Virtual rheoscopic fluids for dense, large-scale fluid flow visualizations

Fluid flow visualizations typically employ a combination of isosur-facing and particle tracing, such as streamlines. These techniques capture surface flow well, but they can obscure deeper features from view. Further, these techniques present visualizations that are far removed from physical methods, which complicates their analysis. However, virtual rheoscopic fluids (VRF) provide a physically-based visualization that is similar to, and can be similarly analyzed as, physical rheoscopic fluid. However, we need a scalable VRF algorithm to apply the technique to large-scale data. In this paper, we present a scalable VRF algorithm based on ray casting implemented in a general visualization toolkit. We demonstrate the application of these techniques to large-scale flow problems beyond what would be physically feasible in the laboratory, and we show that our approach scales through 256 cores.