[vtkusers] question about vtkMeshQuality.cxx

Clinton Stimpson clinton at elemtech.com
Tue Jul 20 10:28:35 EDT 2004


How about just plugging a quality library with the metrics already 
implemented into VTK?  Instead of reimplementing them.
This "verdict" library is LGPL and much research has gone into it, as 
well as work to make it efficient.
http://cubit.sandia.gov/verdict/
It supports, hexahedrons, tetrahedrons, triangle, quadrilateral, as well 
as simple information about wedges and pyramids.
For example, for tetrahedrons, it does calculations for aspect, volume, 
condition number, jacobian, shear, shape, relative size, and shape and 
size.  There's many calculations it can do for most element types.

Clint

>
>------------------------------
>
>Message: 4
>Date: Tue, 20 Jul 2004 15:03:09 +0200
>From: Alexandre gouaillard <alexandre.gouaillard at insa-lyon.fr>
>Subject: Re: [vtkusers] question about vtkMeshQuality.cxx
>To: vtkusers at vtk.org
>Message-ID:
>	<5.0.2.1.2.20040720145653.00c2c778 at pop3.creatis.insa-lyon.fr>
>Content-Type: text/plain; charset="us-ascii"; format=flowed
>
>Indeed , vtkMeshQuality was not enough for us too.
>
>The problem with the "quality" of a cell is that many measures exist in the 
>litterature depending on the application.
>As philippe seems interested in making something additions, I implemented 
>one of the triangle shape measures existing :
>1/2 perimeter * longest edge / area
>normalized by the value for an equilateral triangle.
>It measure the "stretching factor" of a triangle.
>
>please find the code attached.
>it's not perfect, some checking steps were never implemented, but in those 
>cases, a comment line exists to remind that a check should be made.
>If it can be of any help, use it.
>
>alex.
>
>
>
>  
>




More information about the vtkusers mailing list