A failure criterion to explain the test specimen thickness effect on fracture toughness in the transition temperature region

This paper considered the test specimen thickness (TST) effect on the fracture toughness of a material Jc in the transition temperature region for 3 point bending (3 PB) specimens. Fracture toughness tests and elastic–plastic finite element analyses (FEA) with non-standard test specimens, which are non-standard because the specimen thickness-to-width ratio B/W was varied in the range of 0.25–1.5, were conducted. Based on these tests and the FEA results, it was demonstrated that the “planar” (4δt, σ22c) failure criterion—which states that cleavage fracture after significant plastic deformation occurs when the crack opening stress σ22 at a distance from the crack-tip that is equal to four times the crack-tip opening displacement δt exceeds a critical value σ22c—was verified to effectively explain the TST effect. This (4δt, σ22c) criterion also successfully predicted the tendency of Jc to saturate to some bounding value for B/W = 1.0. This tendency was similar to that of the T33-stress, which is the out-of-plane elastic crack-tip constraint parameter. Because the (4δt, σ22c) criterion could predict the TST effect on Jc and because the criterion could predict the bounded behavior of Jc for large B/W, the TST effect was concluded to be mainly mechanical in nature, which the weakest link model fails to predict. The mechanical cause of the TST effect on Jc was considered to be an out-of-plane crack-tip constraint, and one of its measures of magnitude is the T33-stress.