SPC/E water potential implementation

Dear lammps users,

I am setting up a small test case of NVT equilibrium of water ( 132 molecules ) in a domain size 54 x 27 x 27 angs.

Further:

  1. water model is spc/e ;

  2. pairs_style lj/cut/coul/cut

  3. using fix nvt and fix shake.

When the HH and OH interaction are set to non zero ( ie. eps,sig in pair_coeff are non zero),

then the fix nvt and fix shake seems to work: Temp fluctuates around set value of 300k; the pressure

fluctuates ( not yet equilibrium ). Please see output in section 1 below.

However, if the HH and OH interactions are set to zero for SPC/E simulation , then the

temp and press are now β€œ-nan-” values.

Please see section 2 in output below.

Question: Any suggestion why this β€œnan” error occurs ?

The lammps code segment is in section 3 below.

regards

shankar

Dear lammps users,

I am setting up a small test case of NVT equilibrium of water ( 132
molecules ) in a domain size 54 x 27 x 27 angs.

Further:
1. water model is spc/e ;
2. pairs_style lj/cut/coul/cut
3. using fix nvt and fix shake.

When the HH and OH interaction are set to non zero ( ie. eps,sig in
pair_coeff are non zero),
then the fix nvt and fix shake seems to work: Temp fluctuates around set
value of 300k; the pressure
fluctuates ( not yet equilibrium ). Please see output in section 1 below.

However, if the HH and OH interactions are set to zero for SPC/E simulation
, then the
temp and press are now "-nan-" values.
Please see section 2 in output below.

Question: Any suggestion why this "nan" error occurs ?

who knows? likely there is an error somewhere.

the excessive use of variables makes input practically unreadable and
it is incomplete to boot, so it not likely that anybody can or want to
debug it for you.

i have no problems with an SPC/E test input posted at the URL below.
regardless of whether using long-range electrostatics or modify it to
use only cutoffs.
http://git.lammps.org/git/?p=lammps-icms.git;a=tree;f=bench-accel/bench_spce;hb=HEAD

axel.

I understand. Thanks for the link.