yade-users team mailing list archive

[Jan Stránský <question698605@xxxxxxxxxxxxxxxxxxxxx>]

Question #698605 on Yade changed:
https://answers.launchpad.net/yade/+question/698605

Jan Stránský posted a new comment:
> Despite the apparent randomness in variation of contact forces,
regular trends emerge when normal and tangential components of
interparticle forces are averaged over groups of contacts with similar
orientations.

ok. Just have in mind the case that in one bin you have one force 9 N,
in another bin 3 forces 3 N. Depending on what you really want, plain
sum or average gives (of course) totally different results

Anyway, in your case, you assign full force to the bins, no matter their relation to the plane.
So if the force is perpendicular to the plane, you still assign its full value, even if it sounds logical not to include it at all.
Consider assigning just the projection of the force to the plane.
Or just select "not too much perpendicular" interactions.
Or ...
There are really many options of the evaluation.

I suggest:
- using (trying at least) the same method for both forces and direction
- investigate why the 3 forces plot is exactly the same
- investigate if wall interactions were used for direction plot (they are used in "standard" utils.plotDirections)
- take into account force direction w.r.t. plane
- test the method on artificial input with known results

For further reasonable discussion, please provide a MWE (final packing
and plot evaluation, no simulation).

Cheers
Jan

-- 
You received this question notification because your team yade-users is
an answer contact for Yade.