K. Shimoda et aL/Composites Science and Technology 69(2009)1623-1628 The objective of this article is to provide the results of recent matrix formation. Characterizations of Sic nano-powder employed ctivities in our group at Kyoto University, aimed at developing ad- in this study was described elsewhere [22]. For the preparation of vanced Sic/Sic composites with tailoring of mechanical perfor- prepreg sheets, unidirectional Py C-coated Tyranno-SA fibers were mances by NITE process. Fig. 1 shows two concepts on typical impregnated in Sic 'nano-slurry, which consists of the tensile fracture behavior for CFCCs [1, 18]. A primary need in their of B-Sic na der and sintering additives(Al203+Y as-producted condition is to display as high a proportional limit wt%(Al203: Y203=60: 40)and Sio2=3 wt%)in isopropyl stress(PLS)as possible. The high PLS value will allow the materials The prepreg sheets were dried at room-temperature and cut into to carry high combinations of mechanical and thermal tensile 40 mm x 40 mm square. Prepared prepreg sheets were unidire stress without cracking, and be considered as the useful index of tionally stacked in a graphite die, and then hot-pressed at 1800- laterial/component design base on elastic mechanical behavior 1900C for 1 h in Ar atmosphere under the pressure of 20 MPa a high stress. And thus, the composites, as shown in Fig. 1a, with two kinds of fiber volume fraction. One is low fiber volume are potentially considered as suitable for fusion blankets, heat fraction (low-VA, appropriately 30 vol%. Another is high fiber vol- exchangers and turbine disks. However, during service time, unex- ume fraction(high-V) above 50 vol% pectedly higher stresses may arise that can locally crack the ma- trix, causing immediate material failure if the composites do not 2.2. Characterization of fabricated Sic/sic composites display a large ductile fracture behavior with a high ultimate ten- sile strength(UTS). The composites, as shown in Fig. 1b, are poten The bulk density and open porosity of fabricated composites were measured by the Archimedes'method, using distilled water protection tiles and after burner flaps. Although several approaches as the immersion medium. Theoretical density of fabricated com- have been taken towards two concepts, interface bonding between posites was calculated using the rule of mixtures, which consist fiber and matrix is frequently focused as a technical key because of Sic matrix with sintering additives and Py c-coated Tyranno ad transfer from the matrix to the fibers and vice versa occurs SA fiber. Fabricated composites were subsequently cut into through the interface [19-21. The high strength type requires effi- 3.0 mm x 1.5 mm x 40 mm for monotonic tensile test and cient load transfers from matrix to fibers which are obtained with 3.0 mm x 1.5 mm x 25 mm for three-point bending test, with par strong interface On the other hand, the high ductility type requires allel direction to fiber axis Monotonic tensile test was carried out high fracture toughness contributed by crack bridging and fiber at room-temperature in an INSTRoN 5581 test machine with a pull-out which is obtained with weak interface. For two concepts, crosshead displacement rate of 0.5 mm/min. On both edges of the effects of fiber volume fraction incorporating fabrication tem he tensile bars, the aluminum tabs were affixed to each side using perature were investigated on densification, microstructure and a kind of standard Araldite binder. The gauge length was desig- mechanical properties. Furthermore, tensile fracture behaviors nated to be 20 mm. Tensile strain was recorded from the exten- were discussed based on the interface bonding someter fixed on both sides of the gauge areas. For comparison, three-point bending test was also performed at room-temperature 2. Experimental procedure in the same test machine, with a crosshead displacement rate of 0.5 mm/min and outer support span of 18 mm. Each test number 2.1. NTE process for Si was at least three specimens. Both the polished cross-section and the tensile fracture surface were observed by field emission elec- Pyrolytic carbon(PyC)-coated Tyranno-SArd SiC fibers(Ube tron microscopy (FE-SEM). for SiC/SiC fabrication. Our previous study revealed that the thick- 3. Results and discussion ness of induced-Pyc interface strongly affected the density al microstructure, and therefore dominated fracture behaviors [17]. 3.1. Densification process In the present study, Py C coating was appropriately chosen at the thickness of 0.5 um through chemical vapor deposition(CvD)pro- Before studying the densification process, it is necessary to cessB-SiC nano-powder(Sumitomo Osaka Cement Co Ltd, Japan, investigate densification of the matrix itself without fibers. The T-1 grade) and sintering additives with Al2O3( Kojundo Chemical classic liquid phase sintering densifies in three overlapping stages Laboratory Co. Ltd Japan, mean diameter of 0.3 um. 99.99% pure),(1)rearrangement stage. ( 2)solution-reprecipitation stage and (3) Y203(Kojundo Chemical Laboratory Co Ltd, Japan, mean diameter solid skeleton stage. Fig. 2 shows the effects of the fabrication tem- of 1.0 um, 99.99% pure)and SiO2( Kojundo Chemical Laboratory Co. perature on the densification of monolithic SiC and SiC/Sic com- Ltd, Japan, mean diameter of 1.0 um, 99.9% up pure) were used for posites. For monolithic Sic corresponding to the matrix in NITE b Hah。 via strong fiber bonding Maximum use of 9 bundled fiber strengt High PLS wE。 Larger strain LOw PLS Strain Strain Fig. 1. Concept on tensile fracture behavior for continuous fiber-reinforced ceramic composites: (a) high strength type and(b) high ductility type.The objective of this article is to provide the results of recent activities in our group at Kyoto University, aimed at developing ad￾vanced SiC/SiC composites with tailoring of mechanical perfor￾mances by NITE process. Fig. 1 shows two concepts on typical tensile fracture behavior for CFCCs [1,18]. A primary need in their as-producted condition is to display as high a proportional limit stress (PLS) as possible. The high PLS value will allow the materials to carry high combinations of mechanical and thermal tensile stress without cracking, and be considered as the useful index of material/component design base on elastic mechanical behavior to a high stress. And thus, the composites, as shown in Fig. 1a, are potentially considered as suitable for fusion blankets, heat exchangers and turbine disks. However, during service time, unex￾pectedly higher stresses may arise that can locally crack the ma￾trix, causing immediate material failure if the composites do not display a large ductile fracture behavior with a high ultimate ten￾sile strength (UTS). The composites, as shown in Fig. 1b, are poten￾tially considered as suitable for combustor liners, thermal protection tiles and after burner flaps. Although several approaches have been taken towards two concepts, interface bonding between fiber and matrix is frequently focused as a technical key because load transfer from the matrix to the fibers and vice versa occurs through the interface [19–21]. The high strength type requires effi- cient load transfers from matrix to fibers which are obtained with strong interface. On the other hand, the high ductility type requires high fracture toughness contributed by crack bridging and fiber pull-out which is obtained with weak interface. For two concepts, the effects of fiber volume fraction incorporating fabrication tem￾perature were investigated on densification, microstructure and mechanical properties. Furthermore, tensile fracture behaviors were discussed based on the interface bonding. 2. Experimental procedure 2.1. NITE process for SiC/SiC composites Pyrolytic carbon (PyC)-coated TyrannoTM-SA3rd SiC fibers (Ube Industrials Ltd., Yamaguchi, Japan) were used as reinforcement for SiC/SiC fabrication. Our previous study revealed that the thick￾ness of induced-PyC interface strongly affected the density and microstructure, and therefore dominated fracture behaviors [17]. In the present study, PyC coating was appropriately chosen at the thickness of 0.5 lm through chemical vapor deposition (CVD) pro￾cess. b-SiC nano-powder (Sumitomo Osaka Cement Co. Ltd., Japan, T-1 grade) and sintering additives with Al2O3 (Kojundo Chemical Laboratory Co. Ltd. Japan, mean diameter of 0.3 lm, 99.99% pure), Y2O3 (Kojundo Chemical Laboratory Co. Ltd., Japan, mean diameter of 1.0 lm, 99.99% pure) and SiO2 (Kojundo Chemical Laboratory Co. Ltd., Japan, mean diameter of 1.0 lm, 99.9% up pure) were used for matrix formation. Characterizations of SiC nano-powder employed in this study was described elsewhere [22]. For the preparation of prepreg sheets, unidirectional PyC-coated Tyranno-SA fibers were impregnated in SiC ‘nano’-slurry, which consists of the mixture of b-SiC nano-powder and sintering additives (Al2O3 + Y2O3 = 9 wt% (Al2O3:Y2O3 = 60:40) and SiO2 = 3 wt%) in isopropyl alcohol. The prepreg sheets were dried at room-temperature and cut into 40 mm 40 mm square. Prepared prepreg sheets were unidirec￾tionally stacked in a graphite die, and then hot-pressed at 1800– 1900 C for 1 h in Ar atmosphere under the pressure of 20 MPa with two kinds of fiber volume fraction. One is low fiber volume fraction (low-Vf), appropriately 30 vol%. Another is high fiber vol￾ume fraction (high-Vf) above 50 vol%. 2.2. Characterization of fabricated SiC/SiC composites The bulk density and open porosity of fabricated composites were measured by the Archimedes’ method, using distilled water as the immersion medium. Theoretical density of fabricated com￾posites was calculated using the rule of mixtures, which consist of SiC matrix with sintering additives and PyC-coated TyrannoTM￾SA fiber. Fabricated composites were subsequently cut into 3.0 mm 1.5 mm 40 mm for monotonic tensile test and 3.0 mm 1.5 mm 25 mm for three-point bending test, with par￾allel direction to fiber axis. Monotonic tensile test was carried out at room-temperature in an INSTRON 5581 test machine with a crosshead displacement rate of 0.5 mm/min. On both edges of the tensile bars, the aluminum tabs were affixed to each side using a kind of standard Araldite binder. The gauge length was desig￾nated to be 20 mm. Tensile strain was recorded from the exten￾someter fixed on both sides of the gauge areas. For comparison, three-point bending test was also performed at room-temperature in the same test machine, with a crosshead displacement rate of 0.5 mm/min and outer support span of 18 mm. Each test number was at least three specimens. Both the polished cross-section and the tensile fracture surface were observed by field emission elec￾tron microscopy (FE-SEM). 3. Results and discussion 3.1. Densification process Before studying the densification process, it is necessary to investigate densification of the matrix itself without fibers. The classic liquid phase sintering densifies in three overlapping stages: (1) rearrangement stage, (2) solution-reprecipitation stage and (3) solid skeleton stage. Fig. 2 shows the effects of the fabrication tem￾perature on the densification of monolithic SiC and SiC/SiC com￾posites. For monolithic SiC corresponding to the matrix in NITE Fig. 1. Concept on tensile fracture behavior for continuous fiber-reinforced ceramic composites: (a) high strength type and (b) high ductility type. 1624 K. Shimoda et al. / Composites Science and Technology 69 (2009) 1623–1628