Lattice Boltzmann Models for Micro-tomographic Pore-spaces

TL;DR

This study systematically evaluates relaxation parameters in lattice Boltzmann models to improve accuracy, convergence, and parameter-independence in simulating flow through micro-tomographic porous media.

Contribution

It identifies optimal relaxation parameter sets in the MRT model that enhance accuracy and convergence while reducing parameter dependence in permeability predictions.

Findings

Certain relaxation parameter sets yield accurate permeability estimates.

Optimal au range is between 1.0 and 1.3 for best results.

Parameter independence improves with specific relaxation parameter choices.

Abstract

The lattice Boltzmann method (LBM) is a popular numerical framework to investigate single and multiphase flow though porous media. For estimation of absolute permeability based on micro-tomographic images of the porous medium, the single-relaxation time (SRT) collision model is the most widely-used, although the multiple-relaxation-time (MRT) collision model also has recently acquired wider usage, especially for industrial applications. However, the SRT collision model and a sub-optimal choice of the MRT collision parameters can both lead to permeability predictions that depend on the relaxation time, \tau. This parametric dependence is nonphysical for Stokes flow in porous media and also leads to much larger number of iterations required for convergence. In this paper, we performed a systematic numerical evaluation of the different sets of relaxation parameters in the D3Q19-MRT model for modeling Stokes flow in 3-D microtomographic pore-spaces using the bounceback scheme. These sets of parameters are evaluated from the point of view of accuracy, convergence rate, and an ability to generate parameter-independent permeability solutions. Instead of tuning all six independent relaxation rates that are available in the MRT model, the sets that were analyzed have relaxation rates that depend on one or two independent parameters, namely \tau and \Lambda. We tested elementary porous media at different image resolutions and a random packing of spheres at relatively high resolution. We observe that sets of certain specific relaxation parameters (Sets B, D, or E as listed in Table 2), and \tau in the range \tau\in[1.0,1.3] can result in best overall accuracy, convergence rate, and parameter-independent permeability predictions.