Hi,
i have setup a simulation with 6000 spheres using the "hertz tangential history" contact model and the atom_style granular. Each sphere has a specific radius, initial position and a my_val property (see below). my_val is related to the radius (external lookup table). After running 500000 steps, i change Youngs modulus from 5.00000e+006 to 5.10000e+006 using the fix-line below. In the next dump-file saved afterwards (501000), my_val and radius of 484 spheres (out of 6000) no longer matches.
Why?
The spheres dont move that much and i can reconstruct where each sphere went between timestep 500000 and 501000. The radius of the spheres is constant, but my_val changes. It looks like a random permutation. The changes are localized at 2 specific z-coordinate regions, but seem uncorrelated in x and y-coordinates. Each additional change in Youngs modulus causes another random permutation.
Thanks,
Laurent
Setup my_val
fix trackID all property/atom my_val scalar yes no no 0
Sphere creation with my_val setup (example):
create_atoms 1 single 51.6240 64.9049 89.3060
set atom 5988 diameter 5.50018
set atom 5988 property/atom my_val 2188.00
youngs modulus set:
run 500000
fix m1 all property/global youngsModulus peratomtype 5.10000e+006 5.10000e+006
run 100000