Robust identification of an augmented Gurson Model for elasto-plastic porous media

Abstract

In the paper we investigate the robust identification approach to identify the material parameters in the augmented Gurson model for the elasto-plastic porous media. We consider the robust loss function given by Huber [9], Beaton and Tuckey [38] and the loss function based on the l₁-norm. The resulting minimization problem is solved by means of our own implementation of the Boender et al. global minimization method. Our aim is to compare the results with our earlier standard least squares estimates. In the paper, the effects of nucleation and growth of voids in the plastic porous media are investigated. Three different forms of the model are considered: the augmented Gurson model (total porosity model) with variable nucleation and growth material function, the same model with constant growth material function and the separated porosity model. The identification of the material functions parameters is based on Fischer's experimental data set for axisymmetric tension of steel specimens.