[lammps-users] Coefficient determine by Long-range Solver with Pair_style Table

Hello All,

I’m using the pair_style table with the long-range solver. I have a force field that behaves ~ const_factor / r at large distances, and its short-ranged behavior is different from ~ const_factor / r . I tabulated the short-range part of it for pair_style table such that it’s compatible with the k-space solver. And it seems to work! At least no error is reported. The output from pair_write is also OK.

My question is, can k-space solver determine my const_factor correctly in computing the long-ranged interaction?

Best regards,
Renjie

No

Thank you!

So what leading factor does the long-ranged solver assume in this case? Does it depend on the charges carried by the two atom types?

Like the coulomb parts of the corresponding pair styles, there is no regard for atom types. The interactions in both cases are computed based on the actual per atom charges.
You can introduce a global scaling factor by changing the dielectric constant.

no. tabulated potentials are not treated like coulomb, even if they contain coulomb contributions. that is why you have to use the additional keyword to indicate that.

the “cleanest” was to implement a modified and scaled coulomb with long-range support would be to treat only the short range difference between a scaled 1/r and your potential with pair style table and the use that in combination with pair style hybrid/overlay and coul/long plus kspace.

axel.

Many thanks for your email!

Does the dielectric constant also apply to the short-ranged part tabulated for pair_style table?