Grzegorz Karol Karch

Visualization of Advection-Diffusion in Unsteady Fluid Flow

Advection has been the standard transport mechanism in flow visualization. Diffusion, in contrast, has not been considered important in visual flow field analysis so far, although it is inherent to many physical processes. We present a novel technique that allows for interactive 3D visualization of both advection and diffusion in unsteady fluid flow. We extend texture‐based flow visualization, which is advection‐oriented, by diffusion. Our finite volume approach based on WENO (weighted essentially non‐oscillatory) reconstruction is well parallelizable and features low numerical diffusion at interactive rates. Our scheme contributes to three different applications: (a) high‐quality dye advection at low numerical diffusion, (b) physically‐based dye advection accounting for diffusivity of virtual media, and (c) visualization of advection‐diffusion fluxes in physical media where the velocity field is accompanied by a concentration field. Interactive rendering of the virtual dye is accomplished by ray casting. We apply our GPU implementation to CFD examples of thermal convection and evaporation phenomena.

Visualization of molecular structures using state-of-the-art techniques in WebGL

The potential of web browsers as a platform for remote visualization has long been recognized. One of the motivations behind remote visualization has mainly been large data that cannot be easily transferred to the client. Further impeding factors for local visualization at the client side included lack of compute power or capabilities as well as lack of native support for hardware-accelerated APIs in the browser. However, moving the visualization to a remote site is challenging because of high latency and low available bandwidth affecting interactivity and visual quality. Although bandwidth is continuously improving, latency remains an issue. To obtain optimal interactivity, the rendering needs to be performed on the client machine. With the introduction of WebGL, it is now possible to exploit the power of the GPU inside the browser without third-party additions. We present a GPU-based ray-casting technique for the visualization of molecular structures that adheres to the restrictions of WebGL. We describe implementation details with respect to these restrictions and present the results and performance. Biomolecules with tens of thousands of atoms can be rendered at interactive rates. This shows the potential of establishing the web as an attractive platform for interactive visualization using GPU-accelerated algorithms.

Visualization of piecewise linear interface calculation

Piecewise linear interface calculation (PLIC) is one of the most widely employed reconstruction schemes for the simulation of multiphase flow. In this visualization paper we focus on the reconstruction from the simulation point of view, i.e., we present a framework for the analysis of this reconstruction scheme together with its implications on the overall simulation. By interpreting PLIC reconstruction as an isosurface extraction problem from the first-order Taylor approximation of the underlying volume of fluid field, we obtain a framework for error analysis and geometric representation of the reconstruction including the fluxes involved in the simulation. At the same time this generalizes PLIC to higher-order approximation. We exemplify the utility and versatility of our visualization approach on several multiphase CFD examples.

Visualization of 2D unsteady flow using streamline-based concepts in space-time

Treating time as the third dimension of 2D time-dependent flow enables the application of a wide variety of visualization techniques for 3D stationary vector fields. In the resulting space-time representation, 3D streamlines represent 2D pathlines of the original field. In this paper, we investigate the application of different streamline-based visualization concepts to the 3D space-time representation of 2D time-dependent flow. As a consequence, we obtain from each streamline-based concept a Galilean-invariant counterpart that takes the time dependence of the original field explicitly into account. We show the advantages of the overall approach for vortex analysis and the analysis of the dynamics of material lines. In particular, we employ the concept for the extraction of vortex centers, vortex core regions, and the visualization of material line dynamics using streamsurface integration and line integral convolution in the space-time field. We exemplify the utility of our visualization approach using two 2D time-dependent datasets that exhibit vortical flow.Graphical Abstract

Topological Features in Time-Dependent Advection-Diffusion Flow

Concepts from vector field topology have been successfully applied to a wide range of phenomena so far—typically to problems involving the transport of a quantity, such as in flow fields, or to problems concerning the instantaneous structure, such as in the case of electric fields. However, transport of quantities in time-dependent flows has so far been topologically analyzed in terms of advection only, restricting the approach to quantities that are solely governed by advection. Nevertheless, the majority of quantities transported in flows undergoes simultaneous diffusion, leading to advection-diffusion problems. By extending topology-based concepts with diffusion, we provide an approach for visualizing the mechanisms in advection-diffusion flow. This helps answering many typical questions in science and engineering that have so far not been amenable to adequate visualization. We exemplify the utility of our technique by applying it to simulation data of advection-diffusion problems from different fields.

Visual Analysis of Inclusion Dynamics in Two-Phase Flow

In single-phase flow visualization, research focuses on the analysis of vector field properties. In two-phase flow, in contrast, analysis of the phase components is typically of major interest. So far, visualization research of two-phase flow concentrated on proper interface reconstruction and the analysis thereof. In this paper, we present a novel visualization technique that enables the investigation of complex two-phase flow phenomena with respect to the physics of breakup and coalescence of inclusions. On the one hand, we adapt dimensionless quantities for a localized analysis of phase instability and breakup, and provide detailed inspection of breakup dynamics with emphasis on oscillation and its interplay with rotational motion. On the other hand, we present a parametric tightly linked space-time visualization approach for an effective interactive representation of the overall dynamics. We demonstrate the utility of our approach using several two-phase CFD datasets.

Dye-Based Flow Visualization

Computational dye advection helps engineers understand fluid dynamics simulations by providing interactive tools that mimic physical experiments.