Hi all,
I am quite new at lammps and I am trying to face the problem of having a
Fe2O3 lattice wall which interacts with H atoms. For the interaction
between the H and the wall I would like to have something like:
sigmaeq = (1/2)*(sigmaAVG + sigmaH) where sigmaAVG is a weighted average
sigmaAVG = (1/5)*(2*sigmaF + 3*sigmaO)
Is there a way to do such a thing in lammps? I have only found the
pair_modify command in lammps but weighting is not allowed.
Also, is it possible to specify an hybrid pair_style with two styles
that are the same (lets say lj/cut) but use different mixing rules? (one
geometric and the other one arithmetic)
Thanks you very much!
Eduardo
Hi all,
I am quite new at lammps and I am trying to face the problem of having a
Fe2O3 lattice wall which interacts with H atoms. For the interaction
between the H and the wall I would like to have something like:
sigmaeq = (1/2)*(sigmaAVG + sigmaH) where sigmaAVG is a weighted average
sigmaAVG = (1/5)*(2*sigmaF + 3*sigmaO)
Is there a way to do such a thing in lammps? I have only found the
pair_modify command in lammps but weighting is not allowed.
LAMMPS allows you to provide the pair coefficients explicitly for each
pair of atom types individually. thus you can input them from any kind
of rule you want.
Also, is it possible to specify an hybrid pair_style with two styles
that are the same (lets say lj/cut) but use different mixing rules? (one
geometric and the other one arithmetic)
no. see above. in general, automated mixing is only meaningful when
using "package" force fields like Amber, CHARMM, OPLS or similar. even
those have occasional exceptions that need to override the automatic
mixing to achieve consistent results.
in general, it is beneficial to specify all pair coefficients
explicitly, that is doubly true when using hybrid styles, since there
you can run into ambiguities very easily. if needed, the table of
coefficients can be generated from a small script program and then
used in LAMMPS via include.
axel.
I see. Well, actually I am simulating fluid hydrocarbons (I wanted to
simplify things in my last email) using OPLS force field with an
automatic geometric rule for mixing. Anyway, as far as I understood,I
can still use it for the hydrocarbon-hydrocarbon interaction but
manually override this behaviour for the wall-hydrocarbon interactions
using pair_coeff. Am I right?
Thanks you Axel,
Eduardo
I see. Well, actually I am simulating fluid hydrocarbons (I wanted to
simplify things in my last email) using OPLS force field with an
automatic geometric rule for mixing. Anyway, as far as I understood,I
can still use it for the hydrocarbon-hydrocarbon interaction but
manually override this behaviour for the wall-hydrocarbon interactions
using pair_coeff. Am I right?
i don't know. depends on whether you do it correctly. the best way to
be certain is always to set up a few very small test systems, output
forces and energies and compare the results against what you can
compute manually. or turn off individual parts of the interactions and
see, if you get consistent results.