ovember 2006 Oxide Fiber Composites 3319 Sidebar B. Fiber Push-In Testing of porous Matrix CFCCs he interfacial debonding and sliding properties are most conveniently probed by fiber push-in testing. The test is performe using an instrumented indenter, usually with a sharp diamond tip (Berkovich or Vickers), resulting in plastic deformation of the with fiber yielding must be measured on a reference(non-sliding) fiber. Typically, the reference state is produced by fabricating a composite without a fiber coating and subjecting it to a heat treatment that produces strong interfacial bonds As porous-matrix CFCCs do not rely on fiber coatings, the al technique for producing a reference fiber is impractical Furthermore because of its low stiffness. the matrix surrounding the pushed fiber undergoes an indeterminate amount of elastic displacement during fiber push-in concurrent with plastic displacement of the fiber sur ace a sliding along the interface. Con tly, if the reference fibers were embedded in a dense matrix. the test results would capture all of the extraneous displacement light of these problems, a variant on the established push-in technique has been developed for use with porous ye2卜 matrix CFCCs. Two changes have been implemented: (i)a blunt(spheroconical) indenter that produces only elastic 10 deformation is used for fiber pushing; and (i) the hysteresis loop width is used for analysis, rather than the absolute displacement. The loop width is obtained by subtracting the measured displacements on loading and unloading at each load level; as all extraneous displacement is elastic, it does not 050100150200250300350400 contribute to the displacement difference. Displacement(nm) A full analysis of the hysteresis loops is presented in W et al. The pertinent solutions are summarized in Table BI Representative measurements for a system with a particulate lullite-alumina matrix are in Fig. Bl Table bl. Solutions for Fiber -Push-in Test Definitions Er--fiber modulus ATRIE Fc-critical force for debond initiation T=0 J/m2 t FM-maximum force R--fiber radius k≡ T-debond energy Fc=2x√RE Cycle 2 oop Width Cycle I .=k(1-k) Normalized Force, F/F (0≤F≤FC) (0S FS FM) Fig B1.(a) Force-displacement curves from a fiber push-in test on a porous mullite-alumina matrix continuous-fiber ceramic composite The inset shows the test configuration.(b)Loop width measurements (0≤F≤Fc) along with analysis to ascertain debond energy and sliding stress the maximum force in the test) VI. Thermomechanical Stability of Porous-Matrix CFCCs The alumina-silica system undergoes the most rapid degrada- ion, starting at temperatures below 1000.C. In addition to the atings is vital to composite durability. A useful metric for as- no pullout. These changes are correlated with coarsening of sessment of stability is the strength retention following extended natrix porosity and extensive sintering to the fiber surface. high-temperature exposure. The results of this type for three Analogous changes are obtained at elevated temperatures and families of porous-matnix CFCCs following 1000-h exposures in notched specimens. The retained strength of the system are plotted in Fig. 18(a). All are reinforced with Nextel"720 with the mullite-alumina matrix remains essentially unchanged fibers in an eight-harness satin weave. They are distinguished upto 1200oC. Limited sintering is evidenced by elevations from one another by matrix composition and topology: () par stiffness, especially beyond 1000oC, as well as observations of ticulate alumina bonded by nanoporous silica gl ( i slightly reduced fiber pullout and larger amounts of remnant 80%mullite-20% alumina particle mixture, strengthened by matrix on the fiber surfaces. The all-mullite matrix system ex- about 4% of a precursor-derived alumina" and (iii)bimodal hibits the best performance. Its retained strength remains stable particulate mullite, with particle sizes of about I and 0. 1 um. 4 up to 1200 C (actually increasing slightly). The stiffness increas- Some additional insights into the extent of sintering are obtained es over this temperature range but not as rapidly as that of the from changes in Young,'s modulus, plotted in Fig. 18(b). The mullite-alumina matrix composite Beyond 1200C, the stiffness increases at a greater rate and the strength begins to diminish.VI. Thermomechanical Stability of Porous-Matrix CFCCs The stability of the pore structure in a CFCC without fiber coatings is vital to composite durability. A useful metric for as￾sessment of stability is the strength retention following extended high-temperature exposure. The results of this type for three families of porous-matrix CFCCs following 1000-h exposures are plotted in Fig. 18(a). All are reinforced with Nextelt 720 fibers in an eight-harness satin weave. They are distinguished from one another by matrix composition and topology: (i) par￾ticulate alumina bonded by nanoporous silica glass52; (ii) an 80% mullite–20% alumina particle mixture, strengthened by about 4% of a precursor-derived alumina12; and (iii) bimodal particulate mullite, with particle sizes of about 1 and 0.1 mm.43 Some additional insights into the extent of sintering are obtained from changes in Young’s modulus, plotted in Fig. 18(b). The notable trends follow. The alumina–silica system undergoes the most rapid degrada￾tion, starting at temperatures below 10001C. In addition to the strength loss, fracture occurs in a brittle manner with virtually no pullout. These changes are correlated with coarsening of matrix porosity and extensive sintering to the fiber surface. Analogous changes are obtained at elevated temperatures and in notched specimens.53,54 The retained strength of the system with the mullite–alumina matrix remains essentially unchanged upto 12001C. Limited sintering is evidenced by elevations in stiffness, especially beyond 10001C, as well as observations of slightly reduced fiber pullout and larger amounts of remnant matrix on the fiber surfaces.12 The all-mullite matrix system ex￾hibits the best performance. Its retained strength remains stable up to 12001C (actually increasing slightly). The stiffness increas￾es over this temperature range, but not as rapidly as that of the mullite–alumina matrix composite. Beyond 12001C, the stiffness increases at a greater rate and the strength begins to diminish. Sidebar B. Fiber Push-In Testing of Porous Matrix CFCCs The interfacial debonding and sliding properties are most conveniently probed by fiber push-in testing.50 The test is performed using an instrumented indenter, usually with a sharp diamond tip (Berkovich or Vickers), resulting in plastic deformation of the fiber beneath the tip. To ascertain the sliding displacement from the measured (total) value, the plastic displacement associated with fiber yielding must be measured on a reference (non-sliding) fiber. Typically, the reference state is produced by fabricating a composite without a fiber coating and subjecting it to a heat treatment that produces strong interfacial bonds. As porous-matrix CFCCs do not rely on fiber coatings, the usual technique for producing a reference fiber is impractical. Furthermore, because of its low stiffness, the matrix surrounding the pushed fiber undergoes an indeterminate amount of elastic displacement during fiber push-in, concurrent with plastic displacement of the fiber surface and sliding along the interface. Consequently, if the reference fibers were embedded in a dense matrix, the test results would not capture all of the extraneous displacement. In light of these problems, a variant on the established push-in technique has been developed for use with porous matrix CFCCs.51 Two changes have been implemented: (i) a blunt (spheroconical) indenter that produces only elastic deformation is used for fiber pushing; and (ii) the hysteresis loop width is used for analysis, rather than the absolute displacement. The loop width is obtained by subtracting the measured displacements on loading and unloading at each load level; as all extraneous displacement is elastic, it does not contribute to the displacement difference. A full analysis of the hysteresis loops is presented in Weaver et al. 51 The pertinent solutions are summarized in Table BI. Representative measurements for a system with a particulate mullite–alumina matrix are in Fig. B1. Fig. B1. (a) Force–displacement curves from a fiber push-in test on a porous mullite–alumina matrix continuous-fiber ceramic composite.51 The inset shows the test configuration. (b) Loop width measurements along with analysis to ascertain debond energy and sliding stress. (FM is the maximum force in the test). Table BI. Solutions for Fiber-Push-in Test Definitions Ef—fiber modulus FC—critical force for debond initiation FM—maximum force R—fiber radius t—sliding stress G—debond energy D  F2 M 4p2R3tEf g  4p2GR3Ef F2 M k  F FM FC ¼ 2p ffiffiffiffiffiffiffiffiffiffiffiffiffiffi R3EfG p Loop Width Cycle 1 Cycle 2 D1 D ¼ ð1 gÞ 1 1 2 ð1 kÞ 2  ð0  F  FCÞ Dn D ¼ kð1 kÞ ð0  F  FMÞ D1 D ¼ 1 2 þ k 3k2 2 ð0  F  FCÞ November 2006 Oxide Fiber Composites 3319