Flow in complex domains simulated by Dissipative Particle Dynamics driven by geometry-specific body-forces

JOURNAL OF COMPUTATIONAL PHYSICS

卷:305

Article

Yazdani, Alireza¹  Deng, Mingge¹  Caswell, Bruce²  Karniadakis, George Em¹ 

[1] Brown Univ, Div Appl Math, Providence, RI 02912 USA

[2] Brown Univ, Sch Engn, Providence, RI 02912 USA

关键词: Particle methods;    Pressure-driven flows;    Stagnation-type flows;    DPD;    Mesoscopic simulations;    3D simulations;   

DOI  :  10.1016/j.jcp.2015.11.001

来源: Elsevier

【 摘 要 】

We demonstrate how the quality of simulations by Dissipative Particle Dynamics (DPD) of flows in complex geometries is greatly enhanced when driven by body forces suitably tailored to the geometry. In practice, the body force fields are most conveniently chosen to be the pressure gradient of the corresponding Navier-Stokes (N-S) flow. In the first of three examples, the driving-force required to yield a stagnation-point flow is derived from the pressure field of the potential flow for a lattice of counter-rotating line vortices. Such a lattice contains periodic squares bounded by streamlines with four vortices within them. Hence, the DPD simulation can be performed with periodic boundary conditions to demonstrate the value of a non-uniform driving-force without the need to model real boundaries. The second example is an irregular geometry consisting of a 2D rectangular cavity on one side of an otherwise uniform channel. The Navier-Stokes pressure field for the same geometry is obtained numerically, and its interpolated gradient is then employed as the driving-force for the DPD simulation. Finally, we present a third example, where the proposed method is applied to a complex 3D geometry of an asymmetric constriction. It is shown that in each case the DPD simulations closely reproduce the Navier-Stokes solutions. Convergence rates are found to be much superior to alternative methods; in addition, the range of convergence with respect to Reynolds number and Mach number is greatly extended. (C) 2015 Elsevier Inc. All rights reserved.

【 授权许可】

Free