High-fidelity scaling relationships for determining dissipative particle dynamics parameters from atomistic molecular dynamics simulations of polymeric liquids

• Published in: Scientific reports Vol. 10; no. 1; p. 4458
• Main Authors: Nafar Sefiddashti, M. H., Boudaghi-Khajehnobar, M., Edwards, B. J., Khomami, B.
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 10-03-2020; Nature Publishing Group
• Subjects: 639/166/898; 639/301/923/1028; 639/301/923/1029; 639/766/189; Humanities and Social Sciences; Molecular dynamics; multidisciplinary; Polyethylene; Scaling; Science; Science (multidisciplinary); Simulation; Viscosity
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/s41598-020-61374-8

An optimized Dissipative Particle Dynamics (DPD) model with simple scaling rules was developed for simulating entangled linear polyethylene melts. The scaling method, which can be used for mapping dimensionless (reduced units) DPD simulation data to physical units, was based on scaling factors for three fundamental physical units; namely, length, time, and viscosity. The scaling factors were obtained as ratios of equilibrium Molecular Dynamics (MD) simulation data in physical units and equivalent DPD simulation data for relevant quantities. Specifically, the time scaling factor was determined as the ratio of longest relaxation times, the length scaling factor was obtained as the ratio of the equilibrium end-to-end distances, and the viscosity scaling factor was calculated as the ratio of zero-shear viscosities, each as obtained from the MD (in physical units) and DPD (reduced units) simulations. The scaling method was verified for three MD/DPD model liquid pairs under several different nonequilibrium conditions, including transient and steady-state simple shear and planar elongational flows. Comparison of the MD simulation results with those of the scaled DPD simulations revealed that the optimized DPD model, expressed in terms of the proposed scaling method, successfully reproduced the computationally expensive MD results using relatively cheaper DPD simulations.