(230p) Multiscale Modeling of Reacting Polymer Foams Via Computational Fluid Dynamics and Molecular Dynamics

Reacting polymer foams are very complex multiphase systems constituted by a continuous liquid phase, undergoing polymerization, and a disperse gaseous phase, that find important applications in the chemical industry. Among many practical examples, reacting polyurethane (PU) foams are particularly interesting, as they are widely used insulation materials. PU foams are obtained by mixing a solution of isocyanates with a solution of polyols, surfactants, catalysts, chemical and physical blowing agents. As isocynates and polyols undergo polymerization forming the final elastic solid polymer, the chemical and physical blowing agents produce gas bubbles that drive the expansion of the foam, and that are subjected to the processes of bubble growth and coalescence. This expansion is often carried out via mold filling processes, in order to confer specific (often very complex) shapes to the final PU foam.

Mathematical models, based on computational fluid dynamics (CFD), can be profitably used to simulate these mold filling expansion processes, especially when the volume-of-fluid method (VOF) is used, allowing tracking the evolution of the interface of the expanding foam within the mold. Recently one such a model was developed and applied to reacting PU foam simulation; the model included a simple but realistic representation of the polymerization kinetics and it also accounted for the presence of physical or chemical blowing agents. Notwithstanding its simplicity, fair agreement with experimental data was obtained.

In this work some of the limiting assumptions previously adopted are now relaxed. In particular:

(1) the density of the polymerizing liquid mixture, previously assumed fixed to a constant value, is now considered as a function of temperature and degree of polymerization (or cross-linking),

(2) the temperature dependence of the solubility of the blowing agents (e.g., R-11, in the case of physically-blown foams, and of the produced carbon dioxide, in the case of chemically-blown foams), ignored in the previous model, is now fully considered.

As the experimental investigation of these PU foams is very complex, density and solubility dependence are investigated by using molecular dynamics (MD) simulations. The results obtained are then wrapped into simple surrogate models and linked to the CFD simulations, resulting in a significant improvement of the agreement with experiments.