Back to Search Start Over

Simplified hydrodynamic-wave particle method for the multiscale rarefied flow.

Authors :
Liu, W.
Yang, L.M.
Zhang, Z.L.
Teo, C.J.
Shu, C.
Source :
Applied Mathematical Modelling. Apr2023, Vol. 116, p469-489. 21p.
Publication Year :
2023

Abstract

• A unified CFD-particle method is proposed for multiscale rarefied flows. • By introducing the simplified hydrodynamic-wave flux, the particle solver can transfer to the CFD solver adaptively. • Error analysis proves that the simplified treatment preserves the second-order accuracy of the present method. • Tens of times speedup ratio could be achieved compared to the DVM. Coupling the Computational Fluid Dynamics (CFD) with the stochastic particle method to address multi-scale rarefied flows remains of great interest to many researchers. Rather than involving the field decomposition and the information exchange, the simplified hydrodynamic-wave particle method (SHWPM) has been proposed in this paper to automatically switch from the stochastic particle solver to the conventional CFD method in the multi-scale simulation. The weights to couple the respective numerical fluxes are derived based on the integral solution of the Boltzmann equation. By introducing the simplified hydrodynamic-wave flux that can be computed by the linear combination of inviscid and viscous fluxes from the conventional CFD solver, the collision effects of particles are calculated in a macroscopic approach and the number of sampling particles could be significantly reduced in the near-continuum regime. Error analysis demonstrates that the simplified treatment can preserve the second-order accuracy and asymptotic preserving property of the present method in both the continuum and free molecular limits. Several numerical cases over a wide range of Knudsen and Mach numbers exhibit the capability of SHWPM for modeling the multiscale flow accurately and efficiently. The example of hypersonic flow passing a cylinder shows that the SHWPM could achieve tens of times speedup ratio compared to the discrete velocity method (DVM) in the continuum regime. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
0307904X
Volume :
116
Database :
Academic Search Index
Journal :
Applied Mathematical Modelling
Publication Type :
Academic Journal
Accession number :
161817196
Full Text :
https://doi.org/10.1016/j.apm.2022.11.041