Modeling of porous metal deformation by SPH method and construction of constitutive equations using artificial neural networks

This study aims to develop a macroscopic continuum model of dynamic deformation of porous metals based on the application of artificial neural networks. Data sets obtained by modeling the compression of representative volumes of porous media by smoothed particle hydrodynamics based on a dislocation plasticity model previously parameterized for copper are used to train the ANN. Such modeling is used both for training data sets and to investigate the deformation physics of porous copper with micrometer and millimeter scale pores.