CERAMICS INTERNATIONAL ELSEVIER Ceramics International 33(2007)905-909 www.elsevier.com/locate/ceramint Effects of polycarbosilane infiltration processes on the microstructure and mechanical properties of 3D-Cf/Sic composites Ke Jian", Zhao-Hui Chen, Qing-Song Ma Hai-feng Hu, Wen-Wei Zhen ey Laboratory of Advanced Ceramic Fibers Composites, National Universiry of Defense Technology. Changsha 410073, People's Republic of China Received 7 September 2005: received in revised form 29 January 2006: accepted 20 February 2006 Available online 18 April 2006 Abstract Three-dimensional braided carbon fiber-reinforced silicon carbide (3D-CrSiC) composites were prepared through eight cycles of infiltration of polycarbosilane(PCS)/divinylbenzene (DvB)and subsequent pyrolysis under an inert atmosphere. The effects of infiltration processes on the microstructure and mechanical properties of the Cr/Sic composites were investigated. The results showed that increasing temperature could reduce the viscosity of the PCS/DVB solution, which was propitious to the infiltration processes. The density and flexural strength of 3D-CpSic composites fabricated with vacuum infiltration were 1.794 g cm and 557 MPa, respectively. Compared to vacuum infiltration, heating and pressure infiltration could improve the infiltration efficiency so that the composites exhibited higher density and flexural strength, i.e 1.944 g cm and 662 MPa. When tested at 1650C and 1800C in vacuum, the flexural strength reached 647 MPa and 602 MPa, respectively C 2006 Elsevier Ltd and Techna Group S.r.L. All rights reserved. Keywords: B. Interfaces; Infiltration processes; Density; Viscosity; C/SiC composites 1. Introduction PIP), etc. [ 3-5]. Of these methods, the PIP route is being actively developed because it offers many potential advantages SiC materials fibers from the high-temperature cap- such as low processing temperature, controllable ceramic ability of carbon nd high modulus and oxidation compositions, and near-net-shape technologies. resistance of the atrix 1. As a fiber-reinforced mechanical properties of CaSic composites are composite, the mechanical and thermal properties can be determined by their microstructures. It is well demonstrated tailored by adjusting fiber volume and placement to meet the that weak interfacial bonding and high densities are preferred in needs of many applications. Additional flexibility can be order to obtain high performance CSiC composites [6, 7]. The achieved with the wide variety of carbon fibers available for microstructures of C Sic composites are governed by the reinforcement. Silicon carbide has good high-temperature fabrication process. The manufacturing of C Sic composites strength that is maintained to around 1700C, sufficient via PIP includes three stages In stage l, the fiber preform is thermal shock resistance, a low thermal expansion coefficient, infiltrated with precursor solution. In stage 2, the precursor is and good thermal conductivity [2]. As a result, CrSic cured into a solid. In stage 3, the cured solid is pyrolyzed in an composites are considered as desirable high-temperature inert or reactive atmosphere. Several infiltration-cure-pyrolysis structural materials and have wide application in various fields. cycles are required to densify composites because evolution of There are several methods to fabricate CrSiC composites, gaseous products and volume shrinkage occur during pyrolysis, such as chemical vapor infiltration(CVi), slurry infiltration resulting in micro-cracks and pores in the matrix. The influence combined with hot-pressing and polymer-infiltration-pyrolysis of the pyrolysis processes have been discussed in detail in evious papers [8, 9, but there are still relatively few reports fects of the different infiltration processes on E-omesi adrs: inke nud sina com (K. Jiany mechanical properties and microstructure of CSic prepared by PIP route. 2-8842/$3200C 2006 Elsevier Ltd and Techna Group S.r.L. All rights reserved 10.1016 1-ceramint.2006.0200
Effects of polycarbosilane infiltration processes on the microstructure and mechanical properties of 3D-Cf/SiC composites Ke Jian *, Zhao-Hui Chen, Qing-Song Ma, Hai-feng Hu, Wen-Wei Zheng Key Laboratory of Advanced Ceramic Fibers & Composites, National University of Defense Technology, Changsha 410073, People’s Republic of China Received 7 September 2005; received in revised form 29 January 2006; accepted 20 February 2006 Available online 18 April 2006 Abstract Three-dimensional braided carbon fiber-reinforced silicon carbide (3D-Cf/SiC) composites were prepared through eight cycles of infiltration of polycarbosilane (PCS)/divinylbenzene (DVB) and subsequent pyrolysis under an inert atmosphere. The effects of infiltration processes on the microstructure and mechanical properties of the Cf/SiC composites were investigated. The results showed that increasing temperature could reduce the viscosity of the PCS/DVB solution, which was propitious to the infiltration processes. The density and flexural strength of 3D-Cf/SiC composites fabricated with vacuum infiltration were 1.794 g cm3 and 557 MPa, respectively. Compared to vacuum infiltration, heating and pressure infiltration could improve the infiltration efficiency so that the composites exhibited higher density and flexural strength, i.e., 1.944 g cm3 and 662 MPa. When tested at 1650 8C and 1800 8C in vacuum, the flexural strength reached 647 MPa and 602 MPa, respectively. # 2006 Elsevier Ltd and Techna Group S.r.l. All rights reserved. Keywords: B. Interfaces; Infiltration processes; Density; Viscosity; Cf/SiC composites 1. Introduction Cf/SiC materials benefit from the high-temperature capability of carbon fibers and high modulus and oxidation resistance of the SiC matrix [1]. As a fiber-reinforced composite, the mechanical and thermal properties can be tailored by adjusting fiber volume and placement to meet the needs of many applications. Additional flexibility can be achieved with the wide variety of carbon fibers available for reinforcement. Silicon carbide has good high-temperature strength that is maintained to around 1700 8C, sufficient thermal shock resistance, a low thermal expansion coefficient, and good thermal conductivity [2]. As a result, Cf/SiC composites are considered as desirable high-temperature structural materials and have wide application in various fields. There are several methods to fabricate Cf/SiC composites, such as chemical vapor infiltration (CVI), slurry infiltration combined with hot-pressing and polymer-infiltration-pyrolysis (PIP), etc. [3–5]. Of these methods, the PIP route is being actively developed because it offers many potential advantages such as low processing temperature, controllable ceramic compositions, and near-net-shape technologies. The mechanical properties of Cf/SiC composites are determined by their microstructures. It is well demonstrated that weak interfacial bonding and high densities are preferred in order to obtain high performance Cf/SiC composites [6,7]. The microstructures of Cf/SiC composites are governed by the fabrication process. The manufacturing of Cf/SiC composites via PIP includes three stages. In stage 1, the fiber preform is infiltrated with precursor solution. In stage 2, the precursor is cured into a solid. In stage 3, the cured solid is pyrolyzed in an inert or reactive atmosphere. Several infiltration-cure-pyrolysis cycles are required to densify composites because evolution of gaseous products and volume shrinkage occur during pyrolysis, resulting in micro-cracks and pores in the matrix. The influence of the pyrolysis processes have been discussed in detail in previous papers [8,9], but there are still relatively few reports about the effects of the different infiltration processes on mechanical properties and microstructure of Cf/SiC prepared by PIP route. www.elsevier.com/locate/ceramint Ceramics International 33 (2007) 905–909 * Corresponding author. E-mail address: jianke_nudt@sina.com (K. Jian). 0272-8842/$32.00 # 2006 Elsevier Ltd and Techna Group S.r.l. All rights reserved. doi:10.1016/j.ceramint.2006.02.005
K Jian et al. /Ceramics International 33(2007)905-909 In this paper, carbon fiber-reinforced silicon carbide composites were fabricated with polycarbosilane(PCS)and divinylbenzene (VB)as precursors, and the effects of the infiltration processes on the microstructure and mechanical properties of 3D-C Sic composites were investigated 2. Experimental procedure 2 Three-dimensional braided carbon fiber(3D-Ce) was used as he reinforcement. It was two-step braided in Nanjing Fiberglass Research Design Institute. Fiber volume fraction Fig. 1. Viscosity-temperature curve of PCS/DVB solution. n x, y, z directions of three-dimensional braided carbon fiber preform was 8: 1: 1. The fibers were Jitan fibers (Jilin Carbon Corp, China) and the tensile strength and elastic modulus of 2200). For the mechanical properties test, three 3 mm x 4 hem were about 3000 MPa and 210 GPa, respectively mm x 70 mm specimens were measured for each composite Polycarbosilane whose average molecular weight and The load-displacement curves were recorded simultaneously softening point were about 1300 and 210C was used as the by using load-cell and laser extensometer. The elastic modulus precursor to SiC ceramics. It was synthesized by ourselves. of the composites were measured from the slope of the linear Divinylbenzene was used as solvent and cross-linking reagent portion of the load-displacement curves for PCS In previous work, the PCS/DVB solution with a mass Fracture surfaces of the specimens after three-point ratio of 1: 0. 4 was demonstrated to be proper to prepare CSic bending tests were examined by scanning electron micro- composites [10] scopy(SEM) 2.2. Preparation of 3D-B CAsic 3. Results and discussion Two samples which were denoted as 3D-VI and 3D-Pl were 3. 1. The infiuence of temperatures on the viscosity of the prepared with different infiltration processes. First, the 3D PCS/DVB solution carbon fiber preforms were put into an airtight vessel, and then the preforms were infiltrated with PCS/DVB solution in In a previous work, the PCS/DVB solution with a mass ratio vacuum conditions. Then the 3D-VI was put into the oven of 1: 0.4 was demonstrated to be proper to prepare CSiC directly and heated to 150C. The temperature was maintained composites. Nevertheless, the solution had high viscosity at for about 6 h to ensure PCS/DVB solution to cure into a solid. room temperature resulting in a great difficulty to impregnate On the other hand, 3D-PI was diverted into an autoclave and the solution into the carbon fiber preforms. The viscosity heated to 80C, and then 6 MPa nitrogen atmosphere was depends on the temperature are shown in Fig. 1. When the purged. About 60 min later, the autoclave was heated to 150C temperature was increased to 80C, the viscosity of and left at this temperature for about 6 h. Afterwards, the cured solution reduced to 0.104 Pa s and it exhibited a good 3D-VI and 3D-PI were pyrolyzed at high temperature in an flowability. Nevertheless with a temperature above 80C, inert atmosphere. To densify the composites, seven infiltration- the solution will cure rapidly which is greatly disadvantageous cure-pyrolysis cycles were repeated in the same conditions as to the infiltrative processes those of the initial cycle 2.3. Characterization of the composites The viscosities of the Pcs/dvb solution were tested with a calibrated Ubbelohde-type capillary viscometer. The bulk densities of the composites were computed from the weight-to-volume ratio Three-point bending tests were used to evaluate flexural strength of the composites with the span/height ratio of 15 and a crosshead speed of 0.5 mm/min. The flexural strength of 3D-PI composites at room temperature was tested at Central South Theoretical data University, China, using a universal testing machine(Instron- 1342). The flexural strength of composites at high temperature in vacuum was tested at Northwestern Polytechnical University, Pyrolysis Cycles China, using a flexural and tensile testing machine (Y KM Fig. 2. Density-infiltration time curves of 3D-CHSiC
In this paper, carbon fiber-reinforced silicon carbide composites were fabricated with polycarbosilane (PCS) and divinylbenzene (DVB) as precursors, and the effects of the infiltration processes on the microstructure and mechanical properties of 3D-Cf/SiC composites were investigated. 2. Experimental procedure 2.1. Raw materials Three-dimensional braided carbon fiber (3D-Cf) was used as the reinforcement. It was two-step braided in Nanjing Fiberglass Research & Design Institute. Fiber volume fraction in x, y, z directions of three-dimensional braided carbon fiber preform was 8:1:1. The fibers were Jitan fibers (Jilin Carbon Corp., China) and the tensile strength and elastic modulus of them were about 3000 MPa and 210 GPa, respectively. Polycarbosilane whose average molecular weight and softening point were about 1300 and 210 8C was used as the precursor to SiC ceramics. It was synthesized by ourselves. Divinylbenzene was used as solvent and cross-linking reagent for PCS. In previous work, the PCS/DVB solution with a mass ratio of 1:0.4 was demonstrated to be proper to prepare Cf/SiC composites [10]. 2.2. Preparation of 3D-B Cf/SiC Two samples which were denoted as 3D-VI and 3D-PI were prepared with different infiltration processes. First, the 3D carbon fiber preforms were put into an airtight vessel, and then the preforms were infiltrated with PCS/DVB solution in vacuum conditions. Then the 3D-VI was put into the oven directly and heated to 150 8C. The temperature was maintained for about 6 h to ensure PCS/DVB solution to cure into a solid. On the other hand, 3D-PI was diverted into an autoclave and heated to 80 8C, and then 6 MPa nitrogen atmosphere was purged. About 60 min later, the autoclave was heated to 150 8C and left at this temperature for about 6 h. Afterwards, the cured 3D-VI and 3D-PI were pyrolyzed at high temperature in an inert atmosphere. To densify the composites, seven infiltrationcure-pyrolysis cycles were repeated in the same conditions as those of the initial cycle. 2.3. Characterization of the composites The viscosities of the PCS/DVB solution were tested with a calibrated Ubbelohde-type capillary viscometer. The bulk densities of the composites were computed from the weight-to-volume ratio. Three-point bending tests were used to evaluate flexural strength of the composites with the span/height ratio of 15 and a crosshead speed of 0.5 mm/min. The flexural strength of composites at room temperature was tested at Central South University, China, using a universal testing machine (Instron- 1342). The flexural strength of composites at high temperature in vacuum was tested at Northwestern Polytechnical University, China, using a flexural and tensile testing machine (YKM- 2200). For the mechanical properties test, three 3 mm 4 mm 70 mm specimens were measured for each composite. The load–displacement curves were recorded simultaneously by using load-cell and laser extensometer. The elastic modulus of the composites were measured from the slope of the linear portion of the load–displacement curves. Fracture surfaces of the specimens after three-point bending tests were examined by scanning electron microscopy (SEM). 3. Results and discussion 3.1. The influence of temperatures on the viscosity of the PCS/DVB solution In a previous work, the PCS/DVB solution with a mass ratio of 1:0.4 was demonstrated to be proper to prepare Cf/SiC composites. Nevertheless, the solution had high viscosity at room temperature resulting in a great difficulty to impregnate the solution into the carbon fiber preforms. The viscosity depends on the temperature are shown in Fig. 1. When the temperature was increased to 80 8C, the viscosity of the solution reduced to 0.104 Pa s and it exhibited a good flowability. Nevertheless with a temperature above 80 8C, the solution will cure rapidly which is greatly disadvantageous to the infiltrative processes. 906 K. Jian et al. / Ceramics International 33 (2007) 905–909 Fig. 1. Viscosity–temperature curve of PCS/DVB solution. Fig. 2. Density–infiltration time curves of 3D-Cf/SiC.
K Jian et al /Ceramics International 33(2007)905-909 (n≥1) (4) H-3D-PI Theoretical data where the superscript E means experimental data and the sub script denotes the number of the fabrication cycles; m, the mass of he C/SiC composites; Psic, the density of PCS/DVB solution and SiC matrix derived from PCS/DVB pyrolysis (about 2.55 g cm ) and V, the volume of the composites. Because the volume of the composites changes slightly, Vis used to denote the volume of the composites during the whole fabrication processes The theoretical densities and porosities are given by Pyrolysis Cycles Pn=Pi-I+0r-71Ppcs/DVB (n21) Fig. 3. Porosity-infiltration time curves of 3D-C/ SiC. 3.2. The influence of infiltration processes on the densities OT=oT_8n-InPpCs/DVB (n21) and porosities of the C/sic composites Psi-c where the superscript T means theoretical data; PpCS/DVB, the According to the viscosity-temperature curve, the para- density of PCS/DVB solution(about 1.02 g cm); and n, the fabricated by vacuum infiltration curve of the 3D-Pi was closer to the theoretical curve than that Density is one of the most important factors for the mechanical of the 3D-VI, i.e., the density of 3D-PI was increased much properties of the CsIc composites. The experimental and faster than for 3D-VI, i.e., compared to vacuum infiltration, theoretical densities and porosities of the samples in relation to heating and pressure infiltration could improve the infiltration the number of cycles are shown in Figs. 2 and 3, respectively. In efficiency Figs 2 and 3, Po and Bo are the density and porosity of the 3D-Cr preforms. In this case, po and Bo are given by 3.3. The infiuence of infiltration processes on the flexura strength of the CSic composite Po 6=1-2 The properties of the CSic composites fabricated with (2) different infiltration processes are shown in Table 1. The density and flexural strength of 3D-Vi were 1.794 g cm and where me is the mass of the carbon fiber performs: Vo, the volume of the carbon fiber performs; and Po, the density of the Table I carbon fiber(about 1.76 g cm) Properties of 3D-CrSiC composites fabricated by different infiltration pro- The experimental densities of the composites were cesses computed from the weight-to-volume ratio and the experi- Sample Density (g cm- Porosity (%) Flexural strength(MPa) mental porosities were obtained from 3DVI1.794 16.48 1944 A=v(n21) Fig. 4. SEM photos of fracture surface of the 3D-VI composites (a) Sample 3D-VI(x40)and(b)sample 3D-VI(x1000)
3.2. The influence of infiltration processes on the densities and porosities of the Cf/SiC composites According to the viscosity–temperature curve, the parameters of the heating and pressure infiltration to fabricate 3D-PI were determined. In order to make a comparison, 3D-VI was fabricated by vacuum infiltration. Densityis one ofthe mostimportant factors forthe mechanical properties of the Cf/SiC composites. The experimental and theoretical densities and porosities of the samples in relation to the number of cycles are shown in Figs. 2 and 3, respectively. In Figs. 2 and 3, r0 and u0 are the density and porosity of the 3D-Cf preforms. In this case, r0 and u0 are given by r0 ¼ mf V0 (1) u0 ¼ 1 r0 rf (2) where mf is the mass of the carbon fiber performs; V0, the volume of the carbon fiber performs; and r0, the density of the carbon fiber (about 1.76 g cm3 ). The experimental densities of the composites were computed from the weight-to-volume ratio and the experimental porosities were obtained from rE n ¼ mE n V ðn 1Þ (3) uE n ¼ uE n1 mE n mE n1 VrSiC ðn 1Þ (4) where the superscript E means experimental data and the subscript denotes the number ofthe fabrication cycles;m, the mass of the C/SiC composites; rSi–C, the density of PCS/DVB solution and SiC matrix derived from PCS/DVB pyrolysis (about 2.55 g cm3 ); and V, the volume of the composites. Because the volume of the composites changes slightly, Vis used to denote the volume of the composites during the whole fabrication processes. The theoretical densities and porosities are given by rT n ¼ rT n1 þ uT n1hrPCS=DVB ðn 1Þ (5) uT n ¼ uT n1 uT n1hrPCS=DVB rSiC ðn 1Þ (6) where the superscript T means theoretical data; rPCS/DVB, the density of PCS/DVB solution (about 1.02 g cm3 ); and h, the ceramic yield of the cured PCS/DVB (about 0.58). Figs. 2 and 3 clearly show that the density–number of cycles curve of the 3D-PI was closer to the theoretical curve than that of the 3D-VI, i.e., the density of 3D-PI was increased much faster than for 3D-VI, i.e., compared to vacuum infiltration, heating and pressure infiltration could improve the infiltration efficiency. 3.3. The influence of infiltration processes on the flexural strength of the Cf/SiC composites The properties of the Cf/SiC composites fabricated with different infiltration processes are shown in Table 1. The density and flexural strength of 3D-VI were 1.794 g cm3 and K. Jian et al. / Ceramics International 33 (2007) 905–909 907 Fig. 3. Porosity–infiltration time curves of 3D-Cf/SiC. Fig. 4. SEM photos of fracture surface of the 3D-VI composites. (a) Sample 3D-VI (40) and (b) sample 3D-VI (1000). Table 1 Properties of 3D-Cf/SiC composites fabricated by different infiltration processes Sample Density (g cm3 ) Porosity (%) Flexural strength (MPa) 3D-VI 1.794 16.48 557 3D-PI 1.944 10.60 662
908 K Jian et al. /Ceramics International 33(2007)905-909 Fig. 5. SEM photos of fracture surface of the 3D-Pl composites (a) Sample 3D-PI (x50)and (b) sample 3D-PI(x 1000). 557 MPa, respectively. The 3D-PI composites exhibited higher Table 2 density and flexural strength, 1.944 g cmand 662 MPa Mechanical properties of 3D-C.SiC at high temperatures Tested temperature (C) Flexural The fracture surfaces of the 3D-VI and 3 D-PI composites are odulus( GPa) iven in Figs. 4 and 5. Both kinds of composites show extensive fiber-debonding and pull-out, which indicates a relatively weak fiber/matrix interfacial bonding. It was well 1800 328 demonstrated that weak interfacial bonding could result in high mechanical properties. As a result, 3D-VI and 3D-PI also flexural modulus at high temperature is mainly to be attributed exhibited high flexural strength. at the same time it could be to the softening of the sic matrix. observed that the number and length of the pull-out fibers of the composites were similar for the 3D-VI and 3D-PI composites. 4. Conclusions Consequently, the difference of the flexural strength did not result from different interfacial structures (1) With the increase of temperature, the viscosity of the PCS/ Density is another important factor to affect the mechanical DVB solution reduced sharply. Nevertheless with too high properties of the CSiC composites. Fig. 4 shows that in the temperatures, the solution would cure in a too short time. 3D-VI composites, big sized pores exist within the fasciculi and 80C was a proper temperature for the impregnation of PCS/ many small sized pores exist among the fibers. Pores derive DVB solution from(i)the evolution of gaseous products occurring during (2)The density and flexural strength of the composites fabricated PCS/DVB pyrolysis, (ii) the incomplete infiltration of the with vacuum infiltration were only 1.794 g cm and precursor solution, and (iii)the difference of the coefficient of 557 MPa, respectively. The composites fabricated with thermal expansion between carbon fiber and Sic matrix. Fig. 5 heating and pressure infiltration exhibited higher density shows that the number and size of the pores in the 3D-PI and flexural strength, 1.944 g cm and 662 MPa, respec composites decrease markedly compared to the 3D-VI tively. Both composites exhibited a weak fiber-matrix composites. By the heating and pressure infiltration processes interfacial bonding and the difference of the strength did e PCS/DVB solution could be impregnated into the pores not result from the different interfacial structures The more easily than for vacuum infiltration. As a result, compared increase of the flexural strength was mainly attributed to the to vacuum infiltration, heating and pressure infiltration could decrease of the porosity that resulted from the improvement improve the infiltration efficiency so that the 3D-PI had higher of the infiltration efficiency density, which agrees with the results of Fig. 2. Therefore, 3D-(3)Tested at 1650C and 1800C in vacuum, the flexural PI exhibited higher flexural strength than 3D-VI. strength reached 647 MPa and 602 MPa, respectively. The 3D-CSiCcomposites exhibited good flexural strength athigh 3. 4. Mechanical properties of Csic at high temperature temperatures. Nevertheless, the residual flexural modulus was only 48.6% and 25.0%o at 1650C and 1800C, respectively The flexural strength of the 3D-PI composites tested at The decrease in the flexural modulus at high temperature different temperatures in vacuum is shown in Table 2. The was mainly attributed to the softening of the Sic matrix flexural strength show a slight decrease when tested at 1650C and 1800C, the residual strength being 97.7% and 92.0%, References respectively. On the contrary, the flexural modulus reduced greatly at high temperature, the residual flexural modulus being [11 C. Thomas, T. Joseph, A. Peter, M. James, Ceram. Eng. Sci. Proc. 21(3) only 48.6% and 25.0%0, respectively. The decrease in the (2000517-524
557 MPa, respectively. The 3D-PI composites exhibited higher density and flexural strength, 1.944 g cm3 and 662 MPa, respectively. The fracture surfaces of the 3D-VI and 3D-PI composites are given in Figs. 4 and 5. Both kinds of composites show an extensive fiber-debonding and pull-out, which indicates a relatively weak fiber/matrix interfacial bonding. It was well demonstrated that weak interfacial bonding could result in high mechanical properties. As a result, 3D-VI and 3D-PI also exhibited high flexural strength. At the same time it could be observed that the number and length of the pull-out fibers of the composites were similar for the 3D-VI and 3D-PI composites. Consequently, the difference of the flexural strength did not result from different interfacial structures. Density is another important factor to affect the mechanical properties of the Cf/SiC composites. Fig. 4 shows that in the 3D-VI composites, big sized pores exist within the fasciculi and many small sized pores exist among the fibers. Pores derive from (i) the evolution of gaseous products occurring during PCS/DVB pyrolysis, (ii) the incomplete infiltration of the precursor solution, and (iii) the difference of the coefficient of thermal expansion between carbon fiber and SiC matrix. Fig. 5 shows that the number and size of the pores in the 3D-PI composites decrease markedly compared to the 3D-VI composites. By the heating and pressure infiltration processes, the PCS/DVB solution could be impregnated into the pores more easily than for vacuum infiltration. As a result, compared to vacuum infiltration, heating and pressure infiltration could improve the infiltration efficiency so that the 3D-PI had higher density, which agrees with the results of Fig. 2. Therefore, 3DPI exhibited higher flexural strength than 3D-VI. 3.4. Mechanical properties of Cf/SiC at high temperature The flexural strength of the 3D-PI composites tested at different temperatures in vacuum is shown in Table 2. The flexural strength show a slight decrease when tested at 1650 8C and 1800 8C, the residual strength being 97.7% and 92.0%, respectively. On the contrary, the flexural modulus reduced greatly at high temperature, the residual flexural modulus being only 48.6% and 25.0%, respectively. The decrease in the flexural modulus at high temperature is mainly to be attributed to the softening of the SiC matrix. 4. Conclusions (1) With the increase of temperature, the viscosity of the PCS/ DVB solution reduced sharply. Nevertheless with too high temperatures, the solution would cure in a too short time. 80 8C was a proper temperature for the impregnation of PCS/ DVB solution. (2) The density and flexural strength ofthe composites fabricated with vacuum infiltration were only 1.794 g cm3 and 557 MPa, respectively. The composites fabricated with heating and pressure infiltration exhibited higher density and flexural strength, 1.944 g cm3 and 662 MPa, respectively. Both composites exhibited a weak fiber-matrix interfacial bonding and the difference of the strength did not result from the different interfacial structures. The increase of the flexural strength was mainly attributed to the decrease of the porosity that resulted from the improvement of the infiltration efficiency. (3) Tested at 1650 8C and 1800 8C in vacuum, the flexural strength reached 647 MPa and 602 MPa, respectively. The 3D-Cf/SiCcompositesexhibited goodflexural strengthat high temperatures. Nevertheless,the residual flexural modulus was only 48.6% and 25.0% at 1650 8C and 1800 8C, respectively. The decrease in the flexural modulus at high temperature was mainly attributed to the softening of the SiC matrix. References [1] C. Thomas, T. Joseph, A. Peter, M. James, Ceram. Eng. Sci. Proc. 21 (3) (2000) 517–524. 908 K. Jian et al. / Ceramics International 33 (2007) 905–909 Fig. 5. SEM photos of fracture surface of the 3D-PI composites. (a) Sample 3D-PI (50) and (b) sample 3D-PI (1000). Table 2 Mechanical properties of 3D-Cf/SiC at high temperatures Tested temperature (8C) Flexural strength (MPa) Flexural modulus (GPa) 20 662 131 1650 647 63.7 1800 609 32.8
K Jian et al /Ceramics International 33(2007)905-909 [2]SM. Zhu, S.Q. Ding, H.A. Xi, R D Wang, Mater. Lett. 59(5)(2005)595- [7] H.Q. Ly, R. Taylor, R.J. Day, J Mater. Sci. 36(15)(2001)4027-4035 18 Z.H. Ke Jian, Q. Chen, S Ma, w.w. Zheng, Mater. Sci. Eng. A390(1-2) [3] J Suo, Z.H. Chen, J.Y. Xiao, w.w. Zheng, Ceram. Int. 31(2005)447-452 (2005)154158 4 M.Z. Berbon, D R. Dietrith, D B. Marshall, J. Am. Ceram Soc. 84(10 [9 Q.S. Ma, Z.H. Chen, w.w. Zheng, H F Hu, Ceram. Int. 31(2005)305- (2001)2229-2234. [5] T Noda, H. Araki, H. Suzuki, J Nucl. Mater. 212-1(9)(1994)823-829. [10] B Su, XJ. Wu, Z.H. Chen, Polymer. Mater. Sci Eng. 10(2)(1994)121- [6] J.F. Desprec, M. Monthioux, J. Eur. Ceram Soc. 15 (3)(1995)209-224 123 (in Chinese)
[2] S.M. Zhu, S.Q. Ding, H.A. Xi, R.D. Wang, Mater. Lett. 59 (5) (2005) 595– 597. [3] J. Suo, Z.H. Chen, J.Y. Xiao, W.W. Zheng, Ceram. Int. 31 (2005) 447–452. [4] M.Z. Berbon, D.R. Dietrith, D.B. Marshall, J. Am. Ceram. Soc. 84 (10) (2001) 2229–2234. [5] T. Noda, H. Araki, H. Suzuki, J. Nucl. Mater. 212–1 (9) (1994) 823–829. [6] J.F. Desprec, M. Monthioux, J. Eur. Ceram. Soc. 15 (3) (1995) 209–224. [7] H.Q. Ly, R. Taylor, R.J. Day, J. Mater. Sci. 36 (15) (2001) 4027–4035. [8] Z.H. Ke Jian, Q. Chen, S. Ma, W.W. Zheng, Mater. Sci. Eng. A390 (1–2) (2005) 154–158. [9] Q.S. Ma, Z.H. Chen, W.W. Zheng, H.F. Hu, Ceram. Int. 31 (2005) 305– 314. [10] B. Su, X.J. Wu, Z.H. Chen, Polymer. Mater. Sci. Eng. 10 (2) (1994) 121– 123 (in Chinese). K. Jian et al. / Ceramics International 33 (2007) 905–909 909