LIGGGHTS® - User Forum

I am using insert/stream for particle insertion. But for some reason, particles do not come out of the region created by specifying extrusion length. And, due to this, I am not getting streaming flow of particles.

Hello Team, I want to specify frictionless primitive wall. I'm having a triclinic box, with particles and primitive walls defined as type 1 and 2, respectively, and important lines in input script are as follows.

boundary p f p

fix m3 all property/global coefficientRestitution peratomtypepair 2 ${EC} 1.0 1.0 1.0
fix m4 all property/global coefficientFriction peratomtypepair 2 ${MU} 0.0 0.0 0.0

Hi
I am using LIGGGHTS-flexible fiber version
I have a step-wise simulation process for which I create several restart files. Later in the simulation, I need to create some atoms, However, I am not able to do that as LIGGGHTS does not allow it.
Is there a way to get around this issue?

Thanks
Shubham

I'm using superquadric particles in my simulation and I want to group the particles having the same sizes after insertion so that each size is put in a certain group how could I code this issue?

Hello everyone!

I want to use hybrid/overlay, and I don't know how to use gran models. When I use:

pair_style hybrid/overlay gran model hertz tangential history

I got:
pair_style hybrid/overlay gran model hertz tangential history colloid 1.54e-4 yukawa/colloid 1.0408e6 1.54e-4
ERROR on proc 0: unknown contact model (.../CFDEM/LIGGGHTS/src/pair_gran_proxy.cpp:82)

Did anyone experience this error in using hybrid/overlay? and how to fix this!

Thanks!

Hello Team, I have a quick question: if I'm starting second part of a simulation using read_restart, can I append thermo data from second part into already written thermo file from first part, automatically?

Thank you,
Jagan Mohan.

Hello Team, I'm building a script and due to some strange reason, error message suggests me to use update_on_run_end yes flag in compute(s). For example,

variable vY atom mass*vy*vy
compute vYSum all reduce update_on_run_end yes sum v_vY

Let us say I've to simulate 1000000 time-steps, I can do it either in one go (Method: 1) or break 1000000 into two parts in which second part is started using read_restart from first part (Method: 2). My observation is that, everything else remaining absolutely same, the results are very very different from Method 1 and Method 2.

Can you please suggest how to install LIGGGHTS on a cluster? Our university HPC administrators ask for any guide or instruction, is there are any available? Thank you in advance!

Hello,

*This is a new version of an edited post because I originally thought this was a liggghts versioning issue, but it certainly isn't.*

I'm using this input data: https://1drv.ms/u/s!Am5GkbZS_98Xto1HN3LSdxzKI-x0fw?e=RBzhV9

Hi Guys

I am new to LIGGGHTS and and DEM analysis, my background is structural engineering and dynamics.
I would like to use LIGGGHTS to study a new type of vibrating screen which I am working on.

Current screens have either a linear, circular or elliptical stroke, with the stoke being constant at all points of the screen.
My proposed screen model will have a stroke which varies on various locations along the deck length.