a b s t r a c tThe stability structure for lattice Boltzmann schemes has been introduced in Banda et al. (2006) [16], Junk and Yong (2007) [14] to analyze the stability of numerical algorithms. The first purpose of this paper is to discuss the stability structure from the perspective of matrix analysis. Its second goal is to illustrate and apply the results to different classes of lattice Boltzmann collision operators. In particular we formulate an equivalence condition -just recently also reported in Yong (2008) [18] -that guarantees the existence of a pre-stability structure. It is then illustrated by several examples, how this equivalence condition can be effectively employed for the systematic verification and construction of stable collision operators. Finally, we point out some shortcomings of the stability structure approach arising in certain cases.
a b s t r a c tWe show that the Chapman-Enskog expansion can be viewed as a special instance of a general expansion procedure which also encompasses other methods like the regular error expansion and multi-scale techniques and that any two expansions which properly describe the lattice Boltzmann solution necessarily coincide up to higher order terms. For a model problem, both the regular error expansion and the Chapman-Enskog expansion are carried out. It turns out that the classical Chapman-Enskog method leads to an unstable equation at super-Burnett order in a parameter regime for which the underlying lattice Boltzmann algorithm is stable. However, our approach naturally allows us to consider variants of the super-Burnett equation which do not suffer from instabilities. The article concludes with a detailed comparison of the Chapman-Enskog and the regular error expansion.
In this paper we consider a two-population lattice-Boltzmann algorithm to approximate the advection equation. First, the stability of this model algorithm is examined. The analysis is based on the analytic computation of the spectrum pertaining to the evolution matrix. After proving a necessary stability condition, the stability of the evolution matrix is shown, which is related to the CFL-condition. We use the model algorithm to demonstrate that formal stability criteria based on a multiscale expansion may fail to predict instability.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.