Phenomenon of fracture and fracture morphology of glass tested in tension flexure and torsion

A method of testing glass rods in tension and torsion was devised. Two types of glass, Pyrex and Soda-lime, were tested in tension, flexure and torsion and the data analyzed by both Normal and Weibull statistics. A significant difference in strength was found to exist between the tension and flexural techniques of testing of similar specimens of Soda-lime glass. This difference is explained by the size effect brought about by the pre-existing flaw concept. A significant difference was also found between similar specimens of Pyrex and Soda-lime when both were tested in tension. This is explained by a difference in intrinsic strengths of the two glasses. Explanations are put forward concerning the formation of certain fracture markings and light and electron micrographs are presented to indicate the various markings that may be observed on glass fracture surfaces. The fact that a Griffith type equation can be used to predict the nominal stress at fracture, with the mirror being considered as the flaw, is explained, using the concepts of Irwin, by suggesting that in any one glass there is a specific strain energy release rate, Gm, associated with the mirror boundary.U.S. Naval Research Laboratory, Contract No. Nonr 2947 (02) (x), NRL Project 62 R05 19A, Technical Memo No. 19