Homogenization-based chemomechanical properties of dissipative heterogeneous composites under transient mass diffusion

The chemomechanical properties of heterogeneous composites under mass diffusion are of significance in modern advanced technology and engineering applications. A homogenization-based two-scale chemomechanical model is developed for heterogeneous composites undergoing chemical mass diffusion. A two-scale incremental variational formulation is established for heterogeneous composites consisting of multiconstituents featuring local dissimilar diffusion-deformation properties. The minimization problems for coupled chemomechanical behaviors are solved for both macrostructure and microstructure contexts, where the macroscopic material properties are extracted from the results of local boundary value problem on the nested representative volume elements (RVEs). Through a staggered finite element method (FEM) implementation procedure, the proposed homogenization-based two-scale solution algorithm is implemented in the FEM package ABAQUS (V6.14). The developed variational model and tangential algorithm is checked by solving chemomechanical properties of particles enforced composite, where several numerical examples are conducted applying two-scale solution algorithm and validated by full-scale simulations. Parametric studies are carried out on the size effects of RVEs, with respect to the ‘inertia effect’ associated with ‘moment of mass concentration’, and the coupling mechanisms are discussed for mechanical and chemical solutions. To the end, the inelastic dissipations are solved on subscale BVPs and their effects on the mechanical deformation and chemical mass diffusion are checked. The contributions of this work are mainly two-folds. One is the theoretical advance for self-consistent homogenization modeling of the coupled multi-physics of heterogeneous composites, and a rigorous FE² solution procedure. The other is providing numerical reference for evaluation of approximation algorithm as well as advanced data-driven method, which is needed for high-efficient material design.