Challenges trying to Voxelize VTP unstructured data from DEM Simulation

ParaView Support
1

Hi,

I am having trouble finding a solution for this problem.

I have a VTP dataset from a DEM simulation in MFIX. These are particle positions and other metrics as a function of time.

I would like to map this information to a grid so I can plot contours of local solid volume fraction, create slices, and overall manipulate the data.

With VTP data I cant even threshold the particles to take a look at the particles inside the core of the particle cloud (its a fluidized bed).

I have tried Resample with Dataset in Paraview 5.13.0 to resample the VTP data into the CFD grid, but Paraview always crashes.

I have tried other things like applying sections directly to the VTP data, or extract some particles to try and use that as a way to filter some particles out… But havent made anything work.

Is there a solution to this?

2

Please share your .vtp data

Best,

3

PARTICLE_2311.vtp (3.0 MB)

Here is an example.

4

Indeed, ResampleWithDataSet is not designed to work with point clouds, but you can use GaussianResampling filter for that:

image
image1117×625 31.3 KB

5

Okay,

Yea i had used GaussianResampling, but I thought there might have been another way of doing it.

6

I’m not sure what you are trying to. Gaussian Resampling will let you do the contours and thresholding you talked about in the original description. It would be easier to answer if you could be more specific if what you are trying to do or at least what Gaussian Resampling is not doing.

You can also try Point Volume Interpolator, which is similar to Gaussian Resampling but provides more controls. There is also Delaunay 3D, which will connect the points together in a tetrahedral mesh.

Perhaps you are looking for Convert To Point Cloud. That will give you points that you can run points like threshold and clip. Things like slice or contour won’t work though because these are still just discrete points in space.

1 Like
7

Kenneth,

I think GaussianResampling does what I needed to do. I was just trying to find a way to maybe make it more visually nice, less jagged and such, but I can play with the settings on the filter for that. We are basically comparing some of those contours to time averaged PIV contours from experiments.

But GaussianResampling will do, I think my expectations were too high on what the image I could get from resampling point data into a structured grid.