A comparative study of lattice Boltzmann models for incompressible flow

Liangqi Zhang, Zhong Zeng, Haiqiong Xie, Yongxiang Zhang, Yiyu Lu, Akira Yoshikawa, Hiroshi Mizuseki, Yoshiyuki Kawazoe

Research output: Contribution to journalArticlepeer-review

10 Citations (Scopus)


For incompressible flow, a comparative study on the four lattice Boltzmann (LB) models, the standard model, the He-Luo model, Guo's model, and the present model, is performed. Theoretically, the macroscopic equations derived from the involved LB models are compared by the Chapman-Enskog analysis. Then, the analytical framework proposed in M. Junk's work is applied to investigate the finite difference stencils and the equivalent moment systems pertaining to the concerned LB models. Conclusions are drawn from the theoretical derivations that the truncated error terms, which differ among the concerned LB models, have effects on the accuracy of the modeled deviatoric stress. Moreover, the cavity flow in two dimensions is adopted as a benchmark test to confirm the theoretical demonstrations. The resulting velocity fields from the present model are more in line with the reference solutions in the region of high deviatoric stress than other three LB models, which is consistent with the theoretical expectations and is further confirmed by the comparisons of the truncation error terms. In addition, we also conclude from the numerical tests that the present model has the advantage of better convergence efficiency but suffers from the worse stability.

Original languageEnglish
Pages (from-to)1446-1466
Number of pages21
JournalComputers and Mathematics with Applications
Issue number10
Publication statusPublished - 2014 Nov 1


  • Deviatoric stress
  • Finite difference stencils
  • Incompressible flow
  • Lattice Boltzmann model

ASJC Scopus subject areas

  • Modelling and Simulation
  • Computational Theory and Mathematics
  • Computational Mathematics


Dive into the research topics of 'A comparative study of lattice Boltzmann models for incompressible flow'. Together they form a unique fingerprint.

Cite this