Liquid crystalline polymers (LCPs) have found a multitude of applications because they exhibit exceptional physical properties. However, the major uses all involve LCPs as fibers; either synthetic materials such as Kevlar, which is produced by Dupont, or naturally occurring materials like coal and petroleum pitches, which are converted from a liquid crystalline state into high strength carbon fibers. In spite of these successes, however, other potentially more important applications have not been realized. The excellent mechanical properties of LCPs derive from the spontaneous orientational ordering that they display in the liquid state. This tendency for orientational ordering is enhanced by the spinning process used to make fibers. However, other forms of polymer processing, such as injection molding, that are used to make sheets, or objects of more complex shape, involve shear flows, or combinations of shear and extensional flows. For LCPs that “tumble” in shear flows (including all lyotropic materials and, apparently, many of the commercially interesting main-chain thermotropic LCPs (TLCPs)) these flows seriously degrade or even destroy the orientational order, leading to a “polydomain” structure of microdomains and defects (disclinations) that is isotropic (or nearly so) from a mesoscopic perspective. The resulting materials do not realize the special potential of LCPs, and in view of the high cost of synthetic LCPs, they have little commercial value.
The question, then, is whether the formation of disclinations and the polydomain structure can be controlled sufficiently by some modification of the processing procedure (for example by modest changes in the mold geometry), or by changes in the chemical structure of the LCP (for example by using more flexible spacers in the case of a synthetic main-chain LCP) to minimize the tendency of the LCP to “tumble” in shear flows. The present research project is focused on the first of these possibilities. A necessary first step is to understand how flow fields interact with an LCP to change both the orientation, and the degree of alignment at the molecular level, and to develop a predictive capability that reproduces experimental observations. This includes the formation and/or proliferation of orientational defects (the disclinations). The major source of difficulty is that one must predict both the flow and the coupled material structure. The majority of preceding attempts to address this problem have not considered this coupling. Instead, they have focused on predictions of changes in the material structure, assuming that the flow adopts a simple predetermined form (such as simple shear flow). In contrast, the objective of this research is the analysis of the full coupled problem for 3D flows that occur in a simple shear cell, and in pressure-driven channel flows for both 2D and 3D channel geometries, where the channel may either have a converging or diverging cross-section, or may be a parallel-wall channel.
Our most recent progress and results can be found in a recent paper published in Proceedings of the 12th Computational Techniques and Applications Conference CTAC-2004, volume 46, pages C210-C244, April 2005 and poster presented at the National Science Foundation 2005 IGERT Project Meeting. Another paper was recently submitted to Physics of Fluids.