W. Dressler, R. Riedel observed the chemistry of the grain boundary Experimental results. 2 showed that Si,N phase to be an important parameter in the field ceramics having exclusively grains with aspect of toughening of SiaNa-ceramics. Crack propa ratios below 4 reveal fracture toughness values gation experiments in p-Si3NA-whisker doped of about 5.5 MPa m /. This led to the concl aluminium-yttrium oxynitride glasses with sion that the grain fraction with an aspect ratio fixed nitrogen content revealed that the deb- smaller than 4 can be regarded as the matrix onding length and, hence, the toughening and grains having an aspect ratio higher than 4 response of the whiskers decreases with risin can be considered as the reinforcing particles. Al2 O3 content due to an increase in interface By means of the above described quantitative energy. This experiment points out that fracture microstructural analysis method the volume toughness tailoring requires both the micro- weighted diameter of grains(2, Vr Dmin )has structure design and control of grain boundary been calculated from the measured grain size chemistry distributions According to Irwin the strength of ceramic In Fig. 7 the dependence of fracture tough- materials(oB) showing no plastic deformation ness on the volume weighted diameter of elon- can be correlated with the materials fracture gated grains(aspect ratio >4)is shown for toughness(Kle), a geometric factor(Y) and the Si,n4 ceramics densified by 10.7 wt% defec Y, O, and 3 6 wt% Al,O3 as sintering aids This plot shows that the fracture toughness grain diameter rises as described by eqn(6)and that fracture toughness values of 10-3 MPa'm can be achieved. It was concluded that the This reveals that the strength of brittle ceramics pull-out model fits in principle the experimental is not constant but depends on the defect size data. The deviations are explained by the influ- distribution incorporated in the ceramics due to ence of the neglected crack deflection which has processing defects and microstructure features been calculated from Bengisu et al. 4 to be It has been shown that the failure probability about 2 MPa m/2 even in composite materials (P)of brittle ceramics can be described by the having a low difference in the elastic properties two parametric Weibull-distribution. 00 Io1 between the matrix and the reinforcing phase Thus, very high strengthened Si,Na ceramics can be produced by grain coarsening due to p long heat treatments at high temperatures additionally, Becher et al. 95 and others b,, 96-9K Here(m)is the Weibull modulus and (oo) is a parameter. By plotting InIn [1/(1-P)] vs(oB) 200 straight line having the slope (m) results Therefore, the Weibull parameter (m)describes the strength variation of the particular material In Fig. 8 the four point bending strength di tributions of gas pressure(10 MPa N2 pressure) sintered Al2O/Y2O3(10 vol%) containing Si, -ceramics with fine and coarse microstruc tures derived from different starting powders are shown. i In the case of the material derived from UBE SN-E10 starting powder(UBE E-10 0,6 0.8 UBE Industries, Japan) the coarsening of the Dmin(a>4) um microstructure leads to a decrease of the maxi- mum and mean strength o an ig.7. Relation between fracture toughness and volume increase of the Weibull modulus from 13. 5 to weighted diameter of elongated(aspect ratio >4) grains for Si, N. ceramics densified by liquid phase sinteri 46, which means that the reliability of the (107wt%Y2O3+36wt%Al2O3)112 material can be improved by high temperature22 W. Dressier, R. Riedel Experimental results TM showed that Si3N4 - ceramics having exclusively grains with aspect ratios below 4 reveal fracture toughness values of about 5.5 MPa'm ~/2. This led to the conclu￾sion that the grain fraction with an aspect ratio smaller than 4 can be regarded as the matrix and grains having an aspect ratio higher than 4 can be considered as the reinforcing particles. By means of the above described quantitative microstructural analysis method the volume weighted diameter of grains (Z,~'DL~,) has been calculated from the measured grain size distributions. In Fig. 7 the dependence of fracture tough￾ness on the volume weighted diameter of elon￾gated grains (aspect ratio >4) is shown for Si3N4 ceramics densified by using 10.7wt% Y20~ and 3.6wt% A1203 as sintering aids. ''''2 This plot shows that the fracture toughness increases by square if the volume weighted grain diameter rises as described by eqn (6) and that fracture toughness values of 10.3 MPa. m '/2 can be achieved. It was concluded" that the pull-out model fits in principle the experimental data. The deviations are explained by the influ￾ence of the neglected crack deflection which has been calculated from Bengisu et al. 94 to be about 2 MPa.m '/2 even in composite materials having a low difference in the elastic properties between the matrix and the reinforcing phase. Thus, very high strengthened Si3N4 ceramics can be produced by grain coarsening due to long heat treatments at high temperatures. Additionally, Becher et al. 95 and others 8"'~6-'~" 200 E 150 -L, n 100 50 Fig. 7. 00 I , i ' 0,2 0,4 0 6 0,8 Dmin(a>4) [pro] Relation between fracture toughness and volume weighted diameter of elongated (aspect ratio >4) grains for Si,N4-ceramics densified by liquid phase sintering (10"7 wt% Y203 +3"6 wt% A1203).' i, J2 observed the chemistry of the grain boundary phase to be an important parameter in the field of toughening of Si3N4-ceramics. Crack propa￾gation experiments "5 in fl-Si3Nn-whisker doped aluminium-yttrium oxynitride glasses with a fixed nitrogen content revealed that the deb￾onding length and, hence, the toughening response of the whiskers decreases with rising A1203 content due to an increase in interface energy. This experiment points out that fracture toughness tailoring requires both the micro￾structure design and control of grain boundary chemistry. According to Irwin ~° the strength of ceramic materials (o-,3) showing no plastic deformation can be correlated with the materials fracture toughness (K,c), a geometric factor (Y) and the defect size (a): glc a,= (7) Y" £d:a This reveals that the strength of brittle ceramics is not constant but depends on the defect size distribution incorporated in the ceramics due to processing defects and microstructure features. It has been shown that the failure probability (P) of brittle ceramics can be described by the two parametric Weibull-distribution.'"" '"' E P=l-exp - -- . (8) \(70/ J Here (m) is the Weibull modulus and (a,,) is a parameter. By plotting lnln [1/(I-P)] vs (aB) a straight line having the slope (m) results. Therefore, the Weibull parameter (m) describes the strength variation of the particular material. In Fig. 8 the four point bending strength dis￾tributions of gas pressure (10 MPa N2 pressure) sintered AI203//Y203 (10 vol%) containing Si3Na-ceramics with fine and coarse microstruc￾tures derived from different starting powders are shown." In the case of the material derived from UBE SN-E10 starting powder (UBE E-10: UBE Industries, Japan) the coarsening of the microstructure leads to a decrease of the maxi￾mum and mean strength as well as to an increase of the Weibull modulus from 13.5 to 46, which means that the reliability of the material can be improved by high temperature