[vtkusers] Clipping: "limping tetrahedra"

[David Pont]

A geometrical query,

  I have a rough first attempt at clipping working for vtkTetra, based on a 
"marching tetrahedra" approach. Hence the "limping tetrahedra" label for the 
moment. There are some minor problems with point ordering causing badly 
formed cells, and/or normals. But now that I have something working it is 
time to address a fundamental issue. As a first step I solved the clipping 
by output of tetrahedra or wedges.

The wedges were a shortcut to get moving, but I need to know if this is 
valid. For example when 4 edges of the tetra are intersected, 2 original 
tetra points are clipped away, 2 are left, and these together with 4 new 
points (from the intersections with the 4 edges) are used to define a wedge. 
In contouring, the analagous case generates TWO triangles. Does this imply I 
should generate 3 tetrahedra (1 wedge can be decomposed as 3 tetra)?

I can see possible issues with non-planar surfaces if I generate wedges. 
Also I want to make the solution general in the sense that the same approach 
could be applied to other cell types: wedge, pyramid (others?). The eventual 
goal is to implement a new filter: vtkClipUnstructuredGrid, which requires 
working Clip methods for all cell types. Inspection of vtkCell methods shows 
me that in general, contouring outputs triangles for various cell types. So 
this suggests clipping should output tetra in general.

Can any of the vtk experts (or geometry gurus) advise me?.

  regards
    Dave Pont
_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.

Share information about yourself, create your own public profile at 
http://profiles.msn.com.