Bottom-up Multiscale Approach to Estimate Viscoelastic Properties of Entangled Polymer Melts with High Glass Transition Temperature

A multiscale computational method is presented for the prediction of the viscoelastic properties of entangled homopolymer melts with high glass transition temperatures. Starting from an atomistic model of a polymer, two coarser representations are introduced – a coarse-grained model and a slip-spring representation, that successively operate at longer time and length scales. The three models are unified by renormalizing the time and modulus scales, which is achieved through matching their normalized chain mean squared displacement and stress relaxation modulus, respectively. To facilitate the relaxation of entangled chains, the simulations are performed at temperatures higher than those accessible in experiments. Time-temperature superposition is then applied to extrapolate the viscoelastic properties calculated at high temperatures to experimentally accessible lower temperatures. This proposed approach can predict the linear rheology of the melt starting from an atomistic model and does not require experimental parameters as an input. Here it is demonstrated for syndiotactic polystyrene, where good agreement with experimental measurements is achieved.