Global and Local Approaches of Fracture — Transferability of Laboratory Test Results to Components

This paper is devoted to the application of local micromechanisms of failure to the prediction of the macroscopic fracture toughness properties of metallic materials — a field pioneered by F. McClintock. The global approach of fracture assumes that failure can be described in terms of a single parameter, such as K IC or J IC . This approach may yet prove to be questionable in complex situations. This is the reason why local approaches of fracture have also to be developed. An attempt is made here to review the application of these local approaches, especially those which are micro-mechanistically based and to indicate a number of research fields which necessitate further studies. The paper is divided into two main parts. In the first part, the three basic fracture modes encountered in metallic materials, i.e. cleavage, intergranular fracture, and ductile rupture are reviewed. For cleavage fracture, a statistical model based on the Weibull weakest link concept is introduced and applied to a number of low-alloy ferritic steels. Some questions relating to intergranular fracture are also discussed. For ductile rupture, the emphasis is laid on the discontinuous or continuous character of the nucleation of cavities from second phase particles, on cavity growth and coalescence. The mechanics of plastic porous materials is briefly introduced to model this mode of failure and the statistical aspects of ductile rupture. It is shown that, in spite of the large research effort devoted to ductile rupture over the past few decades, it is still necessary to use an empirical fracture criterion based on the concept of critical void growth, that was originally introduced by McClintock. The second part of the paper is devoted to the application of these local fracture criteria to predict the fracture toughness of specimens or the fracture load of components. The concept of characteristic distances related to the microstructure of materials and that of the “process zone” are briefly discussed. Then theoretical expressions between the fracture toughness (K IC or J IC ) for 2D cracks, tested under small-scale yielding conditions, and local criteria are introduced. The local criterion for brittle cleavage fracture is based on Weibull statistics, which gives rise to a size effect, while the criterion for ductile rupture is based on critical void growth. Finite element method (FEM) numerical simulations of compact tension (CT) and center-cracked panel (CCP) specimens under large-scale yielding conditions were used in conjunction with these local fracture criteria to show that the ligament size requirements for “valid” J IC measurements are not only dependent on the specimen geometry (crack length, tension versus bending), but also on material work-hardening exponent and, more importantly, on the microscopic modes of failure. Further applications of the local approach of fracture are also presented, including fracture toughness testing of 3D cracks, the ductile-to-brittle transition behavior for ferritic steels, and fracture tests of specimens and components under non-isothermal conditions. Finally, further developments are briefly discussed.