Yade

Intraction between wall and particles

Asked by Hossein

Hi every body

I tend to consider Hertz Mindlin contact law as particle's interactions and consider Cundall contact law in terms of wall-particle interaction.

Is the below-mentioned code are working truly according the above note?

O.engines=[
 ForceResetter(),
 InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Box_Aabb()]),
 InteractionLoop(
  [Ig2_Sphere_Sphere_ScGeom6D(),Ig2_Box_Sphere_ScGeom()],
  [Ip2_FrictMat_FrictMat_FrictPhys(),Ip2_FrictMat_FrictMat_MindlinPhys()],
  [Law2_ScGeom_FrictPhys_CundallStrack(),Law2_ScGeom_MindlinPhys_Mindlin(includeAdhesion=False,includeMoment=False,label = 'Mindlin_Law')]
 ),
 GlobalStiffnessTimeStepper(active=1,timeStepUpdateInterval=100,timestepSafetyCoefficient=0.8),
 VTKRecorder(iterPeriod=1000,fileName="./export",recorders=['spheres','intr','color'],dead=1,label='VTK'),
 triax,
 newton,
 ]

Question information

Language:
English Edit question
Status:
Solved
For:
Yade Edit question
Assignee:
No assignee Edit question
Solved by:
Hossein
Solved:
Last query:
Last reply:
Revision history for this message
Jérôme Duriez (jduriez) said : 		#1

Hi,

You may note the following remarks:

1.
As somewhat explained in https://yade-dem.org/doc/user.html#law2-functor-s (to open with another web browser than Mozilla Firefox at the moment), the Law2_* functor will be chosen according to

- the IGeom (cont.geom # with cont an Interaction) type of the interaction
- and its IPhys type

The latter itself depends on Materials assigned to bodies (and on the chosen Ip2), which you did not precise here.

2.
Assuming everything is FrictMat-material, I'm not expecting the "[Ip2_FrictMat_FrictMat_FrictPhys(),Ip2_FrictMat_FrictMat_MindlinPhys()]" part to work well / as you intend. I would speculate that one Ip2 superseeds the other here, because Ip2 are not designed to care about the Bodies shapes. It is easy to check by yourself though, asking for cont.phys and checking what you got.

3.
It should not be too difficult to also check directly if your script conforms your intended behavior. For instance, it should be possible to ask for a penetration depth at a particle-particle interaction (Sphere-Sphere shapes ?) and a particle-wall (Sphere-Box shapes, actually ?), together with the normal force value, and check whether linear elasticity or non-linear elasticity applies.

(On a very personal note, I'm not really optimistic about the possibility obtaining the intended behavior, and wonder why you would like to get that behavior)

Revision history for this message
Hossein (hossein75) said : 		#2

Thank you for your practical comment