D-K. Kim, W.M. Kriven/Materials Science and Engineering 4 380 (2004)237-244 3. Results and discussion Inner mullite layer SEM micrographs of the mullite and the 50 vol. alu mina: 50 vol. YAG in situ composite matrix powders are seen in Fig. 2. The mullite powder had a granular morphol- ogy, with an average particle size and surface area of 0.8 um and 8.5m-/g, respectively. The in situ alumina-YAG com- posite powder showed an irregular shape with sharp edges and it had a specific surface area of 2.7 m/g after I h attri- tion milling. Alumina platelets had a hexagonal plate shape Outer AlPo. layer of 5-10 um diameter, and I um thickness with a specific surface area of 1. 4m/g(Fig 3(a). An alumina platelet pel let CIPped at 413.7 MPa and then sintered at 1600C 20125130135140145150155160 3h, had a density of 80% of theoretical density and a 3-point bend strength of 205 7 MPa. This bending strength was Temperature (C) much higher than 1.5+0.2 MPa of AlPO4 [30]. This means Fig. 4. The variation of mixing torques with temperature for the mullite that alumina platelets can possibly function as a stronger inner rod and AlPO4 interphase layer. debonding material, resulting in a higher strength and tough- ness composite material. AlPO4 powder had an irregular shape with sharp edges(Fig. 3(b)). The I h attrition-milled AlPO4 powder had a particle size and specific surface area Connecticut, USA). The temperature ranged of 0.9 um and 87 m2/g, respectively emperature to 900C, and the heating rate The variations of the mixing torque for the mullite in- was Flexural strengths were measured with a ner matrix rod and AlPO4 interphase layer are plotted in testing machine(Model 4502, Instron Corp, Fig 4. To prevent intermixing between the inner matrix layer Canton, MA) in 3-point bending. The samples were cut and interphase layer during extrusion and, finally, to obtain from sintered pellets where the length of the testing sample a fibrous monolithic texture after extrusion, the viscosity was aligned with the longitudinal direction of the filaments. of the inner layer should be higher than that of the inter- The samples were ground with diamond paste and finally phase region. Fig. 4 shows that the inner layer had about polished with 600 grit SiC paper, after which the four edges a five-times higher mixing torque than did the aluminum of each sample were bevelled. The flexural strengths were phosphate, interphase layer. This means that the viscosity determined after testing 3-5 samples The supporting span requirement to obtain fibrous monolithic composite texture 30mm, the cross head speed was 0. I mm/min and was well satisfied for the formulations in Table 1. The mixing sample size was3mm(H)×4mm(W)×40mm(L torque decreased with increasing temperature. The binder Temperature ( C) Fig. 5. The binder burnout TGa profile from room temperature to 900C, for the mullite-AlPO4 two-layer, fibrous monolithic composite240 D.-K. Kim, W.M. Kriven / Materials Science and Engineering A 380 (2004) 237–244 Fig. 4. The variation of mixing torques with temperature for the mullite inner rod and AlPO4 interphase layer. Perkin Elmer, Connecticut, USA). The temperature ranged from room temperature to 900 ◦C, and the heating rate was 5 ◦C/min. Flexural strengths were measured with a screw-driven testing machine (Model 4502, Instron Corp., Canton, MA) in 3-point bending. The samples were cut from sintered pellets where the length of the testing sample was aligned with the longitudinal direction of the filaments. The samples were ground with diamond paste and finally polished with 600 grit SiC paper, after which the four edges of each sample were bevelled. The flexural strengths were determined after testing 3–5 samples The supporting span was 30 mm, the cross head speed was 0.1 mm/min and sample size was 3 mm (H) × 4 mm (W) × 40 mm (L). Fig. 5. The binder burnout TGA profile from room temperature to 900 ◦C, for the mullite-AlPO4 two-layer, fibrous monolithic composite. 3. Results and discussion SEM micrographs of the mullite and the 50 vol.% alu￾mina:50 vol.% YAG in situ composite matrix powders are seen in Fig. 2. The mullite powder had a granular morphol￾ogy, with an average particle size and surface area of 0.8
m and 8.5 m2/g, respectively. The in situ alumina-YAG com￾posite powder showed an irregular shape with sharp edges and it had a specific surface area of 2.7 m2/g after 1 h attri￾tion milling. Alumina platelets had a hexagonal plate shape of 5–10
m diameter, and 1
m thickness with a specific surface area of 1.4 m2/g (Fig. 3(a)). An alumina platelet pel￾let CIPped at 413.7 MPa and then sintered at 1600 ◦C for 3 h, had a density of 80% of theoretical density and a 3-point bend strength of 205 ± 7 MPa. This bending strength was much higher than 1.5 ± 0.2 MPa of AlPO4 [30]. This means that alumina platelets can possibly function as a stronger debonding material, resulting in a higher strength and tough￾ness composite material. AlPO4 powder had an irregular shape with sharp edges (Fig. 3(b)). The 1 h attrition-milled AlPO4 powder had a particle size and specific surface area of 0.9
m and 87 m2/g, respectively. The variations of the mixing torque for the mullite in￾ner matrix rod and AlPO4 interphase layer are plotted in Fig. 4. To prevent intermixing between the inner matrix layer and interphase layer during extrusion and, finally, to obtain a fibrous monolithic texture after extrusion, the viscosity of the inner layer should be higher than that of the inter￾phase region. Fig. 4 shows that the inner layer had about a five-times higher mixing torque than did the aluminum phosphate, interphase layer. This means that the viscosity requirement to obtain fibrous monolithic composite texture was well satisfied for the formulations in Table 1. The mixing torque decreased with increasing temperature. The binder