September 1999 Hi-Nicalon/SiC Minicomposites with(Pyrocarbon/SiCn Nanoscale Multilayered Interphases 2471 -(2050)10 12 -(1050)30 -(3/30)10 )10 -(350)10 02 Deformation(%) Fig. 7. Elastic modulus versus applied deformation during the tensile test(E denotes the elastic modulus of the microcracked minicomposite, and Matrin defection z160 with Nicalon fibers Fig 8. SEM image showing the double deflection of a matrix crack. Deformatlon (Sp) (Table II). A finite-element analysis of the thermally induced arison of the tensile- force-deformation curves mea. esidual stresses in minicomposites with various fiber arrange ments shows that, despite a larger thermal expansion mis match in the Nicalon-fiber-reinforced minicomposites, the re sidual stresses are not tremendously larger. Axial stresses that are 100 MPa larger have been computed in the (3)Influence of Initial TowTiwisting ngitudinal matrix cracks may result from the stress state cte is balanced by that of the Young s modulus. However, the that is induced by the initial twisting of the fiber tows. Longi- esidual stresses contribute to the presence of a lower propor- tional limit in the Nicalon-fiber-reinforced minicomposites out tow twisting. Initial tow twisting may influence the stress field that operates on the fibers and the matrix, which leads to (2)I nence of Residual Stresses a multiaxial stress state that involves a radial stress component As mentioned previously, the CTEs of the fiber and the in the internal matrix and interphases, as the bent fibers try to matrix are identical in the Hi-Nicalon-fiber-reinforced mini stretch under tensile loads composites. As a consequence, no significant thermal expat The radial stress component that is induced by the curved sion mismatch should be expected between the fiber and the fibers that try to stretch under a tensile load may increase the matrIx F/M interactions in the interior of the minicomposites. This Finite-element computations for various fiber arrangements, effect is supported by the SEM observations, which show that as well as measurements of matrix-crack-opening displacement the crack-spacing distance is significantly shorter in the inter under tensile loads, confirmed the presence of small residual nal matrix. It may be responsible for the discrepancy that is stresses Axial stresses of -50 MPa in the matrix and +28 MP observed between the T values determined from the hysteresis in the fiber were computed, using the thermoelastic properties loop width and those determined from the crack-spacing dis- given in Table Il (the interphase was assumed to consist of a tance at the surface of the minicomposites Unlike the cracks in single PyC layer, with a thickness of 0.5 um). The radial the surface, the hysteresis loops reflect the F/M interactions in stresses in the matrix were dependent on the fiber-spacing dis- the interior of the minicomposites. Therefore, they provide the tance. A maximum stress of 150 MPa was obtained when the higher T values. However, the hysteresis T values provide prob- fibers are in contact. The contribution of residual stresses can- ably underestimations, because this value is the number of matrix cracks at the surface of the minicomposites that was(Table II). A finite-element analysis of the thermally induced residual stresses in minicomposites with various fiber arrange￾ments21 shows that, despite a larger thermal expansion mis￾match in the Nicalon-fiber-reinforced minicomposites, the re￾sidual stresses are not tremendously larger. Axial residual stresses that are 100 MPa larger have been computed in the Nicalon-fiber-reinforced minicomposites.21 The effect of the CTE is balanced by that of the Young’s modulus. However, the residual stresses contribute to the presence of a lower propor￾tional limit in the Nicalon-fiber-reinforced minicomposites. (2) Influence of Residual Stresses As mentioned previously, the CTEs of the fiber and the matrix are identical in the Hi-Nicalon-fiber-reinforced mini￾composites. As a consequence, no significant thermal expan￾sion mismatch should be expected between the fiber and the matrix. Finite-element computations for various fiber arrangements, as well as measurements of matrix-crack-opening displacement under tensile loads, confirmed the presence of small residual stresses. Axial stresses of −50 MPa in the matrix and +28 MPa in the fiber were computed, using the thermoelastic properties given in Table II (the interphase was assumed to consist of a single PyC layer, with a thickness of 0.5 mm). The radial stresses in the matrix were dependent on the fiber-spacing dis￾tance. A maximum stress of 150 MPa was obtained when the fibers are in contact. The contribution of residual stresses can￾not be regarded as significant. (3) Influence of Initial Tow Twisting Longitudinal matrix cracks may result from the stress state that is induced by the initial twisting of the fiber tows. Longi￾tudinal matrix cracks are not observed on minicomposites with￾out tow twisting. Initial tow twisting may influence the stress field that operates on the fibers and the matrix, which leads to a multiaxial stress state that involves a radial stress component in the internal matrix and interphases, as the bent fibers try to stretch under tensile loads. The radial stress component that is induced by the curved fibers that try to stretch under a tensile load may increase the F/M interactions in the interior of the minicomposites. This effect is supported by the SEM observations, which show that the crack-spacing distance is significantly shorter in the inter￾nal matrix. It may be responsible for the discrepancy that is observed between the t values determined from the hysteresis loop width and those determined from the crack-spacing dis￾tance at the surface of the minicomposites. Unlike the cracks in the surface, the hysteresis loops reflect the F/M interactions in the interior of the minicomposites. Therefore, they provide the higher t values. However, the hysteresis t values provide prob￾ably underestimations, because this value is the number of matrix cracks at the surface of the minicomposites that was Fig. 7. Elastic modulus versus applied deformation during the tensile test (E denotes the elastic modulus of the microcracked minicomposite, and E0 represents the initial elastic modulus). Fig. 8. SEM image showing the double deflection of a matrix crack. Fig. 9. Comparison of the tensile-force–deformation curves mea￾sured on SiC/SiC minicomposites reinforced with Hi-Nicalon or Nicalon fiber. September 1999 Hi-Nicalon/SiC Minicomposites with (Pyrocarbon/SiC)n Nanoscale Multilayered Interphases 2471