A Tool for Visualizing Large-Scale Interactions between Turbulence and Particles in 3D Space through 2D Trajectory Visualization

Abstract

Particle-laden turbulent flows are relevant to many industrial, biological, and environmental applications. In these flows, there are many interactions that occur between the carrier-fluid turbulence and the dispersed particles inside; visualization tools aid in understanding the complex physical phenomena at various spatial and time scales in such flows. For a large scale 3D particle-laden turbulence, visualizing the vector data in a dynamic environment makes the computation expensive. Furthermore, the interaction between the turbulence and particles is difficult to observe in 3D. The particle-turbulence interaction is easy to observe using 2D visualization, yet 2D visualization cannot properly capture the dynamics of the interactions. In this paper, we explore a method which can visualize the 3D effect of particle-turbulence interaction as a 2D trajectory visualization. We divide the 3D particle trajectory into several segments and, for each, we compute the plane crossing the line segment. As the turbulence vectors across the plane have a significant impact on the particle motion, we only visualize the planes where the particles are localized. Since we select the salient planes to visualize instead of the whole 3D vector cube, considerable improvements can be achieved with regards to the memory and time performance. For each plane, we use Line Integral Convolution to visualize the turbulence. Originally, Line Integral Convolution only visualizes the vectors’ direction. Instead, we not only visualize the vector directions but also the magnitude of the flow by utilizing a more meaningful texture, which generates a more compelling visualization.KeywordsDirect Numerical SimulationParticle TrajectoryInput TextureParticle PathDirect Numerical Simulation StudyThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.