Grid refinement for entropic lattice Boltzmann models

TL;DR

This paper introduces a new multi-domain grid refinement method for entropic lattice Boltzmann models, enhancing simulation accuracy for complex flows like turbulence and supersonic flow, validated against experiments and direct numerical simulations.

Contribution

It presents a novel grid refinement technique specifically designed for entropic lattice Boltzmann models, improving their efficiency and accuracy in simulating various flow regimes.

Findings

Validated the method against DNS and experimental data.

Demonstrated improved accuracy in turbulent and supersonic flows.

Showcased adaptive features for multi-grid simulations.

Abstract

We propose a novel multi-domain grid refinement technique with extensions to entropic incompressible, thermal and compressible lattice Boltzmann models. Its validity and accuracy are accessed by comparison to available direct numerical simulation and experiment for the simulation of isothermal, thermal and viscous supersonic flow. In particular, we investigate the advantages of grid refinement for the set-ups of turbulent channel flow, flow past a sphere, Rayleigh-Benard convection as well as the supersonic flow around an airfoil. Special attention is payed to analyzing the adaptive features of entropic lattice Boltzmann models for multi-grid simulations.