Orowan
Index
In a paper in Zeitschrift fur Physik 82 (1933) 235, 
E. Orowan described experiments in which mica sheets were tested in tension using samples of different configuration as shown in the diagram. He found that the conventional tensile test samples (a) failed at a tensile stress ten times less than the same material tested in the (b) configuration. This behavior was ascribed to the presence of edge cracks in (a) that were placed in tension, whereas those in (b) were not loaded by the grips A. This was a further extension of the Griffith concept of incipient cracks being responsible for brittle fracture.

In 1949, Orowan applied a "modified Griffith idea" to the brittle fracture of metals that showed some ductility. He found that the plastic deformation work done during crack growth near the crack surface was very much larger than the crack surface energy. If this work replaced the surface energy term, g, in the Griffith expression: 
sc = (2Eg/pc)0.5, the experimental fracture stress was in good agreement with the modified fracture model.
From: Timoshenko,
"History of Strength of Materials,"
McGraw Hill (1953)