(291c) Multiscale Simulation of Flow-Induced Crystallization in Polymers

The structure and properties of most semicrystalline polymers depend sensitively on both the chemical physical nature of the polymer and the manner in which it has been processed. These phenomena in turn are operative on different, but inter-dependent, length and time scales. Whereas crystallization and the dynamics of chains engaged in crystallization are sensitive to details of interaction between atoms, chemical groups and short chain segments, the response of the polymer to flow depends on distribution of lengths of entire chains, the dynamics of entanglement between chains, and the networks that arise with crystallization. Modeling flow-induced crystallization in polymers thus requires a two-way coupling between atomistic and molecular scales – to capture not only the effect of flow on crystallization kinetics, but also the effect of crystallization on the rheology of the flowing polymer melt. Taking advantage of significant advances in recent years in both the modeling of crystal nucleation and growth processes, and in the rheological description of industrially relevant polymers, a coupled, multiscale description of polymer crystallization in process flows is now emerging.

In this talk, we discuss first the use of nonequilibrium molecular dynamics (NEMD) simulations with atomistic resolution to characterize the nucleation of a new crystal phase from a melt of linear polyethylene-like chains under homogeneous flow conditions. A method based on mean first-passage times is used to characterize the kinetics and physics of nucleation enhanced by flow, in both simple shear and uniaxial extension. These results are used to assess several of the existing models in the literature for flow-enhanced nucleation, and several new models are proposed, based on physical insight into the nucleation phenomenon made possible by such simulations. Evidence for the emergence of a flow-stabilized nematic phase in simple polyolefins may also be presented.

In the second part, we present a variation of the discrete slip-link model (DSM) for the rheology of entangled polymer melts that accounts for partial crystallinity through blending of free chains with crosslinked chains that resemble the bridge and/or tail (“dangling”) segments between developing crystallites. The evolution of the crosslinked chain population describes the development of the physical network associated with the nucleation and growth of crystallites within the flowing melt. Combined with a suspension model to reflect the stiffening associated with a growing crystal phase, this modified DSM simultaneously captures the evolution of viscosity and elasticity over the whole range of frequencies in the linear regime for a number of different polyolefins. Kinetic modeling at the two scales is coupled through the dependence of nucleation kinetics on molecular conformation, as predicted by the modified DSM, and through the dependence of rheology on the nucleation and growth of crystallites predicted by atomistic NEMD.

We believe such simulations represent a significant advancement in the chemical engineer’s ability to model polymer process flows commensurate with solidification, and thereby to develop a deeper understanding for the essential physics that underlie the coupling between fluid mechanics and phase change in an industrially important class of materials. Ramifications for similar coupling between flow and solidification in other important systems can be imagined.