Abstract
Historically, size criterions have been developed with the aim of guaranteeing specimen size independent fracture toughness values. The basis for the criterions have been different assumptions regarding the specimen constraint. Presently, it has become evident that for cleavage type fracture there exists a statistical size effect affecting the measured toughness in addition to constraint. In this work existing data bases are analyzed, considering the statistical size effect, but neglecting any corrections for constraint. It is shown that the validity of small bend specimen fracture toughness estimates is excellent, even with specimens that clearly violate even the least stringent size requirements.