Problem Using CMesh Class

Jarvis · Post by **Jarvis** » Mon Dec 15, 2008 9:35 pm

I have been experimenting with creating meshes of yarns using the CMesh class. I have used http://texgen.sourceforge.net/index.php ... xport_Mesh as a guide. I find that sometimes after I add the volume to the mesh many of the node coordinates are as follows:

mesh.GetNode(1)
-1.#IND, -1.#IND, -1.#IND

What does the #IND mean, and how do I get rid of it?

Jarvis

wr · Post by wr » Mon Dec 15, 2008 10:34 pm

is the node inside the volume?
Does the .vtk conversion work?

Jarvis · Post by **Jarvis** » Fri Dec 19, 2008 5:52 pm

I don't know where the #IND Nodes are as I can't get their coordinates. They are generally the really early nodes i.e. the nodes with low index numbers like 0 through 50, and after that all of the nodes are fine.

I don't know about .vtk, I don't have the software to view the results, and I haven't really experimented with it. The conversion completes in TexGen with no errors, but I don't know if it produces valid results. The SaveToABAQUS() works, but the input file lists the nodes with the same strange coordinates.

Thanks for the help!

Jarvis

wr · Post by wr » Tue Dec 30, 2008 6:45 pm

Hi Jarvis,

I was asking to see if the nodes are outside of the domain, maybe something to do with periodicity or maybe the interference correction not coming up with a proper coordinate, if you have a basic reproduction of the error post the script so we can have a look.

If you want to look at vtk files check out paraview:
http://www.paraview.org/

W

Jarvis · Post by **Jarvis** » Mon Jan 12, 2009 11:05 pm

It seems to be independent of the domain and periodicity as I can get the error without specifying repeats or domain. I just took a simple script that was causing the error and changed the way that the sections were defined, and the error went away.

So, I am not sure exactly what the error is, but I believe it has to do with section assignments.

Jarvis