Materials Science and Engineering A 497(2008)235-238 Contents lists available at science direct Materials Science and Engineering A ELSEVIER urnalhomepagewww.elsevier.com/locate/msea Effect of thermal cycling on modulus and tensile strength of 3D needled C/Sic composite in controlled environments Jingjiang Nie, Yongdong Xu*, Junqiang Ma, Litong Zhang, Laifei Cheng, Xiaowei Yin National Key Laboratory of Thermostructure Composite Materials, Northwestem Polytechnical University, Xi'an 710072, China ARTICLE INFO A BSTRACT A needled C/SiC composite was subjected to thermal cycling with a temperature interval om700 to 1200 C, in a pure Ar gas and a simulated air atmosphere. The change of the resonant Received in revised form 1 July 2008 Accepted 7 August 2008 used to determine the reduction in the modulus of the composite Results show that the decreased with increasing thermal cycles. And after thermal cycled in pure Ar and simulated air for 90 cles, the residual tensile strengths were reduced by 46. 2% and 66.5%, being caused by thermo-mismatch Keywords and combining effects of thermo-mismatch and oxidation, respectively O 2008 Elsevier B V. All rights reserved. Resonant frequency 1. Introduction form has been widely applied to fabricate C/sic composite of re-entry vehicles, rocket engine components, advanced braking Carbon fiber reinforced silicon carbide(C/SiC)composites are system, etc. [1-3, 12-14. considered as one of the most potential thermal structural mate In most thermo-structural applications, Sic-matrix compos- rials[1-3. In last decades, a lots of efforts have been achieved ites are subjected to thermal cycling and exposed to oxidizing the development and application of the 2D laminated [4-6. 2.5D atmospheres at elevated temperatures. The thermal cycling or [7-9]and multi-D [10-12](3D, 4D, etc. ) braided C/SiC composites. thermo-oxidation may severely degrade the stiffness and strength in plane but are susceptible to delamination. 2. 5D and multi-D fore, the monitoring of their damage during use is desirable so as composites exhibited improved isotropy, good delamination resis- to provide remedies or changes in service conditions before catas- tance and thick part manufacturing capability, but have a complex trophic failure takes place but those damages especially reduction fabrication process with high cost and are not suited for produ in modulus cannot be observed directly. Fortunately, the changes ing complex parts combining both thick and thin areas [3. 12. To in modulus can be determined by monitoring the changes in modal improve the interlaminar properties, save cost and combine the frequencies [15-17. Srivastava et al. [16]reported that the changes advantages of 2D and multi-D preform types, a novel needled pre- in flexural modulus can be determined by measuring the flexural form, Novoltex, with fibers reinforced in the thickness direction resonant frequency The modulus(e) has a fixed relationship with was developed by SEP (Societe Europeenne de Propulsion, now resonant frequency () as the following equation [17] named as Snecma) at the end of 1970s [ 3, 12]. which was pro- luced from woven fabrics and non-woven webs made of carbon E=0.9465x10-(6)(B)/T eedling process, a thermal treatment was used to convert the car- where L, h and b are the specimen dimensions of length, thick- bon precursor to carbon fibers. To avoid thermal shrinkage during ness and width, M is the specimen mass and t is a correction the thermal treatment, carbon fibers are directly used to fabricate factor. Although lots of efforts on thermal cycling behavior of C/ Sic preforms by needling technique. Now, this needled carbon pre- composites have been carried out, the effect of thermal cycling in oxidizing environment on the modulus and tensile strength is still lack of c In this work, a needled C/SiC composite fabricated by chemi- 4619;fax:+862988494620. cal vapor infiltration(Cvi) was subjected to thermal( E-mail address: ydxu07epgmailco temperature interval ranging from 700 to 1200 C, in a pure Ar gas 5093/s-see front matter o 2008 Elsevier B V. All rights reserved. 0016/msea200808009
Materials Science and Engineering A 497 (2008) 235–238 Contents lists available at ScienceDirect Materials Science and Engineering A journal homepage: www.elsevier.com/locate/msea Effect of thermal cycling on modulus and tensile strength of 3D needled C/SiC composite in controlled environments Jingjiang Nie, Yongdong Xu∗, Junqiang Ma, Litong Zhang, Laifei Cheng, Xiaowei Yin National Key Laboratory of Thermostructure Composite Materials, Northwestern Polytechnical University, Xi’an 710072, China article info Article history: Received 2 April 2008 Received in revised form 1 July 2008 Accepted 7 August 2008 Keywords: CVI Needled C/SiC composites Thermal cycling Tensile Resonant frequency abstract A needled C/SiC composite was subjected to thermal cycling with a temperature interval ranging from 700 to 1200 ◦C, in a pure Ar gas and a simulated air atmosphere. The change of the resonant frequencies was used to determine the reduction in the modulus of the composite. Results show that the modulus linearly decreased with increasing thermal cycles. And after thermal cycled in pure Ar and simulated air for 90 cycles, the residual tensile strengths were reduced by 46.2% and 66.5%, being caused by thermo-mismatch and combining effects of thermo-mismatch and oxidation, respectively. © 2008 Elsevier B.V. All rights reserved. 1. Introduction Carbon fiber reinforced silicon carbide (C/SiC) composites are considered as one of the most potential thermal structural materials [1–3]. In last decades, a lots of efforts have been achieved on the development and application of the 2D laminated [4–6], 2.5D [7–9] and multi-D [10–12] (3D, 4D, etc.) braided C/SiC composites. The traditional 2D composites have high mechanical performance in plane but are susceptible to delamination. 2.5D and multi-D composites exhibited improved isotropy, good delamination resistance and thick part manufacturing capability, but have a complex fabrication process with high cost and are not suited for producing complex parts combining both thick and thin areas [3,12]. To improve the interlaminar properties, save cost and combine the advantages of 2D and multi-D preform types, a novel needled preform, Novoltex®, with fibers reinforced in the thickness direction was developed by SEP (Societe Europeenne de Propulsion, now named as Snecma) at the end of 1970s [3,12], which was produced from woven fabrics and non-woven webs made of carbon precursor [12] (such as pre-oxidized polyarylonitrile fibers). After needling process, a thermal treatment was used to convert the carbon precursor to carbon fibers. To avoid thermal shrinkage during the thermal treatment, carbon fibers are directly used to fabricate preforms by needling technique. Now, this needled carbon pre- ∗ Corresponding author. Tel: +86 29 8849 4619; fax: +86 29 88494620. E-mail address: ydxu07@gmail.com (Y. Xu). form has been widely applied to fabricate C/SiC composites for TPS of re-entry vehicles, rocket engine components, advanced braking system, etc. [1–3,12–14]. In most thermo-structural applications, SiC–matrix composites are subjected to thermal cycling and exposed to oxidizing atmospheres at elevated temperatures. The thermal cycling or thermo-oxidation may severely degrade the stiffness and strength of composites and in some cases lead to catastrophic failure. Therefore, the monitoring of their damage during use is desirable so as to provide remedies or changes in service conditions before catastrophic failure takes place, but those damages especially reduction in modulus cannot be observed directly. Fortunately, the changes in modulus can be determined by monitoring the changes in modal frequencies [15–17]. Srivastava et al. [16] reported that the changes in flexural modulus can be determined by measuring the flexural resonant frequency. The modulus (E) has a fixed relationship with resonant frequency (f) as the following equation [17]: E = 0.9465 × 10−9 L h 3 M b f 2T (1) where L, h and b are the specimen dimensions of length, thickness and width, M is the specimen mass and T is a correction factor. Although lots of efforts on thermal cycling behavior of C/SiC composites have been carried out, the effect of thermal cycling in oxidizing environment on the modulus and tensile strength is still lack of comprehension. In this work, a needled C/SiC composite fabricated by chemical vapor infiltration (CVI) was subjected to thermal cycling with a temperature interval ranging from 700 to 1200 ◦C, in a pure Ar gas 0921-5093/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2008.08.009
36 J Nie et al Materials Science and Engineering A 497(2008) 235-238 Fig. 1. Drawing of as-prepared C/SiC specimens for thermal cycling tests in con- trolled environments (all dimensions in mn and a simulated air atmosphere the resonant frequencies were monitored to determine the reduction in the modulus of the com- osite with increasing thermal cycles. And after thermal cycled for 90 cycles, residual tensile strength was measured at room tempe ature to investigate the effect of the thermal cycling in different environments on the mechanical properties of the needled C/Sic Fig. 2. Schematic drawing cooling water, (5)the measurement for temperature, 6) the induction coil for cyclic 2. Experimental procedures temperature, (7) the grip holder and 8)the pressure and flow of the controlled Z.1. Needled C/Sic composite preparation 2.3. Measurement of resonant frequency to determine the relative In the present work, the needled C/SiC composite was fabri cha ated using a three-step and easily controlled process. At first the preform was fabricated by alternatively stacked non-woven cloth The thermal cycling tests were interrupted to measure the res- layers and short-cut-fiber webs using a needling technique and two onant frequency at the given cycles of 10, 30, 50, uccessive non-woven cloth layers were oriented at an angle of vibration tests were conducted at room temperature. One end of 90, which has been described in detail elsewhere [18]. The car- spec was fixed and the other end was excited to vibrate hori- bon fibers used were PAN-based carbon fibers(T700, 12K, from zontally by a modally tuned hammer(PCB 086C05).The responses Toray, Japan). The fiber contents of the non-woven cloth layers, the were monitored using a non-contact laser displacement sensor(LK- ebs and the needling fibers were 21.0, 7.0 and 2.0 vol % respec- G80H, KEYENCE Corp. Japan). Thus, the vibration frequencies of the ively. Secondly, to protect the carbon fibers from damage in the specimens were gained. For a given sample, dimensions, mass, and CVI process and to weaken the interfacial bonding between the shape factors are supposed to be fixed before and after thermal carbon fibers and the Sic matrix[2, 19 a pyrolytic carbon(Pyc) cycling tests. Then, from Eq (1), the relative changes in modulus lyer was deposited on the surface of carbon fibers as fiber/matrix can be derived and expressed as following terphase prior to the densification of SiC matrix. Finally, Cvi was En tions for CvI process were the same as that described in Refs. Eo 18 where Eo and En are the modulus of the composite specimen unquenched and thermal cycled for n cycles, respectively 2. 2. Thermal cycling tests in two controlled environments 2.4. Residual tensile strength measurement and microstructure The dimensions and shape of the specimens are shown in Fig. 1 nalysis nd three specimens were used for thermal cycling test in each Residual tensile strengths of the specimens after thermal cycled Fig. 2, which has been detailed described in Ref (20 In this work, for 90 cycles in both environments were measured on an INSTRON thermal cycling was carried out from 700 to 1200oC controlled by device(8801. INSTRON Ltd. England)according to ASTM C1275 programmable microprocessor. The temperature was measured standard (211. with a crosshead speed of 0.001 mm/s Strains were an infrared pyrometer through a small window in the wall of the recorded using an exter ter with a gauge length of 25 furnace and the wall was internally cut out to enable the circulat- The density and open porosity of the composites were measured ing cold water to reach all over the surfaces. Thermal cycling was by Archimedes' method at room temperature. The microstructures carried out between two selected temperatures and the period was of the fracture surfaces were observed using a scanning electron 180s: holding for 60 s at 700C, heating to 1200C in 60 s and hold- microscope (SEM, JSM-6700F, JEOL, Japan). difference ATA 500.C). Only the middle parts of specimens (about 3. Results and discussion 40-mm long. 3.5-mm wide and 3.5-mm thick ]were kept in the hot zone and were quenched in two typical environments: The density and open porosity of the as-received needled gas; and(ii) a simulated air which was the mixture of Ar(79 vol%) C/Sic composite were 2.1 g/cm and 15%, respectively. And the and o(21 vol%)gas. The flux of gases was accurately controlled microstructure of the composite is shown in Fig 3. It can be found by a mass flow controller (5850 i series of BROOKS in Japan)and its that the adjacent layers were perfectly interlinked together by the precision could reach 0.1 sccm. needling fiber tow
236 J. Nie et al. / Materials Science and Engineering A 497 (2008) 235–238 Fig. 1. Drawing of as-prepared C/SiC specimens for thermal cycling tests in controlled environments (all dimensions in mm). and a simulated air atmosphere. The resonant frequencies were monitored to determine the reduction in the modulus of the composite with increasing thermal cycles. And after thermal cycled for 90 cycles, residual tensile strength was measured at room temperature to investigate the effect of the thermal cycling in different environments on the mechanical properties of the needled C/SiC composites. 2. Experimental procedures 2.1. Needled C/SiC composite preparation In the present work, the needled C/SiC composite was fabricated using a three-step and easily controlled process. At first the preform was fabricated by alternatively stacked non-woven cloth layers and short-cut-fiber webs using a needling technique and two successive non-woven cloth layers were oriented at an angle of 90◦, which has been described in detail elsewhere [18]. The carbon fibers used were PAN-based carbon fibers (T700, 12K, from Toray, Japan). The fiber contents of the non-woven cloth layers, the webs and the needling fibers were 21.0, 7.0 and 2.0 vol.%, respectively. Secondly, to protect the carbon fibers from damage in the CVI process and to weaken the interfacial bonding between the carbon fibers and the SiC matrix [2,19], a pyrolytic carbon (PyC) layer was deposited on the surface of carbon fibers as fiber/matrix interphase prior to the densification of SiC matrix. Finally, CVI was employed to deposit PyC interphase and SiC matrix. The conditions for CVI process were the same as that described in Refs. [11,18]. 2.2. Thermal cycling tests in two controlled environments The dimensions and shape of the specimens are shown in Fig. 1, and three specimens were used for thermal cycling test in each controlled environment. The thermal cycling apparatus is shown in Fig. 2, which has been detailed described in Ref. [20]. In this work, thermal cycling was carried out from 700 to 1200 ◦C controlled by a programmable microprocessor. The temperature was measured by an infrared pyrometer through a small window in the wall of the furnace and the wall was internally cut out to enable the circulating cold water to reach all over the surfaces. Thermal cycling was carried out between two selected temperatures and the period was 180 s: holding for 60 s at 700 ◦C, heating to 1200 ◦C in 60 s and holding for 60 s, then cooling back to 700 ◦C immediately (temperature difference T ≈ 500 ◦C). Only the middle parts of specimens (about 40-mm long, 3.5-mm wide and 3.5-mm thick) were kept in the hot zone and were quenched in two typical environments: (i) pure Ar gas; and (ii) a simulated air which was the mixture of Ar (79 vol.%) and O2 (21 vol.%) gas. The flux of gases was accurately controlled by a mass flow controller (5850 i series of BROOKS in Japan) and its precision could reach 0.1 sccm. Fig. 2. Schematic drawing of atmosphere chamber with a detailed view of the grip holders, the furnace and the specimen (eight major critical points are indicated)[20]. (1) The load alignment, (2) the configuration of the specimen, (3) the heater, (4) the cooling water, (5) the measurement for temperature, (6) the induction coil for cyclic temperature, (7) the grip holder and (8) the pressure and flow of the controlled atmosphere. 2.3. Measurement of resonant frequency to determine the relative change in modulus The thermal cycling tests were interrupted to measure the resonant frequency at the given cycles of 10, 30, 50, 70 and 90. The vibration tests were conducted at room temperature. One end of specimen was fixed and the other end was excited to vibrate horizontally by a modally tuned hammer (PCB 086C05). The responses were monitored using a non-contact laser displacement sensor (LKG80H, KEYENCE Corp., Japan). Thus, the vibration frequencies of the specimens were gained. For a given sample, dimensions, mass, and shape factors are supposed to be fixed before and after thermal cycling tests. Then, from Eq. (1), the relative changes in modulus can be derived and expressed as following: En E0 = fn f0 2 (2) where E0 and En are the modulus of the composite specimen unquenched and thermal cycled for n cycles, respectively. 2.4. Residual tensile strength measurement and microstructure analysis Residual tensile strengths of the specimens after thermal cycled for 90 cycles in both environments were measured on an INSTRON device (8801, INSTRON Ltd., England) according to ASTM C1275 standard [21], with a crosshead speed of 0.001 mm/s. Strains were recorded using an extensometer with a gauge length of 25 mm. The density and open porosity of the composites were measured by Archimedes’ method at room temperature. The microstructures of the fracture surfaces were observed using a scanning electron microscope (SEM, JSM-6700F, JEOL, Japan). 3. Results and discussion The density and open porosity of the as-received needled C/SiC composite were 2.1 g/cm3 and 15%, respectively. And the microstructure of the composite is shown in Fig. 3. It can be found that the adjacent layers were perfectly interlinked together by the needling fiber tows.
J Nie et al. Materials Science and Engineering A 497(2008)235-238 Table 1 Tensile properties of the needled C/SiC composite before and after cycled for 90 Environments Strain(‰) Tensile strength(MPa) As-received composites Short-cut-fiber web Ar gas 0.38 53.2 90° unidirectional I Table 2 The Youngs modulus(E)and expansion coefficient()[5, 22-24 Materials trength(MPa) GPa) (×10-6K-1) T700cr4900 038 Fig. 3. Microstructure of the as-received needled C/SiC composite. of the needled C/SiC composite, including strength, fracture strain and modulus. After quenched in pure Ar gas for 90 cycles, the retain ing rates of strength and strain are 53.5% and 53.8%, respectively. The lower retaining rates of 25. 4% and 33.5% were found for the specimens cycled in the simulated air. This indicates that a more 08 severe degradation was caused by the combination of the thermal cycling and oxidation. 3. 2. Effect of thermal cycling in pure Ar gas It is well known that, SiC matrix, PyC interface and carbon fiber have different expansion coefficients, as shown in Table 215, 22-24]. The differences between these expansion coefficients are very great and cannot beignored. The needled C/Sic composite was fabricated Thermal cycles at 1100 C. When it was cooled to room temperature, the expan- sion coefficient mismatch between carbon fiber, Pyc interface and Fig. 4. Relative reduction in modulus for the specimens cycled in both environ- ents SiC matrix would induce residual thermal stress During the ther- mal cycling tests, these microcracks were promoted by the thermal 3. 1. Degradations in modulus and tensile properties stress in Sic matrix created by coefficients mismatch at low tem- perature. The thermo-stress in SiC matrix, o th(m), can be estimated The relative changes in modulus calculated from the expression by the following expression[25] (2)for specimens quenched in both environments are plotted in oth(m)=Ec x(am-ai)AT thermal cycles. And the reduction rate in modulus of the specimen where Ec is the Young's lus of the composites, AT is the thermal cycled in the simulated air was much greater than that in temperature difterence ue of Ec and AT are 75 GPa [18 the pure Ar gas and500°C,res (3), the thermo-stress in the sic After thermal cycled for 90 cycles in the pure Ar gas and sim- matrix is estimated to be 67.5 MPa. It will promote the microc ulated air, the residual tensile strength were measured, and the racks extending in the Sic matrix, weaken the interfacial bonding. sults are presented in Fig. 5 and Table 1. It is obvious that the or even lead to matrix cracking at the stress concentration sites such as pores and needling potions within the needled c/sic com- thermal cycling plays a great degradation on the tensile propertes posite. Fig. 6 shows the tensile fracture surface of the specimen as-reseived composite simulated a 0 LLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLu 40.5060.7 Strain(%) 100um ig ed oeso ac tesile properties of the needed clsic compost Fig. 6. Tensile fracture surface of the specimen after quenched in pure Ar gas
J. Nie et al. / Materials Science and Engineering A 497 (2008) 235–238 237 Fig. 3. Microstructure of the as-received needled C/SiC composite. Fig. 4. Relative reduction in modulus for the specimens cycled in both environments. 3.1. Degradations in modulus and tensile properties The relative changes in modulus calculated from the expression (2) for specimens quenched in both environments are plotted in Fig. 4. In both environments, modulus decreased with increasing thermal cycles. And the reduction rate in modulus of the specimen thermal cycled in the simulated air was much greater than that in the pure Ar gas. After thermal cycled for 90 cycles in the pure Ar gas and simulated air, the residual tensile strength were measured, and the results are presented in Fig. 5 and Table 1. It is obvious that the thermal cycling plays a great degradation on the tensile properties Fig. 5. Residual tensile properties of the needled C/SiC composite after thermal cycled for 90 cycles. Table 1 Tensile properties of the needled C/SiC composite before and after cycled for 90 cycles in different environments Environments Strain (%) Tensile strength (MPa) As-received composites 0.71 158.9 Ar gas 0.38 85.5 Simulated air 0.18 53.2 Table 2 The Young’s modulus (E) and expansion coefficient (˛) [5,22–24] Materials Strength (MPa) E (GPa) ˛l (×10−6 K−1) ˛r (×10−6 K−1) T700 Cf 4900 230 −0.38 7 PyC – 30 3 26 SiC – 430 4.8 4.8 of the needled C/SiC composite, including strength, fracture strain andmodulus. After quenched in pure Ar gas for 90 cycles, the retaining rates of strength and strain are 53.5% and 53.8%, respectively. The lower retaining rates of 25.4% and 33.5% were found for the specimens cycled in the simulated air. This indicates that a more severe degradation was caused by the combination of the thermal cycling and oxidation. 3.2. Effect of thermal cycling in pure Ar gas It is well known that, SiC matrix, PyC interface and carbon fiber have different expansion coefficients, as shown inTable 2 [5,22–24]. The differences between these expansion coefficients are very great and cannot be ignored. The needled C/SiC composite was fabricated at 1100 ◦C. When it was cooled to room temperature, the expansion coefficient mismatch between carbon fiber, PyC interface and SiC matrix would induce residual thermal stress. During the thermal cycling tests, these microcracks were promoted by the thermal stress in SiC matrix created by coefficients mismatch at low temperature. The thermo-stress in SiC matrix, th(m), can be estimated by the following expression [25]: th(m) = Ec × (am − ai) T (3) where Ec is the Young’s modulus of the composites, T is the temperature difference. The value of Ec and T are 75 GPa [18] and 500 ◦C, respectively. From Eq. (3), the thermo-stress in the SiC matrix is estimated to be 67.5 MPa. It will promote the microcracks extending in the SiC matrix, weaken the interfacial bonding, or even lead to matrix cracking at the stress concentration sites, such as pores and needling potions within the needled C/SiC composite. Fig. 6 shows the tensile fracture surface of the specimen Fig. 6. Tensile fracture surface of the specimen after quenched in pure Ar gas.
38 J Nie et al Materials Science and Engineering A 497(2008) 235-238 Fig. 7. Tensile fracture surface of the specimen quenched in the simulated air. (a) Macro-structure and (b) the gaps between fibers and matrix due to the oxidation of Pyc nterface. carbon fiber and sic matrix after quenched in the pure Ar gas. The long fibers pullout can be with increasing thermal cycles. The degradation in modulus, tensile observed, but the failure strain is much lower than that of the as- strength and failure strain caused by the combination of thermo- received composite (as listed in Table 1). This indicates that the mismatch and oxidation in the simulated air was much more severe terfacial bonding was weakened by the thermal cycling and then than that caused by only thermo-mismatch in the pure Ar gas And reduced the load transfer ability between the fiber and the matrix. the changes in resonant frequency can well reveal the degradatior Combining effects both of internal thermo-mismatch stress and in modulus of the needled C/SiC composite stress concentration caused by matrix cracking lead to the decrease of the modulus and tensile strength of the needled C/SiC composite. Acknowledgement 3.3. Effect of the combination of thermal cycling and oxidation in the simulated air ural Science Foundation of kthe financial support of the Nat- 90405015)and the National Young Elitists Foundation(Contract Fig. 7 shows the tensile fracture surface of the specimen after No. 50425208 hed in the simulated air. Within a bundle, the fiber pullout at the centre (as shown in Fig. 7a) and the gaps between fibers Reference and matrix at the edge(as shown in Fig. 7b)can be observed. The fiber pullout indicates that the bonding of the PyC interface was [11 R. Naslain, Int J AppL. Ceram. TechnoL. 2(2005)75-84. weakened by the thermal cycling. And the gaps indicate that the b E christin. dw En. Mater. 4 012)2002)90 F. Cheng Coat. Technol. 200(2006) mal cycling tests. Therefore, when specimen thermal cycled in 5] G. Camus, L Guillaumat, S Baste, Compos. Sci. TechnoL. 56(1996)1363-1372. unavoidably occurred. At high temperatures(1000-1200oC), the 7 Mo M l. xu dr Zha.. L E Cheng, Se N. .0.ng. Scripta Mater. 54(2006) microcracks were closed and the kinetics of formation of silica is and the condensed oxide scale is protective and tends to seal [8]A. Dalmaz, D. D日Cu9m.题ac or/and fill the residual pores and microcracks, stopping(or at least 910.Siron, j. Pailhes, ). Lamon, Compos. Sci. Technol. 59( wing down the in-depth diffusion of oxygen [2]. Conversely, at [10] Y.D. Xu, LF. Cheng. LT Zhang, H.F. Yin, xW. Yin, Mater Sci Eng. A 300(2001 w temperatures, 700<T<900 C, microcracks were opened and this provided the channel for O2 to aggress the Pyc interface and [11 Y.D. Xu, LT Zhang, LF. Cheng. D T Yan, Carbon 36(1998)1051-1056. 12 F.A. Christin then the carbon fibers And the kinetics of oxidation of the PyCinter- [13] Y D Xu, Y.N. Zhang LF. Cheng. LT. Zhang, J]. Lou, JZ. Zhang Ceram.Int33 phase and carbon fibers in the composite is fast whereas that of Sic (2007)439-445. is almost negligible [2]. As shown in Fig. 7b, PyC interphase, cal [14 S.W. Fan, LT. Zhang, Y.D. Xu, LF Cheng, J. Lou, J-Z. Zhang, L Yu, Compos. Sci. echnol67(20072390-2398 bon fiber and Sic matrix were oxidized, leading to gaps between [15].H. Diazvaldes, C. Soutis, ]. Sound Vibrat. 228(1999)1-9 ber and matrix. The oxidation of carbon fibers resulted in the [16 vK srivastava, K Maile, K Bothe, A Udoh, Mater. Sci Eng. A 354(2003 soor load transfer between the fiber and the matrix. Therefore, the n bas sd tec nl 1b2 01 19pm g3 B. Patelb, C. cCofhinb, J. Eldridge. damage caused by the combination of thermo-mismatch and oxi- [18] J.J. Nie Y.D. Xu, LT Zhang. LE Cheng J.Q. Ma. ]. Mater. Proc. Technol, in press. dation in the simulated air was more severe than that in the pure[19] R. Naslain, Comp Ar environment. A29(1998)1145-1155. 4. Conclusions 22 h 3D needled C/SiC composite fabricated by CVI process subjected [23 A. Eckel, R.C. Brad, J Am Ceramic. Soc. 73(1990)1334-1 to thermal cycling from 700 to 1200C in a pure Ar gas and a L Bobet, ]. Lamon, Acta Metall. Mater. 43(1995)2241-22 simulated air. In both environments, the modulus was decreased 12515R Qiao, D Han, G.Q. Luo, Key Eng Mater. 297-300(2005)435-439
238 J. Nie et al. / Materials Science and Engineering A 497 (2008) 235–238 Fig. 7. Tensile fracture surface of the specimen quenched in the simulated air. (a) Macro-structure and (b) the gaps between fibers and matrix due to the oxidation of PyC interface, carbon fiber and SiC matrix. after quenched in the pure Ar gas. The long fibers pullout can be observed, but the failure strain is much lower than that of the asreceived composite (as listed in Table 1). This indicates that the interfacial bonding was weakened by the thermal cycling and then reduced the load transfer ability between the fiber and the matrix. Combining effects both of internal thermo-mismatch stress and stress concentration caused by matrix cracking lead to the decrease of the modulus and tensile strength of the needled C/SiC composite. 3.3. Effect of the combination of thermal cycling and oxidation in the simulated air Fig. 7 shows the tensile fracture surface of the specimen after quenched in the simulated air. Within a bundle, the fiber pullout at the centre (as shown in Fig. 7a) and the gaps between fibers and matrix at the edge (as shown in Fig. 7b) can be observed. The fiber pullout indicates that the bonding of the PyC interface was weakened by the thermal cycling. And the gaps indicate that the oxidation of PyC interface and fibers occurred during the thermal cycling tests. Therefore, when specimen thermal cycled in the simulated air, in addition to thermo-mismatch, oxidation was unavoidably occurred. At high temperatures (1000–1200 ◦C), the microcracks were closed and the kinetics of formation of silica is fast and the condensed oxide scale is protective and tends to seal or/and fill the residual pores and microcracks, stopping (or at least slowing down) the in-depth diffusion of oxygen [2]. Conversely, at low temperatures, 700 < T < 900 ◦C, microcracks were opened and this provided the channel for O2 to aggress the PyC interface and then the carbon fibers. And the kinetics of oxidation of the PyC interphase and carbon fibers in the composite is fast whereas that of SiC is almost negligible [2]. As shown in Fig. 7b, PyC interphase, carbon fiber and SiC matrix were oxidized, leading to gaps between fiber and matrix. The oxidation of carbon fibers resulted in the loss of fiber strength, and the oxidation of PyC interface led to the poor load transfer between the fiber and the matrix. Therefore, the damage caused by the combination of thermo-mismatch and oxidation in the simulated air was more severe than that in the pure Ar environment. 4. Conclusions 3D needled C/SiC composite fabricated by CVI process subjected to thermal cycling from 700 to 1200 ◦C in a pure Ar gas and a simulated air. In both environments, the modulus was decreased with increasing thermal cycles. The degradation in modulus, tensile strength and failure strain caused by the combination of thermomismatch and oxidation in the simulated air was much more severe than that caused by only thermo-mismatch in the pure Ar gas. And the changes in resonant frequency can well reveal the degradation in modulus of the needled C/SiC composite. Acknowledgements The authors acknowledge the financial support of the Natural Science Foundation of China (Contract Nos. 50672076 and 90405015) and the National Young Elitists Foundation (Contract No. 50425208). References [1] R. Naslain, Int. J. Appl. Ceram. Technol. 2 (2005) 75–84. [2] R. Naslain, Compos. Sci. Technol. 64 (2004) 155–170. [3] F. Christin, Adv. Eng. Mater. 4 (12) (2002) 903–912. [4] S.J. Wu, L.F. Cheng, L.T. Zhang, Y.D. Xu, Surf. Coat. Technol. 200 (2006) 4489–4492. [5] G. Camus, L. Guillaumat, S. Baste, Compos. Sci. Technol. 56 (1996) 1363–1372. [6] M. Wang, C. Laird, Acta Mater. 44 (1996) 1371–1387. [7] J.Q. Ma, Y.D. Xu, L.T. Zhang, L.F. Cheng, J.J. Nie, N. Dong, Scripta Mater. 54 (2006) 1967–1971. [8] A. Dalmaz, D. Ducret, R. El Guerjouma, P. Reynaud, P. Franciosi, D. Rouby, G. Fantozzi, J.C. Baboux, Compos. Sci. Technol. 60 (2000) 913–925. [9] O. Siron, J. Pailhes, J. Lamon, Compos. Sci. Technol. 59 (1999) 1–12. [10] Y.D. Xu, L.F. Cheng, L.T. Zhang, H.F. Yin, X.W. Yin, Mater. Sci. Eng. A 300 (2001) 196–202. [11] Y.D. Xu, L.T. Zhang, L.F. Cheng, D.T. Yan, Carbon 36 (1998) 1051–1056. [12] F.A. Christin, Int. J. Appl. Ceram. Technol. 2 (2005) 97–104. [13] Y.D. Xu, Y.N. Zhang, L.F. Cheng, L.T. Zhang, J.J. Lou, J.Z. Zhang, Ceram. Int. 33 (2007) 439–445. [14] S.W. Fan, L.T. Zhang, Y.D. Xu, L.F. Cheng, J.J. Lou, J.Z. Zhang, L. Yu, Compos. Sci. Technol. 67 (2007) 2390–2398. [15] S.H. Diazvaldes, C. Soutis, J. Sound Vibrat. 228 (1999) 1–9. [16] V.K. Srivastava, K. Maile, K. Bothe, A. Udoh, Mater. Sci. Eng. A 354 (2003) 292–297. [17] N. Chawlaa, K.K. Chawlab, M. Koopmanb, B. Patelb, C. Coffinb, J.I. Eldridgec, Compos. Sci. Technol. 61 (2001) 1923–1930. [18] J.J. Nie, Y.D. Xu, L.T. Zhang, L.F. Cheng, J.Q. Ma, J. Mater. Proc. Technol., in press, doi:10.1016/j.jmatprotec.2008.02.035. [19] R. Naslain, Composites A 29 (1998) 1145–1155. [20] H. Mei, L.F. Cheng, L.T. Zhang, Scripta Mater. 54 (2006) 163–168. [21] ASTM C1275-00, Standard test method for monotonic tensile behavior of continuous fiber-reinforced advanced ceramics with solid rectangular crosssection test specimens at ambient temperature. [22] http://www.torayusa.com/cfa/pdfs/T700SDataSheet.pdf. [23] A.J. Eckel, R.C. Brad, J. Am. Ceramic. Soc. 73 (1990) 1334–1338. [24] J.L. Bobet, J. Lamon, Acta Metall. Mater. 43 (1995) 2241–2253. [25] S.R. Qiao, D. Han, G.Q. Luo, Key Eng. Mater. 297–300 (2005) 435–439