CARBON PERGAMON Carbon41(2003)707711 Thermal diffusivity of 3d C/ Sic composites from room temperature to1400°C Laifei Cheng", Yongdong Xu, Qing Zhang, Litong Zhang State Key Laboratory of Solidification Processing, Northwestern Polytechnical University, Xi an Shaanxi 710072, China Received 1 April 2002; accepted 17 October 2002 Abstract A 3D C/SiC composite and a bulk CVd SiC material were prepared. The effects of the CVD SiC coating and the heat reatment on the longitudinal and transverse thermal diffusivity of the C/Sic composites were investigated. The thermal diffusivity of the C/SiC composites could be well fitted by a multinomial function from room temperature to 1400C which ncludes a power term, an exponential term and a constant term. The exponential term affected the thermal diffusivity and led to its increase above 1200C with activation energy of 77 kcal/ mol. The microstructure change in the composites was the reason that the thermal diffusivity was increased above 1200C. The longitudinal thermal diffusivity of the composite was twice or more than the transverse one and increased more rapidly by the exponential term. The former was decreased by the CVD SiC coating, but the latter was increased by it. The heat treatment could increase the thermal diffusivity and make the exponential term disappeared in the functions. The functional curve before the treatment intersected that after the treatment at the treatment temperature C 2002 Elsevier Science Ltd. All rights reserved Keywords: A. Carbon composites; D. Thermal diffusivity; D. Microstructure 1. Introduction where p(kg m)is the density and c,o Kgk )the specific heat at constant pressure. Because the SiC matrix, Carbon fiber reinforced silicon carbide composites(C/ PyC interlayer, fibers and pores are included in a 3D SiC)have been developed and tested for thermostructural C/SiC composite, its thermal diffusivity can be calculated plications such as the components of turbine engines, the by [4 reentry thermal protection system of spacecraft, ultra-light- of the most important properties for all these application Although some data is reported in literatures on thermal where u, is the volume fraction of any phase, a" is the diffusivity of C/SiC composites [3], it has not been thermal diffusivity of the composite and a is that of the The thermal diffusivity (ms)of a material is direction of the thermal flow. The value of m s Adthe systematically investiga phase, n is related to the preform structure an ween related to its thermal conductivity k(wmk )by the -1 and ollowing equation [4 The thermal diffusivity of a 3D C/SiC composite is very sensitive to changes of its microstructure and composition. k t is possible to determine the mechanisms for the micro- structural and compositional changes of the composite quantitatively by systematically measuring its thermal diffusivity. Obviously, this is significant to microstructure R Corresponding author. Tel: +86-29-849-4616; fax:+86-29. analysis of the composite. In this paper, a 3D C/Sic composite and a bulk CVD SiC material were prepared E-mail address: chenglfnwpu. edu.cn(L. Cheng). The effects of the Cvd Sic coating and the heat treatment 0008-6223/02/S-see front matter 2002 Elsevier Science Ltd. All rights reserved doi:1o.1016S0008-6223(02)00382-2
Carbon 41 (2003) 707–711 T hermal diffusivity of 3D C/SiC composites from room temperature to 1400 8C Laifei Cheng , Yongdong Xu, Qing Zhang, Litong Zhang * State Key Laboratory of Solidification Processing, Northwestern Polytechnical University, Xi’an Shaanxi 710072, China Received 1 April 2002; accepted 17 October 2002 Abstract A 3D C/SiC composite and a bulk CVD SiC material were prepared. The effects of the CVD SiC coating and the heat treatment on the longitudinal and transverse thermal diffusivity of the C/SiC composites were investigated. The thermal diffusivity of the C/SiC composites could be well fitted by a multinomial function from room temperature to 1400 8C which includes a power term, an exponential term and a constant term. The exponential term affected the thermal diffusivity and led to its increase above 1200 8C with activation energy of 77 kcal/mol. The microstructure change in the composites was the reason that the thermal diffusivity was increased above 1200 8C. The longitudinal thermal diffusivity of the composite was twice or more than the transverse one and increased more rapidly by the exponential term. The former was decreased by the CVD SiC coating, but the latter was increased by it. The heat treatment could increase the thermal diffusivity and make the exponential term disappeared in the functions. The functional curve before the treatment intersected that after the treatment at the treatment temperature. 2002 Elsevier Science Ltd. All rights reserved. Keywords: A. Carbon composites; D. Thermal diffusivity; D. Microstructure 23 21 21 1. Introduction where r (Kg m ) is the density and c (J Kg K ) the p specific heat at constant pressure. Because the SiC matrix, Carbon fiber reinforced silicon carbide composites (C/ PyC interlayer, fibers and pores are included in a 3D SiC) have been developed and tested for thermostructural C/SiC composite, its thermal diffusivity can be calculated applications such as the components of turbine engines, the by [4] reentry thermal protection system of spacecraft, ultra-light- n n weight mirrors and so on [1,2]. Thermal diffusivity is one a 5 Ov a (2) C i i of the most important properties for all these applications. n where v is the volume fraction of any phase, a is the Although some data is reported in literatures on thermal i C n thermal diffusivity of the composite and a is that of the diffusivity of C/SiC composites [3], it has not been i phase, n is related to the preform structure and the systematically investigated. 2 21 direction of the thermal flow. The value of n is between The thermal diffusivity a (m s ) of a material is 21 21 21 and 11. related to its thermal conductivity k (W m K ) by the The thermal diffusivity of a 3D C/SiC composite is very following equation [4] sensitive to changes of its microstructure and composition. k It is possible to determine the mechanisms for the micro- a 5 ] (1) rc structural and compositional changes of the composite p quantitatively by systematically measuring its thermal diffusivity. Obviously, this is significant to microstructure *Corresponding author. Tel.: 186-29-849-4616; fax: 186-29- analysis of the composite. In this paper, a 3D C/SiC 849-4620. composite and a bulk CVD SiC material were prepared. E-mail address: chenglf@nwpu.edu.cn (L. Cheng). The effects of the CVD SiC coating and the heat treatment 0008-6223/02/$ – see front matter 2002 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0008-6223(02)00382-2
L. Cheng et al./ Carbon 41(2003)707-711 on the longitudinal and transverse thermal diffusivity of 3. Results and discussion the C/SiC composite with and without a CVD SiC coating were investigated 3. I. Thermal difusivity of CVD Sic The thermal diffusivity as of the CVd Sic prepared in this paper was compared with that a s of the CVD SiC 2. Experimental procedure prepared by rohm and Haas Company and was shown in Fig. 1. Although the latter was much larger than the 2.I. Fabrication of the specimens ormer, they could be well fitted by a simple function over the full temperature range Two kinds of specimens were machined from a 3I C/SiC composite prepared by low-pressure chemical vapor 994 deposition method (LPCVD), one being along the fiber 1.35 +00285 (3) axial and the other cross the axial.a bulk cvd sic specimen prepared by usual pressure chemical vapor 48800 deposition method (UPCVD) was used to investigate the a p1.8+00700 thermal diffusivity of C/SiC composites The deposition conditions of bulk SiC material were as where T is the Kelvin temperature. It could be seen that a follow. temperature 1 100C, time 200 h, H, flow 400 ml decreased much rapidly than as with increasing tempera min. Ar flow 400 ml min and the molar ratio of h ture. The higher the temperature, the smaller the difference and MTS was 3-5. The 3D preforms were deposited with between as and a s pyrolytic carbon (PyC) and Sic using butane and The bulk CVD SiC material used in this paper was very methyltrichlorosilane(MTS). The deposition conditions of different from that prepared by rohm and Haas Company PyC interlayer were as follow: temperature 960C, pre in microstructure and composition due to different depo sure 5 KPa, time 20 h, Ar flow 200 ml min, butane flow tion conditions [5, 6]. When the molar ratio of H, and MTs 15 ml min. The deposition conditions of Sic matrix is about 10, stoichiometric silicon carbide can be eventual- were as follow: temperature 1000C, pressure 5 KPa, time ly obtained. Because this ratio under the UPCVd was only 120 h, H, flow 350 ml min Ar flow 350 ml min and 3-5, the Sic deposited was rich in carbon. The XRD the molar ratio of H, and MTs was 10. A CVD Sic analysis confirmed that the free carbon existed in graphit coating was prepared on the C/SiC specimens after (Fig. 2). The composition of the Sic was calculated to be machining under the same conditions as the sic matrix to 88% B-SiC, 4% a-SiC and 7% graphite. Some defects investigate its effect on the thermal diffusivity. Although where the Sic was loosely packed could be found owing to the deposition conditions of the bulk SiC material were the rapid deposition under UPCVD(Fig. 3). The defects different from those of the SiC matrix and the Sic coating, and the graphite made as much lower than a s. Of course, thermal diffusivity measurement of the bulk material will it would take very short time to prepare a bulk si be helpful to investigate that of the c/Sic composit material by UPCVD 2.2. Diffusivity measurements Laser Flash Apparatus LFA 427 made by NETZSCH company was employed for measurements of the 1.4 o CVD SiC Ref diffusivity. The required specimen size was 12.5X2 mm. All measurements were conducted in Argon atmos- 1400C. In order to investigate the effect of the heat treatment on the thermal diffusivity, all specimens were measured two or three times. The thermal diffusivity at each appointed temperature was measured three times Below 1000C. the relative error Aala is less than 0.5% and above 1000C it is less than 0. 1%. So the measuring 500 00 1500 error have not effect on the relations of thermal diffusivity to temperature. Because the uncertainty of sample thick- ness is less than 1%, it leads to about 2% uncertainty of Fig. 1. Thermal diffusivity of CVD SiC as a function of tempera- ture
708 L. Cheng et al. / Carbon 41 (2003) 707–711 on the longitudinal and transverse thermal diffusivity of 3. Results and discussion the C/SiC composite with and without a CVD SiC coating were investigated. 3 .1. Thermal diffusivity of CVD SiC The thermal diffusivity aS of the CVD SiC prepared in this paper was compared with that a9 S of the CVD SiC 2. Experimental procedure prepared by Rohm and Haas Company and was shown in Fig. 1. Although the latter was much larger than the 2 .1. Fabrication of the specimens former, they could be well fitted by a simple function over the full temperature range Two kinds of specimens were machined from a 3D C/SiC composite prepared by low-pressure chemical vapor 994 a 5 1 ]] 0.0285 (3) S 1.35 deposition method (LPCVI), one being along the fiber T axial and the other cross the axial. A bulk CVD SiC specimen prepared by usual pressure chemical vapor 48800 a9 5 1 ]] 0.0700 (4) S 1.8 deposition method (UPCVD) was used to investigate the T thermal diffusivity of C/SiC composites. The deposition conditions of bulk SiC material were as where T is the Kelvin temperature. It could be seen that a9 S follow: temperature 1100 8C, time 200 h, H flow 400 ml decreased much rapidly than a with increasing tempera- 2 S 21 21 min , Ar flow 400 ml min and the molar ratio of H ture. The higher the temperature, the smaller the difference 2 and MTS was 3–5. The 3D preforms were deposited with between a and a9. S S pyrolytic carbon (PyC) and SiC using butane and The bulk CVD SiC material used in this paper was very methyltrichlorosilane (MTS). The deposition conditions of different from that prepared by Rohm and Haas Company PyC interlayer were as follow: temperature 960 8C, pres- in microstructure and composition due to different deposi- 21 sure 5 KPa, time 20 h, Ar flow 200 ml min , butane flow tion conditions [5,6]. When the molar ratio of H and MTS 2 21 15 ml min . The deposition conditions of SiC matrix is about 10, stoichiometric silicon carbide can be eventualwere as follow: temperature 1000 8C, pressure 5 KPa, time ly obtained. Because this ratio under the UPCVD was only 21 21 120 h, H flow 350 ml min , Ar flow 350 ml min and 3–5, the SiC deposited was rich in carbon. The XRD 2 the molar ratio of H and MTS was 10. A CVD SiC analysis confirmed that the free carbon existed in graphite 2 coating was prepared on the C/SiC specimens after (Fig. 2). The composition of the SiC was calculated to be machining under the same conditions as the SiC matrix to 88% b-SiC, 4% a-SiC and 7% graphite. Some defects investigate its effect on the thermal diffusivity. Although where the SiC was loosely packed could be found owing to the deposition conditions of the bulk SiC material were the rapid deposition under UPCVD (Fig. 3). The defects different from those of the SiC matrix and the SiC coating, and the graphite made a much lower than a9. Of course, S S thermal diffusivity measurement of the bulk material will it would take very short time to prepare a bulk SiC be helpful to investigate that of the C/SiC composite. material by UPCVD. 2 .2. Diffusivity measurements Laser Flash Apparatus LFA 427 made by NETZSCH company was employed for measurements of the thermal diffusivity. The required specimen size was f12.532 2 mm . All measurements were conducted in Argon atmosphere at different temperatures from room temperature to 1400 8C. In order to investigate the effect of the heat treatment on the thermal diffusivity, all specimens were measured two or three times. The thermal diffusivity at each appointed temperature was measured three times. Below 1000 8C, the relative error Da/a is less than 0.5%, and above 1000 8C it is less than 0.1%. So the measuring error have not effect on the relations of thermal diffusivity to temperature. Because the uncertainty of sample thickness is less than 1%, it leads to about 2% uncertainty of Fig. 1. Thermal diffusivity of CVD SiC as a function of temperadiffusivity. ture
L. Cheng et al. /Carbon 41(2003)707-71I UOU thermal diffusivity acs of the C/Sic and that a scs of the SiC/C/SiC could be also well fitted by a function over the full temperature range 16.4 00100·cxp(-14×10-7272)+0484 (5) Fig. 2. XRD patterns of the bulk Sic material prepared by ascs =0.9-0.0074exp(-1.4X10-T2)+0.0416 UPCVD (6) It could seen tha a l was a little larger than al This indicated that the diffusivity of the CVD Si coating is lower than longitudinal one of the fibers, and then it could not increase the thermal diffusivity of the composite It should be noted that there appeared an exponential term in Eqs. (5)and(6) compared with Eqs. (3)and (4), which corresponds to a new mechanism and led to the thermal diffusivity increase after rapid decrease with increasing temperature. The exponential term showed that the mechanism affected the thermal diffusivity above 1200C with activation energy of 77 kcal/ mol. The mechanism should be considered to be the microstructure change in the composites at high temperatures, otherwise 39429KVX2,9910yW039 the thermal diffusivity should be monotone decreasing Fig. 3. S.E.M. photograph of the bulk SiC material prepared by UPCVD 3.3. Transverse thermal diffusivity of the composite Fig. 5 showed the relations of the transverse thermal 3. 2. Longitudinal thermal diffusivity of the composite diffusivity of the C/SiC and SiC/C/SiC to temperature Similarly, the thermal diffusivity acs of the C/SiC and Fig. 4 showed the relations of the longitudinal thermal that a scs of the Sic/C/SiC could be well fitted by a diffusivity of the C/SiC and SiC/C/Sic composite to function over the full temperature range temperature. Although they were more complicated, the o 3D-C/SiC= 0.12 o 3D-SiC/C/SiClI(g) g0.08 500 1000 500 1000 Temperature(c) Temperature(C) Fig. 4. The longitudinal thermal diffusivity of the C/SiC and Fig. 5. The transverse thermal diffusivity of the C/SiC and SiC/C/SiC composite as a function of temperature SiC/C/SiC composite as a function of temperature
L. Cheng et al. / Carbon 41 (2003) 707–711 709 i i thermal diffusivity a of the C/SiC and that a of the CS SCS SiC/C/SiC could be also well fitted by a function over the full temperature range 16.46 i ] 272 23 a CS 5 2 ] 0.0100 ? exps d 2 1.4 3 10 T 1 0.0484 0.9 T (5) 15.13 i ] 272 23 Fig. 2. XRD patterns of the bulk SiC material prepared by a SCS 5 2 ] 0.0074 ? exps d 2 1.4 3 10 T 1 0.0416 0.9 T UPCVD. (6) i i It could be seen that a was a little larger than a . CS SCS This indicated that the thermal diffusivity of the CVD SiC coating is lower than longitudinal one of the fibers, and then it could not increase the thermal diffusivity of the composite. It should be noted that there appeared an exponential term in Eqs. (5) and (6) compared with Eqs. (3) and (4), which corresponds to a new mechanism and led to the thermal diffusivity increase after rapid decrease with increasing temperature. The exponential term showed that the mechanism affected the thermal diffusivity above 1200 8C with activation energy of 77 kcal/mol. The mechanism should be considered to be the microstructure change in the composites at high temperatures, otherwise the thermal diffusivity should be monotone decreasing. Fig. 3. S.E.M. photograph of the bulk SiC material prepared by 3 .3. Transverse thermal diffusivity of the composite UPCVD. Fig. 5 showed the relations of the transverse thermal 3 .2. Longitudinal thermal diffusivity of the composite diffusivity of the C/SiC and SiC/C/SiC to temperature. 5 Similarly, the thermal diffusivity a CS of the C/SiC and 5 Fig. 4 showed the relations of the longitudinal thermal that a of the SiC/C/SiC could be well fitted by a SCS diffusivity of the C/SiC and SiC/C/SiC composite to function over the full temperature range temperature. Although they were more complicated, the Fig. 4. The longitudinal thermal diffusivity of the C/SiC and Fig. 5. The transverse thermal diffusivity of the C/SiC and SiC/C/SiC composite as a function of temperature. SiC/C/SiC composite as a function of temperature
710 L. Cheng et al./ Carbon 41(2003)707-711 treatment, the longitudinal thermal diffusivity was found to 13-00020exp(-14×10-272)+00130 be increased and could be well fitted by the following (7) 1549 19.38 0.0020·exp(-14×10-2723)+0.0163 709+00396 The experimental curve of the thermal diffusivity obtained by the second measurement was coincident with that Certainly, the microstructure change affected the trans- obtained by the third measurement. As a result, the thermal diffusivity could not be increased by further treatment any verse appeared in Eqs. (7) and = ne threshold more, and then it could be given by the same function. The exponential term disappeared in this function, and the first curve intersected the second curve at 1400C which was tudinal thermal diffusivity increased the treatment temperature. These would not be difficult to increasing temperature than the transverse one. Therefore, understand if the microstructure change in the composit the former was more sensitive to the microstructure change caused by the heat treatment was non-reversible in the composites than the latter The transverse thermal diffusivity of the SiC/c/Sic Because ascs was large acs, the thermal dif. obtained by the first measurement was compared with that fusivity of the coating wa obtained by the second measurement in Fig. 7. Similarl the fibers. and then it could the thermal diffusivity the thermal diffusivity after the treatment could be wel of the composite. Consequently, the effect of the coating fitted by the following function on the transverse thermal diffusivity of the C/Sic wa contrary to that of longitudinal one 25.76 +0.0167 Figs. 4 and 5 showed that the longitudinal thermal diffusivity of the composite was much larger than trans- For the same reason, the exponential term disappeared in verse one. This is reasonable because the axial thermal this function, and the first curve intersected the second At roo ity of fibers is half an order larger than radical one. curve at 1400C. Although the thermal diffusivity was At room temperature, both the longitudinal and transverse increased, the difference of the transverse thermal dif- thermal diffusivity was a little lower than the reported data. fusivity produced by the treatment was much smaller than that of the longitudinal one over the whole temperature 3.4. Effect of heat treatment on the thermal diffusivity range. This should be easy to understand because the increase of the former produced by the exponential term The longitudinal thermal diffusivity of the SiC/C/Sic was much smaller than that of the latter obtained by the first measurement was compared with The thermal diffusivity of the CVD Sic was not those obtained by the second and the third measurement in increased by the treatment(Fig. 1), and that of the Fig. 6. The first measurement acted on the second one as a would not be changed by the treatment also heat treatment, so did the second on the third. After the 1400 C. Consequently, the microstructure change of the Pyc interlayer produced by the treatment should 0,16 p3 D-SiC/C/SiC‖(g) o 3D-siC/C/(g) o 30./C/SiC= 005 F0.01 0 1c00 1500 500 Fig. 6. Effect of the heat treatment on the longitudinal thermal Fig. 7. Effect of the heat treatment on the transverse thermal diffusivity of the Sic/C/SiC composite diffusivity of the SiC/C/SiC composite
710 L. Cheng et al. / Carbon 41 (2003) 707–711 treatment, the longitudinal thermal diffusivity was found to 5 2 10.60 ] 72 23 a CS 5 2 ] 0.0020 ? exps d 2 1.4 3 10 T 1 0.0130 be increased and could be well fitted by the following 1.15 T function (7) i 15.49 a 9 5 1 ]] 0.0396 (9) SCS 0.9 5 2 19.38 72 23 T a SCS 5 2 ]] 0.0020 ? exps d 2 1.4 3 10 T 1 0.0163 1.1 T The experimental curve of the thermal diffusivity obtained (8) by the second measurement was coincident with that obtained by the third measurement. As a result, the thermal Certainly, the microstructure change affected the trans- diffusivity could not be increased by further treatment any verse thermal diffusivity, and then an exponential term more, and then it could be given by the same function. The appeared in Eqs. (7) and (8) with the same threshold exponential term disappeared in this function, and the first temperature and activation energy. However, the longi- curve intersected the second curve at 1400 8C which was tudinal thermal diffusivity increased more rapidly with the treatment temperature. These would not be difficult to increasing temperature than the transverse one. Therefore, understand if the microstructure change in the composite the former was more sensitive to the microstructure change caused by the heat treatment was non-reversible. in the composites than the latter. 5 5 The transverse thermal diffusivity of the SiC/C/SiC Because a was larger than a , the thermal dif- SCS CS obtained by the first measurement was compared with that fusivity of the coating was larger than the transverse one of obtained by the second measurement in Fig. 7. Similarly, the fibers, and then it could increase the thermal diffusivity the thermal diffusivity after the treatment could be well of the composite. Consequently, the effect of the coating fitted by the following function on the transverse thermal diffusivity of the C/SiC was 5 25.76 contrary to that of longitudinal one. a 9 5 1 ]] 0.0167 (10) SCS 1.2 Figs. 4 and 5 showed that the longitudinal thermal T diffusivity of the composite was much larger than trans- For the same reason, the exponential term disappeared in verse one. This is reasonable because the axial thermal this function, and the first curve intersected the second diffusivity of fibers is half an order larger than radical one. curve at 1400 8C. Although the thermal diffusivity was At room temperature, both the longitudinal and transverse increased, the difference of the transverse thermal difthermal diffusivity was a little lower than the reported data. fusivity produced by the treatment was much smaller than that of the longitudinal one over the whole temperature 3 .4. Effect of heat treatment on the thermal diffusivity range. This should be easy to understand because the increase of the former produced by the exponential term The longitudinal thermal diffusivity of the SiC/C/SiC was much smaller than that of the latter. obtained by the first measurement was compared with The thermal diffusivity of the CVD SiC was not those obtained by the second and the third measurement in increased by the treatment (Fig. 1), and that of the fibers Fig. 6. The first measurement acted on the second one as a would not be changed by the treatment also below heat treatment, so did the second on the third. After the 1400 8C. Consequently, the microstructure change of the PyC interlayer produced by the treatment should be Fig. 6. Effect of the heat treatment on the longitudinal thermal Fig. 7. Effect of the heat treatment on the transverse thermal diffusivity of the SiC/C/SiC composite. diffusivity of the SiC/C/SiC composite
L. Cheng et al. /Carbon 41(2003)707-71I responsible for the thermal diffusivity increase of the The heat treatment could increase the thermal diffusivity C/SiC composite. It could be inferred that the graphitiza- and make the exponential term disappeared in the func- tion of the PyC interlayer took placed partially in the tions. The functional curve before the treatment intersected composite. Because graphite had a much larger thermal that after the treatment at the treatment temperature diffusivity than both the Pyc interlayer and the fibers, the graphitization could increase the thermal diffusivity of the composite. Unfortunately, no obvious graphitization after Acknowledgements the treatment was observed by the TEM analysis. Thermo- dynamically, the graphitization of the Pyc interlay The authors acknowledge the support of the Chinese possible in the treatment. A reasonable explanation is National Foundation for Natural Sciences under Contract the interfacial stress produced by thermal expansion mis- No.59772023. match of the fibers and the matrix induced the graphitize- tion. The effect of heat treatment on the thermal diffusivity of a C/SiC composite could be further enhanced by References increasing thickness of the PyC interlayer, it is limited owing to its low volume fraction. Although the micro- [1 Jamet JM, Lmicq P, Schiroky GH. Composite thermo- structure change in a C/SiC composite can be confirmed structures: an overview of the French experience. In: Naslain by its thermal diffusivity increase at high temperatures, it R, editor, High temperature ceramic matrix composites, eminds to be explained by further research. Bordeaux: Woodhead Publications, 1993, pp. 215-29 2] Laux T, Ullmann T, Auweter-Kurtz M, Hald H, Kurz A Investigation of thermal protection materials along an X-38 4. Conclusions re-entry trajectory by plasma wind tunnel simulations, Sec- ond Intermational Symposium on Atmospheric Reentry Ve- The longitudinal and transverse thermal diffusivity of hicles and Systems. Arcahon( France): 2001: 1-9 3] Tawil H, Bentsen LD, Baskaran S, Hasselman DPH. Thermal the C/SiC composite with and without a CVd SiC coating diffusivity of chemically vapor deposited silicon carbide could be well fitted by a multinomial function from room reinforced with silicon carbide or carbon fibers. j Mater Sci temperature to 1400C which includes a power term,an 1985;20:3201-12 exponential term and a constant term. 14] Parker WJ, Jenkins R, Butler CP, Abbott GL. Flash method The exponential term affected the thermal diffusivity of determining thermal diffusivity, heat capacity, and thermal and led to its increase above 1200C with activation conductivity. J Appl Phys 1961; 32(9): 1679-84 energy of 77 kcal/mol. The microstructure change in the 5]Kim Y, Zangvil A, Goela JS, Taylor RL. Microstructure composite was the reason that the thermal diffusivity was mparison of transparent and opaque CVD SiC. JAm Ceram Soc1995;78:1571-9 increased above 1200° [6] Goela JS, Pichering MA, Burns LE. Chemical The longitudinal thermal diffusivity of the C/SiC Sited Sic for high heat flux applications omposite was twice or more than the transverse one and 1996;28552-13 increased more rapidly by the exponential term. The former was decreased by the CVD SiC coating, but the latter was increased by it
L. Cheng et al. / Carbon 41 (2003) 707–711 711 responsible for the thermal diffusivity increase of the The heat treatment could increase the thermal diffusivity C/SiC composite. It could be inferred that the graphitiza- and make the exponential term disappeared in the function of the PyC interlayer took placed partially in the tions. The functional curve before the treatment intersected composite. Because graphite had a much larger thermal that after the treatment at the treatment temperature. diffusivity than both the PyC interlayer and the fibers, the graphitization could increase the thermal diffusivity of the composite. Unfortunately, no obvious graphitization after Acknowledgements the treatment was observed by the TEM analysis. Thermodynamically, the graphitization of the PyC interlayer is The authors acknowledge the support of the Chinese impossible in the treatment. A reasonable explanation is National Foundation for Natural Sciences under Contract the interfacial stress produced by thermal expansion mis- No. 59772023. match of the fibers and the matrix induced the graphitization. The effect of heat treatment on the thermal diffusivity of a C/SiC composite could be further enhanced by References increasing thickness of the PyC interlayer, it is limited owing to its low volume fraction. Although the micro- [1] Jamet JM, Lmicq PJ, Schiroky GH. Composite thermostructure change in a C/SiC composite can be confirmed structures: an overview of the French experience. In: Naslain by its thermal diffusivity increase at high temperatures, it R, editor, High temperature ceramic matrix composites, Bordeaux: Woodhead Publications, 1993, pp. 215–29. reminds to be explained by further research. [2] Laux T, Ullmann T, Auweter-Kurtz M, Hald H, Kurz A. Investigation of thermal protection materials along an X-38 re-entry trajectory by plasma wind tunnel simulations, Sec- 4. Conclusions ond International Symposium on Atmospheric Reentry Vehicles and Systems. Arcahon (France): 2001:1–9. The longitudinal and transverse thermal diffusivity of [3] Tawil H, Bentsen LD, Baskaran S, Hasselman DPH. Thermal the C/SiC composite with and without a CVD SiC coating diffusivity of chemically vapor deposited silicon carbide could be well fitted by a multinomial function from room reinforced with silicon carbide or carbon fibers. J Mater Sci temperature to 1400 8C which includes a power term, an 1985;20:3201–12. exponential term and a constant term. [4] Parker WJ, Jenkins RJ, Butler CP, Abbott GL. Flash method of determining thermal diffusivity, heat capacity, and thermal The exponential term affected the thermal diffusivity conductivity. J Appl Phys 1961;32(9):1679–84. and led to its increase above 1200 8C with activation [5] Kim Y, Zangvil A, Goela JS, Taylor RL. Microstructure energy of 77 kcal/mol. The microstructure change in the comparison of transparent and opaque CVD SiC. J Am composite was the reason that the thermal diffusivity was Ceram Soc 1995;78:1571–9. increased above 1200 8C. [6] Goela JS, Pichering MA, Burns LE. Chemical vapor deThe longitudinal thermal diffusivity of the C/SiC posited SiC for high heat flux applications. SPIE Proc composite was twice or more than the transverse one and 1996;2855:2–13. increased more rapidly by the exponential term. The former was decreased by the CVD SiC coating, but the latter was increased by it