2 produced noticeably more crack deflection along the interphase (Fig. 4(b)). Again, Fig. 9 demonstrates more-extensive crack deflection along the platelet interphase that contains only 3 vol% of 3Al2O3z2SiO2 powder. Following the concept of a bimodal microstructure, it is postulated that regions with one layer of high matrix:interphase ratio (high strength) should be alternated with regions of low matrix:interphase ratio (high toughness), which consist of several such thin layers—not just a couple of layers, as was fabricated here. The crack, after it has passed through a thick matrix layer, has been hypothesized to have high kinetic energy. The role of the several layers with a low matrix:interphase ratio is to slow the crack, by causing it to deflect along a tortuous interphase path, where some of its energy is dissipated, hence imparting toughness to the composite. Further work investigating such a mechanism of intimate mixing of strength and toughness, on a microstructural scale, through a tailored microstructure, is underway. However, preliminary data—such as, for example, that illustrated in Figs. 8 and 9 (for an alternating sequence of just one layer with a 12:1 ratio, followed by layers with a 5:1 ratio (Table III))—show the potential for such a system to exhibit sustained toughness and flaw tolerance, as well as a significant WOF. If this mechanism is coupled with an intrinsically strong matrix (e.g., 3Y-TZP), the absolute strength of the composite can be enhanced further. Additional work to investigate these speculations and hypotheses is currently underway. In addition, the high-temperature stability of the pores and the high-temperature mechanical properties will be investigated and described in a future publication. IV. Conclusions A novel mechanism of interphase debonding in an all-oxide composite system has been introduced. This concept is based on the engineering of a suitably weak interphase, through the use of relatively unsinterable alumina platelets 10–15 mm (or 5–10 mm) in diameter and 1 mm thick. The room-temperature strength of the interphase can be adjusted using minor additions of matrix powders (on the order of 1–3 vol%). Laminated composites of both mullite and alumina have been fabricated. Modified fibrous monoliths of alumina that consist of a triple layer “core/interphase/ matrix” arrangement have been designed. Although optimization of the processing and microstructural-design parameters has not fully conducted, preliminary mechanical data and scanning electron microscopic observation of the crack profiles nevertheless demonstrate this procedure to be a viable high-temperature, oxidation-resistant, toughening mechanism in chemically compatible oxide composites. In terms of chemistry, the platelets provide a simple and effective mechanism for debonding in air, independent of temperature, up to the melting point of the alumina or the matrix. A mechanism that intimately mixes strength and toughness, on a microstructural scale, through an optimally tailored bimodal microstructure has been introduced. It has been postulated that regions with one layer of high matrix:interphase ratio (high strength) could be alternated with regions of low matrix:interphase ratio (high toughness), and these regions may consist of several such thin layers. After the crack passes through a thick matrix layer, it is believed to have high kinetic energy. The role of the several layers with a low matrix:interphase ratio is to slow the crack, by causing it to deflect along a tortuous interphase path, where some of its energy is dissipated and imparts toughness to the composite. References 1 A. G. Evans, “Perspective on the Development of High-Toughness Ceramics,” J. Am. Ceram. Soc., 73 [2] 187–206 (1990). 2 D. B. Marshall, B. N. Cox, and A. G. Evans, “The Mechanics of Matrix Cracking in Brittle Fiber Composites,” Acta Metall., 33, 2013–21 (1985). 3 R. W. Davidge, “Fiber-Reinforced Ceramics,” Composites (Guildford, UK), 18 [2] 92–98 (1987). 4 M. Y. He and J. W. Hutchinson, “Crack Deflection at an Interface Between Dissimilar Elastic Materials,” Int. J. Solids Struct., 25 [9] 1053–67 (1989). 5 A. G. Evans, M. Y. He, and J. W. Hutchinson, “Interface Debonding and Fiber Cracking in Brittle Matrix Composites,” J. Am. Ceram. Soc., 72 [12] 2300–303 (1989). 6 J. W. Hutchinson and H. M. Jensen, “Models of Fiber Debonding and Pullout in Brittle Composites with Friction,” Mech. Mater., 9, 139–63 (1990). 7 R. J. Kerans, R. S. Hay, N. J. Pagano, and T. A. Parthasarathy, “The Role of the Fiber–Matrix Interface in Ceramic Composites,” Am. Ceram. Soc. Bull., 68 [2] 429–42 (1989). 8 A. G. Evans, “The Mechanical Performance of Fiber-Reinforced Ceramic Matrix Composites,” Mater. Sci. Eng., A, A107, 227–39 (1989). 9 A. G. Evans and F. W. Zok, “Review of the Physics and Mechanics of Fiber-Reinforced Brittle Matrix Composites,” J. Mater. Sci., 29, 3857–96 (1994). 10W. J. Clegg, K. Kendall, N. M. Alford, D. Birchall, and T. W. Button, “A Simple Way to Make Tough Ceramics,” Nature (London), 347, 455–57 (1990). 11W. J. Clegg, “The Fabrication and Failure of Laminar Ceramic Composites,” Acta Metall., 40 [11] 3085–93 (1992). 12Y. Shigegaki, M. E. Brito, K. Hirao, M. Toriyama, and S. Kanzaki, “Processing of a Novel Multilayered Silicon Nitride,” J. Am. Ceram. Soc., 79 [8] 2197–200 (1996). 13W. S. Coblenz, “Fibrous Monolithic Ceramic and Method for Production,” U.S. Pat. No. 4 772 524, Sept. 20, 1988. 14S. Baskaran, S. D. Nunn, D. Popovic, and J. W. Halloran, “Fibrous Monolithic Ceramics: I, Fabrication, Microstructure, and Indentation Behavior,” J. Am. Ceram. Soc., 76 [9] 2209–16 (1993). 15S. Baskaran, S. D. Nunn, D. Popovic, and J. W. Halloran, “Fibrous Monolithic Ceramics: II, Flexural Strength and Fracture Behavior of the Silicon Carbide/Graphite System,” J. Am. Ceram. Soc., 76 [9] 2217–24 (1993). 16S. Baskaran and J. W. Halloran, “Fibrous Monolithic Ceramics: III, Properties and Oxidation Behavior of the Silicon Carbide/Boron Nitride System,” J. Am. Ceram. Soc., 77 [5] 1249–55 (1994). 17D. Kovar, B. H. King, R. W. Trice, and J. W. Halloran, “Fibrous Monolithic Ceramics,” J. Am. Ceram. Soc., 80 [10] 2471–87 (1997). 18P. E. D. Morgan and D. B. Marshall, “Functional Interfaces for Oxide/Oxide Composites,” Mater. Sci. Eng., A, A162, 15–25 (1993). 19P. E. D. Morgan, D. B. Marshall, and R. M. Housley, “High Temperature Stability of Monazite–Alumina Composites,” Mater. Sci. Eng., A, A195, 215–22 (1995). 20P. E. D. Morgan and D. B. Marshall, “Ceramic Composites of Monazite and Alumina,” J. Am. Ceram. Soc., 78 [6] 1553–63 (1995) and erratum, J. Am. Ceram. Soc., 78 [9] 2574 (1995). 21D. H. Kuo and W. M. Kriven, “A Strong and Damage-Tolerant Oxide Laminate,” J. Am. Ceram. Soc., 80 [9] 2421–24 (1997). 22D. H. Kuo and W. M. Kriven, “Fracture of Multilayer Oxide Composites,” Mater. Sci. Eng., A, A241, 241–50 (1998). 23D. H. Kuo and W. M. Kriven, “Oxide Laminates with High Strength and Work-of-Fracture”; pp. 477–88 in Materials Research Society Symposium Proceedings, Vol. 458, Interfacial Engineering for Optimized Properties. Edited by C. L. Briant, C. B. Carter, and E. L. Hall. Materials Research Society, Pittsburgh, PA, 1997. 24W. M. Kriven and D. H. Kuo, “High-Strength, Flaw-Tolerant, Oxide Ceramic Composite,” U.S. Pat. No. 5 948 516, Sept. 7, 1999. 25W. M. Kriven and S. J. Lee, “Mullite/Cordierite Laminates with b 3 a Cristobalite Transformation Weakened Interphases,” Ceram. Eng. Sci. Proc., 19 [3] 305–16 (1998). 26W. M. Kriven and S. J. Lee, “Toughening of Mullite/Cordierite Laminates by Transformation Weakening of b-Cristobalite Interphases,” to be submitted to J. Am. Ceram. Soc. 27W. M. Kriven, C. M. Huang, D. Zhu, and Y. Xu, “Toughening of Titania by Transformation Weakening of Enstatite (MgSiO3) Interphases,” submitted to Acta Metall. Mater. 28J. B. Davis, J. P. A. Lo¨fvander, A. G. Evans, E. Bischoff, and M. L. Emiliani, “Fiber Coating Concepts for Brittle-Matrix Composites,” J. Am. Ceram. Soc., 76 [5] 1249–57 (1993). 29T. May, K. Keller, T. A. Parthasarathy, and J. Guth, “Fugitive Interface Coating in Oxide–Oxide Composites: A Viability Study,” Ceram. Eng. Sci. Proc., 14 [9–10] 922–30 (1993). 30C. G. Levi, J. Y. Yang, B. J. Dalgleish, F. W. Zok, and A. G. Evans, “Processing and Performance of an All-Oxide Ceramic Composite,” J. Am. Ceram. Soc., 81 [8] 2077–86 (1998). 31F. F. Lange, “Constrained Network Model for Predicting Densification Behavior of Composite Powders,” J. Mater. Res., 2 [1] 59–65 (1987). 32R. K. Bordia and G. W. Scherer, “On Constrained Sintering—I: Constitutive Model for a Sintering Body,” Acta Metall. Mater., 36 [9] 2393–97 (1988). 33R. K. Bordia and G. W. Scherer, “On Constrained Sintering—II: Comparison of Constitutive Models,” Acta Metall. Mater., 36 [9] 2399–407 (1988). 34R. K. Bordia and G. W. Scherer, “On Constrained Sintering—III: Rigid Inclusions,” Acta Metall. Mater., 36 [9] 2411–16 (1988). 35O. Sudre and F. F. Lange, “Effect of Inclusions on Densification: I, Microstructural Development in an Al2O3 Matrix Containing a High Volume Fraction of ZrO2 Inclusions,” J. Am. Ceram. Soc., 75 [3] 519–24 (1992). April 2001 Toughened Oxide Composites Based on Porous Alumina-Platelet Interphases 773
74 Journal of the American Ceramic Society-Lee and Kriven Vol 84. No 4 bO. Sudre, G, Bao, B. Fran, F. F. Lange, and A. G. Evans, "Effect of Inclusions oS. Prochazka and F J. Klug, "Infrared-Transparent Mullite Ceramic,"J.Am. Densification: Il, Numerical Model, J. Am. Cera. Soc., 75 [3] 525-31 Ceran.Soc,66[12187480(1983) F J. Klug, S. Prochazka, and R H. Doremus, "Alumina-Silica Phase Diagram in "l. K. Cherian, M. D. Lehigh, 1. Nettleship, and W. M. Kriven, "Stereological the Mullite Region,"J. Am. Ceram. Soc., 70 [10]750-59(1987) Sommers, A. Barda, an J. In. Ceran. Soc,7923273-81(1996 G. Zywicki, "Process for Preparing Textured Ceramic Composites, "U.S. Pat. Ne tructure and Mechanical Properties of Alumina 5645781,July8,1997 Platelet Reinforced 3Y-TZP and Mullite Composites", Ph D W. M. Riven, 4H. G. Tattersall and G. Tappin, "The Work of Fracture and Its Measurement in ". K. Cherian and W. M. Kriven, "Suspension Processing, Microstructure an Metals, Ceramics and Other Materials, "J. Mater. Sci, 1, 296-301(1966). C M. Huang, Y. Xu, D. Zhu, and w. M. Kriven,"Combustion Synthesized Mechanical Properties of Alumina Platelet Reinforced 3Y-TZP, submitted to Am B'-SiAION Composites Reinforced with SiC Monofilaments, Mater. Sci. Eng, A, 4188,341-51(1994)
36O. Sudre, G. Bao, B. Fran, F. F. Lange, and A. G. Evans, “Effect of Inclusions on Densification: II, Numerical Model,” J. Am. Ceram. Soc., 75 [3] 525–31 (1992). 37I. K. Cherian, M. D. Lehigh, I. Nettleship, and W. M. Kriven, “Stereological Observations of Platelet-Reinforced Mullite- and Zirconia-Matrix Composites,” J. Am. Ceram. Soc., 79 [12] 3273–81 (1996). 38I. K. Cherian, “Processing, Microstructure and Mechanical Properties of Alumina Platelet Reinforced 3Y-TZP and Mullite Composites”; Ph.D. Thesis. (W. M. Kriven, advisor.) University of Illinois at Urbana–Champaign, Urbana, IL, 1995. 39I. K. Cherian and W. M. Kriven, “Suspension Processing, Microstructure and Mechanical Properties of Alumina Platelet Reinforced 3Y-TZP,” submitted to Am. Ceram. Soc. Bull. 40S. Prochazka and F. J. Klug, “Infrared-Transparent Mullite Ceramic,” J. Am. Ceram. Soc., 66 [12] 874–80 (1983). 41F. J. Klug, S. Prochazka, and R. H. Doremus, “Alumina–Silica Phase Diagram in the Mullite Region,” J. Am. Ceram. Soc., 70 [10] 750–59 (1987). 42D. Popovic, W. Halloran, G. E. Hilmas, G. A. Brady, S. Sommers, A. Barda, and G. Zywicki, “Process for Preparing Textured Ceramic Composites,” U.S. Pat. No. 5 645 781, July 8, 1997. 43H. G. Tattersall and G. Tappin, “The Work of Fracture and Its Measurement in Metals, Ceramics and Other Materials,” J. Mater. Sci., 1, 296–301 (1966). 44C. M. Huang, Y. Xu, D. Zhu, and W. M. Kriven, “Combustion Synthesized b9-SiAlON Composites Reinforced with SiC Monofilaments,” Mater. Sci. Eng., A, A188, 341–51 (1994). M 774 Journal of the American Ceramic Society—Lee and Kriven Vol. 84, No. 4