Y-H. Koh et al. Journal of the European Ceramic Society 24(2004)699-703 且 三苏 BN Concentration in Slurry Iwt %l Cell Boundary Thickness lumI Fig. 2. Cell boundary thickness of the specimens as a function of BN Fig. 4. Flexural strength of the specimens as a function of cell concentration in the slurry. boundary thickness For structural ceramic applications, energy dissipa- tion capacity during fracture is one of the most impor 700(B)Thickness=18 um tant criteria. The energy dissipated by the sample during (C)Thickness=37 the non-catastrophic failure is estimated by the appar ent work-of-fracture, which corresponds to the area under the load-deflection curve after the first load drop (inelastic region). Despite the decrease in strength, the ∽ work-of-fracture increased markedly for fibrous mono- liths, as shown in Fig. 4. When the cell boundary was very thin(10 um), the specimen fractured in a brittle manner. However, with increasing the cell boundary thickness. the work-of-fracture also increased. max imum work-of-fracture in the present research was observed for the specimen with a boundary thickness of 37 um. When the cell boundary was thicker than this, the work-of-fracture decreased sharply, as shown in Fig. 5. These results suggest that the work-of-fracture is Fig 3. Flexural response of specimens with cell boundary thicknesses influenced by the strength of the material as well as by of(A)0(B)18. and (C)37 um. the crack propagation pattern. When the strength is too high, i.e. when too much strain energy is stored in the material, the crack propagates through the cells or cell The maximum stress recorded in Fig 3 corresponds boundaries without any significant crack interactions to the flexural strength of the material. The strength of On the other hand when the strength is too low, the fibrous monoliths with respect to the cell boundary work-of-fracture is low because the material can with thickness is shown in Fig 4. As expected, the strength of stand only limited stress fibrous monoliths was lower than that of monolithic The influence of cell boundary thickness on the crack Si3N4, and decreased steadily with increasing cell pattern is demonstrated in the sEM micrographs, in boundary thickness. This decrease in flexural strength Fig. 6. The fracture pattern of the fibrous monolithic was due to the increase in flaw size which might be pre- specimen with boundary thickness of 18 um is shown in sent in cell boundaries. However, the strengths were not Fig. 6(A). Unlike the monolithic materials, extensive very sensitive to the boundary thickness except for o-um crack deflections occurred at the cell boundaries. When cell boundary thickness (i.e. monolithic Si3 N4), pre- the cell boundary thickness was increased to 37 um, not sumably because the size of the critical flaws responsible only the crack deflections but also extensive crack dela for fracture was not strongly dependent on the cell minations occu rred, as illustrated in Fig. 6(B). The boundary thickness. improved work-of-fracture with increasing the cellThe maximum stress recorded in Fig. 3 corresponds to the flexural strength of the material. The strength of fibrous monoliths with respect to the cell boundary thickness is shown in Fig. 4. As expected, the strength of fibrous monoliths was lower than that of monolithic Si3N4, and decreased steadily with increasing cell boundary thickness. This decrease in flexural strength was due to the increase in flaw size which might be pre￾sent in cell boundaries. However, the strengths were not very sensitive to the boundary thickness except for 0-mm cell boundary thickness (i.e. monolithic Si3N4), pre￾sumably because the size of the critical flaws responsible for fracture was not strongly dependent on the cell boundary thickness. For structural ceramic applications, energy dissipa￾tion capacity during fracture is one of the most impor￾tant criteria. The energy dissipated by the sample during the non-catastrophic failure is estimated by the appar￾ent work-of-fracture, which corresponds to the area under the load–deflection curve after the first load drop (inelastic region). Despite the decrease in strength, the work-of-fracture increased markedly for fibrous mono￾liths, as shown in Fig. 4. When the cell boundary was very thin (10 mm), the specimen fractured in a brittle manner. However, with increasing the cell boundary thickness, the work-of-fracture also increased. Max￾imum work-of-fracture in the present research was observed for the specimen with a boundary thickness of 37 mm. When the cell boundary was thicker than this, the work-of-fracture decreased sharply, as shown in Fig. 5. These results suggest that the work-of-fracture is influenced by the strength of the material as well as by the crack propagation pattern. When the strength is too high, i.e. when too much strain energy is stored in the material, the crack propagates through the cells or cell boundaries without any significant crack interactions. On the other hand, when the strength is too low, the work-of-fracture is low because the material can with￾stand only limited stress. The influence of cell boundary thickness on the crack pattern is demonstrated in the SEM micrographs, in Fig. 6. The fracture pattern of the fibrous monolithic specimen with boundary thickness of 18 mm is shown in Fig. 6(A). Unlike the monolithic materials, extensive crack deflections occurred at the cell boundaries. When the cell boundary thickness was increased to 37 mm, not only the crack deflections but also extensive crack dela￾minations occurred, as illustrated in Fig. 6(B). The improved work-of-fracture with increasing the cell Fig. 3. Flexural response of specimens with cell boundary thicknesses of (A) 0, (B) 18, and (C) 37 mm. Fig. 2. Cell boundary thickness of the specimens as a function of BN concentration in the slurry. Fig. 4. Flexural strength of the specimens as a function of cell boundary thickness. Y.-H. Koh et al. / Journal of the European Ceramic Society 24 (2004) 699–703 701