An efficient multiscale model of damping properties for filled elastomers with complex microstructures

This work proposes an efficient framework for prediction of filled elastomer damping properties based on imaged microstructures. The efficiency of this method stems from a hierarchical multiscale modeling scheme, in which the constitutive response of subcell regions, smaller than a representative volume element (RVE), are determined using micromechanics; the resulting constitutive parameters then act as inputs to finite element simulations of the RVE, from which damping properties are extracted. It is shown that the micromechanics models of Halpin-Tsai and Mori-Tanaka are insufficient for modeling subcells with many filler clusters, and thus these models are augmented by an additional interaction term, based on stress concentration factors. The multiscale framework is compared to direct numerical simulations in two dimensions and extended to predictions for three dimensional systems, which include the response of matrix-filler interphase properties. The proposed multiscale framework shows a significant improvement in computational speed over direct numerical simulations using the finite element method, and thus allows for detailed parametric studies of microstructural properties to aid in the design of filled elastomeric systems.