Recent ASTM Subcommittee E08.08 on Elastic-Plastic Fracture Technology activities have concentrated on the development of a common fracture test standard that allows the evaluation of elastic and elastic-plastic fracture toughness properties from tests on a single specimen. Many technical changes have been made to allow this process to go forward, but one area in which progress has not been made is in clarifying the various specimen size requirements that were part of the earlier standards. Instead, the size requirements in the new toughness testing standard E1820 have been extracted from the earlier standards, like ASTM E399 and ASTM E813 and extended to the new standard without change, in spite of their clear inconsistency.
A review of the development of the E399 size criteria  seems to make it clear that the E399 size criterion was meant by its originators to represent only a requirement that allowed the 0.95 secant offset method to adequately evaluate a consistent measure of crack initiation. The later J-integral-based size criteria of E813, E1152, E1737, E1820, and the recent To Reference Temperature standard E1921 are based on finite element analysis , which ensures that a minimal deviation from small-scale yielding (SSY) occurs. This is intended to generate consistent crack tip constraint as the specimen size decreases or the crack length increases.
In this work, specimens from three materials, a 6061-T651 aluminum, a 6Al-4V titanium alloy, and a 4340 steel alloy, were tested to investigate what the proper size criteria should be for a combined fracture mechanics test standard. Larger specimens from each material were tested to obtain KIc values meeting the stringent size criteria of ASTM E399. Smaller specimens of each material were tested that intentionally failed the E399 size criteria, but satisfied the less stringent criteria of ASTM E813 and E1820. The unloading compliance method has been used in the testing of the smaller specimens so that an estimate of the true crack initiation point on the specimen load displacement record can be measured.
The fracture toughness at the onset of ductile crack extension is measured in terms of the J integral and converted to an equivalent stress intensity KJ value using the standard relationship. The fracture toughness at this crack initiation point can then be compared with the KQ obtained using the 0.95 secant offset initiation point specified by the E399 KIc procedure, and the validity of the various measures of initiation toughness can be determined with respect to the different size criteria.