Non-segmentation image-based finite element modeling (NSIB-FEM) for mechanical characterization and design of composites of fine particles

Abstract

Image-based finite element modeling (IB-FEM) is a powerful tool for the characterization and design of various composite materials. However, conventional IB-FEM methods suffer from a number of deficiencies. To eliminate these deficiencies, a so-called non-segmentation image-based finite element modeling (NSIB-FEM) approach is proposed in this paper. The major difference between NSIB-FEM and conventional IB-FEM is that a finite element is allowed to contain more than one material phase, and the key step in NSIB-FEM is to determine element effective material properties. Theoretically, any micromechanics model of particulate composites can be applied for the calculation of element effective properties. A number of such micromechanics models, including Voigt’s model, Reuss’ model and Mori-Tanaka method, were tested to study how they affect the accuracy of NSIB-FEM. NSIB-FEM was validated against reported experimental data and compared with predictions by voxel-based finite element modeling. The results show that NSIB-FEM is a promising approach for the characterization of fine-particulate composites, the micromechanics model used to calculate element effective material properties has significant effect on the accuracy of NSIB-FEM. Further effort needs to be devoted to develop a more accurate method for the calculation of element-level effective material properties with the ability to deal with high volume-fraction of inclusion and element-level material anisotropy.