Pair distance < table inner cutoff error and some other problems

Dear everyone,

I am working on a simulation box of bicrystal containing a grain boundary. I use pair_style table command to apply the pairwise potential to the box. After running it shows pair_distance < table inner cutoff error. And I attach my files here. When I try to change the value of rlo and rhi, then I got

Step Temp TotEng Pxx Pyy Pzz Lx Ly Lz
0 800 -4.5323653e+15 -3.0165052e+09 -1.0767133e+11 -6.5259168e+19 11.669 39.0174 22.4367
1000 -nan -nan -nan -nan -nan -nan -nan -nan
2000 -nan -nan -nan -nan -nan -nan -nan -nan
3000 -nan -nan -nan -nan -nan -nan -nan -nan

Please inform me how I can solve this problem.

Yours sincerely,
Wangtg

bc.go.data (36.5 KB)

bc.go.in (2.36 KB)

pair.table (1.63 MB)

Dear everyone,

        I am working on a simulation box of bicrystal containing a grain
boundary. I use pair_style table command to apply the pairwise potential to
the box. After running it shows pair_distance < table inner cutoff error.
And I attach my files here. When I try to change the value of rlo and rhi,
then I got

     Step Temp TotEng Pxx Pyy Pzz Lx Ly Lz
       0 800 -4.5323653e+15 -3.0165052e+09 -1.0767133e+11
-6.5259168e+19 11.669 39.0174 22.4367
    1000 -nan -nan -nan -nan -nan
-nan -nan -nan
    2000 -nan -nan -nan -nan -nan
-nan -nan -nan
    3000 -nan -nan -nan -nan -nan
-nan -nan -nan

     Please inform me how I can solve this problem.

your potential energy is *extremely* high. that usually indicates are
problem in the initial geometry.
you can verify this by trying to use a simple analytical potential,
e.g. lj/cut or morse and pick parameters so that the potential is
similar to your tabulations.
if the potential energy is still high, you first need to run a
miminimization or better remove any close contacts. the latter can be
done using the delete_atoms command with the overlap option. if the
potential energy with lj/cut or morse is "normal" and extreme only
with your tabulated potential, then the problem is with your
tabulation.

axel.

Dear Axel,
       Thanks for your advice. I tried the lj/cut style, as it worked, but the structure is terrible yet. But it's hard to hold the grain boundary(GB) and the atoms near the GB by using the lj/cut or other style. I wander what can I do to my tabulation.
       Thanks again.
Yours sincerely,
Wangtg

Dear Axel,
       Thanks for your advice. I tried the lj/cut style, as it worked, but the structure is terrible yet. But it's hard to hold the grain boundary(GB) and the atoms near the GB by using the lj/cut or other style. I wander what can I do to my tabulation.

you are not paying attention to what i was suggesting. i didn't
suggest to actually do the simulation with lj/cut, but rather to use
it to figure out whether your input geometry is meaningful (i.e.
normal potential energy) or has close contacts. tabulated potentials
are much more tricky to handle when starting from a bad initial
geometry. also, this would inform you about whether your tabulation is
in the right ballpark or not.

axel.