[lammps-users] nan shake stats using tip4p/ppm


I am trying to setup a system of TIP4P water molecules in a cubic volume, using pppm. I get the below output which has "nan" values in the SHAKE stats section. The input data file contains water molecules that have been set up according to the bond and angle requirements of TIP4P, so I think that should be okay. Has anyone else seen this error before?


Hi Joyce,

I am not sure if you can use SHAKE with TIP4P. The manual says:

"LAMMPS allows for the following kinds of clusters to be constrained: one
central atom bonded to 1 or 2 or 3 atoms, or one central atom bonded to 2
others and the angle between the 3 atoms also constained."

I don't think TIP4P falls in either of those categories, since you have a
central atom (the LJ site) bonded to 3 other atoms (the charges), but you
have to fix the lengths of those 3 bonds AND various angles. I might be
wrong, though.

You could try to use "fix rigid" instead for TIP4P molecules, or use a
different water model (TIP3P, SPC, SPC/E should work).


Lutz, Joyce,

To my knowledge we added a tip4p package to lammps to work with a real space
lj/cut/coul and a kspace pppm. I believe the respective calls to be used

    pair_style lj/cut/coul/long/tip4p #r_cross
    kspace_style pppm/tip4p #precision

where #r_cross is the value for the radius at which real space crosses to
kspace, and #precision is the value for the desired precision. These
packages fill in the extra site on tip4p, thus allowing for shaking the
remaining bonds.