← Back to team overview

yade-users team mailing list archive

Re: [Question #181411]: cylinder larger than half of the periodic cell

 

Question #181411 on Yade changed:
https://answers.launchpad.net/yade/+question/181411

    Status: Answered => Open

Giulia Macaro is still having a problem:
I'm using 3036, but I've commented the 'if(scene->Periodic)' condition in the Bo1_ChainedCylinder_Aabb (line 328).
I'm sending you my script.

1.  Thanks for the detailed explanation. The example of the sphere on the inclined plane is quite useful.
But how many cells are reproduced during the simulation? Is it an infinite number? I assume that, if I have a sphere falling under gravity inside a periodic cell without any plane, it keeps on going out and in the primary cell, in loop for ever, is that right?

2.  I've run some more tests, and now it works fine with a large number of chains (Ne=30 in generatePipe() of my script) and blocking the rotations and lat translations.
If then I  make the cylinder with only one chain (Ne=1), and still block xXYZ, the cylinder doesn't move (as expected). Plus it seems to be periodic only on one side.
If I remove the constrain to the DOFs, and make the cylinder with only one chain, I've Segmentation Fault.
Could I just use the long chain with blockDOFs='xXYZ'? The cylinder seems to behave still dynamically..

3. Is there any difference between creating a periodic cell (O.periodic=True and O.cell.setBox()) and creating a periodic packing (particleSD(periodic=True))? 
My script works fine with O.periodic=True and particleSD(periodic=False), but gives weird behaviour when also particleSD(periodic=True).

4.  In the periodicSandPile example, it's not clear to me why the bottom
floor is not periodic. In my simulation I always have a periodic bottom
floor, and so the cylinder.  How do I make them non-periodic?

Thank you

-- 
You received this question notification because you are a member of
yade-users, which is an answer contact for Yade.