Embedded Boundaries and AMR Vector Fields
(Summer Work Report)

Soon Tee Teoh
James Chen

UC Davis

Monday, September 25, 2000
1:00 - 2:00 PM
50F Conference Room

Currently the Visualization Group works with researchers who are doing embedded boundary, EB, computations and/or adaptive mesh refinement, AMR, computations. Soon Tee has been working on visualizing embedded boundaries and James has been working on visualizing vector fields defined on AMR grids. Both of these projects are part of a larger effort to extend visualization algorithms and techniques to EB and AMR data.

In some of the fluid dynamics simulations being done by the Applied Numerical Algorithms Group, ANAG, there are data sets which have embedded boundaries. Embedded boundaries divide the domain of the computation into regions where fluids are and are not present. The embedded boundary is not represented explicitly as geometry. Instead, cells at the boundary contain the information related to the boundary that is needed to carry out the simulation, e.g. volume fractions, boundary normals, area fractions, centroids. The long-term goal of Soon Tee's project is to visualize such a data set. One important task is to generate an approximation to geometry of the boundary.

Soon Tee will give a description of the data sets and the methods he has developed to generate the embedded boundary geometry. He will also describe future work.

The visualization and exploration of time varying vector fields is a difficult and largely unsolved problem. For specific computations in specific areas of science and engineering notable progress has been made. We would like to address visualizing and exploring time varying vector fields defined on AMR grids. This is a large problem and James has begun some of the work necessary to do this.

Specifically, James will present his summer work on vector field visualization on AMR grids with VTK. First, a C++ class was written to convert the HDF5 file to VTK Structured Points format. Scalar and vector data were then visualized with VTK. The VTK streamline function was also modified to utilize the higher resolution data available in the finer grids. James will also discuss future work.

Snacks will be provided.

See Conundrum Talks for more information about this series.