Characterization and Design of Three-Phase Particulate Composites: Microstructure-Free Finite Element Modeling vs. Analytical Micromechanics Models

Abstract

Three-phase particulate composites offer greater design flexibility in the selection of phase materials and have more design variables than their two-phase counterparts, thus providing larger space for tailoring effective properties to meet intricate engineering requirements. Predicting effective elastic properties is essential for composite design. However, experimental methods are both expensive and time intensive, whereas the scope of analytical micromechanics models is limited by their inherent assumptions. The newly developed microstructure-free finite element modeling (MF-FEM) approach has been demonstrated to be accurate and reliable for two-phase particulate composites. In this study, we investigate whether the MF-FEM approach can be applied to three-phase particulate composites and, if applicable, under which conditions. The study commences with a convergence analysis to establish the threshold ratio between the element size and the RVE (representative volume element) dimension. We then validate the MF-FEM approach using experimental data on three-phase composites from the existing literature. Subsequently, the MF-FEM method serves as a benchmark to assess the accuracy of both traditional and novel analytical micromechanics models, in predicting the effective elasticity of two distinct types of three-phase particulate composites, characterized by their small and large phase contrasts, respectively. We found that the threshold element-to-RVE ratio (1/150) for three-phase composites is considerably smaller than the ratio (1/50) for two-phase composites. The validation underscores that MF-FEM predictions align closely with experimental data. The analytical micromechanics models demonstrate varying degrees of accuracy depending on the phase volume fractions and the contrast in phase properties. The study indicates that the analytical micromechanics models may not be dependable for predicting effective properties of three-phase particulate composites, particularly those with a large contrast in phase properties. Even though more time-intensive, the MF-FEM proves to be a more reliable approach than the analytical models.