Dear users and researchers:
I am woking with tae process of carbonization now, but i face a difficult that is i can not achieve periodic remove volatile gaes (such as H2O,CO,H2) from the simulation cell using LAMMPS. I was tried the command delete_atoms, but it do not success. Could you tell me which commmand will achieve it or need to call of external program.
Thanks
There are no ready to use commands for such complex tasks (and there cannot be except for specialized application software).
The major question here is how you can identify atoms during an MD run that you would want to remove.
What do you mean by “unsuccesful”?
It is extremely difficult to give helpful and specific advice on such a very vague description of your simulation.
such as the pyrolysis of resin, it will create volatile gases in the heating pyrolysis process. Using delete_atomscommand has four ways to delete atoms in region or group. But, deletion of specific molecules cannot be achieved.
That is because LAMMPS is a computer program and as such can only do things that it has programmed to do. For LAMMPS your system is not a resin or molecules, but just a bunch of atoms that are categorized by their atom type number which it then relates to the force field based on those numbers. So what is a “molecule” and identifying specific kinds to molecules is not something that plays a role here.
In fact, identifying specific molecules is a very complex task (not only for a computer program), so you will have to first add some functionality to LAMMPS that can identify exactly which group(s) of atoms constitute the molecules of interest. So what you are asking for is a very challenging task. This is further complicated by the fact that what constitutes a molecule is usually an input parameter (as the bond topology) to LAMMPS. I infer from your description that you are using ReaxFF (again, vague descriptions → unspecific advice) where this is all not defined by computed on the fly based on bond orders and other internal parameters and settings.
Thank you akohlmey. May be i should try to achieve this task by other programs.
If you do use ReaxFF (did I mention that it is very annoying to not know any specifics about your simulation), you may want to check out fix reaxff/species command — LAMMPS documentation
You can post-process the result or add functionality to it that would translate the detected species information into some numbers, (e.g. molecule ids) that would allow to identify the molecules of interest.
i also think so. May I ask if the teacher has an example of this?
No. If I had it, I would have already done the study and you would not have a research project.
for reference, this functionality is now available as the ‘delete’ keyword of fix reaxff/species (currently available in the ‘develop’ branch on Github)