Understanding and predicting structure-property relations in polymeric materials through molecular simulations

The development of computational methods for predicting thermal, mechanical and rheological properties of polymers from chemical constitution calls for hierarchical strategies, which are capable of addressing the broad spectra of length and time scales governing the behaviour of these materials. This paper reviews three recently developed strategies that appear particularly promising: (a) use of connectivity-altering Monte Carlo algorithms for rapid equilibration of atomistic models of long-chain polymer systems and calculation of their structural and thermodynamic properties; (b) mapping of molecular dynamics trajectories onto the Rouse and reptation models for the prediction of linear viscoelastic properties; (c) kinetic Monte Carlo simulations of large network specimens, generated on the basis of self-consistent field theoretical analysis, for tracking large-scale deformation and fracture of polymer-polymer interfaces. How connections can be established between the atomistic, mesoscopic (entanglement network) and macroscopic (continuum) descriptions is discussed. Validations of the simulation results against experiment are presented and questions pertaining to materials design are addressed.