Journal of the American Ceramic Society--Bitterlich and Heinrich Vol. 88. No. 10 black area= radiated ar ea E32363 ■ parallel to layer Cer50 MMw wA Fig 14. Thermal conductivities of stacks made by compression and by using precursor-containing pastes, respectively 0 350 10 2e(deg Fig 12. X-ray diffraction patterns of sintered stacks measured in dif- er than that of Cer30, but this is not the case. Other factors may ferent orientations to the cing directio lso have an influence on the strength of the stacks. Although the microstructure of the interlayers of the various paste com- positions is not significantly different, their chemical composi- on may differ from each other. Also, stresses because of the by SEM at lower magnification(Fig 9). Vickers hardness data different shrinkage behavior, which act in the direction parallel (HVO.5)differ from 1090 in the porous interlayer to 1650 in the to the layers, could be present and seem to be most favorable for dense tape microstructure, which is another hint of defects. the composition Cer50. However, as mentioned above, one has These defects seem to be introduced by dust or cutting debri to consider that the processing has a great influence on the on the green tapes hindering the lamination process. The influ- strength results: the stacks consist of many layers and even if one ence of these defects on strength is strongest when the effective layer has not been laminated perfectly, the strength of all the forces are perpendicular to the layers-causing the observed an- samples cut out of this stack will be low. isotropy in strength. To confirm this supposition, lamination (B) Thermal Conductivity: Thermal conductivity of sili- planned to be carried out in a future work under clean room con nitride is affected mostly by the grain size distribution, the conditions and a very careful purification of the tapes to be aspect ratio of B-grains, the composition of the grain-boundary phase, the purity of the silicon nitride grains, and the overall Stacks laminated with composition Cer50 have almost the density. As no orientation of elongated B-grains parallel to the same strength and Weibull modulus as those laminated by com- casting direction has been observed, it was expected that the pression(Fig. 13). Other paste compositions lead to a decrease thermal properties of stacks laminated by compression should in mechanical properties, which may be because of an unfavor- be isotropic because of the good joint between the tapes and the able combination of shrinkage and thickness ratio of the tape absence of any interlayer with different properties. a few pores and the interlayer, resulting in internal stresses and reduction in the interlayers, which can have a strong influence on strength, the strength of the multi-layer stack. In the previous section, it ave almost no effect on thermal conductivity. For these com- has already been mentioned that the density of the stacks de- ponents, the measured thermal conductivity was in fact--unlike creases slightly with increasing precursor content of the paste the strength -almost the same in the two measured directions A lower density or higher number of pores typically results in a (Fig. 14). The thermal conductivity of the laminated stacks is in 600 E=003 086420 86420 er50 Cer30 Cer60 compression Cer50 Fig. 13. Graphical summary of the strength measurements of stacks made both by compression and by using precursor-containing pastes.by SEM at lower magnification (Fig. 9). Vickers hardness data (HV 0.5) differ from 1090 in the porous interlayer to 1650 in the dense tape microstructure, which is another hint of defects. These defects seem to be introduced by dust or cutting debris on the green tapes hindering the lamination process. The influ￾ence of these defects on strength is strongest when the effective forces are perpendicular to the layers—causing the observed an￾isotropy in strength. To confirm this supposition, lamination is planned to be carried out in a future work under clean room conditions and a very careful purification of the tapes to be laminated. Stacks laminated with composition Cer50 have almost the same strength and Weibull modulus as those laminated by com￾pression (Fig. 13). Other paste compositions lead to a decrease in mechanical properties, which may be because of an unfavor￾able combination of shrinkage and thickness ratio of the tape and the interlayer, resulting in internal stresses and reduction in the strength of the multi-layer stack. In the previous section, it has already been mentioned that the density of the stacks de￾creases slightly with increasing precursor content of the pastes. A lower density or higher number of pores typically results in a decrease in strength. Thus, the strength of Cer50 should be low￾er than that of Cer30, but this is not the case. Other factors may also have an influence on the strength of the stacks. Although the microstructure of the interlayers of the various paste com￾positions is not significantly different, their chemical composi￾tion may differ from each other. Also, stresses because of the different shrinkage behavior, which act in the direction parallel to the layers, could be present and seem to be most favorable for the composition Cer50. However, as mentioned above, one has to consider that the processing has a great influence on the strength results: the stacks consist of many layers and even if one layer has not been laminated perfectly, the strength of all the samples cut out of this stack will be low. (B) Thermal Conductivity: Thermal conductivity of sili￾con nitride is affected mostly by the grain size distribution, the aspect ratio of b-grains, the composition of the grain-boundary phase, the purity of the silicon nitride grains,22,23 and the overall density. As no orientation of elongated b-grains parallel to the casting direction has been observed, it was expected that the thermal properties of stacks laminated by compression should be isotropic because of the good joint between the tapes and the absence of any interlayer with different properties. A few pores in the interlayers, which can have a strong influence on strength, have almost no effect on thermal conductivity. For these com￾ponents, the measured thermal conductivity was in fact—unlike the strength—almost the same in the two measured directions (Fig. 14). The thermal conductivity of the laminated stacks is in Fig. 12. X-ray diffraction patterns of sintered stacks measured in dif￾ferent orientations to the stacking direction. Fig. 13. Graphical summary of the strength measurements of stacks made both by compression and by using precursor-containing pastes. Fig. 14. Thermal conductivities of stacks made by compression and by using precursor-containing pastes, respectively. 2720 Journal of the American Ceramic Society—Bitterlich and Heinrich Vol. 88, No. 10