M.L. Antti et al. Journal of the European Ceramic Society 24(2004)565-578 parallel which is evidence of the constraint imposed on matrix shrinkage mentioned above since such constraint would lead to a tensile stress parallel to the fibres 45. XRD and Raman spectroscopy X-ray diffraction of the as-received material revealed corundum and mullite peaks and a broad background peak indicative of a glassy phase. The sample exposed 1100oC for 100 h. still exhibited mullite and corund peaks but the background peak had disappeared and peaks corresponding to cristobalite had appeared. It was also observed that the( 120)and(210)mullite peaks had shifted somewhat. These observations suggest that the silica in the matrix, originally in the amorphous Fig. 12. Fracture surfaces of samples with +45 fibre orientation after was either already partly segregated or segregated durin 100 h heat-treatment at 1 100oC thermal exposure and then crystallised to cristobalite Density, open porosity, average pore area and hardness values of the matrix as a function of heat-treatment time and temperature Sample Density [g/cm1 Average hardness[Hv Porosity [vol%] Pores per unit area ±45° ±45° Asreceive 21.8 204(22) 11.0 100h at I 213(54) 3240 h at 323(22) 20 h at 1100°C 334(33) l00hatl100°2.6 22.8 21.8 457(116) 8.1 13.6 Standard deviations in parentheses. Average taken of 10 indents Fig 13. Inhomogeneous matrix, after heat-treatment 100 h at 1 100C. Fig 14. Indentation cracking after heat-treatment 20 h at 1100C.LT LT surface (SEM, BED) surface (SEM, BED).parallel which is evidence of the constraint imposed on matrix shrinkage mentioned above since such constraint would lead to a tensile stress parallel to the fibres. 4.5. XRD and Raman spectroscopy X-ray diffraction of the as-received material revealed corundum and mullite peaks and a broad background peak indicative of a glassy phase. The sample exposed at 1100
C for 100 h, still exhibited mullite and corundum peaks but the background peak had disappeared and peaks corresponding to cristobalite had appeared. It was also observed that the (120) and (210) mullite peaks had shifted somewhat. These observations suggest that the silica in the matrix, originally in the amorphous state, was either already partly segregated or segregated during thermal exposure and then crystallised to cristobalite. Table 4 Density, open porosity, average pore area and hardness values of the matrix as a function of heat-treatment time and temperature Sample Density [g/cm3 ] Open porosity [%] Average hardnessa [HV] Porosity [vol%] Pores per unit area 0/90
45
0/90
45
As-received 2.64 2.68 21.8 20.9 204 (22) 4.3 11.0 100 h at 1000
C 2.60 2.67 23.2 20.9 213 (54) – – 3240 h at 1000
C 2.64 2.67 22.7 20.7 323 (22) 5.3 14.0 20 h at 1100
C 2.61 – 22.3 – 334 (33) 6.6 16.4 100 h at 1100
C 2.64 2.62 22.8 21.8 457 (116) 8.1 13.6 a Standard deviations in parentheses. Average taken of 10 indents. Fig. 12. Fracture surfaces of samples with 45
fibre orientation after 100 h heat-treatment at 1100
C. Fig. 13. Inhomogeneous matrix, after heat-treatment 100 h at 1100
C. LT surface. (SEM, BEI). Fig. 14. Indentation cracking after heat-treatment 20 h at 1100
C. LT surface (SEM, BEI). M.-L. Antti et al. / Journal of the European Ceramic Society 24 (2004) 565–578 575