So in trying to convert a MS file to LAMMPS I have used a couple of methods including the [VMD -> topo writelammpsdata -> output] and msi2lmp.exe method as well. They both have their own set of problems that are listed as follows:
VMD:
- Doesn't carry over velocity/energy data.
- Creates unnecessary and unwanted bonds and also doesn't define bonds in CNT.
but mostly the fact that it doesn't carry over the energy data for the system, after completing multiple days worth of computation time for dynamics in MS. The unwanted bonds causes a lot of other problems as well so no small issue.
msi2lmp.exe:
Now with this method I have not been able to convert to a LAMMPS data file without using the
[-ignore] command when converting. I later figured out that msi2lmp doesn't support the new COMPASS (II and III) forcefields, so i switch to cvff and COMPASS I (two different versions of my system to test). But still while trying to convert I get errors like:
- Unable to find mass for n3a/h1n/n2a
- Unable to find equivalent type for n3a/h1n/n2a
- Unable to find bond data for n3a c3a/n3a h1n
I have looked at various guides in converting using the msi2lmp tool but most are outdated or don't work, for example:
https://nerdfuns.wordpress.com/2014/01/01/msi2lmp-tools-how-to-use/
I am assuming that a certain calculation is required in MS to convert to LAMMPS or some combination of simulations required. Please HELP!
VMD Plugin for Accelrys/Material Studio files was designed for visualization purposes only (i.e. coors and bonds), so no surprise it doesn't convert velocities and FF related data. msi2lmp was written over a decade ago and is currently maintained only for backwards compatibility with quite outdated free/published version of COMPASS FF. You can try it with some newer FFs like pcff or COMPASS II provided you have full .frc files for those FFs, but even then it's limited:
- there is no support to select morse bonds over harmonic bonds
- there is no support for auto-equivalences to supplement fully
parameterized interactions with heuristic ones
- there is no support for bond increments
- there is no support for coordinates defined by symmetry operations,
i.e. the .mdf file has to be set up for space group P1
If you really need to convert your molecular structures and newer FFs from Materials Studio to LAMMPS , you should look at MedeA from Materials Design (https://www.materialsdesign.com/medea-software)
Hi Ernst Ellis ,
I'm having the same issue. I had managed to convert the MS file to LAMMPS using msi2lmp few month ago with some minor errors. But recently (using new Material Studio version) I'm getting same errors as you mentioned.
Please share if you found any solution.
Thanks
Ali
Hi Ali Tousi ,
So as Vitalie Botan explained above, the best option is to switch to MedeA. You can copy your MS program directly to MedeA and it has a wide range of different modules that you can use including various LAMMPS modules. I gave up on the msi2lmp.exe because its outdated and barely works, and VMD also has its own set of issues. You can contact the people at Material design (they own MedeA) and ask for a trail version including lammps application and see if the program works for your needs.
Hope this helps
Than you Ernst Ellis , I would definitely try this approach. I was just wondering if you could advice me whether the MedeA helps me to convert my model(structure) from Material Studio to a LAMMPS Data file or I need to use MedeA instead of Material Studio.
I ask this, because I have done part of my simulation in Material Studio (Crosslinked Polymers) and I just need to continue the rest of the work in LAMMPS.
Thanks so much.
- Badges
- Science topic
- Similar topics
- Filing