40TH ANNIVERSARY J MATER SCI41(006)823-839 A review of the development of three generations of small diameter silicon carbide fibres A.R. BUNSELL A. IANT Ecole des Mines de Paris, Centre des Materiaux, BP87, 91003 Evry Cedex, France Three generations of small diameter ceramic fibres based on polycrystalline silicon carbide have been developed over a period of thirty years. This has been possible due to studies into the relationships between the microstructures and properties of the fibres a variety of techniques have been employed by research teams on three continents the fibres are made by he conversion of polymer precursors to ceramic fibres and all three generations are presently produced commercially. The nature of the precursor and the techniques used for cross-linking have been varied in order to optimise both properties and cost of manufacture. It has been possible to improve the characteristics of the fibres as the processes involved in the cross-linking of the precursor fibres have been better understood and the mechanisms governing both room temperature and high temperature behaviour determined. The result is that, although first generation fibres were limited by a low Young s modulus at room limiting the behaviour of bulk silicon carbide, the third generation fibres shows many ot tho oo temperature and by creep and instability of the structure at temperatures far lower than thos characteristics of stoichiometric silicon carbide. This remarkable improvement in characteristics has been due to a thorough understanding of the materials science governing the behaviour of these fibres which are reinforcements for ceramic matrix composite materials 2006 Springer Science Business Media, Inc. 1 Introduction The PCs obtained in this way could be melt spun to give Silicon carbide fibres, with diameters of around 15 um, weak fibres. Stabilisation was initially by cross-linking were first produced commercially in 1982 by Nippon Car- of the polymer by heating in air, just as in the carbon bon. This industrial production was the direct result of fibre production route using PAN precursors. This was research started in the 1970s and carried out by Professor followed by heating in vacuum at temperatures, generally Yajima and his team at the Tohoku University in Japan. around 1200oC and allowed the first generation of small The approach adopted to produce SiC fibres owed much diameter Sic fibres to be produced. The availability of to the experience gained from the development of carbon these Sic fibres brought rapid interest from the aerospace fibres which involves the spinning of polyacrylonitrile and aero-engine industries as they offered the possibil PAN) precursor fibres, which are stabilised by cro ity of producing ceramic fibre reinforced carbon and ce linking and then pyrolysed under controlled conditions ramic matrix composites materials, capable of being used to give carbon fibres. The starting polymer for produc- as structural materials to higher temperatures than those ing SiC fibres, by necessity, needed to contain silicon and attainable with the best nickel based super-alloys. The at- carbon atoms and the polycarbosilane(PCS), which was traction of silicon carbide is that it is a ceramic, which in chosen as the starting material, comprised these elements bulk form, has a Youngs modulus twice that of steel for arranged in a cyclic form consisting of six atoms, which less than half the density and can be used up to 1600c suggested their arrangement in B-SiC. The synthesis of Although oxidised at high temperature, bulk Sic under the PCS used dimethyldichlorosilane( CH3)2 SiCI which goes surface passive oxidation which protects the bulk of was converted into polydimethylsilane [(CH3) Si]n, by the specimen. However it was found that the characteris dechlorination with metal sodium and which in turn was tics of these first generation fibres were not those of bulk converted into a polycarbosilane polymer by heating in SiC. The fibres possessed a Youngs modulus less than an inert atmosphere at 4000C [1]. The chemical compo- half that expected. The fibres crept at 1000C and above sition of PCS can be simplified as -[SiCH3H-CH2In and degraded above 1250%C. The understanding of the 0022-2461 2006 Springer Science+Business Media, Inc DOI:10.1007/10853-0066566-z 82340TH ANNIVERSARY J MATER SCI 4 1 (2 0 0 6 ) 8 2 3 –8 3 9 A review of the development of three generations of small diameter silicon carbide fibres A. R. BUNSELL, A. PIANT Ecole des Mines de Paris, Centre des Materiaux, BP87, 91003 Evry Cedex, France ´ Three generations of small diameter ceramic fibres based on polycrystalline silicon carbide have been developed over a period of thirty years. This has been possible due to studies into the relationships between the microstructures and properties of the fibres. A variety of techniques have been employed by research teams on three continents. The fibres are made by the conversion of polymer precursors to ceramic fibres and all three generations are presently produced commercially. The nature of the precursor and the techniques used for cross-linking have been varied in order to optimise both properties and cost of manufacture. It has been possible to improve the characteristics of the fibres as the processes involved in the cross-linking of the precursor fibres have been better understood and the mechanisms governing both room temperature and high temperature behaviour determined. The result is that, although first generation fibres were limited by a low Young’s modulus at room temperature and by creep and instability of the structure at temperatures far lower than those limiting the behaviour of bulk silicon carbide, the third generation fibres shows many of the characteristics of stoichiometric silicon carbide. This remarkable improvement in characteristics has been due to a thorough understanding of the materials science governing the behaviour of these fibres which are reinforcements for ceramic matrix composite materials.
C 2006 Springer Science + Business Media, Inc. 1. Introduction Silicon carbide fibres, with diameters of around 15 µm, were first produced commercially in 1982 by Nippon Car￾bon. This industrial production was the direct result of research started in the 1970s and carried out by Professor Yajima and his team at the Tohoku University in Japan. The approach adopted to produce SiC fibres owed much to the experience gained from the development of carbon fibres which involves the spinning of polyacrylonitrile (PAN) precursor fibres, which are stabilised by cross￾linking and then pyrolysed under controlled conditions to give carbon fibres. The starting polymer for produc￾ing SiC fibres, by necessity, needed to contain silicon and carbon atoms and the polycarbosilane (PCS), which was chosen as the starting material, comprised these elements arranged in a cyclic form consisting of six atoms, which suggested their arrangement in β-SiC. The synthesis of the PCS used dimethyldichlorosilane (CH3)2SiCl2 which was converted into polydimethylsilane [(CH3)2Si]n, by dechlorination with metal sodium and which in turn was converted into a polycarbosilane polymer by heating in an inert atmosphere at 400◦C [1]. The chemical compo￾sition of PCS can be simplified as –[SiCH3H-CH2]n –. The PCS obtained in this way could be melt spun to give weak fibres. Stabilisation was initially by cross-linking of the polymer by heating in air, just as in the carbon fibre production route using PAN precursors. This was followed by heating in vacuum at temperatures, generally around 1200◦C and allowed the first generation of small￾diameter SiC fibres to be produced. The availability of these SiC fibres brought rapid interest from the aerospace and aero-engine industries as they offered the possibil￾ity of producing ceramic fibre reinforced carbon and ce￾ramic matrix composites materials, capable of being used as structural materials to higher temperatures than those attainable with the best nickel based super-alloys. The at￾traction of silicon carbide is that it is a ceramic, which in bulk form, has a Young’s modulus twice that of steel for less than half the density and can be used up to 1600◦C. Although oxidised at high temperature, bulk SiC under￾goes surface passive oxidation which protects the bulk of the specimen. However it was found that the characteris￾tics of these first generation fibres were not those of bulk SiC. The fibres possessed a Young’s modulus less than half that expected. The fibres crept at 1000◦C and above and degraded above 1250◦C. The understanding of the 0022-2461
C 2006 Springer Science + Business Media, Inc. DOI: 10.1007/s10853-006-6566-z 823