Finite-size effects in dissipative particle dynamics simulations

Abstract

We have performed dissipative particle dynamics (DPD) simulations to evaluate the effect that finite size of transversal area has on stress anisotropy and interfacial tension. The simulations were carried out in one phase and two phases in parallelepiped cells. In one-phase simulations there is no finite-size effect on stress anisotropy when the simulation is performed using repulsive forces. However, an oscillatory function of stress anisotropy is found for attractive-repulsive interactions. In the case of liquid-liquid interfaces with repulsive interaction between molecules, there is only a small effect of surface area on interfacial tension when the simulations are performed using the Monte Carlo method at constant temperature and normal pressure. An important but artificial finite-size effect of interfacial area on surface tension is found in simulations in the canonical ensemble. Reliable results of interfacial tension from DPD simulations can be obtained using small systems, less than 2000 particles, when they interact exclusively with repulsive forces.

Hyperscaling relationship between the interfacial tension of liquids and their correlation length near the critical point

“…The simulations were performed with 4500 DPD particles in a cubic box with L * = 11.4. Previous studies [25,26] have…”

Study of Polymer-Solvent Interactions of Complex Polysiloxanes Using Dissipative Particle Dynamics

“…R * c was taken as equal to 1, as was also the mass [27]. An additional advantage of DPD over other numerical modeling techniques is its robustness to finite size effects [26], which is a consequence of the shortrange nature of its interactions; therefore, one can obtain meaningful results with relatively small systems. The dimensions of the box were a cube of a volume of 1000 with each side L *…”

“…The parameters defining the dissipative and random forces intensities were γ * = 4.5 and σ * = 3.0, respectively. The equations of motion were integrated numerically using the so-called velocity-Verlet DPD algorithm, with a time step, t*, equal to 0.03 [12,26,27]. The simulations were run in blocks of 10 4 time steps, with five blocks used for reaching equilibrium and fifty more for the production phase.…”

“…Español and Warren [12] later developed the statistical mechanics of DPD, which incorporated the fluctuation-dissipation theorem into the DPD model. In the DPD formulation, all particles interact through three forces: a conservative force F C , a dissipative force F D , and a random force F R [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28]. This kind of simulation opens up the way to bridge the gap from atomistic simulations, where solubility parameters can be calculated, to mesoscopic simulations where mesophases and network formation can be studied.…”

“…The mechanical response of cancerous cells has also been modeled recently with DPD [29], which shows the versatility and usefulness of the technique. For further details and applications of DPD, the reader is referred to references [13][14][15][16][17][18][19][20][21][22][23], and [26][27][28][29][30][31]. Since DPD has been a very successful tool for the study of the physicochemical properties of polymer solutions, we used it here to model a series of polysiloxanes, namely polydimethylsiloxane (PDMS), polymethylhexilsiloxane (PMHS), and polydimethyl-A-1,1´, dimethylpropylnaminemethylsiloxane (P2DMPAS), in solvents that are out of theta conditions and also in theta conditions.…”