Dear LAMMPS developers and users,
Greetings. I am working on a water-between-coarse grained slabs case, and I am trying to adapting the slab function in LAMMPS. Conditions are summarized below:
initial condition of water (as in the attached file: 95.28500000):
- pre-equilibrated SPC/E rigid water (in bulk condition)
- 512 molecules
- density = 1 g/cm^3
- cubic box (L = 24.8344 A)
- T = 362.8 K
walls:
- applied in z direction at zlo and zhi
- tried two 9-3 LJ potentials (marked in the input file)
The program stops at the PPPM initialization. I thought that I may need to minimize my system after the slab being inserted, but the SHAKE algorithm does not allow that. I also tried to enlarge the system in z direction by imposing change_box in order to leave more gap between the walls, but this trick does not help neither. I would greatly appreciate if anyone is willing to share his/her expertise on this…my script and data file are attached for review. Thank you for reading my email.
LC Liu
95.28500000 (302 KB)
in.spce (1.94 KB)
Dear LAMMPS developers and users,
Greetings. I am working on a water-between-coarse grained slabs case, and I
am trying to adapting the slab function in LAMMPS. Conditions are summarized
below:
initial condition of water (as in the attached file: 95.28500000):
- pre-equilibrated SPC/E rigid water (in bulk condition)
- 512 molecules
- density = 1 g/cm^3
- cubic box (L = 24.8344 A)
- T = 362.8 K
walls:
- applied in z direction at zlo and zhi
- tried two 9-3 LJ potentials (marked in the input file)
The program stops at the PPPM initialization. I thought that I may need to
minimize my system after the slab being inserted, but the SHAKE algorithm
does not allow that. I also tried to enlarge the system in z direction by
you can approximate SHAKE very well during minimization by using high
force constants on the otherwise constrained bonds and angles.
imposing change_box in order to leave more gap between the walls, but this
trick does not help neither. I would greatly appreciate if anyone is willing
to share his/her expertise on this....my script and data file are attached
for review. Thank you for reading my email.
you should check whether the problem is with pppm or with your system
at all. so i would suggestion you first try running it with
lj/cut/coul/cut and without pppm and see, if you can make it work and
perhaps even minimize it with this setting. if you are far away from
equilibrium, it doesn't matter if your potential is not very accurate
and once you are close, you can switch to the proper setup.
there should be no need to enlarge the box, since kspace_modify should
do that automatically, if i remember correctly.
axel.
Hi, Axel,
Thank you so much for the prompt reply. Really appreciated. I will definitely try the comments you made.
Yet, the current script even does not pass the PPPM initialization process, namely, the program stops at:
PPPM initialization …
G vector (1/distance) = 0.290408
grid = 15 15 45
stencil order = 5
estimated absolute RMS force accuracy = 0.0256293
estimated relative force accuracy = 7.7182e-05
using double precision FFTs
3d grid and FFT values/proc = 7252 2240
So, is there any possible mistake made with the kspace setting?
LC Liu
Hi, Axel,
Thank you so much for the prompt reply. Really appreciated. I will
definitely try the comments you made.
Yet, the current script even does not pass the PPPM initialization process,
namely, the program stops at:
PPPM initialization ...
G vector (1/distance) = 0.290408
grid = 15 15 45
stencil order = 5
estimated absolute RMS force accuracy = 0.0256293
estimated relative force accuracy = 7.7182e-05
using double precision FFTs
3d grid and FFT values/proc = 7252 2240
So, is there any possible mistake made with the kspace setting?
why do you assume, it has to do with kspace? there is no error here
and it looks as if the pppm initialization worked fine. there *has* to
be an error message or you should be able to generate a stack trace
from a core dump or from running under a debugger. *that* should tell
you where the error is happening. i would bet a large sum of money
that it is *not* due to pppm or kspace_modify.
axel.
Hi, Axel,
Thank you so much for the comments. I am debugging aligning per your suggestions.
LC Liu