pair hybrid

Dear Lammps users,

I am trying to use pair hybrid style for comb and LJ potential. I have a system of CO2 and U-metal. I would like to look at the interaction between CO2 atoms with LJ potential and U-metal…O with comb potential. Here is the part of input file and the error I am getting. In my input file U is atom type 1, C is atom type 2 and O is atom type 3.

Reading potential file ffield.comb with DATE: 2011-02-22
ERROR: Potential file is missing an entry (…/pair_comb.cpp:795)

units metal
atom_style full
dimension 3
boundary p p p

bond_style harmonic
angle_style harmonic
dihedral_style none
improper_style none
kspace_style ewald 0.000001

read_data u-pure-co2.data

pair_style hybrid lj/cut/coul/long 10.0 comb
pair_coeff * * comb ffield.comb U NULL O

pair_coeff 1 2 none
pair_coeff 2 2 lj/cut/coul/long 0.056 2.757 10.0
pair_coeff 2 3 lj/cut/coul/long 0.066 2.895 10.0
pair_coeff 3 3 lj/cut/coul/long 0.16 3.033 10.0

neighbor 0.5 bin
timestep 0.2
group u_metal type 1
group co2 type 2 3
fix NVT all nvt temp 360. 355. 100. nreset 100
fix 1 u_metal qeq/comb 100 0.0001 file feq.out
thermo_style custom step temp press pe ke etotal enthalpy vol pxx pyy pzz
thermo 100

run 1000

Thank you,

-Yasaman

Yasaman Ghadar
Postdoctoral Research Associate
Chemistry Department
Washington State University
Pullman, WA, 99164

Dear Lammps users,

I am trying to use pair hybrid style for comb and LJ potential. I have a
system of CO2 and U-metal. I would like to look at the interaction between
CO2 atoms with LJ potential and U-metal....O with comb potential.

you cannot do that. COMB is a manybody potential and thus interactions
are not pairwise additive, yet what you want to do requires pairwise
additive interactions.

Here is
the part of input file and the error I am getting. In my input file U is
atom type 1, C is atom type 2 and O is atom type 3.

Reading potential file ffield.comb with DATE: 2011-02-22
ERROR: Potential file is missing an entry (../pair_comb.cpp:795)

a quick look suggests that COMB has only force field entries for plain
U and not U-O pairs and triples.

do you have any reference, that COMB is actually able to represent
this kind of interactions properly?

axel.

Dear Axel,

There is no potential available for what I want to do. I was trying to see if this could work. My next attempt would be to use MEAM potential however at the current moment lammps does not have MEAM Uranium potential. Hence I was trying to see if I could use what is available before I start looking at other potentials.

Thank you for your time,

Yasaman

Dear Axel,

There is no potential available for what I want to do. I was trying to see
if this could work. My next attempt would be to use MEAM potential however
at the current moment lammps does not have MEAM Uranium potential. Hence I
was trying to see if I could use what is available before I start looking at
other potentials.

you seem to be neglecting how little transferable most (empirical)
potentials are. unless somebody has parameterized a potential for a
specific application that is equal of very similar to what you want to
do, you are likely to get bad results. you could practically pick the
parameters for any other element and get the same quality of result
(i.e. nonsense).
so *before* even trying anything, you should check the literature for
what has been done for your system and how this has been modeled and
how accurate the results are. beware, there are plenty of less careful
people around and it is getting increasingly easier to get
"not-quite-reliable" results published these days.

if you want to model substances based on the identify of elements
alone, you will have to settle for quantum chemistry calculations, but
let me tell you from my own past experience, even than is very
daunting for the case of uranium (or other heavy elements), which
requires to consider relativistic correction and efficient inclusion
of correlation effects, or that you get very, very lucky.

axel.

Dear Axel,

Thank you for your advise.

-Yasaman