Influence of stress intensity and crack speed on fracture surface topography: mirror to mist to macroscopic bifurcation

Abstract

The development of roughness on the fracture surfaces of a brittle, glassy, epoxy resin from the mirror-to-mist transition to macroscopic bifurcation has been investigated using optical microscopy, scanning electron microscopy (SEM) and contact and non-contact laser profilometry. Most of the observations were made on specimens fractured in edge-notched tension. In a series of tests the initial crack length was varied to obtain fracture surfaces formed by accelerating and decelerating cracks without macroscopic bifurcation (specimen A) and by cracks which accelerated continuously to macroscopic bifurcation (specimen B). Some observations were made on specimens tested in compact tension to study changes in fracture surface topography associated with crack arrest in stick-slip fracture. There was a close correlation between the topographical detail revealed by the different techniques. In specimen A the roughness increased progressively from the mirror-to-mist transition and reached a maximum before decreasing as the crack decelerated. The topographical features revealed by optical microscopy and SEM were the same for accelerating and decelerating cracks at the same roughness value. In specimen B the roughness increased continuously to macroscopic bifurcation. There was a close similarity between the topographical features at all levels of roughness. A simple model for the basic step involved in roughness formation is presented which involves an element of the crack tip tilting out of the plane of the main crack before stopping (micro-bifurcation). The scale of micro-bifurcation ranged from 3 μm in the early stages of mist, when the crack velocity was close to 10% of the shear wave velocity, to the full width of the specimen (6 mm) at macroscopic bifurcation. The micro-bifurcation process develops from crack surface undulations and does not involve micro-crack nucleating ahead of the main crack. It is concluded that the relationships between crack velocity and dynamic stress intensity, and the value of the limiting crack velocity, must be interpreted in terms of micro-mechanical processes at the crack tip which are strongly dependent on specific material characteristics.