[MOAB-dev] MOAB question

[shontz at cse.psu.edu]

Dear Jason,

You were right about this -- I saw the 4 vertices and had assumed they
were tets.  But closer inspection shows they were quads.  Your
postprocessing step (using mbconvert with the -3 option) worked.  So now I
have the hex meshes with the corresponding boundary information I need.

Thanks,
Suzanne


> Suzanne Shontz wrote:
>> Jason & Tim,
>>
>> Actually, on closer inspection of the resulting meshes, I see that
>> mbskin did *not* yield the desired result.  mbconvert yielded hex meshes
>> (but without boundary information), whereas mbskin (when called as
>> indicated below) yielded the desired boundary information, but it
>> returned hybrid meshes with tet and hex elements.  Any idea why this is
>> happening?  The meshes are definitely hex meshes I'm trying to convert.
>> Any suggestions
>> would be appreciated.
>
> It definitely should not be creating any tetrahedral elements.  If it
> is, that is certainly a bug.  Are you sure that it is creating
> tetrahedral elements and not quadrilateral ones?  The mbskin utility
> might be creating the quadrilateral elements that define the skin of the
> mesh.  If that is the case, you can use mbconvert to remove those from
> the file:
>    mbconvert -3 input.vtk output.vtk
>
> - jason
>