Aluminum die casting components are widely used in vehicle constructions because they satisfy the conflicting requirements between weight reduction and mechanical property improvement. However, the analysis of deformation and damage behavior of aluminum cast components is very complex, since local mechanical properties in the components are inhomogeneous as a consequence of spatial distribution of microstructure. For crash simulation it is necessary to well predict the damage behavior which is strongly influenced by micro-defects especially by cast pores. The conventional continuum mechanics approaches often fail due to the statistical character of cast pores. In this work the Markov random field model (Ising) is used to describe the pore morphology. Markov random field classes are defined by porosity (macroscopic property) and equivalent pore size (microscopic property) and determined by micro computer tomography (CT) analysis.A multi scale approach was applied to map the results of the stochastic model to the FE models, which results in a distribution of porosity. A porosity dependent continuum model was developed based on results of representative volume elements with variation of porosity. It was shown that the continuum model with porosity distributions from the Ising model as initial conditions captures well the spatial material properties (i.e. fracture strain) and their variations in the bridging scale.
A Multi-Scale Lip Segmentation Method Based on Markov Random Field
