SOME MICROSTRUCTURAL CONDITIONS FOR EVALUATING FRACTURE-TOUGHNESS AND R-CURVE BEHAVIOR IN PLATELET-REINFORCED COMPOSITES

The validity of K(Ic) determinations by various techniques usually employed and compared with each other is discussed in detail for two ceramic materials. They are a simple Si3N4 material fracturing with transgranular crack propagation and a Si3N4-25vol% SiC-platelet measured by four different methods: single-edge V-notched beam (SEVNB), single-edge precracked beam (SEPB), indentation fracture (IF), and chevron-notched beam (CNB). It is generally recognized that, in platelet composites, K(Ic) values can be reliably obtained only by procedures adopting notched specimens in which the wake-zone of precracking is mechanically removed (SEVNB or renotched SEPB). In the present investigation the CNB method, in which the wake contribution to fracture resistance is generally small but not negligible, was also found to give stress-intensity factors at maximum load whose value was almost coincident with the K(Ic) determined by SEVNB method. Using stereological concepts, fractographic observations, and acoustic emission (AE) experiments, the microstructural conditions for negligible wake-contribution during the subcritical stable crack extension in the CNB experiment were worked out and applied to rationalize the behavior of the present composite. Apparent toughness values markedly higher than the ''true'' K(Ic) of the composite were measured both by the standard SEPB and IF method. Such a discrepancy is considered due, in the former case, mainly to traction forces between the precracked surfaces while, in the latter, to an insufficient account of the residual stress field associated with the indentation.