Dear users of the granular package
I have found a problem when I run the file I supply below (using the 15Mar11 version with the granular package enabled). I have added comments in to make it easier to inderstand, but I basically create a packing of overlapping spheres bounded by granular walls. I then apply viscous damping to remove any unwanted forces, and I am monitoring the grain z coordinate. Then I remove damping completely.
What happens is that I register different behaviour if I do run 10000 instead of run 5000, run 5000. This behaviour might be difficult to spot unless you run with zero damping, but is probably a bug.
So what I think is happening is that the run command deletes all the forces and recalculates them, and I think that something is not done properly in the way the shear force gets calculated for each interaction pair (ie in the way the last value of shear gets passed on to the pairs at the initial step after the command run Nsteps is found). I looked into the code a bit and indeed the function Verlet::setup seems to be running, which as I understand it deletes all forces and recalculates them, but also builds the neighbour list again and maybe (not sure) remaps the atoms to processors etc..
I am emailing because I have spent quite a lot of time trying to figure out how to do the force recalculation properly and I got lost- maybe recalculation needs to be disabled? I am not sure.. Help is gratefully appreciated.
# Created on 17/3/2011 for LAMMPS 15Mar version