Automated Workflow for Coarse-Grained MD Simulations of Block Copolymer Self-Assembly

We have developed an automated workflow for equilibrium molecular dynamics (MD) simulations of coarse-grained models to study block copolymer self-assembly. Using Python, we have integrated two widely used MD simulation packages (HOOMD-blue and LAMMPS) with our in-house FORTRAN 90 code for efficient data analysis. This workflow not only allows automated determination of several critical simulation parameters (e.g., timestep, equilibration time, sampling frequency, and total simulation time), but also enables efficient simulation case planning and submission by automatically generating the input data files for the MD simulations and summarizing the important simulation results (e.g., ensemble averages and statistical errors) of various cases into an Excel spreadsheet. As an example for demonstrating this automated workflow, we calculated the bulk periodicity of various ordered structures (i.e., lamellae, cylinders, gyroid, and spheres) formed by the dissipative particle dynamics chain model of diblock copolymer melts and computed their corresponding free-energy densities.