Gauss-Hermite quadratures and accuracy of lattice Boltzmann models for non-equilibrium gas flows

Jian-Ping Meng, Yonghao Zhang

Research output: Contribution to journalArticlepeer-review

59 Citations (Scopus)
289 Downloads (Pure)

Abstract

Recently, kinetic theory-based lattice Boltzmann (LB) models have been developed to model nonequilibrium gas flows. Depending on the order of quadratures, a hierarchy of LB models can be constructed which we have
previously shown to capture rarefaction effects in the standing-shearwave problems. Here, we further examine the capability of high-order LB models in modeling nonequilibrium flows considering gas and surface interactions and their effect on the bulk flow. The Maxwellian gas and surface interaction model, which has been commonly used in other kinetic methods including the direct simulation Monte Carlo method, is used in the LB simulations. In general, the LB models with high-order Gauss-Hermite quadratures can capture flow characteristics in the Knudsen layer and higher order quadratures give more accurate prediction. However, for the Gauss-Hermite quadratures, the present simulation results show that the LB models with the quadratures obtained from the even-order Hermite polynomials perform significantly better than those from the odd-order polynomials. This may be attributed to the zero-velocity component in the odd-order discrete set, which does not participate in wall
and gas collisions, and thus underestimates the wall effect.
Original languageEnglish
Pages (from-to)Article 036704
Number of pages10
JournalPhysical Review E: Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics
Volume83
Issue number3
DOIs
Publication statusPublished - 11 Mar 2011

Keywords

  • high-order lattice Boltzmann models
  • nonequilibrium gas flows
  • Gauss-Hermite quadratures
  • even-order Hermite polynomials

Fingerprint

Dive into the research topics of 'Gauss-Hermite quadratures and accuracy of lattice Boltzmann models for non-equilibrium gas flows'. Together they form a unique fingerprint.

Cite this