Polymer Molecular Dynamics toolkit, easy to learn, fast to code, ready for polymer property production

---

Documentation: https://polymer-molecular-dynamics.netlify.app

Source Code: https://github.com/Ramprasad-Group/Polymer-Molecular-Dynamics

---

PMD is a modern, fast, python framework for building LAMMPS input and data files for predicting polymer properties

The key properties are:

• Glass transition temperature (Tg) - [Guide] [Scripts]
• Gas diffusivity - [Guide] [Scripts]
• Solvent diffusivity - [Guide] [Scripts]
• Viscosity - [Guide] [Scripts]
• Young's modulus and tensile strengths - [Guide] [Scripts]
• Thermal conductivity - In-progress
• Solubility - Planned
• Melting temperature (Tm) - Planned

Installation

pip install pmd

Example

Below is an example where we use PMD to generate LAMMPS data and input files for Tg measurement with a list of SMILES strings.

From a python script

example.py

import pmd

# A list of polymer SMILES strings to create simulations for
smiles_list = ['*CC*', '*CC(*)CC', '*CC(*)CCCC', '*CC(*)c1ccccc1']

for smiles in smiles_list:
    # Define polymer and system specs
    syst = pmd.System(smiles=smiles,
                      density=0.8,
                      natoms_total=5000,
                      natoms_per_chain=150,
                      builder=pmd.EMC(force_field='pcff'))

    # Customize LAMMPS simulation
    lmp = pmd.Lammps(read_data_from=syst,
                     procedures=[
                         pmd.Minimization(min_style='cg'),
                         pmd.Equilibration(Teq=600, Tmax=800),
                         pmd.TgMeasurement(Tinit=600, Tfinal=200)
                     ])

    # Create job scheduler settings
    job = pmd.Torque(run_lammps=lmp,
                     jobname=smiles,
                     project='Your-project-id',
                     nodes=2,
                     ppn=24,
                     walltime='48:00:00')

    # Generate all necessary files at each SMILES folder
    run = pmd.Pmd(system=syst, lammps=lmp, job=job)
    run.create(output_dir=smiles, save_config=True)

From the command line

PMD can generate config file in YAML format out of the box, which helps you keep track of all the parameters used for each simulation. At the same time, you can build PMD systems directly via the config file from the command line. For example, run the pmd-load command with the following config.yaml to get exact same setup as the above example python script (but only for '*CC*').

$ pmd-load config.yaml [-o output_dir]

config.yaml

pmd.System:
  smiles: "*CC*"
  density: 0.8
  builder:
    pmd.EMC:
      force_field: pcff
  natoms_total: 5000
  natoms_per_chain: 150
  data_fname: data.lmps
pmd.Lammps:
  read_data: data.lmps # This file name has to match data_fname if build from a yaml file
  get_functional_form_from: # A PMD Builder has to be provided if build from a yaml file
    pmd.EMC:
      force_field: pcff
  procedures:
    - pmd.Minimization:
        min_style: cg
    - pmd.Equilibration:
        Teq: 600
        Tmax: 800
    - pmd.TgMeasurement:
        Tinit: 600
        Tfinal: 200
  lmp_input_fname: lmp.in
pmd.Torque:
  run_lammps: lmp.in # This file name has to match the above lmp_input_fname if build from a yaml file
  jobname: "*CC*"
  project: Your-project-id
  nodes: 2
  ppn: 24
  walltime: "48:00:00"
  job_fname: job.pbs