Elastoplastic constitutive modeling of metallic materials at finite strains: FEM numerical responses from different constitutive models

This work reports the main results of a campaign of analyses on available finite element codes that was carried-out at the University of Bergamo and TenarisDalmine R&D in 2010/2011. The aim of this work is to enquire the potentialities of main FEM codes in the context of large strain constitutive modeling, in view of assessing their capabilities, by checking both positive and negative aspects. The present work addresses only quasi-static large strain elastoplastic constitutive models, without considering dynamic behavior, temperature effects, damage and fracture. However, the obtained results appear to be of relevance also for the further contexts above, by assessing core basic features, as investigated and revealed here. The considered FEM codes are the following: Dassault Systèmes, ABAQUS 6.10, both Implicit and Explicit Packages; MSC Software, MARC 2008; COMSOL Group, COMSOL Multyphysics 3.5; LSTC, LSDyna 9.71 Revision 59419, both Implicit and Explicit Packages; ANSYS, ANSYS 13.0 both Implicit and Explicit Packages; De Souza Neto et al. [1], HYPLAS research code. The achieved results clearly show how the present FEM tools may lead to simulations of great usefulness but also to outcomes that appear completely incoherent and untrustworthy, which may even lead potentially to dangerous structural consequences. Severe oscillatory incoherencies have been found for much of the computed output. Errors derived from their presence clearly lead to completely incoherent results. In order to avoid such problems, it may be recommended to follow the indications that are pointed-out in this document and briefly resumed in the Conclusions. In fact, this report may act as a guide, useful to avoid output incoherencies possibly coming from one of the FEM codes that have been analyzed here.