CERAMIC COMPOSITE INTERFACES 509 fiber pullout test), does allow for tensile fiber loading analogous to actual com- posite fracture(versus the fiber push-in test), and affords a continuous measure of force-displacement measurements(compared with those from optical im- ages). The test provides a means by which all of the pertinent interfacial properties can be discerned, i.e. the interfacial toughness, the coefficient of friction, and the roughness amplitude, by coupling the total force-displacement curve with the appropriate analysis of Equation 1. Marshall et al (38) conducted similar tests by etching away a portion of the matrix and measured displacement using a high resolution matrix mapping technique based upon optical methods Microcomposite Tests Large-diameter fibers offer the flexibility of single-fiber tests described above However, for smaller-diameter fibers or those fibers serving as reinforcements in matrices demonstrating densities significantly less than theoretical densities the pullout and push-in tests are difficult, if not impossible, to perform. Lamon et al (39) have proposed a microcomposite test for such systems. The test con- sists of a concentric cylindrical element containing a fiber, coating, and matrix and is fabricated under identical conditions to the actual composite. The mi- rocomposites are tested in uniaxial tension and do not require precracking of the matrix. The load and axial displacement measurements are accompanied by acoustic emission monitoring for evidence of matrix cracking. Once matrix cracking commences, the specimen is unloaded and reloaded, and the hyster- sis loop associated with the loading and unloading is monitored. Similar to the Hutchinson Jensen(5)and Marshall(6) analyses described for the pullout problems above, the frictional sliding resistance t, the misfit strain e, and the interface debond energy Gic can be obtained through a parabolic curve-fitting routine as shown in Reference 40. Morscher et al (4 1)used the technique for de- termining interfacial properties of large-diameter fibers (140 um SiC monofil aments )in chemical vapor-deposited(CVD)SiC matrices. These results could be conveniently compared with pullout and push-out experiments, where rea- sonable agreement was seen. Attempts to use the test at elevated temperature were less successful because matrix crack saturation was not reached (42) Because of the difficulty in applying the microcomposite test to high tem- peratures, as well as applying cyclic loading to such configurations, Evans and co-workers (43-45)developed a methodology for extracting interface prop- erties from composite specimens, and from them, predicting the stress-strain response of the materials. Such constituent properties, when coupled with the appropriate analysis, can be used to examine the consequence of strain con- centrations on the composite. Specifically, the composite is pulled in tension, unloaded, and reloaded subsequent to saturation of matrix cracks to achieve t L pue <orP1: ARK/MBL/rkc P2: MBL/vks QC: MBL/agr T1: MBL May 16, 1997 13:47 Annual Reviews AR034-16 CERAMIC COMPOSITE INTERFACES 509 fiber pullout test), does allow for tensile fiber loading analogous to actual com￾posite fracture (versus the fiber push-in test), and affords a continuous measure of force-displacement measurements (compared with those from optical im￾ages). The test provides a means by which all of the pertinent interfacial properties can be discerned, i.e. the interfacial toughness, the coefficient of friction, and the roughness amplitude, by coupling the total force-displacement curve with the appropriate analysis of Equation 1. Marshall et al (38) conducted similar tests by etching away a portion of the matrix and measured displacement using a high resolution matrix mapping technique based upon optical methods. Microcomposite Tests Large-diameter fibers offer the flexibility of single-fiber tests described above. However, for smaller-diameter fibers or those fibers serving as reinforcements in matrices demonstrating densities significantly less than theoretical densities, the pullout and push-in tests are difficult, if not impossible, to perform. Lamon et al (39) have proposed a microcomposite test for such systems. The test con￾sists of a concentric cylindrical element containing a fiber, coating, and matrix and is fabricated under identical conditions to the actual composite. The mi￾crocomposites are tested in uniaxial tension and do not require precracking of the matrix. The load and axial displacement measurements are accompanied by acoustic emission monitoring for evidence of matrix cracking. Once matrix cracking commences, the specimen is unloaded and reloaded, and the hystere￾sis loop associated with the loading and unloading is monitored. Similar to the Hutchinson & Jensen (5) and Marshall (6) analyses described for the pullout problems above, the frictional sliding resistance τ , the misfit strain εT, and the interface debond energy Gic can be obtained through a parabolic curve-fitting routine as shown in Reference 40. Morscher et al (41) used the technique for de￾termining interfacial properties of large-diameter fibers (140 µm SiC monofil￾aments) in chemical vapor–deposited (CVD) SiC matrices. These results could be conveniently compared with pullout and push-out experiments, where rea￾sonable agreement was seen. Attempts to use the test at elevated temperature were less successful because matrix crack saturation was not reached (42). Because of the difficulty in applying the microcomposite test to high tem￾peratures, as well as applying cyclic loading to such configurations, Evans and co-workers (43–45) developed a methodology for extracting interface prop￾erties from composite specimens, and from them, predicting the stress-strain response of the materials. Such constituent properties, when coupled with the appropriate analysis, can be used to examine the consequence of strain con￾centrations on the composite. Specifically, the composite is pulled in tension, unloaded, and reloaded subsequent to saturation of matrix cracks to achieve τ , Gic, and σ T the misfit stress