Particle Penetration & ID selection

Hi
I have two questions:
(i)
I am trying to model a problem as attached in the figure. A soft web is bonded to a rigid substrate which is moving at constant speed. Whwn web hits the wall, it should delaminate. However as shown, in my case particles are penetrating the wall. I have tried increasing the contact stiffness, but its not helping.
Can someone please help me with this, I have also attache the input file.

(ii)
My second question is related to above
After few particles get delaminated, i want to apply a contant velocity boundary condition to the last particle (right-most) of the web. However, I dont know the ID for the same. How can I extract ID of the particles, satisfying a certain criteria.

Thanks

Attachment	Size
problem def	98.74 KB
input file	10.67 KB

Forums:

LIGGGHTS® - User Forum

mschramm | Sat, 10/10/2020 - 18:00

some issues

Hello,
I noticed some issues with the file.

When adding the blade, you need to set a new lattice size as the one you were using before hand will space out the spheres too far. I suspect this is why you spheres were able to "penetrate" the blade, as the blade spheres are actually much smaller and there is a large gap between the spheres on the blade. You had the idea to reset the lattice but you forgot to use the blade particle radius for the cylinder.

variable dummy2 equal 1.1*${beam_length}
region bc_blade cylinder z ${dummy2} 0.00 ${blade_particle_radius} 0.0 ${blade_length} units box
lattice sc ${blade_particle_diameter}
create_atoms 3 region bc_blade # Insert spheres of type 3 into region blade

Another issue I saw was when you were creating the bonds. You used the same bond_skin length for all of the bonds. After fixing the above issue, you need to reset the bond skins to match the prescribed bond lengths.
If you are using one of my newer releases, the bond_create command takes an extra flag 'doNorm'. You would use it as follows
fix bondcr web bond/create/gran 1 1 1 ${bond_skin_norm} 1 6 doNorm yes
where bond_skin_norm is a normalized distance value of the two spheres trying to be bonded.

SHUBHAM AGARWAL | Sat, 10/10/2020 - 22:13

ID detection

Thanks a lot mshramm
Can you please also help me with the second doubt i.e extract ID of the particles, satisfying a certain criteria.

mschramm | Sun, 10/11/2020 - 23:29

use regions and groups

Hello,
I would try to assign the target atom to a group by regions.
It may take multiple adds but you can union these groups together to get a singular atom.
https://www.cfdem.com/media/DEM/docu/group.html

SHUBHAM AGARWAL | Tue, 10/13/2020 - 09:00

Thanks a lot, it worked

SHUBHAM AGARWAL | Sun, 10/11/2020 - 07:51

Function of donorm cmd

Thanks mschramm, for the suggestion
Can you please help to understand what difference does donorm yes argument make exactly. Sorry I didn't understand it

Thanks

mschramm | Sun, 10/11/2020 - 23:17

doNorm Command

Hello,
Normally, the bond creation command simply uses the distance between two atoms (r).
The code would then check to see if r < ${bond_skin}.
If all of your atoms have the same diameter then this is not an issue.
The doNorm yes flag takes the distance r and divides it by (r1+r2) where r1 is the radius of atom 1 and r2 is the radius of atom 2.
Let r' = r/(r1+r2).
Now a bond is created if r' < ${bond_skin_norm}.