Abstract
This study examines the effect of artificially introduced spherical voids on the bend strength of a polycrystPalline brittle solid and compares the results to traditional empirical and phenomenological porosity-strength relationships. No correlation is found between these relationships and the data. The statistical approach by Weibull is examined particularly with regard to the implications to microstructural features. A direct test of Weibull statistics showed deviations from theory, perhaps due to the very small stressed volumes employed. Microstructural evidence suggests that some of the basic assumptions in Weibull's formulation may restrict its use in this and perhaps other systems. Based on the similarity between McClintock's statistical model and the observed microstructural features a failure criterion for systems with randomly distributed flaws is proposed. 25 references.

This publication has 0 references indexed in Scilit: