Preparation of fibre reinforced Si3 N, composites 2.6 24 2.2 Drum Winding Cutting and Curing Firing N2100300cN2,1350°C 18 1.4 Impregnation and Curing Firing Product 12 N2,1350° npregnation and Firing Times Fig. 1. Fabrication process of continuous fibre reinforced Si-N composite Fig. 2. Densification curve of UD composite and shear failure. In order to fracture samples by a As is obvious from Fig. 2, the steps of PCPI have tensile stress, the ratios of span to thickness were theoretically proceeded et to 25-30. In addition, for UD composites, an Mechanical properties of UD composites are interlaminar shear strength(ILSS)test was also shown in Table 3. The stress-deflection curve performed in accordance with JIS K7057 For the shown in Fig 3. The SNF/SNC composite demon 2D samples the direction of the fibre on a surface strates a flexural strength of over I GPa at room layer was aligned to the orientation of the tensile temperature. We have reported the effect of coating stress. a true density was measured with a pycno- on the fibre, 4 in the carbon coated case, 2D (plain meter by using n-decane as the medium. The total weave laminate)silicon carbide fibre reinforced porosity was calculated by the true density for 2D composite has demonstrated a non-brittle failure with fibre pull-out and a high flexural strength of 294 MPa. On the contrary, in the non-coated case, the composite has shown catastrophic failure and a 3 RESULTS AND DISCUSSION relatively low flexural strength of 60 MPa. In this study, the SNF reinforced composites have also 3.1 UD (Unidirectional fibre reinforced composites) shown a non-brittle failure, similar to the one in the carbon coated case. Therefore, this suggests that The densification curves are shown in Fig. 2. the coating on the fibre is essential for an adequate Because of the shrinkage of precursors in firing, interface shear strength in CMCs reimpregnation of polymers is required to densify The auger depth profiles of coated SNF surface the product. The curve for the product agrees well show that a carbon-rich layer, which is con with the theoretical curve calculated from eqn(1). aminated with silicon, is deposited(Fig. 4). In an (The impregnation efficiency is assumed to be 0.9. attempt to perform auger analysis on the fractu Equation(1)is defined as follows: Pc(n)=Pm(1 -Vr)+prVr-pmPe Table 3. The properties of CMcs prepared by the PCPI method Firing condition 1350C: 8 times 1350.C: 7 times Pm)=Pn-1)(1-ICYPp/pm) Fibre SNF SNF SNF SNBF P(0)=(1-H) Matrix precursor SNC SNC where n is the number of firing times, pc, pr, Pm and Total porosity(%) Flexural strength R.T. 649 1049 618 627 Pp are the densities of composites, fibres, matrix (fired polymers)and polymers, respectively, Vr and ILSS (MI P are the volume fraction of fibres and porosities, A(%a) 808595.5945 y is the char yield of polymers, I is the impregna- R: apparent effectiveness =0c/(o,V,A): A is assumed to be tion efficiency and C is the polymer concentration 10UD)and0.5(2D)Preparation of jibre reinforced S&N4 composites 181 Drum Winding Cutting and Curing Firing N2.100-300°C N2,13!X”C L+, + ;: :. ;; : t Polysilazane T’ Impregnation and Curing \ fifing Nz, loo-300°C N2,13WC i83ma f .“zAhr Product Fig. 1. Fabrication process of continuous fibre reinforced Si-N composite. and shear failure. In order to fracture samples by a tensile stress, the ratios of span to thickness were set to 25-30. In addition, for UD composites, an interlaminar shear strength (ILSS) test was also performed in accordance with JIS K7057. For the 2D samples, the direction of the fibre on a surface layer was aligned to the orientation of the tensile stress. A true density was measured with a pycno￾meter by using n-decane as the medium. The total porosity was calculated by the true density for 2D composites. 3 RESULTS AND DISCUSSION 3.1 UD (Unidirectional fibre reinforced composites) The densification curves are shown in Fig. 2. Because of the shrinkage of precursors in firing, reimpregnation of polymers is required to densify the product. The curve for the product agrees well with the theoretical curve calculated from eqn (1). (The impregnation efficiency is assumed to be 0.9.) Equation (1) is defined as follows: Pc(n) = Pnl(l - Vf) + PfVf - PnJ(n-I) (1) with P(n) = P(n-I)(1 - ~CY&hn) P(0) = (1 - Vf) (24 (2b) where n is the number of firing times, pc, pr, pm and pi, are the densities of composites, fibres, matrix (fired polymers) and polymers, respectively, Vr and P are the volume fraction of fibres and porosities, Y is the char yield of polymers, I is the impregna￾tion efficiency and C is the polymer concentration. 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 Polycarbosilant I 1 3 5 7 Impregnation and Firing Times Fig. 2. Densification curve of UD composite. As is obvious from Fig. 2, the steps of PCPI have theoretically proceeded. Mechanical properties of UD composites are shown in Table 3. The stress-deflection curve is shown in Fig. 3. The SNFjSNC composite demon￾strates a flexural strength of over 1 GPa at room temperature. We have reported the effect of coating on the fibre,i4 in the carbon coated case, 2D (plain weave laminate) silicon carbide fibre reinforced composite has demonstrated a non-brittle failure with flbre pull-out and a high flexural strength of 294 MPa. On the contrary, in the non-coated case, the composite has shown catastrophic failure and a relatively low flexural strength of 60 MPa. In this study, the SNF reinforced composites have also shown a non-brittle failure, similar to the one in the carbon coated case. Therefore, this suggests that the coating on the fibre is essential for an adequate interface shear strength in CMCs. The auger depth profiles of coated SNF surface show that a carbon-rich layer, which is con￾taminated with silicon, is deposited (Fig. 4). In an attempt to perform auger analysis on the fracture Table 3. The properties of CMCs prepared by the PCPI method Type UD 2D Firing condition 1350°C : 8 times 1350°C : 7 times Fibre SNF SNF SNF Matrix precursor PHPS SNC SNC Bulk density 2.54 2.35 2.36 Vf @I 43 54 57 Total porosity (%) - 7.2 Flexural strength R.T. 649 1049 618 (MPa) 1250°C 546 ILSS (MPa) 55.1 45.4 - Ra (%) 80 85 95.5 SNBF SNC 2.38 58 5.8 627 595 94.5 “I?: apparent effectiveness = o,/(ofV,A); A is assumed to be 1 .O (UD) and 0.5 (2D)