Pore-scale study of rarefied gas flows using low-variance deviational simulation Monte Carlo method

Ferdin Don Bosco, Yonghao Zhang

Research output: Contribution to journalArticlepeer-review

Abstract

Gaseous flow through ultra-tight porous media, e.g. shale and some high-performance insulation materials, is often rarefied, invalidating an analysis by the continuum flow theory. Such rarefied flows can be accurately described by the kinetic theory of gases which utilizes the Boltzmann equation and its simplified kinetic models. While discrete velocity methods (DVM) have been successful in directly solving these equations, the immense potential of a particle-based solution of the variance-reduced Boltzmann-BGK (Bhatnagar-Gross-Krook) equation for rarefied flows in porous media has not been exploited yet. Here, a parallel solver based on the low variance deviational simulation Monte Carlo (LVDSMC) method is developed for 3D flows, which enables pore-scale simulations using digital images of porous media samples. The unique advantage of this particle-based formulation is in providing additional insights regarding the multi-scale nature of the flow and surface/gas interactions via two new parameters, i.e. pore and surface activity respectively. Together, these two parameters can identify key flow properties of the porous media. The computational efficiency and accuracy of the current method has also been analysed, suggesting that this new solver is a powerful simulation tool to quantify flow properties of ultra-tight porous media.
Original languageEnglish
Pages (from-to)25–48
JournalTransport in Porous Media
Volume138
Early online date1 Apr 2021
DOIs
Publication statusPublished - May 2021

Keywords

  • rarefied flow
  • knudsen diffusion
  • Low-variance deviational simulation Monte Carlo method
  • Porous Media
  • Pore-scale simulation

Fingerprint

Dive into the research topics of 'Pore-scale study of rarefied gas flows using low-variance deviational simulation Monte Carlo method'. Together they form a unique fingerprint.

Cite this