310 International Journal of Applied Ceramic Technolog-Liut, et al. Vol.8,No.2,2011 another two cycles of CVI SiC were essential for cov- followed by ion beam milling. The samples were exam- ering the porosity and the ends of fibers to form the final ined used field emission transmission electron micro- composites. Three kinds of composites were prepared. scope(Tecnai F30, FEI Company, Hillsboro, OR)with One was SiC matrix com posite reinforced with both an operated voltage of 300 kV SiC and carbon fibers([SiC-C]/Py C/SiC), which was named as hybrid composite and denoted as sample A C/PyC/SiC and C/PyC / SiC composites, denoted p The other two kinds of composites were named as Oxidation Tests Oxidation tests were conducted in static air in a samples B and C, respectively. The superscript HT tube furnace at 700C for 10h, and the weight changes means the PyC interphase was heat treated at 1800c of the samples were obtained by analytical balance(AG before the deposition of SiC matrix. It should be noted 204, Mettler Toledo, Schwerzenbach, Switzerland)and that PyC interphase in the hybrid composite was not be recorded as a function of oxidation time.Cumulative heat treated, because the strength of Hi-Nicalon SiC fi- weight change and the strength retained ratio of the ber deteriorate tes at temperatures exceeding 1400.C. samples were calculated according to the weight and flexural strength of the samples before and after oxida- tion tests, respectively Flexural Strength Tests and Microstructur Observation Results and Discussion Density and open porosity of the composites were btained by the archimedes method. The fexural Microstructure strength of the composites before and after oxidation was measured by the three-point Flexural method, and As shown in Table II, sample a had a higher den- the fracture toughness was determined by the single- sity than the other two kinds of composites. This result ge-notched-beam method using a fexural testing ma- was due to the significantly higher density of SiC fiber hine (SANS CMT 404, Sans Materials Testing, (2.74g/cm) than that of the carbon fiber(1.76g/cm) Shenzhen, China). The exural modulus was calculated Moreover, the diameter of SiC fber was larger than that using the slope of the load-displacement curves of the of carbon fiber, which made the spaces among the SiC composites according to the ASTM C1341-00 stan- fibers and SiC tows larger than those among carbon dard. The density, porosity, and Flexural strength of fibers and carbon tows, and hence it was easier for SiC the three kinds of the composites were measured after matrix to be deposited on the surfaces of SiC fibers and the sixth, seventh, and eighth cycles of CVI SiC matrix, SiC tows in the inner of the preform. As a result, the respectively. The microstructure of the composites and average value of open porosity the fracture surface morphologies of the tested samples lower than those of the other two kinds of composites, were analyzed by scanning electron microscopy (S- as shown in Table II 2700, Hitachi, Tokyo, Japan). For transmission elec The polished morphologies of SiC matrix in three on microscope (TEM) observation, the composites kinds of composites are shown in Fig. 1. There were were cut into samples with a thickness of 1000 um, no matrix microcracks found in sample A, as shown in and then mechanically thinned to a thickness of 30 um Figs. la and b. However, both samples B(Fig. Ic)and Table Il. Properties of the Three Kinds of Composites Flexural Fr acture Flexural ens porosity strengt modulus MPa) (MPam (GPa) ratio(%) A 6 530(16 175(2) 96(4.9) B 117(16) 33(0.5) 6(6.3) 18.6(1) 64(3.4 andard deviations are given in parentheses.another two cycles of CVI SiC were essential for cov￾ering the porosity and the ends of fibers to form the final composites. Three kinds of composites were prepared. One was SiC matrix composite reinforced with both SiC and carbon fibers ([SiC–C]/PyC/SiC), which was named as hybrid composite and denoted as sample A. The other two kinds of composites were named as C/PyC/SiC and C/PyCHT/SiC composites, denoted as samples B and C, respectively. The superscript HT means the PyC interphase was heat treated at 18001C before the deposition of SiC matrix. It should be noted that PyC interphase in the hybrid composite was not be heat treated, because the strength of Hi-Nicalon SiC fi- ber deteriorates at temperatures exceeding 14001C.9,10 Flexural Strength Tests and Microstructural Observation Density and open porosity of the composites were obtained by the Archimedes method. The flexural strength of the composites before and after oxidation was measured by the three-point flexural method, and the fracture toughness was determined by the single￾edge-notched-beam method using a flexural testing ma￾chine (SANS CMT 4304, Sans Materials Testing, Shenzhen, China). The flexural modulus was calculated using the slope of the load-displacement curves of the composites according to the ASTM C1341-00 stan￾dard.17 The density, porosity, and flexural strength of the three kinds of the composites were measured after the sixth, seventh, and eighth cycles of CVI SiC matrix, respectively. The microstructure of the composites and the fracture surface morphologies of the tested samples were analyzed by scanning electron microscopy (S- 2700, Hitachi, Tokyo, Japan). For transmission elec￾tron microscope (TEM) observation, the composites were cut into samples with a thickness of 1000 mm, and then mechanically thinned to a thickness of 30 mm followed by ion beam milling. The samples were exam￾ined used field emission transmission electron micro￾scope (Tecnai F30, FEI Company, Hillsboro, OR) with an operated voltage of 300 kV. Oxidation Tests Oxidation tests were conducted in static air in a tube furnace at 7001C for 10 h, and the weight changes of the samples were obtained by analytical balance (AG 204, Mettler Toledo, Schwerzenbach, Switzerland) and recorded as a function of oxidation time. Cumulative weight change and the strength retained ratio of the samples were calculated according to the weight and flexural strength of the samples before and after oxida￾tion tests, respectively. Results and Discussion Microstructure As shown in Table II, sample A had a higher den￾sity than the other two kinds of composites. This result was due to the significantly higher density of SiC fiber (2.74 g/cm3 ) than that of the carbon fiber (1.76 g/cm3 ). Moreover, the diameter of SiC fiber was larger than that of carbon fiber, which made the spaces among the SiC fibers and SiC tows larger than those among carbon fibers and carbon tows, and hence it was easier for SiC matrix to be deposited on the surfaces of SiC fibers and SiC tows in the inner of the preform. As a result, the average value of open porosity of hybrid composites was lower than those of the other two kinds of composites, as shown in Table II. The polished morphologies of SiC matrix in three kinds of composites are shown in Fig. 1. There were no matrix microcracks found in sample A, as shown in Figs. 1a and b. However, both samples B (Fig. 1c) and Table II. Properties of the Three Kinds of Composites Samples Density (g/cm3 ) Open porosity (%) Flexural strength (MPa) Fracture toughness (MPa m1/2) Flexural modulus (GPa) Strength retained ratio (%) A 2.56 6 530 (16) 17.5 (2) 96 (4.9) 99 B 2.23 10 117 (16) 3.3 (0.5) 66 (6.3) 67 C 2.25 9 505 (37) 18.6 (1) 64 (3.4) 78 Standard deviations are given in parentheses. 310 International Journal of Applied Ceramic Technology—Liu, et al. Vol. 8, No. 2, 2011