ournal 1.Am, Conan Sen,871lO1912-1918(2004 Shear Strength as a Function of Test Rate for SiC /BSAS Ceramic Matrix Composite at Elevated Temperature Sung R Choi"and Narottam P Bansal National Aeronautics and Space Administration, Glenn Research Center, Cleveland, Ohio 44135 Both interlaminar and in-plane shear strengths of a unidirec- studies have been done on this subject of slow crack growth of tional Hi-Nicalon M-fiber-reinforced barium strontium alumi- CFCCs under shear at elevated temperatures nosilicate(SiC/ BSAS) composite were determined at 1100 C In previous studies, ultimate tensile strength of several in air as a function of test rate using double-notch shear test different CFCCs was determined as a function of test rate at 1 100o specimens. The composite exhibited a significant effect of test to 1200 C in air. It was shown that the ultimate tensile strength of rate on shear strength, regardless of orientation. The shear those CFCCs depended significantly on the test rate: their strength degraded by about 50%o as the test rate decreased strengths increased as the test rate increased and decreased as the from 3.3 x 10 to 3.3 x 10- mm/s. The rate dependency of rate decreased. The leading mechanism has been understood as trength at 1100 C for the two-dimensional (2-D) SiCPBSAS low crack growth, supported by the results of both stress rupture and preloading testing. This understanding has suggested that life composite, in which tensile strength decreased by about 60% prediction parameters of CFCCs could be estimated, at least for a nological, power-law slow crack growth model is proposed and strength and test rate, termed"dynamic fatigue. "that has been formulated to account for the rate dependency of shear used in many monolithic brittle materials including glasses, glass rength of the composite. The proposed model has been ceramics, and advanced ceramics. The purpose of the current validated with additional results of both constant stress-rate work, extending those previous studies, was to determine the rate and constant stress testing in shear at 1100C using a 2-D dependency of shear strength at 1 C in air using a unidirec- Nicalon-fiber-reinforced crossply magnesium aluminosilicate tional Hi-Nicalon-fiber-reinforced barium strontium aluminosili- (SiCPMAS-5) ceramic matrix composite. cate(SiC /BSAScomposite. Both interlaminar and in-plane shear trengths of the SiC /BSAS composite were determined with louble-notch shear(DNS) test specimens as a function of the test I. Introduction rate ranging from 3.3 x 10 to 3. 3 X 10 mm/s. The effect of the test rate on the shear strength of the 1-D SiC, BSAS composite E successful development and design of continuous-fiber einforced ceramic matrix composites(CFCCs) are dependen similar two-dimensional (2-D) SiC/BSAS composite. A phenom on understanding their basic properties such as deformation enological, slow crack growth model is proposed to describe the fracture. and delayed failure (fatigue, slow crack growth, or damage accumulation) behavior. Particularly, accurate evaluation of delayed failure behavior under specified loading/environment rate-dependency effect of shear inducted on SiC/MAS-5 ceramic ditions is prerequisite to ensure accurate life prediction of matrix composite at 1 100C to verify the proposed model structural components at elevated temperatures Although CFCCs have shown improved resistance to fracture and increased damage tolerance compared with monolithic ceram IL. Experimental Procedure ics, inherent material/processing defects or cracks in the matrix rich interlaminar regions can still cause delamination under inter- The processing of SiC,/BSAS composite can be found else- laminar normal or shear stress, resulting in loss of stiffness or in where. Hi-Nicalon fibers with an average diameter of 14 um were used as reinforcement. The fiber surfaces were coated by some cases structural failure Strength behavior of CFCCs in shear chemical vapor deposition with 0.4 um BN followed by OI um has been characterized in view of their unique interfacial architec ures and its importance in structural applications. Because of SiC. The BN interfacial layer acts as a weak, crack deflection he inherent nature of ceramic matrix composites, it would be phase, while the SiC overcoat acts as a barrier to diffusion of boron highly feasible that interlaminar defects or cracks are susceptible from BN into the oxide matrix and also prevents diffusion of to slow crack growth or damage accumulation in certain environ- matrix elements into the fiber. The precursor to the celsian matrix ments(mostly air) particularly at elevated temperatures, resulting of 0.75Ba0-0 25Sro Al,O' 2SiO,(BSAS) was made by solid in strength degradation or time-dependent failure. Although slow state reaction. The precursor powder consisted mainly of SiO, and crack growth is one of the important life-limiting phenomena, few BaAl,O, with small amounts of Ba,SiOa, a-Al,O,. and Ba Sr,Al,O,. This powder was made into a slurry with an organic solvent with various additives, Tows of BN/SiC-coated fibers were impregnated with the matrix precursor by passing them through the slurry. The resulting prepreg tape was dried and cut into pieces. R J. Kerans-contributing edite ay-up (20 plies)followed by warm pressing at 150%C to fom Unidirectional fiber-reinforced composite was prepared by tape mposite. Finally, relatively mposites wer obtained by hot pressing under vacuum at 1500.C for 2 h under 28 Apnl27.2004 正ET MPa in a graphite die. X-ray diffraction showed that the precursor was fully converted into the desired monoclinic celsian phas through solid-state reacti e composite laminate thus fabr Resident Principal Scientist, Ohio Aerospace Institute cyogt-choie cated was about 4.2 mm thick and had a fiber volume fraction of eland, Ohio hould be addressed e-mail about 0.42 and a porosity of about I% 1912journal S7 | HI] 2004) Shear Strength as a Function of Test Rate for SiCf/BSAS Ceramic Matrix Composite at Elevated Temperature Sung R. Choi*-^ and Narottam P. Bansal* National Aeronautics and Space Administration, Glenn Research Center. Cleveland. Ohio 44135 Both interlaminar and in-plane shear strengths of a unidirec￾tional Hi-Nicalon'^-fiber-reinforced barium strontium alumi￾nosilicate (SiC/BSAS) composite were determined at I1OO"C in air as a function of test rate using double-notch shear test specimens. The composite exhibited a significant elTect of test rate on shear stren}>th, regardless of orientation. The shear strength degraded by about 50% as the lest rate decreased from 3.3 x 10"' to 2<.i x 10"*^ mm/s. The rate dependency of shear strength was similar to that observed for ultimate tensile strength at 1100"C for the two-dimensional (2-D) SiC/BSAS composite, in which tensile strength decreased by about 60% when the test rate varied from 5 lo 0.005 MPa/s. A phenome￾nological. power-law slow cratk growth model is proposed and formulated to account for the rale dependency of shear strength of the composite. The proposed model has been validated with additional results of hoth constant stress-rate and constant stress testinj> in shear at 1100°C using a 2-D Nicalun-fiber-reinforced crossply magnesium aluminosilicate (SiC|/MAS-5) ceramic matrix composite. I. Introduction T ill, successful development and design of continuous-fiber￾reinforced ceramic matrix composites (CFCCs) are dependent on understanding their basic properties such as deformation, fracture, and delayed failure (fatigue, slow crack growth, or damage accumulation) behavior. Particularly, accurate evaluation of delayed failure behavior under specified loading/environment conditions is prerequisite to ensure accurate life prediction of struclurul components at elevated temperatures. Although CFCCs have shown improved resistance to fracture and increased damage tolerance compared with monolithic ceram￾ics, inherent material/processing defects or cracks In the matrix￾rich interlaminar regions can still cause delamination under inter￾laminar normal or shear stress, resulting in loss of stiffness or in some cases structural failure. Strength behavior of CFCCs in shear has been characterized in view of their unique interfacial architec￾tures and its importance in structural applications.'^ Becau.se of the inherent nature of ceramic matrix composites, it would be highly feasible that interlaminar defects or cracks are susceptible to slow crack growth or damage accumulation in certain environ￾ments (mostly air) particularly al elevated temperatures, resulting in strength degradation or time-dependent failure. Although slow crack growth is one of the important life-limiting phenomena, few R. J. Kerans—contributing editor Manuscript No. iU61ll. Reteived October 20. :iX)3; iipproved April 27. 2(XM. This work was,suppi)ned in pan by the Ultra-Efficient Engine Technology lUEETl Project. NASA Glenn Research Center. Cleveland. OH, 'Member, American Ceramic Society. NASA Resident Principal Scientist. Ohio Aerospace tnstitute, Cleveland, Ohio, Author to whom eorrespondence shouid be addressed, e-mail: stjng.r.choi@ grc,na.sa,giiv. Studies have heen done on this subject of slow crack growth of CFCCs under shear at elevated temperatures. In previous studies.''"' ultimate lensile strength of several different CFCCs was determined as a function of test rate at 1100° to I2OO''C in air. It was shown that the ultitnate tensile strength of those CFCCs depended significantly on the test rate: their strengths increased as the test rate increased and decreased as the rate decreased. The leading mechanistn has been understood as slow crack growth, supported by the results of both stress rupture and preloading testing. This understanding has suggested that life prediction parameters of CFCCs could be estimated, at least for a short range of lifetimes, by a relationship between ultimate tensile strength and test rate, termed "dynamic fatigue." that has been used in many monolithic brittle materials including glasses, glass ceramics, and advanced ceramics. The purpose of tbe current work, extending those previous studies, was to determine the rate dependency of shear strength at 1100"C in air using a unidiree￾tional Hi-Nicalon-fiber-reinforced barium strontium aluminosili￾cate (SiC|/BSAS) composite. Both interlaminar and in-plane shear strengths of the SiC/BSAS composite were detennined with double-notch shear (DNS) test specimens as a function of the test rate ranging from 3.3 X 10"^ to 3.3 X iO~' mnVs. The effect of the test rate on the shear strength of the I -D SiC,/BS AS composite is compared with that of the test rate on the tensile strength of a similar two-dimensional (2-D) SiC/BSAS composite. A phenom￾enological, slow crack growth model is proposed to describe the rate-dependency effect of shear strength of the composite. Addi￾tionally, shear testing was also conducted on SiC,/MAS-5 ceramic matrix composite at 1 IOO"C to verify the proposed model. II. Experimental Procedure The processing of SiC/BSAS composite can be found else- '** Hi-Nicalon fibers with an average diameter of 14 jim were used as reinforcement. The liber surfaces were coated by chemical vapor deposition with 0.4 \x.m BN followed by 0,1 jim SiC. The BN intertacial layer acts as a weak, crack deflection phase, while the SiC overcoat acts as a barrier to diffusion of boron from BN into the oxide matrix and also prevents diffusion of matrix elements into the fibcr.*^ The precursor to the celsian matrix of 0.75BaO-0.25SrO-Al,O,-2SiO, (BSAS) was made by solid￾state reaction. The precursor powder consisted mainly of SiO, and BaAI-,04 with stiiall amounts of Ba^SJOj, a-AUOi, and Ba-,Sr2Al2O7. This powder was made into a slurry with an organic solvent with various additives. Tows of BN/SiC-coated ftbers were impregnated with the matrix precursor by passing them through the slurry. The resulting prepreg tape was dried and cut into pieces. Unidirectional tiber-reinlorced composite was prepared by tape lay-up (20 plies) followed by warm pressing at 150°C to form a "green" composite. Finally, relatively dense composites were obtained by hot pressing under vacuum at 1 .'>00^C for 2 h under 28 MPa in a graphite die. X-ray diffraction showed that the precursor was fully converted into the desired monocHnic celsian phase through solid-state reaction.^ The composite laminate thus fabri￾cated was about 4,2 mm thick and had a fiber volume fraction of about 0.42 and a porosity of about 1%. 1912