Accompanying dataset for the paper "An implicit staggered algorithm for CPFEM-based analysis of aluminum"

Contributions

Pedro Areias did contribute to general programming and theory

Charles dos Santos did contribute to investigation, specific programming and validation

Rui Melicio did contribute to the text and typesetting

Nuno Silvestre did contribute to the Scientific validation

Funding sources

FCT Fundação para a Ciência e a Tecnologia project LAETA Base Funding (DOI: 10.54499/UIDB/50022/2020)

Data structure and information

data - dataset directory

convertFig2Eps - script to convert xfig sources to EPS

driftfigures - data to assess drifting in \xi and effective strain

cuboid.gid - directory for a single cube analysis, to assess drifting

epcfigures - these are the sources for the cylinder test with localization

cylindersinglecrystalcoarse.gid - cylinder with ( $\theta=0.25\pi,\phi=0$)

cylindersinglecrystalcoarseotherangles.gid - cylinder with ($\theta=0.304\pi,\phi=0.25\pi$")

errorlogstrain - contains the mathematica sheet for the plots in the logstrain error analysis

originalfigures - these are the original figures in the paper

padeerrorgraf - mathematica sheet for the analysis of padé approximation error

reactions - reactions sources for the localization problem

workflows - reproducibility of some of the computational results

Dataset Description

Contains:

Data sources from SimPlas (txt and order)

Gnuplot files (gp)

Tikz files (tikz)

XFig files (fig)

Mathematica scripts (nb)

Script to convert Xfig in Eps: figtex2eps.sh

Original figures are also included.

Paper Description

In this paper, we propose an implicit staggered algorithm for crystal plasticity finite element method (CPFEM) which makes use of dynamic relaxation at the constitutive integration level. An uncoupled version of the constitutive system consists of a multi-surface flow law complemented by an evolution law for the hardening variables. Since a saturation law is adopted for hardening, a sequence of nonlinear iteration followed by a linear system is feasible. To tie the constitutive unknowns, the dynamic relaxation method is adopted. A Green-Nagdhi plasticity model is adopted based on the Hencky strain calculated using a [ 2/2 ] Padé approximation. For the incompressible case, the approximation error is calculated exactly. A enhanced-assumed strain (EAS) element technology is adopted, which was found to be especially suited to localization problems such as the ones resulting from crystal plasticity plane slipping. Analysis of the results shows significant reduction of drift and well defined localization without spurious modes or hourglassing.