RandomDensePack and spheres ejections

Hello Anna,

a short script producing the error is always useful :-) could you please
provide us with some?
Thanks
Jan
Dne 17.5.2012 3:03 "anna effeindzourou" <
<email address hidden>> napsal(a):

> New question #197520 on Yade:
> https://answers.launchpad.net/yade/+question/197520
>
> Hi all,
> I generate a sphere packing using randomdensepacking in a box opened on
> the top. I just apply gravity. But even if I use a small time step (1e-8)
> and a high damping (0.9) some spheres are ejected. The spheres are in
> CohFrictMat.
> I don't understand why.
> Could you please help me?
> If you don’t understand my question, please don’t hesitate to let me know.
>
>
> Cheers,
> Anna
>
> --
> You received this question notification because you are a member of
> yade-users, which is an answer contact for Yade.
>
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-users
> Post to : <email address hidden>
> Unsubscribe : https://launchpad.net/~yade-users
> More help : https://help.launchpad.net/ListHelp
>

Hi,
I have prepared a short script .Maybe the problem is due to the fact that the is an overlap between spheres, when the randomdensepack is generate.
The contact stiffness is really important. I am trying to obtain a stress-free packing.
Thanks,
Anna

Here is the script.

# -*- coding: utf-8
from yade import ymport, utils,pack

O.materials.append(FrictMat(young=16e9,poisson=0.25,frictionAngle=27,density=1576.))

O.bodies.append(geom.facetBox(center=(0.0,0.0,0),extents=(.2,.2,.1),orientation=Quaternion((0.,0,0),0),wallMask=31,wire=True,dynamic=False,color=(0,1,1)))

pred=pack.inAlignedBox((-.2,-.2,-.1),(.2,.2,.1))
spheres=pack.randomDensePack(pred,spheresInCell=2000,radius=0.015,rRelFuzz=.5)
O.bodies.append(spheres)

O.engines=[
 ForceResetter(),
 InsertionSortCollider([Bo1_Sphere_Aabb(label='aabb'),Bo1_Facet_Aabb()]),
 InteractionLoop(
 [Ig2_Sphere_Sphere_ScGeom(label='Ig2ssGeom'),Ig2_Facet_Sphere_ScGeom(label='Ig2fsGeom')],
 [Ip2_FrictMat_FrictMat_FrictPhys(label='block_particles'),
 ],
 [Law2_ScGeom_FrictPhys_CundallStrack(label='Law_2'),
 ]),
 NewtonIntegrator(gravity=[0,0,-9.81],damping=0.5,label='Newton_Integrator'),
]

Hi,
I have prepared a short script .Maybe the problem is due to the fact that
the is an overlap between spheres, when the randomdensepack is generate.
The contact stiffness is really important. I am trying to obtain a
stress-free packing.
Thanks,
Anna

Hi,
Has someone have an ideas about it?
I think maybe the randomDensePack creates the pack with overlapping between particles.

Thanks in advance,
Anna

Hi Anna,

I would say you are right - the particles are slightly overlapping and
therefore the packing is not 100% stress free.. However, the magnitude of
forces induced by this should not be large.. Have you tried e.g. to add
gravity to your simulation?
Jan
Dne 28.5.2012 5:17 "anna effeindzourou" <
<email address hidden>> napsal(a):

> Question #197520 on Yade changed:
> https://answers.launchpad.net/yade/+question/197520
>
> anna effeindzourou posted a new comment:
> Hi,
> Has someone have an ideas about it?
> I think maybe the randomDensePack creates the pack with overlapping
> between particles.
>
> Thanks in advance,
> Anna
>
> --
> You received this question notification because you are a member of
> yade-users, which is an answer contact for Yade.
>
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-users
> Post to : <email address hidden>
> Unsubscribe : https://launchpad.net/~yade-users
> More help : https://help.launchpad.net/ListHelp
>

Hi Jan,
I take account of the gravity in the Newton Integrator.
I was thinking that the magnitude of this forces is linked to the value of the contact stiffness, in my case kn=25e7.
What do you think about that?

Thanks in advance,
Anna

This function is only a frontend to TriaxialTest, and many parameters are hardcoded in it, so that that it is less flexible than the triaxial test itself.
Among hidden parameters in the confining pressure, this one gives the initial overlaps.
In short, you confine the packing under, let's say 10kPa (I would have to check the precise value), then you remove the confining pressure and replace it by gravity, which is much smaller in magnitude (it seems). Suddenly, elastic energy is released and the particles jump away.

If I were you, I would directly use Triaxial scripts to generate the packing. It is not difficult and you have many example scripts. It would let you define the loading history more precisely.