volume rendering for unstructured grid returns error, how to solve that?

ParaView Support
1

Hi,

I created a vtu file, which originally was csv-like file, its format: x y z + point data. However, when trying to do volume rendering, I encountered error: vtkOpenGLProjectedTetrahedraMapper (000002413E153CE0): Encountered non-tetrahedra cell!

How can I fixed this problem except doing interpolator? Doing this will lose some high frequency information. And is the method paraview use is pre-integrated transfer function? Someone told me this generate higher quality result than other method.Thanks!

2

Use Tetrahadralize filter ?

4

I’ve tried, but the error is still there.

5

Please share your data.

6

mine.vtu (7.3 KB)

This is part of the complete version and shares all the problem I met in the complete one
Original one is too big to upload and test

7

Your dataset contains only points with all the same scalars values.
You cannot volume render this.

You can convert your points into a volumic datasets using Delaunay3D filter, that will fix your initial issue, but your data is still identical everywhere, which cant be volume renderered.

1 Like
8

sorry, that part was too small to show the differences, this is a larger samplemine2.vtu (3.7 MB)
And the original one has over 10 million points, is delaunay3d possible to produce result? Or are there lighter way?

9

is delaunay3d possible to produce result?

It should work. But ResampleToImage could be a more efficient way.

Or are there lighter way?

Providing the topology in your original dataset if you already have it

1 Like
10

Actually, I’m looking at your second dataset, delaunay 3d will give you unsatisfactory results with this kind of disconnected unstrucutred grid.

Even Resampling to an image wont work great.

In order to volume render, you need to provide a topology.

Allternatively, do not use vol

1 Like