Hallo,
I got a problem when inserting different sized particles. After the process of settling, I expect a somewhat plane bed of particles (which is what I get when i use monosized particles). I attach a picture of the problem to clarify the problem.
As you can see, the particles on the right side are on a higher level then the ones at the left side, which is kind of strange.
Particles are inserted via
fix pts1 all particletemplate/sphere 1 atom_type 1 density constant ${Partikeldichte} radius uniform number ${uni_low} ${uni_high}
fix pdd1 all particledistribution/discrete 1 1 pts1 1
within the region
region bc cylinder z 0 0 ${Box} ${radius} ${box_ins} side in units box
Like i said, if I do the same simulation with a constant radius, the particles are on a quite plane level.
Does someone know the reason for this or can explain to me what I am doing wrong?
Kind regards
Philipp
| Attachment | Size |
|---|---|
 particle_distribution.jpg | 200.22 KB |
 in.rRMuniformV.txt | 5.65 KB |
cstoltz | Tue, 03/05/2013 - 12:03
Can you attach the entire
Can you attach the entire input deck? From your original posting, there isn't enough to go on.
Regards,
Chris
knoe_ph | Tue, 03/05/2013 - 13:27
Hallo Chris,
Hallo Chris,
I attached the input file (I simplified it a little)
cstoltz | Tue, 03/05/2013 - 16:34
I tried to run your input
I tried to run your input deck, but there are a number of problems:
* forgot to include a create_box command
* the dump mesh/vtk file doesn't apply since you only have primitives and can be commented out to eliminate the warnings
* your timestep is quite large - ~85% of Rayleigh time when using the uniform particles sizes - recommend to keep < 20% for stability
* you have gravity applied to a group that has no particles defined, so no settling is happening
I didn't try to go any further.
Regards,
Chris
knoe_ph | Thu, 03/07/2013 - 09:32
I have to respond to your
I have to respond to your noted points:
*I deleted the create_box command at the input file I attached in my opening post. It's included in my oringial input file though, sorry
*Same goes for dump mesh/vtk, in my original input there is a stl file included
*I will lower my timestep accordingly, thanks
*I'm not sure what you are meaning. The gravity is applied for group "nve_group", isn't it?
#define particles insertion into domain
fix pts1 all particletemplate/sphere 1 atom_type 1 density constant ${Partikeldichte} radius uniform number ${uni_low} ${uni_high}
fix pdd1 all particledistribution/discrete 1 1 pts1 1
#region and insertion
region bc cylinder z 0 0 ${Box_halbe} ${Partikelradius} ${box_ins} side in units box
group nve_group region bc
fix gravi nve_group gravity 9.81 vector 0.0 0.0 -1
fix ins nve_group insert/pack seed 1 distributiontemplate pdd1 insert_every once overlapcheck yes particles_in_region ${n_Partikel} region bc
cstoltz | Thu, 03/07/2013 - 13:43
Regarding gravity, unless you
Regarding gravity, unless you have a reason to apply it only to some particles, just make it
fix gravi all gravity 9.81 vector 0.0 0.0 -1.0
The way you have it set up now, it will be applied to a group, but you're defining the group before you've inserted any particles. Thus the group is empty. I think you will need to insert the particles first, then declare the group, then you can apply gravity. Easier though to just define gravity as acting on 'all'. Same goes for the integrator.
Also, note that your insertion region is the same size as your cylinder. Use the 'all_in yes' option, or your particles will be created overlapping the wall, which will blow out your system.
Regards,
Chris
knoe_ph | Thu, 03/07/2013 - 09:03
Hallo Chris,
Hallo Chris,
thank you, I will look into this.
Philipp