Direct microscopic evidence is presented in support of an explanation of R-curve behavior in monophase ceramics by grain-localized bridging across the newly formed crack interface. In situ observations are made of crack growth in tapered cantilever beam and indented flexure specimens of a coarsegrained alumina. The fractures are observed to be highly stable, typical of a material with a strongly increasing resistance characteristic, but are discontinuous at the microstructural level. Associated with this discontinuity is the appearance of overlapping segments in the surface fracture trace around bridging grains; the mean spacing of such "activity sites" along the trace is about 2 to 5 grain diameters. These segments link up with the primary crack beneath the specimen surface, and continue to evolve toward rupture of the bridge as fracture proceeds. The bridges remain active at large distances, of order 100 grain diameters or more, behind the crack tip. Scanning electron microscopy of some of the bridging sites demonstrates that secondary (interface-adjacent) microfracture and frictional tractions are important elements in the bridge separation process. Evidence is sought, but none found, for some of the more popular alternative models of toughening, notably frontal-zone microcracking and cracktiphnternal-stress interaction. It is suggested that the crackinterface bridging mechanism may be a general phenomenon in nontransforming ceramics.

Stable crack growth strongly influences both the fracture strength of brittle rocks and some of the phenomena precursory to catastrophic failure. Quantification of the time and environment dependence of fracture propagation is attempted with the use of a fracture mechanics technique. We examine some of the difficulties encountered when applying techniques originally developed for simple synthetic materials to complex materials like rocks. A picture of subcritical fracture propagation is developed that embraces the essential ingredients of the microstructure, a microcrack process zone, and the different roles that the environment plays. To do this we examine the results of (1) fracture mechanics experiments on five rock types, (2) optical and scanning electron microscopy, (3) studies of microstructural aspects of fracture in ceramics, and (4) exploratory tests examining the time‐dependent response of rock to the application of water.

No abstract

A zone of distributed microcracking is often suggested to accompany tensile macrocrack propagation in rocks and ceramics. The microcracking is said to be largely responsible for (1) high values of fracture energy, (2) increasing resistance to fracture with crack extension and (3) the dependence of fracture mechanics data on the experimental setup. In the present paper, the material breakdown processes in imperfectly elastic Westerly granite are investigated using ultrasonic wave probing and in situ microscopy during mode I fracture experiments. These observations are compared with an in situ reflection/transmission microscopy investigation of mode I fracture in a near‐ideal elastic polycrystalline alumina (Al2O3). As defined by the spatial distribution of longitudinal and surface wave attenuation in wedge‐loaded double‐cantilever beam specimens of Westerly granite, the fracture process zone is elongate in the direction of fracture propagation (15–40 mm long by 1–2 grain dimensions wide; grain size 0.75 mm). As revealed by in situ reflection microscopy, the ultrasonic wave energy is partially transmitted through the developing fracture surfaces via two sources of crack interface traction: (1) remnant islands of unfractured material left behind the advancing fracture front and (2) geometrical interlocking of the microstructurally rough fracture surfaces. A similar zone of crack flank tractions is found in the alumina (greater than 2000 μm long; grain size 20–100 μm). No evidence of a diffuse kidney‐shaped cloud of microcracking distributed ahead of the main fracture tip (predicted by many fracture models) was found in either material. Instead, interface‐localized microcracking was observed to operate at positions where the tractions, or restraining forces, are transmitted across the nascent fracture surfaces. Crack flank tractions shield the main crack tip from high levels of stress and are relieved by friction‐induced microcracking and microcrack rupture of intact‐material bridges. As a consequence of the crack history dependence of the crack tip shielding, it is proposed that, even under small‐scale inelastic deformation conditions appropriate to linear elastic fracture mechanics, neither R‐curve behavior nor applied‐KI/subcritical crack velocity relationships are intrinsic properties of these and similar materials.

Holographic interferometry has been used to observe the surface deformation of Westerly granite during creep experiments at 50 MPa (500 bars) confining pressure. Interference fringes on the holograms produced by this technique constitute contour maps of the sample surface shape change between exposures. Several distinctive characteristics of the surface shape change have been observed. In general, surface deformation during primary and secondary creep, while highly inhomogeneous, is distributed fairly uniformly. No large scale pattern is seen in the surface shape changes. At approximately the onset of tertiary creep, surface displacements become more locallized. Eventually (∼85‐90% of time‐to‐failure) a distinct ridge is formed which exactly coincides with the location of the final fracture.