《复合材料 Composites》课程教学资源（学习资料）第五章 陶瓷基复合材料_9Properties of reinforced boron carbide laminar composites

Availableonlineatwww.sciencedirect.com SCIENCE DIRECT● E噩≈S ELSEVIER Journal of the European Ceramic Society 25(2005)3639-364 www.elsevier.com/locate/jeurceramsoc Properties of reinforced boron carbide laminar composites S. Tariollea. F. Thevenota. T Chartier, * J L. Be a Departement Ceramiques Speciales, UMR CNRS 5146, Ecole Nationale Superieure des Mines de saint-Etienne, 158 Cours Fauriel, F.42023 Saint-Erienme Cedex 2. fra b SPCTS, UMR CNRS 6638, Ecole Nationale Superieure de Ceramique Industrielle, 47-73 avenue Albert Thomas, F-87065 Limoges Cedex, france Received 3 June 2004: received in revised form 15 September 2004; accepted 24 September 2004 Available online December 2004 Abstract The reinforcement by crack deflection in boron carbide laminar composites is obtained by both controlling macrostructure and microstruc- ture. This structure had never been studied before in boron carbide materials Composites were prepared using tape-casting technique. Different composites with either porous interlayers obtained by pore forming agent, or weak interlayers obtained without adding sintering aid, or weak interlayers obtained by a mixture of boron carbide and boron nitride or weak graphite or boron nitride interfaces have been elaborated and characterized. Reinforcement by crack deflection was observed in most of these composites. In comparison to the work of rupture of the dense material, i. e. 23.09kJm', the following values were obtained for the minar composites: 38 kJm-3for composites with interlayers with corn starch(55 vol %) 40kJ m-3for composites with B4 C-BN interlayers 30kJ m' for composites with weak interlayers in BN and 39 km for composites with weak interlayers in graphite C 2004 Elsevier Ltd. All rights reserved. Keywords: Tape casting; Composites, Porosity; Mechanical properties, B4 C, BN 1. Introduction energetic criterion. It is well known that the interface crack deflection is influenced by the fracture energy and by the A way to reinforce ceramics, often characterized by their Youngs modulus of materials constituting each side of the euwtoughness that induces catastrophic rupture of the mate- interface. These two properties are dependent on the porosity rials, is to use laminar materials. The use of weak interfaces In the case of a weak graphite interface in SiC material, He or interlayers in functionally graded materials could be effi- and Hutchinson established that the ratio between fracture cient to improve toughness in non-oxide ceramics. The weak energies of the weak interface Gi and of the strong layer Gs interface could be obtained by the incorporation of graphite, must fulfil the following criterion to allow the crack to deflect boron nitride2or oxide ceramics(LaPO4 or YPO4). Another at this interface way to reinforce ceramic materials was to introduce porous and alternate dense-porous SiC(solid phase sintering) have Gs 0.57 ceramic interlayers. Thus, alternate dense-porous alumina the interface between porous and dense layers leading to an For dense -porous laminates, Clegg and coworkers,5ex- increase of the fracture energy pressed this energetic criterion(Eq. (1)considering that the Clegg and coworkers4.5 analyzed crack deflection mech- interface energy G could be replaced by the ligament energy anism in alternate dense-porous ceramic materials using an Glig (ligament of ceramic material between pores in which Corresponding author. Tel. +33555 45 22 25; fax: +33 555 79 0998 Ge $0.57 0955-2219/S-see front matter c 2004 Elsevier Ltd. All rights reserved

Journal of the European Ceramic Society 25 (2005) 3639–3647 Properties of reinforced boron carbide laminar composites S. Tariollea, F. Thevenot ´ a, T. Chartierb,∗, J.L. Bessonb a D´epartement C´eramiques Sp´eciales, UMR CNRS 5146, Ecole Nationale Sup´erieure des Mines de Saint-Etienne, 158 Cours Fauriel, F-42023 Saint-Etienne Cedex 2, France b SPCTS, UMR CNRS 6638, Ecole Nationale Sup´erieure de C´eramique Industrielle, 47-73 avenue Albert Thomas, F-87065 Limoges Cedex, France Received 3 June 2004; received in revised form 15 September 2004; accepted 24 September 2004 Available online 8 December 2004 Abstract The reinforcement by crack deflection in boron carbide laminar composites is obtained by both controlling macrostructure and microstruc￾ture. This structure had never been studied before in boron carbide materials. Composites were prepared using tape-casting technique. Different composites with either porous interlayers obtained by pore forming agent, or weak interlayers obtained without adding sintering aid, or weak interlayers obtained by a mixture of boron carbide and boron nitride, or weak graphite or boron nitride interfaces have been elaborated and characterized. Reinforcement by crack deflection was observed in most of these composites. In comparison to the work of rupture of the dense material, i.e. 23.09 kJ m−3, the following values were obtained for the laminar composites: 38 kJ m−3 for composites with interlayers with corn starch (55 vol.%), 40 kJ m−3 for composites with B4C-BN interlayers, 30 kJ m−3 for composites with weak interlayers in BN and 39 kJ m−3 for composites with weak interlayers in graphite. © 2004 Elsevier Ltd. All rights reserved. Keywords: Tape casting; Composites, Porosity; Mechanical properties; B4C; BN 1. Introduction A way to reinforce ceramics, often characterized by their low toughness that induces catastrophic rupture of the mate￾rials, is to use laminar materials. The use of weak interfaces or interlayers in functionally graded materials could be effi- cient to improve toughness in non-oxide ceramics. The weak interface could be obtained by the incorporation of graphite,1 boron nitride2 or oxide ceramics (LaPO4 or YPO4).3 Another way to reinforce ceramic materials was to introduce porous ceramic interlayers. Thus, alternate dense-porous alumina4 and alternate dense-porous SiC (solid phase sintering)5 have been studied. In these materials, crack deflection occurred at the interface between porous and dense layers leading to an increase of the fracture energy. Clegg and coworkers4,5 analyzed crack deflection mech￾anism in alternate dense-porous ceramic materials using an ∗ Corresponding author. Tel.: +33 555 45 22 25; fax: +33 555 79 09 98. E-mail address: t chartier@ensci.fr (T. Chartier). energetic criterion. It is well known that the interface crack deflection is influenced by the fracture energy and by the Young’s modulus of materials constituting each side of the interface. These two properties are dependent on the porosity. In the case of a weak graphite interface in SiC material, He and Hutchinson6 established that the ratio between fracture energies of the weak interface Gi and of the strong layer Gs must fulfil the following criterion to allow the crack to deflect at this interface: Gi Gs ≤ 0.57 (1) For dense-porous laminates, Clegg and coworkers4,5 ex￾pressed this energetic criterion (Eq. (1)) considering that the interface energy Gi could be replaced by the ligament energy Glig (ligament of ceramic material between pores in which the crack propagates): Glig Gs ≤ 0.57 (2) 0955-2219/$ – see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.jeurceramsoc.2004.09.014

S. Tariolle et al /Journal of the European Ceramic Sociery 25(2005)3639-3647 Therefore, Eq (2)can be expressed in relation with porosity F G 0.57 Ga(I-p where Gp is the fracture energy of the porous layer and Gd that of the dense laver Fig. 1. Schematic representation of notched specimen tested in 3 point According to Eq. (3), a porosity of 37 vol. is red bending fracture test Span L=20mm; W: thickness; b: width to initiate crack deflection at the interface between porous depth. F indicates the direction of the applied load. The layers are perpen- and dense layers, this value of porosity was experimentally dicular to this direction confirmed in SiC4 and alumina specimens 2.2 Technical characterizations In this context. we have studied boron carbide laminar opposites for their potential applications in armor and in nu- Density of materials was calculated from the measured clear energy fields 'Different boron carbide composites with weight and the geometrically determined volume. Image various weak interlayers or weak interfaces were elaborated analysis was used for characterization of grain and pore to study their reinforcement properties by crack defection. size using micrographs obtained by optical microscopy (for The energetic criterion and the level of porosity reported by porosity)and SEM(for grain size) Clegg and coworkers, to achieve crack deflection in the case of porous weak interlayers have been verified for boron 2.2.I. Description of the technique of observation of carbide laminar composites ed using 3 point-bending fracture tests on notched speci 2. Experimental procedure e o Crack propagation in multi-layered materials was evalu- hens(Fig. 1). A mechanical testing machine was used (IN STRON 8562)with a cross-head speed of0.025 mm/min The 2.. Preparation displacement was measured by an LVDT sensor The different composites have been elaborated using the 2.2.2. Measure of the work of rupture and of the crack tape-casting technique. The complete description of the pro- deviation cess has already been described. 9, 0 Composites were com- he work of rupture was evaluated using load-displace- posed of alternate dense boron carbide layers(94% of theo- ment curves and the crack deviation was measured on frac etical density )and weak layers or weak interfaces. Starting tographies of composites from boron carbide powder(Tetrabor 3000F, Wacker Ceran The work of rupture WR was calculated using Eq (4) ics, mean diameter 0.75 um), solid state sintering of boron where C()corresponds to the hatched area below the curve carbide was performed by pressureless sintering(2150C/1h (Fig. 2)until the load reached a plateau for a load C under argon)using phenolic resin as sintering aid.8, I0Com posites with different types of interlayers and interfaces have WR- Lb(w -a. c(x)dx been elaborated (Table 1). Porous interlayers were obtained using corn starch as pore forming agent, under-sintered in- An apparent friction stress can be calculated using Eq (5) terlayers using no sintering aid and weak interlayers using a mixture of boron carbide and boron nitride(55/45 in vol Fe ume). Weak interfaces were obtained using different sprays b(w-ae) (graphite or boron nitride) which have been pulverized on dense layers before thermocompression 0 Work of rupture Table I Denomination of the various composites elaborated Composite with Friction stress denomination B,C-BN Interlayers with mixing B, C-BN (55/4 Interlayers prepared with 50 vol. of( erlayers prepared with 55 vol % of 000,020,040.060,080,100.12014 IBN Interfaces with boron nitride Fig. 2. Load-displacement curve obtained on lamellar composites

3640 S. Tariolle et al. / Journal of the European Ceramic Society 25 (2005) 3639–3647 Therefore, Eq. (2) can be expressed in relation with porosity p: Gp Gd(1 − p) ≤ 0.57 (3) where Gp is the fracture energy of the porous layer and Gd that of the dense layer. According to Eq. (3), a porosity of 37 vol.% is required to initiate crack deflection at the interface between porous and dense layers, this value of porosity was experimentally confirmed in SiC4 and alumina5 specimens. In this context, we have studied boron carbide laminar composites for their potential applications in armor and in nu￾clear energy fields.7 Different boron carbide composites with various weak interlayers or weak interfaces were elaborated to study their reinforcement properties by crack deflection.8 The energetic criterion and the level of porosity reported by Clegg and coworkers4,5 to achieve crack deflection in the case of porous weak interlayers have been verified for boron carbide laminar composites. 2. Experimental procedure 2.1. Preparation The different composites have been elaborated using the tape-casting technique. The complete description of the pro￾cess has already been described.9,10 Composites were com￾posed of alternate dense boron carbide layers (94% of theo￾retical density) and weak layers or weak interfaces. Starting from boron carbide powder (Tetrabor 3000F, Wacker Ceram￾ics, mean diameter 0.75
m), solid state sintering of boron carbide was performed by pressureless sintering (2150 ◦C/1 h under argon) using phenolic resin as sintering aid.8,10 Com￾posites with different types of interlayers and interfaces have been elaborated (Table 1). Porous interlayers were obtained using corn starch as pore forming agent, under-sintered in￾terlayers using no sintering aid and weak interlayers using a mixture of boron carbide and boron nitride (55/45 in vol￾ume). Weak interfaces were obtained using different sprays (graphite or boron nitride) which have been pulverized on dense layers before thermocompression. Table 1 Denomination of the various composites elaborated. Composites denomination Composite with NSA Interlayers with no sintering aid B4C-BN Interlayers with mixing B4C-BN (55/45 in voume) CS45 Interlayers prepared with 45 vol.% of corn starch CS50 Interlayers prepared with 50 vol.% of corn starch CS55 Interlayers prepared with 55 vol.% of corn starch I-BN Interfaces with boron nitride I-G Interfaces with graphite Fig. 1. Schematic representation of notched specimen tested in 3 point￾bending fracture test. Span L = 20 mm; W: thickness; b: width; ae: notch depth. F indicates the direction of the applied load. The layers are perpen￾dicular to this direction. 2.2. Technical characterizations Density of materials was calculated from the measured weight and the geometrically determined volume. Image analysis was used for characterization of grain and pore size using micrographs obtained by optical microscopy (for porosity) and SEM (for grain size). 2.2.1. Description of the technique of observation of crack propagation Crack propagation in multi-layered materials was evalu￾ated using 3 point-bending fracture tests on notched speci￾mens (Fig. 1). A mechanical testing machine was used (IN￾STRON 8562) with a cross-head speed of 0.025 mm/min. The displacement was measured by an LVDT sensor. 2.2.2. Measure of the work of rupture and of the crack deviation The work of rupture was evaluated using load–displace￾ment curves and the crack deviation was measured on frac￾tographies of composites. The work of rupture WR was calculated using Eq. (4) where C(x) corresponds to the hatched area below the curve (Fig. 2) until the load reached a plateau for a load C. WR = 1 Lb(W − ae)
C(x)dx (4) An apparent friction stress can be calculated using Eq. (5). Ff = C b(W − ae) (5) Fig. 2. Load–displacement curve obtained on lamellar composites

S. Tariolle et al /Journal of the European Ceramic Sociery 25 (2005)3639-3647 An apparent fracture toughness has been calculated using was measured For each specimen, the maximum deviation the maximum load withstand by the composite. The fracture and the mean deviation for each interface or interlayer were toughness was measured using SENB method by Eq (6) given and correlated with the real fracture toughness by Eq. (7) Indeed. the value measured with SenB method overestimates the value of the fracture toughness when the notch root radius 3. Description of each type of composite: increases. A correction of the senb value has to be done macrostructure and mechanical properties kNB=va∑A Different types of interlayers or interfaces were elabo- rated. The denomination of each composite is indicated in Table 1. Whatever the material, the layers had uniform thick where ness and were parallel each other. The dense layers are sim- A0=1.9+0.075 ilar in the different composites. The toughness of a material made of dense layers is equal to 2.9 MPa m-and a work of rupture equal to 23 Jm-3 A1=-3.39+0.08 3.1. Weak interlayers A2=15.4-0.2175 3.1.1. Macrostructures and fractographies of the composites containing weak interlayers Different weak interlayers were studied, some had poros- A3=-2624+0.282 ity obtained with corn starch(CS), others were obtained us- ing no sintering aid (NSA), others mixing boron carbide and L nitride(B4C-BN). The grain size in the porous layers was A4=26.38-0.145 similar to that of the dense ones(0.8 um). The porosity in the layers with corn starch was interconnected and had a mean size of 10 um. In the porous layer without sintering aid, the KIC= KSENB tanh(2 V pe (7) material is under-sintered and the porosity was finer(around the micrometer). The weak layers obtained by a mixture of where pe is the notch root radius, ae the size of critical defect boron carbide and boron nitride have a skeleton of boron and y a geometrical factor equal to 1. 12 for a sharp crack carbide containing boron nitride grains. Macrostructures Lengths of crack deviation were measured on fractogra- of the different composites elaborated were represented in phies using the method developed by Kovar et al. 2 This Fig4 method consists in obtaining the delamination distances mea- Considering the characterization of the reinforcement tak suring the distance between through-thickness crack in adja- ing place in these composites, the typical results of 3 point- cent dense layers. An example of the method used is given bending tests were represented for each type of composite in Fig 3. The length of deviation at each interface or interlayer Figs. 5-7. Reinforcement by crack deflection was observed in the case of interlayers with corn starch for a porosity larg than 051(CS55)(Fig. 5)and for the interlayers made with a mixing of B C-BN(Fig. 7). In addition, as can be observed on the load-displacement curves, in these two cases, there was a friction stress due to the succession of crack deflections that induced a load resistance. In opposition, no crack deflection was observed in the composite(NSA)where interlayers are under-sintered. The rupture was brittle(Fig. 6) 3. 1.2. Weak porous interlayers in composites CS The influence of the porosity in the porous layers and the relative thickness between the dense and the porous layers on the work of rupture and on the lengths of crack deflection was studied in the composites(CS) with weak interlayers with corn starch 500 um 3.1.2.1. Infuence of the porosity on crack defection in CS. Fig 3. Fractography of a lamellar composite( CS55), the white lines corre- First, let us see the results concerning the influence of the bond to different crack deviations that have been measured level of porosity on the crack deflection properties. When the

S. Tariolle et al. / Journal of the European Ceramic Society 25 (2005) 3639–3647 3641 An apparent fracture toughness has been calculated using the maximum load withstand by the composite. The fracture toughness was measured using SENB method11 by Eq. (6) and correlated with the real fracture toughness by Eq. (7). 11 Indeed, the value measured with SENB method overestimates the value of the fracture toughness when the notch root radius increases. A correction of the SENB value has to be done. KSENB IC = σr √ae 4 i=0 Ai ae W  i (6) where A0 = 1.9 + 0.075 L W , A1 = −3.39 + 0.08 L W , A2 = 15.4 − 0.2175 L W , A3 = −26.24 + 0.2825 L W , A4 = 26.38 − 0.145 L W . KIC = KSENB IC tanh  2Y ac ρe  (7) where ρe is the notch root radius, ac the size of critical defect and Y a geometrical factor equal to 1.12 for a sharp crack. Lengths of crack deviation were measured on fractogra￾phies using the method developed by Kovar et al.12 This method consists in obtaining the delamination distances mea￾suring the distance between through-thickness crack in adja￾cent dense layers. An example of the method used is given Fig. 3. The length of deviation at each interface or interlayer Fig. 3. Fractography of a lamellar composite (CS55), the white lines corre￾spond to different crack deviations that have been measured. was measured. For each specimen, the maximum deviation and the mean deviation for each interface or interlayer were given. 3. Description of each type of composite: macrostructure and mechanical properties Different types of interlayers or interfaces were elabo￾rated. The denomination of each composite is indicated in Table 1. Whatever the material, the layers had uniform thick￾ness and were parallel each other. The dense layers are sim￾ilar in the different composites. The toughness of a material made of dense layers is equal to 2.9 MPa m−1/2 and a work of rupture equal to 23 kJ m−3. 3.1. Weak interlayers 3.1.1. Macrostructures and fractographies of the composites containing weak interlayers Different weak interlayers were studied, some had poros￾ity obtained with corn starch (CS), others were obtained us￾ing no sintering aid (NSA), others mixing boron carbide and nitride (B4C-BN). The grain size in the porous layers was similar to that of the dense ones (0.8
m). The porosity in the layers with corn starch was interconnected and had a mean size of 10
m. In the porous layer without sintering aid, the material is under-sintered and the porosity was finer (around the micrometer). The weak layers obtained by a mixture of boron carbide and boron nitride have a skeleton of boron carbide containing boron nitride grains. Macrostructures of the different composites elaborated were represented in Fig. 4. Considering the characterization of the reinforcement tak￾ing place in these composites, the typical results of 3 point￾bending tests were represented for each type of composite in Figs. 5–7. Reinforcement by crack deflection was observed in the case of interlayers with corn starch for a porosity larger than 0.51 (CS55) (Fig. 5) and for the interlayers made with a mixing of B4C-BN (Fig. 7). In addition, as can be observed on the load–displacement curves, in these two cases, there was a friction stress due to the succession of crack deflections that induced a load resistance. In opposition, no crack deflection was observed in the composite (NSA) where interlayers are under-sintered. The rupture was brittle (Fig. 6). 3.1.2. Weak porous interlayers in composites CS The influence of the porosity in the porous layers and the relative thickness between the dense and the porous layers on the work of rupture and on the lengths of crack deflection was studied in the composites (CS) with weak interlayers with corn starch. 3.1.2.1. Influence of the porosity on crack deflection in CS. First, let us see the results concerning the influence of the level of porosity on the crack deflection properties. When the

S. Tariolle et al /Journal of the European Ceramic Sociery 25(2005)3639-3647 500um Fig. 4.(a-c) Macrostructures of the different composites: (a) interlayers with corn starch(CS50); (b) interlayers under-sintered (NSA),(c)interlayers of B4 C-BN. Black layers: porous, grey layers: dense porosity increased in the interlayers, an important increase of deflection in such type of materials. In boron carbide com- the work of rupture(Fig 8)and of the length of crack deflec- posites, deflection appeared from a porosity of 0.51. There tion( Fig. 9)was observed for a porosity of 0.51. Moreover is a great difference between our observations and those pre- the presence of a friction stress indicated some reinforcement dicted by the theory of Clegg. This difference can be explain in the case of a porosity equal to 0.46 and 0.51(Table 2). The by studying the energetic criterion that is fully developed apparent fracture toughness was also increased with porosity in a further article. I (Table 2). More the interlayers were porous, so brittle, better was the reinforcement of the composite. Then, the porosity 3. 1.2.2. Infuence of the relative thickness of the layers in in the porous interlayers is an important criteria for crack de- CS55. The influence of the relative thickness of the dense and flection in composites. In the introduction, a porosity of 0.37 of the porous layers ea/ep was studied in composites contain- was required by Clegg and coworkers+,to observe crack ing a porosity of 0.51 in the porous layers(CS55 ). This pa- rameter was varied between 0.27 and 257. Macrostructure Table 2 of these types of composites were represented in Fig. 10 Values of apparent toughness and friction stress in function of the porosity As we can observe for work of rupture(Fig. 11), lengths in the interlayers in composites( CS)with interlayers with corn starch of crack deflection(Fig. 12)and also for the apparent friction Composites Porosity in porous KIC(MPam 2) Fr(MPa) stress and the apparent fracture toughness, there was a great dispersion of the values in function of ed/ep. This criterion CS45 1.54±0.16 seemed to have no significant influence on reinforcement by CS50 195±0.34 crack deflection. All the composites tested( CS55) presented CS55 3.44±0.95 reinforcement

3642 S. Tariolle et al. / Journal of the European Ceramic Society 25 (2005) 3639–3647 Fig. 4. (a–c) Macrostructures of the different composites: (a) interlayers with corn starch (CS50); (b) interlayers under-sintered (NSA); (c) interlayers of B4C-BN. Black layers: porous; grey layers: dense. porosity increased in the interlayers, an important increase of the work of rupture (Fig. 8) and of the length of crack deflec￾tion (Fig. 9) was observed for a porosity of 0.51. Moreover the presence of a friction stress indicated some reinforcement in the case of a porosity equal to 0.46 and 0.51 (Table 2). The apparent fracture toughness was also increased with porosity (Table 2). More the interlayers were porous, so brittle, better was the reinforcement of the composite. Then, the porosity in the porous interlayers is an important criteria for crack de- flection in composites. In the introduction, a porosity of 0.37 was required by Clegg and coworkers4,5 to observe crack Table 2 Values of apparent toughness and friction stress in function of the porosity in the interlayers in composites (CS) with interlayers with corn starch Composites Porosity in porous interlayers KIC (MPa m1/2) Ff (MPa) CS45 0.42 1.54 ± 0.16 0 CS50 0.46 1.95 ± 0.34 1.26 ± 0.78 CS55 0.51 3.44 ± 0.95 1.81 ± 0.59 deflection in such type of materials. In boron carbide com￾posites, deflection appeared from a porosity of 0.51. There is a great difference between our observations and those pre￾dicted by the theory of Clegg. This difference can be explain by studying the energetic criterion,8 that is fully developed in a further article.13 3.1.2.2. Influence of the relative thickness of the layers in CS55. The influence of the relative thickness of the dense and of the porous layers ed/ep was studied in composites contain￾ing a porosity of 0.51 in the porous layers (CS55). This pa￾rameter was varied between 0.27 and 2.57. Macrostructures of these types of composites were represented in Fig. 10. As we can observe for work of rupture (Fig. 11), lengths of crack deflection (Fig. 12) and also for the apparent friction stress8 and the apparent fracture toughness8, there was a great dispersion of the values in function of ed/ep. This criterion seemed to have no significant influence on reinforcement by crack deflection. All the composites tested (CS55) presented reinforcement

S. Tariolle et al /Journal of the European Ceramic Sociery 25 (2005)3639-3647 005 15 0 displacement(mm) 0.1 displacement(mm) Fig. 7.(a) Load-displacement curve of osite(BC-BN) ture of boron carbide and boron nitride (b) Fractograp of a composite(B C-BN) with interlayers with a mixture of boron carbide porosity of0.51(b)Fractography of a composite and boron nitride. The arrow next to the fractography indicates the direction S55 )with ith corn starch and with a porosity of 0.51. The of the crack propagation rrow next to the fractography indicates the direction of the crack propaga 20 CS45 I 10 048 Porosity in porous interlayers Fig 8. Work of rupture in function of the porosity in porous interlayers in composites(CS)with interlayers made with corn starch. 00250.05 0.1 displacement(mm) E ● max mun △ mean per layer Fig. 6.(a) Load-displacement curve of a composite(NSA)with interlay Porosity in porous interlayers ers under-sintered. (b) Fractography of a composite(NSA)with interlayers under-sintered. The arrow next to the fractography indicates the direction of Fig. 9. Lengths of crack deflection in function of the porosity in porous the crack propagation nterlayers in composites( CS)with interlayers made with corm star

S. Tariolle et al. / Journal of the European Ceramic Society 25 (2005) 3639–3647 3643 Fig. 5. (a) Load–displacement curve of a composite (CS55) with interlayers with corn starch and with a porosity of 0.51. (b) Fractography of a composite (CS55) with interlayers with corn starch and with a porosity of 0.51. The arrow next to the fractography indicates the direction of the crack propaga￾tion. Fig. 6. (a) Load–displacement curve of a composite (NSA) with interlay￾ers under-sintered. (b) Fractography of a composite (NSA) with interlayers under-sintered. The arrow next to the fractography indicates the direction of the crack propagation. Fig. 7. (a) Load–displacement curve of a composite (B4C-BN) with inter￾layers with a mixture of boron carbide and boron nitride. (b) Fractography of a composite (B4C-BN) with interlayers with a mixture of boron carbide and boron nitride. The arrow next to the fractography indicates the direction of the crack propagation. Fig. 8. Work of rupture in function of the porosity in porous interlayers in composites (CS) with interlayers made with corn starch. Fig. 9. Lengths of crack deflection in function of the porosity in porous interlayers in composites (CS) with interlayers made with corn starch

3644 S. Tariolle et al /Journal of the European Ceramic Sociery 25(2005)3639-3647 没转 500pm 500pm (a)ed/ep=0.27 (b)ed/ep=0.9 (c)ed/ep=1.85 Fig. 10.(a-c)Macrostructure of composites( CS55)with interlayers obtained by the use of com starch(p=0. 51)with different relative thicknesses 3.2. Weak interfaces Two different weak interfaces were studied: some with graphite spray (l-G)and some with boron nitride spray (I- BN). Dense layers had a thickness of approximately 100 um and were separated by thin interfaces of graphite or boron nitride(less than 5 um). The vaporization of the spray did not lead to a constant thickness of interfaces (Fig. 13). Long 60 crack deflections and sometimes layer delaminations were bserved on these two types of composites( Figs. 14 and 15) An increase of the work of rupture was also observed. A significant reinforcement by crack deflection was observed in composites with weak interfaces, either for weak inter- faces in graphite(l-G)or in boron nitride(I-BN). The prob- lem encountered in this type of composite is linked in the difficulties to reproduce the weak interfaces. This was re- Fig. Il. Work of rupture in function of the relative thickness in composites lated to the method used to pulverize the spray that could be (CS55) with interlayers with corn starch. The dote line correspond to the value of work of rupture for the dense boron carbide

3644 S. Tariolle et al. / Journal of the European Ceramic Society 25 (2005) 3639–3647 Fig. 10. (a–c) Macrostructure of composites (CS55) with interlayers obtained by the use of corn starch (p = 0.51) with different relative thicknesses. 3.2. Weak interfaces Two different weak interfaces were studied: some with graphite spray (I-G) and some with boron nitride spray (I￾BN). Dense layers had a thickness of approximately 100
m and were separated by thin interfaces of graphite or boron nitride (less than 5
m). The vaporization of the spray did not lead to a constant thickness of interfaces (Fig. 13). Long crack deflections and sometimes layer delaminations were observed on these two types of composites (Figs. 14 and 15). An increase of the work of rupture was also observed.8 A significant reinforcement by crack deflection was observed in composites with weak interfaces, either for weak inter￾faces in graphite (I-G) or in boron nitride (I-BN). The prob￾lem encountered in this type of composite is linked in the difficulties to reproduce the weak interfaces. This was re￾lated to the method used to pulverize the spray that could be improved. Fig. 11. Work of rupture in function of the relative thickness in composites (CS55) with interlayers with corn starch. The dote line correspond to the value of work of rupture for the dense boron carbide

S. Tariolle et al /Journal of the European Ceramic Sociery 25 (2005)3639-3647 E￥o Fig. 12. Lengths of crack deflection in function of the relative thickness in composites(C$55)with interlayers with com starch. displacement (mm) Fig. 14. Load-displacement curve and fractography of a composite (l-G) with interfaces of graphite. The arrow next to the fractography indicates the direction of the crack propagation. 500gm displacement(mm) Fig. 13. (a)Macrostructure of a composite with interfaces of graphite(l-G). (b)Macrostructure of a composite with interfaces of boron nitride(I-BN) 4. Comparison of the reinforcement by crack deflection in the different composites tested In order to compare the different composites, the different esults of work of rupture (Fig. 16), length of crack deflection Fig 15. Load-c ent curve and fractography of a composite(I-BN) (Fig. 17), apparent fracture toughness and apparent friction with interfaces of boron nitride. The arrow next to the fractography indicates stress were analyzed the direction of the crack propagation

S. Tariolle et al. / Journal of the European Ceramic Society 25 (2005) 3639–3647 3645 Fig. 12. Lengths of crack deflection in function of the relative thickness in composites (CS55) with interlayers with corn starch. Fig. 13. (a) Macrostructure of a composite with interfaces of graphite (I-G). (b) Macrostructure of a composite with interfaces of boron nitride (I-BN). 4. Comparison of the reinforcement by crack deflection in the different composites tested In order to compare the different composites, the different results of work of rupture (Fig. 16), length of crack deflection (Fig. 17), apparent fracture toughness8 and apparent friction stress8 were analyzed. Fig. 14. Load–displacement curve and fractography of a composite (I-G) with interfaces of graphite. The arrow next to the fractography indicates the direction of the crack propagation. Fig. 15. Load–displacement curve and fractography of a composite (I-BN) with interfaces of boron nitride. The arrow next to the fractography indicates the direction of the crack propagation

3646 S. Tariolle et al. Journal of the European Ceramic Society 25(2005)3639-3647 Conclusion Different types of boron carbide composites have been 60 elaborated by tap se types of C555 composites had never been studied before with boron car bide materials. Different composites with either porous in- 40 terlayers obtained by layers obtained without addition of sintering aid, or weak B4c-BN interlayers obtained by a mixture of boron carbide and Cs45 C550 boron nitride, or weak interfaces in graphite or boron ni- tride have been realized. Most of them presented reinforce- ment by crack deflection as we can see on the values of Fig. 16. Work of rupture in the different composites work of rupture: 38.07kJm-3for composites with interlay ers with corn starch(55 vol %) 40.25 kJ for compos- ites with B4C-BN interlayers, 30 11 kJm-3 for composites The composite NSA presented againof 175%for the work with weak interlayers in bN and 38.64kJm-3 for compos- of rupture compared with dense material whereas no crack ites with weak interlayers in graphite. An increase is ob- deflection was observed: in this case, there was no reinforce- served compare with the value for a dense boron carbide ment by crack deflection. The composite with interlayers ob-(25.09kJm-) tained by the lowest quantity of corn starch(CS45) presented Concerning the most promising materials, further re- no gain of work of rupture and small crack deflection: there searches should be carried out concerning(i) the influence is no reinforcement. In these two cases, no friction stress was of relative thickness on weak interlayers obtained with lower contents of pore forming agent and(ii)mixtures of B4 C-BN All other composites presented reinforcement by crack varying the content of BN. Concerning weak interfaces, the deflection Concerning composites with interlayers obtained technique of deposition of the interlayers should be better by the use of corn starch(Cs), the reinforcement was bet- mastered ter for a large value of porosity in the interlayers: a gain of 50% for the work of rupture compared with dense ma- terial was observed and long crack deflection were mea- sured for CS55 containing 51 vol. of porosity. A slight Acknowledgement gain of apparent fracture toughness (16%)was also no- ticed for the composite. The composite with interlayers This work is a part of the thesis of s. tariolle made with a mixture of boron carbide and boron nitrido presented a significant reinforcement by crack deflection: a gain of 60% in work of rupture and significant crack de- References flections. Both the composites with weak interfaces with graphite (l-G)and boron nitride(l-BN) presented a gain in 1. Clegg, w.J., The fabrication and failure of laminar ceramic compos- work of rupture (54% and 20%, respectively)and a gain ites. Acta Metall. Mater, 1992, 40(11), 3085-3093 in apparent fracture toughness(13% and 105 2. Liu, H. Y. and Hsu, S. M, Fracture behavior of multilayer sili- on nitride/boron nitride ceramics. J. Am. Ceram. Soc., 1996, 79(9) tively) 2452-2457 3. Kuo, D. H and Kriven, W M, Fracture of multilayer oxide compos- ites. Mater. Sci. Eng. 4, 1998 4. Blanks, K. S, Kristoffersson, A, Carlstrom, E. and Clegg, w. J Crack deflection in ceramic laminates using porous interlayers u: Ceram.Soc,1998,18,1945-1951. 5. Davis, J B, Kristoffersson, A, Carlstrom, E and Clegg, W. J, Fab- mean per layer rication of laminates with crack deflecting porous interlayers J.Am. Cera.Soc.,2000,83(10),23692374 6. He, M.-Y. and Hutchinson, J. W. Kinking of a crack out of the xoo interface.J. Appl. Mech., 1989, 56, 270-278 CS50 7. Thevenot, F, Boron carbide: a comprehensive review. J. Eur Ceram Soc,1990,6(4),205-225 8. Tariolle B4C-BN H-BN 1-G osites lamellaires. Elaboration, propri Ph.D 502 thesis. 328TD. Ecole des mines de saint-Etienne. france NSA 9. Reynaud, C, Thevenot, F and T, Processing and microstruc- ture of Sic laminar J. Refract. Met. Hard Mater Fig. 17. Lengths of crack deflection in the different composites 2001,19,425-435

3646 S. Tariolle et al. / Journal of the European Ceramic Society 25 (2005) 3639–3647 Fig. 16. Work of rupture in the different composites. The composite NSA presented a gain of 175% for the work of rupture compared with dense material whereas no crack deflection was observed: in this case, there was no reinforce￾ment by crack deflection. The composite with interlayers ob￾tained by the lowest quantity of corn starch (CS45) presented no gain of work of rupture and small crack deflection: there is no reinforcement. In these two cases, no friction stress was observed. All other composites presented reinforcement by crack deflection. Concerning composites with interlayers obtained by the use of corn starch (CS), the reinforcement was bet￾ter for a large value of porosity in the interlayers: a gain of 50% for the work of rupture compared with dense ma￾terial was observed and long crack deflection were mea￾sured for CS55 containing 51 vol.% of porosity. A slight gain of apparent fracture toughness (16%) was also no￾ticed for the composite. The composite with interlayers made with a mixture of boron carbide and boron nitride presented a significant reinforcement by crack deflection: a gain of 60% in work of rupture and significant crack de- flections. Both the composites with weak interfaces with graphite (I-G) and boron nitride (I-BN) presented a gain in work of rupture (54% and 20%, respectively) and a gain in apparent fracture toughness (13% and 105%, respec￾tively). Fig. 17. Lengths of crack deflection in the different composites. 5. Conclusion Different types of boron carbide composites have been elaborated by tape casting and lamination. These types of composites had never been studied before with boron car￾bide materials.8 Different composites with either porous in￾terlayers obtained by pore forming agent, or weak inter￾layers obtained without addition of sintering aid, or weak interlayers obtained by a mixture of boron carbide and boron nitride, or weak interfaces in graphite or boron ni￾tride have been realized. Most of them presented reinforce￾ment by crack deflection as we can see on the values of work of rupture: 38.07 kJ m−3 for composites with interlay￾ers with corn starch (55 vol.%), 40.25 kJ m−3 for compos￾ites with B4C-BN interlayers, 30.11 kJ m−3 for composites with weak interlayers in BN and 38.64 kJ m−3 for compos￾ites with weak interlayers in graphite. An increase is ob￾served compare with the value for a dense boron carbide (25.09 kJ m−3). Concerning the most promising materials, further re￾searches should be carried out concerning (i) the influence of relative thickness on weak interlayers obtained with lower contents of pore forming agent and (ii) mixtures of B4C-BN, varying the content of BN. Concerning weak interfaces, the technique of deposition of the interlayers should be better mastered. Acknowledgement This work is a part of the thesis of S. Tariolle. References 1. Clegg, W. J., The fabrication and failure of laminar ceramic compos￾ites. Acta Metall. Mater., 1992, 40(11), 3085–3093. 2. Liu, H. Y. and Hsu, S. M., Fracture behavior of multilayer sili￾con nitride/boron nitride ceramics. J. Am. Ceram. Soc., 1996, 79(9), 2452–2457. 3. Kuo, D. H. and Kriven, W. M., Fracture of multilayer oxide compos￾ites. Mater. Sci. Eng. A, 1998, 241, 241–250. 4. Blanks, K. S., Kristoffersson, A., Carlstrom, E. and Clegg, W. J., ¨ Crack deflection in ceramic laminates using porous interlayers. J. Eur. Ceram. Soc., 1998, 18, 1945–1951. 5. Davis, J. B., Kristoffersson, A., Carlstrom, E. and Clegg, W. J., Fab- ¨ rication of ceramic laminates with crack deflecting porous interlayers. J. Am. Ceram. Soc., 2000, 83(10), 2369–2374. 6. He, M.-Y. and Hutchinson, J. W., Kinking of a crack out of the interface. J. Appl. Mech., 1989, 56, 270–278. 7. Thevenot, F., Boron carbide: a comprehensive review. J. Eur. Ceram. Soc., 1990, 6(4), 205–225. 8. Tariolle, S., Carbure de bore monolithique poreux et com￾posites lamellaires. Elaboration, propri´et´es, renforcement. Ph.D. thesis, 328TD, Ecole des Mines de Saint-Etienne, France, 2004. 9. Reynaud, C., Thevenot, F. and Chartier, T., Processing and microstruc￾ture of SiC laminar composites. Intern. J. Refract. Met. Hard Mater., 2001, 19, 425–435

S. Tariolle et al. Journal of the European Ceramic Society 25 (2005)3639-3647 3647 10. Tariolle, S, Reynaud, C, Thevenot, F, Chartier, T. and Besson, J 12. Kovar, D, Thouless, M. D. and Halloran, J. w, Crack deflection and paration, microstructure and mechanical properties of SiC-s propagation in layered silicon nitride boron nitride ceramics. JAm. and B.C-B4C laminates J. Solid State Chem., 2004, 177, 487-492 Ceran.Soc.,l998,81(4)1004-1012 I1. Damani, R.J., Gstrein, R. and Danzer, R, Critical notch-root radius 13. Leguillon, D, Tariolle, S, Martin, E, Chartier, T. and Besson, J. L. effect in SENB-S fracture toughness testing. J. Eur: Ceram. Soc Prediction of crack deflection in porous/dense ceramic laminates. J 1996,16(7),695-702 Eur. Ceram Soc., in press

S. Tariolle et al. / Journal of the European Ceramic Society 25 (2005) 3639–3647 3647 10. Tariolle, S., Reynaud, C., Thevenot, F., Chartier, T. and Besson, J. L., Preparation, microstructure and mechanical properties of SiC-SiC and B4C-B4C laminates. J. Solid State Chem., 2004, 177, 487–492. 11. Damani, R. J., Gstrein, R. and Danzer, R., Critical notch-root radius effect in SENB-S fracture toughness testing. J. Eur. Ceram. Soc., 1996, 16(7), 695–702. 12. Kovar, D., Thouless, M. D. and Halloran, J. W., Crack deflection and propagation in layered silicon nitride boron nitride ceramics. J. Am. Ceram. Soc., 1998, 81(4), 1004–1012. 13. Leguillon, D., Tariolle, S., Martin, E., Chartier, T. and Besson, J. L., Prediction of crack deflection in porous/dense ceramic laminates. J. Eur. Ceram. Soc., in press