LIGGGHTS® - User Forum

Hi,

I just want to say there is a new Voronoi package for Lammps by Daniel Schwen. It supports now polydispers packings, calculation of the cell surface and much more. It's already included in the Lammps git repository. You can install and compile in Liggghts without major problems, but the delivered vorotest script has to be changed, Liggghts doesn't support inline variables and the abs() function for variables yet. Done that, the tests should run without any error.

Regards,

Paul

Hi,

I need the average pressure between the particles. I tried two ways with exact the same results. compute pressure for all atoms using the temperature from a compute temp and compute stress/atom for every single particle and sum these with compute reduce. What would be faster? Is there a difference in speed between compute reduce sum and the variable sum() command?

Regards,

Paul

Hi,

I want to generate a dense packing with fix insert/pack and particle growing. This is very common and no problem with fix adapt. The growing ends when the particles have a reached a given pressure. The pressure is calculated by compute pressure. Since I have a lot of particles and slightly CPU I wonder if there is something new in interrupting a run when a special event occurs to save CPU time? Or is run N every n, check with if and going on else dump results and quit still the preferred solution?

Regards,

Paul

I am using LIGGGHTS on a bluegene/p cluster and when I attempt to run a simulation I get an error that reads:

Could not locate a fix/property storing value(s) for youngsModulus as requested by gran/hooke. (modify_liggghts.cpp:247)

I checked with the sysadmin and he has told me that there is nothing wrong with the installation. I have version 2.3.5 installed. Any help or direction at all would be greatly appreciated.

Hi,

I am wondering how the adhesion between the particles and the wall is treated. I have set up my simulation with 3 different types, 1 is the particles, 2 is the walls, and 3 is for a no friction/no adhesion symmetry plane. I have specified the cohesion as:

fix m6 all property/global cohesionEnergyDensity peratomtypepair 3 &
1000000 1000000 0000001 &
1000000 0000001 0000001 &
0000001 0000001 0000001

I am using the sphere type for the granular simulation, how could I include the potential force in the simulation? I tired us the Pair_style hybrid gran/hertz/history lj/cut 2.5. But I get an error message” pair hybrid sub-style is not used”. Can someone help me out? thanks in advance

I typically use the fix_nve integration (only updates position and linear velocity) for my granular simulations as this gives a good approximation of blocky (angular) material behaviour. I am wondering if the sjkr cohesion model can be used with this integrator or do I need to use fix_nve_sphere to get cohesion?

In the mean time, I am making a little test model.

Thank you,

Kirk

Hi,

playing around with the dump image command I came over a restiveness. I inserted some atoms in a lattice with create_atoms, and changed diameter to 0.2 by set command. Computes and dump local/custom are working like expected. The command dump image ignores diameter and uses 1.0, after dump_modify adiam it works as well. Is it because of using create_atom?

Edit:

Dear guys:

I am new to this forum, and I would like to seek some help in the compilation of LIGGGHTS

When I compiled LIGGGHTS using "make openmpi", I encounter the following error messages:

Good morning,

I would like to make a model using the fix mesh/surface/stress/servo command. I was wondering if someone would be willing to share a little example simulation that includes this command?

Thank you in advance,\

Kirk Fraser