|From:||Axel Kohlmeyer <akohlmey@...24...>|
|Date:||Tue, 19 Sep 2017 13:22:12 -0400|
On Tue, Sep 19, 2017 at 1:08 PM, Stephan Grein <stephan.grein@...7106...
Am 15/09/17 um 20:56 schrieb Axel Kohlmeyer:
> On Fri, Sep 15, 2017 at 1:48 PM, Stephan Grein
Thanks for pointing that out.> <mailto:stephan.grein@...7128...7...
> > 3) create_atoms can create/distribute
> > random atoms. I need however to create
> > small molecules and distribute them
> > randomly. The small molecule structure
> > is known a-priori, however I need to
> > insert let's say 10 randomly in my
> > simulation box.
> > with a recent enough version of LAMMPS, create_atoms can also create
> > molecules that have been previously defined with the molecule command
> > and read from a suitable file.
> Follow up on this:
> If I would add a new ion type to my system,
> I would have to specify obviously all the
> interactions with the other types in my system.
> Correct? So would it be better to prepare the
> ion in a let's say dummy PDB file and convert
> it to a LAMMPS input data with e.g. topotools
> and then use read_data?
> the read_data command takes options like: extra/atom/types
> where you can reserve space for these atom types.
> the basic rules in LAMMPS related to this are simple:
> - number of atom/bond/angle/dihedral/impro
per types per atom, max
rs per atom, max special per atom are
> locked in, when the simulation cell is created. this happens with either
> create_box, read_data or read_restart.
> - this cannot be changed after that as it would require to redo the box
> creation. a new box can only be created after the clear command, which
> reinstantiates the LAMMPS class from scratch.
> no need to use a dummy atom, tho.
This might seem like a silly question:
If I don't need extra atom types for my molecule
I want to add via the molecule command I could just
use the atom types provided from my "main molecule"
which I loaded before with the read_data command.
There is probably no automation I could make use of.
Correct? (Just want to make sure I don't miss
something or do unneccesary work). TIA.correct. there is no way to automate this. by the time you could check whether there are redundant atom types through comparing pair_style and pair_coeff settings, the number of atom types is already locked in. for most simple force fields, you only need to provide the per type parameters and then the mixed parameters can be generated from the selected mixing rule. anything beyond that (e.g. many-body or hybrid pair style setups, you will have to figure it out "manually" on a case by case basis.axel.
> Best regards,
> > axel
> > Thanks in advance.
> > Best regards,
> > Stephan
> > -----------------------------
------------------------------> > <mailto:lammps-users@...93... -------------------
> > _____________________________
> > lammps-users mailing list
> > lammps-users@...42...
> > https://lists.sourceforge.net
> > --
> > Dr. Axel Kohlmeyer akohlmey@...24... <mailto:akohlmey@...24...>
> <mailto:akohlmey@...24... <mailto:akohlmey@...24...>>
> > http://goo.gl/1wk0
> > College of Science & Technology, Temple University, Philadelphia PA, USA
> > International Centre for Theoretical Physics, Trieste. Italy.
> Stephan Grein
> Department of Mathematics
> Temple University
> 1805 N Broad Street, Room 513
> Philadelphia, PA 19122-6094, USA
> phone: +1 (215) 434-5843