CERAMICS INTERNATIONAL ELSEVIER Ceramics International 30(2004)213-217 www.elsevier.com/locate/ceramint R-curve behavior of laminated SiC/Bn ceramics Dongyun Li Guanjun Qiao, Zhihao Ji State Key Laboratory for Mechanical Behavior of Materials, School of Materials Science and Engineering Xian Jiaotong University, Xi an 710049, PR China Received 22 November 2002: received in revised form 7 March 2003: accepted 7 April 2003 Abstract Laminated Sic/Bn ceramics with a perfect layered structure was fabricated by tap-casting process and hot-pressing sintering. Damage resistance and r-curve be of laminated Sic/Bn ceramics were evaluated using the indentation-strength-in-bending technique, and compared with those of monolithic SiC ceramics. The results showed that the indentation strengths for monolithic ceramics decreased steeply with indentation load, while for laminated SiC/bn ceramics reduced slightly with indentation loac suggesting an exceptional damage resistance. Moreover, a rising R-curve behavior was demonstrated for laminated Sic/Bn cera- mics, a plateau R-curve behavior for monolithic Sic ceramics. The excellent damage resistance and pronounced R-curve behavior of laminated SiC/BN ceramics were attributed to crack defection, crack branching and crack delamination at the Sic/Bn weak C 2003 Elsevier Ltd and Techna S.r.L. All rights reserved Keywords: B Interface: C. Fracture: C. Mechanical properties: D. Sic 1. Introduction achieved with strengths of 729 MPa and fracture toughness of 20.5 MPa m2. For the latter, damage When ceramic materials are applied to structural capability and oxidation resistance were improved in components, the lack of damage-tolerance is one of the comparison with microcomposites with a carbon inter most crucial problems. Ceramic composites with phase. These properties make this system attractive for layered structure have been considered to offer one of applications the most important approaches to this problem, and a In this paper, laminated SiC/BN ceramics(hereafter number of studies have been conducted so far in several denoted by LSB)with a perfect layered structure were systems including alumina zirconia [1-4], alumina /alu- fabricated by tap-casting process and hot-pressing sin minium titanate [5], mullite/alumina [6], silicon carbide tering. Damage resistance and R-curve behavior of LSB [7-10] and silicon nitride [11-13]. These laminated cera- are evaluated using the indentation-strength-in-bending mics have been reported to exhibit increased apparent technique(ISB), and compared with those of monolithic fracture toughness and fracture energy as well as a non- SiC ceramics(hereafter MS), the difference between the catastrophic fracture behavior. However, little work has two being analyzed been performed on the flaw tolerance and fracture resistance behavior of the layered ceramics Silicon carbon (SiC) layered ceramics with weak 2. Experimental procedure boron nitride (BN) interphases have been previously manufactured in a conventional two-dimensional 2. materials layered structure [14, as well as in microcomposites [15]. For the former, impressive properties were The flow chart for the fabrication process of LSB is shown in Fig. 1. Briefly, submicrometer silicon carbide 4 Corresponding author. Tel: +86-29-266-7942: fax: +86-29-266 powders with sintering additives 6 wt. alumina and 4 wt% yttria and the organic additives were mixed and E-mailaddressdongyunli(@hotmail.com(d.Li) cast into the green tapes, then coated with BN-containing 0272-8842/03/$30.00@ 2003 Elsevier Ltd and Techna S.r.l. All rights reserved. doi:10.1016/S0272-8842(03)00091-9
R-curve behavior of laminated SiC/BN ceramics Dongyun Li*, Guanjun Qiao, Zhihao Jin State Key Laboratory for Mechanical Behavior of Materials, School of Materials Science and Engineering, Xi’an Jiaotong University, Xi’an 710049, PR China Received 22 November 2002; received in revised form 7 March 2003; accepted 7 April 2003 Abstract Laminated SiC/BN ceramics with a perfect layered structure was fabricated by tap-casting process and hot-pressing sintering. Damage resistance and R-curve behavior of laminated SiC/BN ceramics were evaluated using the indentation-strength-in-bending technique, and compared with those of monolithic SiC ceramics. The results showed that the indentation strengths for monolithic SiC ceramics decreased steeply with indentation load, while for laminated SiC/BN ceramics reduced slightly with indentation load, suggesting an exceptional damage resistance. Moreover, a rising R-curve behavior was demonstrated for laminated SiC/BN ceramics, a plateau R-curve behavior for monolithic SiC ceramics. The excellent damage resistance and pronounced R-curve behavior of laminated SiC/BN ceramics were attributed to crack deflection, crack branching and crack delamination at the SiC/BN weak interfaces. # 2003 Elsevier Ltd and Techna S.r.l. All rights reserved. Keywords: B. Interface; C. Fracture; C. Mechanical properties; D. SiC 1. Introduction When ceramic materials are applied to structural components, the lack of damage-tolerance is one of the most crucial problems. Ceramic composites with a layered structure have been considered to offer one of the most important approaches to this problem, and a number of studies have been conducted so far in several systems including alumina/zirconia [1–4], alumina/aluminium titanate [5], mullite/alumina [6], silicon carbide [7–10] and silicon nitride [11–13]. These laminated ceramics have been reported to exhibit increased apparent fracture toughness and fracture energy as well as a noncatastrophic fracture behavior. However, little work has been performed on the flaw tolerance and fracture resistance behavior of the layered ceramics. Silicon carbon (SiC) layered ceramics with weak boron nitride (BN) interphases have been previously manufactured in a conventional two-dimensional layered structure [14], as well as in microcomposites [15]. For the former, impressive properties were achieved with strengths of 729 MPa and fracture toughness of 20.5 MPa m1/2. For the latter, damage capability and oxidation resistance were improved in comparison with microcomposites with a carbon interphase. These properties make this system attractive for applications. In this paper, laminated SiC/BN ceramics (hereafter denoted by LSB) with a perfect layered structure were fabricated by tap-casting process and hot-pressing sintering. Damage resistance and R-curve behavior of LSB are evaluated using the indentation-strength-in-bending technique (ISB), and compared with those of monolithic SiC ceramics (hereafter MS), the difference between the two being analyzed. 2. Experimental procedure 2.1. Materials The flow chart for the fabrication process of LSB is shown in Fig. 1. Briefly, submicrometer silicon carbide powders with sintering additives 6 wt.% alumina and 4 wt.% yttria and the organic additives were mixed and cast into the green tapes, then coated with BN-containing 0272-8842/03/$30.00 # 2003 Elsevier Ltd and Techna S.r.l. All rights reserved. doi:10.1016/S0272-8842(03)00091-9 Ceramics International 30 (2004) 213–217 www.elsevier.com/locate/ceramint * Corresponding author. Tel.: +86-29-266-7942; fax: +86-29-266- 5443. E-mail address: dongyun_li@hotmail.com (D. Li)
D. Liet al. Ceramics International 30(2004)213-217 SiC+additive 2. 2. SB test To evaluate damage resistance and r-curve behavi Milling(24h)+ persant the hot-pressed billets of both LSB and Ms were cut and ground into rectangular specimens; the prospectiv tensile surfaces of the specimens were perpendicular to Plasticizer Second milling(24h)+ Binder the hot-pressing direction, normal to the layer plane The nominal dimensions of the test specimens were 3x4x30 mm. The test specimens were indented at the Degassing Tamer center of the polished prospective tensile surface(4-mm- wide side) using a Vickers diamond pyramid indenter under loads ranging from 0. 1 to 300 N. Care was taken Tape-casting to orient one set of the indentation cracks to be parallel to the longitudinal axis of the rectangular specimens After indentation, the specimens were tested in three point bending with a span of 24 mm and a crosshead speed of 0.5 mm/min(Testing machine: instron 1195) Four tests were performed for each indentation load Conversion of vickers ISB data o(P), to generate a toughness curve, T(c) [T(c)=KgI is as follows [16] Punching and stacking Conversion is accomplished using an objective indenta tion-strength K-field analysis. Under the action of applied stress oA, radial cracks of size c produced at an Binder removal indentation load P extend according to the equilibrium ↓ condition ot-pressing sintering KA(c)=roAc/2+xPc-3/2=T(c) Fig. 1. Flow chart of fabrication process of LSB where KA(c) is a global applied stress intensity factor corresponding to an applied stress oA, T(c) the tough ness curve, and y the crack shape factor which depend slurry. After coating, the green laminates were dried, on crack and specimen geometry. x the residual contact stacked, and the organic additives were removed. coefficient which is often expressed in terms of Young Finally, they were sintered under a 30 MPa pressure at modulus E and hardness H[7]x=5(E/H 5 being 1850C for I h in a nitrogen atmosphere. The obtained factor of proportionality equals to 0.016[18]. The value layered structure is shown in Fig. 2. As can be seen, the of E/H can be measured by the Knoop indentation Sic layers(dark regions) and bn(gray regions) are method proposed by Marshall et al. 19 reasonably uniform and the interfaces are straight and For a given indentation load P, failure occurs at tha cated by stacking SiC green tapes without BN coating condill tress o well-distinguishable. For comparison, MS were fabri f which satisfies the " tangency using the same fabrication process as for LSB dKa(c)/dc= dT(c)/dc (3) Accordingly, given an appropriate calibration of the coefficients y and x, families of K'a(c) curves can be generated from the o p)data. T(c) then can be deter- mined objectively as envelopes of tangency points to these families of curves [16 3. Results and discussio 3. 1. Load-displacement curves Fig. 3 shows the load-displacement curves of the indented specimens for LSB and MS. Evidently, the fracture behavior of LSB was quite different from that Fig. 2. SEM micrograph of the cross section of LSB of MS. Ms fractured catastrophically, while LSB
slurry. After coating, the green laminates were dried, stacked, and the organic additives were removed. Finally, they were sintered under a 30 MPa pressure at 1850 C for 1 h in a nitrogen atmosphere. The obtained layered structure is shown in Fig. 2. As can be seen, the SiC layers (dark regions) and BN (gray regions) are reasonably uniform and the interfaces are straight and well-distinguishable. For comparison, MS were fabricated by stacking SiC green tapes without BN coating using the same fabrication process as for LSB. 2.2. ISB test To evaluate damage resistance and R-curve behavior, the hot-pressed billets of both LSB and MS were cut and ground into rectangular specimens; the prospective tensile surfaces of the specimens were perpendicular to the hot-pressing direction, normal to the layer plane. The nominal dimensions of the test specimens were 3430 mm. The test specimens were indented at the center of the polished prospective tensile surface (4-mmwide side) using a Vickers diamond pyramid indenter under loads ranging from 0.1 to 300 N. Care was taken to orient one set of the indentation cracks to be parallel to the longitudinal axis of the rectangular specimens. After indentation, the specimens were tested in threepoint bending with a span of 24 mm and a crosshead speed of 0.5 mm/min (Testing machine: instron 1195). Four tests were performed for each indentation load. Conversion of Vickers ISB data f (P), to generate a toughness curve, T(c) [T(c)=KR] is as follows [16]. Conversion is accomplished using an objective indentation-strength K-field analysis. Under the action of applied stress A; radial cracks of size c produced at an indentation load P extend according to the equilibrium condition, K0 AðcÞ ¼ Ac1=2 þ Pc 3=2 ¼ TðcÞ ð1Þ where K0 A(c) is a global applied stress intensity factor corresponding to an applied stress A, T(c) the toughness curve, and the crack shape factor which depends on crack and specimen geometry. the residual contact coefficient which is often expressed in terms of Young’s modulus E and hardness H [17], ¼ ðE=HÞ 1=2 being a factor of proportionality equals to 0.016 [18]. The value of E/H can be measured by the Knoop indentation method proposed by Marshall et al. [19]. For a given indentation load P, failure occurs at that applied stress A=f which satisfies the ‘‘tangency condition’’, dK0 AðcÞ=dc ¼ dTðcÞ=dc ð3Þ Accordingly, given an appropriate calibration of the coefficients and , families of K0 A(c) curves can be generated from the f(P) data. T(c) then can be determined objectively as envelopes of tangency points to these families of curves [16]. 3. Results and discussion 3.1. Load–displacement curves Fig. 3 shows the load–displacement curves of the indented specimens for LSB and MS. Evidently, the fracture behavior of LSB was quite different from that of MS. MS fractured catastrophically, while LSB Fig. 1. Flow chart of fabrication process of LSB. Fig. 2. SEM micrograph of the cross section of LSB. 214 D. Li et al. / Ceramics International 30 (2004) 213–217
D. Li et al.Ceramics International 30(2004)213-217 215 700 600 z500 R400 0 LSB 300 b G=756P0289 200 G622P014 100 020406080100120 100 Displacement /um Indentation load/N ig. 3. Load-displacement curves of the two types of specimens with Fig. 5. Strength plotted against indentation load for the two materi- an indentation in the tensile surface:(a)MS (b) LSB als Curves for the two materials deviate slightly from slope of -1/3 loading, LSB also deformed in a linear elastic fashion with a low indentation load, the strength was not loid showed a progressive failure behavior. Upon initial Fig. 5 into two regions. In the left region, for specimens until the crack reached the same stress intensity as ms dependent, because it was microstructure-controlled However, rather than traveling right across the speci On the other hand, in the right region, where specimens men the crack was deflected at the first bn interface had a higher indentation load, strength was inversely that it reached. as shown in Fig. 4. Crack deflection related to the indentation load, because strength was along the weak interface allowed the load to continue controlled by external flaws. rising. Failure of the second SiC layer gave rise to the Fig. 5 shows that in the low-indentation-load region first load drop. This process was repeated until all the the strengths were higher for Ms than for LSB, in the SiC layers had cracked, resulting in a step-like load- high-indentation-load region the strengths for the two displacement response. The total area under the load- materials decreased linearly with indentation load, but displacement curve represented the work-of-fracture the slope being much steeper for MS than for LSB. This WOF. Evidently, woF of LSB was much higher than means that LSB have a higher retained strength than hat of ms MS for an equivalent indentation load, hence an improved damage tolerance in comparison with Ms. It 3.2. Indentation strengths and R-curves also suggested that LSB might have a higher fracture resistance than ms Indentation fracture strength, of, is plotted logarith R-curves of two materials were obtained from inden- mically against indentation load, P, in Fig. 5 for both tation-strength data. Linear regression was used to MS and LSB. For each of the two materials a knee in the obtain the best fit for the experimental data in Fig. 5. It corresponding curve(at P=P*)separated the plot in showed that the slopes of MS and LSB were-0. 289 and -o142, respectively. Griffith materials, for which the R-curve is flat, would follow the power law, or x P-k with k=1/ 3. The fact that k is lower than 1/3, suggests a rising R-curve behavior for both materials. If the Vick ers crack geometry is considered to be material-inde- pendent, the values y= 1. 24 [20]. x=0.071 can be calculated for the experimental value E/H=20 for MS we use the same values of y and x for the two materials Two families of Ka(c)curves can now be constructed from the indentation-strength data in Fig. 5, insertin 0A=Or at each value of indentation load P in Eq (1) The envelopes of tangency points for two materials are hown in Fig. 6. It can be seen from Fig. 6 that the envelope of tangency points for MS was approximately horizontal, indicative of a plateau R-curve be Fig. 4. Propagation of a major crack through the specimen of LSB. while, for LSB the envelope of tangency points yields a Note that crack deflection occurs along the SiC/BN interfaces. rising R-curve. This suggests that LSB possessed excellent
showed a progressive failure behavior. Upon initial loading, LSB also deformed in a linear elastic fashion until the crack reached the same stress intensity as MS. However, rather than traveling right across the specimen the crack was deflected at the first BN interface that it reached, as shown in Fig. 4. Crack deflection along the weak interface allowed the load to continue rising. Failure of the second SiC layer gave rise to the first load drop. This process was repeated until all the SiC layers had cracked, resulting in a step-like load– displacement response. The total area under the load– displacement curve represented the work-of-fracture (WOF). Evidently, WOF of LSB was much higher than that of MS 3.2. Indentation strengths and R-curves Indentation fracture strength, f, is plotted logarithmically against indentation load, P, in Fig. 5 for both MS and LSB. For each of the two materials a knee in the corresponding curve (at P=P*) separated the plot in Fig. 5 into two regions. In the left region, for specimens with a low indentation load, the strength was not loaddependent, because it was microstructure-controlled. On the other hand, in the right region, where specimens had a higher indentation load, strength was inversely related to the indentation load, because strength was controlled by external flaws. Fig. 5 shows that in the low-indentation-load region the strengths were higher for MS than for LSB, in the high-indentation-load region the strengths for the two materials decreased linearly with indentation load, but the slope being much steeper for MS than for LSB. This means that LSB have a higher retained strength than MS for an equivalent indentation load, hence an improved damage tolerance in comparison with MS. It also suggested that LSB might have a higher fracture resistance than MS. R-curves of two materials were obtained from indentation-strength data. Linear regression was used to obtain the best fit for the experimental data in Fig. 5. It showed that the slopes of MS and LSB were 0.289 and 0.142, respectively. Griffith materials, for which the R-curve is flat, would follow the power law, f / P k with k=1/3. The fact that k is lower than 1/3, suggests a rising R-curve behavior for both materials. If the Vickers crack geometry is considered to be material-independent, the values =1.24 [20], =0.071 can be calculated for the experimental value E/H=20 for MS. we use the same values of and for the two materials. Two families of K0 A(c) curves can now be constructed from the indentation-strength data in Fig. 5, inserting A=f at each value of indentation load P in Eq. (1). The envelopes of tangency points for two materials are shown in Fig. 6. It can be seen from Fig. 6 that the envelope of tangency points for MS was approximately horizontal, indicative of a plateau R-curve behavior, while, for LSB the envelope of tangency points yields a rising R-curve. This suggests that LSB possessed excellent Fig. 3. Load–displacement curves of the two types of specimens with an indentation in the tensile surface: (a) MS, (b) LSB. Fig. 4. Propagation of a major crack through the specimen of LSB. Note that crack deflection occurs along the SiC/BN interfaces. Fig. 5. Strength plotted against indentation load for the two materials. Curves for the two materials deviate slightly from slope of 1/3. D. Li et al. / Ceramics International 30 (2004) 213–217 215
D. Liet al. Ceramics International 30(2004)213-217 86420 200N 864 0①cg3 86420864 2 100200300400500600700 100200300400500600700 Crack Crack size c/um Fig. 6. Toughness curve diagrams for(a) MS, (b) LSB. Families of solid curves are plots of KA(c)in Eq(1)using strength data in Fig. 5. Shaded lines are T(c) functions, plotted as locus of tangency points to Ka(c). damage tolerance and a rising R-curve behavior inin Fig. 8, further absorbed the fracture energy of mate- comparison with Ms rials, relaxed the stress at the crack tip and improved the fracture resistance and damage resistance 3.3. The main toughening mechanisms The above results showed that LSB possessed excel D27 lent damage tolerance and a rising R-curve behavior in comparison with Ms. This was thought be related to their different toughening mechanisms. Fig. 7 showed the crack propagation path of Ms, which introduced by Vickers indentation, crack deflection along weak grain interface is its main toughening mechanism. The main toughening mechanism of LSB is crack deflection along weak BN interface continually. As shown in ig. 4, when a crack propagates through LSB, crack deflection occurred along weak interface continually leading to a step-like crack propagation path which greatly extended effective crack length and absorbed more fracture energy, with an improvement of fracture resistance. Besides, crack branching and crack delani (a) nation which appeared at the weak interface, as shown D26 6um (b) Fig. 8. Two cracking mode of LSB for(a) crack branching, (b) crack Fig. 7. Crack propagation path of Ms
damage tolerance and a rising R-curve behavior in comparison with MS. 3.3. The main toughening mechanisms The above results showed that LSB possessed excellent damage tolerance and a rising R-curve behavior in comparison with MS. This was thought be related to their different toughening mechanisms. Fig. 7 showed the crack propagation path of MS, which introduced by Vickers indentation, crack deflection along weak grain interface is its main toughening mechanism. The main toughening mechanism of LSB is crack deflection along weak BN interface continually. As shown in Fig. 4, when a crack propagates through LSB, crack deflection occurred along weak interface continually, leading to a step-like crack propagation path which greatly extended effective crack length and absorbed more fracture energy, with an improvement of fracture resistance. Besides, crack branching and crack delamination which appeared at the weak interface, as shown in Fig. 8, further absorbed the fracture energy of materials, relaxed the stress at the crack tip and improved the fracture resistance and damage resistance. Fig. 6. Toughness curve diagrams for (a) MS, (b) LSB. Families of solid curves are plots of K0 A(c) in Eq. (1) using strength data in Fig. 5. Shaded lines are T(c) functions, plotted as locus of tangency points to K0 A(c). Fig. 7. Crack propagation path of MS. Fig. 8. Two cracking mode of LSB for (a) crack branching, (b) crack delamination. 216 D. Li et al. / Ceramics International 30 (2004) 213–217
D. Liet al. Ceramics International 30(2004)213-217 4. Conclusion [6 H. Katsuki, Y. Hirata, Coat of alumina sheet with needle-like mullite, J Ceram Soc. Jpn. 98(1990)1114-1119 The fracture behavior of LSB was quite different from [7 J. She. T. Inoue, K. Ueno. Multilayer Al203/SiC ceramics wit hat of MS. Ms fractured catastrophically, while LSB S roved mechanical behavior, Ceram. Int. 20(2000)1771- showed a non-catastrophic failure behavior. Due to the [8 N.P. Padture, D.C. Pender, S. Wuttiphan, B.R. Lawn, In situ deflection, branching and delamination of the transverse processing of silicon carbide layer structures. J. Am. Ceram Soc cracks at the Sic/Bn interfaces, the indentation 78(11)(1995)3160-3162. strengths of LSB were observed to be insensitive to the [9 W.J. Clegg, K. Kendall, N.M. Alford, A simple way to make increases in the indentation load. indicative of excellent tough ceramics, Nature(London)347(1990)455-461 [0 w.J. Clegg, The fabrication and failure of laminar ceramic com- damage resistance, compared with Ms. Furthermore, a posites. Acta Metall. Mater. 40(11)(1992)3085-3093 rising R-curve behavior was demonstrated from ISB test. [11] H Liu, B.R. Lawn, S M. Hsu, Hertzian contact response of tailored silicon nitride multilayers, J Am Ceram Soc. 79(4)(1996)1009 [12 H. Liu, S.M. Hsu, Fracture behavior of multilayer silicon nitrid boron nitride ceramics, J. Am. Ceram. Soc. 79(9)(1996)2452- Acknowledgements [ T Ohji, Y Shigeaki, T Miyaji The authors wish to acknowledge State Key labora of multilayered silicon nitride, J. Am. Ceram. Soc. 80(4)(1997) tory for Mechanical Behavior of Materials for financial 991-994. nd technical support [14 GJ. Yuan, Y.M. Luo, D M. Chen, Selection of interfacial layer of laminated silicon carbide matrix composites, J. Chinese Ceram Soc. 29(3)(2001)226-231(in Chinese) [5 F. Rebillat, A. Guette, L. Espitalier, Oxidation resistance of References nly crystallised Bn interphase, J. Eur. Ceram Soc. 18(1998)1809-1819. [P Sarkar, X. Haung, P.S. Nicholson, Structural ceramic [16 S K. Lee, S. Wuttiphan, C.J. Fairbanks, Role of microstructure- minates by electrophoretic deposition, J. Am. Ceram Soc. strength properties in ceramics: I, effect of crack size on toug es,J.Am. Ceram.Soc.80(9)(1997)2367-2381 []D B. Marshall, J.J. Ratto, F.F. Lange, Enhanc [7R F. Cook, Direct observation and analysis of indentation toughness in layered microcomposites of Ce-ZrO2 and AlO3, acking in glass and ceramics, J. Am. Ceram Soc. 73(1990) J.Am. Ceram.Soc.74(12)(1991)29792987. 3J. Requena, R Moreno, J.S. Moya, Alumina and alumina/zirco- [18]GT. Antis, A critical evaluation of indentation techniques for ia multilayer composites obtained by slip casting, J. Am. Ceram. neasuring fracture toughness-direct crack measurements. J. Am Soc.72(8)(1989)151l-1513 Ceram.Soc.64(2)(1981)533 (K.P. Plucknett, C.H. Caceres, C. Hughers, D.S. Wilkinson, Pro- [19] D B. Marshall, T Noma, A.G. Evans, A simple method for deter cessing of tape-cast laminates prepared from fine alumina/zir ning elastic-modulus-to-hardness ratios using Knoop indenta nia powders, J Am Ceram Soc. 77(8)(1994)2145-2153 n measurements, J Am Ceram Soc. 65(10)(1982)175-176. 5C. Russo, M.P. Harmer, H M. Chan, G.A. Miller, Design of [20]R F. Cook, Microstructure-strength properties in ceramics: e laminated of crack size on toughness, J. Am. Ceram. Soc. 68(11)(1 ness,J.Am. Ceram.Soc.75(12)(1992)3396-3400 604615
4. Conclusion The fracture behavior of LSB was quite different from that of MS. MS fractured catastrophically, while LSB showed a non-catastrophic failure behavior. Due to the deflection, branching and delamination of the transverse cracks at the SiC/BN interfaces, the indentation strengths of LSB were observed to be insensitive to the increases in the indentation load, indicative of excellent damage resistance, compared with MS. Furthermore, a rising R-curve behavior was demonstrated from ISB test. Acknowledgements The authors wish to acknowledge State Key Laboratory for Mechanical Behavior of Materials for financial and technical support. References [1] P. Sarkar, X. Haung, P.S. Nicholson, Structural ceramic microlaminates by electrophoretic deposition, J. Am. Ceram. Soc. 75 (10) (1992) 2907–2909. [2] D.B. Marshall, J.J. Ratto, F.F. Lange, Enhanced fracture toughness in layered microcomposites of Ce-ZrO2 and Al2O3, J. Am. Ceram. Soc. 74 (12) (1991) 2979–2987. [3] J. Requena, R. Moreno, J.S. Moya, Alumina and alumina/zirconia multilayer composites obtained by slip casting, J. Am. Ceram. Soc. 72 (8) (1989) 1511–1513. [4] K.P. Plucknett, C.H. Caceres, C. Hughers, D.S. Wilkinson, Processing of tape-cast laminates prepared from fine alumina/zirconia powders, J. Am. Ceram. Soc. 77 (8) (1994) 2145–2153. [5] C.J. Russo, M.P. Harmer, H.M. Chan, G.A. Miller, Design of laminated ceramic composite for improved strength and toughness, J. Am. Ceram. Soc. 75 (12) (1992) 3396–3400. [6] H. Katsuki, Y. Hirata, Coat of alumina sheet with needle-like mullite, J. Ceram. Soc. Jpn. 98 (1990) 1114–1119. [7] J. She, T. Inoue, K. Ueno, Multilayer Al2O3/SiC ceramics with improved mechanical behavior, Ceram. Int. 20 (2000) 1771– 1775. [8] N.P. Padture, D.C. Pender, S. Wuttiphan, B.R. Lawn, In situ processing of silicon carbide layer structures, J. Am. Ceram. Soc. 78 (11) (1995) 3160–3162. [9] W.J. Clegg, K. Kendall, N.M. Alford, A simple way to make tough ceramics, Nature (London) 347 (1990) 455–461. [10] W.J. Clegg, The fabrication and failure of laminar ceramic composites, Acta Metall. Mater. 40 (11) (1992) 3085–3093. [11] H. Liu, B.R. Lawn, S.M. Hsu, Hertzian contact response of tailored silicon nitride multilayers, J. Am. Ceram. Soc. 79 (4) (1996) 1009. [12] H. Liu, S.M. Hsu, Fracture behavior of multilayer silicon nitride/ boron nitride ceramics, J. Am. Ceram. Soc. 79 (9) (1996) 2452– 2457. [13] T. Ohji, Y. Shigegaki, T. Miyajima, Fracture resistance behavior of multilayered silicon nitride, J. Am. Ceram. Soc. 80 (4) (1997) 991–994. [14] G.J. Yuan, Y.M. Luo, D.M. Chen, Selection of interfacial layer of laminated silicon carbide matrix composites, J. Chinese Ceram. Soc. 29 (3) (2001) 226–231 (in Chinese). [15] F. Rebillat, A. Guette, L. Espitalier, Oxidation resistance of SiC/SiC micro and minicomposites with a highly crystallised Bn interphase, J. Eur. Ceram. Soc. 18 (1998) 1809–1819. [16] S.K. Lee, S. Wuttiphan, C.J. Fairbanks, Role of microstructurestrength properties in ceramics: I, effect of crack size on toughness, J. Am. Ceram. Soc. 80 (9) (1997) 2367–2381. [17] R.F. Cook, Direct observation and analysis of indentation cracking in glass and ceramics, J. Am. Ceram. Soc. 73 (1990) 787–817. [18] G.T. Antis, A critical evaluation of indentation techniques for measuring fracture toughness-direct crack measurements, J. Am. Ceram. Soc. 64 (2) (1981) 533. [19] D.B. Marshall, T. Noma, A.G. Evans, A simple method for determining elastic-modulus-to-hardness ratios using Knoop indentation measurements, J. Am. Ceram. Soc. 65 (10) (1982) 175–176. [20] R.F. Cook, Microstructure–strength properties in ceramics: effect of crack size on toughness, J. Am. Ceram. Soc. 68 (11) (1985) 604–615. D. Li et al. / Ceramics International 30 (2004) 213–217 217