A VIEW ON DUCTILE-FRACTURE MODELING

Theoretical models of ductile fracture are reviewed in terms of experi mental results from metallurgical studies of ductile fracture in metal s and alloys. It is shown that the plastic limit-load model, which is based on a criterion of void coalescence by internal microscopic necki ng of the intervoid matrix, is fully consistent with scanning electron microscope (SEM) observations of both the ductile-fracture surface an d the microstructure immediately adjacent to the fracture surface. On the other hand, the dilational-plastic models of ductile fracture, whi ch are based on the dilational-growth of spherical voids to some arbit rary critical void-volume fraction, are inconsistent with the microstr uctural observations of ductile fracture. This inconsistency between t he dilational-plastic models and experimental results is shown to be t he combined effect of neglecting the controlling influence of extensio nal void-growth and the failure to incorporate a physically realistic criterion of void coalescence. The problems of modelling the ductile c rack-growth process by both analytical and numerical (finite element) studies, where problems of uniqueness of the plastic velocity field ma y occur, are also considered. The limitations of the finite-element me thod in modelling void-coalescence problems, where the equations of pl asticity are of second-order hyperbolic form, are also discussed.