Microscale FFT-based modeling of artificial porous microstructures and analysis of macroscopic models under varying loading scenarios

Abstract

This article investigates the modeling of metallic isotropic porous materials by evaluating and comparing macroscopic and microstructure-resolved simulation approaches. At the macroscopic level, different variants of the Deshpande-Fleck (DF) model are implemented to describe the elastoplastic behavior of metallic porous materials. Considering the porous microstructure explicitly, FFT-based simulations are performed to compute homogenized responses of synthetically generated microstructures under various loading conditions. Assessing representativeness and morphological variation, different realizations of artificial closed-cell porous microstructures with different relative densities are analyzed. The study systematically compares macroscopic DF model predictions with the results of FFT-based homogenization to assess DF model accuracy and identify limitations in different loading scenarios in a wide range of porosities. The results demonstrate that the original DF model captures key trends in the mechanical response, but needs repeated calibration for varied load paths. Using modified DF models, the identification of varying ellipticity based on uniaxial and hydrostatic compression simulations enables a good representation of multiaxial compression loading. The FFT-based simulations highlight the strong influence of morphological properties, emphasizing the universality of microstructure-based models without the need for recalibration for different loading scenarios, but leading to increased computational effort. The findings provide valuable insights into the capabilities and limitations of the different DF models over a wide range of porosities and demonstrate the universality of microstructure-resolved simulations for reliable prediction of porous material behavior under varying loading conditions.