PLASTICITY AND DUCTILE FRACTURE OF IF STEELS - EXPERIMENTS AND MICROMECHANICAL MODELING

If one aims at the simulation of plasticity and failure of multiphase materials, the choice of an appropriate material law is of major impor tance. Plasticity models for porous metals contain, in addition to the yield surface and the flow potential, also functions describing the v oid nucleation, dependent on some macroscopically observable quantitie s, and the growth of these voids. In this paper, a micromechanically b ased method to develop a void nucleation function for porous plasticit y models is proposed which is valid for all possible microstructures a s long as the amount of second phase particles is low (i.e, the partic les do not interact with respect to the stress and strain fields), and as long as the particles are large enough (above 0.1 mu m) justifying a continuum mechanical approach. The method described consists of two stages: In the first stage, the microstructure is investigated via a finite element model. The FE model implicitly contains the effects of the shape of the precipitates, of the material parameters of both the matrix and the precipitates, of the void nucleation hypothesis (by the assumption of ''nucleation limits'' for characteristic damage-related quantities), and of the applied stress state. In the second stage, du ring postprocessing, the volume fraction of precipitates as well as th e influences of the particle orientation distribution, size distributi on, and size dependence of the damage-related quantities are taken int o account. The model is applied to the microstructure of IF (Interstit ially Free) steel, a material with a ductile matrix and rigid second p hase particles of cubical shape. This microstructure is particularly s uited for investigating shape and size effects. The model shows that e ither the size effect or the shape effect dominate the void nucleation behavior: in the case of particles of roughly the same size, the size distribution will hardly alter the nucleation strain distribution obt ained by taking into account only the shape and orientation effects. F or particles of very different sizes, the size effect will completely override the rather ''sharp'' original distribution regarding particle shape and orientation. (C) 1998 Elsevier Science Ltd. All rights rese rved.