LIGGGHTS® - Developer Forum

Hello all,

Thanks in advance for your time . Im trying to simulate Cylindrical Couette Granular Flows with inner cylinder rotating. But it takes lot of atoms (300000) to simulate my entire domain. hence if i can Simulate a section of the cylindrical couette (by using periodic boundary conditions in cylindrical coordinates) then i can reduce the number of atoms. Is there a possibility in LIGGGHTS or i should edit the code for my purpose. If so any suggestions or ideas are greatly appreciated and welcomed.

Thanks and regards,
krishnaraj

Hi,

I'm looking for reference(s) on the way LIGGGHTS/LAMMPS is doing the domain decomposition to track the history variables such as contact information (whether a contact existed before or if it is a new contact). This is for updating, say the incremental frictional forces. Any pointers?

I think the history dependence makes DEM codes slightly different than MD codes.

Thanks,
Joe

Dear all,

Thanks for your ongoing work with LIGGGHTS it's great to see it being updated all the time. I was wondering what did you mean by:

"Fixed some broken macros in the source code. Thanks to Richard Berger (JKU) for patching this."

in the patch notes. Which macros did you mean? How might they affect the results of previously performed simulations?

Thanks again, I look forward to your response.

Jon

Hi everyone,
I've been using LIGGGHTS pretty successfully for a while now and I'm now at the point where I need to make some new fixes to extend the code to my particular research topic. But without any experience at all with c++ I'm struggling mightily. This is a request for general help because on top of not knowing exactly what's going wrong -- I'm not even knowing the right questions to google to find my answers. I'm not looking for someone to help with each line of code, just maybe point me to more specific locations where I can find answers.

Hi,
I would like to perform simulations with LIGGGHTS using load balance. If I understood correctly from previous forum discussions, it seems this feature is implemented but it is currently not part of the public release :(
However, it seems from previous forum discussions too that it is possible to use this feature "at our own risk". Where do I find instructions for compiling/using load balance with LIGGGHTS?
Thanks so much in advance!
Best
Eric

Hi,

I wrote a fix similar to "fix_cfd_compute_force" that add a constant force to all atoms.
It works well when a constant force is applied at the initialization (no coupling).

I want now to access this property/atom array from python to allow coupling. There is the function "lammps_extract_fix" in library.cpp that seems to do that.

I added the possibility that the restitution coefficient varies with a local Stokes number of the particle (motion of particles in a viscous fluid) to pair_gran_hooke_history and fix_wall_gran_hooke_history.

The code and an example script is provided.

One has to enable a "viscousflag" (similar to dampflag for instance) into pair style or fix wall definition. The viscous flag is :

pair_style gran/hooke/history rolling_friction cdt viscous st

It will work only for hookean (history?) contacts. One has to provide the 3 parameters needed for the viscous damping :

Some recent benchmark studies looking at LIGGGHTS, EDEM 2.4.3, and StarCCM+

The first test case is a short section of rotating drum with 5,000 particles run on 1,2,4 and 8 cores (Intel X5355, Redhat EL5) run from 0 to 10 seconds.

The next case is the same drum geometry extruded to 3m in length with 100,000 particles run on 8 cores

Dear all,

I noticed that the new implementation now uses iMesh to indicate the ID of the mesh, and it still uses iTri to identify each mesh triangle (face). (please note this is from fix_wall_gran.cpp)

However I am not 100% sure what the variable "idTri" is, please could you clarify?

From what I can work out, it's a triangle/face id which is derived from iTri? Is it the ID number for each face - for each processor, when the simulation is run in parallel?

Is there a specific reason that only 32 groups can be defined? Is there any way to relax this limit?

I suppose an alternative to my problem would be a way to define forces on atoms using the atom ID rather than groups (as in addforce, setforce, aveforce). If such methods are available, please let me know!

Any help is greatly appreciated. Thank you!
Zamir