Project Overview

Project Objectives

Project Proposal

Traces of 110 particles tracked in experimentally obtained Lang data set encoded with 2,105 RBFs.

We will solve the CFD visualization problems by developing techniques for creating a procedural abstraction for a very large dataset, developing effective and efficient methods for mapping from the procedural to visual representation, and applying these techniques to the problem of visualizing large CFD simulations. Our new methods will provide interactive visualization of very large datasets on a desktop computer, and will scale gracefully across a range of compute power and bandwidth situations.

Objectives

Our major research objectives are to:
1. Detect important features (e.g. shocks) in complex, highly-detailed flows using topological operators based on critical points and separatrix curves and surfaces.

2. Characterize the immense amount of data relative to these features using a procedural representation consisting of implicit models based on radial basis functions and free-form deformations based on subdivision solids.

3. Adapt the procedural representation to the appropriate level of detail using multi-resolution techniques based on multigrid methods.

4. Encapsulate domain specific knowledge as metadata to explore these extremely large datasets both at the feature level and, more importantly, at the higher level of relationships among features (e.g., tip vortices).

5. Visualize the data directly from the procedural representation, using and extending numerous existing CFD visualization techniques (e.g. cutting planes, isosurfacing, volume splatting, direct volume rendering, particle clouds, streams, rakes, line-integral convolution and glyphs).

6. Verify the accuracy of the procedural representation with careful tracking of approximation error throughout the entire process, including scanning, modeling, reconstruction and visualization.

7. Apply these techniques to the large-scale computational flow simulation problems currently studied at Stanford and at the SimCenter at the NSF Engineering Research Center at Mississippi State University.