Dears
Hope you are doing well
I want to create data file for PDMS and what I understand is that I have to use chain tools from tool directories. My question is how to use the (def.chain.ab , def.chain, chain.f90) and where and when I have to use the chain < def.chain > data.file. Please I need help because am new in lammps
Many thanks for your help
Hi Mohammed
You said you wanted to build a system containing the PDMS polymer. I think the chain.f90 program was intended to be used for creating simple abstract coarse-grained polymer models. I don’t know if it can be used to prepare realistic all-atom simulations. For that, I recommend one of the other molecule builder tools. The PDMS polymer contains silicon, which limits the force fields you can use somewhat. I suspect that the PCFF, COMPASS and DREIDING force fields will include all the parameters you need for PDMS, (The AMBER and OPLSAA force fields do not contain any definitions for the Si atom. I don’t think the CHARMM or GROMOS force fields do either. So I did not mention tools which only work with these force fields.)
There is a partial list of LAMMPS molecule builder tools here. Among the tools I am familiar with, I recommend EMC or moltemplate (or VMD/topotools).
The EMC program has excellent force-field support (including PCFF and a limited subset of COMPASS) and can generate reasonable random polymer structures to use as a starting point for your simulation.
The moltemplate program supports DREIDING and a limited subset of COMPASS and lets you select the starting shape of the polymer. There are several examples of how to build all-atom and coarse-grained polymers here and here. In both cases, you may need to use
Andrew
If your polymer(s) are long, then in order to reach a plausible equilibrium conformation, it is necessary to allow the polymer to pass through itself. In LAMMPS the later can be done using any of the following pair styles:
https://lammps.sandia.gov/doc/pair_fep_soft.html
https://lammps.sandia.gov/doc/pair_table.html
https://lammps.sandia.gov/doc/pair_soft.html
https://lammps.sandia.gov/doc/pair_gauss.html
http://moltemplate.org/lammps_code/pair_lj_charmm_coul_charmm_inter.html (Search for “soft-core”. Note: This feature requires you to download some 3rd-party code and recompile LAMMPS. For this reason, I recommend using one of the other pair styles above instead. Those pair styles are officially supported by the LAMMPS team.)
Hope this helps get you started.
Andrew
Hi Mohammed
You said you wanted to build a system containing the PDMS polymer. I think the chain.f90 program was intended to be used for creating simple abstract coarse-grained polymer models. I don’t know if it can be used to prepare realistic all-atom simulations. For that, I recommend one of the other molecule builder tools. The PDMS polymer contains silicon, which limits the force fields you can use somewhat. I suspect that the PCFF, COMPASS and DREIDING force fields will include all the parameters you need for PDMS, (The AMBER and OPLSAA force fields do not contain any definitions for the Si atom. I don’t think the CHARMM or GROMOS force fields do either. So I did not mention tools which only work with these force fields.)There is a partial list of LAMMPS molecule builder tools here. Among the tools I am familiar with, I recommend EMC or moltemplate (or VMD/topotools).
The EMC program has excellent force-field support (including PCFF and a limited subset of COMPASS) and can generate reasonable random polymer structures to use as a starting point for your simulation.
The moltemplate program supports DREIDING and a limited subset of COMPASS and lets you select the starting shape of the polymer. There are several examples of how to build all-atom and coarse-grained polymers here and here. In both cases, you may need to use
Andrew
If your polymer(s) are long, then in order to reach a plausible equilibrium conformation, it is necessary to allow the polymer to pass through itself.
This might not be necessary if you are using EMC because it tries to generate reasonable random structures. However if you are simulating a polymer phase which is highly ordered (such as a crystalline phase), then I am not sure if EMC can generate those on the first try. (I don’t know. Perhaps it can.) Either way, running a simulation using one of the “soft” pair styles I mentioned in my previous email should help.