We present a method of visualizing topological defects arising in numerical simulations of liquid crystals. The method is based on scientific visualization techniques developed to visualize second-rank tensor fields, yielding information not only on the local structure of the field but also on the continuity of these structures. We show how these techniques can be used to first locate topological defects in fluid simulations of nematic liquid crystals where the locations are not known a priori and then study the properties of these defects including the core structure. We apply these techniques to simulation data obtained by previous authors who studied a rapid quench and subsequent equilibration of a Gay-Berne nematic. The quench produces a large number of disclination loops which we locate and track with the visualization methods. We show that the cores of the disclination lines have a biaxial region and the loops themselves are of a hybrid wedge-twist variety.

Abstract-We present visualization tools for analyzing molecular simulations of liquid crystal (LC) behavior. The simulation data consists of terabytes of data describing the position and orientation of every molecule in the simulated system over time. Condensed matter physicists study the evolution of topological defects in these data, and our visualization tools focus on that goal. We first convert the discrete simulation data to a sampled version of a continuous second-order tensor field and then use combinations of visualization methods to simultaneously display combinations of contractions of the tensor data, providing an interactive environment for exploring these complicated data. The system, built using AVS, employs colored cutting planes, colored isosurfaces, and colored integral curves to display fields of tensor contractions including Westin's scalar cl, cp, and cs metrics and the principal eigenvector. Our approach has been in active use in the physics lab for over a year. It correctly displays structures already known; it displays the data in a spatially and temporally smoother way than earlier approaches, avoiding confusing grid effects and facilitating the study of multiple time steps; it extends the use of tools developed for visualizing diffusion tensor data, re-interpreting them in the context of molecular simulations; and it has answered long-standing questions regarding the orientation of molecules around defects and the conformational changes of the defects.

No abstract