J Mater sci(2007)42:4245-4253 DOI10.1007/s10853-006-0688-1 Carbon fiber/ceramic matrix composites: processing, oxidation and mechanical properties Samanta rafaela de omena pina Luiz Claudio Pardini .Inez Valeria Pagotto Yoshida Received: 16 March 2006/ Accepted: 19 July 2006/Published online: 3 March 2007 Springer Science+Business Media, LLC 2007 Abstract Ceramic matrix composites ( CMC) have (COMPOSITE 2)the values were 14.2+ 4.1 MP been considered in the last two decades to be alterna- 15.0+ 2.0 GPa, respectively. The properties of tive materials for highly demanding thermo-structural latter me ceramic matir rnee by the micro- applications. Pre-ceramic polymers offer significant structure of advantages for manufacturing these composites by the polymer impregnation method. In the present work carbon fiber/silicon oxycarbide(C/SiCrOy) composites were obtained by controlled pyrolysis of carbon fiber/ Introduction bridge polysilsesquioxane composites(COMPOSITE 1)followed by infiltration/pyrolysis cycles with a poly- Damage tolerance and efficiency enhancement are cyclic silicone network. The polysilsesquioxane showed challenges in aero-engine components and other high high wettability and adhesion on the carbon fiber sur- temperature industrial devices, such as gas turbines face. An improvement of the thermo-oxidation resis- and low weight thermal protection shields. These tance and a reduction of the porosity as a function of challenges can be met by using ceramic matrix com e number of polycyclic silicone infiltration cy posites(CMC)[1-4. Candidates for these materials were observed. An extra improvement in the thermo- are mainly based on non-oxide matrixes, such as car- oxidation protection was found when the C/SiC,O, bon, Sic, Si3N4 or certain oxide matrixes, such as composite was coated with a poly (phenylsilsesquiox- SiC,Oy, SiO2-Al2O3, etc.[5-10 ane) layer(COMPOSITE 2). Shear properties for the Among the preparation methods used for CMC composites showed a dependence on the nature of the composites, the polymer pyrolysis route is one of the matrix. The average in-plane shear strength and the most attractive because it allows the production of shear modulus were 44.2+ 1.9 MPa and 2.2+0.5 GPa complex composite shapes and is cost effective. for the polymeric matrix composite(COMPOSITE 1), addition, any of the conventional processing tech respectively. For the ceramic matrix composite niques of the polymeric matrix composite, such as hand lay-up, autoclave molding, resin transfer molding (RTM) and filament winding can be successfully used to obtain CMC. These methods provide near net shaping for many complex geometries, which is con istituto de Quimica, Universidade Estadual de Campinas sidered to be an advantage in making cost effective UNICAMP, CP 6154, 13084-971 Campinas, SP, Brazil [5,8-11] Silicon-based polymers have received a great deal of fo L. C. Pardini ion for the al Instituto de aeronautica e CMC. Polysilanes, polysiloxanes, polysilazanes and a Espaco, AMR, 12228-904 Sao Jose dos Campos, SP, Brazil variety of hybrid polymers containing hetero-atoms
Carbon fiber/ceramic matrix composites: processing, oxidation and mechanical properties Samanta Rafaela de Omena Pina Æ Luiz Claudio Pardini Æ Inez Vale´ria Pagotto Yoshida Received: 16 March 2006 / Accepted: 19 July 2006 / Published online: 3 March 2007 Springer Science+Business Media, LLC 2007 Abstract Ceramic matrix composites (CMC) have been considered in the last two decades to be alternative materials for highly demanding thermo-structural applications. Pre-ceramic polymers offer significant advantages for manufacturing these composites by the polymer impregnation method. In the present work, carbon fiber/silicon oxycarbide (C/SiCxOy) composites were obtained by controlled pyrolysis of carbon fiber/ bridge polysilsesquioxane composites (COMPOSITE 1) followed by infiltration/pyrolysis cycles with a polycyclic silicone network. The polysilsesquioxane showed high wettability and adhesion on the carbon fiber surface. An improvement of the thermo-oxidation resistance and a reduction of the porosity as a function of the number of polycyclic silicone infiltration cycles were observed. An extra improvement in the thermooxidation protection was found when the C/SiCxOy composite was coated with a poly(phenylsilsesquioxane) layer (COMPOSITE 2). Shear properties for the composites showed a dependence on the nature of the matrix. The average in-plane shear strength and the shear modulus were 44.2 ± 1.9 MPa and 2.2 ± 0.5 GPa for the polymeric matrix composite (COMPOSITE 1), respectively. For the ceramic matrix composite (COMPOSITE 2) the values were 14.2 ± 4.1 MPa and 15.0 ± 2.0 GPa, respectively. The properties of the latter composite were also governed by the microstructure of the ceramic matrix. Introduction Damage tolerance and efficiency enhancement are challenges in aero-engine components and other high temperature industrial devices, such as gas turbines and low weight thermal protection shields. These challenges can be met by using ceramic matrix composites (CMC) [1–4]. Candidates for these materials are mainly based on non-oxide matrixes, such as carbon, SiC, Si3N4 or certain oxide matrixes, such as SiCxOy, SiO2–Al2O3, etc... [5–10]. Among the preparation methods used for CMC composites, the polymer pyrolysis route is one of the most attractive because it allows the production of complex composite shapes and is cost effective. In addition, any of the conventional processing techniques of the polymeric matrix composite, such as: hand lay-up, autoclave molding, resin transfer molding (RTM) and filament winding can be successfully used to obtain CMC. These methods provide near net shaping for many complex geometries, which is considered to be an advantage in making cost effective parts [5, 8–11]. Silicon-based polymers have received a great deal of attention for their use as pre-ceramic precursors for CMC. Polysilanes, polysiloxanes, polysilazanes and a variety of hybrid polymers containing hetero-atoms, S. R. de Omena Pina I. V. P. Yoshida (&) Instituto de Quı´mica, Universidade Estadual de Campinas – UNICAMP, CP 6154, 13084-971 Campinas, SP, Brazil e-mail: valeria@iqm.unicamp.br L. C. Pardini Centro Te´cnico Aeroespacial, Instituto de Aerona´utica e Espac¸o, AMR, 12228-904 Sa˜o Jose´ dos Campos, SP, Brazil 123 J Mater Sci (2007) 42:4245–4253 DOI 10.1007/s10853-006-0688-1
J Mater Sci such as boron, are currently of interest [10-16. On the Polymeric matrix composite manufacture other hand. carbon fibers silicon carbide fibers and alumino-silicate fibers have been used as reinforce- Carbon fiber/bridge hybrid polysilsesquioxane com- ment in CMC for highly demanding thermo-mechani- posite was made by the hand lay-up technique by cal applications [17-19 stacking 22 layers of carbon fabric impregnated with a The present work focuses on the preparation and mixture of PBFEAPS, in a stoichiometric NH2: epoxy characterization of a polymeric matrix composite, ratio (1: 1). The amine-epoxy addition reaction was obtained from carbon fibers and a bridge poly- followed by the hydrolysis and condensation reactions silsesquioxane matrix precursor by controlled pyrolysis. of alkoxysilane groups, in situ, promoted by environ a polycyclic silicone network generated in situ by the mental moisture infiltration of a mixture of functional siloxane cyclics The bridge hybrid polysilsesquioxane matrix was and platinum-catalyst, followed by pyrolysis, was cured at room temperature and aged for 24 h, to form as precursor of SiC,Oy. After three infiltration/ the green-body laminate polymeric composite, named lysis cycles the carbon fiber reinforced ceramic matrix COMPOSITE 1. Specimens of this com composite was coated with poly(phenylsilsesquiox- trimmed in appropriate dimensions 4×smm) The structural evolution of the polymeric matrixes to A PBFE/APS monolithic polymeric sample was also the corresponding ceramic products was studied. The obtained as a tough and colorless material, by the same polymeric and ceramic composites were evaluated by procedure, to study the structural evolution from the density and porosity measurements, shear tests, polymer to the corresponding cerar thermo-gravimetry and their morphologies were investigated by field emission scanning electron Ceramic matrix composite manufacture Specimens obtained from the green-body laminate COMPOSITE 1) were pyrolyzed under argon flow witI the following heating cycle: heating to 180C at 5C/ Experimental procedure min, holding for 60 min at this temperature; heating to 500C at 5C/min, holding for 60 min at this temper Starting raw materials ture; and heating to 1000C, at 2 C/min, holding for 120 min at this temperature. Finally the specimens were Plain-weave fabrics made of Toray T300 ex-PAN based cooled to room temperature at 2C/min. These carbon fibers were constituted by 3000 filament-count pyrolyzed specimens were named C/C-SiO2. yarns. The polysilsesquioxane matrix(PBFEAPS)was In order to obtain CMC with a reasonable density prepared in situ by the reaction between amin- and improved mechanical properties, the C/C-Sio opropyltriethoxysilane(APS), H2N( CH2)3Si(OC2H5)3, composite samples were infiltrated with a mixture purchased from Aldrich Chemical Company, Inc(Mil- D4 Vi and DH, 58: 42 (wt%), respectively, and waukee, USA), and an oligomeric epoxy resin derived 0.01 wt% of Pt-catalyst. The polymer was cured at from polybisphenol A-co-epiclorhydrin(PBFE), Mn room temperature giving rise to an in situ polycyclic -600 g/mol, supplied by Dow Chemical do brazil (sao silicone network (DVDH). The composite samples lulo, Brazil) were subsequently pyrolyzed by using the same Commercially available functional siloxane cyclic heating cycle described before. In this step, DVDH oligomers, the 1,3, 5,7-tetramethyl-1, 3, 5, 7-tetravinylcyclo was converted to SiC Oy. Three infiltration/pyrolysis tetrasiloxane,[(CH3)(CH2=CH)SiO4,(D4 Vi), and the cycles were carried out, and the corresponding com- 1, 3, 5, 7-tetramethyl-cyclotetrasiloxane, [(CH3)(H)SiO]4, posite samples were named C/SiC, O, 1, C/SiC,O, 2, (D4H) were supplied by Dow Corning do Brazil (Hort- and C/SiC,O, 3. Finally, the C/SiC,O, 3 composite olandia, Brazil). Platinum divinylcomplex, 2-3% in vinyl sample was coated with a pre-hydrolyzed phenyltri terminated poly(dimethylsiloxane),(Pt-catalyst)was ethoxysilane solution, which gave rise to a thin in situ purchased from Aldrich Chemical Company, Inc(Mil- poly(phenylsilsesquioxane)(POSS)film. This final waukee, USA). These materials were used as received. composite was named COMPOSITE 2(Fig. 1). To The mixture of DaVi, D,H and Pt-catalyst was used in understand the ceramization process, all the compos- the infiltration step with the ceramic matrix composite to ite specimens(COMPOSITE 1, C/C-SiOz, C/SiCrOy produce in situ the polycyclic silicone network as a 1, C/SiCrOy 2, C/SiCr 3 and COmPOSItE 2)were SiCO, precursor characterized 2 Springer
such as boron, are currently of interest [10–16]. On the other hand, carbon fibers, silicon carbide fibers and alumino-silicate fibers have been used as reinforcement in CMC for highly demanding thermo-mechanical applications [17–19]. The present work focuses on the preparation and characterization of a polymeric matrix composite, obtained from carbon fibers and a bridge polysilsesquioxane matrix precursor by controlled pyrolysis. A polycyclic silicone network generated in situ by the infiltration of a mixture of functional siloxane cyclics and platinum-catalyst, followed by pyrolysis, was used as precursor of SiCxOy. After three infiltration/pyrolysis cycles the carbon fiber reinforced ceramic matrix composite was coated with poly(phenylsilsesquioxane) resin in order to enhance its oxidation resistance. The structural evolution of the polymeric matrixes to the corresponding ceramic products was studied. The polymeric and ceramic composites were evaluated by density and porosity measurements, shear tests, thermo-gravimetry and their morphologies were investigated by field emission scanning electron microscopy. Experimental procedure Starting raw materials Plain-weave fabrics made of Toray T300 ex-PAN based carbon fibers were constituted by 3000 filament-count yarns. The polysilsesquioxane matrix (PBFE/APS) was prepared in situ by the reaction between aminopropyltriethoxysilane (APS), H2N(CH2)3Si(OC2H5)3, purchased from Aldrich Chemical Company, Inc. (Milwaukee, USA), and an oligomeric epoxy resin derived from polybisphenol A-co-epiclorhydrin (PBFE), Mn ~600 g/mol, supplied by Dow Chemical do Brazil (Sa˜o Paulo, Brazil). Commercially available functional siloxane cyclic oligomers, the 1,3,5,7-tetramethyl-1,3,5,7-tetravinylcyclo tetrasiloxane, [(CH3)(CH2=CH)SiO]4, (D4Vi), and the 1,3,5,7-tetramethyl-cyclotetrasiloxane, [(CH3)(H)SiO]4, (D4H) were supplied by Dow Corning do Brazil (Hortolaˆndia, Brazil). Platinum divinylcomplex, 2–3% in vinyl terminated poly(dimethylsiloxane), (Pt-catalyst) was purchased from Aldrich Chemical Company, Inc. (Milwaukee, USA). These materials were used as received. The mixture of D4Vi, D4H and Pt-catalyst was used in the infiltration step with the ceramic matrix composite to produce in situ the polycyclic silicone network as a SiCxOy precursor. Polymeric matrix composite manufacture Carbon fiber/bridge hybrid polysilsesquioxane composite was made by the hand lay-up technique by stacking 22 layers of carbon fabric impregnated with a mixture of PBFE/APS, in a stoichiometric NH2:epoxy ratio (1:1). The amine-epoxy addition reaction was followed by the hydrolysis and condensation reactions of alkoxysilane groups, in situ, promoted by environmental moisture. The bridge hybrid polysilsesquioxane matrix was cured at room temperature and aged for 24 h, to form the green-body laminate polymeric composite, named COMPOSITE 1. Specimens of this composite were trimmed in appropriate dimensions (4 · 4 · 5 mm) and evaluated by the Iosipescu shear test. A PBFE/APS monolithic polymeric sample was also obtained as a tough and colorless material, by the same procedure, to study the structural evolution from the polymer to the corresponding ceramic. Ceramic matrix composite manufacture Specimens obtained from the green-body laminate (COMPOSITE 1) were pyrolyzed under argon flow with the following heating cycle: heating to 180 C at 5 C/ min, holding for 60 min at this temperature; heating to 500 C at 5 C/min, holding for 60 min at this temperature; and heating to 1000 C, at 2C/min, holding for 120 min at this temperature. Finally the specimens were cooled to room temperature at 2 C/min. These pyrolyzed specimens were named C/C–SiO2. In order to obtain CMC with a reasonable density and improved mechanical properties, the C/C–SiO2 composite samples were infiltrated with a mixture of D4Vi and D4H, 58:42 (wt%), respectively, and 0.01 wt% of Pt-catalyst. The polymer was cured at room temperature giving rise to an in situ polycyclic silicone network (DVDH). The composite samples were subsequently pyrolyzed by using the same heating cycle described before. In this step, DVDH was converted to SiCxOy. Three infiltration/pyrolysis cycles were carried out, and the corresponding composite samples were named C/SiCxOy 1, C/SiCxOy 2, and C/SiCxOy 3. Finally, the C/SiCxOy 3 composite sample was coated with a pre-hydrolyzed phenyltriethoxysilane solution, which gave rise to a thin in situ poly(phenylsilsesquioxane) (POSS) film. This final composite was named COMPOSITE 2 (Fig. 1). To understand the ceramization process, all the composite specimens (COMPOSITE 1, C/C–SiO2, C/SiCxOy 1, C/SiCxOy 2, C/SiCxOy 3 and COMPOSITE 2) were characterized. 4246 J Mater Sci (2007) 42:4245–4253 123
J Mater Sci(2007)42:4245-4253 4247 were machined with a band power saw. In order to ceramic polymer avoid the direct access of oxygen to the bulk of the carbon fibers in the composites, through the lateral edge, these specimens were trimmed from a COM POSiTE 1 plate and submitted to similar infiltration/ pyrolysis cycles as described above, and also to POSS ③ coating procedure. Density measurements of the composites were performed on a picnometer (Mi- COMPOSITE cromeritics 1305). The samples were exhaustively purged with He before the measurements. The open porosity of each composite sample(small bars trimmed from the COMPOSitE 1 plate)was characterized by COMPOSITE 2 high-pressure mercury intrusion in an Autoscan-3 Quantachrome Porosimeter. The samples were out- Fig 1 Schematic diagram of the preparation steps of the gassed prior to analysis to facilitate filling of the pores Measurement techniques The losipescu test Monolithic polymeric samples Shear tests were conducted according to the astm d 5379 standard [20, 21]. They were performed in The structural evolution from polymeric precursors to Instron 4301 equipment for both COMPOSITE 1 and e corresponding ceramic matrixes was studied on COMPOSITE 2. The crosshead speed was 0.5 mm/ polymeric samples obtained using the same conditions min. The geometry of the losipescu specimen is shown as those used in the preparation of the composites. in Fig 2a. It consists of a beam with two 45 notches Infrared spectra(FTIR) were obtained in a Fourier machined opposite each other at the mid-point of the spectrometer(Bomem MB series), operating between specimen. By applying two coupled forces that gener- 4000and400 at 4 cm resolution. with the ate t' unter-acting moments, a pure and uniform conventional transmission KBr pellet technique. 2Si shear stress state is generated at section ac, between MAS NMR spectra were recorded on a spectrometer the two notches of the beam. The resulting shear and Bruker,model AC300P), operating at 59.62 MHz for moment diagrams are shown in Fig. 2b and c,respec- Si nucleus with a rotation frequency between 3 and tively. The maximum average shear stress is calculated 5 kHz. The Si one-pulse experiments required a by eq recycle delay of 60 s with 30%-pulse angle. Thermo- gravimetry (TGA)of the polymeric samples (-10 mg) ty=bt was p (2950, TA Instruments) at a heating rate of 20 C/min where P is the load, b is the width between the notches from 30 up to 980C, in a controlled dry argon flow of and t is the specimen thickness. The average values of 100 mL/min apparent shear stress were taken from measurements on five samples of each polymeric(COMPOSITE 1) Composite samples and ceramic (COMPOSITE 2)composite. The shear strain was calculated from the strain-gage data by the The thermo-oxidative behavior of all composite spec- relationship imens(10 mg) was studied by thermo-gravimetry in the same equipment as described above. heating from =E1-E2 room temperature to 1000C at 20C/min in a where &1 and e2 are the strains measured by the strain ynthetic air flow(Air Liquide: 80%N2, 20%O2). The gage elements that are +45 to the principal material morphology of the composite specimens was analyzed coordinates by field emission scanning electron microscopy ( FESEM)on a JEOL JSM-6340F equipment after each Results and discussion step of processing, i.e. molding, pyrolysis, infiltration/ pyrolysis and coating cycles. For FESEM analyses, Carbon fiber polymeric composites are among the rectangular shaped specimens (with 5 x 5 3.0 mm) strongest materials yet devised, in which the fibers
Measurement techniques Monolithic polymeric samples The structural evolution from polymeric precursors to the corresponding ceramic matrixes was studied on polymeric samples obtained using the same conditions as those used in the preparation of the composites. Infrared spectra (FTIR) were obtained in a Fourier spectrometer (Bomem MB series), operating between 4000 and 400 cm–1, at 4 cm–1 resolution, with the conventional transmission KBr pellet technique. 29Si MAS NMR spectra were recorded on a spectrometer (Bruker, model AC300P), operating at 59.62 MHz for 29Si nucleus with a rotation frequency between 3 and 5 kHz. The 29Si one-pulse experiments required a recycle delay of 60 s with 30-pulse angle. Thermogravimetry (TGA) of the polymeric samples (~10 mg) was performed on a thermo-gravimetric analyzer (2950, TA Instruments) at a heating rate of 20 C/min from 30 up to 980 C, in a controlled dry argon flow of 100 mL/min. Composite samples The thermo-oxidative behavior of all composite specimens (~10 mg) was studied by thermo-gravimetry in the same equipment as described above, heating from room temperature to 1000 C at 20 C/min in a synthetic air flow (Air Liquide: 80% N2, 20% O2). The morphology of the composite specimens was analyzed by field emission scanning electron microscopy (FESEM) on a JEOL JSM-6340F equipment after each step of processing, i.e. molding, pyrolysis, infiltration/ pyrolysis and coating cycles. For FESEM analyses, rectangular shaped specimens (with 5 · 5 · 3.0 mm) were machined with a band power saw. In order to avoid the direct access of oxygen to the bulk of the carbon fibers in the composites, through the lateral edge, these specimens were trimmed from a COMPOSITE 1 plate and submitted to similar infiltration/ pyrolysis cycles as described above, and also to POSS coating procedure. Density measurements of the composites were performed on a picnometer (Micromeritics 1305). The samples were exhaustively purged with He before the measurements. The open porosity of each composite sample (small bars trimmed from the COMPOSITE 1 plate) was characterized by high-pressure mercury intrusion in an Autoscan-33 Quantachrome Porosimeter. The samples were outgassed prior to analysis to facilitate filling of the pores with mercury. The Iosipescu test Shear tests were conducted according to the ASTM D 5379 standard [20, 21]. They were performed in an Instron 4301 equipment for both COMPOSITE 1 and COMPOSITE 2. The crosshead speed was 0.5 mm/ min. The geometry of the Iosipescu specimen is shown in Fig. 2a. It consists of a beam with two 45 notches machined opposite each other at the mid-point of the specimen. By applying two coupled forces that generate two counter-acting moments, a pure and uniform shear stress state is generated at section ac, between the two notches of the beam. The resulting shear and moment diagrams are shown in Fig. 2b and c, respectively. The maximum average shear stress is calculated by equation 1 [22]: sxy ¼ P bt ð1Þ where P is the load, b is the width between the notches and t is the specimen thickness. The average values of apparent shear stress were taken from measurements on five samples of each polymeric (COMPOSITE 1) and ceramic (COMPOSITE 2) composite. The shear strain was calculated from the strain-gage data by the relationship: c = e1 e2 ð2Þ where e1 and e2 are the strains measured by the strain gage elements that are ±45 to the principal material coordinates. Results and discussion Carbon fiber polymeric composites are among the strongest materials yet devised, in which the fibers Fig. 1 Schematic diagram of the preparation steps of the composites J Mater Sci (2007) 42:4245–4253 4247 123
J Mater Sci provide strength and the polymeric matrix allows the The pyrolysis of dvdh polycyclic silicone at fibers to remain together in the correct orientation, 1000C produces a SiC O, residue with a high ceramic protects them from damage, and enables the successive yield [24, 25]. TGa curves obtained for each constit carbon fiber layers to be laminated. In this study, bridge uent of the composites, in flowing argon, are shown in polysilsesquioxane(PBFE/APS)acted as the polymeric Fig 3. The thermal stability, evaluated as a function of matrix. It was prepared from a mixture of epoxy resin the initial weight loss temperature under an argon and APs, by an addition reaction, as follows: atmosphere, showed the following order: PBFE PBFE H2N(CH2kSi(och (Hs C2O)Si(CH2)3 NHCH2(OH)HC PBFE CH(OH)CH2NH(CH2)3Si(OC2H5) PBFE/APS alkoxysilane showed appropriate prop. APS DVDH carbon fibers. The initial erties as a polymeric precursor to intercalate between ture of weight loss, Ti, for PBFEAPs bridge poly carbon fiber layers, due to its high wettability and silsesquioxane was -100C. The first step of weight adhesion on the carbon fiber surface. In the presence of loss in this polymer, from -100 to -370C, was asso environmental moisture, the hydrolysis of Si-OC2Hs ciated with volatile evolution, mainly H2O and/or groups followed by the condensation of Si-OH ethanol, due to the thermally-induced condensation resulting groups gave rise to the in situ PBFE/APS reactions of residual Si-OH and/or Si-OC2Hs groups bridge polysilsesquioxane Concomitantly to the poly- T for DVDH and carbon fibers were -470C and condensation of Si-OC2Hs and Si-OH groups, part of -750C, respectively. A rapid oxidation of carbon these groups can react with C-OH usually present on fibers in air begins at temperatures above 500C. The the carbon fiber surface by condensation reactions ceramic yields in an argon atmosphere at 980C, for ibe/C—OH+—SH(Oc2Hs fibre/ C +H2O+(C2H5OH) These condensation reactions gave rise to a covalent carbon fibers, DVDH and PBFE-APS were 91.8+ 0.1 carbon fiber-polymeric matrix interface in COMPOS- 86.1+0.1 and 25.5+ 0.1%, respectively. The lower ITE 1. In addition, hydrogen bonds between the car- ceramic yield obtained for the polysilsesquioxane is bon fiber surface and the polar O-H and N-H groups, due to the higher organic content in this polymer in from the polymeric matrix, also contributed to the high relation to DVDH adhesion observed between the fiber surface and The evolution of the molecular structure of the PBFElAPS bridge polysilsesquioxane matrix PBFElAPS monolithic polymeric sample as a function The pyrolysis of ComPoSite l gave rise to a C/C- of the pyrolysis temperature was investigated by Sio2 composite, which was submitted to three infil- infrared spectroscopy, as shown in Fig 4. The broad tration/pyrolysis cycles with a mixture of (DAH and absorption with a maximum in the 1200-1000 cm D4Vi) functional cyclic siloxanes and Pt-catalyst. This range is due to Si-O-Si stretching and it was observed mixture produced in situ, by a hydrosilylation reaction, in all spectra, remaining up to 1000C. The broadening the elastomeric polycyclic silicone network named and intensification of this band with the increase of the DVDH 23, 24]. The hydrosilylation reaction is an pyrolysis temperature were associated to the Si-O-Si addition reaction,where no by-product is formed [23], enrichment in the material promoted by loss of the follows. volatile organic groups. Above 200C the absorption Si-H+ CH2=CH-Si- i-CH2-CH2-Si=+- Si-CH(CH3)-si- B-adduct a -adduct 2 Springer
provide strength and the polymeric matrix allows the fibers to remain together in the correct orientation, protects them from damage, and enables the successive carbon fiber layers to be laminated. In this study, bridge polysilsesquioxane (PBFE/APS) acted as the polymeric matrix. It was prepared from a mixture of epoxy resin and APS, by an addition reaction, as follows: PBFE/APS alkoxysilane showed appropriate properties as a polymeric precursor to intercalate between carbon fiber layers, due to its high wettability and adhesion on the carbon fiber surface. In the presence of environmental moisture, the hydrolysis of Si–OC2H5 groups followed by the condensation of Si–OH resulting groups gave rise to the in situ PBFE/APS bridge polysilsesquioxane. Concomitantly to the polycondensation of Si–OC2H5 and Si–OH groups, part of these groups can react with C–OH usually present on the carbon fiber surface, by condensation reactions: These condensation reactions gave rise to a covalent carbon fiber-polymeric matrix interface in COMPOSITE 1. In addition, hydrogen bonds between the carbon fiber surface and the polar O–H and N–H groups, from the polymeric matrix, also contributed to the high adhesion observed between the fiber surface and PBFE/APS bridge polysilsesquioxane matrix. The pyrolysis of COMPOSITE 1 gave rise to a C/C– SiO2 composite, which was submitted to three infiltration/pyrolysis cycles with a mixture of (D4H and D4Vi) functional cyclic siloxanes and Pt-catalyst. This mixture produced in situ, by a hydrosilylation reaction, the elastomeric polycyclic silicone network named DVDH [23, 24]. The hydrosilylation reaction is an addition reaction, where no by-product is formed [23], as follows: The pyrolysis of DVDH polycyclic silicone at 1000 C produces a SiCxOy residue with a high ceramic yield [24, 25]. TGA curves obtained for each constituent of the composites, in flowing argon, are shown in Fig. 3. The thermal stability, evaluated as a function of the initial weight loss temperature under an argon atmosphere, showed the following order: PBFE/ APS < DVDH < carbon fibers. The initial temperature of weight loss, Ti, for PBFE/APS bridge polysilsesquioxane was ~100 C. The first step of weight loss in this polymer, from ~100 to ~370 C, was associated with volatile evolution, mainly H2O and/or ethanol, due to the thermally-induced condensation reactions of residual Si–OH and/or Si–OC2H5 groups. Ti for DVDH and carbon fibers were ~470 C and ~750 C, respectively. A rapid oxidation of carbon fibers in air begins at temperatures above 500 C. The ceramic yields in an argon atmosphere at 980 C, for carbon fibers, DVDH and PBFE-APS were 91.8 ± 0.1, 86.1 ± 0.1 and 25.5 ± 0.1%, respectively. The lower ceramic yield obtained for the polysilsesquioxane is due to the higher organic content in this polymer in relation to DVDH. The evolution of the molecular structure of the PBFE/APS monolithic polymeric sample as a function of the pyrolysis temperature was investigated by infrared spectroscopy, as shown in Fig. 4. The broad absorption with a maximum in the 1200–1000 cm–1 range is due to Si–O–Si stretching and it was observed in all spectra, remaining up to 1000 C. The broadening and intensification of this band with the increase of the pyrolysis temperature were associated to the Si–O–Si enrichment in the material promoted by loss of the volatile organic groups. Above 200 C the absorption H2C CH O HC CH2 O 2H2N(CH2)3Si(OC2H5)3 (H5C2O)3Si(CH2)3NHCH2(OH)HC CH(OH)CH2NH(CH2)3Si(OC2H5) PBFE 3 PBFE fibre C OH SiOH(OC2H5) fibre C O Si H2O (C2H5OH) Si-H + CH2=CH-Si Si-CH2-CH2-Si + Si-CH(CH3)-Si β-adduct α-adduct 4248 J Mater Sci (2007) 42:4245–4253 123
J Mater Sci(2007)42:4245-4253 4249 d|400° 排 200°C e」 PBFE/APS 40003500300025002000 L Fig 4 Infrared spectra of PBFE/APS at differen emperatures P L/L-b P.L儿L-b residue presented only a very broad and weak +Pb/2 absorption, with maximum at -1080 cm, associated with Si-O-Si groups. The low intensity of this absorption suggests the presence of small amount of Sio2 dispersed in the carbon residue. In addition, absorptions of the carbon phase that are very weak P.b/2 the infrared [10, 24, 25, were not observed The structural evolution of dvdh to the ceramic ig. 2(a)Geometry of the test losipescu specimen (b) Shear phase as a function of the pyrolysis temperature can be diagram.(c) Moment diagra observed in Fig. 5. Absorptions at 3055 and 1610 cm are characteristic of vinyl residual groups, corre- rasiloxane ring of the polysilsesquioxane network. sponding to c-H and C=C stretchings, respectively at 480 cm. associated with the breath mode of tet- The strong band at 2160 cm is typical of si band D1, was observed [26-28] On heating the sample, stretching [10, 23, 24], which suggests the presence of organic groups were released, promoting the decrease residual Si-h groups in the polymeric precursor. As in relative intensities of the absorptions at 2970-2840, can be seen, practically all the C=C groups were con 1410 and 1250 cm", which are due to aliphatic CHi sumed at 400C, indicating the increase in the degree nd CH3 groups, and also at 1600 and 1500 cm due to of connection between the siloxane cycles at this e phenyl rings of the polymeric structure. For the PBFE/APS sample pyrolyzed at 1000C, the dark rbon fabric 600 DVDH 40 PBFEJAPS 30 4000350030002500200015001000500 Temperature(C) Wavenumber(cm) Fig 3 TGA curves of carbon fabric and polymers used in the Fig. 5 Infrared spectra of DVDH polycyclic silicone network at of the composites(Ar flow at 20C/min) different pyrolysis temperatures
at ~480 cm–1, associated with the breath mode of tetrasiloxane ring of the polysilsesquioxane network, band D1, was observed [26–28]. On heating the sample, organic groups were released, promoting the decrease in relative intensities of the absorptions at 2970–2840, 1410 and 1250 cm–1, which are due to aliphatic CH2 and CH3 groups, and also at 1600 and 1500 cm–1 due to the phenyl rings of the polymeric structure. For the PBFE/APS sample pyrolyzed at 1000 C, the dark residue presented only a very broad and weak absorption, with maximum at ~1080 cm–1, associated with Si–O–Si groups. The low intensity of this absorption suggests the presence of small amount of SiO2 dispersed in the carbon residue. In addition, absorptions of the carbon phase that are very weak in the infrared [10, 24, 25], were not observed. The structural evolution of DVDH to the ceramic phase as a function of the pyrolysis temperature can be observed in Fig. 5. Absorptions at 3055 and 1610 cm–1 are characteristic of vinyl residual groups, corresponding to C–H and C=C stretchings, respectively. The strong band at 2160 cm–1 is typical of Si–H stretching [10, 23, 24], which suggests the presence of residual Si–H groups in the polymeric precursor. As can be seen, practically all the C=C groups were consumed at 400 C, indicating the increase in the degree of connection between the siloxane cycles at this Fig. 5 Infrared spectra of DVDH polycyclic silicone network at different pyrolysis temperatures Fig. 2 (a) Geometry of the test Iosipescu specimen. (b) Shear diagram. (c) Moment diagram Fig. 3 TGA curves of carbon fabric and polymers used in the preparation of the composites (Ar flow at 20 C/min) Fig. 4 Infrared spectra of PBFE/APS at different pyrolysis temperatures J Mater Sci (2007) 42:4245–4253 4249 123
J Mater Sci Table 1 Apparent density and porosity of the composites Ig cm-1 Porosity [% C-SiO, (PBFE/APS COMPOSITE 1 1.50 C/SiC.O, 3 COMPOSITE 2 1.76 00 high amount of organic volatiles from the PBFE/APS polymeric phase. However, the apparent density Chemical shift(ppm) increased due to volume contraction. The subsequent Fig.6Si MAS NMR spectra of the ceramic residues derived filtration/pyrolysis cycles contributed to an increase from the pyrolysis of PBFE/APS and DVDH at 1000C under in the density values and a decrease in the porosity, due to enrichment in the SiCx Oy ceramic phase. In fact this behavior was applied to C/SiC,O, I and ciSicroy 2 composites. A distinct behavior was found for the C/ temperature. Residual amounts of Si-H groups still SiC,, 3 composite, due to the coalescence of closed were detected in the product up to 600C. As and isolated pores generating areas with open porosit described by Schiavon et al. [24], after the consumption In COMPOSITE 2 the thin layer of Poss polymeric of Si-H groups by thermal induced hydrosilylation, film reduced considerably the porosity by filling the generated by reorganization reactions of pores and also contributed to a slight dec in the siloxane chains during pyrolysis. Between 200 and density value of this composite 600C, the decreases in the relative intensities of the The thermo-oxidation experiments were carried out aliphatic C-H absorptions in the 2960-2890 cm range up to 1000C in the TGA oven under a synthetic air- and at 1263 cm- was significant, suggesting that the flow. The results are shown in Fig. 7. COMPOSITE 1 mineralization of the dVdh polymeric network star- showed its temperature of maximum degradation rate, ted around 600oC. Two broad bands centered at 1080 Td, at 840C, with a 5%residue at 1000C. The c/ and 800 cm were assigned to Si-o-Si and Si-c C-SiOz composite showed a better thermo-oxidative stretchings, respectively, in the sample pyrolyzed at performance than the former. In addition, the intro- 1000C under an inert atmosphere. The spectrum of duction of the Sic oy phase by infiltration/pyrolysis this sample showed a profile characteristic of a carbi- cycles of dVdh polycyclic silicone (CISiCrOy l, c dic-rich SiC.O, ceramic [24 SiCrOy 2 and C/SiC,oy 3)led to an increase in Td The ceramic matrixes obtained from PbFE/APS and values, in relation to the former due to the effective characterized by 2Si MAS NMR, as shown in Fig. 6. to oxygen diffusion. Similar results were described by The ceramic product derived from PBFE/APS showed only a broad peak centered at-108 ppm, associated to SiO4 sites, characteristic of silica(SiO2). This resulting silica was dispersed in a rich carbon phase [25]. For DVDH, the spectrum showed peaks at-106, -72,-32 16 and 5 ppm, associated with a random distribution of SiO4, SiO3C, SiO2C, SiCa and Sioc3 sites in the SiC,O, residue, respectively [10, 23, 24, 29]. In addition 01=-CC sc9 一 COMPOSITE1 to this, a high abundance of SiC4, compared to usual ceramic products obtained from the pyrolysis of alk oxosilane gels, was also observed [24, 26, 30-32 The apparent densities and porosities of the com- COMPOSITE 2 posites at various stages of the preparation process are given in Table 1. The pyrolysis of COMPOSItE 1 led to the formation of the C/C-sioz sample, which Temperature(°C) showed cracks and pores caused by the evolution of a 7 TGA curves under synthetic air-flow for all composites 2 Springer
temperature. Residual amounts of Si–H groups still were detected in the product up to 600 C. As described by Schiavon et al. [24], after the consumption of Si–H groups by thermal induced hydrosilylation, they are generated by reorganization reactions of siloxane chains during pyrolysis. Between 200 and 600 C, the decreases in the relative intensities of the aliphatic C–H absorptions in the 2960–2890 cm–1 range and at 1263 cm–1 was significant, suggesting that the mineralization of the DVDH polymeric network started around 600 C. Two broad bands centered at 1080 and 800 cm–1 were assigned to Si–O–Si and Si–C stretchings, respectively, in the sample pyrolyzed at 1000 C under an inert atmosphere. The spectrum of this sample showed a profile characteristic of a carbidic-rich SiCxOy ceramic [24]. The ceramic matrixes obtained from PBFE/APS and DVDH polymers by pyrolysis at 1000 C were also characterized by 29Si MAS NMR, as shown in Fig. 6. The ceramic product derived from PBFE/APS showed only a broad peak centered at –108 ppm, associated to SiO4 sites, characteristic of silica (SiO2). This resulting silica was dispersed in a rich carbon phase [25]. For DVDH, the spectrum showed peaks at –106, –72, –32, –16 and 5 ppm, associated with a random distribution of SiO4, SiO3C, SiO2C2, SiC4 and SiOC3 sites in the SiCxOy residue, respectively [10, 23, 24, 29]. In addition to this, a high abundance of SiC4, compared to usual ceramic products obtained from the pyrolysis of alkoxysilane gels, was also observed [24, 26, 30–32]. The apparent densities and porosities of the composites at various stages of the preparation process are given in Table 1. The pyrolysis of COMPOSITE 1 led to the formation of the C/C–SiO2 sample, which showed cracks and pores caused by the evolution of a high amount of organic volatiles from the PBFE/APS polymeric phase. However, the apparent density increased due to volume contraction. The subsequent infiltration/pyrolysis cycles contributed to an increase in the density values and a decrease in the porosity, due to enrichment in the SiCxOy ceramic phase. In fact, this behavior was applied to C/SiCxOy 1 and C/SiCxOy 2 composites. A distinct behavior was found for the C/ SiCxOy 3 composite, due to the coalescence of closed and isolated pores generating areas with open porosity. In COMPOSITE 2 the thin layer of POSS polymeric film reduced considerably the porosity by filling the pores and also contributed to a slight decrease in the density value of this composite. The thermo-oxidation experiments were carried out up to 1000 C in the TGA oven under a synthetic air- flow. The results are shown in Fig. 7. COMPOSITE 1 showed its temperature of maximum degradation rate, Td, at 840 C, with a ~5% residue at 1000 C. The C/ C–SiO2 composite showed a better thermo-oxidative performance than the former. In addition, the introduction of the SiCxOy phase by infiltration/pyrolysis cycles of DVDH polycyclic silicone (C/SiCxOy 1, C/ SiCxOy 2 and C/SiCxOy 3) led to an increase in Td values, in relation to the former, due to the effective protection by SiCxOy of the carbon fibers with respect to oxygen diffusion. Similar results were described by Fig. 6 29Si MAS NMR spectra of the ceramic residues derived from the pyrolysis of PBFE/APS and DVDH at 1000 C under Ar flow Table 1 Apparent density and porosity of the composites Composites Apparent density [g cm–3] Porosity [%] COMPOSITE 1 1.50 3.2 C/SiO2 1.59 19.0 C/SiCxOy 1 1.75 7.0 C/SiCxOy 2 1.82 3.5 C/SiCxOy 3 1.80 4.5 COMPOSITE 2 1.76 3.0 Fig. 7 TGA curves under synthetic air-flow for all composites 4250 J Mater Sci (2007) 42:4245–4253 123
J Mater Sci(2007)42:4245-4253 Manocha et al. [33 in SiC.Ov-coated carbon fibers. Little fractures in the matrix phase surface and at the COMPOSITE 2 showed Td at -890C and a-53% fiber-matrix interface may have been mechanically residue at 1000C. This result also reflects the generated during the cutting of the samples due to saw remarkable effect of the poly(phenylsilsesquioxane) vibration. In order to seal the remaining pores in the C/ coating layer, as an additional barrier to oxygen dif- SiC., 3 composite, a further coating with a POss fusion into the composite layer was made. The micrograph of this coated com Figure 8 shows SEM micrographs of polymeric and posite surface(COMPOSITE 2) revealed that a uni- ceramic composites. Good adhesion at the PBFE/APs form thin POSS layer, with a very smooth surface, was polymeric matrix/carbon fiber interface was observed formed on the analyzed specimen(Fig. &f). In this in COMPOSITE l, as can be seen in Fig 8a. Figure 8b composite the carbon fibers and ceramic phase were illustrates typical pores in the C/C-SiO2 composite shielded by a POSS layer. produced by organic volatile evolution during the Tests for measuring shear strength and modulus of pyrolysis of the PBFE/APS polymeric phase. The composites have been used in the characterization of porosity in the subsequent composites was significantly composites [20, 22, 34-37. In these tests the main dif reduced due to the presence of the SiC.O, phase ficulty is to assure a nearly pure shear stress state in the formed by infiltration/pyrolysis cycles of the DVDH gage section of the specimen at a constant magnitude polycyclic silicone network. A good adhesion of this In fiber-reinforced composites there is an additional ceramic phase on the carbon fiber surface was also difficulty related to the orientation of fibers. Unlike observed, as can be seen in Fig. &c-e, even considering with isou. it is necessary to point out the plane where pic materials, when measuring composite the possible mechanical damages promoted by the propertie power saw in the preparation step for these samples. the load is being applied, and the orientation of the Fig 8 FESEM micrographs of polymeric and ceramic tes COMPOSITE 1:(b)C/C- SiO2:(c)C/SiC,O, 1;(d)Cl Sic,o C/SiC:Oy 3, and (f COMPOSITE 2
Manocha et al. [33] in SiCxOy-coated carbon fibers. COMPOSITE 2 showed Td at ~890 C and a ~53% residue at 1000 C. This result also reflects the remarkable effect of the poly(phenylsilsesquioxane) coating layer, as an additional barrier to oxygen diffusion into the composite. Figure 8 shows SEM micrographs of polymeric and ceramic composites. Good adhesion at the PBFE/APS polymeric matrix/carbon fiber interface was observed in COMPOSITE 1, as can be seen in Fig. 8a. Figure 8b illustrates typical pores in the C/C–SiO2 composite produced by organic volatile evolution during the pyrolysis of the PBFE/APS polymeric phase. The porosity in the subsequent composites was significantly reduced due to the presence of the SiCxOy phase formed by infiltration/pyrolysis cycles of the DVDH polycyclic silicone network. A good adhesion of this ceramic phase on the carbon fiber surface was also observed, as can be seen in Fig. 8c–e, even considering the possible mechanical damages promoted by the power saw in the preparation step for these samples. Little fractures in the matrix phase surface and at the fiber-matrix interface may have been mechanically generated during the cutting of the samples due to saw vibration. In order to seal the remaining pores in the C/ SiCxOy 3 composite, a further coating with a POSS layer was made. The micrograph of this coated composite surface (COMPOSITE 2) revealed that a uniform thin POSS layer, with a very smooth surface, was formed on the analyzed specimen (Fig. 8f). In this composite the carbon fibers and ceramic phase were shielded by a POSS layer. Tests for measuring shear strength and modulus of composites have been used in the characterization of composites [20, 22, 34–37]. In these tests the main dif- ficulty is to assure a nearly pure shear stress state in the gage section of the specimen at a constant magnitude. In fiber-reinforced composites there is an additional difficulty related to the orientation of fibers. Unlike with isotropic materials, when measuring composite properties it is necessary to point out the plane where the load is being applied, and the orientation of the Fig. 8 FESEM micrographs of polymeric and ceramic composites: (a) COMPOSITE 1; (b) C/C– SiO2; (c) C/SiCxOy 1; (d) C/ SiCxOy 2; (e) C/SiCxOy 3, and (f) COMPOSITE 2 J Mater Sci (2007) 42:4245–4253 4251 123
J Mater Sci(2007)42:4245-4253 fiber axis in relation to the applied load. Only in 1992, the v-notch losipescu test was established as a standard p [34] direct comparison of the shear test results between COMPOSITE 1 and COmPOSIte 2 is obviously inappropriate since their matrixes are different. These composites were prepared by stacking Plain weave fabric layers and counter-reacting forces were applied, as shown in Fig. 2. Considering the axis system for the test specimen, the shear modulus to be measured is Gr For COMPOSITE 1(carbon fiber/PBFE/APS poly- meric matrix)the average in-plane shear strength and the shear modulus were 44.2+19 MPa and 2.2 +0.5 GPa, respectively. For comparison, the shear strength of an epoxy resin specimen measured by tor sion can vary from 40 to 90 MPa, depending on the rate of loading [35]. The in-plane shear strength and shear modulus for laminate composites is highly dependent on the matrix and interfacial properties. Similar values of the shear strength and shear modulus were also de- scribed by Odegard and Kumosa [36]and Chiang and Fig 10 Representation of main failure modes found in the Jianmei [37], respectively. Figure 9 shows a typical losipescu shear testing stress-strain curve of the losipescu shear test for COMPOSITE 1. The curve shows a typical non-linear weave fabric. there is an additional failure mechanism behavior up to failure shear stress. For composites besides fiber/matrix interface slipping and interlaminar having reinforcing fibers perpendicular to the shear shear, associated with the fiber bundle splitting in the loading direction, the failure mode occurs mainly by fabric mesh, as shown schematically in Fig. 10c. fiber slipping and debonding at the fiber/matrix inter- Figure 11 shows a typical stress-strain curve of the face, as schematically represented in Fig. 10a. On the losipescu shear test for COMPOSITE 2(C/SiC,O,3/ other hand, in composites having reinforcing fibers POSS). The average shear strength and shear modulus parallel to the shear loading direction, the failure may for COMPOSITE 2 were 14.2+4.1 MPa and 15.0+ occur by an interlaminar crack at the sample gage 2.0 GPa, respectively. Non-linearity was observed to a length, as schematically represented in Fig. 10b. In any minor extent and the deformation level up to failure was case, shear properties are mainly dominated by the low (-0, 2%). Similar results for shear strength matrix and the fiber/matrix interface and the failure found by Duran et al. [38] in C/SiC composites h modes are associated with a shear deformation mech- 10% porosity. In addition, Brondsted et al anism. For COMPOSitE 1 which is made with a Plain 0000 20000 1000125015001750 Fig 9 Typical shear stress curve as a function of shear strain for Fig. 11 Typical shear stress curve as a function of shear strai meric compost for the ceramic co 2 Springer
fiber axis in relation to the applied load. Only in 1992, the V-notch Iosipescu test was established as a standard procedure for shear testing [34]. In this investigation, a direct comparison of the shear test results between COMPOSITE 1 and COMPOSITE 2 is obviously inappropriate since their matrixes are different. These composites were prepared by stacking Plain weave fabric layers and counter-reacting forces were applied, as shown in Fig. 2. Considering the axis system for the test specimen, the shear modulus to be measured is Gxy. For COMPOSITE 1 (carbon fiber/PBFE/APS polymeric matrix) the average in-plane shear strength and the shear modulus were 44.2 ± 1.9 MPa and 2.2 ± 0.5 GPa, respectively. For comparison, the shear strength of an epoxy resin specimen measured by torsion can vary from 40 to 90 MPa, depending on the rate of loading [35]. The in-plane shear strength and shear modulus for laminate composites is highly dependent on the matrix and interfacial properties. Similar values of the shear strength and shear modulus were also described by Odegard and Kumosa [36] and Chiang and Jianmei [37], respectively. Figure 9 shows a typical stress-strain curve of the Iosipescu shear test for COMPOSITE 1. The curve shows a typical non-linear behavior up to failure shear stress. For composites having reinforcing fibers perpendicular to the shear loading direction, the failure mode occurs mainly by fiber slipping and debonding at the fiber/matrix interface, as schematically represented in Fig. 10a. On the other hand, in composites having reinforcing fibers parallel to the shear loading direction, the failure may occur by an interlaminar crack at the sample gage length, as schematically represented in Fig. 10b. In any case, shear properties are mainly dominated by the matrix and the fiber/matrix interface and the failure modes are associated with a shear deformation mechanism. For COMPOSITE 1, which is made with a Plain weave fabric, there is an additional failure mechanism, besides fiber/matrix interface slipping and interlaminar shear, associated with the fiber bundle splitting in the fabric mesh, as shown schematically in Fig. 10c. Figure 11 shows a typical stress-strain curve of the Iosipescu shear test for COMPOSITE 2 (C/SiCxOy 3/ POSS). The average shear strength and shear modulus for COMPOSITE 2 were 14.2 ± 4.1 MPa and 15.0 ± 2.0 GPa, respectively. Non-linearity was observed to a minor extent and the deformation level up to failure was low (~0,2%). Similar results for shear strength were found by Dura´n et al. [38] in C/SiC composites having 10% porosity. In addition, Bro¨ndsted et al. [39] Fig. 9 Typical shear stress curve as a function of shear strain for the polymeric composite Fig. 10 Representation of main failure modes found in the Iosipescu shear testing Fig. 11 Typical shear stress curve as a function of shear strain for the ceramic composites 4252 J Mater Sci (2007) 42:4245–4253 123
J Mater Sci(2007)42:4245-4253 4253 described similar values for the in-plane shear modulus 4. Ohnabe H, Masaki M, Onozuka M. Miyahara M, Sasa T for a C/Sic composite. Apart from the shear failure 999)Compos Part A Appl Sci Manuf 30: 489 5. Mckee Dw(1987) Carbon 25: 551 mechanisms described for polymeric matrix composites, 6. Eherburger P Lahaye J(1981) Carbon 19: 7 the shear behavior of CMC, as in the case of COM- 7. Manoucha LM(1994 )Carbon 32: 213 POSITE 2, is also influenced by the presence of pores 8. Zhou X, Zhang C, Ma J, Zhou A(1999)Key Eng Mater and microcracks in the bulk matrix and at the fiber/ 164:43 matrix interface. These features can explain the low 10.Radovanovic E, Gozzi mF, Goncalves MC, Yoshida IVP shear strength found in these latter composites. Due to (1999)J Non-Cryst Solids 248: 37 the fragile nature of the CMC, the failure in COM- 11. Wonderly C, Grenestedt J, Ferlung G, E Cepus(2005) POSITE2 occurred at the gage section of the specimen, 12.Gozzi MF Goncalves mc. Yoshida IvP (1999)] Mater Sci 34:155 experimental data is relatively high 13. Schiavon MA, Sorari GD, Yoshida IVP(2002)J Non-Cryst Solids 304: 76 14. Schiavon MA, Soraru GD, Yoshida IVP(2004)J Non-Cryst Solids 348: 156 Conclusions 15. Greil P(1995)J Am Ceram Soc 78: 835 16. Kaindl A, Lehner W, Greil P, Kim DJ(1999) Mater Sci Eng Bridge polysilsesquioxane produced by the in situ A Struct 260: 101 reaction between an epoxy resin(PBFE)and APS was 17. Krenkel W, Heidenreich B, Renz R(2002)Adv Eng Mater an excellent polymeric matrix to produce carbon fiber- 18. Davies I, Hamada H(2001) Adv Compos Mater 10: 77 reinforced polymeric composites(COMPOSITE 1), 19. Twitty A, Russellfloyd RS, Cooke RG, Harris B (1995)J due to the high adhesion at the fiber-polymer interface Eur Ceram Soc 15: 455 The pyrolysis of this polymeric composite followed by 20. ASTM D5379-93(1993)In Standard method for shear further infiltration/pyrolysis cycles with a DVDH properties of composite materials by the V-notched beam method. American Society for the Testing of Materials, New polycyclic silicone network, led to a progressive in- crease in the apparent density and in the thermo-oxi- 21. losipescu N(1967)J Mater 2: 537 dation resistance of the ceramic composites, due to 22. Tamopo'skii M, arnautov Ak, Kulakov AvL (1999) SiCr Oy enrichment of the ceramic phase. In addition, a 23 Redondo SUA, Radovanovic E. Torriani IL, Yoshida IvP good adhesion between carbon fibers and the ceramic phase was observed. The highest thermo-oxidative 24 Schiavon MA, Radovanovic E, Yoshida IVP(2002)Powder resistance was found for the ceramic composite coated with poly(phenylsilsesquioxane), named COMPOSITE 25. Schiavon MA. Redondo SUA. Pina Sro. Yoshida IvP 2002)J Non-Cryst Solids 304: 92 2. Shear properties(strength and modulus)obtained by 26. Li X, King TA (1996)J Non-Cryst Solids 204:235 e losipescu method for COMPOSITES 1 and 27. Bornhauser P, Calzaferri G(1996)J Phys Chem 100: 2035 showed a dependence on the nature of the matrix. The 28. Bellamy L(966) In The infrared spectra of complex mol average in-plane shear strength and shear modulus 29. Hurwitz fl. Kacik ta. bu xy. masnovi j. heimann pi were442±1.9 MPa and22±0.5 GPa for the poly Beyene K(1995)J Mater Sci 30. 3130 meric matrix composite(COMPOSITE 1), respec- 30. Belot V, Corriu RJP, Leclerq D, Mutin PH, Vioux A(1992) tively. For the ceramic matrix composite ( COMPOSITE 2)the values were 14.2+ 4.1 MPa and 31. Renlund GM, Prochaska S, Doremus RH (1991)J Mater 15.0+2.0 GPa, respectively. Additionally, for the last 32. Pantano CG, Singh AK, Zhang H(1999)J Sol-Gel Sci composite these properties seemed to be governed by Technol 14: 7 the microstructure of the ceramic matrix. Failure 33 Manocha LM, Manocha S, Patel KB, Glogar P(2000)Car bon38:1481 modes observed for the composites were described by 34. Liu JY(2000) In Shear test fixture design for orthotropic a shear deformation mechanism materials, International Community for Composite Engi neering, Denver, ICCE/7 Acknowledgments We gratefully acknowledge financial 35. Gilat A, Goldberg RK, Roberts GD(2005)In Strain rate port from CNPq and FAPESP(Process 00/06882- sensitivity of epoxy resin in tensile and shear loading, NASA/ TM,2005-213595 36. Odergard G, Kumosa M(2000)Compos Sci References 37. Chiang MYM, Jianmei H (2002)Compos Pa 38. Duran A, Aparicio M, Rebstock K, Vogel w Mater 127- 287 39 Brondsted P, Heredia FE, Evans AG(1994)J Am Ceram 2. Erauzkin E, Llorca J(1997) Key Eng Mater 127: 761 3. Wilshire B, Carreno F(2000)J Euro Ceram Soc 20: 463
described similar values for the in-plane shear modulus for a C/SiC composite. Apart from the shear failure mechanisms described for polymeric matrix composites, the shear behavior of CMC, as in the case of COMPOSITE 2, is also influenced by the presence of pores and microcracks in the bulk matrix and at the fiber/ matrix interface. These features can explain the low shear strength found in these latter composites. Due to the fragile nature of the CMC, the failure in COMPOSITE 2 occurred at the gage section of the specimen, as represented in Fig. 10b, and the scatter of the experimental data is relatively high. Conclusions Bridge polysilsesquioxane produced by the in situ reaction between an epoxy resin (PBFE) and APS was an excellent polymeric matrix to produce carbon fiberreinforced polymeric composites (COMPOSITE 1), due to the high adhesion at the fiber-polymer interface. The pyrolysis of this polymeric composite followed by further infiltration/pyrolysis cycles with a DVDH polycyclic silicone network, led to a progressive increase in the apparent density and in the thermo-oxidation resistance of the ceramic composites, due to SiCxOy enrichment of the ceramic phase. In addition, a good adhesion between carbon fibers and the ceramic phase was observed. The highest thermo-oxidative resistance was found for the ceramic composite coated with poly(phenylsilsesquioxane), named COMPOSITE 2. Shear properties (strength and modulus) obtained by the Iosipescu method for COMPOSITES 1 and 2 showed a dependence on the nature of the matrix. The average in-plane shear strength and shear modulus were 44.2 ± 1.9 MPa and 2.2 ± 0.5 GPa for the polymeric matrix composite (COMPOSITE 1), respectively. For the ceramic matrix composite (COMPOSITE 2) the values were 14.2 ± 4.1 MPa and 15.0 ± 2.0 GPa, respectively. Additionally, for the last composite these properties seemed to be governed by the microstructure of the ceramic matrix. Failure modes observed for the composites were described by a shear deformation mechanism. Acknowledgments We gratefully acknowledge financial support from CNPq and FAPESP (Process 00/06882-5). References 1. Wang C, Zhu Z, Hou X, Li H (2000) Carbon 38:1821 2. Erauzkin E, Llorca J (1997) Key Eng Mater 127:761 3. Wilshire B, Carren˜ o F (2000) J Euro Ceram Soc 20:463 4. Ohnabe H, Masaki M, Onozuka M, Miyahara M, Sasa T (1999) Compos Part A Appl Sci Manuf 30:489 5. Mckee DW (1987) Carbon 25:551 6. Eherburger P, Lahaye J (1981) Carbon 19:7 7. Manoucha LM (1994) Carbon 32:213 8. Zhou X, Zhang C, Ma J, Zhou A (1999) Key Eng Mater 164:43 9. Shimoo T, Okamura K, Toyoda F (2000) J Mater Sci 35:3811 10. Radovanovic E, Gozzi MF, Gonc¸alves MC, Yoshida IVP (1999) J Non-Cryst Solids 248:37 11. Wonderly C, Grenestedt J, Ferlung G, E Cepus (2005) Compos Part B Eng 36:417 12. Gozzi MF, Gonc¸alves MC, Yoshida IVP (1999) J Mater Sci 34:155 13. Schiavon MA, Soraru` GD, Yoshida IVP (2002) J Non-Cryst Solids 304:76 14. Schiavon MA, Soraru` GD, Yoshida IVP (2004) J Non-Cryst Solids 348:156 15. Greil P (1995) J Am Ceram Soc 78:835 16. Kaindl A, Lehner W, Greil P, Kim DJ (1999) Mater Sci Eng A Struct 260:101 17. Krenkel W, Heidenreich B, Renz R (2002) Adv Eng Mater 4:427 18. Davies IJ, Hamada H (2001) Adv Compos Mater 10:77 19. Twitty A, Russellfloyd RS, Cooke RG, Harris B. (1995) J Eur Ceram Soc 15:455 20. ASTM D5379-93 (1993) In Standard method for shear properties of composite materials by the V-notched beam method. American Society for the Testing of Materials, New York 21. Iosipescu N (1967) J Mater 2:537 22. Tarnopo’skii YM, Arnautov AK, Kulakov AVL (1999) Compos Part A Appl Sci Manuf 30:879 23. Redondo SUA, Radovanovic E, Torriani IL, Yoshida IVP (2001) Polymer 42:1319 24. Schiavon MA, Radovanovic E, Yoshida IVP (2002) Powder Technol 123:232 25. Schiavon MA, Redondo SUA, Pina SRO, Yoshida IVP (2002) J Non-Cryst Solids 304:92 26. Li X, King TA (1996) J Non-Cryst Solids 204:235 27. Bornhauser P, Calzaferri G (1996) J Phys Chem 100:2035 28. Bellamy LJ (1966) In The infrared spectra of complex molecules, Methuen, London 29. Hurwitz FI, Kacik TA, Bu XY, Masnovi J, Heimann PJ, Beyene K (1995) J Mater Sci 30:3130 30. Belot V, Corriu RJP, Leclerq D, Mutin PH, Vioux A (1992) J Non-Cryst Solids 144:287 31. Renlund GM, Prochaska S, Doremus RH (1991) J Mater Res 6:2716 32. Pantano CG, Singh AK, Zhang H (1999) J Sol-Gel Sci Technol 14:7 33. Manocha LM, Manocha S, Patel KB, Glogar P (2000) Carbon 38:1481 34. Liu JY (2000) In Shear test fixture design for orthotropic materials, International Community for Composite Engineering, Denver, ICCE/7 35. Gilat A, Goldberg RK, Roberts GD (2005) In Strain rate sensitivity of epoxy resin in tensile and shear loading, NASA/ TM, 2005-213595 36. Odergard G, Kumosa M (2000) Compos Sci Technol 60:2917 37. Chiang MYM, Jianmei H (2002) Compos Part B Eng 33:461 38. Dura´n A, Aparicio M, Rebstock K, Vogel W (1997) Key Eng Mater 127:287 39. Bro¨ndsted P, Heredia FE, Evans AG (1994) J Am Ceram Soc 77:2569 J Mater Sci (2007) 42:4245–4253 4253 123
Copyright of Journal of Materials Science is the property of Springer Science Business Media B.V. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holders express written permission. However, users may print, download or email articles for individual use