S Gustafsson et aL. /Joumal of the European Ceramic Sociery 29(2009)539-550 by the tEM observation of cavitation at multi-grain junctions mullite/sic at1400°C and the presence of strain contours in adjacent mullite grains This is consistent with the increase in the stress exponent from ur n=1.2at1300°Cton=2at1400°C. The mullite/5 vol. SiC nanocomposite has an increased creep resistance which is caused by grain boundary pinning by intergranular SiC particles and a reduced matrix grain size Creep at 1300C, and at lower stresses at 1400C, is dominated by diffusion, and may be controlled by the slower self-diffusion in the Sic particles as compared to the mullite matrix. Multi experimental creep predicted creep rate grain junction cavitation under higher stresses at 1400C. caused by rigid grain boundary sliding facilitated by softening 10 of an intergranular amorphous phase, reduces the positive effect STRESS(MPa) of the SiC nanoparticles. This results in a stress exponent that Fig. predicted nanocomposite creep rate at 1400C based on the Increases with stress at 1400C. hat Sic self-diffusion is rate limiting. 18,I 9 The experimentally deter- Data o tate creep rates of the nanocomposite at 1400 C are also shown. Acknowledgement the mullite matrix and within the SiC particles, and that these Research and from AIST, NEDO, Japan through the Synergy two processes would operate independently. The nanocompos- Ceramics project is gratefully acknowledged ite creep rates predicted at 1400C under these assumptions are shown in Fig. 15 together with the experimentally deter- mined creep rates at this temperature. There is a relatively good References agreement between the predicted and experimental creeprates at 1. Ihara. K New design concept of structural ceramics-ceramic nanocom- conditions may be controlled by self-diffusion through the si2)工1mhm对kn particles. This is also suggested by the observation of Sic par- alumina/silicon carbide nanocomposite. J Am Ceram Soc., 1994, 77(12), TEMinvestigation revealed, however, that there is usually a thin mina/silicon carbide nanocomposite amn. ceram soc. 1gg, 8029y glassy film separating a SiC particle from the mullite matrix This glass would not only provide a more efficient diffusion path, 4. Winn, A J and Todd, R I, Microstructural requirements for alumina-Sic but also be a viscous medium facilitating rigid particle/matrix nanocomposites. Br Ceram. Trans., 1999, 98(5), 219-22 5. Ohji, T, Hirano, T, Nakahira, A and Ihara, K, Particle/matrix interface The nanocomposite specimen crept under a stress of d its role in creep inhibition in alumina/silicon carbide nanocomposites. JAm. Cera.So,1996,791),33-45 50.0 MPa at 1400oC showed a significantly increased cavitation 6. Stearns, L C. Zhao, J and Harmer, M.P., Processing and microstructure at the multi-grain junctions. This suggests that unaccommo- development in Al2O3-SiC nanocomposites. J. Eur. Ceram Soc. 1992, 10, dated grain boundary sliding, facilitated by a reduced viscosi 473-477. of the grain boundary glass, had taken place. These observations 7. Ohji, T. Jeong. Y-K Choa, Y.H. and Ihara, K. Strengthening and tough- may explain the increased stress exponent at higher stresses at 1400C and the discrepancy between the proposed model and 81(6),1453-1460. Sternitzke, M, Review: structural ceramic nanocomposites. J. Eur. Ceram. the experimental creep data at these stresses(Figs. I and 15) Soc.,1997,17,1061-1082 The TEM results indicated that the dislocation activity was 9. Lessing, P A, Gordon, R.S. and Mazdiyasni, K.S. Creep of polycrystalline mullite. J. Am. Ceram. Soc.. 1975 58. 149 material. The low dislocation density was virtually unchanged in 10. Okamoto, Y, Fukudome, H. Hayashi, K and Nishikawa. T. Creep detor- the creep tested specimens. The combined results, hence, point 11. Descamps, P, Poortemann, M. and Cambier, E, Thermomechanical proper- towards diffusion-controlled creep processes at lower stresses idly quenched powders. Key Eng Mater and a transition to creep controlled by diffusion and rigid grain 1997,132-136,595-598 boundary sliding under higher stresses at 1400C. 12. Dokko, P C. Pask, J.A. and Mazdyasni, K S, High-temperature mechanical properties of mullite under compression. J. Am. Ceram. Soc., 1977, 60(3-4) 6. Concluding remarks 13. Hynes, A P and Doremus, R H, High-temperature compressive creep of ycrystalline mullite. J. Am. Ceram Soc., 1991, 74(10), 2469-2475. Creep of polycrystalline mullite at 1300C is controlled 14. Calderon-Moreno, J. M. and Torrecillas, R, High-temperature creep of polycrystalline mullite. Key Eng Mater, 1997, 132-136, 587-590 increases the creeprate as compared to apure mullite ceramic. At 15. Schneider, H, Okada k, and Pask, 1. A, Mullite and Mullite Ceramics 1400C, grain boundary sliding, facilitated by softening of the 16. Kingery, W.E., Bowen, H K and Uhlmann, D R, Introduction to Ceramics intergranular glass, dominates the creep behaviour as suggested (second edition). John Wiley Sons, US, 1976, P. 59S. Gustafsson et al. / Journal of the European Ceramic Society 29 (2009) 539–550 549 Fig. 15. The predicted nanocomposite creep rate at 1400 ◦C based on the assumption that SiC self-diffusion is rate limiting.18,19 The experimentally deter￾mined steady-state creep rates of the nanocomposite at 1400 ◦C are also shown. Data taken from Pitchford18. the mullite matrix and within the SiC particles, and that these two processes would operate independently. The nanocompos￾ite creep rates predicted at 1400 ◦C under these assumptions are shown in Fig. 15 together with the experimentally deter￾mined creep rates at this temperature. There is a relatively good agreement between the predicted and experimental creep rates at low stresses (<25 MPa), which indicates that creep under those conditions may be controlled by self-diffusion through the SiC particles. This is also suggested by the observation of SiC par￾ticle pinning of mullite/mullite grain boundaries (Fig. 14). The TEM investigation revealed, however, that there is usually a thin glassy film separating a SiC particle from the mullite matrix. This glass would not only provide a more efficient diffusion path, but also be a viscous medium facilitating rigid particle/matrix movement.34 The nanocomposite specimen crept under a stress of 50.0 MPa at 1400 ◦C showed a significantly increased cavitation at the multi-grain junctions. This suggests that unaccommo￾dated grain boundary sliding, facilitated by a reduced viscosity of the grain boundary glass, had taken place. These observations may explain the increased stress exponent at higher stresses at 1400 ◦C and the discrepancy between the proposed model and the experimental creep data at these stresses (Figs. 1 and 15). The TEM results indicated that the dislocation activity was very limited also during creep of the mullite/SiC nanocomposite material. The low dislocation density was virtually unchanged in the creep tested specimens. The combined results, hence, point towards diffusion-controlled creep processes at lower stresses and a transition to creep controlled by diffusion and rigid grain boundary sliding under higher stresses at 1400 ◦C. 6. Concluding remarks Creep of polycrystalline mullite at 1300 ◦C is controlled by diffusion, but the smaller volumes of grain boundary glass increases the creep rate as compared to a pure mullite ceramic. At 1400 ◦C, grain boundary sliding, facilitated by softening of the intergranular glass, dominates the creep behaviour as suggested by the TEM observation of cavitation at multi-grain junctions and the presence of strain contours in adjacent mullite grains. This is consistent with the increase in the stress exponent from n = 1.2 at 1300 ◦C to n = 2 at 1400 ◦C. The mullite/5 vol.% SiC nanocomposite has an increased creep resistance which is caused by grain boundary pinning by intergranular SiC particles and a reduced matrix grain size. Creep at 1300 ◦C, and at lower stresses at 1400 ◦C, is dominated by diffusion, and may be controlled by the slower self-diffusion in the SiC particles as compared to the mullite matrix. Multi￾grain junction cavitation under higher stresses at 1400 ◦C, caused by rigid grain boundary sliding facilitated by softening of an intergranular amorphous phase, reduces the positive effect of the SiC nanoparticles. This results in a stress exponent that increases with stress at 1400 ◦C. Acknowledgement Financial support from the Swedish Foundation for Strategic Research and from AIST, NEDO, Japan through the Synergy Ceramics project is gratefully acknowledged. References 1. Niihara, K., New design concept of structural ceramics—ceramic nanocom￾posites. J. Ceram. Soc. Jpn., 1991, 99(10), 974–982. 2. Ohji, T., Nakahira, A., Hirano, T. and Niihara, K., Tensile creep behaviour of alumina/silicon carbide nanocomposite. J. Am. Ceram. Soc., 1994, 77(12), 3259–3562. 3. Thompson, A. M., Chan, H. M. and Harmer, M. P., Tensile creep of alu￾mina/silicon carbide nanocomposite. J. Am. Ceram. Soc., 1997, 80(9), 221–228. 4. Winn, A. J. and Todd, R. I., Microstructural requirements for alumina–SiC nanocomposites. Br. Ceram. Trans., 1999, 98(5), 219–224. 5. Ohji, T., Hirano, T., Nakahira, A. and Niihara, K., Particle/matrix interface and its role in creep inhibition in alumina/silicon carbide nanocomposites. J. Am. Ceram. Soc., 1996, 79(1), 33–45. 6. Stearns, L. C., Zhao, J. and Harmer, M. P., Processing and microstructure development in Al2O3–SiC nanocomposites. J. Eur. Ceram. Soc., 1992, 10, 473–477. 7. Ohji, T., Jeong, Y.-K., Choa, Y.-H. and Niihara, K., Strengthening and tough￾ening mechanisms of ceramic nanocomposites. J. Am. Ceram. Soc., 1998, 81(6), 1453–1460. 8. Sternitzke, M., Review: structural ceramic nanocomposites. J. Eur. Ceram. Soc., 1997, 17, 1061–1082. 9. Lessing, P. A., Gordon, R. S. and Mazdiyasni, K. S., Creep of polycrystalline mullite. J. Am. Ceram. Soc., 1975, 58, 149. 10. Okamoto, Y., Fukudome, H., Hayashi, K. and Nishikawa, T., Creep defor￾mation of polycrystalline mullite. J. Eur. Ceram. Soc., 1990, 6, 161–168. 11. Descamps, P., Poortemann, M. and Cambier, F., Thermomechanical proper￾ties and creep of mullite from rapidly quenched powders. Key Eng. Mater., 1997, 132–136, 595–598. 12. Dokko, P. C., Pask, J. A. and Mazdyasni, K. S., High-temperature mechanical properties of mullite under compression. J. Am. Ceram. Soc., 1977, 60(3–4), 150–155. 13. Hynes, A. P. and Doremus, R. H., High-temperature compressive creep of polycrystalline mullite. J. Am. Ceram. Soc., 1991, 74(10), 2469–2475. 14. Calderon-Moreno, J. M. and Torrecillas, R., High-temperature creep of polycrystalline mullite. Key Eng. Mater., 1997, 132–136, 587–590. 15. Schneider, H., Okada, K. and Pask, J. A., Mullite and Mullite Ceramics. John Wiley & Sons, Chichester, UK, 1994. 16. Kingery, W. E., Bowen, H. K. and Uhlmann, D. R., Introduction to Ceramics (second edition). John Wiley & Sons, US, 1976, p. 595