Journal J Am Ceram Soc, 80 [8]2029-12 (1997) Tensile Stress Rupture of Sic/sicm Minicomposites with Carbon and Boron Nitride Interphases at elevated temperatures in Air Gregory N Morscher" T Case Western Reserve University, Cleveland, Ohio 4413 The stress-rupture properties of precracked minicompos- sealing"the cracked surface of the composite, prohibiting ites were determined in air at temperatures in the range of apid oxidation of the interphase from occurring 7000-1200oC. The minicomposite systems consisted of a At intermediate temperatures, the interphase may be con single tow of Nicalon or Hi-Nicalon fibers with carbon sumed if oxidation in the matrix crack is too slow to promote boron nitride (BN) interphases and a chemical-vapor crack sealing. -8 If the temperature is low enough to volatilize infiltrated silicon carbide (CVI-SiC)matrix. The stress- the interphase and not form stable solid- or liquid-oxidation rupture results were compared to single-fiber stress- products, the fiber-matrix bonding is destroyed, with a loss of rupture data and composite data in the literature. Severe load transfer and possible fiber degradation. If oxidation prod- embrittlement occurred for carbon interphase minicon ucts do form in place of the original interphase material, the posites. However, BN interphase minicomposites showed composite is embrittled, because of the strong bond that is only mild degradation in the rupture properties. This was formed between the fiber and the matrix and/or fiber degrada- true even though the BN interphase reacted and vaporized tion. This intermediate temperature regime poses the greates because of water vapor in the atmosphere at intermediate threat to composite embrittlement, because oxygen and other temperatures(7000-950C)and glass formation occurred vapor species have easy access to the interphase at higher temperatures(950%-12000C). The severe degra At higher temperatures, whether beneficial sealing can occur dation in rupture properties that occurred for carbon in for SiC/SiC under thermomechanical cycling must be deter- terphase composites at intermediate temperatures was due mined. a type of sealing phenomena has been shown to occur to degradation of the Nicalon-fiber properties from the en- for a Nic /BN/BMASm matrix composite. In this system, the vironment. The rupture properties of the BN-interphas outer skin(50 um) of the cracked matrix undergoes a phase minicomposites were controlled by the fib transformation during stress-rupture testing at 1100C, which erties at temperatures of less than -900oC and greater seals the cracks at the surface of the composite, prohibiting -1100C. In the range of -90011000C, most fibers embrittlement. However, this is different than what is hypoth to the matrix because of a glass layer that formed between esized for SiC, because the BMAs matrix itself transforms to the fiber and matrix, resulting in fiber stress concentrations cause a seal, whereas, with SiC, the oxidation product must that led to the mild embrittlement of the BN-interphase seal the cracks incomposite The other factor that may or may not be related to the en vironment is the mechanical properties of the fibers. Ulti L. Introduction mately, the composite mechanical properties can only be as HE success of a silicon carbide(SiC)-fiber-reinforced Sic good as the strength, rupture, and creep properties of the fibers matrix(SiC/SiC)composite as a high-temperature mate The fiber mechanical properties can degrade, because rial will be dependent on the ability of the composite to main- chemical reactions between the fiber, interphase, matrix, and/or tain desirable mechanical properties in an oxidizing environ- environment. The fiber strength also can degrade, because ment. Several factors can lead to deleterious composite intrinsic chemical instability, slow crack growth, or creep The objective of this study was to determine and explain the The most vulnerable component of the SiC/SiC system is stressed-oxidative stability over a wide range of temperatures the interphase that separates the fiber and the matrix. For a and times in air for carbon and BN interphase Sic/Sic ceramic- cracked composite in an oxidizing environment at an elevated matrix composites(CMCs). A minicomposite(single-tow com- temperature, the interphase is exposed to the environment. To posite"approach was used that enabled many specimens with maintain the debonding and sliding properties that are required different fiber/interphase combinations to be fabricated, The for good composite mechanical properties, either the interphase use of minicomposites also enabled several simple stress must be resistant to oxidation or the composite system must pture rigs to be built so that long-term(1000 h)experiment protect the interphase from oxidation. It is most desirable for could readily be performed. The results are discussed in rela- the interphase to be an oxidation-resistant material; however, to tion to the known fiber stress-rupture properties. The micro- date uch interphase has been proven effective at high structural features are described and related to the thermo temperatures for the SiC/SiCm system. The current interphase chemical reactions that occurred over the time/temperature of choice is boron nitride(BN), 2 which reacts with oxygen at for ndings were then related to the ob- intermediate temperatures(.) to form boron oxide served mechanical behavior and discussed in regard to their (B2O3 )liquid. It is hoped that the B2O, liquid will react with relevance to SiC/SiC composites the Sic matrix to form a borosilicate glass that is capable of Il. Experimental R. Naslain--contributing editor was pursued because it re- Several different variations were processed Two different commercial fibers(Nicalon 202 and Hi- Resident Research Associate at NASA Lewis Research Center, Cleveland, OH er, Nippon Carbon Co., Tokyo, Japan.Tensile Stress Rupture of SiCf/SiCm Minicomposites with Carbon and Boron Nitride Interphases at Elevated Temperatures in Air Gregory N. Morscher*,† Case Western Reserve University, Cleveland, Ohio 44135 The stress-rupture properties of precracked minicompos￾ites were determined in air at temperatures in the range of 700°–1200°C. The minicomposite systems consisted of a single tow of Nicalon or Hi-Nicalon fibers with carbon or boron nitride (BN) interphases and a chemical-vapor￾infiltrated silicon carbide (CVI-SiC) matrix. The stress￾rupture results were compared to single-fiber stress￾rupture data and composite data in the literature. Severe embrittlement occurred for carbon interphase minicom￾posites. However, BN interphase minicomposites showed only mild degradation in the rupture properties. This was true even though the BN interphase reacted and vaporized because of water vapor in the atmosphere at intermediate temperatures (700°–950°C) and glass formation occurred at higher temperatures (950°–1200°C). The severe degra￾dation in rupture properties that occurred for carbon in￾terphase composites at intermediate temperatures was due to degradation of the Nicalon-fiber properties from the en￾vironment. The rupture properties of the BN-interphase minicomposites were controlled by the fiber rupture prop￾erties at temperatures of less than ∼900°C and greater than ∼1100°C. In the range of ∼900°–1100°C, most fibers fused to the matrix because of a glass layer that formed between the fiber and matrix, resulting in fiber stress concentrations that led to the mild embrittlement of the BN-interphase minicomposites. I. Introduction THE success of a silicon carbide (SiC)-fiber-reinforced SiC matrix (SiCf /SiCm) composite as a high-temperature mate￾rial will be dependent on the ability of the composite to main￾tain desirable mechanical properties in an oxidizing environ￾ment. Several factors can lead to deleterious composite properties at elevated temperatures in oxidizing environments. The most vulnerable component of the SiCf /SiCm system is the interphase that separates the fiber and the matrix. For a cracked composite in an oxidizing environment at an elevated temperature, the interphase is exposed to the environment. To maintain the debonding and sliding properties that are required for good composite mechanical properties, either the interphase must be resistant to oxidation or the composite system must protect the interphase from oxidation. It is most desirable for the interphase to be an oxidation-resistant material; however, to date, no such interphase has been proven effective at high temperatures for the SiCf /SiCm system. The current interphase of choice is boron nitride (BN),1,2 which reacts with oxygen at intermediate temperatures (∼450°C) to form boron oxide (B2O3) liquid. It is hoped that the B2O3 liquid will react with the SiC matrix to form a borosilicate glass that is capable of ‘‘sealing’’ the cracked surface of the composite, prohibiting rapid oxidation of the interphase from occurring. At intermediate temperatures, the interphase may be con￾sumed if oxidation in the matrix crack is too slow to promote crack sealing.3–8 If the temperature is low enough to volatilize the interphase and not form stable solid- or liquid-oxidation products, the fiber–matrix bonding is destroyed, with a loss of load transfer and possible fiber degradation. If oxidation prod￾ucts do form in place of the original interphase material, the composite is embrittled, because of the strong bond that is formed between the fiber and the matrix and/or fiber degrada￾tion. This intermediate temperature regime poses the greatest threat to composite embrittlement, because oxygen and other vapor species have easy access to the interphase. At higher temperatures, whether beneficial sealing can occur for SiC/SiC under thermomechanical cycling must be deter￾mined. A type of sealing phenomena has been shown to occur for a Nicf ‡ /BN/BMASm matrix composite.2 In this system, the outer skin (∼50 mm) of the cracked matrix undergoes a phase transformation during stress-rupture testing at 1100°C, which seals the cracks at the surface of the composite, prohibiting embrittlement. However, this is different than what is hypoth￾esized for SiC, because the BMAS matrix itself transforms to cause a seal, whereas, with SiC, the oxidation product must seal the cracks. The other factor that may or may not be related to the en￾vironment is the mechanical properties of the fibers. Ulti￾mately, the composite mechanical properties can only be as good as the strength, rupture, and creep properties of the fibers. The fiber mechanical properties can degrade, because of chemical reactions between the fiber, interphase, matrix, and/or environment. The fiber strength also can degrade, because of intrinsic chemical instability, slow crack growth, or creep. The objective of this study was to determine and explain the stressed-oxidative stability over a wide range of temperatures and times in air for carbon and BN interphase SiC/SiC ceramic￾matrix composites (CMCs). A minicomposite (single-tow com￾posite9 ) approach was used that enabled many specimens with different fiber/interphase combinations to be fabricated. The use of minicomposites also enabled several simple stress￾rupture rigs to be built so that long-term (1000 h) experiments could readily be performed. The results are discussed in rela￾tion to the known fiber stress-rupture properties. The micro￾structural features are described and related to the thermo￾chemical reactions that occurred over the time/temperature range for testing. These findings were then related to the ob￾served mechanical behavior and discussed in regard to their relevance to SiC/SiC composites. II. Experimental Procedure (1) Minicomposites Tested The minicomposite approach was pursued because it re￾quired relatively short lengths (tens of feet) of coated tow. Several different minicomposite variations were processed. Two different commercial fibers (Nicalon 202 and Hi￾R. Naslain—contributing editor Manuscript No. 191885. Received April 15, 1996; approved December 26, 1996. Supported by the NASA HITEMP Program. *Member, American Ceramic Society. † Resident Research Associate at NASA Lewis Research Center, Cleveland, OH. ‡ Ceramic-grade Nicalon™ fiber, Nippon Carbon Co., Tokyo, Japan. J. Am. Ceram. Soc., 80 [8] 2029–42 (1997) Journal 2029