Two- and three-dimensional micromechanical viscoelastic finite element modeling of stone-based materials with X-ray computed tomography images

This paper presents 2D and 3D micromechanical finite element (FE) models to predict the viscoelastic properties including dynamic modulus and phase angle of stone-based materials (using an example of asphalt mixtures). Heterogeneous asphalt mixtures are consisted of very irregular aggregates, asphalt matrix and a small amount of air voids. The internal microstructure of asphalt mixtures was captured with X-ray computed tomography (CT) imaging techniques. The 2D and 3D digital samples were created with the reconfiguration of the scanned horizontal surface images. The FE mesh of digital samples was generated with the locations of image pixels within each aggregate and asphalt matrix. Along the boundary of these two phases, the aggregate and matrix FEs share the nodes to connect the deformation. The micromechanical FE model was accomplished by incorporating the captured microstructure and ingredient properties (viscoelastic asphalt matrix and elastic aggregates). The generalized Maxwell model was applied for viscoelastic asphalt matrix with calibrated parameters from the nonlinear regression analysis of the lab test data. The displacement-based FE simulations were conducted for the uniaxial compression under sinusoidal cyclic loading. Overall, the predicted dynamic modulus and phase angle from 2D and 3D micromechanical models were compared favorably with lab test data of the asphalt mixture specimens. The 3D simulation with digital samples generated better prediction than the 2D models. These results indicate that the developed micromechanical FE models have the ability to accurately predict the global viscoelastic properties of the stone-based materials.