Wear prediction

Submitted by alexander.polson on Mon, 12/06/2010 - 18:05

Hi Everyone

I dont know if this problem is a post-processing or a LIGGGHTS problem. I hope this is the right forum.

I ran the "chute_wear" example in (LIGGGHTS 1.2.1). During visualizing the results in Paraview, I get an error which looks like the wear prediction function does not work right.

I show in the attached 3x files a Paraview view of the model from above (at the same timestep for all 3x files):
- "wear.png"
- "shear.png"
- "pressure.png"

Looking at "wear.png", it appears that:
1) there is no cumulative wear caused by the region of the initial impact of each of the spheres in the vertical/downwards stream.
2) all wear is confined to the bottom left part of the picture

In comparison, for both "shear.png" and "pressure.png" it appears that shear and pressure "events" registers for triangular faces where spheres touch, and also that this happens both above and below the red line (x-axis) in the mentioned images.

A further problem arises at Paraview snapshot no 240 (i.e. at timestep = 240*200 = 48,000):
From this point till the end of the visualization in Paraview there is no further shear or pressures registering anywhere a sphere touches the geometry. Compared to the wear visualisations, there is at least something that looks right untill frame 240, whereas for wear it looks wrong the whole time.

Any thoughts, please?

Kind regards

Alexander

AttachmentSize
Image icon wear.png77.59 KB
Image icon pressure.png76.13 KB
Image icon shear.png75.61 KB
ckloss's picture

ckloss | Mon, 12/06/2010 - 18:25

>>I show in the attached 3x files
There are no files attached?

>> get an error which looks like the wear prediction function does not work right.
What error do you get?

For me it works fine with 1.2.2 - so I think it is probably related to postprocessing

Cheers,
Christoph

alexander.polson | Mon, 12/06/2010 - 18:30

Please see my update - they didnt upload at my previous attempt.

My apologies

Alexander

ckloss's picture

ckloss | Mon, 12/06/2010 - 18:41

Ok; I see the pics now - and I agree that the wear pic can not be correct. But unfortunately its hard to do remote diagnostics... because it's running fine on all system we have tried it so far. Maybe you can re-try with a second version of paraview and/or look into the VTK file to narrow down the error

Christoph