May 2006 Creep and After-Creep Stress-Strain Behavior 1657 cracking strengths in a turbine blade com ature at relatively low stress conditions and for a short time before insertion into service. Improved matrix strengths would also result in applications involving stresses that are for most part thermally driven, as these can still alter the residual stress state in a similar fashion. Most importantly, the creep results show that matrix cracking properties actually im- rove with use, provided the matrix is more creep prone than the reinforcing fibers and can accommodate a high creep strain without cracking. Both attributes appear to exist for the as-pro- duced CVI-SiC matrix of this study as it was able to deform at least enough to accommodate a relatively high creep strain (>0.2% permanent strain on average) Finally, these results show that one should be cautious when nterpreting the stress for non-linearity(proportional limit)at 138 MPa creep; Fast-fracture at 0.1mm 1315C for these composite systems. Non-linearity is often as- oom temperature aft ociated with matrix cracking, but as it was observed in this study, much non-linearity early in the stress-strain curve was Fig. 6. Micrograph of surface 90 crack. because of creep. Faster loading rates may result in higher elastic dulus and different offset stresses the offset-stress stress in the matrix upon rapid unloading from the applied stress measurements of an after-creep specimen should not be affected ondition. This infers that a portion of the creep strain was by creep as the material has effectively experienced"creep-hard simply because of increased loading of the more creep-resistant ening. fibers and to some extent CVI-SiC portion of the matrix. It also infers that sufficient creep followed by rapid stress removal and specimen cool-down can lead to increased compressive stress on V. Conclusions the matrix and higher off-set stresses for matrix thru-thickness Woven Hi-Nicalon Type S, melt-infiltrated SiC matrix com- cracking under fast or slow fracture conditions at low temper- ites were shown to be relatively creep resistant because of the atures and under fast-fracture conditions near the creep tem- eep-resistant nature of the reinforcing fiber, consistent with perature. The need for rapid stress removal and specimen cool- reep behavior for similar SiC/SiC systems. Tensile creep at down follows from the possibility that upon stress removal at d stresses up to 138 MPa in the primary fiber direction the creep temperature, residual tensile stresses on the fibers ays survived the 100 h limit imposed in this study. Fast could force the matrix to creep in compression, thereby reduc- ing the tensile stresses on the fibers and corresponding com- fracture tensile testing of specimens after 103 or 138 MPa for pressive stresses on the matrix. However, creep rate in 100 h creep showed significantly higher offset stresses, or the compression of siliconized Sic is known to be over an order as-producte -linear stress-strain behavior when compared with ites. This was attributed to the stress relax- of magnitude less than in compression and may minimize the matrix recovery. Because of the limited number of specimens, tion of the matrix during tensile creep resulting in an increased ss to ci to use not enough tests were performed to determine the optimum attribute of composite system to raise matrix cracking stresses, a esign parameter for CMCs, or at least anticipate a stress relaxation will occur under reduced load at the creep tem- in matrix cracking stress for certain applications subjected to perature. However, it is apparent that the small amounts of creep strain for the 103 MPa creep condition can lead to signif- low applied stresses. icant increase in the off-set stresses of the o/e curve ( Tables II and Ill and Fig. 3) More detailed studies are needed to optimize the constituents References nd temperature-time-stress conditions to maximize the matrix Rospars, J. L. Chermant, and P. Ladeveze, ""On a First Creep Model for a nomenon could be used to enhance matrix cracking n applications where this is desirable. For example, Fatigue Behavior in Hi-NicalonTM-Fiber-Reinforced Silicon Carbide Composites at High Temperatures. "J.Am. Ceram. Soc., 81[1] 117-28(1999 3S. Zhu. M. Mizuno. y Kagawa and Y Mut onotonic Tension. Fati Creep Behavior of SiC- Fiber-Reinforced SiC-Matrix Composites: A Re Comp. Sci. Technol, 59 [6]833-51( 1999). R. Bunsell and M-H. Berger. "Fine Diameter Ceramic Fibres. J. Eur. SG.N. Morscher and V. V. Pujar. "Melt-Infiltrated SiC Com urbine gs of the 49th ASME IGTI Turbo Lane H M. Yun and J.A. DiCarlo, Comparison of the Tensile Creep, and Rupture ngth Properties of Stoichiometric SiC Fibers, Ceram. Eng. Sci. Proc., 20[31 G.N. Morscher and J. Z. Gyekenesi "Room Temperature Tensile Behavior and Damage Accumulation of Hi-Nicalon Reinforced SiC Matrix Composites, L. Guillaumat and J. Lamon, "Multi-Fissuration De Composites SiC/SiC. Matrix Cracking in 2D Woven Sic-Fibe Reinforced Melt-Infiltrated SiC Matrix Composites, Comp. Sci. Techno/ 64 Panel A2 IH. M. Yun, J. A. DiCarlo, R. T. Bhatt and J.B. Hurst, "Processing and 175 MPa Creep SiC Fiber for SiC/SiC Components. 1 mm 四g.Sci.Pro,24[3247-53(2003) olmes."Influence of Stress Ratio on the Elevated-Temperature Fa- le Fiber-Reinforced Silicon Nitride Composite, "J.An. ig. 7. Micrograph of unbridged cracks in interior 0 fiber tow ceran.Soc,74[163945(1991)stress in the matrix upon rapid unloading from the applied stress condition. This infers that a portion of the creep strain was simply because of increased loading of the more creep-resistant fibers and to some extent CVI–SiC portion of the matrix. It also infers that sufficient creep followed by rapid stress removal and specimen cool-down can lead to increased compressive stress on the matrix and higher off-set stresses for matrix thru-thickness cracking under fast or slow fracture conditions at low temper￾atures and under fast-fracture conditions near the creep tem￾perature. The need for rapid stress removal and specimen cool￾down follows from the possibility that upon stress removal at the creep temperature, residual tensile stresses on the fibers could force the matrix to creep in compression, thereby reduc￾ing the tensile stresses on the fibers and corresponding com￾pressive stresses on the matrix. However, creep rate in compression of siliconized SiC is known to be over an order of magnitude less than in compression and may minimize the matrix recovery.16 Because of the limited number of specimens, not enough tests were performed to determine the optimum condition for maximizing matrix compression or whether matrix stress relaxation will occur under reduced load at the creep tem￾perature. However, it is apparent that the small amounts of creep strain for the 103 MPa creep condition can lead to signif￾icant increase in the off-set stresses of the s/e curve (Tables II and III and Fig. 3). More detailed studies are needed to optimize the constituents and temperature–time–stress conditions to maximize the matrix cracking stresses. Neverthless, this study shows that this phe￾nomenon could be used to enhance matrix cracking properties in applications where this is desirable. For example, the matrix cracking strengths in a turbine blade component could be im￾proved by subjecting the component to a higher than expected use temperature at relatively low stress conditions and for a short time before insertion into service.12 Improved matrix strengths would also result in applications involving stresses that are for most part thermally driven, as these can still alter the residual stress state in a similar fashion. Most importantly, the creep results show that matrix cracking properties actually im￾prove with use, provided the matrix is more creep prone than the reinforcing fibers and can accommodate a high creep strain without cracking. Both attributes appear to exist for the as-pro￾duced CVI–SiC matrix of this study as it was able to deform at least enough to accommodate a relatively high creep strain (40.2% permanent strain on average). Finally, these results show that one should be cautious when interpreting the stress for non-linearity (proportional limit) at 13151C for these composite systems. Non-linearity is often as￾sociated with matrix cracking, but as it was observed in this study, much non-linearity early in the stress–strain curve was because of creep. Faster loading rates may result in higher elastic modulus and different offset stresses. However, the offset-stress measurements of an after-creep specimen should not be affected by creep as the material has effectively experienced ‘‘creep-hard￾ening.’’ V. Conclusions Woven Hi-Nicalon Type STM, melt-infiltrated SiC matrix com￾posites were shown to be relatively creep resistant because of the creep-resistant nature of the reinforcing fiber, consistent with creep behavior for similar SiC/SiC systems. Tensile creep at ap￾plied stresses up to 138 MPa in the primary fiber direction al￾ways survived the 100 h limit imposed in this study. Fast￾fracture tensile testing of specimens after 103 or 138 MPa for 100 h creep showed significantly higher offset stresses, or the onset of non-linear stress–strain behavior when compared with as-produced composites. This was attributed to the stress relax￾ation of the matrix during tensile creep resulting in an increased stress to cause matrix cracking. It may be possible to use this attribute of composite system to raise matrix cracking stresses, a typical design parameter for CMCs, or at least anticipate a rise in matrix cracking stress for certain applications subjected to low applied stresses. References 1 C. Rospars, J. L. Chermant, and P. Ladeveze, ‘‘On a First Creep Model for a 2D SiCf–SiC Composite,’’ Mater. Sci. Eng. A, A250, 264–9 (1998). 2 S. Zhu, M. Mizuno, Y. Kagawa, J. Cao, Y. Nagano, and H. Kaya, ‘‘Creep and Fatigue Behavior in Hi-NicalonTM-Fiber-Reinforced Silicon Carbide Composites at High Temperatures,’’ J. Am. Ceram. Soc., 81 [1] 117–28 (1999). 3 S. Zhu, M. Mizuno, Y. Kagawa, and Y. Mutoh, ‘‘Monotonic Tension, Fatigue and Creep Behavior of SiC-Fiber-Reinforced SiC–Matrix Composites: A Re￾view,’’ Comp. Sci. Technol., 59 [6] 833–51 (1999). 4 A. R. Bunsell and M.-H. Berger, ‘‘Fine Diameter Ceramic Fibres,’’ J. Eur. Ceram. Soc., 20, 2249–60 (2000). 5 G. N. Morscher and V. V. Pujar, ‘‘Melt-Infiltrated SiC Composites for Gas Turbine Engine Applications’’; Proceedings of the 49th ASME IGTI Turbo Land, Sea and Air Conference, June 14–17, Vienna, Austria, 2004. Paper number: GT2004-53196. 6 H. M. Yun and J. A. DiCarlo, ‘‘Comparison of the Tensile Creep, and Rupture Strength Properties of Stoichiometric SiC Fibers,’’ Ceram. Eng. Sci. Proc., 20 [3] 259–72 (1999). 7 G. N. Morscher and J. Z. Gyekenesi, ‘‘Room Temperature Tensile Behavior and Damage Accumulation of Hi-Nicalon Reinforced SiC Matrix Composites,’’ Ceram. Eng. Sci. Proc., 19 [3] 241–9 (1998). 8 L. Guillaumat and J. Lamon, ‘‘Multi-Fissuration De Composites SiC/SiC,’’ Rev. Comp. Mater. Avances., 3, 159–71 (1993). 9 G. N. Morscher, ‘‘Stress-Dependent Matrix Cracking in 2D Woven SiC-Fiber Reinforced Melt-Infiltrated SiC Matrix Composites,’’ Comp. Sci. Technol., 64 [9] 1311–9 (2004). 10H. M. Yun, J. A. DiCarlo, R. T. Bhatt, and J. B. Hurst, ‘‘Processing and Structural Advantages of the Sylramic-iBN SiC Fiber for SiC/SiC Components,’’ Ceram. Eng. Sci. Proc., 24 [3] 247–53 (2003). 11J. W. Holmes, ‘‘Influence of Stress Ratio on the Elevated-Temperature Fa￾tigue of a Silicon Carbide Fiber-Reinforced Silicon Nitride Composite,’’ J. Am. Ceram. Soc., 74 [7] 1639–45 (1991). Fig. 6. Micrograph of surface 901 crack. Fig. 7. Micrograph of unbridged cracks in interior 01 fiber tow. May 2006 Creep and After-Creep Stress–Strain Behavior 1657