Dear all,
Hi.
what are the differences between phrases “max angles/atom” and “max # of 1-3 neighbors” in LAMMPS? Why the number of angles and bonds in data files (made by ch2lmp) are different from the main psf files? Should I delete some rows of numbers in datafile?
Also, I have built a membrane with two different dimensions in VMD and after using ch2lmp tool, I could simulate on of them without any errors and the other with this error:
Memory usage per processor = 346.293 Mbytes
---------------- Step 0 ----- CPU = 0.0000 (sec) ----------------
TotEng = 4886710969383185408.0000 KinEng = 0.0000 Temp = 0.0000
PotEng = 4886710969383185408.0000 E_bond = 6114.4553 E_angle = 33488.2719
E_dihed = 12614.8005 E_impro = 191.0441 E_vdwl = 4886710969383166976.0000
E_coul = 666584.4928 E_long = -700206.7009 Press = -nan
ERROR: Out of range atoms - cannot compute PPPM
Does the number of atoms cause the problem? I should note that only the dimensions of membrane is changed and nothing is different from the previous simulation!
All the best,
Alireza
Dear all,
Hi.
what are the differences between phrases "max angles/atom" and "max # of 1-3
neighbors" in LAMMPS? Why the number of angles and bonds in data files (made
the number of bonds, angles and dihedrals can be different for two reasons.
a) the conversion failed, and b) to emulate some idiosyncrasies of the
CHARMM force field. this is in theory less efficient, but allows to keep the
lammps code simple and the difference only mattered much in the stone
ages of bio-simulations when computers were less powerful and had much
less memory.
by ch2lmp) are different from the main psf files? Should I delete some rows
of numbers in datafile?
no, but there were several discussions on the subject recently,
and it looks as if ch2lmp is not compatible with many psf files
that people produce these days.
Also, I have built a membrane with two different dimensions in VMD and after
using ch2lmp tool, I could simulate on of them without any errors and the
other with this error:
Memory usage per processor = 346.293 Mbytes
---------------- Step 0 ----- CPU = 0.0000 (sec)
----------------
TotEng = 4886710969383185408.0000 KinEng = 0.0000 Temp
= 0.0000
PotEng = 4886710969383185408.0000 E_bond = 6114.4553 E_angle =
33488.2719
E_dihed = 12614.8005 E_impro = 191.0441 E_vdwl =
4886710969383166976.0000
E_coul = 666584.4928 E_long = -700206.7009 Press =
-nan
ERROR: Out of range atoms - cannot compute PPPM
Does the number of atoms cause the problem? I should note that only the
no. just look at the gigantic energies. your system is blowing up big time.
you probably have overlapping atoms or otherwise garbage in your data file.
i suggest your try using ch2lmp on a small test system first (a few
water molecules?)
and figure out how to make that work and then _gradually_ increase the
complexity of your problem until you know what is happening and can
do the conversion correctly. setting up an MD simulation is not a
"just click here and it will be alright" kind of deal. you have to _know_
what you are doing or you are toast.
axel.
what are the differences between phrases "max angles/atom" and "max # of 1-3
neighbors" in LAMMPS? Why the number of angles and bonds in data files (made
the number of bonds, angles and dihedrals can be different for two reasons.
a) the conversion failed, and b) to emulate some idiosyncrasies of the
CHARMM force field. this is in theory less efficient, but allows to keep the
lammps code simple and the difference only mattered much in the stone
ages of bio-simulations when computers were less powerful and had much
less memory.
Also, as the special_bonds doc page explains, 1-3 neighbors are those
implied by the bond topology. It has nothing to do with angles that
are actually defined.
Steve