Molecular simulation for predicting the rheological properties of polymer melts

Bottom-up prediction that links materials chemistry to their properties is a constant theme in polymer simulation. Rheological properties are particularly challenging to predict because of the extended time scales involved as well as large uncertainty in the stress output from molecular simulation. This review focuses on the application of molecular simulation in the prediction of such properties, including approaches solely based on molecular simulation and its integration with rheological models. Most attention is given to the prediction of quantitative properties, in particular, those most studied such as shear viscosity and linear viscoelasticity. Studies on the fundamental understanding of rheology are referenced only when they are directly relevant to the property prediction. The review starts with an overview of the major methods for extracting rheological properties from molecular simulation, using bead-spring chain models as a sandbox system. It then discusses materials-specific prediction using chemically-realistic models, including systematically coarse-grained models that allow the mapping between scales. Finally, integrating molecular simulation with rheological models extends the prediction to highly entangled polymers. Recent development of several multiscale predictive frameworks allowed the successful prediction of rheological properties from the chemical structure for polymers of experimentally relevant molecular weights.