(272c) Recent Developments of Highly-Scalable Coarse-Grain Modeling Tools Formulated Upon Dissipative Particle Dynamics

Modeling and simulation of materials at the micro- and mesoscales have a wide range of challenges. For example, the mechanical properties of many polymers are dominated by topological constraints or entanglements, which prohibit the use of atomistic-level modeling and simulation. Therefore, the simulation of the morphology and mechanics of microphase separated systems, such as gels and polymer networks require mesoscale modeling and simulation. One such method, Dissipative Particle Dynamics (DPD), has become a common method of choice for use in the study of coarsened soft matter (e.g., surfactants, polymers, biomolecules). While DPD itself is isothermal, the DPD concepts and methodology have been extended to simulate condensed-phase systems under isothermal, isobaric, isoenergetic, and isoenthalpic conditions.
Numerical integration of the DPD equations-of-motion requires special consideration, where the Shardlow-splitting algorithm (SSA) has been found to be the most appropriate integration scheme available to-date. Despite significant DPD algorithmic improvements with the SSA, there are considerable challenges in parallelizing the SSA algorithm. This work focuses on exploiting multi-core architectures by converting the serial SSA algorithm into a highly-scalable application-focused software package. The multi-scale suite of tools has been implemented into the open-source LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator) simulation package. While the suite of DPD methods has major application in polymer and biological sciences, we will demonstrate that they also can be used for molecular solids. In particular, we will demonstrate shock and heating of coarsened RDX (cyclotrimethylenetrinitramine), an energetic material, where dynamic and thermodynamic properties using the constant-energy DPD method agree well with fully atomistic simulations.