[lammps-users] report a bug in normal force calculation of granular contacts --follow up

LAMMPS LAMMPS Mailing List Mirror
1

Hi Steve,

Have you confirmed my bug report on the hertz force calculation of the granular package.

After some further testing, I believe that only one line need to be modified to fix the bug.

in “pair_gran_hertzian.cpp” line 153:

  • rhertz = sqrt(radsum - r);
  • rhertz = sqrt((radsum - r)/radsum);

Of couse, this only fixes the bug.

The code is still only applicable to mono-disperse packing due to the fact that the Kn is a function of radius of the interacting particles.

So, if I were to implement full Hertz contact, that is, making Young’s modulus and Poisson ratio as the input parameters and calculate “xkk” (Kn) on the fly. What are the files to be modified?

BTW, why is Kt hard coded as 2/7 Kn?

Thanks.

yours,
yzp

2

We're still debating about the best way to allow full polydispersity. I should
send you something this week. All the changes will be in
pair_gran_hertzian.cpp.
The 2/7 is just a common model in the literature.

Steve