S.R. Choi et al. /Journal of the European Ceramic Society 25(2005)1629-1636 ing Nicalon or Hi-Nicalon M SiC crossply(2D)fiber- Cooled Wedge Grip reinforced barium strontium aluminosilicate(designated Test Specimen Nicalon/BSAS and Hi-Nicalon/BSAS), Nicalon'unidirec tionally (ID) reinforced magnesium aluminosilicate(des- ignated SiCr/MAS), Nicalon"M plain-woven(2D)silicon carbide(designated SiCr/SiC: "enhanced"), and T-300 car- bon fiber-reinforced plain-woven(2D)silicon carbide(des- ignated C/SiC: "standard"and"enhanced ). The matrices Extensometer of the composites, except for Cr/SiC, were reinforced by Nicalon m or hi-nicalon m fibers with a fiber volume frac Induction Coil tion of about 0. 40. The unidirectional, crossply or plain- woven laminates of the SiC fiber-reinforced composites were typically 12-18 plies thick with a nominal thickness of around Cooled Wedge Grip 3-3.5 mm, depending on material. The Cr/Sic composite had a total of 29 plies, a fiber volume fraction of0. 46, and a nomi- nal laminate thickness of3 mm. The enhanced Sice/SiC com- Fig. 1. Schematics of experimental setup used in tensile testing for ceramic posite was modified from its standard matrix by a proprietary matrix composites at elevated temperatures in air process to increase the oxidation resistance of the composite Sic was also chemically vapor deposited on the composite sile testing was performed in accordance with ASTM Test panels to cover the residual porosity. The enhanced Cf/SiC composite had a boron carbide that was introduced into the Preload or accelerated tensile testing, primarily appl composite prior to deposition of pyrolytic carbon interface to protect the carbon fibers from oxidation Both Nicalon/bSAs conducted at 1 100 or 1200oC the lowest test rate of and Hi-Nicalon/BSAS were fabricated at NASA Glenn Re- 0.005 MPa/s in an attempt to better understand the govern- search Center, 2 SiCr/MAS by Corning, Inc. Corning, NY), 3 ng failure mechanism(s)of the composites. A predetermined SiC /Sic by duPont Company(Newark, DE), 3 and C/Sic preload, corresponding to an% of ultimate tensile strength by Honeywell Advanced Composites, Inc.(Newark, DE). 4 of each composite that was determined at 0.005 MPa/s with cessing can be found elsewhere. 2y omposites and their pro- no preload, was applied quickly (100 MPa/s) to the test Detailed information regarding the specimen prior to testing, and monotonic tensile testing at The dogboned tensile test specimens measuring 152mm 0.005 MPa/s started and continued until the test specimen (length) by 12.7 mm(width) were machined from the com- failed. The corresponding ultimate tensile strength was de- posite laminates, with the gage section of about 30mm termined. One test specimen was used in preload testing for long, 10 mm wide, and 3.0-3.5 mm thick(as-furnished) each composite The Cr/Sic test specimens were supplied with a notch ma- Constant stress("stress rupture") tensile testing was con- chined(in depth=2.5 mm and root radius 1.2 mm)at ducted at 100C in air for the Nicalon/BSAS (wIth two ne side of gage section at the longitudinal center of each different batches"A"and"B") composite using test speci test specimen. After machining, the Ce/Sic test specimens mens with the same geometry and the same test frame and were seal coated with SiC by the chemical vapor infiltration equipment that were employed in constant stress-rate tensile testing. The limited availability of test materials confined the Monotonic tensile testing was conducted in air at 1100oC testing to three to four test specimens, depending on batch for Nicalon/BSAS, Hi-Nicalon/BSAS and SiC/MAS-5 and Two to three different constant stresses were applied to test at 1200C for SiC /SiC and Cr/SiC, using a servohydraulic specimens and corresponding times to failure were deter test frame(Model 8501, Instron, Canton, MA). A schematic test setup is shown in Fig. 1. Each test specimen, located in- side of a SiC susceptor via two upper and lower water-cooled hydraulic grips, was induction-heated by radiation through 3. Results and discussion 15-kw power supply. Two high-temperature extensometers were placed on edges of each test specimen to measure ten- 3.. Ultimate tensile strength sile strain. Detailed descriptions on test setup and induction heating equipment were found in a previous study. 'A total The results of ultimate tensile strength as a function of of three different test rates in force control, corresponding test rate determined for the aforementioned CMCs are pre- to stress rates of 5, 0.16, and 0.005 MPa/s, were employed sented in Fig. 2, where ultimate tensile strength was plot for a given composite. This test method, when applied to ad- ted as a function of applied stress rate using log-log scales vanced monolithic ceramics is called constant stress rate o Each solid line in the figure represents the best-fit regres- dynamic fatigue"testing. Typically, one to three test speci- sion based on the log(ultimate tensile strength) versus log mens were tested at each test rate for a given composite. Ten-(applied stress rate)relation. The strength data determined1630 S.R. Choi et al. / Journal of the European Ceramic Society 25 (2005) 1629–1636 ing NicalonTM or Hi-NicalonTM SiC crossply (2D) fiber￾reinforced barium strontium aluminosilicate (designated Nicalon/BSAS and Hi-Nicalon/BSAS), NicalonTM unidirec￾tionally (1D) reinforced magnesium aluminosilicate (des￾ignated SiCf/MAS), NicalonTM plain-woven (2D) silicon carbide (designated SiCf/SiC: “enhanced”), and T-300 car￾bon fiber-reinforced plain-woven (2D) silicon carbide (des￾ignated Cf/SiC: “standard” and “enhanced”). The matrices of the composites, except for Cf/SiC, were reinforced by NicalonTM or Hi-NicalonTM fibers with a fiber volume frac￾tion of about 0.40. The unidirectional, crossply or plain￾woven laminates of the SiC fiber-reinforced composites were typically 12–18 plies thick with a nominal thickness of around 3–3.5 mm, depending on material. The Cf/SiC composite had a total of 29 plies, a fiber volume fraction of 0.46, and a nomi￾nal laminate thickness of 3 mm. The enhanced SiCf/SiC com￾posite was modified from its standard matrix by a proprietary process to increase the oxidation resistance of the composite. SiC was also chemically vapor deposited on the composite panels to cover the residual porosity. The enhanced Cf/SiC composite had a boron carbide that was introduced into the composite prior to deposition of pyrolytic carbon interface to protect the carbon fibers from oxidation. Both Nicalon/BSAS and Hi-Nicalon/BSAS were fabricated at NASA Glenn Re￾search Center,2 SiCf/MAS by Corning, Inc. (Corning, NY),3 SiCf/SiC by DuPont Company (Newark, DE),3 and Cf/SiC by Honeywell Advanced Composites, Inc. (Newark, DE).4 Detailed information regarding the composites and their pro￾cessing can be found elsewhere.2–4 The dogboned tensile test specimens measuring 152 mm (length) by 12.7 mm (width) were machined from the com￾posite laminates, with the gage section of about 30 mm long, 10 mm wide, and 3.0–3.5 mm thick (as-furnished). The Cf/SiC test specimens were supplied with a notch ma￾chined (in depth = 2.5 mm and root radius = 1.2 mm) at one side of gage section at the longitudinal center of each test specimen. After machining, the Cf/SiC test specimens were seal coated with SiC by the chemical vapor infiltration method.4 Monotonic tensile testing was conducted in air at 1100 ◦C for Nicalon/BSAS, Hi-Nicalon/BSAS and SiCf/MAS-5 and at 1200 ◦C for SiCf/SiC and Cf/SiC, using a servohydraulic test frame (Model 8501, Instron, Canton, MA). A schematic test setup is shown in Fig. 1. Each test specimen, located in￾side of a SiC susceptor via two upper and lower water-cooled hydraulic grips, was induction-heated by radiation through a 15-kW power supply. Two high-temperature extensometers were placed on edges of each test specimen to measure ten￾sile strain. Detailed descriptions on test setup and induction heating equipment were found in a previous study.3 A total of three different test rates in force control, corresponding to stress rates of 5, 0.16, and 0.005 MPa/s, were employed for a given composite. This test method, when applied to ad￾vanced monolithic ceramics, is called constant stress rate or “dynamic fatigue” testing. Typically, one to three test speci￾mens were tested at each test rate for a given composite. Ten￾Fig. 1. Schematics of experimental setup used in tensile testing for ceramic matrix composites at elevated temperatures in air. sile testing was performed in accordance with ASTM Test Standard C 1359.5 Preload or accelerated tensile testing, primarily applied to monolithic ceramics in order to save test time,6 was also conducted at 1100 or 1200 ◦C using the lowest test rate of 0.005 MPa/s in an attempt to better understand the govern￾ing failure mechanism(s) of the composites. A predetermined preload, corresponding to an 80% of ultimate tensile strength of each composite that was determined at 0.005 MPa/s with no preload, was applied quickly (∼100 MPa/s) to the test specimen prior to testing, and monotonic tensile testing at 0.005 MPa/s started and continued until the test specimen failed. The corresponding ultimate tensile strength was de￾termined. One test specimen was used in preload testing for each composite. Constant stress (“stress rupture”) tensile testing was con￾ducted at 1100 ◦C in air for the Nicalon/BSAS (with two different batches “A” and “B”) composite using test speci￾mens with the same geometry and the same test frame and equipment that were employed in constant stress-rate tensile testing. The limited availability of test materials confined the testing to three to four test specimens, depending on batch. Two to three different constant stresses were applied to test specimens and corresponding times to failure were deter￾mined. 3. Results and discussion 3.1. Ultimate tensile strength The results of ultimate tensile strength as a function of test rate determined for the aforementioned CMCs are pre￾sented in Fig. 2, where ultimate tensile strength was plot￾ted as a function of applied stress rate using log–log scales. Each solid line in the figure represents the best-fit regres￾sion based on the log (ultimate tensile strength) versus log (applied stress rate) relation. The strength data determined