Pair coeff modified

Dear lammps users and developers,

I am simulating the liquid-solid interactions, using pair_style lj/cut/coul/long. Data file contents the LJ parameters in Pair Coeffs section. After reading the data file, a mixing rule is used to define all cross interactions. As the interactions of atoms of solid have different interaction parameters between them and with the atoms of liquid, these are modified using pair_coeff comand (after read the
data file and defining all cross interactions).

My question is: This setup modifies only the pair coeffecient interactions using in pair_coeff comando for all simulation, or the pair coefficients are updated, modifying all pair coefficients.

For example, for atoms of solid phase, the pair coefficient in data file
correspond to the parameters of solid-liquid interactions. While in the input data, these parameters are modified for interaction between the atoms of solid phase using pair_coeff comand. The interactions between the liquid and solid are computed using the parameters of data file or are modified by the pair_coeff in the input data?

Thank for your valuable help.

Regards

Please note that LAMMPS is a computer program. It doesn’t know anything about the physics or properties of your system. So it does not know (or care) whether your atoms are solid or not. It will simply follow the programmed “rules”. What are the rules is described in detail in the documentation. The documentation also clearly explains that LAMMPS will always apply the current state and not know about how settings were entered if they were later modified. Consequently, for pair_coeff settings thus, there is only one state. The documentation also explains when exactly and how mixing is applied.

if you have any doubts about this, you can easily figure out what happens yourself by either:

  • setting up minimal test systems with just a few atoms, where you can compute forces and energies manually and compare to what LAMMPS computes
  • use the info coeffs or write_coeff commands to review the state of force field parameter settings

As you have seen from the lack of responses, nobody here will do this for you, and particularly not on such an abstract/hypothetical level as that is what documentation is written for. If you find an inconsistency in the documentation or can provide a specific input deck, where the documentation disagrees with the actual behavior, we are happy to discuss and resolve it.

Axel.

Please correct me if I’m wrong, but it sounds like you’re wondering if the interactions you defined in the input script via pair_coeff will be saved in subsequent data files? I believe the answer to that is yes. If you’re wondering if the pair_coeff commands alter the parameters in a specific data file you just read in, then I believe the answer to that is no.

Does that help?

Will

Please correct me if I’m wrong, but it sounds like you’re wondering if the interactions you defined in the input script via pair_coeff will be saved in subsequent data files? I believe the answer to that is yes.

i disagree. Without additional options LAMMPS will only write the per atom type definitions of the “PairCoeff” section (equivalent to the “pair ii” option to write_data), the remaining settings will be inferred from mixing only. This is to accommodate conventional (bio)molecular force field simulations. If you want to store all pairs of coefficient settings, you need to use the “pair ij” option. Or better yet, use a restart file or do not write out the force field parameters in the data file at all but use the write_coeff command and then include those with the “include” command. For any complex system, using PairCoeff (and BondCoeff etc.) in the data file can easily become clumsy and complex.

Axel.

Thanks for the correction, Axel.