Kent,

You don't say what your application is, but I have a 
situation that may be similar.  I'm running an Eulerian 
finite difference code where space is partitioned into 
regular hexahedra.  Many of those cells contain 
multiple materials and material boundaries can only be 
inferred by the contents of neighboring cells.  I 
create fields containing the relative volume (volume of 
material / volume of cell) of each of the materials in 
the calculation, a single byte value is plenty of 
resolution and provides some space savings.  A single 
isosurface (normally at 0.5) of each of the fields, 
colored differently for each material, provides a good 
representation of the discontinuous regions of material 
concentration.

-Ned Piburn


Kent Eschenberg wrote:
> 
> Hi,
> 
> Has anyone found a way to show 3D regions of a particular value when the
> data is not continuous?
> 
> We have simple structured data where the integer at each location
> represents a material. As a result, interpolation, such as used to find an
> isosurface, is not appropriate. Putting a retangular glyph at the locations
> with the desired value produces the right external surface but  draws a
> great number of interior glyphs that are not seen and not used.
> 
> Thanks in advance for you suggestions.
> Kent
> - - -
> Kent Eschenberg   [EMAIL PROTECTED]   (412)268-6829
> Scientific Visualization Specialist
> Pittsburgh Supercomputing Center, CMU, Pittsburgh, PA
> 
> There are only 10 types of people in this world:
> Those who understand binary, and those who don't.

Reply via email to