A velocity space hybridization-based Boltzmann equation solver

JOURNAL OF COMPUTATIONAL PHYSICS

卷:408

Article

Oblapenko, G.¹  Goldstein, D.²  Varghese, P.^1,2  Moore, C.³ 

[1] Univ Texas Austin, Oden Inst Computat Engn & Sci, 2201 E 24th St,Stop C0200, Austin, TX 78712 USA

[2] Univ Texas Austin, ASE EM Dept, 2617 Wichita St,Stop C0600, Austin, TX 78712 USA

[3] Sandia Natl Labs, POB 5800, Albuquerque, NM 87185 USA

关键词: Rarefied gas dynamics;    Direct simulation Monte Carlo;    Boltzmann equation;    Discrete velocity method;   

DOI  :  10.1016/j.jcp.2020.109302

来源: Elsevier

【 摘 要 】

In the present work, a new method for simulation of rarefied gas flows is proposed, a velocity-space hybrid of both a DSMC representation of particles and a discrete velocity quasi-particle representation of the distribution function. The hybridization scheme is discussed in detail, and is numerically verified for two test-cases: the BKW relaxation problem and a stationary Maxwellian distribution. It is demonstrated that such a velocity-space hybridization can provide computational benefits when compared to a pure discrete velocity method or pure DSMC approach, while retaining some of the more attractive properties of discrete velocity methods. Further possible improvements to the velocity-space hybrid approach are discussed. (C) 2020 Elsevier Inc. All rights reserved.

【 授权许可】

Free