LIGGGHTS® - User Forum

Hi,

I have a question to the combination of Hertz contact model and rolling model. Consider a particle with high overlap rotating in elastic region -> that results in a stiffer k_n and k_t -> results in a stiffer k_r in rolling behaviour. Now the overlap is reduced and the particles roll back -> kn kt is lower, also kr is lower -> results in a larger back rotation because of the lower k_r now. Attached is a sketch of this rolling case. So do I have to use hook if I want to use the epsd rolling model?

kind regards
Joshua

I'm trying to run a compilation using a hybrid granular molecular atom_style, but when i run LIGGGHTS, this error appears:

ERROR: Invalid atom_style command (../atom_vec.cpp:72)

I'm trying to solve this problem but even reading the documentation, i cannot find a solution.

PD: I'm using the 3.2.0 LIGGGHTS version

Cheers, Nicolás

Is there a way to simulate multispheres in LIGGGHTS without the particletemplate/multisphere command?. I'm currently working in a SAG mill simulation with a partner, and our boss ask to do the same simulation with multispheres... please, a hint would be more than useful.

hi
i am a new user of liggghts 3.x and want to pour my particles throgh:
fix ins nve_group pour/dev nparticles 16200 1 distributiontemplate pdd1 vol 0.3 10 massflowrate 20. vel uniform 0. 0. 0. 0. -0.5 region bc
or
fix ins nve_group pour/dev nparticles 16200 1 distributiontemplate pdd1 vol 0.3 10 vel uniform 0. 0. 0. 0. -0.5 region bc

Hello all,
any one came accros this issue? after compilation, then i launched a simulation and found an error : /lib64/libc.so.6: version `GLIBC_2.14' not found
This happend when i updated ligggghts to 3.X version.
thanks,
Zyan

Hello community,

I am interested in the new "modify_timing" command, but I can't find any clue in the manual of version 3.2.1. Is it a standalone fix or belongs "modify_timing" to another command??

Best regards
Niels

Hi,

I am trying to extract the shear stress on every mesh element of a surface/mesh/stress using "dump mesh/vtk" as instructed in the manual. The problem here is that I can't understand the structure of the output file. Do you have any tip or do you know any better way in this regard.

Thank you.
Mohammad

To simulate a scoop I would need the scoop mesh to move over a fixed distance and to rotate for a fixed angle sequentially. I tried to use a timestep calculated like this

variable deltatime equal sqrt(${rho}/($E/(2*(1+$v))))*c_minradi*PI/(0.1631*$v+0.876605)/5

and the set velocity of the mesh used in the

fix movecad1 all move/mesh mesh cad1 linear 0.0 0.0 ${scm1d}

to get the required number of timesteps

Hi, I want to dump simulation images that does not show the whole simulation region rather a part of it.

Mine is a cylindrical region in which the segregation of particles by induced vibration is being studied. When I dump the images with the following syntax of command, it dumps the whole cylinder along with the simulation box. What I want is to clip the image at half of the x-axis something like this http://www.cfdem.com/system/files/sc_10per_overlap_coord.png

Source file "tangential_model_no_history.h" contains string with number 90:
if (vrel != 0.0) Ft = min(Ft_friction, Ft_damping) / vrel;
but this expression does not correlate with the main formula for force between particles (see doc/pair_gran.html), where the tangential part F_t consists of two components:
Ft_friction minus(!) Ft_damping

Simple example of collision of two rough particles without viscosity (gammat = 0 or Ft_damping = 0) shows that the value of tangential force Ft always will be equal zero, that is completely wrong.