Resampling of Large Datasets for Industrial Flow Visualization

Many industrial flow simulations calculated in the car development process are defined on unstructured grids. However, visualization of unstructured grids is computationally expensive and often does not allow for interactive work with large datasets. On the other hand, there are efficient visualization algorithms for Cartesian grids taking advantage of the grid’s regular topology that achieve interactive frame rates even for datasets consisting of millions of cells. In this paper, algorithms are presented and evaluated that resample an unstructured grid on a hierarchy of Cartesian grids. While resampling algorithms are well-known in general, it turns out that tayloring them to a specific application scenario and optimizing them for datasets with tens of millions of tetrahedra is by no means trivial. Using an adapted version of the commercially available visualization tool PowerVIZ, we enable engineers to visualize large unstructured datasets at interactive frame rates without ha...

• Type: Conference
• Year: 2003
• Where: VMV
• Authors: Simon Stegmaier, Martin Schulz, Thomas Ertl
• Keywords: Cartesian Grids, Interactive Frame Rates, Unstructured Grid, VMV 2003, VMV 2008