Three-dimensional numerical modeling of ductile fracture mechanisms at the microscale
Résumé
The present work aims at a better understanding and modeling of ductile fracture during the forming of metallic materials. These materials are formed using series of thermo-mechanical loads where many parameters such as loading type and direction vary, and predictive numerical tools are necessary to understand the occurrence of failure and optimize production costs. Ductile fracture in metallic materials is the result of voids nucleation, growth, and coa-lescence mechanisms, leading to a progressive loss of load carrying capacity, until failure [1]. In this work, a micromechanical approach is developed in order to conduct realistic Finite Element (FE) simulations of ductile fracture at the microscale, accounting explicitly for the microstructure of the studied materials. Such simulations require a robust methodology to discretize the microstructure, and then update this discretization during its deformation and progressive failure. Level-Set (LS) functions are used to represent all interfaces [2], and ease the modeling of void linkage events, that are difficult to handle with standard FE methods. This implicit representation of interfaces raises multiple issues. First, discontinuous material behavior at interfaces remains to be modeled. Second, large plastic strains can only be reached using several remeshing operations, which induce an important diffusion of the volume and morphology of the microstructure when the latter is carried only by LS functions. A mesh generation method is developed to mesh interfaces with a conforming FE discretization, hence explicitly representing discontinuities in material behavior. The quality and accuracy of this discretization is maintained during deformation thanks to a mesh adaption algorithm enhanced with a volume conservation constraint [3]. Additionally, real microstructures typically feature a complex morphology that requires a very fine discretization. Using a local error estimator based on LS functions , the discretization is automatically refined where necessary [4]. This error estimator relies on the geometric properties of LS functions, which are maintained during deformation thanks to a new direct LS reinitialization technique [5]. Finally, a computational
Domaines
Matériaux
Origine : Fichiers produits par l'(les) auteur(s)
Loading...