Constraint and Statistical Adjustment Models Applied to Transition Fracture Toughness Data

Abstract

Constraint and statistical models were applied to fracture toughness data in the transition for steels to determine to what extent either or both are required to deal with size and geometry effects in the transition. A Weibull model was used for statistical adjustment to try to remove size effects from the data. The small-scale yield model was used as a constraint adjustment model to try to do the same. Results showed that no single adjustment worked in every case. To remove size effects a Weibull statistical adjustment taken alone worked best. However, statistical adjustment has not been used to remove geometry effects. To insure accurate transferability of fracture toughness data from one geometry to the next a constraint based model is needed. For this the model based on the J-Q two parameter crack tip stress field characterization was used. Fracture toughness measured on a compact specimen was used to predict the fracture toughness of center cracked and double-edge cracked tension geometries. These predictions were compared with actual data. The results showed that the model was able to predict the fracture behavior of the double-edge cracked geometry but was nonconservative for the center-cracked geometry. These results show that both statistical and constraint models are necessary for dealing with transition fracture toughness data.