Hi dear friendI used TIP4P/2005 potential model to measure Dipole Moment of Water molecule and got a different answer than reported in other paper. Why?
It is notable that I use FIX SHAKE too.
in my opinion , the dipole/chunk command dose not identify the ghost atom in tip4p model
how can i fix this proble?
best regard
Hi dear friend
I used TIP4P/2005 potential model to measure Dipole Moment of Water molecule and got a different answer than reported in other paper. Why?
It is notable that I use FIX SHAKE too.in my opinion , the dipole/chunk command dose not identify the ghost atom in tip4p model
that depends on how you simulate tip4p. if you use any of the
specialized tip4p styles, then - indeed - the rest of LAMMPS does not
know anything about it. the position of the displaced charge is only
computed on-the-fly while computing the forces. this is the price you
have to pay for the (much) improved computational efficiency of this
setup (number of atoms is 3/4). if you want the M point to be
explicitly present, you have to model tip4p differently using fix
rigid (and have much higher computational cost because you need a
smaller time step and have to compute more interactions).
how can i fix this proble?
since the model you use is a rigid model, the difference between the
correct dipole moment and the one that LAMMPS computes is an easy to
determine scaling factor. just take the ratio of the dipole moment
with the charge on the M site and divide it by the unshifted dipole
moment and you have the scaling factor. can be easily done with a
piece of paper and a pencil, even.
axel.