CRACK ORIGINS AND MICROMECHANISMS OF SLOW CRACK GROWTH IN THE DELAYED FRACTURE OF ALUMINA

Direct observation of the early stages of crack extension from inherent flaws is made in order to identify flaws responsible for crack initiation and to study micromechanisms of slow crack growth in the delayed fracture of alumina. Although difficulties are experienced using optical and scanning electron microscopy even with special techniques, a microscopic examination technique is developed to locate the onset of crack extension from inherent flaws using fluorescent dye penetrant and scanning electron microscopy. The specimen surface is immersed in the dye while cracks are forming which allows the dye to penetrate into the fine cracks. Once the cracking sites are located, high magnification scanning electron microscopy is used for detailed identification of the nature and characteristics of the flaws.; An eccentrically loaded column testing system is developed to produce a number of crack initiations on a surface of a specimen. With this mechanical testing technique, some cracks may propagate dynamically while many cracks are initiating without causing catastrophic failure of the specimen. The testing technique has another significance, that of being able to compare delayed fracture phenomena between tension and compression, since the ratio of tensile and compressive stresses can be precisely controlled.; By combining the microscopic examination technique and the mechanical testing technique, crack origins are identified and micromechanisms of slow crack growth in delayed fracture are studied. A surface of a ceramic contained a large number of inherent flaws which, either by themselves or as assemblies, become potential crack origins. Cracking sites are found to be multiple in delayed fracture. The initial stage of delayed fracture is shown to involve the interaction and coalescence of near-by flaws with extensive intergranular cracking. A model for the delayed fracture, the microscopic slow crack growth law and the crack interaction and coalescence, is proposed as opposed to the conventional single-worst-crack approach. The proposed slow crack growth law gives much better prediction compared with the conventional approach. . . . (Author's abstract exceeds stipulated maximum length. Discontinued here with permission of school.) UMI...