[vtkusers] cutting through unstructured grid

Simon J. Rees sjrees at okstate.edu
Wed Dec 13 14:56:29 EST 2000


If I use a vtkGeometryFilter to get the surface of an unstructured mesh or a
plane of a structured mesh (by setting appropriate extents etc.) there is no
problem with additional triangles. Can somebody explain what happens in this
case - in terms of type of data output etc. - that avoids the tiangle
proplem and appears to give polygons?

Thanks,

Simon

----- Original Message -----
From: "Bill Lorensen" <wlorens1 at nycap.rr.com>
To: "Philippe Guerville" <philippe.guerville at free.fr>; <sjrees at okstate.edu>
Cc: <vtkusers at public.kitware.com>
Sent: Saturday, December 09, 2000 3:19 PM
Subject: Re: [vtkusers] cutting through unstructured grid


> Philippe,
>          When I originally implemented Marching Cubes in 1984, I generated
polygons rather than triangles. The problem is that most renderers
(including OpenGL) do not handle non-planar polygons consistently. In 1984,
I wrote my own renderer which I recall did handle non-planar polygons. But
the problem is more difficult if you want to interpolate color and texture
coords.
>
>
> Bill
>
> At 09:25 PM 12/7/00 +0100, Philippe Guerville wrote:
> >Simon:
> >
> >vtkCutter ask each cell to compute the intersection with the plane using
the
> >marching cubes algorithm (The Visualization ToolKit 2nd Edition, p 161).
The 256
> >intersection cases are described using triangles (file
> >common/vtkMarchingCubeCases.h triCases array). What I've done is to
recombine
> >triangles into polys and to change the function vtkHexahedron::Contour
(file
> >common/vtkHexahedron.cxx line 398).
> >
-- snip --





More information about the vtkusers mailing list