Theoretical and numerical issues on ductile failure prediction- An overview

Abstract

The main goal of this paper is to give a general overview of some of the recent advances accomplished in the description of ductile damage, both from a theoretical and numerical point of view. To start with, the classical local theory with regard to the thermodynamics of irreversible processes is reviewed where a general elasto-plastic damage model is established. It is also highlighted the assumptions and limitations behind the classical theory when the constitutive equations are obtained from the solution of a constrained maximisation problem. Recent advances on the non-local modelling of ductile damage are also addressed where we shed some light on the principles and consequences of non-locality in elasto-plastic damage models. The issues regarding the efficient numerical implementation of both local and non-local theories are also discussed where special attention is devoted to the implementation of non-local models. In particular, a novel computational strategy, suitable for implementations in commercial programs, is presented for the explicit finite element code LS-DYNA in detail. A FORTRAN code excerpt is given in which the main steps for the implementation of the model are schematically depicted. The effectiveness of the non-local model is assessed through the simulation of an axisymmetric specimen and a sheet metal forming process. It is shown that in both cases the non-local numerical strategy is able to diminish the pathological mesh dependency inherently present in local elasto-plastic damage models.